I'lllllllllllfi

“68's IIIIIIIIIIIIIIIIIIIIIIIIIIIII IIIIIIIIIII IZIIIIIIIIIII

3 1293 01388

This is to certify that the

dissertation entitled

ASystalsAsswmrtofPhosdnszFer-tilizer
forUseinPhaseoltsm L. Productim
m'I‘anzania

presented by

Cornellavrencemdeyenam

has been accepted towards fulfillment
of the requirements for

 

Ph-D degree in M

MW

‘ Major professor

MS U is an Affirmative Action/Equal Opportunity Institution 0-12771

—w--~—_~W—.—_‘*» -‘-*

‘ “'F‘ 11—-

 

 

 

 

UBRARY
Mlchtgan State
University

 

 

 

 

PLACE N RETURN BOX to remove this chodtwt from your record.

TO AVOID FlNES mum on or him dd. du0.

I—I—JATEPopf. DATE DUE DATE DUEI

‘J

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

ASI'STEM:

 

 

 

A SYSTEMS ASSESSMENT OF PHOSPHOROUS FERTILIZER FOR USE IN
Phaseolus vulgaris L. PRODUCTION IN TANZANIA.

By

Comel Lawrence Rweyemamu

A DISSERTATION
Submitted to
Michigan State University
in partial fulfillment of requirements
for the degree of

DOCTOR OF PHILOSOPHY

Department of Crop and Soil Sciences

1995

 

ASYSTE.I

 

Phosph

 

Imam A I
Soko'me I‘m-e.
”Elm of Tar.
BEIVGRO m
9M With it?
Ilie Smdy aIs
Phosphate ro<
”Watson w;
In c(”Illmflson
hired at 35 I
block The .\
Incorporated

'I.
In ’22

kgPI.

ABSTRACT

A SYSTEMS ASSESSMENT OF PHOSPHOROUS FERTILIZER FOR USE IN
Phaseolus vulgaris L. PRODUCTION IN TANZANIA

By

Cornel Lawrence Rweyemamu

Phosphorous (P) is critical to improving fertility for sustainable agriculture in
Tanzania. A field study was conducted during the 1993 and 1994 growing seasons at the
Sokoine University of Agriculture farm (6" 508, 37° 39'E at 525 m altitude) in the Morogoro
region of Tanzania. The objective of the study was to evaluate the P version of the
BEANGRO model under Tanzanian conditions for sensitivity to P response by a bean crop
grown with difl‘erent sources and levels of P fertilizers under rainfed and irrigated conditions.
The study also assessed the growth and yield of a bean crop as influenced by Minjingu
phosphate rock (MPR) used as a directly applied phosphate rock (DAPR) fertilizer in
comparison with triple superphosphate (TSP) fertilizer, and the residual effectiveness of MPR
in comparison with TSP fertilizer on bean crop performance. Cultivar Canadian Wonder was
PM at 25 plants rn'2 in a complete randomized block design with three replications in each
block. The MPR soft ore (3 0% P205) and TSP (46% P2 0,) fertilizers were broadcast and
incorporated in each plot at planting. The seven treatments used in the study were: 0 kg P
ha“; 22 kg P ha"-TSP; 22 kg p ha"-MPR; 44 kg P ha"; 88 kg P ha"-TSP; 88 kg P ha"-MPR;
and 176 kg P ha"-MPR The irrigated experiment was sown on a soil with a pH of 5.43; the

rainfed experiment was sown on a soil with a pH of 4.50. Model validation was made by

comparing predicted with field measured data Despite incidence of diseases and pests during

the study, the model accurately predicted the grth cycle, yield and P uptake under irrigated
conditions. The model proved to be sensitive to soil initial P and model functions related to
uptake. Laboratory analysis with neutral ammonium citrate (N AC) indicated MPR was a
highly reactive rock with the value of 6.7% P205. The study also showed that MPR
performed well under irrigated conditions, and that the residual effect of TSP was higher than

that of MPR fertilizer one year after application under rainfed conditions.

 

Idationately .

encouraged me I

DEDICATION

I affectionately dedicate this work to my mother, Regina, for her strong support and her
prayers, my daughter, Evangelyn, for her tolerance, my brothers, sisters, and friends who

encouraged me to seek out more knowledge in life.

iv

First,

Professor An

 

gaming me I
01er AP;
Invesrigator (I
WWII who ir
prodded the II

Deck,
my adtism’ f
“31mm of rr
thum"(lent of
Their Comm
POSSIbIe,

I also
MWOM F
ImeTeSt in my

IDS mggeSiIOm

ACKNOWLEDGMENT S

First, I would like to express my sincere appreciation to the Vice-Chancellor,
Professor Anselm B. Lwoga of Sokoine University of Agriculture (SUA), Tanzania for
granting me the leave time to undertake graduate studies at Michigan State University
(MSU). Appreciation is also expressed to Dr. Matt J. Silbernagell, former principal
investigator (USA) for Bean Collaborative Research Support Program (CRSP-Tanzania
Project) who initiated my training program, encouraged me throughout my study period, and
provided the impetus for this work.

Deepest gratitude is expressed to Dr. Ice T. Ritchie, Homer Nowlin Chair Professor,
my advisor, for his diligence in directing my studies at MSU. I am also grateful for the
assistance of my guidance committee members Dr. Boyd G. Ellis, and Dr. Russell D. Freed,
Department of Crop and Soil Sciences, and Dr. Irvin E. Widders, Department of Horticulture.
Their constructive criticisms, suggestions, and insights made completion of this work
possible.

I also express my recognition to Dr. S. H. Chien, senior soil chemist at the
International Fertilizer Development Center (IFDC) in Muscle Shoals, Alabama for genuine
interest in my work, determination of Minjingu phosphate rock (MPR) characteristics, and

his suggestions during the preparation of this manuscript. I am further thankful to Dr. Walter

l Bowttt S

Nowlin Che

computer as

 

lam
helped me C
mime
Wfipt at.
the Bean C RI,

principal Int-e5

 

mm for adn
Gull Gmup u
Initiation “It?

The fina
hammered fi

lama are qr

T. Bowen, systems scientist (IFDC), and Mr. Brian Baer, program analyst for the Homer
Nowlin Chair, at MSU for working tirelessly on the phosphorous sub-model, and for their
computer assistance.

I am also thankful to all field and laboratory staff at SUA, IFDC, and MSU who
helped me during my research work. Special thanks go to Mrs. Sharlene Rotman,
administrative assistant for the Homer Nowlin Chair at MSU for reading and reviewing this
manuscript, and above all for her office assistance during my stay at MSU. I also appreciate
the Bean CRSP office at Washington State University (Puyallup), Dr. Lorna M. Butler,
principal investigator for Bean CRSP (Tanzania Project), Mrs. Kathy Poole, program
assistant for administering my scholarship. Finally, I want to acknowledge the entire Nowlin
Chair Group whose members came fi'om all continents, as students and visiting scholars.
Interaction with them broadened my vision of life.

The financial assistance from the USAID - Title XII Bean CRSP for this program and
that I received from the Rockefeller Foundation, for the dissertation research while I was in

Tanzania are gratefirlly acknowledged.

TABLE 017 Cr.“
LIST OF TAB;
LIST or not
QWERII

BIBLIO
CHIPTER 2. s,

RESPox
C0.\Dn

ABSTM
Um

 

“ITEM,

1m;
Dar
CuIz

Mo:

TABLE OF CONTENTS

TABLE OF CONTENTS .............................................. vii
LIST OF TABLES ................................................... xii
LIST OF FIGURES ................................................... xv
CHAPTER 1: INTRODUCTION ......................................... l

BIBLIOGRAPHY .............................................. 14

CHAPTER 2: SIMULATION OF BEAN (Phaseolus vulgaris L.) YIELD AND BIOMASS
RESPONSE TO PHO SPHOROUS APPLICATION UNDER TANZANIAN

CONDITIONS ................................................ 22
ABSTRACT .................................................. 22
LITERATURE REVIEW ........................................ 24
MATERIALS AND METHODS ................................... 32
Rainfed Experiment ....................................... 32
Irrigated Experiment ....................................... 38
Data Input .............................................. 42
Cultivar Coefficients Calibration .............................. 48
Model Evaluation ......................................... 50

vii

RES.-

sssss

 

M

RESULT

Ge

RESULTS AND DISCUSSION ................................... 50

Soil Profiles Properties ..................................... 50
Weather During the Study .................................. 51
Cultivar Genetic Coefficients ................................ 56

Field Measured Results ..................................... 56

Model Evaluation Results ................................... 57
SUMMARY AND CONCLUSIONS ................................ 77
RECOMMENDATIONS FOR FUTURE RESEARCH .................. 78
BIBLIOGRAPHY .............................................. 79

CHAPTER 3: SIMULATION OF PLANT AND SOIL PHO SPHOROUS DYNAMICS IN
BEAN CROP GROWN UNDER RAINFED AND IRRIGATED CONDITIONS

............................................................ 87
ABSTRACT .................................................. 87
LITERATURE REVIEW ........................................ 88
MATERIALS AND METHODS ................................... 95

Phosphorous Model Calibration and Modification ................. 97

Initial Soil P (Subroutine SOILPI) ....................... 97

Soil Phosphorous Transformations (Subroutine PQHEM) ..... 99

Phosphorous Uptake (Subroutine PUPTAK) .............. 100

Model Testing .......................................... 101
RESULTS AND DISCUSSION .................................. 102
General Observations ..................................... 102

viii

Phosphorous Concentration and Distribution in the Plant .......... 103
Total Phosphorous Accumulation and Removal From the Field ...... 111

Evaluation of Simulated Phosphorous Concentration and Accumulation

................................................ l 13
SUMMARY AND CONCLUSIONS ............................... 121
RECOMMENDATIONS FOR FUTURE RESEARCH ................. 122
BIBLIOGRAPHY ............................................. 124

CHAPTER 4: AGRONOMIC EVALUATION OF MINJIN GU PHOSPHATE ROCK
(MPR)
USING BEAN (Phaseolus vulgaris L.) CROP UNDER
RAINFED AND IRRIGATED CONDITIONS.

........................................................... 13 1
ABSTRACT ................................................. 13 1
LITERATURE REVIEW ....................................... 132
MATERIALS AND METHODS .................................. 141
RESULTS AND DISCUSSION .................................. 144

General Observations ..................................... 144
Minjingu Phosphate Rock Characteristics ...................... 145
Effect of P Treatments on Soil pH, Extractable
Phosphorous and Exchangeable Calcium. ................ 146
Crop Growth, Total Dry Matter and Seed Yield Components
as Influenced by Different P Regimes ................... 150
Relative Agronomic Effectiveness of MPR ..................... 165
SUMMARY AND CONCLUSIONS ............................... 168
RECOMMENDATIONS FOR FUTURE RESEARCH ................. 169

ix

BlBI

CHAPTER .
BEA

 

ABS
Lnri
MATE

RESI'

 

SLIM
REcox
BHMJC

Append.“ Ia SI

temrxn,s
Ammrx2,[
ICUII;

Appendix 20. E
AgfiCUhL

We 3. Dex
Studyi. .

Appendix 4a. ITTI

BIBLIOGRAPHY ............................................. l 70

CHAPTER 5: RESIDUAL EFFECTS OF PHOSPHOROUS APPLICATION ON
BEAN PRODUCTION UNDER RAINFED AND IRRIGATED CONDITIONS

........................................................... 1 78
ABSTRACT ................................................. 178
LITERATURE REVIEW ....................................... 179
MATERIALS AND METHODS .................................. 182
RESULTS AND DISCUSSION .................................. 185

Effects of Residual P Treatments on Soil pH,
Extractable P, and Exchangeable Ca ..................... 185
Phosphorous Concentration ................................ 185

Crop Growth, LeafArea Index, and Total Dry Matter Production . . . 188

Seed Yield Components. .................................. 189

SUMMARY AND CONCLUSIONS ............................... 196
RECOMMENDATIONS FOR FUTURE RESEARCH ................. 197
BIBLIOGRAPHY ............................................. 198
Appendix 1a. Soil profile description at the rainfed experimental site .............. 201
Appendix 1b. Soil profile description at the irrigated experimental site. ........... 202
Appendix 2a. Daily weather data for the 1993 growing season at Sokoine Unversity of
Agriculture, Morogoro. ......................................... 203
Appendix 2b. Daily weather data for the 1994 growing season at Sokoine Unversity of
Agriculture, Morogoro. ......................................... 208
Appendix 3. Developmental stages of determinate bean (Phaseolus vulgaris L.) used in the
study‘. ...................................................... 213
Appendix 4a. Irrigation schedules for the 1993 season. ....................... 215

 

wmme

Mmmfi
regrrr

memesr
@m

 

 

Appendix 4b. Irrigation schedules for the 1994 season. ....................... 216

Appendix 5a. Plant population (plants m") at V3 and R, fiom fresh applied P fertilizer
regimes. ..................................................... 217

Appendix 5b. Plant population (plants m”) at V3 and R, fiom annual applied P fertilizer
regimes. ..................................................... 218

Mk II Are

the exp

rainfed

Table 2.2b Soil
Imgared

 

Table 2.3. Egan

“blaze. Exam

CUIIIIaJ‘S

Weld Surnm
condliIOrL

Table 2_7_ Su

12

Table 2.3 Slim
condliIOHS

LIST OF TABLES

Table 1.1 Average annual national bean production in Tanzania from 1988/89 to 1992/93

. ............................................................. 5

Table 2.1. General pre-plant soil characteristics of 0-12 cm soil layer at blocks A and B at
the experimental sites. ........................................... 34
Table 2.2a. Soil profile characteristics for the isohyperthemic, aridic tropustic Oxisol at the
rainfed experiment ....... ‘ ........................................ 35
Table 2.2b. Soil profile characteristics for the isohyperthemic, typic tropustic Alfisol at the
irrigated site. .................................................. 40

Table 2.3. Example of input file (SUMO9301.BNX) used to run bean model BEANGRO.
............................................................ 44
Table 2.4. Example of the weather variables used to run bean model BEANGRO. . . . . 46

Table 2.5. Estimated genetic coefficients for Canadian Wonder and other Andean type
cultivars used to evaluate the bean model BEAN GRO under Tanzanian conditions.
............................................................ 47

Table 2.6. Summary of simulation outputs of bean performance under nonlimiting P
conditions (P switched ofi) compared with the high P rate for the 1993 season.
............................................................ 59

Table 2.7. Summary of simulation outputs of bean performance under control conditions
(0 kg P ha") for the 1993 season. ................................... 60

Table 2.8. Summary of simulation outputs of bean performance under low P rate
conditions (22 kg P ha") for the 1993 season. ......................... 62

xii

   
  
   

Table 2.9. S
con:

Table 2. IO
cone;

TahIeiI. X;

Table 3.2 Ph-
by dii

Table 3.3. M
differe

Table 34 Am

Tablels con
(95) arr
ha") in
“Rah

 

Table 37 COmI
0‘8 P ha
Phospha

Table 41- SCIet

IabIC 42a. SOII

MmhA

Table 2.9. Summary of simulation outputs of bean performance under medium P rate
conditions (44 kg P ha") for the 1993 season. ......................... 63

Table 2.10. Summary of simulation outputs of bean performance under high P rate
conditions (88 kg P ha" TSP) and medium P rate (88 kg P ha'l MPR) for the 1993
season. ....................................................... 65

Table 3.1. Nitrogen content at five different bean grth stages in the 1993 season. . 103

Table 3.2 Phosphorous concentration (%) in different parts of bean plants as influenced
by different fertilizer sources and levels of P application. ................ 104

Table 3.3. Nutrient concentration in bean plant leaves at early flowering as influenced by
different sources and levels of annual applied P fertilizer in the 1994 season. . 109

Table 3.4 Amount of phosphorous removed from the field by bean harvest (kg P ha") .
........................................................... l 13

Table 3.5. Comparison between simulated and field measured vegetative P concentration
(%) at flowering stage. .......................................... 115

Table 3.6. Comparison of simulated and field measured phosphorous accumulation (kg P
ha") in bean shoots at two growth stages with the application of triple
superphosphate fertilizer. ........................................ 116

Table 3.7. Comparison between simulated and field measured phosphorous accumulation
(kg P ha") in bean shoots at two growth stages with the application of Minjingu
phosphate rock fertilizer. ........................................ 117

Table 4.1. Selected properties of Minjingu phosphate rock .................... 146

Table 4.2a. Soil pH, Bray-1P and calcium levels in pre-plant soil at of 0-12 cm layer in
blocks A and B at the beginning of 1994 growing season. ............... 148

Table 4.2b. Soil pH, Bray-I, and calcium levels in post-harvest soil at 0-12 cm layer in
blocks A and B at the end of the 1993 and block B at the end of 1994 growing
seasons ..................................................... 149

Table 4.3a. Yield components in beans grown with fresh applied P treatments under
rainfed conditions .............................................. 160

Table 4.31). Yield components in a bean crop grown with fresh applied P treatments under
irrigated conditions. ............................................ 161

xiii

 

     

Tbk4ic
bug;

Tablelid
irriea

V

Table 4 4. \
Ifrnji.
matte

Table 4.5. Sr.
Minjin

Table 5.]. Prr
awed

 

Table 5.2. Sol
Blocks

Table 5.3 llliIUr
maumu

‘-
._.

Table 4.3c. Yield components in a bean crop grown with fresh applied P treatments under
irrigated conditions ............................................ 162

Table 4.3d. Yield components in beans grown with annual applied P treatments under
irrigated conditions ............................................ 163

Table 4.4. Values of coemcients (b) and relative agronomic effectiveness (RAE) of
Minjingu phosphate rock and triple superphosphate in terms of bean total dry
matter production ............................................. 166

Table 4.5. Summary of orthogonal comparisons between triple superphosphate and
Minjingu phosphate rock in agronomic efi‘ectiveness ................... 167

Table 5.1. Pro-plant soil characteristics at 0-12 cm layer for Block C at the experimental
sites during the 1994 season. ..................................... 184

Table 5.2. Soil pH, Bray-1P and calcium levels in post-harvest soil at 0-12 cm layer in
Blocks A and C at the end of 1994 growing season. ................... 186

Table 5.3 Influence of residual P on TDM and seed production; P concentration and
maximum leaf area index. ....................................... 187

Table 5.4. Yield components fora bean crop grown with residual P in the 1994 season.
........................................................... 190

Table 5.5. Values of coefficients (fl) and relative agronomic effectiveness (RAE) of
Minjingu phosphate rock and triple superphosphate in terms of bean total dry
matter production from fresh and residual P in the 1994 season ........... 191

Table 5.6. Summary of orthogonal comparisons between triple superphosphate and

Minjingu phosphate rock in agronomic efi‘eCtiveness from freshly applied and
residual P for the 1994 season. .................................... 192

xiv

Ewell. Sc
' 55350;”.

Time 22 Se.

figute 2 3. Seal
Seasons

Figure 2 43 S I
imSate:
figme 2%- SI;
rainfed c

figure 2-4C. Sin

f5 1.
mean ' .

 

HEUICZ4d‘ Slr'r

mean _
FigureZSg $ij
{mailer u;

Mean ..,

rmu
Under rainf

LIST OF FIGURES

Figure 2.1. Seasonal patterns of temperature regimes during the 1993 and 1994 growing
seasons. ...................................................... 53

Figure 2.2. Seasonal patterns of solar radiation during the 1993 and 1994 growing seasons
............................................................ 54

Figure 2.3. Seasonal patterns of rainfall distribution during the 1993 and 1994 growing
seasons ...................................................... 55

Figure 2.4a. Simulated and field measured grain yield with the application of TSP under
irrigated conditions. Vertical bars represent standard errors of the mean ..... 67

Figure 2.4b. Simulated and field measured grain yield with the application of TSP under
rainfed conditions. Vertical bars represent standard errors of the mean ...... 68

Figure 2.4c. Simulated and field measured grain yield with the application of MPR
fertilizer under irrigated conditions. Vertical bars represent standard errors of the
mean ........................................................ 69

Figure 2.4d. Simulated and field measured grain yield with the application of MPR
fertilizer under rainfed conditions. Vertical bars represent standard errors of the
mean ........................................................ 70

Figure 2.5a. Simulated and measured biomass accumulation with application of TSP
fertilizer under irrigated conditions. Vertical bars represent standard errors of the
mean ........................................................ 73

Figure 2.5b. Simulated and measured biomass accumulation with the application of TSP
under rainfed conditions. Vertical bars represent standard errors of the mean . 74

 

Firure 2.6a

 

Erureila, l
I the 19x.
apphed

lime 32b. T
I the 199
applied:

 

5W9“. Be

(C and 1
Smih

IrlTare 4.2 Bea
and D)
870th

Herrera 3,,
Venice
@3844 Be;
Vertlca
Figure 5]. C0
appIICa]
Hg”’°5.2. c0
apleCa}

Figure 2.6a. Simulated and measured biomass accumulation with the application of MPR
under irrigated conditions. Vertical bars represent standard errors of the mean
............................................................ 75

Figure 2.61). Simulated and measured biomass accumulation with the application of MPR
under rainfed conditions. Vertical bars represent standard errors of the mean . 76

Figure 3.1. Relationship between dry matter production and phosphorous concentration at
maturity. Arrows indicate values of critical P concentration .............. 110

Figure 3.2a The simulated distribution of the P soil pools and the P plant uptake during
the 1993 irrigated experiments for various additions of TPS. The fertilizers were
applied on Day 147. ............................................ 119

Figure 3.2b. The simulated distribution of the P soil pools and the P plant uptake during
the 1993 irrigated experiments for various additions of MPR The fertilizers were
applied on Day 147. ............................................ 120

Figure 4.1. Bean dry weight production for the crop grown under (A and B) rainfed and
(C and D) irrigated conditions with fresh applied fertilizer P. Arrows indicate crop
growth stages, and vertical bars represent standard errors (S.E.) .......... 153

Figure 4.2. Bean dry weight production for crop grown under (A and B) rainfed and (C
and D) irrigated conditions with annual applied P fertilizer. Arrows indicate crop
growth stages, and vertical bars represent standard errors (S.E.) .......... 154

Figure 4.3. Bean dry matter production as influenced by triple superphosphate (TSP).
Vertical bars represent standard errors of means (SE) .................. 157

Figure 4.4. Bean dry matter production as influenced by Minjingu phosphate rock (MPR).
Vertical bars represent standard errors of means (SE) .................. 158

Figure 5.1. Comparison of responses of bean crOp to residual, flesh, and annual
application of different levels of TSP under rainfed and irrigated conditions . . 194

Figure 5.2. Comparison of responses of bean crop to residual, flesh, and annual
application of different levels of MPR under rainfed and irrigated conditions . 195

Phrlie’

 

blamed Spec

Inparricular, 1
linear In a
In CemmL am
nguese to,
WWW: 19
W for Cam

Beans

CHAPTER 1

INTRODUCTION

Phaseolus vulgaris L. (F abaceae) a C3 plant is the most widely grown of the four
cultivated species of Phaseolus. The other three species are Phaseolus coccineus L. (scarlet
runner bean), Phaseolus Iunatus L. (lima bean), and Phaseolus acutrfolius A. Gray (tepary
bean). The P. vulgaris L. crop is known by various names such as kidney beans, haricot
beans, snap beans, dry beans, common beans, etc. In Eastern Africa in general, and Tanzania
in particular, they are usually referred to simply as beans in English and ”maharagwe" in
Kiswahili. In this study the name “beans” will be used. Beans are believed to have originated
in Central, and South America in Andean zones. In the 16th century the Spaniards and
Portuguese took beans to Europe, and horn there to Afiica and various parts of the world
(Pulseglove, 1968; Gentry, 1969). In the 1930s, new varieties were introduced in Tanzania
mainly for canning purpose (Macartney, 1966).

Beans are a major staple food worldwide, and different growth habits have been
developed to meet different yield, and yield stability requirements (Silbemagel and Harman,
1988). Been arltivars are classified according to growth habit and period of grth (Voysest
and Dessert, 1991). CIAT (Centro de Internacional de Agricultura Tropical) classifies beans

into determinate, and indeterminate genotypes. The determinate genotypes, have a main stem

and the later
main stem a'
I971). FurI
cmironmenr

Beans

3mm for the

‘3 ImPOrIam r

 

2

and the lateral branches terminate in an inflorescence, while the indeterminate genotypes have
main stem and the lateral branches are topped by vegetative meristem capable of continuous
organogenesis (Debouck, 1991). These growth difi‘erences are genetically controlled (Bliss,
1971). Further, bean varieties are referred to as ”early” or "late” in reference to the
environment where they are grown (V oysest and Dessert, 1991).

Beans are the most important grain legume crop grown in Tanzania. They are mainly
grown for their dry seed, but are also grown for fresh pod, and leaf consumption. They form
an important component of the diets of many people in Tanzania. Referred to as “food for
the poor”, beans are a particularly important source of protein for families with low income
and limited access to animal protein in rural areas. About 30% of produced grain is marketed
for urban populations where animal protein is becoming less afl‘ordable, while the remaining
is consumed at farm level. Although cultivars developed for monoculture cropping systems
are small-seeded, consumers prefer large seeded beans (i.e., >400 mg seed"). Also, most
farmers prefer bean straw to maize stover in feeding their dairy cattle because of bean straw's
high quality and proportion of edible stem portions (U do and Mlay, unpublished report
1984‘).

Beans are grown during the short and the long rainy seasons in those parts of
Tanzania where there is bimodal rainfall. However, they are commonly cultivated during the
long rainy season in most bean producing areas (Karel er al., 1981). Agronomic practices

used in bean production differ from one area to another depending on factors such as

 

1Urio, NA, and 6.1. Mlay. 1984. First progress report on diagnostic survey among smallholder dairy
farmers in Hai District-Tanzania, IDRC Project File No. 3-P-82—0035.

temporal an

constraints.

 

ale 3 Talbot
land preparaz:
such as mold?

For be
”311 as plant 5.
length of the

especially AID!

 

Mortal g

3

temporal and spatial arrangements on the basis of physical, biological, and social economic
constraints. One such practice, slash and burn, is usually used by farmers to clear the land
after a fallow period. Depending on whether zero or minimum tillage is used farmers begin
land preparation at the onset of the rainy season using hand-hoes, and animal traction devices
such as moldboard plows and tractors for those farmers who can afl‘ord the equipment.

For bean crops gown on traditional farms, there are no specific planting arrangements
such as plant spacing changes with cropping system, rainfall, soil type, crop compactions, and
length of the cropping season. Ridges are commonly built in contours across slopes,
especially along the Uluguru mountains in the Morogoro region, to reduce erosion. In
monocultural systems, planting methods usually follow conventional techniques used in high-
input systems. In intercrops or multiple cropping systems, one or two bean seeds are planted
in holes about 2.5 cm in diameter and 5.0 cm deep in the soil. High rainfall variability and low
soil water retention capacity are mainly responsible for poor bean crop establishment.

Few farmers apply irrigation. This is because irrigation water is not always available,
and irrigation facilities are not within economic reach of most farmers. However, when firrrow
irrigation is applied, the beans are usually gown on ridges. If the crop is not irrigated or
planted at the right time, it is mostly afi‘ected by water stress from flowering to early pod set,
and yield components are more likely to be affected by water stress (Stoker, 1974). Overall
response of beans to water stress depends on the cropping system. In Eastern AfiiCa, there
can be severe yield reduction in a bean crop gown in bean-maize intercrops because of the
strong competition of maize for water and nutrients (Fisher, 1977). However, an important

intercropping system found in Kagera region where beans are gown in association with

bananas and
drought stre
Bean

to 26°C. It
Immature u
Tammie bea
“‘0 months to
farmers in Ta:
great diversity
Wleties STOW
climbing beans
know been“

 

4

bananas and cofi‘ee plants. This bean-banana association plays an important role in reducing
drought stress in the associated bean crop, and improves the stability of the system.

Beans are generally gown in environments with air temperatures ranging from 16°
to 26“C. The length of the gowing season depends mainly upon the variety used and
temperature which in turn is affected by altitude. In the northern and southern highlands of
Tanzania, beans take about three months to mature, while in the lowlands, they take about
two months to reach maturity. Improved varieties are gown on a small scale by only a few
farmers in Tanzania. Most farmers gow ”mixed beans" or ”local varieties” which have a
geat diversity of gowth habit, vegetative characters; seed color, size and shape. Most
varieties gown by the farmers are low yielders. Determinate, indeterminate non-climbing and
climbing beans are gown. The exact number of the varieties gown in the country is not
known because often a variety is referred to by different names in difi‘erent parts of the
country (Karel et 01., 1981). There are a number of commercially available varieties which
include Canadian Wonder, Selian Wonder, Lyamungo 85, Uyole 84; and Kabanima supplied
by Tanzania Seed company (TAN SEED).

Bean production statistics published annually by Food and Agriculture Organization
(F AO) vary in reliability fi'om reasonably accurate estimates to imprecise approximations
(Adams, 1984). For example, bean production for 1992/93 from the Ministry of Agriculture
(MOA) was estimated at 233,500 metric tons from 179,600 ha, i.e., 1,300 kg ha'1 (MOA,
1993) (these estimates are on the high side). Similar estimates were given for previous years
(Table 1.1). Farmers' yields usually average 600 kg ha’1 (Silbemagel and Teri, 1990).

However, improved varieties used in research trials have given yields of more than 1000 kg

h'IlKarele
olbcansare

determinate

produced in r
application 0
yields in Tar.

altitudes (> 1

Table I I AVC

\

19m

/

 

5
ha“ (Karel et 01., 1981; Silbemagel and Teri, 1990). In the Arusha region, about 25,000 ha
of beans are gown on a large scale on contract to foreign seed firms. The cultivars gown are
determinate bushy types selected for their acceptability in Europe as snap beans, and are
produced in monoculture. Since the commercial farms receive more inputs, including aerial
application of pesticides, their yields can be as high as 1600 kg ha“. In general terms, bean
yields in Tanzania are higher (i.e., > 1000 kg seed ha") in agricultural zones found in high

altitudes (> 1000 m) than in areas with elevations below 1000 m.

Table 1.1 Average annual national bean production in Tanzania from 1988/89 to 1992/93”.

 

 

m
1988/89 1989/90 1990/91 1991/92 1992/93
Area‘ 198.54 180.65 232.00 197.68 179.60
Prod.” 337.31 319.18 290.80 256.99 233.50

Yield" 1699.00 1767.00 1253.00 1300.00 1300.00

 

* Source: Ministry of Agriculture (MOA, 1993).
'Area is in ‘000 hectares

”Production is in ‘000 metric tons

cYield (seed) is in kg ha"

Karel er a]. (1981) summarized some of the major insects that attack bean plants in
Tanzania. Some of the most important insects include adult chrysomelid beetles (00theca
benm‘ngsem’ Weise) which feed on bean seedlings ,and plants, and can reduce bean yield by

18 to 31% (Karel and Rweyernarnu, 1985) while bean aphids (Aphisfabae Scop.) can reduce

yield by 37%. Yield
systems than in me
preside a uniform b
ofthebmn aphids i
1992). Few farmer

to suppress insect ;

Esidence sh
inmost parts of Ta:
yield and (ii) by re
associated with be;
More and Latin .

Europe (Beebe a:

Phaseolr'cola (Psp

 

6
yield by 37%. Yield decline, however, has been found to be less serious in multiple cropping

systems than in monoculture systems partly because the diversity of crops gown does not
provide a uniform backgound for pest manipulation. This has been shown on the incidence
of the bean aphids in the bean-maize intercrop (Kennedy et al., 1961; Ogenga-Latigo et al.,
1992). Few farmers use insecticides in Tanzania. They rely on the natural mortality factors
to suppress insect populations.

Evidence shows that diseases account for more severe bean losses than insect damage
in most parts of Tanzania. Diseases afl‘ect the bean crop in two ways: (I) by causing loss of
yield and (ii) by reducing the quality of the seed (Karel er al., 1981). More pathogens are
associated with beans in
Afiica, and Latin America than in temperate bean-gowing regions of North America and
Europe (Beebe and Pastor-Corrales, 1991). Hallo blight (Pseudomonas syringae pv
phaseolicola (Psp)); bean common mosaic (BCMV) a viral disease; rust (Uromyces
appendiculatus var. turpendiculams), and angular leaf spot (Phaeoisariopsr‘s griseola) are the
most economically damaging bean diseases in Tanzania (Teri er al., 1990). Intercropping bean
with maize has been reported to reduce disease incidence and severity in the bean crop
(Katmzi er al., 1987). However, in a study that used the Canadian Wonder bean cultivar and ‘
maize intercrop, Mabagala and Saettler (1991) concluded that intercropping may increase
seed infection with halo blight bacteria Avoidance is currently a principal method for disease
management by small bean producers. In Tanzanian situations where farmers rarely use
purchased inputs in bean production, breeding for broad, multiple resistance is suggested as

the best strategy of disease control (Teri er al., 1990).

We
However i
Chemical v
systems. l.
the times t
in the area
between tw
STOW unde
labor require
large size, h

as the mean

bOed and let

7

Weed infestation is another major limiting factor in bean production for Tanzania.
However it is usually less of a problem in intercrOpping than in monocropping systems.
Chemical weed control is virtually nonexistent in intercrop, multiple and mixed cropping
systems. In areas where monoculture is practiced, weed control is done manually two or
threetimesduringtheeariypartofthegowing seasondepending on thetype ofweed species
in the area. Studies in Morogoro have shown that the severity of weed competition is
between two to four weeks after planting which is regarded as the critical period for beans
gown under monoculture conditions (Mbuya er al., 1986). Manual weed control is a high
labor requirement in small-scale bean production systems in Eastern Afiica. Thus the use of
large size, high leaf area index and high yielding bean cultivars is sometimes recommended
as the means of suppressing weeds (Wortmann, 1993). In most cases, weeds are pulled or
hoed and left on the gound to dry.

Soil fertility research in Tanzania was initially commodity oriented with emphasis on
estate crops such as cotton, coffee, sisal, and tea. Since 1974 studies on other crops,
including bean response to fertilizers, have been conducted. Currently, CIAT maintains a
comprehensive progam on bean research, and works cooperatively with the MOA under
national progarnrnes. A special initiative which was launched in 1980 entitled 'The USAID
Title XII Bean/Cowpea Collaborative Research Support Progarn links American Universities
with national research institutions in bean research. For example, Washington State
University (Pullman) collaborates with Sokoine University of Agriculture (Morogoro),
Tanzania in dealing with problems afl‘ecting production , and utilization of beans on

subsistence farms.

   

General be

identified by B

seed yield (Manriq

 

 

General benefits of fertilizer use in developing counties such as Tanzania have been
identified by Baanante er al. (1989). Only one inoculant ”nitrosua" has been developed for
beans in Tanzania, and it is currently being tested under field conditions (Salema, 1987;
Msurnali, 1992). In most cases, the efi‘ects on nitrogen (N), phosphorous (P), and potassium
(K) have been evaluated together in the same studies (Mkeni, 1989).

Tire application of P in beans results in increased plant gowth (Yan et al., 1995a), and
seed yield (Manrique, 1986; Lynch er al., 1991; Ya er al., 1995b), N—fixation (Graham and
Rosas, 1979), and radiation-use-efiiciency (RUE) (Manrique, 1993). However, research
findings on beans reported by Giller et a1. (1988), and Semoka er a]. (1990) on the use of
Minjingu phosphate rock (MPR) as a direct applied P fertilizer have shown conflicting results.
The Diagnosis and Recommendation Integated System (DRSI) shows that the critical
nutrient level (CNL) for the bean crop gown in East Afiica is 0.32% P (Wortmann et al.,
1992).

Using the Bray-I method as recommended by Nandra (1974), the available P in the
Morogoro region is categorized by Singh and Uriyo (1980) as 0-15, 15-25, 25-30, and > 30
ppm as low, medium, high and very high Samki and Harrop (1984) reported the occurrence
of acid soils. in every district of Tanzania. The FAQ-UNESCO (1974) soil map shows that
Ultisols and Oxisols cover a large area of Tanzania. These soil orders are generally
characterized by low available P, high fixing capacity, and low pH (Sanchez and Uehara,
1980). According to Sanchez and Uehara (1980), soils with high P fixing capacity are those
which require addition of at least 200 kg P ha" to provide an equilibrium soil solution

concentration of 0.2 mg P kg" of soil. However, current definition by CIAT shows that such

soils should have
(Molmmye and l-

The variat
sheen that where s
the dominant inorgz
were the dominan
production is pract
most of the other
imllortant, yields c
deficiencies

AlthOUgh f,

very low averagim

 

9

soils should have the clay content of the top soil >3 5%, and free Fe oxide/clay ratio >1.5
(Mokwunye and Hammond, 1992).

The variation of soil types is very high in Tanzania. Uriyo and Singh (1978) have
shown that where soils were young, the parent material rich in P-bearing minerals, Ca-P was
the dominant inorganic-P fiaction and where the soils were highly weathered, Al-P and Fe-P
were the dominant fractions. In the western Arusha region (where commercial bean
production is practiced), the bean crop thrives on the volcanic soils (Allen et a1, 1989). In
most of the other areas such as Tanga, Kagera, and Morogoro where bean production is
important, yields can be as low as 158 kg ha'1 (FAO, 1991) due to low pH, and nutrient
deficiencies.

Although fertilizers play a sigrificant role in increasing crop yields, their use is still
very low averaging 7.6 kg ha'1 of nutrient in Tanzania (Baanante er al., 1989). Low P
availability is a primary constraint to bean production in Afiica, and more than 50% of the
crop is gown on severely P deficient soils (Wortmann and Allen, 1994). Tanzania depends
on fertilizer imports to meet all of its domestic requirements. The rate of recommended P for
a bean crop which is 50 kg P20, ha‘l (Samki and Harrop, 1984; Semoka et al., 1990) is too
costly for most of the farmers who produce more than 80% of the beans consumed in the
country. Further, the only fertilizer production plant in Tanzania (Tanga based, Tanzania
Fertilizer Company) ceased production in 1991 due to defective machinery. On the other
hand, the indigenous phosphate deposits of soft MPR found at Minjingu in the Arusha region
in Tanzania, is reported to be suitable as a directly applied phosphate rock (DAPR) fertilizer

on various crops (Mwambete, 1991; Van Kauwenbergh, 1991). Phosphate rock (PR) is a

general term used.
concentration of p
phosphate ore, an.‘

Phosphorc
application, and szlI

IBaITow, 1980,80?

    
   

in the literature on

Pinapamrreexper

 

MPR comlmeci to

10

general term used to describe naturally occurring mineral assemblages that contain a high
concentration of phosphate mineral. This term, PR is used to describe both unbeneficiated
phosphate ore, and concentration products (Van Kauwenbergh, 1992).

Phosphorous fertilizer provides the P element required for gowth in the year of
application, and subsequent years. Thus, P fertilizer residual values have been recorded
(Barrow, 1980; Bolland, 1993; Selles et a1. 1995), however, reports on such effects are scanty
in the literature on Tanzanian soils. Marandu er a1. (1975) observed the residual efi‘ects of
P in a pasture experiment conducted in Morogoro. Other field studies using cotton by Scaife
(1968), maize by Mowo and Gama (1988), and IFDC (1991) indicated good performance of
MPR compared to triple superphosphate (TSP) or double superphosphate (DSP) particularly
after the first two years following phosphate application.

Although systems research (analysis) is often thought of as computer work, this
approach comprises, first, an analysis of components and relationships of the system, and
secondly, a synthesis phase. The synthesis phase may involve either development of systems
or the more efficient use of the original system (Wright, 1971). According to Penning de
Vries er a1. (1993 ), systems research approach involves the identification of systems,
subsystems and important processes in agricultural production. Therefore, systems research
approach brings together researchers through multi disciplinary research approach, and this
may lead to appropriate conclusions and recommendations. In order to come up with
appropriate conclusions and recommendations, there is a need to firlly understand various

subsystems (components) in the system.

Swmur
bmgmmmm
aseeathcr, allure
Such approach he
decision support 3
International Ben;
00368, 1993, L'eh;

hmawh

defined as a proce

 

 

mi. mots etc, dt

seed VICId. BeCaU

11

Systems analysis is not an alternative to the traditional methods of research but a tool
to integate information fiom single-fictor methods with information about other factors such
as: weather, cultivar, soil characteristics and crop management which may limit crop gowth.
Such approach has been taken by a team of international scientists who have developed a
decision support system for agotechnology transfer (DSSAT) crop models through the
International Benchmark Sites Network for Agotechnology Transfer (IBSNAT) project
(Jones, 1993; Uehara and Tsuji, 1993).

In an excellent review on crop simulation by Whisler er a1. (1986) crop simulation is
defined as a process by which a model acts like a real crop by gadually gowing leaves,
stems, roots, etc, during the season, and finally predicts some final state such as biomass and
seed yield. Because of magnitude of possible management x genotype x environmental
interactions, traditional field experimentation has proved to be limited in its ability to identify
improved crop production practices. This is especially true in the tropics where each farmer
and each firm is unique, and its results are site-, season-, cultivar-, and management-specific
(Nix, 1980; IBSNAT, 1990). Thus the approach of matching the crop requirements to the
land characteristics through systems analysis and crop simulation is important (IB SNAT,
1990).

A number of published crop simulation models have been listed by Whisler er a].
(1986), and Ritchie (1991 and 1994). Various levels at which crop simulation models may
be applied have been identified as field, farm, regional, and national level (Thornton, 1991).
Using the available models, such as those by the IBSNAT project, and the DSSAT, making

simplified representations that give similar results to those of actual cropping systems (Singh,

unpublished rep
work in problem
matter areas (Rite
rpowerful meth
1994).

Crop mod
381W technc
because they redu
199‘” Although
Am“ (Rating er c
MOI any documei
there is also limiter
under field conditit
hm beat reported \
ILil) related gm“
determined under I

Therefore {

 

BEANGRO Crop Sir

response; (2) T0 el

 

' 12
unpublished report 1989’) is possible. This also establishes opportunities for scientists to
work in problem-solving interdisciplinary goups because of the linkages between subject
matter areas (Ritchie, 1989). When models are evaluated for a specific region, they provide
a powerful method for replacing much trial-and-error type experimental research (Ritchie,
1994).

Crop models are powerful tools that can be used to facilitate identification of suitable
agricultural technologies. This is important for resource-poor firmers in developing countries,
because they reduce time, and may reduce cost of doing repetitive experiments (Muchena,
1994). Although there has been a number of crop model simulation application studies in
Afiica (Keating er al., 1992; Singh et al., 1993; Muchena, 1994; Thornton er al., 1995), there
is not any documented work on use of crop modeling experiments in Tanzania. In general,
there is also limited published work on crop simulation using bean crop simulation models
under field conditions (Gutierrez et al., 1994). Further, no field studies on bean irrigation
have been reported whereby the total dry matter (TDM), plant part dry matter, leaf area index
(LAI), related gowth analysis, P removal from the field, and residual P effectiveness were
determined under Tanzanian conditions.

Therefore the objectives of this study were: (1) To evaluate the P version of the
BEANGRO crop simulation model for Tanzanian conditions for sensitivity to phosphorous
response; (2) To evaluate the P uptake, partitioning and its accumulation by a bean

(Pinseolw vulgaris L.) crop gown under different sources, and levels of P fertilizers under

 

2Singh, U. 1989. Introduction to modeling Report for training progam on computer simulation for
crop gowth and fertilizer responses. May 15-26, 1989. International Fertilizer Development Center, PO. Box
2040, Muscle Shoals, AL. 35660.

rainfed, and irr
influenced by Mr
fertilizer, in com;
effectimess of .‘

grown, and yield

 

 

l3

rainfed, and irrigated conditions; (3) To assess the growth, and yield of a bean crop as
influenced by Minjingu phosphate rock (MPR) as a directly applied phosphate rock (DAPR)
fertilizer, in comparison with triple superphosphate (TSP) fertilizer; (4) To assess the residual
efi‘ectiveness of MPR as a DAPR fertilizer in comparison with TSP fertilizer on a bean crop

grown, and yield under rainfed and irrigated conditions.

 

Adams, MW. 19'.
Bean/C04:
Lansing ,\

”9101.111. Dc
their cons:
PlOductior.

Baanante, CA, 3
develOping

 

Banow, NJ. 1980
358. In FE
Phosphorol

In A Van :
CTOp impro

W180": L

BIBLIOGRAPHY

Adams, M.W. 1984. Bean-Cowpeas: Production constraints and national programs.
Bean/Cowpea Collaborative Research Support Program. Mich. State Univ., East
Lansing, MI.

Allen, D.J., M. Dessert, P. Trutrnann, and J. Voss. 1989. Common beans in Afiica and
their constraints. p. 9-31. In HR Schwartz, and M.A Pastor-Corrales (eds) Bean
production problems in the tropics. 2nd ed. CIAT, Cali, Colombia.

Baanante, CA, B.L. Bumb, and PK. Thornton. 1989. The benefits of fertilizer use in
developing countries. Paper Series P8. 1989 2M. IFDC, Muscle Shoals, AL.

Barrow, NJ. 1980. Evaluation and utilization of residual phosphorous in soils. p. 333-
358.11: RE. Khasawneh, E.C. Sample, E.J. Khamprath (eds) The role of
phosphorous in agriculture. ASA, CSSA, and SSSA Madison, WI.

Beebe, 8.15., and M. Pastor-Corrales. 1991. Breeding for disease resistance. p. 561-617.
In A Van Schoonhoven, and O. Voysest (eds). Common beans: Research for
crop improvement. CAB. International. Wallingford, U.K.

Bliss, RA 1971. Inheritance of growth habit and time of flowering in beans, Phaseolus
vulgaris L. J. Am. Soc. Hort. Sci. 96:715-717.

Bolland, M.D.A 1993. Residual value of superphosphate and Queensland rock phosphate
measured using yields of serradella, burr medic and subterranean clover grown in
rotation with wheat and bicarbonate-extractable soil phosphorous. Commun. Soil
Sci. Plant Anal. 24: 1243-1269.

Debouck, D. 1991. Systematics and morphology. p. 55-118. I_n A Van Schoonhoven and

O. Voysest (eds). Common beans: Research for crop improvement. CAB.
International. Wallingford, UK.

14

15

Fisher, NM 1977 . Studies in the mixed cropping. 1. Seasonal differences in relative
productivity of crop mixtures and pure stands in the Kenya Highlands. Expl. Agric.
13: 177- 1 84.

Food and Agriculture Organization (FAO). 1991. Quarterly bulletin of statistics. F AO,
Rome. Italy.

Food and Agriculture Organization - United Nations Education Scientific and Cultural
Organization (FAG-UNESCO). 1974. Soil map of the world. FAO, Rome. Italy.

Gentry, HS. 1969. Origin of the common bean, Phaseolus vulgaris L. Econ. Bot. 23:55-
69.

Giller, K.E., F. Amijee, and SJ. Brodrick. 1988. Soil factors limiting nitrogen fixation and
production of Phaseolus vulgaris L. in East Afiica. In 1988 Annual Report.
Depart. of Biochemistry and Biological Sciences, Wye College, Univer. of
London. Kent, UK.

Graham, P.H., and J.C. Rosas. 1979. Phosphorous fertilization and symbiotic nitrogen
fixation in common bean. Agron. J. 71 :925-929.

Gutierrez, AP., E.J. Marriot, JR Cure, C.S. Wagner-Ridder, C.K. Ellis, and AM.
Villacorta. 1994. A model of bean (Phaseolus vulgaris L.) growth types I-III.
Factors afl‘ecting yield. Agri. Syst. 44:35-63.

International Benchmark Sites Network for Agrotechnology Transfer (IBSNAT). 1990. p.
7-8. In IBSNAT Project goals and objectives. Honolulu, HI.

International Fertilizer Development Center (IFDC). 1991. Use of indigenous resources of
phosphate rock. p. 10-15. In IFDC Annual report 1991. Muscle Shoals, AL.

Jones, J.W. 1993. Decision support systems for agricultural development. p. 459-471. In
F.W.T. Penning de Vries, P.S. Teng, and K. Metselaar (eds). Systems approaches
for agricultural development. Kluwer Academic Publ. Dordrecht, The
Netherlands.

Karel, AK., B.J. Ndunguru, M. Price, S.H. Semuguruka, and BB. Singh. 1981. Bean
production in Tanzania. p. 122-154. In Potential for field beans in Eastern Afiica:
Proc. Regional Workshop, Lilongwe, Malawi, 9-14 March, 1980. Centro
. Internacinal de Agricultura Tropical (CIAT), Calli, Colombia.

Karel, AK, and
bennmgs
14662-6

Ramm’ ', AL, J
intercrop]
80.

liming, BA, R.

response I
FWI. Pe
for agricul

Kennedy.1$.,c.l
field. II],

 

Limit 1., A. La“,-
resDons»: u

Mabagala. R3, 3;
infection u
Coop. 34. j

 

 

l6

KareL AK., and CL. Rweyemamu. 1985. Resistance of foliar beetle Oatheca
' benningsenni (Coleoptera: Chrisomelidae) in common beans. Environ. Entomol.
14:662-664.

Katunzi, AL., J .M. Teri, K.P. Sibuga, and RN Misangu. 1987. The efi'ect of
intercropping and fertilizer application on bean diseases and yield. Bean Res. 2:66-
80.

Keating, B.A, KL. McCown, and BM. Wafirla. 1992. Adjustment of nitrogen inputs in
response to a seasonal forecast in a region of high climate risk. p. 23 5-251. In
F.W.T. Penning de Vries, P. Teng, and K. Metselaar (eds). Systems approaches
for agricultural development. Kluwer Academic Publ. Dordrecht, The Netherlands.

Kennedy, J .S., C.O. Booth, and W.J.S. Kershaw. 1961. Host finding by aphids in the
field. III. Visual attraction. Ann. App]. Biol. 4921-21.

Lynch, 1., A Lauchli, and E. Epstein. 1991. Vegetative growth of common bean in
‘ response to phosphorous nutrition. Crop Sci. 31:380-387.

Mabagala, RB, and AW. Saettler. 1991. The efi‘ect of cropping systems on bean seed
infection with Pseudomonas syringae pv phaseolicola. Ann. rep. of Bean Improv.
Coop. 34:23-24.

Macartney, J .C. 1966. The selection of haricot bean varieties suitable for canning. E.
Afiica Agri. For. J. 32:214-217.

Manrique, LA 1986. Field bean leaf area development and seed yield as affected by
phosphorous application on a Typic Tropudult in Panama. Commun. Soil Sci.
Plant Anal. 17: 135-147. '

Manrique, LA 1993. Radiation-use-efficiency of field beans as afl‘ected by soil
phosphorous management. Commun. Soil Sci. Plant Anal. 24:623-631.

Marandu, W.Y.F., AP. Uriyo, and H0. Mongi. 1975. Residual effects on NPK fertilizer
application on yield of maize and pasture on ferrisol at Morogoro. E. Afiica Agri.
For. J. 40:400-407.

Mbuya, O.S., K.P. Sibuga, and AN. Minjas. 1986. Critical period for weed competition in
beans (Phaseolus vulgaris L.) under monoculture. Bean Res. 12116-123.

MlfllSUy 0f Agl'lxl
1985/37-

Mkcni, P. 1989.
in Tanzar

 

 

 

MOWO, 1 G_‘ and
cotton yie]
56-58. In 3

meeting 1
Tanzania

MW GP. 19

intercrop;
7: 17-22.

“When P. 199.
predlCIlOr
CHV’ironm‘
deveIOpin

Mwambele. I. lg

425.30

Burns HA 1980.
HQ

17

Ministry of Agriculture (MOA), 1993. Basic data on pulses production by region for
1985/87-1991/93. Tanzania.

Mkeni, P. 1989. A review of soil fertility and fertilizer use research and recommendations
in Tanzania. p. 161-195. In H. Ssali, and LB. Williams (eds). Proc. of East and
Southeast Afiican Fertilizer Management and Evaluation Network Workshop,
IFDC, May 27-29, 1987. Nairobi, Kenya.

Mokwunye, U. and Hammond, LL. 1992. Myths and science of fertilizer use in the
tropics. p. 121-133. In R Lal, and RA. Sanchez (eds). Myths and sciences of

soils in the tropics. Proc. International Symposium, Les Vegas, Nevada, 17
October, 1989. SSSA Spec. Publ. No. 29. ASA, CSSA, SSSA, Madison, WI.

Mowo, J. G., and B. M. Gama. 1988. Effect of application of MRP and TSP on seed
cotton yield and soil properties in acid soils of western cotton growing areas. p.
56-58. In F.B.S., Kaihura, J. Floor, and ST. Ikerra. (eds). Proc. phosphate
meeting, Tanga, TanzaniaMisc. Publ. M9. Mingano Agri. Res. Inst, Tanga,
Tanzania.

Msumali, GP. 1987. Effectiveness of maize in depleting soil mineral nitrogen when
intercropped with beans for the purpose of rhizobium strain evaluation. Bean Res.
7: 1 7-22.

Muchena, P. 1994. Experiences of developing countries in using crop modeling
predictions. In P.F. Uhlir, and 6C Carter (eds) Crop modeling and related
environmental data: A focus on applications for arid and semi-arid regions in
developing countries. 1:125-131. CODATA, Paris, France.

Mwarnbete. I. 1991. Tanzania fertilizer mineral deposits. Fert. Res. 30:181-185.

Nandra, SS. 1974. Evaluation of soil test methods for the estimation of available
phosphorous in some Tanzanian soils. A greenhouse study. B. Afiica Agri. and For. J.
4:25-30.

Nix, HA 1980. Strategies for crop research. Proc. Agron. Soc. of New Zealand. 10:107-
1 10.

Ogenga-latigo, 5+
row spam:I
(Aphisfa?

Penning dc Vries.

agricultural

Pulseglovc, 1W. 1

RichicJI. 1989
nlwm
Proc. Am

R11ch1e,J.T. 1991
In RC. M.
10d manag,
St. Lucia I
lrl'eillingfor;I

Ritchie, IT. Classi

airid and Se
France.

Salem MP. 1937
Producnon.

 

18

Ogenga-Latigo, M.W., J .K.O. Ampofo, and CW. Baliddawa. 1992. Influence of maize
row spacing on infestation and damage of intercropped beans by the bean aphid
(Aphisfabae Scop.) 1. Incidence of aphids. Field Crops Res. 30:111-21.

Penning de Vries, F.W.T., P.S. Teng, and K. Metselaar. 1993. Systems Approach for
agricultural development. Kluwer Academic Publ. Dordrecht, The Netherlands.

Pulseglove, J .W. 1968. Tropical crops: Dicotyledons. Longrnans, London, England.

Ritchie, J .T. 1989. Crop models and decision making: Scenario for the future. p. 106-117.
In A Weiss (ed.) Climate and agriculture systems approaches to decision making.
Proc. Am. Meteor. Soc. 5-7 March 1989. Charleston, SC.

Ritchie, J .T. 1991. Specifications of the ideal model for predicting crop yields. p.97-122.
In RC. Muchow and J .A Bellamy (eds) Climatic risk in crop production: Models
and management for the semi-arid tropics and subtropics. Proc. Intnl. Symposium,
St. Lucia, Brisbane, Queensland, Australia. July 2-6, 1990. C.AB. InternationaL
Wallingford, U.K.

Ritchie, J.T. Classification of crop simulation models. 1994. In P.F. Uhlir and QC. Carter
(eds) Crop modeling and related environmental data: A focus on applications for
airid and semiarid regions in developing countries. 1:3-14. CODATA, Paris,
France.

Salema, MP. 1987. Field evaluation of bean rhizobial strains for use in inoculum
production. Bean Res. 22207-209.

Samki, J.K., and IF Harrop. 1984. Fertilizer recommendations in Tanzania on district
basis. Tech. Paper: URT/73/006 National Soil Sci. Agri. Res. Inst. Mlingano,
Tanzania.

Sanchez, PA, and G. Uehara. 1980. Management considerations for acid soils with high

' phosphorous fixation capacity. p. 471-514. In F.E. Khasawneh, E.C. Sample, and

13.1 Kamprath (eds) The role of phosphorous in agriculture. ASA, CSSA, SSSA,
Madison, WI.

Scaife, MA. 1968. The efl‘ect of cassava fallow and various manurial treatments on cotton
at Ukiriguru, Tanzania. E. Afiica Agri. For. J. 33:231-237.

Selles F, CA. C
plant and

Semoka, J.M.R.,
utilization
559. In J I
12. Cali, (

 

Silbemagel, MI .
of comme
resources

Slbemagel, M],
University
delemiinat
Bean-"CR5.
COllabOrm
E151 Lansi1

 

 

 

Singh BR. and A
SOllS N anc

Singh‘ U. 1989 if
SimlllallOn
Muscle Sh

Singh, U, PK. T}
100] for SO
“let P. 1
dell'ClOme

M 1M, M1. 81

Ofprogres;
Vol. 5‘ CL

19

Selles, F., GA Campbell, and RP. Zentner. 1995. Efl‘ect of cropping and fertilization on
plant and soil phosphorous. Soil Sci. Soc. Am. J. 59:140-144.

Semoka, J.M.R, O.T. Edje, and P.M.S. Mkeni. 1990. Prospects for phosphate rock
utilization in the development of sustainable cropping systems with beans. p. 551-
559. In J .B. Smithson (ed.) Bean Res. Vol. 5 CIAT Afiica Workshop Series NO.
12. Cali, Columbia.

Silbemagel, M.J., and RM. Harman. 1988. Utilization of genetic resources in development
of commercial bean cultivars in the USA p. 541-566. In P. Gept (ed.) Genetic
resources of Phaseon beans. Kluwer Academic Publ., Boston, MA

Silbemagel, M.J., and J.M. Teri. 1990. Tanzania/Washington State University/Sokoine
University of Agriculture: Breeding beans (Phaseolus vulgaris L.) for disease and
determination of socioeconomic impact of smaller-farm families. p. 56-59. In
Bean/CRSP (MSU) Proc. of International Res. Meeting of Bean/Cowpea
Collaborative Res. Support Program, April 30-May 3, 1990. Mich. State Univ.,
East Lansing, MI.

Singh, BR, and AP. Uriyo, 1980. The relationship between response to N and P, and
soils N and P. 'J. Agric. Sci. 94:247-249.

Singh, U. 1989. Introduction to modeling. Report for training program on computer
simulation for crop growth and fertilizer responses. May 15-26, 1989. IFDC,
Muscle Shoals, AL.

Singh, U., P.K. Thornton, AR Saka, and J .B. Dent. 1993. Maize modeling in Malawi: A
tool for soil fertility research and development. p. 254-273. In F .W.T. Penning de
Vries, P. Teng, and K. Metsellar (eds) Systems approaches for agricutural
development. Kluwer Academic Publ. Dordrecht, The Netherlands.

Stoker, R 1974. Effect of water stress on dwarf beans at different phases of growth. NZ.
Exp. Agri. 2:13-15.

Teri, J .M., M.J. Silbemagel, and S. Nchimbi. 1990. The SUA/WSU Bean/CRSP: A review
of progress in disease resistance. p. 166-189. In J .B. Smithson (ed.) Bean Res.
Vol. 5. CIAT Africa Workshop Series, NO. 12. Cali, Columbia.

 

Thornton. PK 19';
developmer-
Shoals, Al

Thornton, PK, A}
Application
Expl. Agric

tom, G. and G 3.
de Vries, P 5
development

l'1i1'0,.~‘iP..and B R
phosphorous
profiles in Ta

Van Kauwenbergh, S
Africa. F en. i

\an Kauwenbergh, S
economic con
eniionment. ll

YOESESL 0., and M I
119-162

 

_ In A
for em impro

20

Thornton, PK. 1991. Application of crop simulation models in agricultural research and
development in the tropics and subtropics. Paper Series P-15. IFDC, Muscle
Shoals, AL.

Thornton, P.K., AR Saka, U.Singh, D.J.T. Kumwenda, J.B. Brink, and J.B. Dent. 1995.
Application of a maize crop simulation model in the central region of Malawi.
Expl. Agric. 31:213-226.

Uehara, G., and G.Y. Tsuji. 1993. The IBSNAT project. p. 505-513. In F.W.T. Penning
de Vries, P.S. Teng, and K. Metselaar (eds) Systems approaches for agricultural
development. Kluwer Academic Publ. Dordrecht, The Netherlands.

Uriyo, AP., and BK Singh. 197 8. Studies on distribution of inorganic forms of
phosphorous and their relationships with the extent of weathering in some soil
profiles in Tanzania. E. Afric. Agric. For. J. 43:344-354.

Van Kauwenbergh, SJ. 1991. Overview of phosphate deposits in East and Southeast
Afiica. Fert. Res. 30:127-150.

Van Kauwenbergh, SJ. 1992. The global phosphate rock resource base-technical and
economic considerations. p. 27-39. In J .J . Schultz (ed.) Phosphate fertilizer and
envionment. IFDC, Muscle Shoals. AL.

Voysest, 0., and M. Dessert. 1991. Bean cultivars: Classes and commercial seed types. p.
119-162. In A. Van Schoonhoven and O. Voysest (eds) Common beans: Research
for crop improvement. C.AB. Intemantional, Wallingford, U.K.

Whisler, F.D., B. Acock, D.N. Baker, RE. Fye, H.F. Hodges, J .R. Lambert, H.E.
Lemmon, J.M. McKinion, and V.R Reddy. 1986. Crop simulatiom models in
agronomic systems. Adv. Agron. 40:141-208.

Wortmann, CS. 1993. Contribution of bean morphological characteristics to weed
suppression. Agron. J. 85:840-843.

Wortmann, CS, and DJ. Allen. 1994. Afiican bean production environments: their
definition, characteristics and contraints. Network on Bean Research in Afiica.
Occassional Paper Series 11. Dar-es-Salaam, Tanzania.

Wortmanrt C5 .
integrate
Nutr. 15

Wright, A 1971
Anderson
Aust.

Yin. X, 1?. Lynr
in commall
35:1086-1'

Yan. X, SE. Bee

0f commo
1099

 

 

21

Wortmann, C.S., J. Kisakye, and CT. Edje. 1992. The diagnosis and recommendation
integrated system for dry bean: Determination and validation of norms. J. Plant
Nutr. 15:2369-2379.

Wright, A 1971. Farming systems, models and simulation. p. 1-22. In J.B. Dent, and JR.
Anderson (eds) Systems analysis in agricultural management. Wiley. Sydney,
Aust.

Yan, X., J .P. Lynch, and SE. Beebe. 1995a. Genetic variation for phosphorous efficiency
in common bean in contrasting soil types: I. Vegetative response. Crop Sci.
35: 1086-1093.

Yan, X., S.E. Beebe, and J .P. Lynch. 199 b. Genetic variation for phosphorous efliciency
of common bean in contrasting soil types: H. Yield response. Crop Sci. 35:1094-
1099.

snrmrrox OE-

rorrrosan

A field Stat:
Soione Unversity i
region- of Tanzania
and evaluate the p.
phosphorous (P) res;

mode} BEASGRO. 1

 

0124 111‘2 for both
WWW block des

91? 11211 Cool

A .
‘4 P835 1111

It, .
unlit-131’ (46% P0
“Specific rates at p

..
AP, 88 kg P ha"~hiz

CHAPTER 2
SIMULATION OF BEAN (Phaseolus vulgaris L.) YIELD AND BIOMASS RESPONSE

TO PHOSPHOROUS APPLICATION UNDER TANZANIAN CONDITIONS

ABSTRACT

A field study was conducted during the 1993 and 1994 growing seasons at the
Sokoine University of Agriculture farm (6° SO'S, 37° 39'E at 525 m altitude) in the Morogoro
region of Tanzania. The objectives of the study were to collect a minimum data set (MDS)
and evaluate the performance of a bean (Phaseolus vulgaris L.) crop for sensitivity to
phosphorous (P) response under rainfed and irrigated conditions using the bean simulation
model BEANGRO. The bean cultivar, Canadian Wonder was planted at 25 plants m'2 in plots
of 24 in2 for both experiments. The experimental plots were planted in a complete
randomized block design with three replications. The Minjingu phosphate rock (MPR) sofl
ore that could pass through a 75 - micron screen and triple superphosphate (TSP) were used.
Both TSP (46% P205) and MPR (3 0% P205) were broadcast and incorporated in each plot
at respective rates at planting. Seven treatments were used in the study as follows: 0 kg P ha"
- control; 22 kg P ha"-low rate-TSP; 22 kg P ha“-low rate-MPR; 44 kg P ha"-medium rate-

TSP; 88 kg P ha"-high rate-TSP; 88 kg P ha“-medium rate-MPR; and 176 kg P hal -high

22

rate-MPR The 1

rainfed experime

 

The cro;
obtained from the "
lust non-limitini
to life cycle, pod
validation was n~
d“northern )ielt
Devin mom

Wm 0f flou

 

under irrigated c
BEKVGRO can

MUM bl differ
the irrigated COnc
major 136105 in be

mode] shOUld be

Confidence in the r

23

rate-MPR The irrigated arperiment was sown on an Alfisol soil with a soil pH 5.43, and the
rainfed experiment was sown on an Oxisol soil with a pH 4.50.

The crop simulation model was calibrated using Canadian Wonder cultivar data
obtained fiom the 1993 well-fertilized (88 kg P ha"-TSP) irrigated experiment (which had the
least non-linriting factors). Adjustments were made in cultivar-specific coemcients related
to life cycle, pod, seed, biomass yield, seed size, and maximum leaf area index (LAI). Model
validation was made by comparing model predicted with field measured data for growth,
development, yield and seed yield components obtained from the 1993 and the 1994 seasons.
Despite incidence of pests and diseases during the study, the model accurately predicted the
occurrence of flowering, physiological maturity, pod, seed, biomass yield, and maximum LAI
under irrigated conditions. The conclusion drawn fi'om this study is that the P version
BEAN GRO can be used to assess growth, development and yields of a bean crop as
influenced by difl‘erent fertilizer P rates, planting dates, plant population, bean cultivar under
the irrigated conditions in Tanzania. Results suggest that because diseases and pests are
major firctors in bean yield reductions, their efl‘ects need to be incorporated in the model. The
model should be tested in areas with high bean production of Tanzania so as to build up

confidence in the model-outputs.

 

Agricultt.
with'n r oomphI
comprehensive 11'
models and simeI
Wmdescri;

Amodelm

 

thing or a System
Of“ mm)’ or St:

Win the so

crops_ Thus. cro

24

LITERATURE REVIEW

Agricultural and environmental research utilizes knowledge of specific processes
within a complex system of interacting and interdependent phenomena to attempt a
comprehensive understanding of the operation of the system as a whole. Currently, crop
models and simulation techniques have been developed to provide comprehensive and
quantitative description of the behavior of dynamic crop growth patterns (Singh et al., 1985).

A model as defined by Hanks and Ritchie (1991) is a small representation of the real
thing or a system of postulates, data and inferences presented as a mathematical description
of an entity or state of affairs. Crop models integrate component submodels of various
processes in the soil-plant-atrnosphere system to provide predictions of growth and yield of
crops. Thus, crop models are principal tools which bring agronomic sciences into the
information age (Jones and Ritchie, 1990; Ritchie 1991, and 1994).

In the review by Whisler et al. (1986), crop simulation is defined as the process
whereby the crop model acts like a real crop by gradually growing leaves, stems, roots, etc.,
during the season and finally predicts some final state such as biomass and seed yield. Model
accuracy may be defined in three progressive stages (Jones et al., 1987). These stages are:
(I) verification process (the programming logic is compared with the programmer’s
intentions); (ii) calibration process (the adjustments made to model parameters so as to give
the most accurate comparison between simulated results and results obtained from field
measurements) and; (iii) validation process (the process by which a simulation model results

are compared to field data not used previously in the development or calibration process).

Model tes
isto obtain a me
refers to the pro
fertilizer type is t
it?“ quesrions.

Use of cm
cllllilmfi'tlt‘ttts of n
Ritchie, 1991).

generate relevant ,

 

team work While 5

 

 

mire and have I
pollcll makers (R1
(momma 199], '1

season (Thorium-1 i
fimrllnmema] lmpi

They llaVe “So be
(Adams 9101., 199,

25

Model testing involves validation and sensitivity analyses. The purpose of validation
is to obtain a measure of confidence in the model for the intended use. Sensitivity analysis
refers to the process where one or more input variable such as cultivar, planting date or
fertilizer type is changed while others are held constant. This answers the "what happens
if. . .7” questions.

Use of crop simulation models in agricultural research has increased dramatically with
enhancements of modeling techniques, and with improvements in microcomputers (Hanks and
Ritchie, 1991). Crop simulation models enable better understanding of a systems and
generate relevant and more precise data for the system(s) under study by multi disciplinary
team work while solving the problem. Crop models predict crop yields sown anywhere at
anytime, and have proved to be a useful tool for decision making by farmers, researchers and
policy makers (Ritchie, 1986; Singh, 1989) at field, farm, regional, and national levels
(Thornton, 1991; Thornton et al., 1991). They have also been developed to analyze single-
season (Thornton and Hoogenboom, 1994), and multiple-season experiments (Thomton et
al., 1995 a). Crop models can also be used in sustainability research, and assessment of
environmental impact (Singh and Thornton, 1992; Bowen et al., 1993; Jones et al., 1993).
They have also been used at national and at global level to provide quantitative estimates
(Adams et al., 1990; Rosenzweig and Parry, 1994).

The CERES (Crop-Environment Resource Synthesis)-family models have been used
more widely than other crop models. This has been mainly due to the firnctions which can
be performed by these models as outlined by Adams et al. (1990). CERES-Maize has been

used to study various crop aspects in sub-Saharan Afiica. Using the CERES-Maize crop

model, Keating 1
accurate predicti
regimes. They \
main crop in the
in the Study condu
which can develOp
composite maize

In Mala“

useful on Planting

 

 

 

litiety Selection,

26

model, Keating et a]. (1991), and McCown er a1. (1991) working in Eastern Kenya made
accurate predictions of grain yield under a wide range of water, N, and crop management
regimes. They were also able to assess the residual efl'ects of fertilizer application on the
maize crop in the previous season Keating and Wafirla (1992) used the CERES-Maize model
in the study conducted at the Katumani Research Station to identify the resource constraints
which can develop under high plant population which limit the leaf size of the open-pollinated
composite maize cultivar, Katumani composite B.

In Malawi, Singh et a1. (1993) used the CERES-Maize model to collect information
useful on planting time periods, optimum plant densities, nitrogen fertilizer management,
variety selection, and the economics of fertilizer use by carrying out simulation experiments
replicated over many different seasons. In a difi‘erent study, done on a regional basis,
CERES-Maize was linked with geographic information system (GIS) data, and an automated
land evaluation system. From this study presenting suggestions concerning different
management strategies such as time, and amount of nitrogen fertilizer applications was
possible, as well as quantifying the. weather-related risks of maize production in the central
region of Malawi. Furthermore, information obtained from this study was shared among
decision/policy makers, researchers, and extension personnel (IFDC, 1994; Thornton er al.,
1995b).

In a study at the International Institute of Tropical Agriculture (IITA), Ibadan,
Nigeria, the CERES-Maize model predicted maize yields that were within 10% of the
observed yield (J agtap et al., 1993). However, in a study conducted at a small research

station at Rubona, in Rwanda the CERES-Maize model did not accurately predict grain yield

of the local ma:
1994). The CI
increasing levels
in climate on or
developed count

Most of tl
01- (1986). and m

usmono (1i

 

model for bean (
allows researcher
conduct sensitisi
““le mi
and maXitnum te
We: are also

"We DSSAT c

The BEA

27

of the local maize cultivar because of poor data collection (Mbabaliye and Wojtkowiski,
1994). The CERES-Maize model has been used to estimate how climate changes and
increasing levels of carbon dioxide may alter yields in the world's crops. Efl‘ects of change
in climate on crop yields were found to be more adverse in developing countries than
developed countries (Muchena, 1994; Rosenzweig and Parry, 1994).

Most of the crop simulation models currently available have been listed by Whisler er
al. (1986), and Ritchie (1991 and 1994). The latest released DSSAT crop simulation model
is BEANGRO (Hoogenboom et al., 1994). BEANGRO is a versatile, user fliendly simulation
model for bean (Phaseolus vulgaris L.) crop. The model has a flexible user interface that
allows researchers to compare simulation results with data measured fi'om the field, and
conduct sensitivity analysis (Hoogenboom et al., 1994). The model responds to the
environmental variables such as soil profile characteristics, rainfall, solar radiation, minimum
and maximum temperatures. Cultivar coefficients, characteristics, and crop management
strategies are also very important. These variables constitute the minimum data required to
run the DSSAT crop models (Ritchie and Dent, 1994).

The BEAN GRO crop simulation model operates at daily time steps, and simulates
bean crop growth, and development from the date of sowing until harvest maturity
(Hoogenboom er al., 1994). The model incorporates features that facilitate its use by dry
bean researchers whose primary activity is not modeling per se (Hoogenboom et al., 1991).
BEANGRO predicts total dry matter (TDM) production (kg ha“), leaf area index (LAI),
developmental stages, seed size, and seed yield (kg ha"). The present BEAN GRO version

was developed by an interdisciplinary research team through numerous changes of grain

 

leginne models 5t
pernut(Arach15 ll
eta], 1993). Ii
analysis are giver,

Globally.
under field cond
include that at C}
10153110: and port

hleUSAnnm

 

PIOdUCtion (Cur!)
Gallegos and Whit:
BEMICRO mod.

plOdUCfiOn in no

28
legume models such as SOYGRO for soybean (Glycine mwe (L.) Men), and PNUTGRO for

peanut (Arachis Inpogea L.) (Boote et al., 1989; Hoogenboom er al., 1992; Hoogenboom
er al., 1993). Details on data collection, input and simulation output files, and sensitivity
analysis are given by Hoogenboom et al. (1994), and Tsuji et al. (1994).

Globally, there is little published work on crop simulation using bean crop models
under field conditions (Gutierrez et al., 1994). The few studies that have been conducted
include that at CIAT, Cali, Colombia using an earlier bean model version to study drought
tolerance, and potential drought adaptation mechanisms of bean (Hoogenboom er al., 1987).
In the USA the model has been used to study the impact of potential climate change on bean
production (Curry et al., 1990; Lal et al., 1993). The model has also been used by Acosta-
Gallegos and White (1995) to estimate the length of the crop growing season in Mexico. The
BEAN GRO model has also been used to evaluate the responses to cultivar differences for
production in tropical environments (White et aI. , 1995). Further, the international bean
modeling nurseries have been established at several sites in Canada, the USA, Central and
South America (Hoogenboom et al., 1994). Similar nurseries are being established in several
bean growing regions in Afiica (Smithson and Grisley, 1992).

To have the ideal model of the behavior of the soil and plant P, a firll
description of the cycling processes is required. This requires an understanding of all
important physical, chemical, and biological processes influencing the various forms of P in
each component of the cycle. The complexity of P chemistry in the soil, and plants is reflected

by its many forms that are distinguished in a comprehensive model of soil and plant P given

simpler and more

 

1
Details of

details in C haplel
leaf area (Freede
l989), leaf area 1.

than 1937, Sa a

 

 

 

influence the num
concentration opt
gmetic variation it
being more respc
Wpenor P efiicier

A phOSphc

and its operation

29

by Jones et al. (1984). Mechanisms for such a model have not been fully described. Thus,
simpler and more empirical approaches are currently being used (Jones et al., 1991).

Details of P nutrient uptake, translocation and assimilation in plants are discussed in
details in Chapters 3 and 4. Phosphorous deficiency decreases photosynthetic rate per unit
leaf area (Freeden er al., 1989; Lauer et al., 1989; Qui and Israel, 1992; Rao and Terry,
1989), leaf area to weight ratio (Israel and Rufly, 1988), and nitrogen fixation in soybeans
(Israel, 1987; Sa and Israel, 1995). Other investigations show that metabolic reactions that
influence the number of pods formed, and those affecting weight per seed have difi‘erent P
concentration optima (Grabau et al., 1986). Recent findings indicate that there is a large
genetic variation for P deficiency in bean cultivars, small seeded (Mesoamerican genotypes)
being more responsive to applied P than large-seeded (Andean genotypes) which have a
superior P eficiency under low available P (Y an et al., 1995 a, b).

A phosphorous submodel for the IBSNAT crop models has been developed at IFDC,
and its operation is described in the IBSNAT network report (IBSNAT, 1990), and IFDC
working manual (IFDC, 1992; Bowen 1994, unpublished). The submodel simulates the
processes of adsorption and desorption of P, organic P turnover (immobilization and
mineralization), and dissolution of soluble fertilizer P. The rates of each of the soil P
processes are assumed to be afl‘ected by the size of the labile P pool and prevailing soil, water,
and temperature conditions. The P uptake process is simulated as a firnction of root length
density, soil water content, plant P demand, and soil N and P concentrations. Plants absorb

P from the labile pool. The model is sensitive to methods of P fertilizer application.

This P s;
to the plant grov
readied on P, .‘
the actual dyna:
components. "I
(Hoogenboom e
comtltlmication)

A numbe
(Blasts 1993).
“Willis: in the

a the begflning

 

b“iglflrling of Sim
application), PB.
(Phosplme F061: (1
PFACTO 1P deficit
on daily basis), PC
P11001301, dill)" bar
(P adsorption isotl

Singh e. .1
with the indltidua] f

ll) mismatches b

30
This P submodel, like the present soil water and N balance model, has been coupled

to the plant growth routines of the BEANGRO crop simulation model. Ifthe P submodel is
switched on, P, N and water balance models automatically function (IFDC, 1992). Therefore,
the actual dynamic section of the BEAN GRO version used in this study consists of five
components. These are crop development, carbon balance, water balance, N balance
(Hoogenboom et al., 1994), and P balance simulations (Dr. W.T. Bowen, 1995, personal
communication). This section operates on daily time steps fiom planting to harvest maturity.

A number of models dealing primarily with P residual effects have been proposed
(Black, 1993). In the P model coupled to the BEANGRO crop model there are ten
subroutines in the current P-model, and these are: SOILPI (initializes soil inorganic P pools
at the beginning of simulation), OCNPI (initializes organic pools for C, N, and P at the .
beginning of simulation), FPLACE (fertilizer placement for N and P on days of fertilizer
application), PBAND (efl‘ective band size calculation, if fertilizer applied in band), PDISS
(phosphate rock dissolution, if PR is applied), PUPTAK (P uptake on daily growth basis),
PFACI‘O (P deficiency factors on daily grth basis), ORGANIC (organic N and P cycling,
on daily basis), PCHEM (calculates the rates of movement between the difl'erent inorganic
P pools on daily basis, then updates each pool based on the calculated rates), and PTHERM
(P adsorption isotherm). .

Singh et al. (1985). stated that the success of agrotechnology transfer rests, in essence,
with the indivldual farmer. The two basic reasons for failures in agrotechnology transfer are:

(I) mismatches between the environmental requirements of a technology and the

 

anironmental ch
technology and t

Althoug‘
Ahica (Keating e
l995h), there is n,
11115 has mainly be
researchers and p
local conditions i
‘21]mele and tall

W {01' M61

 

l0 Wthl'l ”1051 all

31

environmental characteristics of the land, and (ii) mismatches between the requirements of the
technology and the resource capabilities of the farmer.

Although there has been a number of crop model simulation application studies in
Africa (Keating et al., 1991 and 1992; Singh et al., 1993; Muchena, 1994; Thornton et al.,
1995b), there is no documented work on the use of crop modeling experiments in Tanzania.
This has mainly been due to (I) the lack of adequate training, and involvement of Tanzanian
researchers and policy makers in model applications, (ii) lack of appropriate models for the
local conditions in Tanzania, (iii) lack of data (such as cultivar and soil characteristics) for
calibrating and validating the already developed model components and (iv) lack of financial
support for model development and application in Tanzania. Furthermore, N is the nutrient
to which most attention has been devoted, and there are no reported P simulation studies
using DSSAT crop models.

Therefore, the specific objectives of this study were: (I) To conduct a two year field
experimental work on beans (Phaseolus vulgaris L. ), and collect the minimum data set
(MDS) required for evaluation of the P version of the BEAN GRO bean simulation model for
Tanzanian conditions; (ii) To evaluate the performance of a bean crop for sensitivity to
phosphorous response under rain-fed, and irrigated conditions using the P version

BEAN GRO crop simulation model for Tanzanian conditions.

This stud
37° 39E at 525
June rainy seasc
s035011 from Man
rainfall of 900 m 1.

Prior to th
for three pm.
mo“- Plots a
targeting“ More
“We! Rerx

actordietc>1mc

 

32
MATERIALS AND METHODS

Rainfed Experiment

This study was conducted at Sokoine University of Agriculture (SUA) farm (6° SO'S,
37° 3913 at 525 m altitude), in Eastern Tanzania, on sandy loam soils during the March to
June rainy seasons of 1993 and 1994. The rainfall pattern on site is bimodal with a long-
season fi'om March to June, and a short-season fiom October to December with an average
rainfall of 900 nun.

Prior to the initiation of the experiment, the site lay under natural vegetation (weeds)
for three years. Tillage was done with a disk plow two weeks before planting in the first
season. Plots were prepared using hand hoes in both seasons. The amount of natural
vegetation incorporated during the land preparation was estimated as described in Form H
of Technical Report 2 (IBSNAT, 1990). The slope at the site was estimated to be O-2%,
according to Ritchie et a]. (1990). Soils for analysis were collected from a composite of 12
cores to a depth of 12 cm fiom each treatment at pro-planting, and post-harvest periods,
mixed and bulked for analysis. Pre-plant soil physical, and chemical characteristics for each
experiment are shown in Table 2.1. Tables 4.2a and 4.2b show soil chemical characteristics
afler fertilizer application in the 1993 and the 1994 seasons.

Since the model requires a data set on climate, the daily weather data for the two
seasons were collected as shown in Figs. 2.1 to 2.3, and Appendix 2a. Longitude, latitude,
daily solar radiation (MJ m"), minimum, and maximum temperatures (° C) were manually

recorded fi'om the meteorological station at SUA.

lllemeteorolog:

I
installed and ra

, l
GOlltlIlCUVllV, ma;~

Soil inpu
mt haion day

photosynthetic fa

er a1. (1990). c.
(1990). Using the
saturated upper 1;
as described by

 

 

the literature V'alu
00mFilter Prograr
Menage mu
each soil layer (T
method described

Allll‘erldix 1 a,

33

The meteorological station was 400 m fiom the experimental site, therefore, rain-gauges were
installed, and rainfall records (mm) were collected daily on-site.

Soil inputs such as albedo, fraction (SALB), evaporation limit, cm (SLUl), drainage
rate, fl'action day" (SLDR), runofl‘ curve (SLRO), mineralization factor, 0-1 factor, (SLNF),
photosynthetic factor, 0-1 (SLPF), root growth factor, 0-1 (SRGF), and saturated hydraulic
conductivity, macropore, cm h" (S SKS) were estimated using methods described by Ritchie
et a1. (1990). Crop residue dry weight (kg ha") was determined as described by IBSNAT
(1990). Using the growing maize (Zea men’s L.) crop, drained upper limit, cm3 cm‘3 (SDUL),
saturated upper limit, cm3 cm’3 (S SAT) and lower limit, cm’ cm’3 (SLLL) were determined
as described by IBSNAT (1990). However, the values of field-measured soil water
availability determined for SDUL, SSAT and SLLL were found to deviate significantly from
the literature values for soils with similar characteristics (Ratlifi‘ et al., 1983). Therefore, a
computer program SWLIM (J .T. Ritchie, 1995, personal communication) was used to
calallate these values using the sand, clay percentage, and bulk density values determined for
each soil layer (Table 2.2a). Bulk density (SLDM) values were predicted based on the

method described by Rawls (1983). The soil profile descriptions for the site are shown in

Appendix la.

Table2.l.Ger1erl
experimental sin

Soil characterist:

Moisture (%)'
pHHZO
Organic Carbont’
Imogen (%)
Bray-I Plppm)
Bulk density (cm‘
Sand(l’t)

Silt (%)

Ch1' (%)
Textual Class
Classification

 

Excmgeable can

Catmgq 1008.1 SC
Mg 9

K it
A] .
ECEC“
A] Winn (0'0)

34

Table 2.1. General pro-plant soil characteristics of 0-12 cm soil layer at blocks A and B at the
experimental sites.

 

 

Soil characteristic Experimental Site
Rainfed Irrigated
Moisture (%)' 13.09 15.23
szHzo 4.60 5.50
Organic Carbon(%) 0.83 2.27
Nitrogen (%) 0.04 0.09
Bray-I P(ppm) 1.18 2.83
Bulk density (cm’lcm’) 1.17 1.39
Sand(%) 60 55
Silt (%) 20 24
Clay (%) 20 21
Textural Class Sandy loam Sandy clay loam
Classification Isotherrnic Isotherrnic
Aridic tropustic Typic tropustic

Exchangeable cations:
Ca (meq 100g'l soil) 2.20 2.70
Mg “ 1.31 0.80
K “ 1.33 1.45
Al " 1.03 1.05
ECEC “ - 5.87 6.00
Al saturation (%) 17.54 17.50
Drainage class Moderate Poor
Permeability Moderate Poor
Color Reddish Dark grayish

brown

' Volumetric moisture content

'1

 

rainfed experimt

Table 2.2a. $01

 

 

'\
. on a n . . . . . s _ . . . . . on. C. :C
r . g.....: .I . ..3. . c is. . . 4 ~ 4 m
r , a o e...» as... aw... .s.s.._r.~a..wa
. . . . . . . . . . a a:
A... »c .1» A. we .1» a. .U .l. rs rm
8
r“. k. E .5 . a T... _... a a .r. .u. ... “I. A . . c .J r e r . o a .3 .4: as; is e
a an. 8 .E .5 .5 may i. C. :r .3 .1. :3 C. r: r: K. .5 ‘1 5‘ r
a c . . . . . . . . . . . . 0
Mn” n5 3 :- 3 A L . t .1; -l 1. o .A s 4 h f
y.
rm
”1% Ar.» P15 F1V FU A .u a .- fin D v .rl.‘ pr.V a. num- F V A .u A v a 1; F .u A. I A v flu I v
a a «It .5 l. :1. ‘u 9. Ala .’ «H» and cw - . 7k ‘1. 4 {a In.» ‘1 Pl» rd. 9.1
o . \V. P: a; .1. U
vy-
. a. .m“

35

Table 2.2a. Soil profile characteristics for the isohyperthemic, aridic tropustic Oxisol at the
rainfed experiment.

 

 

*SUMO930002 MOROGORO, TZ SALO 100 OXISOL ISOHYPERTHEMIC, ARIDIC TROPUSTIC
@SITE COUNTRY LAT LOG SCS FAMILY
MOROGORO TANZANIA ~6.5 37.3 REDDISH BROWN SAND LOAN
e SCOM SALE SLUI SLDR SLRO SLNF SLPF SMHB SMPX SMKE SRGP
2.5YR 0.12 6.0 0.50 84.0 1.00 1.00 IBOOl 18001 I8001 I8003
@ SL8 SLMH SLLL SDUL SLDA SLRF SLKS SLDM SLOC SLCF SLSI SLCY SLNI SLHW
10 Ap 0.091 0.227 0.382 1.000 16.56 1.55 1.13 60.0 20.0 20.0 0.06 4.56
20 Ap 0.066 0.202 0.382 0.841 16.56 1.55 1.13 65.0 30.0 15.0 0.06 4.50
30 8 0.104 0.240 0.399 0.499 16.16 1.50 0.76 65.0 15.0 30.0 0.05 5.38
40 AB 0.125 0.261 0.365 0.338 16.56 1.60 0.76 53.0 13.0 24.0 0.06 5.50
50 81 0.088 0.224 0.382 0.166 6.98 1.55 0.70 65.0 13.0 22.0 0.05 5.69
60 82 0.165 0.301 0.347 0.134 5.98 1.65 0.70 55.0 10.0 35.0 0.04 4.59
70 82 0.053 0.189 0.417 0.019 5.98 1.45 0.40 68.0 15.0 17.0 0.04 4.23
80 82 0.159 0.295 0.417 0.016 16.55 1.45 0.33 40.0 24.0 36.0 0.05 5.14
90 83 0.154 0.290 0.434 0.012 5.99 1.40 0.33 35.0 30.0 35.0 0.05 4.46
100 83 0.089 0.225 0.472 0.003 5.99 1.29 0.30 38.0 22.0 20.0 0.03 4.61
! PHOSPHORUS DATA
8 SL8 SLPX SLPT SLPO SLCA SLAL SLFE SLMN SLBS SLPA SLPB SLKE SLMG SLNA
10 3.50 270 120.0 1.77 1.5 3.20 ~99 ~99 ~99 ~99 1.6 5.60 0.19
20 3.30 270 120.0 1.59 1.5 3.20 ~99 ~99 ~99 ~99 1.6 5.70 0.19
30 3.33 242 120.0 1.45 1.4 3.00 ~99 ~99 ~99 ~99 1.4 2.80 0.22
40 2.51 170 80.0 ,1.31 1.3 2.87 ~99 ~99 ~99 ~99 1.4 2.60 0.22
50 2.51 170 80.0 1.24 1.3 2.86 ~99 ~99 ~99 ~99 1.4 2.80 0.22
60 -2.51 170 80.0 1.07 1.3 2.20 ~99 ~99 ~99 ~99 0.4 2.30 0.22
70 1.59 170 80.0 0.61 1.3 2.20 ~99 ~99 ~99 ~99 0.4 2.30 0.16
80 1.53 75 50.0 0.33 1.3 1.29 ~99 ~99 ~99 ~99 0.4 1.30 0.16
90 1.42 75 50.0 0.34 1.3 1.87 ~99 ~99 ~99 ~99 0.4 1.20 0.16
100 1.41 75 50.0 0.34 1.2 1.10 ~99 ~99 ~99 ~99 0.4 1.20 0.16
SLB-Soil layer depth (cm), SLMHsMaster horizon, SLLL=Lower limit water content(cm’ cm’)

SDUL-Drained upper limit water content (cm’tnw’), SLDA- Saturated water content (cm? cm’),
SLRF-Root growth factor (0-1.0), SLKS-Sat. hydraulic conductivity, macropore cm h”, SLDM-Bulk
density, moist (g cm”), SLOC-Organic carbon (%), SLPC- Sand (8), SLSI-Silt (%) SLCY-Clay(%)
SLNI- Total nitrogen (%), SLHW-pH (water), SLPX= P extractable (mg kg“) SLPT-Total P (mg kg“)
SLPO- Organic P (mg kg“), SLCAP Ca 9 kg“). Other variables not measured are defined by Tsuji
et a1. (1994).

In both seasons pre- and post-harvest soil chemical characteristics were taken seven
days before planting. However, during the 1994 growing season, pre-planting soil
characteristics were taken 365 days from the time the P treatments were applied. Post-
harvest soil characteristics were taken seven days afler harvest. The methods used for soil

analyses were: mechanical analysis - hydrometer (Day, 1965); and pH measurements using

1:2.5, soiliwatr
available P by 1
organic carbon b
extracted in IM
A] by 1N KCl w;

Plots we
between rows an
about 2.5 cm.
Elemental pic

deemed minim

 

 

15mlong m‘th r0

WW across the s

inwdl Will bloc

36
1:2.5, soil:water suspension by use of a pH electrode; N - micro Kjeldahl (Juo, 1979);
available P by Bray I i.e., 0.025N HC1+0.03N NH,F extractant (Bray and Kurtz, 1945),
organic carbon by Walkley and Black (1965). Exchangeable cations (Ca, K, Mg, etc.) were
extracted in IN NILOAc and measured by atomic absorption spectrophotometer. Extractable
Al by UV KCl was determined colorimetrically.

Plots were sown with a dibble, placing one seed per hole at a spacing of 40 cm
between rows and 10 cm within rows on April 4, 1993, and April 14, 1994 at the depth of
about 2.5 cm. The spacing gave a plant population of 250,000 ha'l or 25 plants in .
Experimental plots were prepared as recommended for the physical dimensions of a
designated minimum data set (MDS) for a bean crop (IBSNAT, 1990). Thus, each plot was
15 m long with four rows resulting in a plot size of 24.0 m2 with rows running from East to
West across the slope. During the 1993 season the experiment had two blocks (A and B),
in which each block had three replications, and each replication consisted of seven treatments
(plots). Only block B was used to collect data for evaluation of the BEANGRO model in
both seasons. Block B from each experiment was also used to assess the influence of annual
applied P on bean growth, and yield (refer to Chapter 4 for details).

The experimental design and data collection procedures were as described in the
Technical Report 1 (IBSNAT, 1988), and Technical Report 2 (IBSNAT, 1990). The
experiments were sown in a complete randomized block design (CRBD). The seven
treatments used in the study were as follows: 0 kg P ha“ - control; 22 kg P ha“-low rate-TSP;
22 kg P ha"-low rate-MPR; 44 kg P ha"-medium rate-TSP; 88 kg P ha"-high rate-TSP; 88

kg P ha"-medium rate-MPR; and 176 kg P ha“-high rate-MPR The MPR ore that could pass

through a 75 -

 

obtained from tl
TSP (46% P20I
rates prescribed
groan bean cul
with determinar
al., 1992; Sextc

To prex‘e
SCRSitive, 80 kg

dressed at plan

 

 

 

aSTOllomic PM
included the cont
1mg the insectic
liters of War er n

(Sc

37

through a 75 - micron screen was used. The commercially distributed PR fertilizer was
obtained fi’om the Minjingu Phosphate Company (MIPCO) based in the Arusha Region. Both
TSP (46% P20,), and MPR (30% P2 0,) were broadcast, and incorporated in each plot at
rates prescribed by treatments at planting. The commercially available, and most commonly
grown bean cultivar ”Canadian Wonder” was used in this study. This is an Andean cultivar
with determinate bush growth habit (I), and intermediate photoperiodic response (White et
al., 1992; Sexton et al., 1994).

To prevent yield reductions from inadequate levels of nitrogen to which the model is
sensitive, 80 kg N ha'1 in the form of urea was split into two equal applications. It was side
dressed at planting, and again at the flowering (R1) stage in all plots. Other standard
agronomic practices were used to obtain optimum growth under protected conditions. These
included the control of the major insects, mainly foliar beetle (Oorheca benningseni (W eise)
using the insecticide ”Karate" at 33.0an 20L'1 H20. Dithane-M45 was applied at 250g 100"
liters of water mainly to control bean rust (Uromyces appendiculatus (Pers), white mold
(Sclerotr’m'a sclerotiorum (Lib.) de Bary), and angular leaf spot (Phaeoisariopsis griseola
(Sacc)). Spraying was done at seven-day intervals, starting at 95 % crop emergence (V 2)
until at R,. Weeds were controlled by hoeing between rows and hand pulling within rows.

Leafarea index (LAI) values were determined at each harvest time by photocopying
all leaves from the harvested plants, cutting out the leaf images, and using a portable leaf area
meter model LI-COR 3000 to determine the leaf areas in cm2 from a subtended land area of
10,000 cm’. At crop establishment growth stage (V 2), five plants were tagged per plot, and

leaf number on main stem were counted at every sampling time. At R, growth stage, leaf

promotion on tl
number plant" 1

There a
were done at V
bass (pa plot) I
andNuland and .
“Win 3. r
Stage most plants
Ar each harvest l

 

 

 

The reco
loom] REPOrr
in 1994. “ grar
ramming all def

38

production on the main stem reached a maximum number. After this stage, the average leaf
number plant" plot“1 was determined.

There were four biomass harvests before the final harvest at (R9) maturity. These
were done at V,, R,, R,, and R, growth stages. Vrsual assessments were made on a daily
basis (pa plot) to record days to difi‘erent growth stages as described by Fehr er a1. (1971),
and Nuland and Schwartz (1989). The key for bean growth stages used in this study is shown
inAppendixB. FmalharvestwasdoneatR,(harvestrrraturity)growth stage. Atthis growth
stage most plants had more than 80% of pods yellowing, and less than 15% of green leaves.
At each harvest, plants were hand pulled fiom a 1.25 m long in the two central rows of each
plot. This gave 1.0 m2 of harvested area.

The recommended biomass, and final harvest procedure followed the description in
Technical Report 2 (IBSNAT, 1990). The crop was harvested 65 DAP in 1993 , and 72 DAP
in 1994. All grain yields were reported as marketable seed, i.e., seeds not damaged. After
removing all defective seeds, the seed weight was adjusted to zero moisture content for

comparison with simulated values.
Irrigated Experiment

Details on location, altitude, associated temperature regimes, solar radiation, rainfall
(mm), experimental design, spacing, plot size, treatments, crop management, sampling
procedures, and LAI determinations are as reported under the rainfed experiment. However,

this experiment was conducted in the area near the SUA crop museum, about 800 m fiom the

rainfed aperime
of May through ,
oomained a sand
soil inputs used
saturated upper
characteristics it

description at th

 

39
rainfed experimental site (for the accessrbility of irrigation water), during the growing seasons
of May through August 1993, and 1994. Rows ran north to south across the slope. This site
contained a sand clay loam type soil, and had lain fallow during the previous two years. The
soil inputs used for the crop model i.e., drained upper limit (SDUL), lower limit (SLL),
saturated upper limit (SSAT), pH, etc., are shown in Table 2.2 (b). Physical and chemical
characteristics were determined as explained in the previous experiment. Soil profile

description at the site is as shown in Appendix 1b.

 

Table 2.2b. Soi
irrigated site.

 

'3M001’8‘" 94::

v. rd. .a‘vUVA
fiSITE CZ'
HGFOSRCRC
ISOHYFERTHEVIC T

I
racer sue s:
2.32:: c.13 .
i 5::- sun-z 5'
1: A;
2: A;
3.: A:
re 5;

:Fl a.
UV

., fl
5.

i (A) ('3 I) (j
o a) (r 1') . _
" “’1' . . . . .

 

‘ A
". as -
‘ 3‘
.‘ .
at. ,
9‘ ”-

- 5..

‘P! a-

a}. 3;

I

P-ASFHREHS ,

’7" an ‘
3 ts: 3-2! 3.,

'r-

‘5 4.33 2,

A!

>4} ‘F H.

;- ‘.“ :7‘

..

r , _

3. 3.5 .

A n. -

1° 2.9. :

5.0 e -

”A" ‘-C.. 1

n. .

C“ a.“. 3

‘A -

.'I 5 _

v "35 1

fit‘ . ,

7‘ *-°‘ 1

.

Pd: Q “ ‘

. " '8 I

‘15... ’ as

up. 1.3‘ . ‘.

‘v
r...
\ ‘xhp ‘
~35 V
F Dhaa .az‘rer A
:E“-~_

.. ..

8. «lines L4”;
7:;- . F} .
'n..‘~ ‘

1:" graftf r
3....5“r, 1
~ 5" misc ,_‘

' h
:r‘Igl II . ‘. .5

\a‘ a.."‘
f. r“‘ v."-
Nib M '

“337‘“: r _
5 ‘ ‘
e. a.

40

Table 2.2b. Soil profile characteristics for the isohyperthemic, typic tropustic Alfisol at the
irrigated site.

 

*SUMO930001 Morogoro,Tz SACLLO 100 ALFSOL, ISOHYPERTHEMIC, TYPIC TROPUSTIC

CSITE COUNTRY LAT LONG SCS FAMILY
MOROGRORO TANZANIA ~6 . 5 37 . 3 DARK GRAYISH BROWN SAND CLAY LOAN
ISOHYPERTHEMIC,TYPIC TROPUSTIC
8 SCOM SAL8 SLU1 SLDR SLRO SLNF SLPF SMHB SMPX SHKE SGRP
2.5YR 0.13 6.0 0.35 84.0 1.00 1.00 I8001 18001 18001 18003
8 SL8 SLMH SLLL SDUL SLSA SLR? SLKS SLDM SLOC SLCF SLSI SLCY SLNI SLHW SLHB
10 Ap 0.081 0.217 0.434 1.000 5.01 1.40 2.27 55.0 24.0 21.0 0.09 5.43 5.30
20 Ap 0.097 0.233 0.382 0.819 6.98 1.55 2.27 57.0 23.0 20.0 0.09 5.09 5.20
30 A8 0.102 0.238 0.382 0.268 7.08 1.55 1.15 59.0 18.0 23.0 0.08 4.25 5.20
40 81 0.101 0.237 0.382 0.268 5.01 1.55 1.64 58.0 24.0 22.0 0.07 5.03 5.00
50 81 0.082 0.218 0.399 0.138 5.01 1.50 1.80 58.0 24.0 18.0 0.07 5.03 5.00
60 82 0.129 0.265 0.295 0.128 5.01 1.50 1.80 58.0 23.0 19.0 0.07 5.04 5.00
70 82 0.096 0.232 0.382 0.128 5.02 1.55 1.60 59.0 20.0 21.0 0.09 5.13 5.05
80 82 0.056 0.192 0.451 0.128 6.99 1.35 1.60 48.0 42.0 10.0 0.06 4.83 5.00
90 83 0.056 0.192 0.451 0.100 6.25 1.35 0.55 48.0 42.0 10.0 0.05 4.80 5.10
100 83 0.056 0.192 0.451 0.100 6.84 1.35 0.55 48.0 42.0 10.0 0.05 4.61 5.10
! PHOSPHORUS DATA
8 SL8 SLPX SLPT SLPO SLCA SLAL SLFEL SLMN SLBS SLPA SLPB SLKE SLMG SLNA SLSU
10 4.33 360 280 2.03 1.10 1.20 ~99 ~99 ~99 ~99 0.60 2.80 0.71 ~99
20 4.10 360 280 2.01 1.10 1.20 ~99 ~99 ~99 ~99 0.60 2.50 0.71 ~99
30 3.57 309 30 1.94 0.80 1.60 ~99 ~99 ~99 ~99 0.30 1.50 0.61 ~99
40 2.91 309 30 1.93 0.80 1.10 ~99 ~99 ~99 ~99 0.40 1.30 0.61 ~99
50 2.02 309 30 1.93 0.30 1.10 ~99 ~99 ~99 ~99 0.30 1.10 0.61 ~99
60 1.47 309 30 1.64 0.30 1.20 ~99 ~99 ~99 ~99 0.30 1.10 0.61 ~99
70 2.05 309 30 1.04 0.30 1.20 ~99 ~99 ~99 ~99 0.30 1.10 0.61 ~99
80 1.64 143 40 0.94 0.30 1.10 ~99 ~99 ~99 ~99 0.30 1.10 0.65 ~99
90 1.71 143 40 0.84 0.30 1.10 ~99 ~99 ~99 ~99 0.30 1.00 0.65 ~99
100 1.54 133 40 0.52 0.30 1.00 ~99 ~99 ~99 ~99 0.30 1.00 0.65 ~99
SLB-Soil layer depth (cm), SLMH-Master horizon, SLLL=Lower limit water content(cm’ cm°l

SDULaDrained upper limit water content (cmfi cm”), SLDA= Saturated water content (cm? cm°),
SLRFhRoot growth factor (0-1.0), SLKS=Sat. hydraulic conductivity, macropore cm h“, SLDM-Bulk
density, moist (g cm”), SLOC-Organic carbon (%), SLFC= Sand (%), SLSI=Silt (%) SLCY-Clay(%)
SLNI- Total nitrogen (%), SLHW=pH (water), SLPX= P extractable (mg kg”) SLPT-Total P (mg kg“)
SLPO- Organic P (mg kg“), SLCA: Ca 9 kg“). Other variables not measured are defined by Tsuji
et a1. (1994).

Hoes were used to prepare the experimental plots in both seasons. Sowing was done
on May 28 for the 1993 growing season, and on May 8 for the 1994 growing season.
Planting was done at the end of the 1993 long-rain season. To ensure rapid and complete
development of the root system, pre-plant irrigation was used as shown in Appendix 4b for

the 1993 season experiment. Irrigation water was provided using portable sprinklers 1.5 m

 

above the soil 5
located at the er;
supply was at r
crop was used asl
oolor of the uprig.
by Thuns (1991 r
Water Vt

Fm (date), amo
applied in both
with river to the
MW 31 the site 1
imgallon spam \
the arm- block a;

Sure m“ when -

adequitely imam

41

above the soil surface. The amount of water applied was measured using the flow meter
located at the experimental site. Irrigation water was applied late in the evenings when water
supply was at maximum, while evapotranspiration, and windshifi were minimal. The bean
crop was used as an indicator of the time to irrigate. Irrigation water was applied when the
color of the upright leaves changed from light green to a pronounced dark green as described
by Thung (1991).

Water was applied within two to three days following this color change. Day of the
year (date), amount of water applied (mm), and crop growth stage at which irrigation was
applied in both seasons are shown in Appendices 4 a and 4 b. Water was pumped from the
nearby river to the storage reservoir. From there, water was conveyed to the flow meter by
gravity at the site through the pipe, and then to the respective sprinklers for application. The
irrigation system was installed with sprinkler heads placed in the rectangular grids to irrigate
theentireblockatthe sametime. Thiswasmoved accordinglywithin the block so as to make
sure that when water was applied, the area receiving the least amount of water was
adequately irrigated to provide good crop growth.

Irrigation efiiciency was calculated as defined by the On-Farm Irrigation Committee
of the American Society of Agricultural Engineers (ASAE, 1978), i.e., the ratio of the volume
of water that is used to the volume of irrigation water applied, expressed as:

e, = Vb/Vf
where:
e, = the irrigation efficiency

V, = the volume of water used

 

Irrigati
measuring the a
at the end of t
regarded as the
meter during t‘rlI
ii). To increase
profile being filie

crop was harves

 

prior to
had grown lots

Midde glypho

the land Preparer

42

V, = the volume of water delivered to the field.

Irrigation efliciency for the sprinkler irrigation system used was determined by
measuring the application depths with catch cans (buckets). Irrigation water w... applied, and
at the end of the determined period the total amount of water in the catch buckets was
regarded as the volume of water used (V,), and the amount of water registered by the flow-
meter during the irrigation session was taken as the volume of water delivered to the field
(V). To increase the application efficiency, the irrigations were small so as to avoid the soil
profile being filled. This prevented deep percolation, and surface runofi‘ was minimized. The
crop was harvested 75 DAP in 1993, and 72 DAP in the 1994 season.

Prior to land preparation for the second crop of the 1994 season, experimental plots
had grown lots of weeds in both experiments. These weeds were killed by the use of
herbicide glyphosate (at 2.2 kg ai ha") four weeks before the plots were prepared. During

the land preparation the decomposing organic matter was incorporated in the soil.
Data Input

Gareral input and output file structures used for crop simulation in this study are as
outlined by Tsuji er a]. (1994). The soil profile characteristics at the two sites were entered
in FILES as shown in Table 2.2a for the rainfed and Table 2.2b for the irrigated site. Table
2.3 shows one of the main files, referred to as FILEX which documents the inputs of the
model for a particular season and location (in this case for irrigated experiment during the

1993 season). The daily solar radiation, maximum and minimum air temperature, and rainfall

m uttered in

 

are illustrated t
Appendices 2a

used for the eva

 

43

were entered in file FILEW (Table 2.4). For this example only the months of May and June
are illustrated for the 1993 season. Details on daily weather characteristics are shown in

Appendices 2a and 2b. Canadian Wonder bean cultivar genetic coefficients estimated and

used for the evaluation of the bean model were entered in FILEC (Table 2.5).

Table 2.3. Era"

.-
”’U “F... I .
fiJIH$.F_.J-
., ..-..vf
' T:; 51.:r. ,
.._. . =h
' fiche. -. .

at-

rxv "_:
J. J... u. . .

 

 

I O
2 ' ' i 9: Va,
' Q O h. ‘1’.
A. A” P r
t ‘ ‘ ‘ h. in,
' . C ‘1‘ Ali.
n. ,

“ ' ' v ‘fih'~'
' A o a a - '9‘

 

.C" 'r~~v,- “t"
9' v5 ‘5‘:th

. 5h IEIZL4 Ca'

 

 

 

 

. ‘~
the: ‘3’
i’ .5
~ ~-
: ‘:- 5:3?
‘ 5 v- _
I ‘7 ant,"
Q 4" ‘ .”
"~.._,<
' <7 --__
. '~ .n~;‘
I ‘ ‘h.

 

- p
m (I!
9..
4
'1"
r

(.t
a
Or
A!

T «—
' . ‘ We .4
. §.~,. -
".2: m.‘
-. «big.
‘ o "a
I 1:. ‘ V“,-

0 ‘5 fi
4 37"., '5‘
“‘50 IF»-

' a c ._
'2
‘ 3i‘p. 5
. "~a A —.
"-. ‘yd‘
‘ 3' :-.
. "ae .
‘u ‘t
. ’..:‘ . 3_‘
r ~‘ul F.,
a
. §:-.‘ ~-{
‘ and" I:,’
l
‘ :‘4'., 'u‘
‘~.:~ .__

‘ \1 *F -
~ Sin. «-4
H!!! v.,

V v f
n. 4. _| ‘
‘ :v".. t."
'(,‘ v...

1 a. ‘3 r
‘ :(‘EAE ‘1“
v "Jc Y .'

‘ t
. 3:-. . .g4
"‘44 *p».
* .
J

44

Table 2.3. Example of input file (SUMO9301.BNX) used to run bean model BEANGRO.

*EXP.DETAILS: PHOSPH-BEAN 1993 IRRIGATED EXPERIMENT AT MOROGORO,
THE EXPERIMENTAL DATA FOR THE 1993 SEASON WERE PROVIDED BY
RWEYEMAMU OF SOKOINE UNIVERSITY OF AGRICULTURE,

THROUGH DR. J.T. RITCHIE,

*TREATMENTS
8N R O C NAME ....................
01 1 1 1 0 KgP/halCONTROL193/IR
02 1 1 1 22 KgP/ha TSP 93/IR
03 1 1 1 22 KgP/ha MPR 93/IR
04 1 1 1 44 KgP/ha TSP 93/IR
05 1 1 1 88 KgP/ha TSP 93/IR
06 1 1 1 88 KgP/ha MPR 93/IR
07 1 1 1 176KgP/ha MPR 93/IR
*CULTIVARS
8C CR INGENO CNAME

1 EN 180014 Canadian Wonder+
*FIELDS
8L ID_FIELD WSTA.... FLSA FLOB
l SUMOOOOl SUMO9301 ~99 O
*INITIAL CONDITIONS
@C PCR ICDAT ICRT ICND ICRN
1 EA 93073 1 -99 1.00
@C ICBL SHZO SNH4 SNOB SAEX
1 10 0.222 2.5 3.5 4.33
1 20 0.282 2.5 2.8 4.10
1 30 0.256 2.0 2.8 3.57
1 40 0.216 2.0 2.8 2.91
1 50 0.197 2.0 2.8 2.02
1 60 0.251 1.5 2.8 1.47
1 70 0.258 1.5 2.8 2.05
1 80 0.229 1.5 2.8 1.64
1 90 0.237 1.5 2.8 1.71
1 100 0.222 1.5 2.8 1.54
*PLANTING DETAILS
@P PDATE EDATE PPOP PPOE PLME
1 93148 93158 25.0 25.0 S
*IRRIGATION AND WATER MANAGEMENT
@I EFIR IDEP ITHR IEPT 10??
1 0.63 -99 ~99 ~99 R7
CI IDATE IROP IRVAL

1 93146 IR004 31.5

1 93147 IR004 27.8

1 93152 IR004 33.8

1 93154 IR004 50.0

1 93161 IR004 30.0

1 93162 IR004 27.5

1 93168 IR004 21.1

1 93174 IR004 21.1

1 93180 IR004 28.2

1 93181 IR004 31.5

1 93182 IR004 20.6

1 93187 IR004 21.1

1 93191 IR004 21.1

1 93195 IR004 24.5

1 93198 IR004 24.5

1 93204 IR004 20.8

1 93208 IR004 21.7

1 93214 IR004 20.0

CORNEL L.

TANZANIA

TANZANIA

BRIAN D. BAER (MSU) and DR. W.T. BOWEN (IFDC), USA.
~~~~~~~~ FACTOR LEVELS~-~~~~-~~~
CU FL SA IC MP MI MP MR MC MT ME MH SM
1 1 1 1 1 1 1 1 0 0 0 0 1
1 1 1 1 1 1 2 1 0 0 0 0 1
1 1 1 1 1 1 3 1 0 0 0 0 1
1 1 1 1 1 1 4 1 0 0 0 0 1
l 1 1 1 1 1 5 1 0 0 0 0 1
1 1 l l 1 1 6 1 0 0 0 0 1
1 1 1 l 1 1 7 1 0 0 0 0 l
FLDT FLDD PLDS FLST SLTX SLDP ID_SOIL
SUOOO 00000 00000 00000 SACLLO 100 SUMO930001
ICRE
1.00
PLDS PLRS PLRD PLDP PLWT PAGE PENV PLPH
R 40 0 1.0 55 ~99 ~99 1.0
IAME IAMT
~99 ~99

Table 2.3. (cc:

. PD:
Aflr~t.-.0V“~

M‘-v- NR"
if rm“: :. -.

v-vi on

'0
M

PAL
Ivv

(6:1

93'.

a
v

rs v
. .
D~
on

pit.

n.9—
«U.

|'\.,
lungfsz S

F...
‘ ‘J 6 W

8:»...
UV‘;

 

{v r» A. K5 Ru :3 {v
{c {u i. {O ‘45 .4. «4.

. - .

1 IA 1 It 1‘ 4‘ 1
AU #8 C AIV any
. . . t o
y a» p. .4
a.» a, FtU
0‘

u‘QO-AV‘

Table 2.3. (cont’d).

*FERTILIZERS (INORGANIC)

88

1
1

NNN

UUU’

010101

OtO‘tOt

7
7
.7

FDATE

93148
93184

93148
93148
93184

93148
93148
93184

93148
93148
93184

93148
93148
93184

93148
93148
93184

93148
93148
93184

FMCD

88005
88005

88005
88014
88005

88005
88021
88005

88005
88014
88005

88005
88014
88005

88021
88005
88005

88021
88005
88005

EACD

A8004
A8004

A8004
A8002
A8004

A8004
A8002
A8004

A8004
A8002
A8004

A8004
A8002
A8002

A8002
A8004
A8004

A8002
A8004
A8004

FDEP

4.
4.

bbb Alb-b hath fiéb
GOO 000 GOO 000 CO

AAA
OOO

bthtb
COO

FAMN

40
40

40

40

40

40

40

40

40

40

40
40

40
40

FAMP

22

22

44

88

88

176

*RESIDUES AND OTHER ORGANIC MATERIALS
OR RDATE RCOD RAMT RESN RESP RESK

1 93048 18001 2000 ~99
*SOIL ANALYSIS
@A SADAT SMOC SMNI SMHW
1 93118 ~99 I8001 SA009
8A SABL SADM SAOC SANI
1 10.0 1.40 2.27 2.5
1 20.0 1.55 2.27 2.5
1 30.0 1.55 1.15 2.0
1 40.0 1.55 1.64 2.0
1 50.0 1.50 1.80 2.0
1 60.0 1.50 1.80 1.5
1 70.0 1.55 1.60 1.5
1 80.0 1.35 1.60 1.5
1 90.0 1.35 0.55 1.5
1 100.0 1.35 0.55 1.5
*HARVEST DETAILS
8L HDATE HSTG HCOM HSIZ
1 93226 R9 HL LRG
*SIMULATION CONTROLS
8N GENERAL NYERS NREPS
1 GE 1 1
EN OPTIONS WATER NITRO
1 08 Y Y
@N METHODS WTHER INCON
1 ME M M
8N MANAGEMENT PLANT IRRIG
1 MA R R
8N OUTPUTS FNAME OVVEW
1 CU N Y

~99

SMHB
~99

SAHW
5.50
5.50
5.30
5.30
5.20
.40
.70
.50
.50
.50

U‘U‘U'U‘U‘

HPC

100

START

SYMBI

LIGHT

FERTI

SUMRY
Y

RINP RDEP
~99 100.0 25.0
SMPX SMKE
18002 SAOOS
SAHB SAEX SAKE
~99 4.33 0.60
~99 4.10 0.60
~99 3.57 0.29
~99 2.91 0.31
~99 2.02 0.27
~99 1.47 0.31
~99 2.05 0.29
~99 1.64 0.28
~99 1.71 0.26
~99 1.54 0.31
YRDAY
93121
PHOSP POTAS DISES
Y N N
EVAPO INFIL PHOTO
R S C
RESID HARVS
R M

FROPT GROTH CARBN WATER NITRO MINER DISES LONG

1

45

RSEED SNAME .....................................
2150 PHOSPH~BEAN 1993 IRRIGATED EXPERIMENT

Y

Y

Y

Y

Y

N

Y

 

Table 2.4. Exa

. . . . . . . . . . . . . . .
.P. .r. .n. C. s i. .n. .4. ~. .8 C. :c 5.. .5.
3‘ ..( 7‘ .4 «4 «xx 7.. .)s ..( .1.

 

 

 

 

 

. . . . . . . . . . . . . . . . . . i . . . .
. . . a. . . .1. L. .n. ... .4. E. ~ .u. .h. A s Ah. ~ .b. ~ ‘
. . _ . .1. .n. — .I. .n. .1. .u. L. . . .b. . .h. . . p , I . . a . , .I o
A.L art ”A A}; ”A A A 1L A 6 EA .5 . A . s A s .(v a A ‘4.» a k .4~ fit .1~ a A AA A A - A .4 A A A A at n‘ A‘ at A1; A‘ at n s A‘ n/t A‘ «IA .4
i Q 4 ‘. .hv a, II. 2 A . «L ...L ‘ rtv ‘ a any {a 6 Ala are a a 6 av .IL fils 9 9 Q. E ‘. AIV - a 7 a... 3 9 9 9, w/
.5 r5 8. .n. 8. 1‘ 5. ~ 3.3 ‘5 A. j... y... .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . .3 .C 1‘ r! r; E E r. 4. a . .1. ale .3 p: t. 4! . n 7 .3 3s «6 n at
9 6 pt: :8 - 0.. {v .4. n4. {C 8. ~ Q .1. s I?» . u .8 N a: .u- .r- 6 u a u a a
uauNp. “an!” sac. use.tarasaaahoatlta-cto.A\aid\aoaoaIa11l 91a alluded ea «In 1A to VA aazhl at. 1:4 t4 score
. 8‘ .14 t a .. r. :4. v- u ‘ .n. (he at 5F. F’ n a v a 7; .(J 4|. 1.- .h. 7. nut. 9 Ar... o a 2 .41. 6. {a 6 7: «is :14 hi
.4“ ”1“ «an .45. nu" .4. "a. .(— .6” 5 1.... .IV ‘9 l‘ :4 .K' .‘v pie pk. .na {NJ pha r0. IF» rhv .5. .AJ rF- ‘4 rh- N r 7, \ V s s . ~ \t S \ n...
u a . o v n . a u a I a t o v o t n t a . a . a u a t a e a . a n 4 t a - a t a u a n a t a era a a t a n a . a u a . a s a a a c a n a t a w a a a u a n u n q t e
.c. .1. new .4. .t. .vc .q. .4. ‘4. .ua .o. .1- .s. std .u. .cv 41. .4. n1- .1. .uv .4. :- 54v .1. .r. .4» «4v 4t. .1.- ‘11 .c- .1. 41‘ n: .1. .vv ‘4. qt. :4 .t.
.61 PM. I‘( AC: gig I.- A... 8" 1'. 1‘; A‘s D‘s n‘. a... I‘o .1: I‘( II. 0“. A, I'- I‘d 0‘4 Ail I‘d I‘d .04 0’. 6‘1 .1: Q: 3 n‘c 1'4 D‘s n'o I..< 0‘4 I‘d F‘: A‘. A‘l

 

46

Table 2.4. Example of the weather variables used to run bean model BEAN GRO.
*WEATHER :Moaoeoao TANZANIA

8 INSI LAT LONG ELEV
SUMO -6.5 37.3 525
COATS SRAD TMAX THIN RAIN
93121 15.7 22.0 21.1 .0
93122 14.3 30.2 21.1
93123 11.6 28.0 20.6
93124 12.3 30.0 21.0
93125 16.1 30.8 21.6
93126 21.1 31.1 19.9
93127 17.7 29.2 19.9
93128 11.9 29.0 21.2
93129 16.2 31.1 20.0
93130 14.3 30.3 21.0

N

N

OOU‘NOOOOOHOOOOONOHOOHOOOOOOOOOANOOONOHdHNNOOOOOOOOO‘O-boooomwd

O

OOmmOOOOONOOOOOOOfiOOOOOOOOOOOOh-ISOOOOOQU‘AQHOOOHU‘OOOU‘OOAOOOOQO

93131 14.3 28.6 21.2
93132 11.8 26.4 18.4
93133 9.8 27.0 19.5
93134 9.8 26.2 20.2
93135 8.8 27.0 19.0
93136 11.4 29.0 17.0
93137 18.8 28.0 16.0
93138 9.7 27.5 15.6
93139 15.3 28.5 19.4
93140 19.5 25.7 19.7 .
93141 9.0 29.5 20.1
93142 16.2 28.0 19.2
93143 15.5 28.4 20.5 .
93144 17.9 30.4 19.9
93145 17.4 28.5 19.5 .

93146 13.4 30.1 19.0
93147 17.6 28.9 19.9
93148 16.9 30.0 17.5
93149 20.4 29.0 17.1
93150 16.2 30.0 19.5
93151 17.0 29.2 18.5
93152 18.2 28.5 16.8
93153 13.2 28.0 16.2
93154 13.4 28.5 17.0
93155 16.0 29.2 17.0
93156 11.1 25.5 16.3
93157 11.8 27.3 14.7
93158 19.5 28.4 14.5
93159 18.6 28.4 14.5
93160 14.8 27.4 14.6
93161 19.8 27.5 13.5
93162 10.1 24.8 13.5
93163 18.6 27.4 16.0
93164 8.8 26.2 14.0
93165 10.2 27.0 18.6
93166 14.2 27.2 16.1
93167 11.9 26.1 17.4
93168 12.9 26.8 16.0
93169 20.8 26.0 16.0
93170 15.8 26.5 17.1
93171 15.4 27.0 15.8
93172 15.8 28.5 16.1
93173 17.1 28.0 16.0
93174 21.7 29.0 17.5
93175 17.1 27.6 16.5
93176 13.3 26.5 13.5
93177 12.9 26.4 14.0
93178 12.9 25.5 18.0
93179 11.9 26.6 17.8
93180 14.7 26.6 17.3

 

AI

 

:— I .F m— : ,L c x K c ... v . ~ .
1.53.. 252% 1.5.3“ £932.: Flux .fEHm z><.:.. XE: $7.: 2.72"» 29.-.: .37.: .ZIIE 2:3: EEL 2.2.4 .......... m2<2m> \t<>z
~

QEQOE OWOOZUKU I WFZMHUHkLBOU HQXEOZMU EMQXKQQ

.mCOZ..VEOQ EEC-«NCQL. L032: ozezvsmm

—0—vOE CIOD 0p: Du¢3_8>0 0» 00¢: 9:31.230 ORA. CIG—ufitx 50.20 Uta £0320)? Sammy-2:39 LO.‘ vJCUmOEDOO UmuOCUW 30:22:33 .WN‘ 0~A~§rh

 

47

Am>mu HmEuwcuouonQ. uncauavcoo answuao nave: UmOA UOQ Hanan comma Ou um>wuazo uOu Dumasvou mafia mDQOm _

.anx.. macauavcoo onwkouo vumucmun umvc: Don uma v00» womuw>< >0mom .

.m>mv HmEuonuOuocQ. macauavcoo :u30uo vumvcmum um uuocoo UOQ new COauwusv unaaaau vmwm macaw _

.0. vowm uwa uco«03 Essaxmz ommhz .

Hausa + nmmm ou vocofiuduuma ma umcu nuzouo aafimv uo cofiuomuu sasdxmz amux .

.NEU. .muoaumwa mwuzu. mama Haau uo muam Eaedxm: mANHm .

.0\NEo. macaudvcoo nuIOuo Oumvcmum nova: um>auaao uo «mum umma oauaooaw m>¢4m _

.m-%::oo. unodd no“: vcm .Noo ea> omm .o on um ouau mfimocucamouoza umma ssefixmz x<zmq _

.m>mu HmEuuLuOuOLQ. codmcmaxv mama uo Uco Ucm .Hm. nozoau umuau cmozuwn weds mquam _

.m>nv HmEumcuouOLQ. .bmv >udu3ums HavaooHOam>ca Ucm 5mm. vow» umuau cmmzuwn fleas zmnom _

.m>m0 HmEumcuOuocQ. .mm. Ummm umuqu new uoIOHu umuau cmmzuwn mafia omnam _

.m>mu Hmsuocuouona. Amm. non umudu new umxoau umufiu cmoxumn asap :muHm .

Am>mv HmEuocuouozav .Hm. uocmummaam uw30~u vcm wocvouvew ucmHQ cmwzumn weds Amlzm .

Auaoc\H. mEHu nu“: vo«qu0uoca ou ucwEQoHo>wn uo mmcoammu o>aumaou may no macaw zmmmm .

Anson. poouum cuocma>m0 oc cufi3 mummwuooua ucoeaon>ou m>auozvouaou coax: 30Hmn cuocoq xmo uuozm Hmoauauo qomo _

.maau oow.¢ own. moconn um>fiuaso macu Loan: Ou ma>uoow mcu new mUOU .Oom .

mco«u«c«uoo .

m.m om.m o.ma omn.o co.“ o.mmH .mmm oo.H oo.h oo.- m.b o.~ o.v~ ooo.o FH.NH emooz< +uwvcoz cmHUmcmo «HoomH
m.n om.m o.aa omm.o oo.H o.mmH .mmm oo.~ oo.> om.m~ o.HH o.~ m.vN ooo.o FH.NH amooz< UnonUom maoomH
o.h om.m o.m~ omm.o oo.H o.mmH .mom co.“ oo.m oo.m~ o.o o.m o.m~ ooo.o p~.NH emooz< nouacaz NHoomH
o.n om.m o.mH omm.o oo.H o.mma .mmN co.” oo.wa oo.- o.oH o.m o.w~ ooo.o hH.NH amooz< «HHmDMMH HaoomH
ma va nH ~H «H OH a o h o n v m N H _
«Doom >Dmom «30mm owns: ammx mANHm ¢>qu x<qu hqnqh zmuom omuqh :muqm Amszm zmmmm ammo .oom ..........mz<zm> .m<>w

ammo: onmomwmu . mazmHoEmmoo muraozmw zdmmwmo.’
20228 5335 323 805mm
.308 52. 05 33:50 9 won: ESE—3 25 5092 85¢ 23 .6283 5:350 .8 356830 0:28» 3385mm .m.~ 03:.

 

Geneti
were calibrate
(1994). Culti'l
milable Came
followed by a 1'

The as

Stimaxetheuc

lvz) 39d flower

 

Offlowering v.

Stages) was the:

48

Cultivar Coefficients Calibration

Genetic coefficients in FILEC for the Canadian Wonder variety used in this study
were calibrated using the methods suggested by Hoogenboom et al. (1991) and Boote
(1994). Cultivar coefficient estimation steps involved initial simulations using the already
available Canadian Wonder coefficients as defaults in the BEAN GRO crop simulation model,
followed by a trial and error approach to estimate the values.

The stepwise process involved in calibration of the genetic characteristics was to first
estirmte the crop life cycle by adjusting the EM-FL (i.e., minimum time3 between emergence
(V2) and flower appearance (R1) growth stages in photothermal days), until the correct date
of flowering was attained. The SD-PM (i.e., minimum time between R, and R, growth
stages) was then adjusted until simulated and field measured dates of physiological maturity
were approximately the same. The following step involved the estimation of dry matter
accmnulation at flowering, pod yield (kg ha"), seed yield (kg ha“), seed m’2 at maturity and
the maxim LAI. This was done by adjusting EM-FL, FL-SH (i.e., time between plant first
flower and first pod R3 growth stage), FL-SD (i.e., time between first flower and first seed
(11,), FL-LF (i.e., minimum time between R, and R, growth stages). The SFDUR (i.e., time
for seed filling period in a pod) and PODUR (i.e., time required for a cultivar to reach final
pod load under optimum conditions) were also adjusted.

LAI was calibrated by manipulating SLAVR (i.e., specific leaf area under field

conditions during the season, cm2 g") and the SIZLF (i.e., maximum size of three leaflets or

 

3Time inthiscontextreferstophotothermal time as defined by Jones et al. (1991).

trifoliolates, c
changes in W

Somes
values. These
(PPRTMX).
(PPSHMX). T
values. Leaf
specific leaf 2

accumulation .

 

adjuaing paran
dill and CA:
glCH201m'2 d‘1
“W ”Wired
3‘ 63% in the e

The ab
field “mined
(kg ha“), seer

Obtained Unde

49

trifoliolates, cm’). Calibration for seed size, seeds per pod was done by making necessary
changes in WTPSD (maximum weight seed g" under nonlimiting field conditions).

Some species parameters in the species file (SPE) were adjusted using field measured
values. These parameters included plant P values for maximum concentration (%) in roots
(PPRTMX), stems (PPSTMX), leaves (PPLFMX), seeds (PPSDM), and in shells
(PPSHMX). The minimum values for the same plant parts were also adjusted using measured
values. Leaf parameters such as maximum (SLAMAX), and minimum (SLAMIN) and
specific leaf area (SLA) were also adjusted using field measured values. Biomass
accumulation during the vegetative and reproductive growth phases were matched by
adjusting parameters CMOBMX (i.e., proportion CHZO reserves that can be mobilized in a
day) and CADSTF (i.e., CHZO added to stem reserves at the end of day after growth
g[CHzO] rn'2 d"). To match the simulated pod yield, seed yield, and seed number (m") with
those measured under field conditions, the maximum threshing percentage (TI-IRSH) was set
at 63% in the ecotype (ECO) file.

The above steps were repeated until reasonable agreements between simulated and
field measmed values of anthesis and physiological maturity dates, maximum LAI, pod yield
(kg ha"), seed yield (kg ha'1 ), and seed in2 , seed size (g seedl ) at harvest maturity were
obtained under optimum field conditions. The genetic coeflicients obtained were used to
simulate other treatments. Since the same bean cultivar was used in both years of study, the
genetic eoeficients generated during the calibration process were tested against data fi'om the
other experiment of the 1993 season, and the experiments of the 1994 season. The equations
and parameters governing the main processes for the P in the P-submodel for both the TSP

and MPR fertilizers used in this study are described in Chapter 3.

 

The m

 

Comparisons c
rambles the mtl
experiments forI

The mo

by switching o

 

simulation com
Each ex

a "onfipen‘me

50

Model Evaluation

The model was tested by comparing simulated values with field collected data.
Comparisons of growth, development, and yield variables obtained from field trials with
variables the model predicted under identical conditions were made for rainfed, and irrigated
experiments for the 1993 and the 1994 growing seasons.

The model was evaluated under P limiting and nonlimiting conditions. This was done
by switching on or switching ofi‘ the P simulation process using the option provided in the
simulation control section in the input file.

Each experiment was simulated separately. Since applied N to the experiment was

a non-experimental variable in both experiments, it was not used for model evaluation.

RESULTS AND DISCUSSION

Soil Profiles Properties

Values for the soil properties used in the model are given in Tables 2 a and 2b. The
drained upper limit for the irrigated site (Alfisol, isohyperthemic, typic tropustic) was between
0.192 to 0.265 cm cm". For the rainfed site (Oxisol, isohyperthemic, aridic tropustic) it was
between 0.189 to 0.301 cm3 cm”. The total profile depth at each site was 100 cm. The
watertable was 85 cm at the irrigated site, and below 110 cm at the rainfed site during the
rainy periods. Both soil profiles had high bulk densities ranging fi'om 1.35 to 1.55 g cm“3 at
the irrigated site, and 1.29 to 1.65 g cm” at the rainfed site. The native soil K reserves were

low. This may have been due to clay minerals at both sites being predominantly kaolinitic,

 

nhich art p00
the top layer 0

water entry ar

 

 

 

content during

shown in App

 

 

 

Monthl
5“mi and Ap
mid-May to m
WlOds averag
14.75 M] m-z i1
exPeriments r
(figure 2.2) 1

the1993seas

51

which are poor in K (Thung, 1991). Before the onset of rain, some cracks were observed in
the top layer of the soil profile at the irrigated site. Such condition may have influenced the
water entry and redistribution in the soil during the first days of the rain. Soil profile water
content during the experimental periods were not measured. Soil profile descriptions are

shown in Appendix la for the rainfed site and Appendix 1b for the irrigated site.

Weather During the Study

Monthly mean values for the weather variables are shown in Appendix 2a for the 1993
season, and Appendix 2b for the 1994 season. Daily temperature patterns were lower from
mid-May to mid July in both seasons (Figure 2.1). Solar radiation during the experimental
periods averaged 14.70 M] rn'2 in 1993, and 14.06 M] m2 in 1994 in the rainfed plots, and
14.75 MI rn'2 in 1993, and 15.14 MJm‘2 in 1994 in the irrigated plots. In general terms, both
experiments received the same amount of solar radiation in the 1993 and 1994 seasons
(Figure 2.2). Due to the intense and long thunderstorms, the irrigated site was flooded during
the 1993 season when the treatments had already been applied, and the crop was at the R1
growth stage. Consequently, the first crop sown on May 8 had to be replanted later. The
second crop was sown at the same site on May 28, 1993 after the plots were dry and ready
for replanting. The rainfed experiments differed markedly in the amount of rainfall received '
during the difl‘erent growth phases. Much of the rainfall received during the rainy seasons
came in the form of intense thunderstorms, resulting in a higher percentage of water being lost

as runofl‘(F”rgure 2.3). Due to the delay of the onset of rains during the 1994 season, planting

was also de

experiment.

Com:
Mcosplnert
attacked by as
quality. The

periods (May-

 

 

loss due to '

52
was also delayed by 10 days compared to the 1993 season planting date of the rainfed
experiment.

Common diseases included rust (Uromyces appendiculatus), angular leaf spot
W110 pinades), and white mold (Sclerotinia sclerotiorum). At harvest, seeds were
attacked by ascochyta blight (Ascochyra spp) which resulted in yield losses, and reduced seed
quality. The disease was more serious in the experiment grown during the relatively cool
periods (May-August) when water (fiom irrigation) and high humidity were abundant. Seed

loss due to this disease was as high as 20% in some plots.

 

li\ llll‘:'ll"."l: -L

.1. Sea

Figure 2

53

 

 

30- . _

 

25- e
20- -
15- —

to. l _

Minimum

 

. - 1994
30 - . ..

25 7 —
2o - ~

Temperature (”C day‘l)

15‘ " ' l-
104 -

 

 

 

5

a
~0$~°§3~逰~$¢ Wig"

Day of the Year

9%

Planting day of the Year:
094 Rainfed 1993
104 Rainfed 1994
148 Irrigated 1993
128 Irrigated 1994

Figure 2.1. Seasonal patterns of temperature regimes during the 1993 and 1994 growing
seasons.

54

S

 

 

 

 

 

1993
30' r
c:
'>.
.g 20 .
e
v
g 0
.5 1994
E 304 _
h
[-1
a
'5 20- _
m I l.
10— l I _
0 I I I I I r T l I T I
0c e
9 ~0¢g33¢~$- “Vol?
DayoftheYear

Planting day of the Year:
094 Rainfed 1993
104 Rainfed 1994
14s Irrigated 1993
128 Irrigated 1994

Figure 2.2. Seasonal patterns of solar radiation during the 1993 and 1994 growing seasons.

 

 

Fl
Se11:62.3. 3e

55

 

 

 

 

 

 

 

 

c."
g; _
"D
E
g _
i
jg 30- .
a: 'I

20‘ .l l b

I I

 

 

I

 

 

O 14 28 42 56 70 . 84
Days after planting

Figure 2.3. Seasonal patterns of rainfall distribution during the 1993 and 1994 growing
Seasons.

 

Table
the bean mod.
ha" (TSP) trea
showed that (
predict seed til.
held, seed size
maximum TH,”

10 be consider

Model

 

 

values are diSC

 

mm of

in Chapters 4 a
Rainfal

growing Seasc

““fa’orabre con

hemmedexpe

56

Cultivar Genetic Coefficients

Table 2.5 shows the genetic coeflicients of Canadian Wonder cultivar used to calibrate
the bean model against the data obtained fiom the 1993 irrigated experiment, with 88 kg P
ha" (TSP) treatment. To get the best results on BEAN GRO model performance, this study
showed that calibration should start by predicting life cycle timing, then calibrate traits to
predictseedfillingduration, ava'ageseedsperpodtimetoreachfinalpod load, podand seed
yield, seed size, and finally, fine tune total dry matter accumulation at harvest maturity. The
maximum THRSH (shelling percentage) in the ECO file was found to be an important trait

to be considered during the calibration process.

Field Measured Results

Model performancefor predicting P concentration (%) and total P uptake (kg ha")
at various growth stages, and grain P (kg ha") at maturity as compared to field measured
values are discussed in Chapter 3. Field measured total dry matter, yield components and
response of bean to difi‘erent forms and amounts of applied fertilizer P are discussed in detail
in Chapters 4 and 5.

Rainfall amount at the location, 523 mm during 1993 and 342 mm during 1994
growing seasons, was poorly distributed, especially during the 1994 season, creating
unfavorable conditions for good crop performance during the reproductive growth stage for

the rainfed experiment (F igure2.3). Low temperatures fi'om June to August (Appendices 2a

and 2b) wh'
reproductive

Plant
germination.
decreased at
orpdiments.
maturity in rr
during the Stu

model perforr

 

The s

Were efiimate.

57

and 2b) when the irrigated experiments were conducted, slightly delayed vegetative and
reproductive developments compared to the rainfed experiments.

Plant population (plants m”) count taken at the V3 growth stage showed that seed
germination, crop. vigor and emergence were very good. However, plant population
decreased at maturity regardless of the fertilizer treatments (Appendices 5a and 5b) in both
experiments. Therefore, plant population in this study was considered suboptimal at harvest
maturity in most plots. Growth, development, yield, and yield components data collected
during the study from each experiment formed an independent data set used to evaluate the

model performance.

Model Evaluation Results

The second objective of this study was to evaluate the ability of the BEANGRO
model to simulate bean growth, development, yield, and seed yield components. Once the
genetic coefficients in the cultivar file and crop constants in the species, and ecotype files
were estimated (see Materials and Methods), they were fixed so the accuracy of the model
could be determined against field measured data.

Under the 1993 irrigated, and nonlimiting P conditions (i.e., P switched ofi), the
model correctly predicted the dates for anthesis, timing of the pod and seed set, seed size (mg
seed") and seeds pod ‘. Simulated pod yield (kg ha! ), and seed yield (kg ha ) at harvest
maturity were over predicted compared to the results measured under field conditions (Table
2.6 a). The model under predicted the crop performance in terms of biomass accumulation

(kg ha") by 4%. The under prediction of biomass accumulation with nonlimiting P conditions

 

were due to
could be due
or due to the

Unde
by no days I

under predicts

 

at harvest mat
the 1993 sea».
hwammr
The simulatio
theorem
acCumulation
APPUI
“0P gromh

developments

58

were due to the high persisting senescence process after R, growth stage. Such condition
could be due to the high demand ofN during the pod filling phase (Sinclair and de Wit, 1976),
or due to the source-sink relationship (White and Izquierdo, 1991).

Under the rainfed, and nonlimiting P conditions the anthesis date was over predicted
by two days, while the physiological maturity was accurately predicted. However, the model
under predicted total biomass accumulation by 13% at anthesis, and over predicted it by 3%
at harvest maturity (Table 2.6 b). The results fiom the 1994 season were similar to those of
the 1993 season except that the model over predicted the biomass production by 54% under
irrigated conditions, and under predicted its accumulation by 16% under rainfed conditions.
The simulation results showed water stress factor of 0.26 at pod formation to 0.75 at
physiological nuturity. Such a condition may be have influenced the low predicted biomass
accumulation observed under the 1994 rainfed conditions.

Application of P fertilizer under irrigation conditions did not affect the timing of the
crop growth stages. The model accurately predicted the vegetative, reproductive
developments and physiological maturity with the 0 kg P ha‘l treatment under irrigation
conditions of the 1993 season. However, seed yield, seed number (m") seed size, maximum
LAI, and biomass accumulation at anthesis were all under predicted by the model. The
biomass accumulation at harvest maturity was over predicted by 20% by the model at harvest
maturity (Table 2.7 a). Under the rainfed experiment, flowering stage was over predicted by
three days. The results on seed yield components were as reported under irrigated conditions,
except that seed in2 was under predicted by 44% while seed pod'l were over predicted by

11%. Biomass at harvest maturity was over predicted by 26% (Table 2.7 b).

"M

.v
.r.

‘l
Irv

a. 'b!‘
s"
.------

- Yr'
S.e-
A F N"
r0. 3.-
a- a- -'
SLE Y.
.n-...
that--.
i .-H-
U2Y‘
...
v‘. n
41.3 3..

'VAIN of“
O

(b) The rainfe

Table 2.6. 5!
conditions (P

.1 .d . a a; T. 3. U. n»
t. . u .L a . rs. .. e 3. Wu .C C. t
A; v . V. v . 9 . .au . I. .\~ a: C
a . r . v. . a H to v” 2. 3 Tc
«2 a up a. .0 an «a va y. y .l
a.” v. a c .L .L . . .L .L .r... K. r. .1.
a . n... v 7. .5. .m. at «a. 1. . . s a {a
at D. . Ce 3. r ,5 A: U. a .5 ..~.

TSP rtfasten

59

Table 2.6. Summary of simulation outputs of bean performance under nonlimiting P
conditions (P switched ofi) compared with the high P rate for the 1993 season.

 

(a) The irrigated experiment.

*MAIN GRWI'H AND DEVELOPMENT VARIABLES TSP

8 VARIABLE PREDICTED MEASURED
ANTHESIS DATE (dap) 31 32
FIRST POD (dap) 34 36
FIRST SEED (dap) 40 41
PHYSIOLOGICAL MATURITY (dap) 66 66
POD YIELD (kg/ha) 2079 1488
SEED YIELD (kg/ha) 1508 1215
SHELLING PERCENTAGE (%) 72.53 81.65
WEIGHT PER SEED (g) .392 .314
SEED NUMBER (SEED/m2) 384 387
SEEDS/POD 3.50 3.47
MAXIMUM LAI (mZ/mZ) 3.04 3.00
BIOMASS (kg/ha) AT ANTHESIS 1240 1298
BIOMASS (kg/ha) AT HARVEST MAT. 3997 4160
HARVEST INDEX (kg/kg) .377 .292

(b) The rarnfed experiment.

*MAIN GROWTH AND DEVELOPMENT VARIABLES TSP

8 VARIABLE PREDICTED MEASURED
ANTHESIS DATE (dap) 31 29
FIRST POD (dap) 33 32
FIRST SEED (dap) 38 33
PHYSIOLOGICAL MATURITY (dap) 6O 60
POD YIELD (kg/ha) 1452 1714
SEED YIELD (kg/ha) 1054 1037
SHELLING PERCENTAGE (8) 72.60 60.5
WEIGHT PER SEED (g) .352 .270
SEED NUMBER (SEED/m2) 300 340
SEEDS/POD 3.50 3.70
MAXIMUM LAI (m2/m2) 2.44 2.20
BIOMASS (kg/ha) AT ANTHESIS 946 1096
BIOMASS (kg/ha) AT HARVEST MAT. 2734 2658
HARVEST INDEX (kg/kg) .386 .390

 

‘ TSP refers to triple superphosphate fertilizer

U
.
rt
TIT-Til;
»
T'-

(a) The '

Table 2.7. S-
lit? ha") fo

 

 

 

I . . . . .
I .II. T: HUll .. . a . . .r. . _ a L :u .. .r. I as.»
I T. .. .L .. s. . u, t. _. .. A. .. . . . 2. t. .. .L . . Hi. ”Us 5. .9. C. .5 ~ .
r\ c a Y. . . .. .hr . . 2. I. .3 n .5 . 3. n; a a V. . . .. V) .I .. .Q -~ 3»
. . a . v. . . .u. I. V“ l. L. T. . i. . .1 a . u . a . v. . . . ”r. C» U” L. an 7.
D D a: A. ._ 1. .c at. .. y NJ Ir m 4M 4. . n— A: C «C a. v.1.ae mu .. AN“ ry .o
.u .R W. «J .L r. . 3 T. .L ”r“ .u. pl . rue . o . in .u. V. A c .L .L a . .L .L V... . :u.
a . a . .H w .L ..—.. T. .L .L a r v . a a u :h an. . N o . . A .4. I.» .L In L F. .r. ru- v a a . a».
.3. of D. .D .3 .R A: .2 U. .5 p5 . e .1 ur . In.» .2 pf 3. n . .3 as. IN n5 Cc M. a: u: I.
«m s
a
.

I60

Table 2.7. Summary of simulation outputs of bean performance under control conditions (0
kg P ha") for the 1993 season.

(a) The irrigated experiment.

*MAIN GROWTH AND DEVELOPMENT VARIABLES

8 VARIABLE PREDICTED MEASURED
ANTHESIS DATE (dap) 31 32
FIRST POD (dap) 34 36
FIRST SEED (dap) 40 41
PHYSIOLOGICAL MATURITY (dap) 66 66
POD YIELD (kg/ha) 393 496
SEED YIELD (kg/ha) 281 329
SHELLING PERCENTAGE (8) 71.66 66.33
WEIGHT PER SEED (g) .310 .330
SEED NUMBER (SEED/m2) 91 100
SEEDS/POD 3.50 3.50
MAXIMUM LAI (m2/m2) 0.99 1.74
BIOMASS (kg/ha) AT ANTHESIS 732 947
BIOMASS (kg/ha) AT HARVEST MAT. 1412 1129
HARVEST INDEX (kg/kg) .199 .291

(b) The rarnfed experiment.

*MAIN GROWTH AND DEVELOPMENT VARIABLES

@ VARIABLE PREDICTED MEASURED
ANTHESIS DATE (dap) 31 28
FIRST POD (dap) 33 30
FIRST SEED (dap) 38 32
PHYSIOLOGICAL MATURITY (dap) 63 63
POD YIELD (kg/ha) 411 444
SEED YIELD (kg/ha) 296 271
SHELLING PERCENTAGE (%) 72.01 61.1
WEIGHT PER SEED (g) .473 .252
SEED NUMBER (SEED/m2) 63 107
SEEDS/POD 3.50 3.13
MAXIMUM LAI (mZ/mZ) 0.90 1.42
BIOMASS (kg/ha) AT ANTHESIS 633 726
BIOMASS (kg/ha) AT HARVEST MAT. 1285 956

HARVEST INDEX (kg/kg) .230 .284

 

Bion

 

to that m
fertilizers ur_
maximum L
tainted condl
from the fit
application 0t i
except that at
the field by 4
The
the Simulatio
Men's and p
to field mm
Conditions 8
1mm "u

prediCIed {h i

61

Biomass accumulation predicted by the model at harvest maturity was high compared
to that measured from the field applied with 22 kg P ha" in form of TSP and low with MPR
fertilizers under irrigated conditions (Table 2.8 a). With TSP application, seeds pod",
maximum LAI, and biomass accumulation at anthesis were accurately predicted. Under
rainfed conditions, biomass accumulation at anthesis agreed well with the values measured
from the field, while that at physiological maturity was over predicted by 27% with
application of TSP. Similar results were obtained by the application of MPR at the same rate
except that at harvest maturity, biomass model prediction was higher than that measured from
the field by 47% (Table 2.8b).

The simulation results firrther showed that medium P rate (44 kg P ha" TSP) used in
the simulations over predicted the crop performance in terms of biomass accumulation at
anthesis and physiological maturity (Tables 2.9 a). The model predicted seed value was closer
to field measured data under rainfed conditions of the 1993 season than that from the irrigated
conditions at 44 kg P ha " (Table 2.9 b). Overall, the results show that apart fiom the
discrepancy in the prediction of biomass accumulation at harvest maturity, the model correctly

predicted the bean yield with application of different P fertilizer types and levels.

62

Table 2.8. Summary of simulation outputs of bean performance under low P rate conditions
(22 kg P ha") for the 1993 season.

 

(a) The u'rrgated experiment.

*MAIN GROWTH AND DEVELOPMENT VARIABLES TSP1 MPR2

8 VARIABLE PREDICTED MEASURED PREDICTED M E.A S U R E D
ANTHESIS DATE (dap) 31 32 31 32
FIRST POD tdap) 34 36 34 36
FIRST SEED (dap) 40 41 40 41
PHYSIOLOGICAL MATURITY (dap) 66 66 66 66
POD YIELD (kg/ha) 795 891 1086 812
SEED YIELD (kg/ha) 576 604 734 570
SHELLING PERCENTAGE (%) 72.47 67.79 72.15 70.20
WEIGHT PER SEED (g) .368 .332 .410 .287
SEED NUMBER (SEED/m2) 158 182 191 199
SEEDS/POD 3.50 3.60 3.50 3.37
MAXIMUM LAI (m2/m2) 1.93 1.76 1.69 1.97
BIOMASS (kg/ha) AT ANTHESIS 1168 1035 1040 1197
BIOMASS (kg/ha) AT HARVEST MAT. 3144 2955 2686 2888
HARVEST INDEX (kg/kg) .183 .204 .292 .197

(b) The rarnfed experiment.

*MAIN GROWTH AND DEVELOPMENT VARIABLES TSP MPR

G VARIABLE PREDICTED MEASURED PREDICTED M E A S U R E D
ANTHESIS DATE (dap) 31 28 31 28
FIRST POD (dap) 33 30 33 30
FIRST SEED (dap) 38 32 38 32
PHYSIOLOGICAL MATURITY (dap) 62 63 62 63
POD YIELD (kg/ha) 908 812 964 621
SEED YIELD (kg/ha) 654 508 698 386
SHELLING PERCENTAGE (%) 72.07 62.6 72.3 62.1
WEIGHT PER SEED (9) .470 .274 .480 .270
SEED NUMBER (SEED/m2) 139 186 145 144
SEEDS/POD 3.50 3.33 3.50 3.73
MAXIMUM LAI (m2/m2) 1.94 1.61 1.65 2.33
BIOMASS (kg/ha) AT ANTHESIS 933 969 894 803
BIOMASS (kg/ha) AT HARVEST MAT. 2422 1756 2211 1155
HARVEST INDEX (kg/kg) .270 .289 .316 .334

 

' TSP refers to triple superphosphate fertilizer
2LAPTKrefiustodNfimdhuprphosphauunxflcfintflhux

Table 2.9. .
conditions (

 

(a) The im'g
was sax—

‘ UV” r
" we}:
--.---
‘ '. r-j.
Akin.
«Ora-
D .\
.. ,-.
vim-u.

.55..

Per...
“.3..L

f'l
1'!

lb) The rainft

 

 

 

NI" ' II-
t- . ..
"“‘ 453“79
a ‘V‘n.
“ m,- 15’
5-

63

Table 2.9. Summary of simulation outputs of bean performance under medium P rate
conditions (44 kg P ha") for the 1993 season.

 

(a) The irrigated experiment.

*MAIN GROWTH AND DEVELOPMENT VARIABLES TSP1

8 VARIABLE' PREDICTED MEASURED
ANTHESIS DATE (dap) 31 32
FIRST POD (dap) 34 36
FIRST SEED (dap) 40 41
PHYSIOLOGICAL MATURITY (dap) 66 66
POD YIELD (kg/ha) 1293 1263
SEED YIELD (kg/ha) 936 1030
SHELLING PERCENTAGE (%) 72.3 81.6
WEIGHT PER SEED (g) .424 .330
SEED NUMBER (SEED/m2) 221 312
SEEDS/POD 3.50 3.67
MAXIMUM LAI (m2/m2) 2.36 2.68
BIOMASS (kg/ha) AT ANTHESIS 1231 1121
BIOMASS (kg/ha) AT HARVEST MAT. 3712 3518
HARVEST INDEX (kg/kg) .252 .293

(b) The rarnfed experiment.

*MAIN GROWTH AND DEVELOPMENT VARIABLES TSP

e VARIABLE PREDICTED MEASURED
ANTHESIS DATE (dap) 33 ’ 29
FIRST POD (dap) 34 32
FIRST SEED (dap) 38 32
PHYSIOLOGICAL MATURITY (dap) 62 63
POD YIELD (kg/ha) 1190 1377
SEED YIELD (kg/ha) 860 922
SHELLING PERCENTAGE (%) 72.25 66.9
WEIGHT PER SEED (g) .470 .267
SEED NUMBER (SEED/m2) 183 341
SEEDS/POD 3.50 4.03
MAXIMUM LAI lm2/m2) 2.27 2.28
BIOMASS (kg/ha) AT ANTHESIS 938 1030
BIOMASS (kg/ha) AT HARVEST MAT. 2666 2472
HARVEST INDEX (kg/kg) .323 .373

 

'TSP refers to triple superphosphate fertilizer

 

Tab
marred Va
predicted \

acoimulatio

——-—

error value.

(model pred:

F—

at all P fertih
at R. growth
the control r
conditions. [

blonly 1%1

 

64

Table 2.10 indicates the model predicted bean performance compared to the field
measured values at equal fertilizer rates of TSP and MPR The results show that the model
predicted values agree well with the field measured values for the see yield and biomass
accurmrlation at physiological maturity. The difl'erences were within the measured standard
error values of the mean. Tables 2.6 to 2.10 also show comparisons between simulated
(model predicted), and field measured maximum LAI. The LAI increased with crop growth
at all P fertilizer levels. Overall, the predicted and field measured LAI values were maximum
at R, growth stage. The field measured LAI values ranged from 1.75 in the 1993 season for
the control rainfed conditions to 3.00 during the 1993 season for the 88 kg P ha“ irrigated
conditions. Under nonlimiting P conditions, the simulated maximum LAI was over predicted
by only 1% for irrigated experiment, and by 8 % for the rainfed experiment.

Results of reduction in LAI due to low P reported in this study fi'om simulated and
field measured values has also been reported by Adu-Gyamfi et al. (1990) in pigeon peas, and
Yan et al. (1995a) in beans. The low field measured LAI values compared to predicted
values reported hue may also have been attributed to pest and disease infestation during the
growing season. The predicted values in all treatments were within the standard error values

associated with observed data.

 

 

n.
F.,

We
3:"

1'? In

(%QPM'
WHC

RWZW

 

r t L U 1 a a; —. 6 .~ . v. aar- AC J; an; v. \
v a 9. Po. . . x A; I; A: L p: . «D A”. . a V. v a u. N I
yL E .3 E. Z 6.. T. . n. 2. . .L .. or .1 V. v~ ., at v.
LSSZIWVWNJV ~n .. .533 3 2.31.1.2:
.L 6 . .L .L V... , . :m E :m . a. .5. .H V. are .L F. ». AL ~L ”J.
Fr. L .L .L In. . . . . 9‘. A... ‘F . V4. . . .. nu. Aw I. ..~.. .8. .L f. A.
.2 vl 2. .3 V. .5 E. E, e .c "c . A. F. .6. 3. D. S A: . Q... E. M.
m vn
6‘
\) an.
® v.
' E

65

Table 2.10. summary of simulation outputs of been performance under high P rate conditions
(88 kg P ha‘l TSP) and medium P rate (88 kg P ha‘l MPR) for the 1993 season.

 

(a) The irrigated experiment.

*MAIN GROWTH AND DEVELOPMENT VARIABLES

8 VARIABLE

ANTHESIS DATE (dap)

FIRST poo (dap)

FIRST SEED (dap)
PHYSIOLOGICAL MATURITY (dap)
POD YIELD (kg/ha)

SEED YIELD (kg/ha)

SHELLING PERCENTAGE (%)
WEIGHT PER SEED (g)

SEED NUMBER (SEED/m2)
SEEDS/POD

MAXIMUM LAI (m2/m2)

BIOMASS (kg/ha) AT ANTHESIS
BIOMASS (kg/ha) AT HARVEST MAT.
HARVEST INDEX (kg/kg)

(b) The rainfed experiment.
*MAIN GROWTH AND DEVELOPMENT VARIABLES

8 VARIABLE

ANTHESIS DATE (dap)

FIRST POD (dap)

FIRST SEED (dap)
PHYSIOLOGICAL MATURITY (dap)
POD YIELD (kg/ha)

SEED YIELD (kg/ha)

SHELLING PERCENTAGE (%)
WEIGHT PER SEED (g)

SEED NUMBER (SEED/m2)
SEEDS/POD

MAXIMUM LAI (m2/m2)

BIOMASS (kg/ha) AT ANTHESIS
BIOMASS (kg/ha) AT HARVEST MAT.
HARVEST INDEX (kg/kg)

TSP1

PREDICTED

31
34
40
66
1800
1304
72.41
.420
310
3.50
2.68
1229
4006
.325

TSP

PREDICTED

MEASURED PREDICTED

32
36
41
66
1488
1215
81.65
.314
387
3.47
3.00
1298
4160
.292

MEASURED

MPR2

M E A S U R E D

31 32
34 36
40 41
66 66
1275 1222
921 967
72.25 79.13
.410 .298
225 324
3.50 3.70
1.89 2.67
1111 1195
3056 2459
.306 .393
MPR
PREDICTED M E A S U R E D
31 29
33 32
38 33
62 63
1038 1378
752 828
72.4 60.0
.473 .266
159 311
3.50 4.30
1.78 2.12
917 1116
2338 2013
.322 .411

 

‘ TSP = Triple superphosphate
’ MPR == Minjingu phosphate rock

fie

mama

(1989) con-

epidermal c
mggtstedln
swam
of the leave
between R,|
Photosyntha

Meas
flowering. T
dadinedtowa

value. Difl‘e

 

 

different p le

Egon
ha4). Signific
“l" exPerim
Worn.
ltg ha" “11h ti
Phi" (TSP )

The 1
fettilizer P a

ftmnOfTSp.

66

There was rapid decline of LAI at R, growth stage which coincided with the increase
in pod weight starting at R, to R, growth stages. Fredeen er a]. (1989) and Rao and Terry
(1989) concluded that low P levels decrease growth primarily through the efi'ect of leaf
epidermal cell expansion rather than on rate of photosynthetic activity per unit area, as
suggested by Whiteaker et a1. (1976) and Le Bot er a1. (1994). The decline in LAI with age
shown by simulated and field measured values was due to senescence, death, and abscission
of the leaves at a faster rate than new leaves were being formed. Furthermore, the period
between R6 and R , grth stages was for rapid grain filling process, therefore, the
photosynthate and metabolites were being trans-located to the developing seeds.

Measured values for specific leaf area (SLA) ranged from 250 to 350 cm 2 g’1 at
flowering. The SLA values started slow, increased to maximum at R, growth stage, then
declined toward maturity. The predicted SLA was about 20% higher than the field measured
value. Difi‘erences between predicted and field measured SLA values may indicate how
different P levels, pests and diseases affected the crop in the field.

Figures 2.4 a, b, c, and d show the simulated and field measured bean grain yield (kg
ha"). Significant difl‘erences were observed among treaments in the grain production for the
two experiments in both seasons. The model predicted grain yield formation with similar
accuracy in various treatments (data not shown). The simulated grain yield ranged fi'om 296
kg ha" with the control treatment from the 1994 rainfed experiment to 1304 kg ha" with kg
P ha" (TSP) treatment fi'om the 1993 irrigated experiment.

The model predicted and field measured seed production increased as the levels of
fertilizer P applied increased. This pattern was more Visible in plots that received P in the

form of TSP fertilizer (Figure 2.48 and b). However, in plots that received P in the form of

 

uv
Emu.-
wi 22»

~—

A 75

.4
ared c,

Figure 2

"Ha

67

 

1250 1
1000 -
750 -

 

 

Grain yield (kg ha'l)

 

 

 

 

0 20 40 60 80 100

Fertilizer levels (kg P ha")

V Field measured values
Simulated values

 

Figure 2.4a. Simulated and field measured grain yield with the application of TSP under
irrigated conditions. Vertical bars represent standard errors of the mean.

 

Are: we: 22% EEO

68

 

1993

 

 

Grain yield (kg ha'l)

 

 

 

 

o 20 4o 60 80 100
Fertilizer levels (kg P ha“)

7 Field measured values
Simulated values

 

Figure 2.4b. Simulated and field measured grain yield with the application of TSP under
rainfed conditions. Vertical bars represent standard errors of the mean.

Ci..- lt: 1.4“; {act-HE

 

69

 

rooo- 1;

 

750 r *

OI
O
O

M
0|
0
1;“

J

 

1994

Grain yield (kg ha")

 

 

 

o - . , - + . ,
0 so 100 150 200

 

Fertilizer levels (kg P ha“)

v Field measured values
Simulated values

 

Figure 2.4c. Simulated and field measured grain yield with the application of MPR fertilizer
under irrigated conditions. Vertical bars represent standard errors of the mean.

 

.4.

F1838 2
under rain

70

 

 

 

 

 

 

 

 

p _

g 1200‘

a 1000 ' 1994 F

E 8003 '
CD 500 i l

4001’? — —

zoo _

o i - . - . . - . - i

0 so 100 150 200

Fertilizer levels (kg P ha“)

v Field measured values
Simulated values

 

Figure 2.4d. Simulated and field measured grain yield with the application of MPR fertilizer
under rainfed conditions. Vertical bars represent standard errors of the mean.

MPR grain

of whether ‘

season mth
infected by a
The higher
W filling ;
of P fertilize;
disease and

Fign

{kg hdl) 3C

”Warmth

 

”teamed da
aCCUmulatio
13mg; essiveli
different new
and field me
DAP and S6
rams Were I
5a and 5b), a

in Which be

71

MPR, grain production tended to level of}~ at either 22 kg P ha‘l or 88 kg P ha" , regardless
of whether the experiment was grown under rainfed or irrigated conditions (Figure 2.4c and
d).

The discrepancy obtained in the comparison of predicted grain yield during the 1994
season with that measured from the field was mainly due to a higher percentage of seed being
infected by aschochyta blight and the rodent attack on the crop toward crop harvest maturity.
The higher yield variability under rainfed conditions is attributed to water stress during the
grain filling phase and P stress which was as high as 0.5 even in plots that received high levels
of P fertilizers. In general, the actual seed yield was constrained by soil water deficit, P stress,
disease and pest attack.

Figures 2.5a and b, and 2.68 and b compare the simulated and field measured biomass
(kg ha“) accumulation in both seasons. Significant difi‘erences were observed among
treatments in the rate of total dry matter production by the bean crop. The simulated and field
measured data shows that the control plots had the lowest values. The rate of dry matter
accumulation prior to flowering, as indicated by the slope of the grth curve, was
progressively lower with all treatments. Afier the flowering stage, the rates became very
difi‘ererrt between treatments indicating the effect of the P treatments applied. The simulated
and field measured biomass accumulation increased as the LAI increased until between 42
DAP and 56 DAP. Thereafter, field measured biomass production started to decrease. These
results were mainly due to the decrease in plant population at harvest maturity (Appendices
58 and 5b), and as reported by Hanway and Weber (1971), also due to rapid leaf and petiole

fall which began after R, growth stage.

The
treatments 1
However, ,
inconsistent
conditions d

This is more

 

 

A! h
largely of p.
Values did it
lite main res.
doc to peas
Considered '

by protein re

72

The model predicted the patterns of dry matter accumulation accurately in most
treatments under irrigated conditions and applied with P in form of TSP (Figs. 2.5a and b).
However, plots treated with P in the form of MPR, under rainfed conditions, gave
inconsistent results (Figs. 2.68 and b). The rainfed conditions compared to irrigated
conditions due to the poor rainfall distribution pattern that occurred after the R, growth stage.
This is more evident duting the 1994 growing season (Figure 2.3).

At harvest maturity (R,), a large proportion of the harvested crop was comprised
largely of pods, followed by stems, and finally leaves. In contrast, the simulated biomass
values did not decrease very much toward maturity, especially under irrigated conditions.
The main reasons for such results are that the current model does not account for yield loss
due to pests and diseases (Hoogenboom et al., 1994), and other microclimatic factors are not
considered in the model. Currently, the biomass accumulation declines toward maturity only

by protein remobilization and leaf senescence.

 

 

ATE- av: Cayuga—DE-flvUfl meEOm:

73

 

5m 1 1 l r l 1 1 1 l . l
1993 .o'. t

3750 — fit; ._

 

 

 

 

154 168 182 196 210 224

 

l 1994 ..

Biomass accumulation (kg ha")
0!

 

 

 

 

140 154 168 182 196 210
DayoftheYear

' Control simulated
- Control measured
a 22 kg P ha-1 simulated
v 22 kg P ha-l measured
0 44 kg P ha-l simulated
° 44 kg P ha-l measured

Figure 2.5a. Simulated and measured biomass accumulation with application of TSP fertilizer
under irrigated conditions. Vertical bars represent standard errors of the mean.

 

z - WQEQMMH
w v Egan—asauuu m
u—
:8:

.Sb,

Figure 2

”unis:

under ‘

74

 

 

 

 

 

Biomass accumulation (kg ha")

 

 

 

 

112 126 140 154 168
Day of the Year

' Control simulated
- Control measured
‘ 22 kg P ha-l simulated
' 22 kg P ha-1 measured
0 44 kg P ha-l simulated
- 44 kg P ha-l measured

Figure 2.5b. Simulated and measured biomass accumulation with the application of TSP
under rainfed conditions. Vertical bars represent standard errors of the mean.

It I\ ll

 

9.5— We: 50:5:Eauuu emu-:35

75

 

 

 

 

 

4000 I P l . l i l L l . l
- 1993 _
3000- /A .
g 2000 ‘ ”I" “
1ooo« '
E _ . .
.2
a o -A--
"g 154 168 182 196 210 224
3 4000 - 4 4
g 1994
E
.2 _
a:

 

 

 

 

140 154 -168 182 196 210
DayoftheYear

Control simulated
Control measured

22 kg P ha-“imulated
22 kg P ha»1 measured
88 kg P ha--l simulated
88 kg P ha»l measured

00a...

Figure 2.6a. Simulated and measured biomass accumulation with the application of MPR
under irrigated conditions. Vertical bars represent standard errors of the mean.

4 F.\ r‘

9.5: we: 50:535-300.» emu—=0:—

76

 

o—o—r
j

 

 

 

 

Biomass accumulatiom (kg ha“)
‘0
m
F6
i3
0)
3
O
61
h
a:
(D

 

 

 

 

1 12 126 140 154 168
Day of the Year

Control simulated
Control measured
22 kg P ha»1 simulated
22 kg P ha-1 measured
44 kg P ha-l simulated
44 kg P ha-l measured

0045-.

Figure 2.6b. Simulated and measured biomass accumulation with the application of MPR
under rainfed conditions. Vertical bars represent standard errors of the mean.

Thebea

 

growth, develo;
Tanzania. Th I
regarding bean
study allowed
interactions on

As state
during this stud
model is SSCnsi

give consistent

 

New in usi .
Umally Worse 1
Such obseR'atic
Simulate Cmp l

Thei'fm
be used co nfid
potential ban 1

the Med fof er

77

SUMMARY AND CONCLUSIONS

The bean model BEANGRO was calibrated, tested, and evaluated for predicting bean
growth, development, yield and seed yield components as influnced by P fertilizer regimes in
Tanzania. The systems simulation procedures used in this study showed how decisions
regarding bean production in Tanzania can be made using the BEANGRO bean model. This
study allowed the assessment of bean production system components effects and their
interactions on bean yield.

As stated by Saka et a1. (1990), slight over prediction of yields by the model, as found
during this study, is generally expected because control of experimental factors to which the
model is ssensitive can never be complete. However, since the BEAN GRO model did not
give consistent crop yield predictions under rainfed condtions, this shows that there can be
problems in using this model in marginal areas of Tanzania. In addition, field conditions are
usually worse than the ideal conditions under which simulations were run on the computer.
Such observations show that the model needs more adjustments so as to be able to correctly
simulate crop growth, development, and yield under rainfed conditions.

Therfore, it is concluded from this study that the bean crop model BEAN GRO can
be used confidently to predict the effects of P management and soil fertility regimes on the
' potential bean growth, development and yields under irrigated conditions. This may reduce

the need for empirical field trials where there is sumcient minimum input data.

Accor.
abstential inter
models in East
by identifying
However, there
way to increa
models in the z

Diseas

blighl (Phomc

 

 

 

WW8 ph

to be incorpor

78

RECOMMENDATIONS FOR FUTURE RESEARCH

According to Dr. Phillip Thornton (systems scientist at IFDC), there appears to be
substantial interest in the formation of a regional working group on the use of crop simulation
models in East Afiica Such a group could support skill development, and modeling activities
by identifying important application of crop models, and facilitating information exchange.
However, there is lack of support by the governments for crop modeling activities, thus, one
way to increase support is by demonstrating the benefits of the applications of the crop
models in the area.

Diseases ' such as rust (Uromyces appendiculatus var. appendicularus), ascochyta
blight (Phoma spp), and pests such as leaf beetle (00theca beningsem' Weise), bean fly
(Ophr'anm'apluseolr) are major factors in bean yield reduction, therefore, their efi‘ects need

to be incorporated in the model.

Acosta-Galle
comm

 

Adams, RM
Curry.
Emmi

Amt-Gym.
in pige
Nutri

A“1611mm Soc
and u

 

bnfius, 1F J,

\v 6,

Plant .
Black CA. 1‘.

30016. KJ. 19

011 00]

IFDC

B008. Kr,
PNu

Flour

30““. WT.
Walla
Eight
Inter

BIBLIOGRAPHY

Acosta-Gallegos, IA, and J .W. White. 1995. Phenological plasticity as an adaptation by
common bean to rainfed environment. Crop Sci. 35:199-204.

Adams, RM, C. Rosenzweig, RM. Peart, J.T. Ritchie, B.A McCarl, J. D. Glyer, R B.
Curry, J.W. Jones, K.J. Boote, and L.H. Allen. 1990. Global climate change and US
agriculture. Nature 35:219-224.

Adu-Gyaamfi, J ., K Fujita, and S. Ogata. 1990. Phosphorous fiactions in relation to growth
in pigeon pea (Cajcmas ccy'an (L) Millsp) at various levels of P supply. Soil Sci. Plant
Nutri. 36:531-543.

American Society of Agricultural Engineers (ASAE). 1978. Describing irrigation efliciency
and uniformity. ASCE J Irrig. and Drainage Div. 104235-41.

Angus, 1F ., J.W. Bowden, and BA Keating. 1993. Modeling nutrient responses in the field.
Plant and Soil. 155/156:57-66.

Black CA. 1993. Soil fertility evaluation and control. Lewis Publ. Boca Raton, FL.

Boote, KJ. 1994. Concepts of calibrating crop grth models. Report for training program
on computer simulation for crop grth and nutrient management. May 8-19, 1995.
IFDC, Muscle Shoals, AL.

Boote, K.J., J.W. Jones, G. Hoogenboom, G.G. Wilkerson, and SS. Jagtap. 1989.
PNUTGRO V1.02: Peanut crop growth simulation model. User's Guide. Univ. of
Florida, Gainesville, FL.

Bowen, W.T., J.W. Jones, and PK. Thornton. 1993. Crop simulation as a potential tool for
evaluating sustainable land management. p. 15-21. 15 J.M. Kimble (ed.) Proc. of the
Eighth International Soil Management Workshops: Utilization of Soil Survey
Information for Sustainable Land Use, July 19, 1992, Sacramento, CA.

79

Bray, RH. m
phospi
Curry, RB, R.
for ane

 

Day, PR 19(2|
DD 1
analys

Tehr, W.R.,
develc
931.

Freden A L.
and ca

Gm“. L], I
Physic

 

 

Gutierrez. an
l994_|
yield.

Hm 1. and
Plant

”may. J.J.,

Mem'

HoogenbOQr
yield
Tran:

Hoogenboo:
Batc
Inter

Hoogenboor
lOler

80

Bray, RH, and LT. Kurtz. 1945. Determination of total, organic, and available forms of
phosphorous in soils. Soil Sci. 59:39-45.

Curry, RB, RM. Peart, J .W. Jones, K]. Boote, and L.H. Allen. 1990. Simulation as a tool
for analyzing crop response to climate change. Trans. ASAE 33:981-990.

Day, PR 1965. Particle fiactionation and particle size analysis. p. 548-566. I_n_ C.A. Black,
D.D Evans, J.L. White, L.E. Eusminger, and F .E. Clark (eds) Methods of soil
analysis (2). ASA, CSSA, SSSA, Madison, WI.

Fehr, W.R, C.E. Caviness, D.T. Burmood, and IS. Pennington. 1971. Stages of
development description of soybeans (Glycine max (L) Merrill. Crop Sci. 11:929-
931.

Freeden, A L., LM Rao, and N. Terry. 1989. Influence of phosphorous nutrition on growth
and carbon partitioning in Glycine max . Plant Physiol. 89:225-230.

Grabau, L.J., D.G. Blevis, and HC. Minor. 1986. P nutrition during seed development. Plant
Physiol. 82: 1008-1012.

Gutierrer, AP, E.J. Mariot, J.R. Cure, C.S. Wagner-Kidder, C.K. Ellis, and AM. Villacorta
1994. A model of bean (Phaseolus vulgaris L.) grth types I-III: Factors afi‘ecting
yield. Agri. Syst. 44:35-63.

Hanks, J., and J.T. Ritchie. 1991. Introduction. I_n J. Hanks and J .T. Ritchie (eds.) Modeling
plant and soil systems. Agronomy 31: 1-3.

Hanway, J.J., and CR Weber. 1971. Dry matter accumulation in soybean (Glycine max [L.]
Merrill) plants as influenced by N, P and K fertilization. Agron. J. 63:406-408.

Hoogenboom, G., J .W. Jones, and KJ. Boote. 1992. Modeling growth, development, and
yield of grain legume using SOYGRO, PNUTGRO, and BEANGRO: A review.
Transactions ASAE. 35:2043-2056.

Hoogenboom, G., J.W. Jones, K.J. Boote, W.T. Bowen, N.B. Pickering, and W.D.
Batchelor. 1993. Advancement in modeling grain legume crops. 1993 ASAE
International Winter Meeting. Dec. 14-17, 1993, Chicago, IL.

Hoogenboom, G., Jones, J .W., and J.W. White. 1987. Use of models in studies of drought
tolerance. p. 192-230. 111 Research on drought tolerance in common bean. J.W.

White.
Progr

Hoogenboom
orient

Hoogenboom
bean c
No. N

 

International

 

81

White, G. Hoogenboom, F. Ibara, and SP. Singh (eds) Working Doc. No.41. Bean
Program, CIAT, Cali, Colombia.

Hoogenboom, G., J.W. White, J.W. Jones, and K]. Boote. 1994. BEANGRO: A process-
oriented dry bean model with a versatile user interface. Agron. J 86:182-190.

Hoogenboom, (3., J.W. White, J.W. Jones, and K.J. Boote. 1991. BEANGRO V1.01. Dry
bean crop growth simulation model. User's guide. Florida Agric. Exp Station Publ.
No. N-00379. Univ. of Florida, Gainesville, FL.

International Benchmark Sites Network for Agrotechnology Transfer (IBSNAT). 1988.
Technical report 1. Experimental design and data collection procedures for IBSNAT.
The minimum data sets for systems analysis and crop simulation. 3rd. Ed. Univ. of
Hawaii, Honolulu, HI.

International Benchmark Sites Network for Agrotechnology Transfer (IBSNAT). 1990.
IBSNAT network report (1987-1990). Univ. of Hawaii, Honolulu, HI.

International Benchmark Sites Network for Agrotechnology Transfer (1B SNAT). 1990.
Technical report 2. Field and laboratory methods for the collection of IBSNAT:
Minimum data set for the decision support system for agrotecnology transfer (DS SAT
v.21). Univ. of Hawaii, Honolulu, HI.

International Fertilizer Development Center (IFDC). 1992. Status report for the phosphorous
submodel. Working Paper #1. IFDC, Muscle Shoals, AL.

International Fertilizer Development Center (IFDC). 1994. Networking to meet nutrient
challenges. IFDC 1993 annual report. Muscle Shoals, AL.

Israel, D.W. 1987. Investigation of the role of phosphorous in symbiotic dinitrogen fixation.
Plant Physiol. 84:835-840.

Israel, D.W., and T.W. Rutfy. 1988. Influence of phosphorous nutrition on phosphorous and
nitrogen utilization efficiencies and associated physiological responses in soybean.
Crop Sci. 28:954-960.

Jagtap, 8.8, M. Morrru, and B.T. Kang. 1993. Simulation of growth, development, and yield
for maize in the transition zone of Nigeria. Agri. Syst. 41 :21 5-229.

Jones, AC., C.V. Cole, AN. Sharpley, and J R “Williams. 1984. A simplified soil and plant
phosphorous model: 1. Documentation. Soil Sci. Soc. Am. J. 48:800-805.

82

Jones, J.W., KJ. Boote, S.S. Jagtap, and J.W. Mishoe. 1991. Soybean development. 1;: JR
Hanks and J.T. Ritchie (eds) Modeling plant and soil systems. Agronomy 31 :71-90.

Jones, J.W., W.T. Bowen, W.G. Boggess, and J .T. Ritchie. 1993. Decision support systems
for sustainable agriculture. p. 123-138. A J. Ragland and R Lal. (eds) Technologies
for sustainable agriculture in the tropics. Proc. of an International Symposia. ASA
Spec. Publ. 56. ASA, CSSA, SSSA Madison, WI.

Jones, J .W., J.W. Mishoe, and KJ. Boote. 1987. Introduction to simulation and modeling.
Tech. Bull. No. 100. Food and Fert. Tech. Center. Taipei, Taiwan.

Jones, J.W., and J.T. Ritchie. 1990. Crop growth models. p. 63-89. I_n G.J. Hofl‘rnan, T.A
Howell, and K.H. Solomon (eds) Management of irrigation systems. ASAE. St.

Joseph, MI.

Jones, A C., AN. Sharpley, and JR Williams. 1991. Modeling phosphorous dynamics in
the soil-plant system. I_n JR Hanks and J .T. Ritchie (eds) Modeling plant and soil
systems. Agronomy 31:323-339.

Juo, ASR 1979. Determination of total N in plant materials (Micro-Kjeldahl Method). p.
47-48. I_n ASR. Juo (ed) Selected methods of soil and plant analysis manual series
No. 1. IITA Ibadan, Nigeria.

Keating, B.A, D.C. Godwin, and J.M. Watiki. 1991. Optimizing nitrogen inputs in response
to climatic risk p. 329-358.13 R. Muchow and J. Bellary (eds) Climatic risk in crop
production: Models and management for semi-arid tropics and sub-tropics. Proc. Int].
Symposium, St. Lucia, Brisbane, Queensland, Australia. July 2-6, 1990. CAB.
International. Wallingford, U.K.

Keating, B.A, RL. McCown, and BM. Wafula. 1992. Adjustment of nitrogen inputs in
response to a seasonal forecast in a region of high climatic risk. p. 235-251. I_n F .W.T.
Penning de Vries, P. Teng, and K. Metselaar (eds) Systems approaches for
agricultural development. Kluwer Academic Publ. Dordrecht, The Netherlands.

Keating BA, and BM. Wafula. 1992. Modeling the fully expanded area of maize leaves.
Field Crops Res. 29: 163-176.

Lal, H., G. Hoogenboom, J.P. Calixte, J.W. Jones, and RH. Beinroth. 1993. Using crop
simulation models and GIS for regional productivity analysis. Transactions ASAE.
36:175-184.

83

Lauer, M.J., S.G. Pallardy, D.G. Blevins, and DD. Randall. 1989. Whole leaf carbon
exchange characteristics of phosphate-deficient soybeans. (Glycine max L.) Plant
Physiol. 64:411-416.

Le Bot, 1., DJ. Pilbearn, and EA Kirkby. 1994. Plant mineral nutrition. p. 33-72. I_n AS.
Basra (ed) Mechanisms of plant grth and improved productivity: Modern
approaches. Marcel Dekker, Inc. New York, NY.

Mbabaliye, T., and PA Wojtkowski. 1994. Problems and perspectives on the use of a crop
simulation model in an Afiican research. Expl. Agic. 30:441-446.

McCown, RL., B.M. Wafirla, L. Moharnmed, J.G. Ryan, and J.N.G. Hargreaves. 1991.
Assessing the value of a seasonal rainfall predictor to agronomic decisions: The case
of response farming in Kenya. p. 383-409. 171 R. Muchow and J. Bellamy (eds)
Climatic risk in crop production: Models and management for semi-arid tropics and
sub-tropies. Proc. Intl.. Symposium, St. Lucia, Brisbane, Queensland, Australia. July
2-6, 1990. CAB. International, Wallingford, U.K

Muchena, P. 1994. Experiences of developing countries in using crop modeling predictions.
Ln P.F. Uhlir and GO Carter (eds) Crop modeling and related environmental data:
A focus on applications for arid and semi-arid regions in developing countries. 1:125-
131. CODATA, Paris, France.

Nuland, D.S., and HF. Schwartz. 1989. A visual description of the common bean plant four
major grth periods. Annual report of Bean Improvement Coop. 32:16-17.

Qui, J ., and D.W. Israel. 1992. Diurnal starch accumulation and utilization in phosphorous-
deficient soybean plants. Plant Physiol. 98:316-323.

Rao, IM, and N. Terry 1989. Leafphosphate status, photosynthesis and carbon partitioning
in sugar beet: 1. Changes in growth, gas exchange, and Calvin cycle enzymes. Plant
Physiol. 90:814-819.

Ratlifl', L.F, J .T. Ritchie, and D.K. Cassel. 1983. A survey of field-measured limits of soil
water availability as related to laboratory-measured properties. Soil Sci. Soc. of Am.
J 47:770-775.

Rawls, WJ. 1983. Estimating soil bulk density from particle size analysis and organic matter
content. Soil Sci. 135:123-125.

84

Ritchie, J.T. 1986. Using computerized crop models for management decisions. p. 27-41. In
Proc. International DLG-Congress for Computer Technology. May 1986. Hannover,
Fed. Rep. of Germany.

Ritchie, J .T. 1991. Specifications of the ideal model for predicting crop yields. p. 97-122.
RC. Muchow, and J.A Bellamy (eds) I_n Climatic risk in crop production: Models
and management for the semiarid tropics and subtropics. Proc. Intl.. Symposium, St.
Lucia, Brisbane, Queensland, Australia. July 2-6, 1990. CAB. International,
Wallingford, UK.

Ritchie, J .T. 1994. Classification of crop simulation models. I_n P.F. Uhlir, and G.C. Carter
(eds) Crop modeling and related environmental data: A focus on applications for arid
and semiarid regions in developing countries. 1:3-14 CODATA, Paris, France.

Ritchie, J .T. and J.B. Dent. 1994. Data requirements for agricultural systems research and
applications. p. 153-166. in P. Goldworthy and F.W.T. Penning de Vries (eds)
Opportunities, use, and transfer of system research methods in agriculture to
developing countries. Kluwer Academic Publ. Dordrecht, The Netherlands.

Ritchie, J.T., D.C. Godwin, and U. Singh. 1990. Soil and weather inputs for IBSNAT crop
models. p. 31-45. I_n Proc. of IBSNAT symposium: Decision support system for
agrotechnology transfer. Part 1: Symposium proceedings. Las Vegas, NV. 16-18
October, 1989. Univ. of Honolulu, HI.

Rosenzweig, C., and ML. Parry. 1994. Potential impact of climate change on world food
supply. Nature. 367: 133-138.

Sa, T-M., and D.W. Israel. 1995. Nitrogen assimilation in nitrogen-fixing soybean plants
during phosphorous deficiency. Crop Sci. 35:814-820.

Saka, AR, J.D.T. Kumwenda, P.K. Thornton, U. Singh, and J.B. Dent. 1990. Modelling of
maize growth and development in Malawi. p. 39-43. I_n, M.E. Probert (ed) A search
for strategies for sustainable dryland cropping in semi-arid Eastern Kenya. Australian
Center for International Agricultural Research (ACIAR). Canberra, Australia.

Sexton, P.J., J .W. White, and KJ. Boote. 1994. Yield-determining processes in relation to
cultivar seed size of common bean. Crop Sci. 34:84-91.

Sinclair, TR, and CT. de Wit. 1976. Analysis of carbon and nitrogen limitations to soybean
yield. Agron. J. 68:319-324.

85

Singh, U. 1989. Introduction to modeling. Report for training program on computer
simulation for crop growth and fertilizer responses. May 15-26, 1989. IFDC, Muscle
Shoals, AL.

Singh, U., V. Chinene, P.C. Ching, H. Ikawa, C.A Jones, and G. Uehara. 1985. Simulation
of maize response to nitrogen application. p. 151-158. _I_r_r J.A Silva (ed) Soil-based
agrotechnology transfer benchmark soils project. Univ. of Hawaii, Honolulu, HI.

Singh, U., and PK Thomton 1992. Using crop models for sustainability and environmental
quality assessment. Outlook on Agri. 21:209-218.

Singh, U., P.K. Thornton, AR Saka, and J.B. Dent. 1993. Maize modeling in Malawi: A
tool for soil fertility research and development, p. 254-273. Ln F.W.T. Penning de
Vries, P. Teng, and K. Metselaar (eds) Systems approaches for agricultural
development. Kluwer Academic Publ., Dordrecht, The Netherlands.

Smithson, J .B., and W. Grisley. 1992. First African bean yield and adaptation nursery: Part
11. Performance across environments. Network on Bean Research in Afiica,
Occasional Publ. Series No. 3B. CIAT, Dar-es-Salaarn, Tanzania.

Thornton, PK. 1991. Application of crop simulation models in agricultural research and
development in the tropics and subtropics. Paper Series P-15. IFDC, Muscle Shoals,
AL.

Thornton, P.K., J.B. Dent, and Z. Basci. 1991. A framework for crop grth simulation
model applications. Agri. Syst. 37:327-320.

Thornton, PK, and G. Hoogenboom. 1994. A computer program to analyze single-season
crop model output. Agron. J. 86:860-868.

Thornton, P.K., G. Hoogenboonr, P.W. Wilkens, and WT. Bowen. 19958. A computer
program to analyze multiple-season crop model outputs. Agron. J 87: 13 1-136.

Thornton, P.K., AR. Saka, U. Singh, J .D.T. Kumwenda, J .E. Brinlg and J .B. Dent. 1995b.
Application of a maize crop simulation model in the Central Region of Malawi. Expl.
Agric. 31:213-226.

Thung, M. 1991. Bean agronomy in monoculture. p. 373-834. In A van Schoonhoven and
O. Voyset (eds) Common beans: Research for crop improvement. CAB.
International, Wallingford, U.K

86

Tsuji, G.Y., G. Uehara, and S. Balas. 1994. A decision support system for agrotechnology
transfer (DSSAT v.3) International Benchmark Sites Network for Agrotechnology
Transfer. Univ. of Hawaii, Honolulu, HI.

Walkley, A, and CA Black 1965. Determination of organic carbon. p. 1372-1375. 13 CA
Black, D.D. Evans, J.H. White, L.E. Eusminger, and FE. Clark (eds) Methods of
analysis. ASA, CSSA, SSSA, Madison, WI.

Whisler, F.D., B. Acock, D.N. Baker, RE. Fye, H.F. Hodges, JR Lambert, H.E. Lemmon,
J.M. McKinion, and VR Reddy. 1986. Crop simulation models in agronomic
systems. Adv. Agron. 40:141-208.

White, J.W., G. Hoogenboom, J.W. Jones, and KJ. Boote. 1995. Evaluation of the dry bean
model BEAN GRO V1.01 for crop production research in a tropical environment.
Expl. Agric. 31:241-254.

White, J .W., and J. Izquierdo. 1991. Physiology of yield potential and stress tolerance. p.
287882.14 A van Schoonhoven and O. Voysest (eds) Common beans: Research for
crop improvement. CAB. International. Wallingford, UK.

White, J.W., S.P. Singh, C. Pino, B. Rios, and I. Buddenhagen. 1992. Efi‘ect of seed and
photoperiod response on crop grth and yield of common bean. Field Crops Res.
28:295-307.

Whiteaker, G., G.C. Gerlofl‘, WH Gabelrnan, and D. Lindren. 1976. Interspecific differences
in grth of beans at stress levels of phosphorous. J Amer. Hort. 101:472-475.

Yan, X, J .P. Lynch, and SE. Beebe. 1995a. Genetic variation for phosphorous efficiency of
common bean in contrasting soil types: I. Vegetative response. Crop Sci. 35:1086-
1093.

Yan, X, S.E. Beebe, and JP. Lynch. 1995b. Genetic variation for phosphorous efliciency of
common bean in contrasting soil types: II. Yield response. Crop Sci. 35: 1094-1099.

CHAPTER 3
SIMULATION OF PLANT AND SOIL PHOSPHOROUS DYNAMICS IN BEAN

CROP GROWN UNDER RAINFED AND IRRIGATED CONDITIONS

ABSTRACT

Knowledge of nutrient uptake, distribution, and accumulation in a plant is important
for basic understanding of its nutrition. The objectives of this study were to evaluate the
efl‘ects of P fertilintion on uptake, partitioning, and accumulation pattern of P throughout the
growth cycle of bean crop grown on acid soils. Also the study results were to be used in
evaluating the P version BEAN GRO model in prediction of the plant P concentration (%),
and accumulation (uptake) at various growth stage by comparing with those measured under
field conditions. Difi‘erent levels of triple superphosphate (TSP), and Minjingu phosphate
rock (MPR) fertilizers were applied to the bean cr0p grown under rainfed and irrigated field
conditions during the 1993 and 1994 growing seasons. Sampling for determining dry
weights, and P concentration was done at V,, R,, R,, R7, and R, growth stages. Phosphorous
concentration was analyzed in stems, leaves, pods, seeds, and shells. Increased P application
resulted in slightly increased P concentration in all bean plant parts, and varied between

CXperiments, treatments, plant parts, and crop age. MPR was not as effective as TSP in

87

88
supplying P to the bean crop. Large concentrations of applied P ranging from 0.27 to 0.47%
were allocated to seeds. High levels of P did not result in Zn or Cu accumulation reduction.
The critical P concentrations on dry weight basis at flowering in leaves were 0.23% under
rainfed, and 0.35% under irrigated conditions.

The model predicted quite well the P accumulation in plots that received TSP under
irrigated conditions. The current model over predicted the P uptake from plots applied with
MPR fertilizer. The model was found to be very sensitive to the soil initial P conditions, and
model parameters such as RCONST (i.e., rate of movement between active P and stable P
pools), FPO, (i.e., P fraction used to calculate the plant P uptake, kg P ha") and RPO,U (i.e.,
the plant potential uptake from the soil). With very low initial soil P, the crop did not absorb

P with the application of low and medium P fertilizer.

LITERATURE REVIEW

In agriculture the phosphorous (P) problem is threefold. First, the total P level of soil
is low, about 200 to 2000 kg P per hectare-firrrow slice (HFS) with an average ofabout1000
kgP HFS“. Second, the native P compounds are mostly unavailable for plant uptake, some
being highly insoluble. Third, when soluble sources of P such as those in fertilizer, and
manure are added to soils, they are fixed or are changed to unavailable forms (Brady, 1990).
Since the amount of plant available P at any given time is low (about 0.01% of total P) in the
soil, to obtain high yields, farmers commonly apply more P in fertilizers than is required by

crops.

89

Initially only the P in the solution can be taken up by the plant. However, as this is
removed, it is replenished by more P coming into solution from the labile pool of P. When
in the soil solution phosphorous is primarily in the form of orthophosphate ions, (H,PO,' and
HPOX'). It is absorbed generally as the monovalent ion, (H PO), and less as the divalent
ion, (HPOh, which is more prevalent of the two at neutral pH or above. However, the rate
at which HzPO; is converted to I-IPO," in solution is so rapid that plants have little difficulty
obtaining the necessary P for growth, even in soils with pH of 8 or higher (Foth and Ellis,
1988). Insolubility of P occurs at both extremes of the soil pH range (Lindsay and Moreno,
1960). Maximum P availability to plants is obtained when the soil pH is maintained in the
range of 6.0 to 7.0.

Phosphorous firnctions in all biological processes, in high energy compounds and in
mechanisms of energy transfer. Phosphorous in crop plants is as a component of important
structures such as phospholipids in membranes, phosphorylated sugars and proteins, and as
an integral part of nucleotides such as DNA (deoxyribonucleic acid), and RNA (ribonucleic
acid) which carry the inheritance characteristics of living organisms. Phosphorous is also a
component of energy transfer molecules such as adenosine diphosphate (ADP), and adenosine
5'-triphosphate (ATP); nicotinarnide adenine dinucleotide (NAD); reduced, nicotinamide
adenine dinucleotide diphosphate (NADPH) (Bieleski and Ferguson, 1983), and
phosphoenolpyruvate (PEP).

Phosphorous is also important in C metabolism in plants. A first step in C fixation is
the capture of light energy by the leaf. Lauer at al. (1989) observed that leaves of soybean
that were deficient or low in P showed paraheliotropisrn (i.e., avoided sunlight by keeping the

leaf edge toward the sun during the daylight. However, plants with adequate P levels

90
exhibited diaheliotnopisrn (i.e., tracked the sun during the daylight hours). They also absorbed
that carboxylation efficiency, and other parameters of photosynthesis decreased in low P
plants. Other roles of P in crop plants are N fixation; crop maturation (flowering and fruiting,
including seed formation); and root development.

Several factors regulate the P uptake in plants under field conditions. These include
soil and climate factors, fertilizer type, crop effects, and management practices. Factors such
as extensive root system, volume the soil in contact with the roots, soil moisture, aeration,
temperature, physical and chemical properties of the soil are very important for nutrient
uptake. Evidence fiom experimental results shows that P absorption takes place freely at all
locations. of the root surface (Clarkson and Hanson, 1980). As plant roots push their way
through the soil they come in contact with the P in solution. Since the roots of the growing
plants have a high demand for P, phosphate is absorbed by the roots at a high rate, and the
soil solution in the direct root vicinity is depleted of P. This depletion creates a gradient
between the P concentration near the root surface, and the P concentration in the soil (Mengel
and Kirkby, 1987). Therefore, diffusion is important for P uptake because of its low
concentration in the soil solution. It has been estimated that the distance of diffusion per day
through soil at field capacity to root is 0.004 cm for the phosphate ion, HZPO; (F oth and
Ellis, 1988). Large genetic variation for P uptake has been reported in tropical bean
genotypes. Andean origin (large-seeded) germ plasrn such as Canadian Wonder cultivar
appears to have superior P efliciency under low P availability, while Mesoarnerican genotypes
are more responsive to applied P (Yan et al. 1995a, b).

There is still some controversy about whether HzPO; enters the cell via symporters

or is driven by primary pumps. According to Ullrich-Eberius er al. (1981), Tu er al. (1990),

91

and Mimura (1995); P is taken up as H,PO,' by means of an active mechanism with H -
ATPase cotransport or HCO,‘ arrtiport properties. Since P is a multi-valent moiety, its charge
changes with pH. In the value where pH is around 5, P is present as H2PO ,' and in the
cytoplasm where pH is around 7.5, P is present as HPOKZ Which ionic species of P is
translocated across the tonoplast Mmura, 1995) is not known. Once HZPO; is absorbed by
the roots it is carried to the shoots by the symplastic pathway. Within a short time of uptake,
the P absorbed is incorporated into organic compounds, mainly hexose phosphates and
uridine diphosphate (Jackson and Hagen, 1960).

Under P stress, plants are capable of morphological, and physiological adaptations
during the development that may substantially improve P acquisition. These adaptations
include changes in P, and dry matter partitioning that favor grth of roots over shoots, and
the induction of a high-affinity P uptake, and transport system in roots (Ogata et al., 1988).
Plants can also excrete acid phosphatase by roots to enhance acidification of the rhizosphere
by plasma membrane-associated ATP-ase which may increase the dissolution of P-containing
compounds, thus improving P uptake from the soil (Clarkson, 1985; Goldstein et al., 1988
a, b). Increase in root formation, extension and root hair formation does also enhance the P
uptake (Clarkson, 1985).

Research work has shown that availability of P from soil and applied P can be
increased by inoculation with bacteria. Microorganisms such as Escherichia coli have a P
scavenging system known as pho regulon. This is the system of genes whose expression are
co-regulated bythe P status in the organism. These genes are activated when the P becomes
limiting, and encode proteins which include phosphatases which are P transporter (Torriani-

Gorrini, 1987). In plants there is a similar system to pho regulon (Goldstein et al., 1989).

92

Phosphorous transport into roots, and cells of plants such as Brassica spp (Lefebvre
er al., 1990), maize (Lee et al., 1990), and water fern (Bieleski and Lauchli, 1992), is
enhanced by P deficiency, and phosphatases. Therefore, under P deficiency there is increase
in acid-phosphatase activity in various parts of the plant (Bieleski, 1973 ), an increase in
secretion of acid phosphatases fi'om roots (Goldstein et al., 1988 a, b). More recent studies
have shown that phol (Poirier et al., 1991), and phosZ (Delhaize and Randall, 1995) in
mutations of Arabidopsrls thalirrm genes are involved in regulating the uptake of P in plants.

The effect of vesicular-arbuscular mycorrhizal fungi (V AMF ) on plant P uptake has
been extensively studied and reviewed (Kucey, 1987; Kucey er al., 1989; O'keefe and Sylvia,
1991; McArthur and Knowles, 1993a, b; Khalil er al., 1994). From greenhouse and field
study results, it appears that the major action of mycorrhizal firngi in facilitating plant P
uptake is to increase the absorptive surface area of the mycorrhizal root system, and to extend
the P depletion zone away fi'om the root surface (Kucey, 1987; Kucey et al., 1987). Working
with soybean, Karunaratne er al. (1986) observed that P efilux at low P solution
concentration was lower for mycorrhizal than nonmycorrhizal roots. They concluded that
mycorrhizal roots can remove P from soil solutions having lower P concentrations than
nonmycorrhizal roots. This was mainly due to the mycorrhizal roots having lower Cd, value
(the solution concentration of an ion at which emux equals influx) as shown by the Silberbush
and Barber model (Silberbush and Barber, 1983). The concentration of P in plant tissue
ranges fiom 0.15% to 1% of the dry weight of most crops with variation depending on crop
species, age, plant parts, part position on the plant, soil fertility, water regimes, seasonal
efi'ects, and crop management (Jones et al., 1991). Phosphorous concentration of less than

0.40% or 4000 ppm in foliar tissue of bean is regarded as being deficient, and 0.40-0.60% or

93

4000-6000 ppm as being adequate (Piggot, 1986), while greater than 1.00% is classified as
excess (Jones et al., 1991). A study conducted in East Afiica by Wortmann et al.(1992)
indicated that the previously recommended foliar critical nutrient level (CNL) of 0.25% for
P was too low to be usefirl in predicting responses to applied fertilizers for bean crops grown
in the test environments of that region. Predictions based on levels of 0.32% P were found
to be more accurate.

There are numerous investigations which report on methods of improving the P
availability, and uptake of P from Prs. These include the granule size and partial acidulation
of PRs (Khasawneh and Doll, 1978; Hammond er al., 1986; Chien and Hammond, 1988).
A mixture of PRs and TSP in P ratio of least 1:1 has also been reported to increase the P
uptake in maize (Chien er al. (1987), and in bean and rice (Menon er al., 1991). Recent
laboratory studies by Kpomblekou-A and Tabatabai (1994) have also shown that organic
acids such as citric and oxalic acids have a potential as amendments for releasing plant
available P and Ca fi'om PRs applied to soils.

In the P model used in this study was adapted from Jones et al., 1991 and Bowen,
1994, personal communication The plant absorbs P from the labile pool. Labile P is defined
as resin-extractable P, which is assumed to establish rapid equilibrium with an active mineral
P pool following the addition or uptake of P. Phosphorus uptake studies fi'om the soil are
complicated by the fact that P is highly immobile element, and transport through soil to root
surface is often a rate-limiting step. This problem is amplified by the dependence of the
difl'usion coeficient upon the soil type, P concentration, and soil moisture content (Amijee

er al., 1991).

94

The original model obtained fi'om the International Fertilizer Development Center
calculates the P value based on method of P extraction used (SMPX), the linear relationships
given by Sharpley (1985); Sharpley er al. (1984, 1989) and soil groups as defined in the
SOILPI subroutine. After the labile P pool (PRESIN) is defined, the active P pool (PACTIV)
is calculated as a firnction of PRESIN, and a fertilizer P availability index (FPAI). Operating
on a daily time step, the model calculates the changes in each of the P pools in the PCHEM
subroutine. Although there are studies reported on the test of P model performance under
field conditions in the USA using crops such as cotton, sorghum, wheat and maize (Jones et
al., 1984; Sharpley et al., 1989; Jones et al., 1991), there is no such study reported using the
bean as test crop.

Bean plants are known to respond to applied fertilizer P more than other nutrients
(Thung, 1991). Furthermore, PRs can be a good source of P (Chien and Hammond, 1988;
Sale and Mokwunye, 1993). In this study, it is suggested that the P simulation model, in
conjunction with measurements of initial P status and pertinent soil and bean crop constants,
may be a better tool for predicting P fertilizer requirements of bean crops than soil or plant
tissue testing.

Therefore, the objectives of this study were: (I) To evaluate the uptake, distribution,
and accumulation of P throughout the bean crop grth cycle, on acid soils applied with
difi‘erent levels of TSP, and MPR fertilizers under rainfed and irrigated field conditions; (ii)
To evaluate and improve the P version BEAN GRO model in predicting the plant P
concentration (%), and accumulation (uptake in kg ha") at various bean growth stages, and

different plant parts by comparing with those measured under field conditions. The

95

information obtained is usefirl in improving the P fertilizer use, estimates of P decreases fiom

the top soil, and consequently bean yield.

MATERIALS AND METHODS

Data were obtained from experiments conducted at Sokoine University of Agriculture
(SUA), in Morogoro, Tanzania. Details on both sites, and experimental management are
provided in Chapter 2. Plant materials used for tissue analysis were sampled fi'om all three
blocks. These included block A (used for assessing fiesh applied P in 1993, and for residual
P in 1994), block B (used for assessing annual applied P), and block C (used for assessing
fi'esh applied P in the 1994 season).

Plant samples fi'om stems, leaves (including petioles), pods, seeds, and shells were
collected during the biomass harvest times. Samples were collected at V,, R,, R,, R1, and R,
grth stages (Fehr er al., 1971; Nuland and Schwartz, 1989) descriptions. Sampling for
plant tissues was done fi'om five plants per plot at each sampling time. Sampling and
preparation for analysis were done as described by Jones et al. (1991). This was done by
cutting the plants at the base of the stem. The samples were dipped in water to remove all
soil, and separated into leaves, stems, and reproductive parts (pods) for N (for 1993 season)
and P determination (for both seasons). At maturity pods were shelled; seeds and shells were
also analyzed for P concentration Only the youngest, fully developed trifoliolate leaves, and
mature (firlly expanded) leaves fi'om the top, and middle part of the plants were sampled.
Leaf samples taken at flowering (K) were also used to determine the contents of P, K, Ca,

Mg, and difl‘erent micronutrients in the bean foliar tissues.

96

- The subsarnples were dried at 70°C, ground to pass through a 40-mesh sieve using a
WILEY MILL model 3383-L20. A ground plant tissue of 0.3 grams was used for wet acid
digestion using 3m] concentrated HISO, and 30% H202 as oxidants, and LiCl at 1000 ppm
for temperature elevation (Parkinson and Allen, 1975). After digestion, aliquots were used
for analysis. Concentrations of P in the same digested samples were determined by the
ammonium molybdate-ascobic acid method (Murphy and Riley, 1962).

Calcium and other nutrients’ concentrations were analyzed only for 1994 season.
Nitrogen and P were analyzed using the Skalar autoanalyzer at IFDC laboratories in Muscle
Shoals, Alabama for 1993 plant tissues. Phosphorous at V,, R,, R, , and R, grth stages
fi'om 1994 season was determined using methods described by Murphy and Riley (1962),
using the Lachat analyzer in Crop and Soil Sciences laboratories at Michigan State University,
East Lansing. The 1994 foliar tissues were analyzed for P, K, Ca, Mg, Fe, A1, Zn, Cu, and
Mn at flowering using the inductively coupled plasma emission spectrometry procedures at
Michigan State University.

Due to the time and the financial constraints, plant sampling was done for all
treatments, but not in all experimental plots. Therefore, statistical analyses for nutrients
analyzed in this study were not conducted. Phosphorous concentration in leaves, stems, pods,
seed, and shells were used with the leaf, stem, pod, seed, shell, and dry weight to calculate
the P total acmmulation. Nutrient concentration (i.e., nutrient content per unit dry matter),
and nutrient acammlation (i.e., nutrient concentration x dry matter per unit area) at difi‘erent
growth stages, and plant parts were determined as a firnction of source, and rate of fertilizer
used for the dates when samples were collected. Further, maximum and minimum P

concentration values determined for leaf, stem, seed, and shell were used for cultivar

97

coeflicient calibration in the species file (SPE) as described in Chapter 2. Data from annual
applied P plots were used for calculation of P accumulation (on per hectare basis). The
critical concentrations were estimated at 95% of the maximum yield after plotting the dry
weight at maturity against the P concentration in plant leaves at early flowering fi'om rainfed,

and irrigated experiments as described by Ogata er al. (1988).

Phosphorous Model Calibration and Modification

The model for phosphorous dynamics added to the BEANGRO model was an
adaptation of that presented by Jones et al. (1984) and Sharpley et al. (1984). Their model
structure was not changed. However, refitting some constants in the equations was necessary
for the model to best describe the results obtained from the field experiments conducted in
Tanzania Thus, some of the constants described below for the various subroutines will not

be the same as those described by Jones et al. (1984) and Sharpley et al. (1984).

Initifl Soil P (Subroutine SOILPI)

This subroutine is used to initialize the soil P pools. First, simulations were run and
calibration of the existing model in terms of P concentration (%), and total accumulation (kg
ha") in shoots and grain was made so that the model simulated bean response to soil P
accurately. Bray I (SMPX IB002) method was used for soil test to estimate labile P, on the

highly weathered soils (SGRP [3003) for the two sites used for this study.

98

After calibrating the model using the genetic coefficients (Chapter 2), the second step
was to do the calibration using the control treatment and the soil initial P conditions of the
1993 irrigated experiment.

Using the PRESIN = 0.14 ‘ BRAY] + 4.2 relationship provided by Sharpley er al.
(1984) for predicting the labile P (PRESIN which is the pool responsible for P absorbed by
the plant) in the model; labile P (kg P ha"), and the plant P uptake (kg P ha") were found to
be over predicted by the model. Therefore, a new relationship with a forced zero intercept
was developed (by Dr. Joe T. Ritchie) using the same data provided by Sharpley er al. (1984)
for highly weathered soil as follows: .

PRESIN = 0.203 *BRA 1’]

where

PRESIN = Labile P which is absorbed by the plant

BRAYI = Amount of P measured by BRAY I for the specific soil layer

After the labile P pool was defined, active P pool (PACTIV) was calculated as a
function of PRESIN and fertilizer P availability index (FPAI). FPAI represents the fraction
of fertilizer P remaining as labile 1’ after incubating fertilizer P and soil for six months (g
extracted / g P added to soil).

For the highly weathered soil used in this study, FPAI was calculated as follows
(Jones et al., 1984):

FPAI = (-0. 047 * CLAY) + (0.0045 ‘PRESIN) - (0. 053 " CC) + 0.39

where, .

CLAY = Amount of clay (%)

0C = Organic carbon (%)

99
The FPAI value from this calculation for the soils in this study came out to be about 0.23.
After testing the model using FPAI = 0.23, results showed that the value needed to be much
smaller for the biomass and soil phosphorous measurements to agree with the simulations.
Using a value of FPAI = 0. 08 provided the best fit. This value is similar to ones reported in
Jones et al., (1984) for less acid soils. Thus, this value was used in all the model calculations
for this experiment.
After the FPAI had been calculated, it was used to determine the initial size of PACTIV as
follows:
PACITV = PRESIN * ((1.0 - FPAI) /FPAI)

The stable inorganic P pool (PSTABLE) was assumed to be four times as large as the

active P pool at equilibrium as described by Jones et al. (1984) and calculated as follows:

PSTABLE = PACHV * 4.0

Sofl Phosphorous Transformations (Subroutine PCHEM)
This subroutine calculates the rates of movement between different inorganic P pools,

then updates each pool based on the calculated values. The rate of movement (PRTOPA)
fiom PRESIN to PACTTV was calculated as follows: "

PROTOPA = 0.18 * (PRESIN - PACHV " (FPAI / (1.0 -FPAI)
Ifthe value of PROTOPA is negative, indicating flow from the active to the resin pool, the
slower movement between these pools is calculated as:

PROTOPA = PROTOPA * 0.1

100

Since PSTABLE was assumed to be four times as large as PACTIV at equilibrium,
the equation describing the rate of movement between the two pools (PATOPS) used was as
follows: 1

PA TOPS = RCONST * ((4.0 * PACHJO - PSTABLE)
where

RCONST = rate of P movement that varies with soil type

For highly weathered soils RCONST was calculated as a firnction of FPAI described
by Sharpley et al. (1984):

RCONST = EM” (-1. 77 * FPAI - 7.05)
Ifthe value of PATOPS is negative, indicating flow fi'om the stable to active pool, the slower
movement between these pools is calculated as:
PATOPS = PATOPS * 0.1

After the rate of P movement between pools had been determined, each pool was updated.

Wfiwmufin P TAK)
This determines the P uptake by plant. The phosphate fi'action (FPO) used for plant

uptake was calculated as follows:
FPO, = 1.0 - EX? {-0.5 * (PRESIN - 0.3))
RP04U = RFAC * FPO, * 0.02
where
RPO,U = potential P uptake from a soil layer (kg P ha“)
RFAC = interim variable describing the effects of water stress and root length density

on potential P uptake fi'om a soil layer calculated as:

101

RFAC = RLV "' MFR“ DLAYR
where
RLV = root length density for soil layer (cm root/cm3 soil)

SMDFR = the deficit factor affecting P uptake at low soil water (unitless) was

calculated as:
MFR = (SW-LL) * 10.00, LL > SW> LL + 0.10
MFR = 0, SW< LL
MFR=L SW>LL+0.10

where:

SW = soil water content

LL = soil water content at the lower limit

DLAYR = thickness of the soil layer (cm)
If TPS is added to the soil, the value of RPO,U as calculated above is multiplied times 2. This
is done to simulate the priming effect by the TPS to increase root activity and increase uptake

at the same P concentration.

Model Testing

After calibrating the P model, simulations were made and predicted results were
compared with field measured values. The model was tested for its performance using data
fi'om the control, 22, 44, and 88 kg P ha" treatments in form of TSP and using the control,

22, 88, and 176 kg P ha" fiom both experiments of the 1993 and 1994 cropping seasons.

102

RESULTS AND DISCUSSION

General Observations

Values for N concentration are shown in Table 3.1. Nitrogen concentration was
highest at V3 in the youngest leaves with the values of 2.29 under rainfed conditions, and
3.89% under irrigated conditions. The leaf N concentrations ranged from 2.90 under rainfed
experiment to 3.89% under irrigated conditions. These values are slightly lower than those
observed by Salema (1987) who reported the values ranging fi'om 3.56 to 4.60% from results
of an experiment conducted at the same location (SUA farm) using a similar cultivar. The
high leaf N values reported by Salema (1987), may have mainly been due to the application
of urea fertilizer at 80 kg N ha" at sowing, and the double strain inoculum of Rhizobium
leguminosamm biovar phaseoli applied to the seeds before sowing. The seed-rhizobial
inoculation may have been an additional factor to the leaf N concentration.

The pod N concentration was highest at R, (in the youngest pods) with values of
4.62% under rainfed, and 3.45% under irrigated experiment. The pod N concentration in this
study was also lower than that reported by Salema (1987). At R, the concentration started
to decrease. The decrease of pod N concentration with age observed in this study compared
favorably with the results by Westerrnann er al, (1985) for beans grown under greenhouse
conditions. Such results were due to the increase of the larger quantity of pod biomass which
resulted in dilution effect.

Seed N concentration was lower under rainfed conditions (3.07%) than under
irrigated conditions (3.75%). The seed N values reported here under irrigated conditions

agree favorably with those given by Thung (1991) for the bean crop grown under field

103

conditions. The shell N content were similar in both experiments, with values of 0.68%.

Table 3.1. Nitrogen content at five difi‘erent bean grth stages in the 1993 season.

 

 

 

Growth stage Nitrogen concentration (%)
imminent
Rainfed Irrigated

LeafN at V, 2.90 3.89
LeafN at R1 2.02 3.80
LeafN at R, 3.19 3.05
LeafN at R, 1.58 2.06
LeafN at R, 1.50 1.56
Pod N at R, 4.62 3.45
SeedNatR, 3.07 3.75
Shell N at R, 0.68 0.68

 

Phosphorous Concentration and Distribution in the Plant

Field measured concentration of P in different parts of the bean crop at different
growth stages are shown in Table 3.2. Results on P concentration fi'om both seasons show
that irrigated experiment had higher P concentration in all plant tissues at all stages of
sampling. These results are supported by Begg and Turner (1976) who reported that there
is more rapid increase in nutrient uptake rates than plant grth rates when adequate water
was supplied to crops. Such results may also have been due to lower temperatures

experienced during the growth of the irrigated experiment (Figure 2.1).

104

Table 3.2 Phosphorous concentration (%) in different parts of bean plants as influenced by
different fertilizer sources and levels of P application.

 

 

Treatment Crop Age Stems Leaves Pods Seed Shells

(DAP)

'nfed 'men

0 kg P ha’1 14 0.09 0.20 - - -

28 0.01 0.21 - - -

42 0.03 0.10 0.09 - -

56 0.06 0.06 0.10 - -

Maturity 0.02 0.08 0.09 0.21 0.04
22 kg P ha"TSP -

14 0.13 0.23 - - -

28 0.06 0.35 - - -

42 0.05 0.38 0.08 - -

56 0.03 0.05 0.12 - -

Maturity 0.02 0.04 0.11 0.25 0.02
22 kg P ha“MPR

14 0.13 0.17 - - -

28 0.13 0.28 - - -

42 0.05 0.02 0.08 - -

56 0.06 0.04 0.17 - -

Maturity 0.02 0.07 0.14 0.25 0.03
44 kg P ha“TSP

14 0.14 0.09 - - -

28 0.15 0.23 - - -

42 0.07 0.02 0.09 - -

56 0.06 0.03 0.17 - -

Maturity 0.02 0.05 0.18 0.34 0.07
88 kg P ha"TSP

14 0.16 0.32 - - -

28 0.05 0.28 - - -

42 0.09 0.13 0.12 - -

56 0.03 0.09 0.18 - -

Maturity 0.03 0.04 0.15 0.27 0.04
88 kg P ha-lMPR

14 0.14 0.14 - - -

28 0.15 0.22 - - -

42 0.04 0.03 0.10 - -

Table 3.2 cont.

56

Maturity
176 kg P ha-IMPR

14
28
42
56
Maturity

0 kg P ha‘1 14
28
42
56
Maturity
22 kg P ha"TSP
14
28
42
56
Maturity
22 kg P ha“MPR
. 14
28
42
56
Maturity
44 kg P ha"TSP
14
28
42
56
Maturity
88 kg P ha'lTSP
14
28
42
56
Maturity

105

0.02 0.04
0.02 0.03
0.34 0.09
0.13 0.25
0.05 0.06
0.06 0.07
0.13 0.02
Irrigotod goeriment
0.12 0.13
0.15 0.15
0.08 0.02
0.02 0.10
0.04 0.04
0.17 0.33
0.06 0.48
0.09 0.23
0.05 0.07
0.04 0.09
0.22 0.32
0.17 0.35
0.11 0.14
0.05 0.05
0.03 0.05
0.22 0.35
0.16 0.37
0.15 0.14
0.06 0.12
0.04 0.08
0.24 0.33
0.08 0.43
0.06 0.24
0.08 0.11
0.06 0.03

0.11
0.11

0.08
0.22
0.13

0.12
0.17
0.16

0.18
0.17
0.14

0.09
0.17
0.14

0.14
0.15
0.13

0.16
0.21
0.17

106

Table 3.2. (cont’d)

 

88 kg P ha"MPR
14 0.12 0.31 - - -
28 0.16 0.30 - - -
42 0.10 0.09 0.13 - -
56 0.07 0.07 0.18 - -
Maturity 0.04 0.06 o. 17 0.33 0.01
176 kgP ha"MPR
14 0.21 0.27 - - -
28 0.18 0.38 - - -
42 0.10 0.25 0.17 - -
56 0.03 0.09 0.20 - -
Maturity . 0.04 0.06 O. 12 0.24 0.01
DAP refers to days after planting
Maturity refers to harvest maturity.

Low temperatures are known to have significant effects on nutrient uptake, and
accumulation through plant growth (Saito and Kato 1994). These conditions, during crop
growth increase the immobility of nutrients such as P which results in concentration efl‘ect
(Jarrell and Beverly, 1981). The present results show that, P concentration increased between
V, and R1 growth stages; similar results have been reported in pigeon pea by Adu-Gyamfi et
a1. (1990). Such results may have been influenced by the characteristics of the P fiactions in
bean plant in response to P supply during the two growth stages. Phosphorous concentration
in bean tops decreased with crop age, but dry weight increased with P levels applied. Similar
observations have been reported by Adu-Gyamfi er al. (1990) in pigeon peas. Phosphorous

concentration was lowest in control plots and highest in plots that received TSP at 44 and 88

107 '

kg P ha". Using the scale suggested by Piggott (1986), at flowering (R,) stage, most plots
indicated inadequate levels of P in bean crop.

The results of high P concentrations found in young bean tissues, and their decrease
with age observed in this study agree with those reported in hardwood foliage by Lea er al.
(1979). Such decrease in concentration with age is referred to as dilution efl'ect (J arrell and
Beverly, 1981). Because of a lesser quantity of pod biomass at the same rate of P uptake,
resultsshowthattheP concentrationinpodsincreased asthe pods grew older. This was true
from R, up to R, growth stage. As expected, at R, grth stage, seeds began to develop in
pods, pod biomass increased, and the value in P concentration decreased due to the dilution
effects. As the seeds developed, P moved fiom the pod walls into the seeds where it is
usually stored in form of phytate (Lolas er al., 1976; Ogata et al., 1988).

The leaves and stems decrease in P concentration started at R, growth stage.
According to Mimura (1995), the concentration of P in the leaves is always higher than in
older leaves under P deficiency due to its re-translocation from older leaves to younger
developing tissues. Starting from R, growth stage, P was being remobilized from vegetative
plant parts to developing reproductive structures such as flowers, pods, and seeds as
suggested by Grabau et al. (1986); and Lauer and Blevins (1989) based on nutrient harvest
indices observed in their studies using soybean. Similar observations have been reported by
Adu-Gyamfi er a1. (.1990) using pigeon pea and soybean plants. In few instances plant tissues
from control plots had higher P concentration than those tissues from plots which received
fertilizers. These results could be due to the nutrient accumulation being higher than crop

growth rate during that time of sampling.

108

Seeds had the highest P concentration compared to other plant tissues. This shows
that the plants allocated the majority of their P reserves to seed development. Similar results
have been reported by Grabau er al. (1986) in soybean, and by Fageria (1989) and Yan et al.
(1995b) in beans. It is now known that in grains and seeds, P element is stored as phytate P,
which accounts for 50 to 70% of total P in grains and seeds (Lolas et al., 1976). Phytin, an
insoluble Ca-Mg salt of phytic acid and Zn, Fe or K salts of phytic acid are formed in
developing seeds during the period of rapid starch synthesis. During germination, phytase
breaks down phytate within a short period releasing P and cations for the formation of
membrane and other cellular constituents (Mukherji er al., 1971).

Accurnulations of K Ca, Mg, Cu, Mn, Zn analyzed at flowering are shown in Table
3.3. Increase in leaf 1K, Ca, and Mg due to fertilizer efi‘ect reported in this study is in
agreement with observations reported by Hallmark and Barber (1984) in soybeans, and those
by Fageria (1989) in beans grown on Oxisol soils in Brazil. Literature shows that high levels
of P application induces Zn deficiency in plants (Lonergan er al., 1982; Singh et al., 1988;
Fageria, 1989; Moraghan, 1994). According to Millikan (1963), high levels of P in plant
tissues inactivate the Zn requirements. However, in this study, high levels of P fertilizer did
not have a significant efl‘ect on Zn and Cu uptake. Such results may have been due to low

concentrations of P in plant tissues as revealed by the plant tissue analyses.

109

Table 3 .3. Nutrient concentration in bean plant leaves at early flowering as influenced by
difi‘erent sources and levels of annual applied P fertilizer in the 1994 season.

 

Rainf 'm n

K Ca Mg Cu Mn Zn

 

 

Treatment ---_-----% -------ppm

0 kg P ha'l 2.11 1.98 0.63 13 149 46
22 kg P ha"-TSP 3.92 2.24 0.59 18 302 51
22 kg P had-MPR 2.37 3.54 0.68 16 216 53
44 kg P ha“-TSP 2.28 2.45 0.50 11 195 36
88 kg P ha"-TSP 3.69 2.73 0.61 17 225 45
88 kg P ha“-MPR 2.19 2.56 0.60 14 148 38
176 kg P had-MPR 3.99 2.61 0.54 17 976 87

Irrigated exporiment

0 kg P ha‘l 2.96 1.65 0.54 22 113 46
22 kg P ha“-TSP 3.45 1.95 0.65 12 446 43
22 kg P ha"-MPR 3.38 2.10 0.74 19 613 53
44 kg P ha"-TSP 3.61 1.82 0.60 23 106 66
88 kg P ha"-TSP 3.36 2.01 0.68 13 109 41
88 kg P ha"-MPR 3.41 2.06 0.60 15 136 52
176 kg P ha"-MPR 3.81 1.84 0.57 24 690 62

 

Fe values were over-range by the method used for nutrient analysis.

Relationships between dry weight (g m") at maturity, and P concentrations of bean
leaves at early flowering are shown in Figure 3.1. Critical P concentration values were 0.23%
under rainfed, and 0.35% under irrigated conditions. The results on critical P concentration
reported from irrigated experimental conditions agree with those reported by CIAT (1977),

and Wortmann er al. (1992) for beans grown under East Afiican conditions.

110

 

 

 

 

 

 

 

g 1 r l r l r l
3000 , Rainfed __
t
A 2000 - _
a!
I
80
V
r: .. _
e 1000
r:
0
:1
E o
E 4000 - Irrigated r
o . t
fl
3 30m -‘ m. ._
E l .
b 2000 ~ —
G
1000 - -
.l
0 . . . , , , . , .
0.0 0.1 0.2 0.3 0.4 0.5

P concentration (%)

Figure 3.1. Relationship between dry matter production and phosphorous concentration at
maturity. Arrows indicate values of critical P concentration.

111

Total Phosphorous Accumulation and Removal From the Field

The amount of P absorbed on per hectare basis was in the order of stem <leaves and
<pods at R, At physiological maturity (R.), and at harvest maturity (R9) the order was stem
< shells <leaves <pods <seeds in both experiments. Phosphorous accumulation on per hectare
basiswastheproductoftheP concentrationinthebeanplantorplant part, and the dry matter
produced per unit area. Although at V3 growth stage the leaves had the highest P
concentration; due to low dry matter production, the total P accumulation was very low. The
total P accumulation was highest between R1 and R, growth stages. This was also the time
when dry matter production was also at maximum. Afler R, growth stage, there was a
decrease in P acaumrlation The decrease in P accumulation can be attributed to the decrease
in dry matter production, and the lower P concentration caused by diluting effect by the
increase dry matter produced (J arrell and Beverly, 1981).

In bean growing areas of Tanzania, such as the Kagera Region young tender leaves
known locally as ”0mushokolo”, and immature pods also known as ”Ebiendwa" in the area,
are usually harvested between R1 and R7 growth stages for vegetable consumption. This is
especially so during the time when the bean grains are in short supply. Such time is usually
afier planting the last lot which was reserved for seed. The immature pods harvested between
R, and R, are usually shelled, and immature seeds prepared for consumption.

At R, growth stage which is time for harvest, farmers harvest all aboveground parts
of bean crop. Usually all the harvested plant parts are taken out of the field, to the homestead
for threshing. Therefore, the nutrient taken out fi'om the field is all that which is contained

in the whole plant parts. Once at the homestead, the harvested crop is shelled, and the bean

112

straw is usually used to feed dairy animals. This practice is more important in areas around
Mt. Kilirmnjaro in the northern parts of Tanzania (Urio and Mlay, 1984; Shem et al., 1988).

Table 3.4 shows the quantity of P nutrient which is removed fi'om the field by
harvesting difi‘erent bean plant parts, at different growth stages. The results show that if the
crop is harvested as green pods i.e., between R, and R. growth stages, the P removal can be
as high as 8.7 kg P ha". However, if the crop is harvested at R9 growth stage, seed harvest
may result in removal of about 6 kg P ha" fi'om the field. Similar results were reported in
Ghana by Rhodes (1995) who observed that with the use of NPK fertilizer, P removal from
the fields was 27% by grain or tuber food crop harvest, and 42% was removed in crop
residues. The amount of P removed fiom the field will depend on the amount of fertilizer
applied, and on whether the crop is irrigated or rainfed as the present study results show.
Nutrient accumulation in relationship to the amount of P applied shows that harvesting of
bean crop at maturity, can result in 2 to 5% P removal from the field depending on the P

fertilizer source, rate applied and final crop yield.

113

Table 3.4 Amount of phosphorous removed from the field by bean harvest (kg P ha") .

 

 

1993 season 1994 season
m 1 ° w 'n 1)
1mm Im‘garg Rainfed _g_te.d.lrri a JRainf
0 kg P ha'l 1.10 0.81 0.59 0.98
22 kg P ha"-TSP 2.11 1.21 3.28 1.58

 

Evaluation of Simulated Phosphorous Concentration and Accumulation

Comparison between simulated and field measured P concentration (%) at flowering
stage is shown in Table 3.5. Comparison of these two sets of data shows that the simulated
values were always larger than field measured values. Daily simulation results (data not
shown) indicated that P concentration increased to about 0.50% at V3 growth stage. After
this growth stage the value dropped to about 0.05% depending on the experiment, treatment
and plant part being assessed. The maximum P concentration was at V3 growth stage while
the minimum concentration was at R, growth stage for all treatments. A fairly accurate
prediction of the grain P concentration in all experiments with TSP application and field
measured results for both seasons was obtained at harvest maturity.

Since precautions were taken to avoid any analytical discrepancies, the low values of

P measured under field conditions could be due to the fact that, although the crop received

114
morethanenoughfertilizer, thePwasnotreadilyavailableto the crop due to most ofit being
fixed by the soil. As discussed earlier, the other explanation could be that the crop diluted its
P to a greater extent under field conditions (Jarrell and Beverly, 1981).

Comparison between simulated and field measured P accumulation (kg ha") by the
bean shoots treated by TSP fertilizer are shown in Table 3.6, while those treated with MPR
fertilizer are shown in Table 3.7. In both experiments the control treatment had the lowest
amount of P accumulation.

The field measured amounts of P accumulation were lower than the model predicted
values at the flowering stage. However, at maturity harvest, model predicted values were
higher than field measured values for plots treated with TSP fertilizer (Table 3.6). For plots
that received P in form of MPR, the model predicted P accumulation was higher than the field

measured values except for the results fiom the 1993 irrigated experiment (Table 3 .7).

115

Table 3.5. Comparison between simulated and field measured vegetative P concentration (%)
at flowering stage.

 

1993 seamn 1994 mn
E rim n Experiment
Rainfed Irrigated Rainfed Irri at
SLMLEME SM EM. .SLM_ EM .S.__ 1M
1mm;
Control 0.12 0.27 0.14 0.29 0.23 0.16 0.16 0.15
22 kg P (TSP) 0.24 0.31 0.25 0.30 0.32 0.21 0.20 0.27
44 kg P (TSP) 0.28 0.30 0.26 0.34 0.32 0.19 0.22 0.23
88 kg P (TSP) 0.29 0.32 0.28 0.32 0.32 0.15 0.23 0.16
22 kg P (MPR) 0.21 0.33 0.20 0.29 0.30 0.23 0.16 0.27
88 kg P (MPR) 0.21 0.41 0.20 0.29 0.31 0.21 0.17 0.23

 

' SIM refers to simulated (model predicted) data
2 FM refers to field measured (observed) data

116

Table 3.6. Comparison of simulated and field measured phosphorous accumulation (kg P ha")
in bean shoots at two growth stages with the application of triple superphosphate fertilizer.

 

(gowth stage
R1. R91:

1993 growing mu

R‘ 1" R I
SM EM SM m SJM. F_M_ SL111 m

Treatment

Control 0.70 1.07 0.90 2.20 1.40 0.48 1.20 1.80
22 kg P ha" 1.80 2.56 2.20 4.71 3.20 1.05 2.90 6.21
44 kg P ha“ 2.00 3.52 2.50 3.41 4.20 1.48 4.30 7.74
88 kg P ha" 2.20 3.42 2.70 4.00 5.20 3.22 6.00 6.21

1994 growing season

Control 1.20 1.80 1.30 1.08 1.90 0.48 1.70 1.25
22 kg P ha" 1.60 2.60 1.90 2.75 3.20 1.22 3.00 1.75
44 kg P ha'l 1.60 3.60 2.20 2.69 4.20 1.02 4.20 2.25
88 kg P ha'l 1.60 3.00 2.30 2.24 5.20 2.10 5.40 2.25

 

‘ R, refers to flowering stage

" R, refers to harvest maturity stage

° R refers to rainfed experiment

" I refers to irrigated experiment

° SIM refers to simulated (model predicted) data
’ FM refers to field measured (observed data)

117

Table 3.7. Comparison between simulated and field measured phosphorous accumulation (kg
P ha") in bean shoots at two gowth stages with the application of Minjingu phosphate rock
fertilizer.

 

R1. R91:

R° I‘l R I
SM EM.’ SM m SIM m SL1! F_M_

Treatment

Control 0.70 1.07 0.90 2.20 1.40 0.48 1.20 1.80
22 kgP ha'l 1.40 3.50 1.60 4.00 3.20 3.50 3.50 5.40
88 kgP ha" 1.50 3.22 1.80 4.60 3.50 3.22 4.20 5.80
176 kg P ha" 1.50 4.25 1.80 3.40 3.50 4.75 4.20 4.90

1994 gowing season

Control 1.20 1.80 1.30 1.08 1.90 0.48 1.70 1.25
22 kg P ha" 1.50 3.20 1.40 3.60 2.40 1.20 3.10 1.25
88 kg P ha-l 1.50 2.50 1.60 3.10 2.50 1.40 3.50 1.60
176 kg P ha’l 1.50 2.10 1.60 4.20 2.50 1.50 3.50 1.70

 

' R, refers to flowering stage

" R, refers to harvest maturity stage

° R refers to rainfed experiment

" I refers to irrigated experiment

' SIM refers to simulated (model predicted) data
‘ FM refers to field measured (observed data) .

118

The model better predicted the P uptake from applied TSP than from the MPR fertilizer.
Such results may have been influenced by the fact that during the simulation process, the
MPR is first placed in the stable P pool, then P moves to the active P pool before it is
accounted for the plant P uptake in the labile pool. The three P pools interact directly or
indirectly and afl‘ect the P uptake by the crop and the efi‘ectiveness of fertilizer use. However,
the quantitative relationships of the processes involved are poorly understood and it is dificult
to apply to such comprehensive crop models (Wolf et al., 1987). The other speculation for
such results is that, the MPR characterization conducted at IFDC laboratories (Table 4.1)
indicated the Minjingu phosphate rock’s complexity of its chemistry due to its difl‘erent
degrees of reactivity. Such rock’s characteristics can lead to the discrepancies in results as
shown in this study.

The P pools calculated in the model for the 1993 irrigated experiment are shown in
Figure 3.2a for the TSP treatments and Figure 3 .2b for the MPR treatments along with the
simulated plant uptake. The results fi'om the TPS addition demonstrate how the stable pool
gradually increases during the season after the fertilizer was added. The active increased
rapidly after the fertilizer was added and then began to decrease when the resin pool was at
a low value. The resin pool has fertilizer added into it on the date of application. It
decreases rapidly, flowing into the active pool. The uptake was independent of the labile
concentrations early in the season when all the pools were relatively high, but the uptake
declined at difl‘erent times depending on the level of the labile pool in the root zone. The P
uptake patterns of the plants as measured in the experiment generally agree with the results

as well as the soil resin P values as estimated by the Bray test.

119

 

52 e Triple Super Phosphate
48 ..
44 _
40 e

Stable (mg/kg)

'v'vvvvvvvvvvv'v

 

60
.AA‘AAAAAA AAAAAAAAAAAAAAAAA
AAAAAA
AAAAA
AAAAAAAAAAAAAAA
4 AA AAAA
AAAA
0 A

Active (mg/kg)

20-

 

20-
16-
12*

Resin (mg/kg)

 

 

 

 

_ “A -AA“‘A--‘AA A‘

A- vvvvvvvvv'vvvv

 

Plant uptake kg/ha
O)
L

 

 

l 1 l I l j l T l

140 150 160 170 180 190 200 210 220 230
Day of the year 1993

Figure 3.2a The simulated distribution of the P soil pools and the P plant uptake during the
1993 irrigated experiments for various additions of TPS. The fertilizers were applied on Day
147.

120

 

\l
O
I

Rock Phosphate

O)
O
1

Stable (mg/kg)
01
O
1

Q
a
U
(I
H
lI
lI
I
I
I
I
I
I
I
I
l
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I

 

cub

'v
'V
'V
'v
Vv
"
Vvv
'v
'v'
"v
v
'v
7"
"
'vvvv'
'Vv
n 'vv
V"
7"
Vv'
V'vv
'77
'7'
'VV
"'

AAAAA
AAAAA
AAAAAAA
AAAAAA
AAAAAA
AAAAAALAA AAAAAAA
AAAAAAA
AAAAAAA
AAAAAAA
ALLA

Active (mg/kg)
a on re‘ a:
O O O O
l l l l

 

  

 

 

 

   
  

 

 

 

 

 

8 _ l l '" iv "1 " Iv " l “vvmifln "'1'
g 6 - ..................... ,,,,,,,,,,,,,,, .m.
.5, 4 -

.E ..............................
a 2 - AAAAAAAAAAAAAAAAAAAAAA
cc

0 _
as —.+ (“(3 r l 1 l I l
g 6 _ —-— 22 kg
f, -*— 88 kg
E 4 _ —'—' 176 kg
5‘ m
E 2 _ ;":w
32
0- o

l l l l l l l j l
140 150 160 170 180 190 200 210 220 230
Day of the year 1993

Figure 3.2b. The simulated distribution of the P soil pools and the P plant uptake during the
1993 irrigated experiments for various additions of MPR The fertilizers were applied on Day

147.

121

The MPR results indicate the difi‘erent reactions required for the rock phosphate
source to influence the resin pool and the plant uptake. The rock phosphate is added to the
active pool at the date of application and then begins to move to the resin pool and the stable
pool. The resin pool is considerably lower during the first part of the season as compared to
the TPS source because of the time required for the transfer from the active pool to the resin
pool. ThesoilPattheendofthe seasonandtheplantPuptake data showgeneral agreement

with the experimental findings.

SUMMARY AND CONCLUSIONS

Field measured data indicated that Minjingu phosphate rock (MPR) was not as
efl‘ective as TSP in supplying available P to the bean crop. Increased P application resulted
in increased P concentration all bean plant parts. Phosphorous concentration varied between
experiments, treatments, plant parts, and crop age. The absorption of P by the bean crop
continued until late in the season, so the applied P fertilizers resulted in dry matter increase
almost to the end of the season. Results show that plots which had the highest P
concentration did not necessarily give the highest crop yield. This means that although the
P concentration in leaf tissues is extensively used as an indicator of P deficiency, there should
be other biochemical events which take place in the plant to help the crop cope with the P
stress.

Thepresent studyresultssuggestthatfannerswhoharvestgreenpodsbetweenR, and
R, growth stages result in the highest removal of P nutrient from the field. Since the seed

contained 27 to 47% P at harvest maturity, it means that if the by-products (crop residues)

122

are reincorporated in the fields, a substantial amount of P can be recycled for the succeeding
crop. This may reduce the mining of plant nutrients from already impoverished soils of
Tanzania

The model predicted quite well the P accumulation in plots treated with TSP during
the 1993 season under rainfed and in the 1994 irrigated experiment. However, there were
disagreements between predicted and field measured P accumulation values fi'om plots that
were applied with the MPR fertilizer. The model was found to be very sensitive to the soil
initial P conditions, and model parameters such as RCONST (i.e., rate of movement between
active P and stable P pools), FPO4 (i.e., P fiaction used to calculate the P uptake) and RPO,U
(i.e., the plant potential uptake fi'om the soil).

The applied P fertilizers were fairly well used by the bean crop, for at maturity about
5% of the applied P had been taken up by the crop in plots that received high rates of P. Such
results are agree with those reported in the literature for beans by Thung (1991). The critical
P value obtained from irrigated conditions are considered to be appropriate for beans grown

at Sokoine University of Agriculture farm.

RECOMMENDATIONS FOR FUTURE RESEARCH

Current literature shows that the versicular-arbicular mycorrhizal fungi (V AMF ) can
increase the P uptake by bean plants (Kucey 1987, Kucey and Janzen 1987, and Kucey et al.,
1989). Therefore, investigations to been responses to VAMF infection on improved P uptake

should be initiated in Tanzania. Different bean cultivars have different P uptake efficiencies

123
(Fageria, 1989; Yan et al., 1995a, b). Therefore, studies on P acquisition mechanisms in bean
cultivars for higher P eficiency in Tanzania is recommended.

Efi‘orts should be made to improve the current P model so that accurate simulations
of P concentration, accumulation (uptake), and ultimately crop yield can be predicted.
Therefore, there is a need to conduct a more detailed studies to determine physical and
chemical characteristics of MPR and the soil properties which afi‘ect the rock’s dissolution.
This will enable researchers to improve the theoretical understanding of interactions between
plant parameters, soil factors and MPR fertilizer characteristics. Further, this may help in the

improvement of the existing P model.

BIBLIOGRAPHY

Adu—Gyamfi, J., K. Fujita, and S. Ogata. 1990. Phosphorous fiactions in relation to growth

in pigeon pea (Cry’anas cajcm (L) Millsp) at various levels of P supply. Soil Sci. Plant
Nutri. 36:531-543.

Amijee, F., P.B. Barraclough, and PB. Tinker. 1991. Modeling phosphorous uptake and
utilization by plants. p. 63-75. In C. Johansen, K.K. Lee, and KL. Saharawat (eds)
Phosphorous nutrition of grain legumes in the semi—arid tropics. ICRISAT,

Patancheru, India.
Begg, J.B., and NC. Tunner, 1976. Crop water deficit. Ad. Agron. 28: 161-217.

Bieleski, RL. 1973. Phosphate pools, phosphate transport, and phosphate availability. Ann.
Rev. Plant Physiol. 24:225-252.

Bieleski, RL., and LB. Ferguson. 1983. Inorganic plant nutrition. p. 422-449. In A Lauchli,
and RL. Bieleski (eds) Encyclopedia of plant physiology. Springer-Verlag, Berlin,
Germany.

Bieleski, R. L., and A Lauchli. 1992. Phosphate uptake, efflux and deficiency in water fern,
Azolla. Plant Cell Environ. 15:665-673.

Brady, NC. 1990. The nature and properties of soils. 10th ed. Macmillan Publ. Co. New
York, NY.

Chien, S.H., F. Adams, F.E. Khasawneh, and J. Henao. 1987. Efl‘ects of combinations of
triple superphosphate and a reactive phosphate rock on yield and phosphorous uptake
by corn. Soil Sci. Soc. Am. J. 51:1656-1658.

Chien, SH, and LL. Hammond. 1988. Agronomic evaluation of partially acidulated

phosphate rocks in the tropics: IFDC's Experience. Paper Series IFDC-P-7. IFDC,
Muscles Shoals, AL.

124

12S

CIAT (Centro de Internacional de Agricultura Tropical). 1977. CIAT annual report of the
bean production program 1976. CIAT, Cali, Colombia.

Clarkson, D.T. 1985. Factors affecting mineral nutrition acquisition by plants. Ann. Rev.
Plant Physiol. 36277-115.

Clarkson, D.T., and J.B. Hanson. 1980. The mineral nutrition of higher plants. Ann. Rev.
Plant Physiol. 31:239-298.

Delhaize, E., and PI. Randall. 1995. Characterization of phosphate-accumulator mutant of
Arabidopsis thaliana. Plant Physiol. 107:207-213.

Fageria, N.K. 1989. Efi‘ects of phosphorous on growth, yield and nutrient accumulation in the
common bean. Trop. Agric. 66:249-255.

Fehr, W.R., C.E. Caviness, D.T. Burrnood, and 1.8. Pennington. 1971. Stages of
development descriptions for Soybeans, Glycine max (L.) Merrill. Crop Sci. 11:929-
931.

Foth, H.D., and B.G. Ellis, 1988. Soil Fertility. John erey & Sons, Inc. New York, NY.

Goldstein, AH, D.A. Baertlein, and A Danon. 1989. Phosphate starvation stress as an
experimental system for molecular analysis. Plant Mol. Biol. Rep. 7:7-16.

Goldaein, AH, D.A Baertlein, and R.G McDanniel. 1988a. Phosphate starvation inducible
metabolism in Lycopersicon esculentum 1. Excretion of acid phosphotase by tomato
plants and suspension cultured cells. Plant Physiol. 87:711-714.

Goldstein, AH, D.A Baertlein, and RG McDanniel. 1988b. Phosphate starvation inducible
metabolism in Lycopersicon esculentum II. Characterization of the phosphate
starvation inducible-excreted acid phosphatease. Plant Physiol. 87:716-720.

Grabau, L.J., D.G. Blevis, and HC. Minor. 1986. P nutrition during seed development. Leaf
senescence, pod retention, and seed weight of soybean. Plant Physiol. 82:1008-1012.

Hallmark, W.B., and SA. Barber. 1984. Root growthand morphology of early grth of
soybeans as affected by soil P and K. Agron. J. 76:206-216.

Hammond, L.L., S.H Chien, and AU. Mokwunye, 1986. Agronomic value of unacidulated
and acidulated phosphate rocks indigenous to tropics. Add. Agron. 40:89-140.

126

Jackson, PC. and CE. Hagen. 1960. Products of orthophosphate absorption by barley roots.
Plant Physiol. 35:326-332.

Jarrell, WM. and RB. Beverly. 1981. The dilution efi‘ect in plant nutrition studies. Add.
Agron. 34:197-224.

Jones, J., J. Benton, B. Wolf, and HA Mills. 1991. Plant Analysis Handbook. A Plant
Analysis and Interpretation Guide. Micro-Macro Publishing, Athens, GA

Jones, C.A, C.V. Cole, AN. Sharpley, and J .R erliams. 1984. A simplified soil and plant
phosphorous model: 1. Documentation. Soil Sci. Soc. Am. J. 48:800-805.

Jones, C.A., AN. Sharpley, and J .R Williams. 1991. Modeling phosphorous dynamics in the
soil-plant system In J. Hanks and J.T. Ritchie (eds) Modeling plant and soil systems.

Agronomy 31:323-339.

Karunaratne, RS., Baker, J.H., and AV. Baker. 1986. Phosphorous uptake by micorrhizal
roots of soybean. J. Plant Nutr. 9: 1303-13 13.

KhaliL 8., TE. Loynacharr, and MA Tabatabai. 1994. Mycorrhizal dependency and nutrient
uptake by improved corn and unimproved corn and soybean cultivars. Agron. J.

86:949-958.

Khasawneh, FE, and EC. Doll. 1978. The use of phosphate rock for direct application to
soils. Add. Agron. 30:159-206.

Kpomblekou, A, and MA Tabatabai. 1994. Effect of organic acids on release of
phosphorous fi'om phosphate rocks. Soil Sci. 158:442-453.

Kucey, RMN. 1987. Increasing phosphate uptake by wheat and field beans inoculated with
a phosphorous-solubilizing Penicilr'um bilajr' strain and with vesicular-arbuscular

mycorrhizal firngi. Appl. Environ. Microb. 53:2699-2703.

Kucey, RMN., and HH Janzen. 1987. Efi‘ects of VAM and reduced nutrient availability of
growth and phosphorous and micronutrient uptake of wheat and field beans under

greenhouse conditions. Plant and Soil. 104271-78.

Kucey, RM.N., H.H. Janzen, and ME. Leggett. 1989. Microbially mediated increases in
plant-available phosphorous. Ad. Agron. 42: 199-228.

127

Lauer, M.J., and B.G. Blevins. 1989. Dry matter accumulation and phosphate distribution of
soybean grown on varying levels of phosphate nutrition. J. Plant Nutr. 12: 1045-1060.

Lauer, M.J, S.G. Pallardy, D.G. Blevins, and DD. Randall. 1989. Whole leaf carbon
exchange characteristics of phosphate deficient soybeans (Glycine max L.). Plant
Physiol. 91:848-854.

Lea, R., W.C. Tierson, D.H. Bickelhaupt, and LA Leaf. 1979. Stand treatment and sampling
time hardwood foliage. I. Macro-element analysis. Plant and Soil. 51:515-533.

Lee, RB., RG. Ratchlifi‘e, and TE. Southon. 1990. 31P NMR Measurements of cytoplasmic
and vacuolar Pi content of mature maize Roots: Efi‘ects of phosphate deprivation and
reduction of influx with provision of orthophosphate. Physiol. Plant. 54:199-206.

Lefebvre, D.D., S.M.G. Duff, C.A Fife, C. Julian-Inalsingh, and WC. Plaxton. 1990.

Response to phosphate deprivation on Brass'ica nigra suspension cell. Plant Physiol.
93 : 504-5 1 l .

Lindsay, W.L., and EC. Moreno, 1960. Phosphate equilibria in soils. Soil Sci. Soc. Amer.
Proc. 24:177-182.

Lolas, G.M., N. Palarnidas, and P. Markakis. 1976. The phytic acid total phosphorous
relationship in barley, oats, soybeans, and wheat. Cereal Chem. 53:867-870.

Lonergan, J .F ., D.L. Grunes, R.M. Welch, E.A Aduayi, A. Fengah, V.A. Lazar, and BE.
Cary. 1982. Phosphorous accumulation and toxicity in leaves in relation to zinc
supply. Soil Sci. Soc. Am. J. 46:345-352.

McArthur, DA, and NR Knowles. 1993a. Influence of vesicular-arbuscular mycorrhizal

fungi on the response of potato to phosphorous deficiency. Plant Physiol. 101:147-
160.

McArthur, DA, and RN. Knowles. 1993b. Influence of species of vesicular-arbuscular
mycorrhizal firngi and phosphorous nutrition on growth, development and mineral
nutrition of potato (Solanum tuberosum 1..) Plant Physiol. 102:771-782.

Merrgel. K, and Kiddry, EA 1987. Principles of plant nutrition. 4th ed. International Potash
Institute, Beme, Switzerland.

128

Menon, R.G., S.H. Chien, and Abdel Nabi Gadalla. 1991. Phosphorous rocks compacted
with superphosphate vs. partially acidulated rocks for bean and rice. Soil Sci. Soc.

Am. J. 55:1480-1484.

Millikan, CR 1963. Efl‘ects of different levels of the zinc and phosphorous on the growth of
subterranean clover (Trifolium subterraneum L.). Aust. J. Agric. Res. 14: 180-205.

Mimura, T. 1995. Homoestasis and transport of inorganic phosphate in plants. Plant Cell
Physiol. 36:1-7.

Moraghan, J .T. 1994. Accumulation of zinc, phosphorous, and magnesium by navy bean
seed. J. Plant Nutr. 17:1111-1125.

Mukherji, 83., B. Dey, AK Paul, and SM. Sirear. 1971. Changes of phosphorous fiactions
and phytase activity of rice seeds during germination. Physiol. Plant. 25:94-97.

Murphy, J ., and JP. Riley. 1962. A modified single solution method for the determination of
phosphate in natural waters. Anal. Chem. 27:31-36.

Nuland, D.S., and HF. Schwartz. 1989. A visual description of common bean plant: Four
major growth periods. Ann. Rep. Bean Irnpr. Coop. 32:16-17.

Ogata, 8.1, J. Adu-Gyamfi, and K Fujita 1988. Efl‘ect of phosphorous and pH on dry matter
production, dinitrogen fixation and critical phosphorous concentration in pigeon pea
(Cajanas cajan (L) Millsp). Soil Sci. Plant Nutr. 34:55-64

O'keefe, D.M., and D.M. Sylvia. 1991. Mechanisms of vesicular-arbuscular mycorrhizal
plant-growth response. p. 35-53. In D.K. Arora, B. Rai, KG. Mukerji, and GR
Knudsen (eds) Handbook of applied mycology. Vol. 1. Soil and plants. Marcel
Deker, Inc. New York, NY.

Parkinson, J .A, and SE. Allen. 1975. A wet oxidation procedure suitable for determination
of nitrogen and mineral nutrients in biological material. Comm. Soil Sci. and Plant

Anal. 6:1-11.

Piggott, T.J. 1986. Vegetable crops. p. 148-189. In DJ. Reuter and J .B. Robinson (eds).
Plant analysis: An interpretation manual. Inkata Press, Melbourne, Australia.

Poirier, Y., S. Thomas, C. Somerville, and J. Schiefelbein. 1991. A mutant of Arabidopsis
deficient in xylem loading phosphate. Plant Physiol. 97: 1087-1093.

129

Rhodes, ER 1995. Nutrient depletion by food crops in Ghana and soil organic nitrogen
management. Agic. Syst. 48:101-118.

Saito, M., and T. Kato. 1994. Relationships between dry matter dry matter production,
nitrogen and phosphorous accumulation in soybean grown under cool weather
conditions. Soil Sci. Plant Nutr. 40:73-81.

Sale, P.W.G., and AU. Mokwunye. 1993. Use of phosphate rocks in the tropics. Fert. Res.
35:33-45.

Salema, RM 1987. Partitioning of plant nitrogen in common beans (Phaseon vulgaris L.).
Bean Res. 2: 195-206. Benedictine Publications, Ndanda, Peramiho, Tanzania.

Sharpley, AN. 1985. Phosphorous cycling in unfertilized and fertilized agricultural soils. Soil
Sci. Am. J. 49:905-911.

Slmpley, AN., C.A, C. Gray, and CV. Cole. 1984. A simplified soil and plant phosphorous
model: 11. Prediction of labile, organic, and sorbed phosphorous. Soil Sci. Soc. Am.

J. 48:805-809.

Sharpley, AN., U. Singh, G. Uehara, and J. Kimble. 1989. Modeling soil and plant
phosphorous dynamics in Calcareous and highly weathered soils. Soil Sci. Soc. Am.
J. 53:153-158.

Shem, MN, N.A Urio, and N. Mdoe. 1988. Bean straw as a potential feed for dairy cattle
on smallholder dairy farms. Bean Res. 3:116-121.

Silberbush, M., and SA. Barber. 1983. A sensitivity of simulated phosphorous uptake to
parameters used by a mechanistic-mathematical model. Plant and Soil. 74293-100.

Singh, J.P., RE. Kararnanos, and J .W.B. Stewart. 1988. The Mechanism of phosphorous-
irrdiced zinc deficiency in bean (Phaseolus vulgaris L.). Can. J. Soil Sci. 68:345-358.

Torriani-Gorrini, A 1987. The birth of the Pho regulon. p. 3-11. In A Torriani-Gorrini, F.G.
Rothmarr, S. Silver, A. Wright, E. Yagil (eds) Phosphate metabolism and cellular
regulation in rrricroorganisms. Am. Soc. Microbiol. Washington, DC.

Thung, M. 1991. Bean agronomy in monoculture. p. 737-834. In A van Schoonhoven and
O. Voysest (eds) Common beans: Research for crop improvement. CAB.
International, Wallingford, UK.

130

Tu, S., J .R. Cavanaugh, and RT. Boswell. 1990. Phosphate uptake by excised maize root
tips studies in Vivo l3P nuclear magnetic resource spectroscopy. Plant Physiol.
93:778-784.

Ullrich-Eberius, CL, A Novacky, E. Fischer, and U. Luttge. 1981. Relationship between
energy-dependent phosphate uptake and the electrical membrane potential in Lemma
gibba G1. Plant Physiol. 67:797-801.

Urio, N. A, and GI. Mlay. 1984. First progress report on diagnostic survey among
smallholder dairy farmers in Hai District-Tanzania, IDRC Project File No. 3-P-82-
0085. Tanzania.

Westermann, D.T., L.T. Porter, and W.A O'Deen. 1985. Nitrogen partitioning and
mobilization patterns in bean plants. Crop Sci. 25:225-229.

Wolf, J., C.T. de Wit, B.H Jansen, and DJ. Lathwell. 1987. Modeling long-term crop
response to fertilizer phosphorous. I. The model. Agron. J. 79:445-451.

Wortmann, C.S., J. Kisakye, and CT Edje. 1992. The diagnosis and recommendation
integrated system for dry bean: Determination and validation of norms. J. Plant Nutr.
15:2369-2379.

Yan, X, J .P. Lynch, and SE. Beebe. 1995a. Genetic variation for phosphorous efficiency of
common bean in contrasting soil types: I. Vegetative response. Crop Sci. 3521086-
1093.

Yan, X, JP. Lynch, and SE. Beebe. 1995b. Genetic variation for phosphorous efficiency of
common bean in contrasting soil types: 11. Yield response. Crop Sci. 35:1094-1099.

CHAPTER 4
AGRONOMIC EVALUATION OF MINJIN GU PHOSPHATE ROCK (MPR)
USING BEAN (Phaseolus vulgaris L.) CROP UNDER

RAINFED AND IRRIGATED CONDITIONS.
ABSTRACT

Phosphorous (P) is critical to improving soil fertility for sustainable agriculture in
Tanzania. This study was conducted to assess the growth and yield of a bean crop as
influenced by triple superphosphate (TSP) in comparison with Minjingu phosphate rock
(MPR) applied as direct applied fertilizer under rainfed and irrigated conditions. The MPR
reactivity was determined using the neutral ammonium citrate (N AC) extraction method.
Biomass harvests and growth parameters were determined at V,, R,, R,, R, ,and R, growth
stages. At maturity, total dry matter, seed yield, and yield components were analyzed.
laboratory analysis results with NAC indicated that MPR is highly a reactive phosphate rock
(PR) with the value of 6.7% P20s second extraction. However, the degree of C9
substitution for P0, in the apatite indicated the rock was only a medium to low reactive PR
Soil pH, extractable P, and exchangeable Ca increased in plots where P fertilizers were

applied regardless of the type, and rate of fertilizer used. These increases were higher in plots

131

132
that received higher rates of MPR The crop growth stages, and maturity periods depended
mainlyonthe cool nighttemperatures duringthe growth periods, and not on the P treatments.
Statistical analyses using Duncan Multiple Range Test (P=0.05) showed that TSP and MPR
had equal agronomic efi‘ectiveness Additional analysis using orthogonal comparisons
between TSP and MPR showed similar results under rainfed conditions. However, under
irrigated conditions of 12 comparisons, TSP outperformed MPR four times, and MPR
outperformed TSP five times. The relative agronomic efi‘ectiveness (RAE) of MPR with
respect to TSP (100%), in terms of increasing total dry matter, ranged from 20.66 to 64.59%
under fi'esh P, and 18.15 to 64.15% under annual P. The study also showed that MPR can
have liming efl‘ects when applied under high rates. The conclusion drawn from this study is
that apart fi'om the soil characteristics, such as low soil pH, low P, and low Ca, the use of

MPR should be recommended in areas where irrigation can be applied for bean production.
LITERATURE REVIEW

Phosphorous (P) is a key plant nutrient. It is mobile in plants, being redistributed from
older to younger parts, but non-mobile in the soil. It is derived fiom soil organic and
inorganic fiactions such as orthophosphate; and P-containing minerals such as the apatites of
Ca-, F e-, and Al-phosphates. Also, fiom the labile pool consisting of P adsorbed on soil
colloids and of Fe- and Al-phosphates in equilibrium with P in solution (Mengel and Kirkby,
1987). Phosphorous provides the best example of specifically adsorbed anions in negatively

charged soil particles.

133
Many soils fix large quantities of P by converting readily soluble P to forms less

available to plants. Phosphorous forms diflicult soluble compounds with Fe" and A1" at low

pH to give compounds such as strengite [Fe(OH),H,PO.] and variscite [Al(OH),H,PO.] (Juo

and Ellis, 1968). At higher pH values it forms difiicult soluble compounds with Ca2+ such as

octacalcium phosphate [Ca.H(PO,),; and hydroxyapatite [Ca m(PO ‘) (OH); (Lindsay and

Moreno, 1960).

In the field, the application of P to a bean crop results in increased plant growth and

seed yield (Manrique, 1986; Lynch et al., 1991; Yan et al., 1995 a and b), radiation use
efliciency (RUE) (Manrique, 1993), and pods per plant (Thung, 1991). The enhanced growth
of bean plants with P fertilization is brought about by a promotion of metabolic activity
(Whiteaker et al., 1976). Similar results are reported on soybean by Howell (1984), and
Lauer and Blevins (1989). Applied P has also shown some influence on leaf senescence in
soybean (Crafi-Brander, 1992a and b). Studies conducted at IFDC have shown that
application of P in the form of phosphate rock (PR) does increase the biological nitrogen
fixation (BNF) in soybean plants (Chien et al., 1993). Sa and Israel (1995) working with
soybean plants, observed that P deficiency impairs N fixation by decrease of hexose and
sucrose concentration in nodules, indicating low carbohydrate use. This low carbohydrate
use may result in the decreased movement of ureide and other N constituents in the plants.
Most of the bean-growing areas in the tropics normally have highly weathered acidic soils,
that are low in soil pH (less than 5.0). The soils also have low CEC which is usually less than
10 cmol (+) kg", low base saturation, and high aluminum saturation, often above 20%.

Although liming is an old agricultural practice, going back 2000 years in agriculture (Barber,

134
1967), it is still not commonly practiced in Tanzania. This is mainly because the resource-
poor farmers in Tanzania cannot afi‘ord such a high-input solution.

High "free" Al3+ in soil solution can produce phytotoxicity in beans because it is
absorbed rapidly and builds up a high concentration of Al in cells. This creates cell disorders
by afi‘ecting the membrane structure and function, DNA synthesis and mitosis, cell elongation
and metabolism (Wright, 1989). Bean cultivars such as ”Romano” and 'Dade" are resistant
to Al phytotoxicity. Studies have shown that the mechanism of Al-tolerance is by exudation
of citric acid into the rhizosphere (Mryasaka er al., 1991). Other legumes such as pigeon pea
[Cajanus cry'an (L) Millsp] have been shown to be more efiicient in using iron-bound (F e-P)
P than other crop species. This ability is also reported to be on the root exudates, in
particular piscidic acid and its p-O—methyl derivative, which releases P from F e-P by chelating
Fe” (Ae et al., 1990).

The reddish Oxisols and Ultisols occur in most parts ofTanzania. Such soils are poor
in organic matter, have low pH and crops grown on these soils show significant P deficiency
(Uriyo and Kesseba, 1973). To ameliorate P deficiency, fertilizer use is being encouraged by
the Tanzanian government. However, the relatively slow growth of fertilizer production in
the country compared with fertilizer use, has made the country increasingly dependent on
fertilizer imports. As pointed out by Bumb (1993 ), because of foreign exchange shortages
and debt crises, a significant proportion of the fertilizer imported is financed through aid.

Although TSP has been the most popular phosphate fertilizer in Tanzania (IFDC, 1990)
currently it has become too expensive for most farmers. This is because the only fertilizer
plant in the country ceased production in 1991 through defective machinery and lack of spare

parts. Thus, all required fertilizer has to be imported (Dr. Surjit Sidhu, IFDC principal

 

 

 

135

economist, 1995, personal communication). Due to these problems the government has been
encouraging the use of locally available PR fertilizers.

Important P deposits in Tannnia are summarized by Mchihiyo (1991), and Mwambete
(1991). There are three types of phosphate deposits found in Tanzania: guano, sedimentary,
and igneous deposits. Among these, the only significant deposit in the country currently being
mined is of the sedimentary type at Minjingu in the Arusha Region (Van Kauwenbergh,
1991). Its deposits are estimated to be 10 million metric tons (Hammond er al., 1986).
Samples of Minjingu ore, concentrate, and bird-bone concentrate have been characterized at
IFDC (Van Kauwenbergh, 1985; 1991).

The apatites that are found most often in commercial PRs are generally in the form
of carbonate fluorapatite (francolite). The francolite is defined as a carbonate containing more
than 1% F, and an appreciable amount of CO3 (Sandell et al., 1939). Based on statistical
models, McClellan and Lehr (1969) suggested that the empirical formula for francolite be
generalized as:

Carawwgraoa 400950,.
In MPR, apitite is the major phase ranging from 68 to 88%. The apatite of MPR is complex
and unusual (Van Kauwenbergh, 1991).

The key measure that indicates the speed at which the P fertilizer can increase the
supply of plant available P ions in the soil solution, in the short term, is its solubility in water.
Phosphate forms that are water soluble can be expected to release a large proportion of the
P immediately upon wetting, although a significant portion will remain in less soluble forms
as in the case of calcium phosphate fertilizers (Lindsay and Stephenson, 1959). An indication

of the rate at which a fertilizer can release P is usually given by its solubility in water and/or

136

NAC, and 2% formic acid (Mullins, 1988; Sale and Mokwunye, 1993; Mullins er al., 1995).

Reactive PRs are those in which some of the phosphate groups in the apatite crystal
structure has been substituted by a carbonate group during its formation. In general, the
greater the carbonate substitution, the smaller the crystal size, resulting in a more soluble PR
The most reactive PRs are those having a molar PO./CO, ratio less than 5.0 (Sale and
Mokwunye, 1993). Chien (unpublished report‘) outlines the efi‘ect of CO3 substitution for
PO4 on chemical reactivity of apatite as due to: (i) The decrease in unit cell a-axis that results
in a decrease in crystallite size of apatite, (ii) Substitution of planar CO,” for tetrahedral PO
, 3‘ causes structure instability that results in an increase in reactivity and, (Iii) The resulting
highly substituted fiancolites are less thermodynamically stable than the flourapatite.

According to the report by Van Kauwerrbergh (1991), MPR has a high NAC reactivity
ranging from 5.6 to 12.9% P20, However, Kpomblekou-A and Tabatabai (1994), reported
of MPR having medium NAC reactivity of 1.49-1.77% P,O,. Such contradicting results
indicate how unusual and complex MPR is.

There are a number of phosphate rock’s, environmental, management and plant
factors that influence the agronomic effectiveness of directly applied PRs (Sale and
Mokwunye, 1993; Allen et al., 1995). Phosphate rocks must be finely ground to obtain a
sufficiently rapid dissolution rate. Fine grinding increases the particle surface areas of PRs
that are exposed to the soil. This increases the potential dissolution rate (Wilson and Ellis,

1984; Sale, 1989; Le Mare, 1991), but fine grinding is not a substitute for reactivity.

 

48.1-1 Chien, 1995. Chemical mineral, and solubility characteristics of phosphate rock for direct
applicaiton. International Training Workshop on Direct Application of Phosphate Rock and Appropriate
Technology Fertilizers in Asia: What Hinders Acceptance and Growth. Kandy, Sri Lanka. February 20-14,
1995. IFDC, PO. Box 2040, Muscle Shoals, AL 35660.

137
Reducing PR particles to a size less than 150 microns by grinding is generally not warranted
as fine particles do not increase efi‘ectiveness greatly (Khasawneh and Dell, 1978). MPR
concentrate is produced at 70% passing a 75 - micron screen, thus further grinding is not
needed for direct application (Van Kauwenbergh, 1985; 1991). Soil propertieswherethe

PRs are directly applied do afi‘ect the rock’s efi'ectiveness. The dissolution process of PRs

may be described as:

Ca,,,(P0,),0r‘), + H‘ <=> 10Ca" + 6H,P0; + 2F’

Thus, soil conditions that favor the dissolution of the PR are those which enable the products
of dissolution, the Ca”, HzPO,’ and P ions, to rapidly difi‘use away fi'om the surface of the
dissolving PR particle. This drives the reaction to the right in favor of PR dissolution (Wilson
and Ellis, 1984; Sale, 1989; Sale and Mokwunye, 1993). In general terms, PR dissolution in
soil solution is favored when the soil pH, soil exchangeable Ca, and soil solution P
concentration are low (Khasawnah and Doll, 1978; Le Mare, 1991; and Chien, unpublished
report’). As the P sorption capacity increases, the dissolution rates of the PRs placed in those
soils also increase (Smyth and Sanchez, 1982; Syers and Mackay, 1986). This is due to the
removal of phosphate ions from the soil solution by rapid sorption on Fe and Al compounds.
The other important soil factor in PR dissolution is soil organic matter. Organic matter
forms complexes with soil Ca” and those released fiom PR dissolution, thereby diminishing

Ca” ion concentration in the soil solution and enhancing dissolution of the PR applied (Chien

 

’Chieu, 8.1-1. 1990. Reactions of phosphate rocks with acid soils of humid tropics. Workshop on
Phosphate Sources for Acid Soils of Humid Tropics of Asia. Kuala Lurnpur, Malaysia. Nov. 6-7, 1990.
International Fertilizer Development Center, PO. 2040, Muscle Shoals, AL 35660.

138

et al., 1990; Le Mare, 1991). Hence, PRs tend to be more efi‘ective in acid soils containing
high amounts of organic matter than in soils where the organic matter content is low (Le
Mare, 1991). In general terms, no single soil characteristic appears to have a consistent and
predominant influence on P release from PRs (Anderson et al., 1985).

Fertilizer eficiency of PRs increases with increasing rainfall, and the response to PR
is more erratic under low-rainfall conditions (Hammond er al., 1986). Films of moisture
arr-rounding the PR particle enable the products of dissolution, the Caz’, HzPO; and F ions,
to diffuse away from the dissolving surface of the particle. Adequate levels of soil moisture
also arable the plants to have well-developed roots that are suited to using the P fiom the PRs
(Hammond et al., 1986). Temperature has been found to have no significant effect on the
dissolution of PR in the soil. This implies that the influence of temperature on the agronomic
effectiveness of finely ground PR is most likely an indirect result of the influence of
temperature (climate) on the rate of physiological development of the crop (Hammond et al.,
1986; Sale, 1989).

The application method of PR fertilizer has a marked impact on its efi‘ectiveness.
Applying the PRs in a concentrated band results in a reduction in the dissolution rate of the
fertilizer due to the increase in levels of Ca” and P ions around the dissolving particles
(Hughes and Gilkes, 1986; Sale, 1989). Thus, for a PR to be most effective it should be
broadcast on the soil surface and incorporated into the soil at the time of planting (Sale, 1989;
Chien et al., 1990 b; Sale and Mokwunye, 1993). The use of lime reduces PRs dissolution
rates because liming reduces the H" and increases the Ca 2’ in the soil solution. To overcome

this problem, Foth and Ellis (1988) recommend applying PR about six months in advance of

 

139

lime. This would allow significant amounts of PR to dissolve before the pH and the calcium
status of the soil are raised by lime application (Sanchez and Salinas, 1981).

Difl‘aences among crops in their utilization of PRs depend on their requirements for
both P and Ca, and on how the uptake of these elements afl‘ects the composition of the
rhizospere. Crops that take up large amounts of Ca to lessen the ambient concentration in
the soil arhance the dissolution of PRs (Le Mare, 1991). Crops with extensive root systems,
such as cassava (Manr'hor esculenta Crantz) which is commonly grown in acid soils in the
tropics, are infected with hyphae of vesicular-arbuscular mycorrhizae (V AM), which helps
the transfer of P fi'om poorly soluble well distributed sources to plants, thus increasing the P
from the PRs (Le Mare, 1991; O'Keefe and Sylvia, 1991). In leguminous plants, there is
usually an acidification of the rhizosphere which results in an increase on the effect on PR
dissolution (Kirk and Nye, 1986). This is usually due to H-ions that build up in the
rhizosphere around root surfaces, which results from the excess uptake of cations by the
legume plant roots (Bekele er al., 1983). Such efl‘ects occur due to legume plants active
nitrogen fixation process (Bolan and Hedley, 1991). In general terms, crops do vary in their
ability to utilize the PRs and these differences contribute to their relative agronomic
efi‘ectiveness.

Some PRs are not suitable for direct application because of their low chemical
reactivity (Hamrtrond er al. , 1986). Therefore, partial acidulation of such PRs may improve
the agronomic value at a Iowa cost than would be required to manufacture the conventional,
firlly acidulated fertilizers fi'om that same PR (Hammond et al., 1986; Chien and Hammond,
1988). Acidulation of low quality PRs does release metallic impurities such as Fe, Al, K, and

Mg which react to form precipitates in the final fertilizer product (Mullins, 1988). In some

 

140
studies, agronomic effectiveness of partially acidulated phosphate rocks (PAPR) has been
shown to be as good as, or better than, that of water soluble P fertilizer such as TSP on soils
with high P-retention capacity (McLean and Wheela, 1964; McLean and Logan, 1970; Chien
and Hammond, 1989).

Few studies have been conducted in Tanzania using MPR on beans as a test crop.
Chesworth et al. (1988) compared the effects of the MPR and TSP yields of bean crops
grown on a plots with Ultisols soils. Although the soil tested low in pH (4.6) and low P (5.6
ppm), there was no significant response. These results were attributed to considerable
amounts of exchangeable Al found below 15 cm soil depth (Al saturation values >50%).
Similar results were reported by Giller er a1. (1988), and Smithson er a1. (1990), and were
attributed to the deficiency of K in the soil. Such conclusions show-how poor performance
of MPR in comparison to soluble P fatilizers has mostly been attributed to reasons other than
the rock’s characteristics.

In addition, most of the studies reported on the evaluation of MPR in Tanzania using
various crops were conducted at a single application rate. Where several application rates
were used, the seasons in which MPR was applied were different. Further, all field studies
have been conducted only under rainfed conditions. Conclusions fi'om such studies make
proper evaluation of , MPR as a direct applied fertilizer dificult. Since most studies have been
inconsistent, the main purpose of part of this study was to: (i) To assess the growth and
yield of bean crop as influenced by MPR in comparison with TSP fertilizer, and (ii) also to
assess the agronomic effectiveness of MPR as a direct applied fertilizer under rainfed, and

irrigated conditions.

141

MATERIALS AND METHODS

The experiments were conducted at Sokoine University of Agriculture, in the
Morogoro Region during the 1993 and the 1994 growing seasons. Details on location,
altitude, nreteorological variables, soil characteristics, and experimental procedures are fully
described in Chapter 2. The MPR chemical characteristics were determined at IFDC
laboratories (Table 4.1). Pre-plant soil characteristics were taken seven days before planting
in the 1993 growing season (Table 2.1), and 365 days alter the treatments’ application in the
1994 growing season (Table 4.2a). Post-harvest soil characteristics were taken seven days
afia harvest in both seasons (Tables 4.2b). Soil extractable P at all sampling times, including
those at post-harvest times, was analyzed by using the Bray-I method as suggested by Chien
(1978), and Chien er al. (1987). Soil sampling was done for all treatments, but not in all
experimartal plots. Therefore, soil chemical characteristics determined were not statistically
analyzed. '
Block A was used to assess the influence of fresh applied P on growth, and yield of
bean during the 1993 growing season in both experiments. In the following season, block A
was used to assess the influence of residual P (refer to Chapter 5 for details). Fresh applied
P in this study refers to P fertilizer applied at sowing in plots that had not received P in the
previous season. Although block B was used to collect data for evaluating the BEAN GRO
crop model (as described in Chapter 2), this-block was also used to assess the influence of
ammal applied P on bean crop performance. Annual applied P as used in this study refers to

P fertilizer applied at sowing in each season for two or more successive seasons.

 

. 142
Plant heights (cm) were measured at R. growth stage in each experiment by taking
the average heights of 10 randomly picked plants per plot from the two central rows. Crop
growth rate (CGR) between difl‘erent harvest times for each treatment were calculated as
described by Radford (1967). Maximum CGR values were recorded in all experiments for
both seasons. Harvest index (HI) was calculated fi'om marketable dry bean seed weight as

percentage of total dry matter per plot.
At final harvest, 10 plants were picked randomly fi'om each plot and used to calculate

the average number of pods per plant (N,). A subsample of 20 pods was used to determine
the number of seeds per pod (N,) as well as seed weight (W,). Seed weight (mg seed“) was
calculated by taking the average weight of a hundred seeds. The total seed weight (g) per
plant (Y) was obtained by multiplying the number of pods per plant by the number of seeds
per pod by the weight of individual seed (mg) i.e., (Y = Np * Ns * Ws). Seed yield per plot
was deta'mined after removing all the defective seeds and adjusting to a uniform level of 14%
moisture content.

Data were analyzed using MSTAT-C (Michigan State Univ., 1993) statistical
program Data fi'om the two growing seasons showed highly significant (P=0.05) season X
interaction for all variables. Experimarts wae analyzed separately for each season. Duncan's
Multiple Range Test (DMRT) was used to compare seven treatment means. Since there
were no clear difi’aences between TSP and MPR fatilizer levels using the DMRT, additional
statistical analyses using orthogonal comparisons between TSP and MPR in agronomic
effectiveness were performed. Orthogonal comparison is more powerful than DMRT in

detecting differences between two treatments.

 

143

To describe the non-linear relationship between total dry matter yield, and rate of P
applied fi'om MPR and TSP sources, a semilog function was used as described by Chien et

al. (1990).

y,=Yo+B/nX,X>1

Y, = Yield obtained with source (,),
Y, = Yield obtained with no P application (common to all P sources),
fl = Regresssion coefficient of source 0,

X = Rate of P applied,

To evaluate the RAE of the MPR source with respect to TSP, the RAE was defined
as the ratio of two slopes,

b
RAE index(%)=——“£’ixtoo.
bi?

To compare treatmart effects among P sources, the standard errors (SE) of estimate
for b, were used to evaluate whetha a givar P source was statistically different from the other
source in terms of increasing total dry matter production. This method has been used by
Hellums et al. (1989) to compare the agronomic effectiveness of various PRs with respect

CaCO, and by Chien et al. (1993) to compare the RAE of various PRs with respect to TSP

in soybean.

144

RESULTS AND DISCUSSION

General Observations

The experimental sites contained low pH, low extractable P, and low exchangeable
Ca Atbothexpelimaltal sitestlreinitial soil N, P, Ca, K, Mg, and Al saturation were below
the critical levels for bean production suggested by CIAT (Flor and Thung, 1989). The high
BD of 1.35- 1.55 g cm'3 at the experimental sites (Tables 2a and 2b) may have afi‘ected root
penetration, and therefore resulted in growth and yield reduction in the bean crop (Asady
etal,l985)

Most of the plants were nodulated in both experimarts, but did not have a large mass
of pinkish nodular tissues. At emergence plant population was >90%, but at harvest it was
reduced to as low as 65% in some plots due to beanfly (Ophr'omyr'a ssp) and termite
(Microcerotermes ssp) and rodent attack (Appendices 5a and b). Nutsedges (Operates
rotundus and C. esculentus) were the most important weed species at the irrigated site. At
the rainfed site, wondering jew (Commelr'na benghalensis), and wild spinach (A maranthus
spp) were the dominant weed species. Due to high seed infection by ascochyta blight, the
infected seed were removed. Thaefore comparison of the TSP and MPR fertilizer effect on

bean performance are discussed using only the TDM production values in this chapter.

145

Minjingu Phosphate Rock Characteristics.

Table 4.1 indicates some of the selected chanical characteristics of the MPR ore used
during this study. The length of the axis was found to be 9.3586 +/- 0.006 angstroms (unit
cell a-dimension).

The soluble P results indicate the rock is highly-reactive. However, the degree of
CO, substitution for P0, in the apatite structure indicates the rock is only a medium to low
PR. Similar results were reported by Van Kauwenbergh (1991), and Kpomblekou-A and
Tabatabai (1994). This suggests that non-apatitic P minerals such as Fe-Al-P, which are more
soluble than apatite in NAC may be present. Further, the MPR may contain a small fiaction
of hydroxyapatite that is more soluble than the fiancolite. This small amount of
hydroxyapatite may not affect x-ray difli'action peaks, but afl‘ects the solubility measurements

(Dr. S.H. Chien, 1995 personal communication).

146

Table 4.1. Selected properties of Minjingu phosphate rock '

 

Empirical formula C39.reNao.roM3oor(P 04):.ss(C03)e42F 2.1

 

 

 

Selected properties Value

pH 9.0
Total (P%) 13 .5
Reactivities (% P205)

Neutral ammonium citrate (N AC)
First extraction 5.3
Second extraction 6.7

 

’Determined at the International Fertilizer Development Center (IFDC) AL, laboratories.

Effect of P Treatments on Soil pH, Extractable

Phosphorous and Exchangeable Calcium.

At harvest of the 1993 experiment, the soil pH, Bray-1P (ppm), and exchangeable Ca
(meq 100g'l soil) in plots that received I? had increased (Table 4.2b) in comparison with the
values at pre—planting time, i.e., before the initiation of the experiment during the 1993
growing season (Table 2.1). The increases observed were higher with high rates of MPR than
with TSP applications. The change in soil pH due to the P fertilizer application was higher
in irrigated plots than in rainfed. Significant soil pH increases were registered at the end of
the 1994 season especially in plots which had received high rates of MPR application
compared to the results indicated in Table 4.2a. At the rainfed site soil remained more acidic

during the crop growing period. Such conditions may have enhanced MPR dissolution. The

147

present results show that with time, the applied P fertilizers were releasing more Ca (and
probably Mg) thus raising the soil pH. At the end of the 1993 growing season, increases in
extractable P were observed in plots which had received fertilizer treatments. Phosphorous
increases in plots which received no fertilizer, and those which received TSP may be due to
mineralization of organic matter incorporated into the soil during land preparation.

Obsavations reported here are similar to those reported by Chien er al. (1987) who
worked with a number of PRs in a Colombian Oxisol with pasture grass (Brachiaria
decranbes), and Casanova and Solorzano (1994) working with sorghum and soybean on acid
soils using TSP and different PRs in Venezuela.

The effects of increase in soil Ca by application of the increasing PR rates has also
been observed by Hellums et a1. (1989) in a greenhouse study, where an increase in Ca uptake
by the maize plants, and in the soil exchangeable Ca with the application of higher rates of the

PRs obtained fi'om South America and West Afiica were reported.

148

Table 4.2a; Soil pH, Bray-1P and calcium levels in pro-plant soil at of 0-12 cm layer in blocks
A and B at the beginning of 1994 growing season.

 

Treatment Soil pH Bray I Ca

 

(Ppm) (meq)

Rainfed experiment
0 kg P ha" 4.47 1.21 2.13
22 kk P ha" TSP 4.56 3.87 2.92
22 kg P ha'l MPR ’ 4.61 6.89 3.96
44 kg P ha" TSP 4.49 5.45 3.86
88 kg P ha“ TSP 4.58 9.35 4.61
88 kg P ha" MPR 4.44 10.65 5.01
176 kg P ha" MPR 4.48 11.36 6.94

Irrigated experiment
0 kg P ha" 5.51 2.53 2.82
22 kk P ha" TSP 5.41 4.91 4.94
22 kg P ha“ MPR 5.83 5.61 5.77
44 kg P ha" TSP 5.42 5.95 4.19
88 kg P ha" TSP 5.66 7.31 4.54
88 kg P ha‘l MPR 6.81 8.66 6.19

176 kg P ha'l MPR 6.11 12.42 7.19

 

149

Table 4.2b. Soil pH, Bray-I, and calcium levels in post-harvest soil at 0-12 cm layer in blocks
A and B at the end ofthe1993 and block B at the end of 1994 growing seasons.

 

 

Treatment Soil pH Bray I Ca
(Ppm) (meq)
Rainfed experiment
A B A B A B
0 kg P ha" 4.14 4.67 1.63 1.98 2.47 2.01
22 kk P ha" TSP 4.53 4.56 3.33 3.97 2.49 4.91
22 kg P ha" MPR 4.38 4.63 3.51 4.06 3.49 4.78
44 kg P ha" TSP 4.59 4.72 5.75 6.01 3.24 4.86
88 kg P ha" TSP 4.79 4.75 6.96 6.32 3.12 5.32
88 kg P ha" MPR 4.7-1 4.78 8.94 9.26 3.87 3.19
176 kg P ha" MPR 4.65 4.76 9.19 9.82 4.36 6.27
Irrigated experiment
0 kg P ha" 5.32 5.32 2.31 1.64 2.38 2.51
22 k P ha" TSP 5.62 5.40 5.35 3.39 2.89 4.98
22 kg P ha“ MPR 5.65 5.69 4.86 4.41 2.94 3.04
44 kg P ha" TSP 5.77 5.82 6.79 5.29 2.78 3.18
88 kg P ha" TSP 5.83 5.99 6.94 8.62 2.81 4.81
88 kg P ha" MPR 5.74 6.24 7.16 6.44 2.61 5.96
176 kg P ha" MPR 5.80 6.11 9.94 8.21 3.27 6.46

 

A represents values from blocks A and B in the 1993 season.
B represents values fi'om block B in the 1994 season.

150

The higher increase in soil pH, and exchangeable Ca reported in this study due to the
application of MPR, is associated with fertilizer characteristics of high pH (9.0) because of
the high Ca content (29.2%) in the apatite (Kpomblekou-A and Tabatabai, 1994) . Increase
of Ca by TSP application was due to its high Ca content of 13.8% (Young et al., 1985).
Further, the reaction of free Fe and Al (abundant in the soil) with P ions fiom the fertilizer,
can result inthe increase in soil pH (Coleman and Thomas, 1967). The Bray-1P increase was
mainly due to the high concentration of total P (15%) in the MPR (Kpomblekou-A and
Tabatabai, 1994), and the intaaction of the fertilizer with the initial low (1 . 18-2.83 ppm) soil
P. Therefore, the soil clrarmtaistics (low pH, low P and low Ca) may have contributed to the
observed results in plots that received MPR treatments as indicated by Khasarnneh and Doll

(1978); Le Mare, (1991), and Sale and Mokwunye (1993).

Crop Growth, Total Dry Matter and Seed Yield Components

as Influenced by Different P Regimes

Figures 4.1 and 4.2 illustrate total dry weights at five growth stages of bean crop
grown under difl‘erent P regimes. The initial crop growth was slow, especially between V2
and V3 growth stages. This slow growth may be due to a small number of cells that were
dividing, and small leaf area available for sunlight interception and photosynthesis. During
this growth stage a large percentage of carbohydrates were being partitioned to the roots
(Brown, 1984). With time, total dry weights increased mainly due to an increase in dry
weights of stems and pods. The decline in total dry weight was after R, growth stage. At

this growth stage, leaf and petiole fall fiom the plants increased. This occurred because at R.

l 5 1
growth stage carbohydrates, and argars produced were being translocated to the young pods
that were formed. Studies with sorghum have shown that unless adequate nutrients are
available during seed formation, this translocation may cause deficiencies in leaves, and
premature leaf loss which reduces leaf area duration, and reduces yield (V anderlip, 1972).

As obsaved by Saito and Kato (1994) in soybean, and by Acosta-Gallegos and White
(1995) in beans, the duration within the growth stage in this study depended mainly on the
temperature during the growth period, and not on the fertilizer treatments. Although the
rainfed experiments received total precipitation of 475.0 for the 1993, and 341.4 mm for the
1994 season, wata stress occurred for about 14 days during the grain filling periods starting
at R. growth stage. This was more distinct during the 1994 growing season (Figure 2.3 and
Appendix 2.b). For the rainfed experiment, total dry matter was lowest in the experiment
grown under fresh applied P in the 1994 season, giving the maximum TDM of 202 g m".
This low yield among all treatments was mainly due to poor rainfall distribution in the period
which coincided with the crop reproductive phase.

Canadian Wonder cultivar is a large seeded cultivar (Andean type I), therefore high
temperatures during the reproductive phase (Figure 2.1) may have resulted in low rates of
canopy photosynthesis due to low leaf N as postulated by Sexton et al. (1994). Due to a
better rainfall pattern, the 1993 season rainfed expaiment with fiesh applied P resulted in high
TDM production of 302 g m”2 (Figure 4.2).

The 1993 irrigated experiment started with a high level of wata supply to assure rapid
and complete development of the crop root system during the early growth. This was
necessarysincetheexperimartwassownattlreendoftherainyseasonwhen the soil moisture

contart was already very low, i.e., 15.23% (Table 2.1). Day of the year (date), water applied,

152

and crop growth stage at which irrigation was applied in both seasons are shown in
Appendices 4a and 4b. The amount of irrigation for the 1993 and that of the 1994 growing
season difi‘ered by 39% because the 1993 crop received a total of 539.0 mm while that of
1994 received a total of 327.0 mm. This difi‘erence was due to planting dates, and water
deficits during the growing seasons. The 1993 experiment was planted in late May (on
93148) while the 1994 experiment was planted in early May (on 94126) when the soil was
still moist (>70% soil available water), and precipitation was still high. While irrigation was
applied throughout the growing season of 1993, it was necessary only between 14 and 56
DAP during the 1994 growing season.

Leaf number ranged between 6.00 and 7.33 at the main stem per plant. Leaf
production stopped when plants were at the R, growth stage as there was no new node
production on the main stems at this growth stage. Patterns of treatment effects on leaf
production were inconsistart betwear expaiments. In general, the irrigated experiments had
a higher number of leaves on the main stem than rainfed experiments. Difi‘erences in plant
height due to treatments applied were obvious at the R. growth stage. The control plots had

the shortest plants.

 

153

 

400 ~
300 ~
200 -
roo -

 

300 .

200 r

Bean dry weight production (g m")
o

 

100 .

 

 

 

 

 

 

o - . . . . - a. - . . - . 2 -
0 14 28 42 56 70 0 14 28 42 56 70 84
Days after planting

° Control

- 44 kg P ha-l TSP

‘ 88 kg P ha-l TSP

' 88 kg P ha-l MPR
0 176 kg P ha-1 MPR

Figure 4.1. Bean dry weight production for the crop grown under (A and B) rainfed and (C
and D) irrigated conditions with fi'esh applied fertilizer P. Arrows indicate crop growth
stages, and vertical bars represent standard errors (S.E.).

154

llllllllllllllllllllll

 

 

 

 

 

 

 

400- A R’ If R, C R, R’
a“ R 1 l i i
5300 . '
3?
=200-

.2
$100-
‘a, o
3400-
q£300-
>5
I-r
1:200-
G
8
m100r
o

0 1428425670 0142842567084
Days after planting

° Control

- 44 kg P ha-l TSP

‘ 88 kg P ha-l TSP

' 88 kg P ha-l MPR
° 176 kg P ha-l MPR

Figure 4.2. Bean dry weight production for crop grown under (A and B) rainfed and (C and
D) irrigated conditions with annual applied P fertilizer. Arrows indicate crop growth stages,
and vertical bars represent standard errors (S.E.).

155

The LAI increased with crop growth at all P fertilizer levels and sources regardless
of the fertilizer regimes. Overall, theLAWAI was maximum at the R, growth stage.
Application of 88 kgP ha" (TSP) resulted in the highest LAW values and as expected, the
control plots had the lowest values in both seasons. Similar results have been reported in
soybean where leaf senescence was decreased by P application due to increase in leaf area
duration (LAD) (Grabau et al., 1986).

Gardner et al. (1985) pointed out that for a C,-plant such as bean, the appropriate
CGR value is about 20 g rn'2 day". The CGR values ranged fi'om 6.21 g m‘2 day" in control
plots of the rainfed experiment with a fiesh P fertilizer regime, to 15.65 g rn‘2 day'1 in plots
that received 88 kg P ha" TSP in the rainfed experiments with an annual P fertilizer regime.
The CGR values increased to a maximum between the early and mid-pod filling formation
growth stages, and declined thaeafier. Thaefore, maximum CGR values coincided with mid-
pod filling period. During this period the crop was mature enough to exploit most of the
environmental factors such as moisture, nutrients, solar radiation. and temperature.
Maximum CGR value of 15.65 g m '2 day " (or 156.5 kg ha" day ") was obtained at the LAW
of 2.65. The CGR values of this study suggest that the crop did not have a favorable
environment for maximum growth.

Due to the soil and MPR charactaistics, and crop nranagemart, significant differences
were not evidart among most treatments for both the CGR and the TDM production during
the seasons. At harvest, the TDM production was highest in plots with highest plant
population Total dry matter production is a function of environmental factors, cultivar and
crop managemart. Response to TSP was relatively consistent, and highly significant in both

seasons. Unda irrigated conditions, application of 88 kg P ha" (TSP) resulted in the highest

156

production of TDM (4.2 t ha") as shown in Figure 4.3. Bean response to MPR was sporadic,
inconsistent, and unexplainable ( Figure 4.4). In 1993, MPR had the highest yield under
rainfed conditions. However, in the following year, response was the lowest. Such results
have lead to improper conclusions (IFDC, 1990), and recommendations on the use of MPR
in Tanzania. The control plots had the lowest TDM values due to the decrease in crop
growth. Such results may have been due to P deficiency in bean plants as indicated in the
literature (Y an et al., 1995 a and 1995 b). High levels of applied P may have increased TDM
through processes which resulted into higher efiiciarcy in photosynthesis (Lauer et al., 1989),
and increased in leaf area (Fredeen er a1, 1989).

Since bean is an annual crop, it requires P in early periods of its grth (F redeen er
al., 1989). This is supported by the fact that, the TSP fertilizer which has a higher water
solubility (>85%) than the water insoluble MPR supplied relatively the low P at V3, and R,
growth stages (Table 3 .2) required by the crop which resulted in higher TDM production.
Low response of bean crop to MPR may have been due to the low availability of P fiom the
fertilizer. This may have been caused by factors such as, soil texture at the experimental sites
(Table 2.1). Sandy soils such as those found at the sites used for this study are known to
accumulate P (Ritchie and Weava, 1993). This is because sandy soils have low CEC (which
is also closely related to soil texture), and do not provide a sink for Ca ions released from PR.
Harce, the PR dissolution is slowed which may result in reduced in agronomic effectiveness

(Khasawneh and Doll, 1978; Kanabo and Gilkes, 1988).

157

 

 

 

 

 

 

 

 

 

Total dry matter production (kg ha‘l)

 

 

 

100

Rate of P applied (kg P "a '1)

TDM rainfed 1993
TDM rainfed 1994
TDM irrigated 1993
TDM irrigated 1994

{DIG

Figure 4.3. Bean dry matta production as influarced by triple superphosphate (TSP). Vertical
bars represent standard errors of means (SE).

 

158

l r l a R

3000 - (a) Fresh P

 

    

 

2000-

 

 

 

 

 

 

 

Total dry matter production (kg ha'l)

 

 

0 I ' I ' I ‘ l '
0 5O ' 100 150 200

 

Rate of P applied (kg P "a '1)

TDM rainfed 1993
TDM rainfed 1994
TDM irrigated 1993
TDM irrigated 1994

45!.

Figure 4.4. Bean dry matter production as influenced by Minjingu phosphate rock (MPR).
Vertical bars represent standard errors of means (SE).

159

The variation in HI values ranged between 21% in the control plots with freshly
applied P during the 1993 rainfed expaimart to 49% in plots that received 88 kg P ha" MPR
with a crop grown under irrigation in the 1993 growing season. The HI values generally
ina'easedwiththeincreaseinseedyield, and irrigated experimentshad higherHIvaluesthan
rainfed experiments in both seasons. Similar results have also been reported by Sirait er al.
(1994) in lima bean (Phaseolus Iunatus L.). Interpreting the HI values in this study is difiicult
because of the defoliation, and petiole loss during maturation (after R,,) growth stage.

Expecting to include all plant tissues and obtain realistic values is impossible. Thus
modified HI values such as those proposed by Shibata et a1. (1980) would be appropriate.
These would include the dry weight of abscised organs such as flowers, pods and leaves.
From the obtained results, however, the assumption can be made that the increase in seed
yield recorded in this study was due to the increase in HI. Also, at certain growth periods
plants were not producing any more seed.

Seed yield plant" in beans can be expressed as a product of three components: pods
per plant, seed per pod, and seed weight (Fageria et al., 1991). Among these yield
componarts, pods pa plant has often been recommended as an indirect selection criterion for
increasing yield, because of its higher and more consistent correlation with yield (Bennet et
al., 1977; Sarafi, 1978).

In their study, Nienhuis and Singh (1986) concluded that high seed yield in bean is
associated with high number of pods m", seeds podl , and all architectural traits except
branches plant". Data for yield componarts are presented in Tables 4.3a to 4.3d. Pods plant'
‘ was the component found that consistently affected by the applied P. Similar results have

been reported by Fageria (1989) and Yan et al. (1995 b). Some studies using indeterminate

 

160
bean arltivars (Acosta-Gallegos and Shibata, 1989) revealed that the number of pods plant"

was the yield component most adversely affected by water stress.

Table 4.3a. Yield components in beans grown with fresh applied P treatments under rainfed
conditions.

 

 

 

 

 

Treatments Pods plant" Seeds pod" Seed wt Total seed wt
(g 100 seed") (g plant")
1993 Growing Season
0 kg P ha" 5.10 (0.29)1 3.40 (0.06) 42.60 (2.08) 7.39 (0.69)
22 kgP ha" -TSP 6.20 (0.35) 4.00 (0.12) 41.17 (0.76) 10.21 (0.77)
22 kgP ha" -MPR 6.90 (0.38) 3.70 (0.12) 42.13 (2.55) 10.76 (0.64)
44 kgP ha" -TSP 7.10 (0.74) 3.90 (0.23) 40.03 (0.83) 11.08 (1.39)
88 kgP ha"-TSP 5.90 (0.21) 4.10 (0.09) 43.97 (1.48) 10.64 (0.29)
88 kgP ha"-TSP 5.30 (0.31) 3.80 (0.06) 41.43 (1.09) 8.34 (1.04)
176kgP ha"-MPR 6.30 (0.54) 3.50 (0.10) 43.87 (2.77) 9.67 (0.36)
1994 Growing Season
0 kg P ha" 4.50 (0.27) 2.60 (0.52) 28.54 (0.83) 3.34 (0.44)
22 kgP ha" -TSP 6.40 (0.27) 3.00 (0.44) 31.86 (3.57) 6.12 (0.62)
22 kgP ha" -MPR 6.40 (0.64) 2.70 (0.32) 29.52 (2.66) 5.10 (1.73)
44 kgP ha" -TSP 6.70 (0.65) 4.10 (0.41) 35.67 (2.41) 9.80 (1.86)
88 kgP ha"-TSP 5.80 (0.82) 4.00 (0.43) 35.47 (0.16) 8.23 (1.41)
88 kgP ha"-TSP 6.60 (1.80) 2.70 (0.06) 31.64 (0.54) 5.64 (1.51)
176kgP ha"-MPR 6.50 (0.95) 2.30 (0.63) 27.56 (1.86) 4.12 (1.39)

 

 

1Figures in parentheses show the standard error values.

 

161

Table 4.3b. Yield components in a bean crop gown with flash applied P treatments under
irrigated conditions.

 

 

 

 

 

Treatment Pods plant" Seeds pod" Seed wt Total md wt
(g 100 seed") (g plant")
1993 Growing Season
0kgPha" 6.9 (0.52)1 3.6 (0.36) 41.83 (1.68) 10.39 (1.24)
22kgP ha"-TSP 7.4 (0.70) 4.7 (1.01) 44.70 (1.01) 15.55 (1.77)
22kgP ha"-MPR 6.9 (0.23) 3.5 (0.30) 44.57 (0.29) 10.76 (1.29)
44kgP ha"-TSP 7.4 (0.46) 4.1 (0.09) 42.67 (0.86) 12.94 (0.86)
88kgP ha"-TSP 7.4 (0.29) 4.0 (0.24) 42.07 (0.91) 12.45 (0.63)
88kgP ha"-MPR 7.4 (0.40) 3.5 (0.08) 41.77 (0.47) 10.82 (0.57)
176kgP ha" -MPR 7.5 (0.34) 4.1 (0.15) 42.73 (3.12) 13.14 (0.75)
1994 Growing Season
0kgPha" 4.6(0.76) 3.3 (0.21) 37.30 (1.54) 5.66 (1.15)
22kgP ha"-TSP 6.8 (1.05) 3.0 (0.12) 36.01 (1.97) 7.35 (0.56)
22kgP ha"-MPR 6.4 (1.46) 3.5 (0.34) 34.49 (6.55) 7.73 (2.00)
44kgP ha"-TSP 5.0 (1.01) 3.6 (0.12) 39.12 (2.46) 7.04 (1.37)
88kgP ha"-TSP 6.5 (0.23) 3.6 (0.42) 42.53 (1.02) 9.95 (0.74)
88kgP ha"-MPR 4.8 (0.32) 3.3 (0.17) 38.75 (3.22) 6.14 (0.32)
176 kgP ha" -MPR 5.2 (0.42) 3.3 (0.35) 36.32 (4.16) 6.23 (1.32)

 

 

‘Figures in parentheses show the standard error values.

 

162

Table 4.3c. Yield components in a bean crop gown with fi'esh applied P treatments under
irrigated conditions.

 

 

 

 

 

Treatment Pods plant" Seeds pod" Seed wt Total seed wt
(g 100 seed") (g plant")
1993 Growing Season
0kgPha" 6.9 (0.52)1 3.6 (0.36) 41.83 (1.68) 10.39 (1.24)
22kgP ha"-TSP 7.4 (0.70) 4.7 (1.01) 44.70 (1.01) 15.55 (1.77)
22kgP ha"-MPR 6.9 (0.23) 3.5 (0.30) 44.57 (0.29) 10.76 (1.29)
44kgP ha"-TSP 7.4 (0.46) 4.1 (0.09) 42.67 (0.86) 12.94 (0.86)
88kgP ha"-TSP 7.4 (0.29) 4.0 (0.24) 42.07 (0.91) 12.45 (0.63)
88kgP ha"-MPR 7.4 (0.40) 3.5 (0.08) 41.77 (0.47) 10.82 (0.57)
176kgP ha" -MPR 7.5 (0.34) 4.1 (0.15) 42.73 (3.12) 13.14 (0.75)
1994 Growing Season
0kgPha" 4.6(0.76) 3.3 (0.21) 37.30 (1.54) 5.66 (1.15)
22kgP ha"-TSP 6.8 (1.05) 3.0 (0.12) 36.01 (1.97) 7.35 (0.56)
22kgP ha"-MPR 6.4 (1.46) 3.5 (0.34) 34.49 (5.55) 7.73 (2.00)
44kgP ha"-TSP 5.0 (1.01) 3.6 (0.12) 39.12 (2.46) 7.04 (1.37)
88kgP ha"-TSP 6.5 (0.23) 3.6 (0.42) 42.53 (1.02) 9.95 (0.74)
88kgP ha"-MPR 4.8 (0.32) 3.3 (0.17) 38.75 (3.22) 6.14 (0.32)
l76kgPha"-MPR 5.2 (0.42) 3.3 (0.35) 36.32 (4.16) 6.23 (1.32)

 

 

1Figures in parentheses show the standard error values.

 

163

. Table 4.3d. Yield components in beans gown with annual applied P treatments under
irrigated conditions.

 

 

 

 

 

 

Treatment Pods plant" Seed pod" Seed wt Total seed wt
(g 100 seed") (g plant")
1993 Growing Season
0kg ha" 5.93 (0.92)l 3.50 (0.15) 45.77 (0.63) 9.49 (1.43)
22kgP ha"-TSP 7.50 (0.35) 3.60 (0.15) 44.97 (2.37) 12.14 (0.48)
22kgP ha"-MPR 7.47 (0.56) 4.37 (1.14) 46.80 (2.06) 15.27 (1.67)
44kgP ha"-TSP 6.97 (0.30) 3.67 (0.43) 44.47 (1.07) 11.38 (1.27)
88kgP ha"-1-TSP 8.47 (0.23) 3.47 (0.07) 42.43 (0.61) 12.47 (0.43)
88kgP ha"-MPR 8.37 (0.69) 3.70 (0.31) 37.53 (2.79) 11.62 (0.91)
176kgP ha"-MPR 6.87 (0.35) 3.23 (0.13) 42.10 (1.27) 9.34 (0.38)
1994 Growing Season
0kgPha" 3.93 (0.09) 4.30 (0.10) 35.69 (0.71) 6.03 (0.09)
22kgP ha"-TSP 4.70 (0.61) 4.13 (0.34) 35.60 (1.71) 6.91 (1.08)
22kgP ha"-MPR 5.17 (0.95) 4.25 (1.40) 31.69 (0.74) 6.96 (1.23)
44kgP ha"-TSP 5.47 (0.75) 4.07 (1.39) 35.71 (0.55) 7.95 (1.40)
88kgP ha"-TSP 6.25 (0.87) 3.53 (0.20) 34.97 (1.53) 7.69 (0.24)
88kgP ha"-MPR 4.93 (0.35) 4.77 (0.86) 35.16 (0.53) 8.27 (1.39)
176kgP ha"-MPR 5.93 (0.35) 3.83 (0.49) 34.99 (1.05) 7.95 (0.58)

 

‘Figures in parentheses show the standard error values.

164

Present results do not show sigrificant efi‘ects on pods plant" within experiments.
However, the difi‘erences were observed between seasons. The 1994 season had a lower
average number of pods plant" than those in the 1993 season, regardless of whether the
experiment was rainfed or irrigated. Seeds pod", and seed size (g see‘d ) were also
sigrificantly afi‘ected by P levels. Total seed weight per plant (g plant") was lower in the
control plots. The lowest value was from the rainfed conditions under fi'esh applied P during
1994 season

Plants in control plots aborted a larga number of pods at R, gowth stage ( i.e., pods
less than 10 mm in length) than plants in plots with applied P. This may have resulted fiom
competition among developing pods for photosynthate, and metabolites, which in turn were
affected by available P to the plants. The control plots had higher abortion of individual
ovules, thus resulting in lower seeds per pod. In the 1994 gowing season for example, the
control plots under rainfed conditions gave 3.34 g plant", while other plots with applied P
ranged betwear 4.1 and 15.3 g plant". However, plots that had the highest seed yield plant"
did not necessarily give the highest seed yield per hectare. These results were also afi‘ected
by the TDM loss due to fallar leaves, petiole, and some stems (branches) at maturity harvest.
Furtha, there was a reduction in plant population at harvest time (Appendices 58 and 5b) due

to attack of pests such as termites and some rodents.

 

165

Relative Agononric Effectiveness of MPR

The relative agonomic efi‘ectivaress (RAE) values from this study are shown in Table
4.4. On the basis of the regession coeficiarts of the P response curves, the values calculated
for RAE of MPR (TSP=100%) in increasing bean TDM production difi‘ered with seasons and
experimarts. All RAE values of MPR were below 50% except that of the 1993 fiesh applied
P and rainfed experiment (RAE=64.59), and that of the 1994 annual applied P irrigated
experiment (RAE=64. 15). The RAE values reported in this study are lower than values
reported with highly reactive sources such as Gafsa PR (Tunisia), but are similar to values
reported for medium reactive Huila PR (Colombia) using bean as the test crop (Chien and
Hammond, 1978), and Central Florida PR (USA) using soybean crop (Chien et al. (1990).

The results based on DMRT (P=0.05) showed that TSP was generally as effective as
MPR under rainfed, and irrigated conditions. However, based on orthogonal comparisons,
the results showed that TSP was as effective (P=0.05) as MPR in 13 (three from irrigated
conditions) out of 24 comparisons (Table 4.5). Among the six comparisons where TSP
showed more sigrificant effect than MPR, four were fiom the irrigated experiment. MPR
was sigrificantly more effective in five comparisons, from the irrigated experiments.

The low agononric efi‘ectiveness reported in this study may have been due to bean
TDM production being higher with control treatments than those usually reported under
ordinary expaimartal conditions (in Tanzania). As indicated earlier, the soil texture at both
experimental sites exhibit high P retention capacity. Using the scale suggested by Chien et
al. (1990), MPR showed the charactaistics of low to medium agononric effectiveness. This

observation is also supported by the NAC reactivity characteristics shown in Table 4.1.

 

Table 4.4. Values of coeficiarts (b) and relative agonomic efl‘ectiveness (RAE) of Minjingu
phosphate rock and triple superphosphate in terms of bean total dry matter production.

166

 

 

 

 

 

 

 

 

 

 

Fresh applied P Annual applied P
Rainfed experiment
1993 1994 1993 1994
b, RAE b, RAE b, RAE b, RAE
MPR 9.12 64.59 3.56 34.46 3.15 20.65 5.59 34.40
TSP 14.12 100.00 10.33 100.00 15.25 100.00 16.25 100.00
Irrigated experiment
MPR 3.15 20.66 5.59 34.44 2.97 18.15 1.79 64.15
TSP 15.25 100.00 16.25 100.00 16.36 100.00 2.79 100.00

 

Y,=Y,+b,1nx,x>1

where

Y, = Yield obtained with source (i)

Y, = Yield obtained with no P application

b, = Regession coeficient source

167

Table 4.5. Summary of orthogonal comparisons betwear triple superphosphate and Minjingu

phosphate rock in agonomic efl‘ectiveness.

 

 

Fresh applied P
Season
Rainfed
1993 Treatment compared
22 kgPTSP vs 22 kgPMPR A= °
88 kg P TSP vs 88 kg P MPR A=B°
44 kg P TSP vs 176 kg P MPR A=B
1994 22 kg P TSP vs 22 kg P MPR A=B'
88 kg P TSP vs 88 kg P MPR A=B'
44 kg P TSP vs 176 kg P MPR A>B
Annual applied P
1993 22 kg P TSP vs 22 kg P MPR A=B°
88 kg P TSP vs 88 kg P MPR A=B'
44kgPTSP vs 176 kgPMPR A<B
1994 22 kg P TSP vs 22 kg P MPR A=B'
88 kg P TSP vs 88 kg P MPR A=B'
44 kg P TSP vs 176 kg P MPR A=B

 

Irrigated

A<B.
A<B.

A<B°
A>B'
A:

A<B.
A>B
A>B

A>B°
A<B
A= '

 

° A=B based on DMRT (P=0.05)
A=Triple superphosphate (TSP)
B=Minjingu phosphate rock (MPR)

168

SUMMARY AND CONCLUSIONS

Soil pH, extractable P, and exchangeable Ca increased in plots whae P fertilizers were
applied regardless of type or rate of fertilizer used. However, increases were higher in plots
that received high rates of MPR Crop gowth stages and maturity periods depended mainly
on the night cool temperatures during the gowth period and not on P treatments. The
irrigated expaimarts accumulated geata amounts ofTDM The decline ofTDM production
(kg ha") at harvest maturity was attributed mainly to the decrease in plant population.

Despite irrigation and other optimum agonomic conditions provided to the
experiments, seed yields in this study were still low when compared with yields fi'om medium
altitude, i.e., 900 to 2100 m This was mainly due to disease (ascochyta blight) attack to the
seed. Also, beans do not do well in locations below the altitude of 600 m. High temperatures
ranging between 20° and 30° C, such as those observed during the rainfed experimental
periods, result in poor fi'uit set.

This study shows that MPR can result in raising the soil pH (liming effect) when
applied at high rates such as 88 and 176 kgP ha". However, such high rates would not be
economically feasible for Tanzanian resource-poor farmers. This may be due to high
transportation costs per unit of P delivered to the farm because of MPR's low nutrient
content compared to TSP. The RAE of MPR with respect to TSP, in terms of increasing
total dry matter, were lower than TSP (RAE=100%). Overall analyses show that, despite
some suggestions that MPR can be as effective as superphosphate fertilizers in bean

production, TSP is a supaior fertilizer in areas with soil pH values allowing bean production.

169

Since MPR is known to be a complex mixture of apatites (Van Kauwenbergh, 1991),
before the ore is packed for commercial distribution, characterization studies should be
paforrned from each sample of ore. This will identify the best material that is effective as a
P source for direct application when used under appropriate conditions.

Finally, more laboratory, lgeenhouse, and field research should be conducted to
delineate all factors influarcing the efl'ectiveness of MPR for direct application. Furthermore,
apart fiorn low soil pH, low soil P, and low soil Ca, MPR use should be advocated in areas

where irrigation can be applied in bean production.
RECOMMENDATIONS FOR FUTURE RESEARCH

The information obtained in this study needs to be complemented. Therefore, using
the experience gained in this study, the following is suggested for firture research.

1. Research on partially acidulated MPR should be initiated in Tanzania to determine
whether this technology can provide a means whereby MPR could be an
agonomically better fertilizer for bean and other crops’ production.

2. Since there is a problan of manual direct application of fine (dusty) MPR as observed
during this study, research" on compacted dust free materials of MPR for farmers
should be initiated.

3. Different bean cultivars should be used for firrther evaluation of MPR as different
bean varieties have different characters of P eficiency which are heritable (Y an et
al., 1995 a and 1995 b). Also, different locations including those with high bean yield

potential e.g., >1000 m.a.s.l.) should be included in future studies.

BIBLIOGRAPHY

Acosta-Gallegos, J .A, and J .K. Shibata. 1989. Efi‘ects of water stress on gowth and yield
of indeternrirrate dry-bean (Phaseolus vulgaris L.) cultivars. Field Crops Res. 20:81-
93.

Acosta-Gallegos, J .A, and J .W. White. 1995. Phenological plasticity as an adaptation by
common bean to rainfed environments. Crop Sci. 35:199-204.

Ae, N., J. Arihara, K Okada, T. Yoshihara, and C. Johansen, 1990. Phosphorous uptake by
pigeon pea and its role in cropping systans of Indian subcontinent. Science. 2482477-

248.

Allen, ER, D.W. Ming, LR Hossner, and D.L. Henniger. 1995. Modelling transport
kinetics in clinoptilolite-phosphate rock systems. Soil Sci. Soc. Am. J. 59:248-255.

Anderson, D.L., WR Kusson, and RB. Corey. 1985. Phosphate rock dissolution in soils:
indications fiom plant gowth studies. Soil Sci. Soc. Am. J. 49:918-925.

Asady, G.H, AJ.M. Smucker, and M.W. Adams 1985. Seedling test for quantitative
measurement of root tolerance to compacted soils. Crop Sci. 25:802-806.

Barber, sa. 1967. Liming materials and particles. In R w. Person and F. Adams (eds) Soil
acidity and liming. Agonomy 12:125-160.

Bennet, J.P., M.W. Adams, and C. Burga. 1977. Pod yield component variation and
intercorrelation inPhasequs vulgaris L. as afi‘ected by plant density. Crop Sci. 17273-
75.

Bekele, T, B.J. Clino, P.AI. Elhert, AA Van der Mass, and A Van Deist. 1983. an

Evaluation of plant borne factors promoting the solubilization of alkaline rock
phosphates. Plant and Soil. 75:361-378.

170

171

Bolan, N.S., and M.J. Hedley. 1991. Processes of soil acidification during nitrogen cycling
with emphasis on legume based pastures. p. 169-179.171 RJ. Wright, V.C. Baligar,
and RP. Murrmann (eds) Plant and soil interactions at low pH. Kluwer Academic,
Dordrecht, The Netherlands.

Brown, RH. 1984. Growth of the geen plant. p. 153-174. In M.B. Teaser (ed.)
Physiological basis of crop gowth and development. ASA, CSSA, SSSA, Madison,
WI.

Bumb, BL. 1993. Trards in fatiliza use and production in Sub-Saharan Afiica, 1970-1995:
An overview. Fert. Res. 28:41-48.

Casanova, E., and PR Solorzano. 1994. Sorghum and soybean response to natural and
modified phosphate rock on acid soils in Venezuela. Comm. Soil Sci. Plant Anal.
25:215-224.

Chesworth, W., J.MR Semoka, P. van Straateen, P.N.S. Mkeni, J.A Kamasho, and E.P
Mchihiyo. 1988. Tanzania-Canada agogeology project. A report on completion of
the first phase. Univ. of Guelph. Guelph, Canada.

Chien, SH. 1978. Interpretation of Bray-1 extractable phosphorous from acid soil treated
with phosphate rocks. Soil Sci. 126234-39.

Chien, S.H., G. Carmona, RG. Menon, and D.T. Hellums. 1993. Effect of phosphate rock
sources on biological nitrogen fixation by soybean. Fert. Res. 34:153-159.

Chien, SH, and LL. Hammond. 1978. A comparison of various laboratory methods for
predicting the agonomic potential of phosphate rocks for direct application. Soil Sci.
Soc. Am. J. 42:935-939.

Chien, SH, and L. L. Hammond. 1988. Agonomic evaluation of partially acidulated
phosphate rocks in the tropics: IFDC’s experience. Paper Series IFDC-P-7. IFDC,
Muscle Shoals, AL.

Chien, SH, and LL. Hammond. 1989. Agonomic effectiveness of partially acidulated
phosphate rock as influenced by soil phosphorous fixing capacity. Plant and Soil
120: 159-164.

Chiar, S.H., L.L. Hammond, and LA Leon. 1987. Long-term reactions of phosphate rocks
with an Oxisol in Colombia. Soil Sci. 44:257-265.

172

Chien, S.H., P.W.G. Sale, and D.K. Friesen. 1990. A discussion of the methods for
comparing the relative efi‘ectivaress of phosphate fertilizers varying in solubility. Fert.
Res. 24:149-157.

Colemam, NT, and G.W. Thomas. 1967. The basic chemistry of soil acidity. In RW. Person
and F. Adams (eds) Soil acidity and liming. Agonomy 12:187-190.

Crafis-Brandner, S.J. 1992a Phosphorous nutrition influence on leaf senescence in soybean.
Plant Physiol. 98:1128-1 132.

Crafis-Brandner, S.J. 1992b. Sigrificance of leaf phosphorous remobilization in yield
production in soybean. Crop Sci. 32:420-424.

Fageria, NR 1989. Efi‘ects of phosphorous on gowth, yield and nutrient accumulation in the
common bean. Trop. Agi. 66:249-255.

Fageria, N.K., V.C. Baligar, and CA Jones. 1991. Growth and mineral nutrition of field
crops. Marcel Dekker, Inc. New York, NY.

Flor, CA, and MT Thung. 1989. Nutrient disorders. p. 571-616. In HF. Schwartz and
MA Pasto-Conales (eds) Bean production problems in the tropics. 2nd ed. CIAT.
Cali, Colombia.

Foth, H, and B.G. Ellis. 1988. Soil Fertility. John “friey and Sons. New York, NY.

Fredeen, AL., LM Rao, and N. Terry. 1989. Influence of phosphorous nutrition on gowth
and carbon partitioning in Glycine max [L] Merr. Plant Physiol.89:225-230.

Gardner, F.P., RB. Pearce, and RL. Mitchell. 1985. Physiology of crop plants. Iowa State
Univ. Press, Ames, IA

Giller, K.E., F. Amjee, and SJ. Brodrick. 1988. Soil factors limiting nitrogen fixation and
production of Phaseolus vulgaris L. in East Afiica. Annual report-1988. Wye
College, Univ. of London, Kent, UK.

Grabau, L.J., D.G. Blevis, and HC. Minor. 1986. P nutrition during seed development. Leaf
senescarce, pod retention, and seed weight of soybean. Plant Physiol. 82:1008-1012.

Hammond, L.L., S.H Chien, and AU. Mokwunye, 1986. Agonomic value of unacidulated
and partially acidulated phosphate rocks indigenous to tropics. Adv. Agon. 40:89-
140.

173

Hellums, D.T., S.H Chien, and J .T. Touchton. 1989. Potential agonomic value of calcium
in some phosphate rocks from South Arnerca and West Afiica. Soil Sci. Soc. Am. J.
53:459-462.

Howell, RW. 1984. Phosphorous nutrition of soybeans. Plant Physiol. 29:477-483.

Hughes, J .C., and RT. Gilkes. 1986. The efi‘ect of rock phosphate properties on the extent
of fertilizer dissolution in soils. Aust. J. Soil Res. 24:209-217.

International Fertilizer Development Center (IFDC). 1990. Minjingu phosphate rock
marketing study. A market plan for Minjingu Phosphate Company (MIPCO), Arusha,
Tanzania. IFDC, Muscle Shoals, AL.

Juo, AS.R,and B.G. Ellis. 1968. Chemical and physical properties of iron and aluminum
phosphates and their relation to phosphorous availability. Soil Sci. Amer. Proc.
32:216-221.

Kanabo, I.AK., and RJ. Gilkes. I988. The effect of soil texture on the dissolution of North
Carolina phosphate rock. J. Soil Sci. 39:191-198.

Khasawneh, F .E., and EC. Doll. 1978. The use of phosphate rock for direct application to
soils Adv. Agon. 302159-206.

Kirk, G.J.D., and PH. Nye. 1986. A simple model for predicting the rate of dissolution of
sparingly soluble calcium phosphate in soil 1. The basic model. J. Soil Sci. 37 2529-
540.

Kpomblekou-A, K, and M.A Tabatabai. 1994. Metal contents of phosphate rocks. Comm.
Soil Sci. Plant Anal. 2522871-2882.

Lana, MJ., and D.G. Blevins. 1989. Dry matter accumulation and phosphate accumulation
in soybean gown on varying levels of phosphorous nutrition. J. Plant Nutr. 1221045-
1060.

Lauer, M.J., S.G. Pallardy, D.G. Blevins, and DD Randall. 1989. Whole leaf carbon
exchange characteristics of phosphate deficient soybeans (Glycine max L.). Plant
' Physiol. 91:848-854.

Le Mare, RH 1991. Rock phosphate in agriculture. Expl. Agic. 27 2413-422.

174

Lindsay, W.E., and EC. Moreno. 1960. Phosphate equilibria in soils. Soil Sci. Soc. Am.
Proc. 242177-182.

Lindsay, W.L., and HF. Stephenson. 1959. Nature of the reactions of monocalcium
phosphate mono-hydrate in soils: 1. The solution that reacts with the soil. Soil Sci.
Soc. Proc. 23:12-19.

Lynch, J., A Lauchi, and E. Epstein. 1991. Vegetative gowth of common bean in response
to phosphorous nutrition. Crop Sci. 31 :380-387.

Manrique, LA 1986. Field bean leaf developmart and seed yield as afl‘ected by phosphorous
application in Panama. Comm. Soil Sci. Plant Anal. 17:135-147.

Manrique, LA 1993. Radiation-use-efiiciency of field beans as affected by soil phosphorous
management. Comm. Soil Sci. Plant Anal. 242623-631.

Mchihiyo, E.P. 1991. Phosphate potential in Tanzania. Fert. Res. 30:177-180.

McClellan, G.H, and JR Lehr. 1969. Crystal chemical investigation of natural apatites.
Amer. Mineralogist. 54:1374-1391.

McLean, E.O., and T.J. Logan. 1970. Source of phosphorous for plants gown in soils with
difi‘erent phosphrous fixation tendencies. Soil Sci. Soc. Am. Proc. 34:907-911.

McLean, E.O., and RW. Wheela. 1964. Partially acidulated rock phosphates as a source of
phosphOrous to plants. I Grth chamber studies. Soil Sci. Soc. Am. Proc. 28:545-
550.

Margel, K, and EA Kirldry. 1987. Principles of plant nutrition. 4th ed. International Potash
Institute, Beme, Switzerland.

Michigan State University. 1993. User’s guide to MSTAT-C. A software for desigr,
management and analysis of agonomic research experiments. Mich. State Univ., East
Lansing, MI.

Miyasaka, S.C., J.G. Buta, RK. Howell, and CD. Foy. 1991. Mechanism of aluminum
tolerance in snapbeans: Root exudation of citric acid. Plant Physiol. 96:73 7-743.

Mullins, G.L. 1988. Plant availability of P in commercial superphosphate fertilizers. Comm.
Soil Sci. Plant Anal. 19:1509-1525.

17S

Mullim, G.L., F .T. Sikora, and J .C. Williams. 1995. Efl‘ect of water soluble phosphorous on
the efl‘ectivaress of triple supap'hosphate fatiliza. Soil Sci. Soc. Am. J. 59:256-260.

Mwambete, I. 1991. Tanzania fertilizer mineral deposits. Fat. Res. 302181-185.

Nienhuis, J., and RS. Singh. 1986. Combining ability analysis and relatioship among yield,
yield components and architectural traits in dry beans. Crop Sci. 26:21-27.

Radford, RJ. 1967. Growth analysis-their use and abuse. Crop Sci. 72171-175.

Ritchie, G.S.P., and D.M Weava. 1993. Phosphorous retention and release fiom sandy soils
of the Peel-Harvey catchment. Fert. Res. 36:115-122.

88, T-M., and D.W. Israel, 1995. Nitrogen assimilation in nitrogen-fixing soybean plants
during phosphorous deficiency. Crop Sci. 352814-820.

Saito, M., and T. Kato. 1994. Relationships between dry matter producrion, nitrogen and
phosphorous accumulation in soybean gown under cool weather conditions. Soil Sci.
Plant Nutr. 40273-81.

Sale, P.W.G. 1989. The agonomic efficiency, the econonrics and the quality of direct applied
gound rock and partially acidulated phosphate rock. p. 186-202. In The role of
phosphates in balanced fertilization. Proc. Senrinar, Marrakech, Morocco, 24-27
October, 1989. IFA-IMPHOS, France.

Sale, P.W.G., and AU. Mokwunye. 1993. Use of phosphate rocks in the tropics. Fert. Res.
35233-45.

Sanchez, RA, and J.G. Salinas. 1981. Low-input technology for managing Oxisols and
Ultisols in tropical America. Adv. Agon. 34:279-406.

Sandell, 15.13., MH Hey, and D McConnell. 1939'. The composition of fiancolite. Min. Mag.
25:395-401.

Sarafi, A 1978. A yield component selection experiment invoving American and Iranian
cultivars of common bean. Crop Sci. 1825-7.

Sexton, P.J., J .W. White, and KJ. Boote. 1994. Yield-determining processes in relation to
cultivar seed size of common bean. Crop Sci. 34284-91.

176

Shibata, J.K., and J.C. da Costa, and CS. Miranda 1980. Harvest index in Phaseolus
vulgaris L. Ann Rep. Bean Improv. Coop. 23 287-89.

Sirait, Y., W.G. Pill, and we Kee, Jr. 1994. Lima bean (Phaseon lunatus L.) response to
irrigation regime and plant population density. Hort Sci. 29:71-73.

Smithson, J.B., O.T. Edje, I. Kullaya, J. Maimu, C.S. Mushi, S. Brodrick, and K.E. Giller.
1990. Diagrosis and correction of soil nutrient problems of bean in the Usambara
Mountains of Tanzania. p. 560-575. In J .B. Smithson (ed.) Bean Res. Vol. 5. Afiica
Workshop Series. No.12. CIAT, Cali, Columbia

Smyth, T.J., and RA Sanchez. 1982. Phosphate rock dissolution and availability in Carado
soils as affected by phosphorous sorption capacity. Soil Sci. Soc. Am. J. 46:339-345.

Syers, J .K., and AD. Mackay. 1986. Reactions of Sechura phosphate rock and single
Ssperphosphate in soils. Soil Sci. Soc. Am. J. 50:480-485.

Thung, M 1991. Bean agonomy in monoculture. p. 737-834. In A Van Schoonhoven, and
O. Voysest (eds) Common beans: Research for crop improvement. CAB.
International, Wallingford, U.K

Uriyo, AP., and A Kesseba. 1973. Phosphate fractions in some Tanzania Soils. Geoderma.
10:181-192.

Van Kauwenbergh, SJ. 1985. Cursory characterization of phosphate ore and concentrate
sample fiom Minjingu, Tanzania. Report for the Development of Indigenous Deposits
Training Progarn, May 20- June 6, 1985. IFDC, Muscle Shoals, AL.

Van Kauwenbergh, SJ. 1991. Overview of phosphate deposits in East and South-East Afiica.
Fert. Res. 302127-150.

Vanderlip, RL. 1972. How a sorghum plant develops. Coop. Ext. Serv., Kansas State Univ.,
Manhattan, KS.

Whiteaker, G., G.C. Gerloff, W.H Gabelman, and D. Lindren. 1976. Interspecific difi‘erences
in gowth of beans at stress levels of phosphorous. J. Amer. Hort. 101:472-475.

Wilson, M.A, and B.G. Ellis. 1984. Influence of calcium solution activity and surface area
of the solubility of selected rock phosphates. Soil Sci. 138:354-359.

177

Wright, RJ. 1989. Aluminum toxicity and plant gowth. Comm. Soil Sci. Plant Anal.
20: 1479-1497.

Yan, X, JP. Lynch, and SE. Beebe. 1995a. Genetic variation for phosphorous efficiency of
common bean in contrasting soil types: I. Vegetative response. Crop Sci. 3521086-
1093.

Yan, X, S.E. Beebe, and JP. Lynch 1995b. Genetic variation for phosphorous efficiency of
common bean in contrasting soil types: 11. Yield response. Crop Sci. 35:1094-1099.

Young, RD., G.D. Westfall, and G.W. Colliver. 1985. Production, marketing and use of
phosphorous fertizas. p. 323-376. In O.P. Englelatad (ed.) Fertilizer technology and

use. 3rd ed. ASA CSSA, SSSA, Madison, WI.

CHAPTER 5
RESIDUAL EFFECTS OF PHOSPHOROUS APPLICATION ON

BEAN PRODUCTION UNDER RAINFED AND IRRIGATED CONDITIONS

ABSTRACT

There is no documented information on residual efl‘ectiveness due to P fertilizer
application on beans in Tanzania. Therefore, a two year experiment was set up at Sokoine
University of Agiculture in Morogoro, Tanzania to evaluate the efi‘ects of direct, and residual
values of triple superphosphate (TSP) and Minjingu phosphate rock (MPR) on bean gowth
and yield under rainfed and irrigated conditions. For the direct efi‘ects plots received annual
fertilizers in the 1993 and the 1994 gowing seasons. Different plots received fi'esh P during
both gowing seasons. For residual effects, P applications were made once in the 1993
gowing season. Comparison of bean performance was done using parameters from fi'eshly
applied P, annual P and residual P in 1994. Each block had three replications, and each
replication received seven P treatments. All fertilizer treatments were broadcast and
incorporated in each plot at planting. Basal N application was made at 80 kg ha" in form of
the urea. The N was split applied at planting and at flowering. Growth parameters were

takar at a two week interval. Phosphorous concentration in plant tissues was analyzed at the

178

1 79
same intaval. At maturity, total dry matta, seed yield, and yield components were analyzed.
Results showed that P uptake increased with increasing application of P rates The TSP
fertilizer effect on leaf P concentration at R, gowth stage was similar to that of MPR
fertilizer. Using Duncan multiple range test (DMRT at P=0.05) TSP had similar effects on
total dry matter (I'MD) production under flesh and residual P. However, under orthogonal
comparisons (P=0.05) results showed that TSP performed better than MPR at 44 and 88 kg
P ha" under rainfed conditions. Under irrigated conditions, TSP and MPR had no significant
agonomic efi‘ectiveness. The relative agonomic efl‘ectiveness (RAE) values of MPR with
respect to TSP (RAE=1,00%) in terms of influencing TDM production were 12% under
rainfed conditions, and 41% under irrigated conditions. Results showed that the effect of
soluble P (TSP) fertilizer were higher in terms of influencing TDM production than that of
MPR fatilizer one year after their application under rainfed conditions when equal amounts

were compared.
LITERATURE REVIEW

Residual efi‘ects of plant nutriarts are defined as influence on crops following the crop
to which the nutrients are applied (Black, 1993). When P fertilizer is applied in most soils,
P is retained close to the site of application. Unlike nitrate, which has little amnity for soil
particles, fatiliza P is rapidly adsorbed or precipitated in acid mineral soils. Much of the P
that is applied to the soil is rapidly, and irreversibly fixed by adsorption hysteresis (Pierzynski
et al.,1994). Adsorbed P is normally considered to be slowly available, and capable of

gadually replarishing the soil solution in response to plant uptake by desorption Mechanisms

180

for P 1088 are physical erosion of the top soil, and loss of solution P in surface runofi‘, while
leaching is minimal (Schuman et al., 1973). Therefore, P is a long-lasting nutrient because
of its strong interaction with soil (Black, 1993).

Thaearegenerallythreezonesofreactionsbetweenthe soil and aP fertilizer ganule
(Barrow, 1980). For example, when monocalcium phosphate (Ca(H,PO.),.H,O) is applied
to the soil the fertilizer particle is surrounded by soluble P available to the plants. The
insoluble phosphates are then formed with Ca”, or Fe 3’, and Al 3* cations depending on the
soil pH (Lindsay and Moreno, 1960). At this stage, the P is still readily available to plant
roots since most of it is located at the surface where exudates from the plant can encourage
exchange. In time, the P penetrates the crystal and a small portion is found near the surface.
Under such conditions P becomes fixed (Brady, 1990). The second zone of reaction which
has complexes of Ca”, or Fe” or AI’ and P, contains the residue of P fertilizer (Barrow,
1980)

Organic matter may decrease or increase the ability of soils to adsorb P. This ability
depends on the organic matta and the association with cations such as Fe, Al, and Ca, which
are capable of adsorbing P while associated with organic matter (Moshi et al., 1974; Sample
er al., 1980; Kalkafi er al., 1988). Under practical conditions, the residual effectiveness of
the applied fertilizer is influenced by fertilizer type, nutrient type, time, management,
environmental fact0rs, the crop that removes some of the nutrient supplied and finally, the
inter-relationships of these factors (Devine et al., 1968; Bolland etal., 1988).

Experimental results by Juo and Ellis (1968) showed that when soluble P is applied
to acid soils, P precipitated rapidly to form colloidal Al-P and Fe-P. Because of their smaller

particle sizes, geater surface area and amorphous structure, these colloidal forms were

1 8 1

readily available to plants. Howeva, these colloidal phosphates crystallized to form hydrated
compounds such as varisite (AIPO,.2H,O), and strargite (F ePO,.2H,O) which were much less
available for plant uptake. From their study results, they concluded that, during the first stage
of P fixation, the rate of crystallization was the factor that controlled the relative availability
of P from the variscite and strengite in acid soils. With time, strengite crystallized at a much
faster rate than variscite under favorable arvironmental conditions. Thus, in the second stage
of P fixation, the relative pOrtion of the colloidal form to the crystalline form was the factor
that controlled the relative availability of the veriscite and strengite in the acid soils. In the
third and final stage, when both forms had completed the process of crystallization, specific
surface and crystal structure became important in controlling P availability. Devine
er a1. (1968) obtained results that showed that the effectiveness of ganular dicalcium
phosphate was geater one year after application than when freshly applied. Some studies
have shown that phosphate rock fertilizers persist in the soil up to 40 years longer than
sohrble P fertilizers. However, most experimental work show that residual effects of soluble
P fertilizers are usually geater than those from PRs in the first three - four years after
application when equal amounts ofP are applied (Khasawneh and Doll, 1978; Le Mare, 1991;
Choudhary et al., 1994).

Reports on residual efl‘ectiveness due to P fertilizers in Tanzania are scarce, and where
such studies have been conducted P uptake has not been measured. Among the few studies
reported is that by Marandu et a1. (1975) on NPK fertilizer applied on pasture and maize on
ferrisol soils at Morogoro. In other studies using double superphosphate (DSP), and MPR
fertilizers, positive residual effects have been reported by Scaife (1968) with cotton on

granitic, sandy soil at Ukirigunr. Higher residual efi'ects by MPR than water soluble P

182
fertilizers in the second or third season after application in a maize crop was reported by
Mowo and Gama (1988). At Katumani in Kenya, MPR also showed a substantial residual
value on a maize crop compared to TSP and single superphosphate (SSP) two to five years
after application (Bromfield et al., 1981; IFDC, unpublished report, 1990‘).

A major problem encountered by bean producing farmers in Tanzania is the high
amounts of P fertilizer application needed to meet the crop's requirements. The most
common practice of correcting this P deficiency is by applying a heavy initial P fertilizer
(>50kg P20, ha") to the crop every gowing season. Such a recommendation is expensive
to the resource poor-farmers of Tanzania. Therefore, the hypothesis in this study is that
because (I) the P fatilizer types and rates recommended for application, and (ii) the soil type
on which the bean crop is gown in Tanmnia, the P fertilizers applied can result in substantial
residual effects for the crop. Thus, this study was initiated with the aim of assessing the
residual effectiveness of MPR relative to TSP on bean (Phaseolus vulgaris L.) crop gowth,
and yield during the successive season Results on residual efi’ects of fertilizer P fi'om a single

cropping season following its application are reported.

MATERIALS AND METHODS

Details on materials and methods are as given in Chapter 2 (for blocks A and B).
When this study was initiated in the 1993 gowing season, an extra block C was also set up

so that it could be used as a control block in the 1994 gowing season (i.e., to provide a

 

6Minjingu phosphate rock marketing study: A marketing plan for Minjingu Phosphapte Co. (MIPCO)
in Arusha, Tanzania International Fertilizer Deveth Center, PO. 2040, Muscle Shoals, AL 35660.

183

comparison with the flesh P application). Therefore, block C received fertilizer P only in the
successive gowing season of 1994. Blocks A and B received the P treatments at sowing in
1993.

Bean response assessment for residual P was made by taking crop gowth and yield
measuranents one year following the P application. This was done by taking measurements
fi'om the second-year crop response on the plots in block A, in relationship to the plots in the
control block (C) that received fiesh P fertilizer in the 1994 gowing season. Plots in block
B also served as controls in the first (1993) gowing season. However, an additional
application of P fatilizer was made in the second gowing season of the experiment to block
B.

Initial soil pH, extractable P, and exchangeable Ca levels were analyzed in blocks A
and B at the beginning of the 1993 gowing season (Tables 2.1), and at the beginning of the
1994 gowing season (Table 4.2 a); and for block C as indicated in Table 5.1. The soil
analysis of samples takar immediately before planting the first crop in 1993 for blocks A and
B (Table 2.1) and in 1994 for block C (Table 5.1), are considered to be more representative
of the soil chemical environment during the experiments reported in this study. In the 1994
gowing season, P fertilizer treatments were broadcast, and incorporated at planting at both
sites in blocks B and C. Post-harvest soil chemical characteristics for blocks A and C were

taken seven days afier harvest (Table 5.2).

184

Table 5.1. Pro-plant soil charactaistics at 0-12 cm layer for Block C at the experimental sites
during the 1994 season.

 

 

 

Experimental site
Soil characteristic Rainfed Irrigated
Moisture (%)' 17.14 21.10
pH:H,O 4.80 5.30
Bray 1 (ppm) . 1.20 2.20
Ca (meq 100g" soil) 2.39 3.99

 

'Volumetric moisture content.

Grth parameters such as total dry matter (TDM kg ha "), leaf area index (LAI),
plant height (cm), and crop gowth rate (CGR) were determined as described in Chapter 2
and Chapter 4. Plant samples were collected, prepared, and analyzed for nutrients as
described in Chapter 3. TDM, seed yield (kg ha"), and yield components; and relative
agononric efi‘ectiveness (RAE) were determined as given in Chapter 4. Statistical analyses
were conducted using MSTAT-C Statistical Package (Michigan State University, 1993).
Lastly, crop performance from block A (residual P), block B (annual P), and block C (fresh
P) were compared at the end of the 1994 gowing season. Orthogonal comparisons were

used to compare the agonomic influence of TSP versus that of MPR.

185

RESULTS AND DISCUSSION
Effects of Residual P Treatments on Soil pH,

Extractable P, and Exchangeable Ca.

The efl‘ects of TSP and MPR application were similar to those discussed in Chapter
4. With the exception of the control plots, soil pH, available P, and exchangeable Ca
increased in plots that received P fertilizers (Table 5.2). Such results may have been due to
the interaction between the P fertilizer used, and the soil characteristics as indicated by
Anderson et al. (1985), Chiar er al. (1987), Robinson et al. (1992), and Sale and Mokwunye

( 1993).
Phosphorous Concentration

The uptake of P was higher in fieshly applied P blocks than in residual P conditions;
and higher in irrigated experiment than in the rainfed experiment (Tables 3.2 and 5.3).
According to Piggot (1986), high levels of P fertilizer resulted in what can be regarded as
adequate P concentration at flowering. Fresh P had higher P concentration than residual P
under rainfed conditions. Patterns of P concentration, distribution, and accumulation were
similar to those discussed in Chapter 3. High concentration values of P observed in this study
may have been due to the small leaf size (area) as indicated by valuesof LAI (Table 5.3).
Such a condition may sometimes result in accumulation of P concentration values of the

limiting nutrient (Hiatt and Massey, 1958, Adu-Gyamfi er al., 1990). Although these results

186

show that plants had high leaf P concentration, there is no indication that the high P

application in either fiesh, annual or residual applied P satisfied the crop requirement for P.

Table 5.2. Soil pH, Bray-I P and calcium levels in post-harvest soil at 0-12 cm layer in
Blocks A and C at the end of 1994 gowing season.

 

 

Treatment Soil pH. Bray I Ca
(rpm) (meq)
Rainfed experiment
A B A B A B

0 kg P ha" 4.75 4.81 1.43 1.65 1.91 2.01
22 kk P ha" TSP 4.81 4.92 3.12 2.35 2.29 2.78
22 kg P ha" MPR 4.78 5.01 3.01 3.37 3.82 3.63
44 kg P ha" TSP 4.69 5.12 3.90 3.01 2.57 3.77
88 kg P ha" TSP 4.82 5.01 3.23 3.69 2.92 3.59
88 kg P ha" MPR 4.65 4.89 5.40 4.94 3.31 3.89
176 kg P ha" MPR 4.64 4.98 7.32 6.91 2.96 3.94

Irrigated experiment
0 kg P ha" 5.49 5.40 1.80 1.64 2.16 3.31
22 kk P ha" TSP 5.50 5.43 2.07 2.35 2.07 3.81
22 kg P ha" MPR 5.74 5.47 2.42 2.37 3.54 3.89
44 kg P ha" TSP . 5.65 5.65 2.01 3.01 3.99 3.93
88 kg P ha" TSP 5.87 5.60 3.17 3.69 3.24 3.81
88 kg P ha" MPR 5.96 5.53 4.10 4.94 4.05 4.08
176 kg P ha" MPR 5.53 5.53 6.31 6.91 5.05 4.89

 

A: represents data fiom block A
B: represents data fiom block C

187

Table 5.3 Influence of residual P on TDM and seed production; P concentration and
maximum leaf area index.

 

 

Treatment Residual P TDM Seed LeafP at R, Max.
Bray I (ppm) kg ha" kg ha" (%) LAI
Rainfed experiment
Triple superphosphate
0 kg P ha" 1.21 1060 (544)° 289 e 0.29 1.05
22 kg P ha" 3.89 1950 (234) 571 de 0.35 1.45
44 kg P ha" 5.45 2499 (458) 852 abc 0.29 1.49
88 kg P ha" 9.35 3314 ( 99) 815 a 0.36 1.54
Minjingu phosphate rock
0 kg P ha" 1.21 1060 (544) 289 e 0.29 1.05
22 kg P ha" 6.89 2124 (165) 693 bcd 0.32 1.85
88 kg P ha" 10.65 2608 (318) 833 ab 0.42 1.50
176kg P ha" 11.94 2006 (297) 688 cd 0.33 1.24
Irrigated experiment
Triple superphosphate
0 kg P ha" 2.53 1232 (109) 347 b 0.25 1.03
22 kg P ha" 4.03 1670 (251) 672 a 0.30 1.05
44 kg P ha" 5.61 1921 (415) 761 a 0.40 1.49
88 kg P ha" 7.31 1879 (127) 857 a 0.41 1.13
Minjingu phosphate rock
0 kg P ha" 2.53 1232 (109) 347 b 0.25 1.03
22 kg P ha" 5.91 1907 (361) 703 a 0.38 1.54
88 kg P ha" 8.66 2124 (523) 757 a 0.39 1.33
176kg P ha" 12.45 1910 (137) 703 a 0.41 1.55

 

' Values in parartheses represent standard errors of the means. Means within the column for
seed yield not followed by the same letter are sigrificantly different at 0.05 level using
Duncan’s Multiple Range Test.

188

Crop Growth, LeafArea Index, and Total Dry Matter Production.

The seasonal pattern of gowth and development were similar to those discussed in
Chapta 4. Under rainfed conditions, the CGR value ranged from 3.15 g m'2 day" in the
control plots to 10.4 g m’2 day" in the plots that received 22 kg P hi? (MPR). Under
irrigated conditions, the CGR values ranged fi'om 3.25 g m’2 day" in the control plots to 12.4
g m "day " in plots that received 44 kg P ha " (TSP). The lowest CGR values in the control
plots had the lowest LAI, TDM and seed yield. However, the highest CGR values did not
resultinhighestLAJ, TDMorseedyield (datanot shown). Residual P effects on TDM, seed
yield, and LAI production are shown in Tables 5.3. The irrigated experiment resulted in
higher TDM production than the rainfed experiment. Other variables such as plant population
(plants m"), leaf number on main stem and plant height (cm), all followed similar trends as
discussed in Chapter 2 and Chapter 4.

The LAI values fi'om the residual P blocks were lower than those measured from the
annual, and fieshly applied P plots. Maximum LAI values ranged fiom 1.05 in control plots
to 1.85 in plots that received 22 kg P ha" in the form of MPR under rainfed conditions.
Under irrigated conditions, maximum LAI values ranged fi'om 1.03 in control plots to 1.55
in plots that received 176 kg P ha" in the form of MPR Such low values of leaf area may
have influenced the high values of P concentrations observed in this study as suggested by
Hiatt and Massey (1958). Furtha, since leaf area is critical for crop light interception (Shibles
and Weba, 1965), the low LAI observed here may have substantially influenced crop gowth

and total dry matter yield.

189

Seed Yield Components.

Efi‘ects of residual P on seed yield plant", and yield components are shown in Table
5.4. Fertilizer TSP had the highest yield plant" under rainfed conditions, ranging from 6.7
to 11.1 3 plant". On the other hand, MPR resulted in the seed yield plant" ranging from 8.2
to 10.3 g plant ".Since seed yield in bean is expressed as a product of pods plant", seeds pod“
‘ and seed weight (F ageria et al., 1991), current results show that the low yield per plant
observed under irrigation may have been due to the low number of seeds pod".

Seed yield (kg ha") was statistically non-sigrificant within treatments, except in the
control plots which had lowest values (Table 5.3). Such results may have been influenced by
the diseases that attacked the seeds while the crop was still in the field. Ascochyata blight,
a firngal disease caused by Phoma spp,was an important disease especially in the irrigated
experimart that matured under cool tanperature, high humidity and had abundant water fi'om
irrigation. All these made the environment conducive for the disease.

Table 5.5 illustrates the RAE of flesh and residual P fi'om MPR as compared to TSP.
The influence of flesh P in terms of TDM production were similar in both experiments.
However, unda residual P the rairrfed experiment had a lower RAE (12%), and the irrigated
expaiment had a higher RAE (41%). Such results confirm earlier observations reported in
Chapta 4 that MPR efl‘ectiveness in TDM production was higher under irrigation conditions.
These results, howeva, represart a relatively short period of time for an evaluation of residual
effectiveness since the second crop was gown in a successive season limiting the period to

only one year.

190

Table 5.4. Yield components fora bean crop gown with residual P in the 1994 season.

 

 

 

 

 

 

Treatments Pods plant" Seeds pod" Seed wt Total seed wt
(g 100 seed") (g plant")
Rainfed experiment
0kgPha" 5.5 (0.7)' 4.0 (1.0) 28.4 (1.5) 6.2 (2.8)
22kgP ha"-TSP 5.4 (0.9) 3.8 (0.3) 32.2 (2.3) 6.7 (1.2)
22kgP ha"-MPR 7.2 (0.9) 4.0 (0.4) 28.3 (0.2) 8.2 (2.3)
44kgP ha"-TSP 7.8 (0.7) 4.3 (0.6) 33.2 (2.9) 11.1 (1.3)
88kgPha"-TSP 7.7 (0.6) 4.1 (0.6) 35.6 (1.5) 11.1 (1.9)
88kgP ha"-MPR 6.3 (5.5) 5.0 (0.6) 32.6 (0.7) 10.3 (1.9)
176kgP ha"-MPR 7.0 (0.2) 4.5 (0.3) 31.1 (2.1) 9.8 (0.4)
Irrigated experiment
0kgPha" 4.9 (0.7) 3.6 (0.3) 35.3 (1.5) 6.2 (1.6)
22kgP ha"-TSP 6.8 (0.1) 4.0 (0.3) 33.1 (2.7) 8.9 (0.6)
22kgP ha"-MPR 4.9 (0.9) 3.9 (0.3) 37.4 (0.6) 7.2 (0.6)
44kgP ha"-TSP 7.1 (1.4) 3.9 (0.4) 36.9 (2.1)10.3 (1.9)
88kgP ha"-TSP 8.1 (0.8) 4.0 (0.4) 33.9 (2.5) 10.9 (1.5)
88kgP ha"-MPR 6.5 (1.1) 3.7 (0.4) 37.2 (2.7) 8.9 (1.7)
176kgPha"-MPR 7.0 (0.8) 3.3 (0.3) 38.6 (2.1) 8.9 (1.9)

 

'Figures in parentheses show the standard error values.

191

Table 5.5. Values of coeficients (fl) and relative agronomic effectiveness (RAE) of Minjingu
phosphate rock and triple superphosphate in terms of bean total dry matter production fiom
flesh and residual P in the 1994 season.

 

 

 

Rainfed experiment
Fresh applied P Residual P
13. we r. RAB
MPR—37 34.5 — 3. _ 12.
TSP 10.3 100. 24.8 100.
Irrigated experiment
MPR 5.6 34.4 2.8 41.

TSP 16.3 100. 6.7 100.

 

Y, = Y, + flInX, X>1
where

Y, = Yield obtained with source (i)

Y, = Yield obtained with no P application
A = Regession coefiicient source

X = Rate of P applied

Under rainfed conditions orthogonal comparisons showed that TSP had a higher
agonomic efi‘ect than MPR on TDM production at 44 and 88 kg P ha", but lower efi'ect at
22 kg P ha" (Table 5.6). However, under irrigated conditions, TSP had similar efi‘ects as
MPR on TDM production The higher residual effects on yield fi‘om soluble P fertilizer (TSP)
than MPR in the first year afia application shown in this study agee with the results reported

by Khasawneh and Doll (1978); Le Mare (1991); and Choudhary et al. (1994). Fresh applied

192

P, in the form of TSP, resulted in a significant effect in three of the nine comparisons (two of

these unda irrigated conditions), while MPR performed better than TSP at only 22 kg P ha"

under irrigated conditions.

Table 5.6. Summary of orthogonal comparisons between triple superphosphate and Minjingu
phosphate rock in agonomic effectiveness fi'om fieshly applied and residual P for the 1994

 

Freshly applied P

 

Treatment comparisons

 

22 kgPTSP vs 22 kgPMPR
88kgPTSva88 kgPMPR
44kgPTSva 176kgPMPR

Residual applied P
22 kg P TSP vs 22 kg P MPR

88 kgP TSP vs 88 kgP MPR
44 kgP TSP vs 176 kgP MPR

 

 

Experiment

Rainfed Irrigated
A= ' A<B‘
A=B' A>B
A>B A>B
A<B° A= °
A>B' A=B‘
A>B° A=B'

 

' A=B based on DMRT (P=0.05)
A represents triple superphosphate (TSP)
B represents Minjingu phosphate rock (MPR)

 

193

Comparison of responses of a bean crop to residual, fresh, and annual P application
are shown in Figure 5.1 for TSP fertilizer and in Figure 5.2 for MPR fertilizer. The response
curves of residual applied P were initially steep for rainfed experiments in. both seasons, but
were almost flat and no response under the irrigated conditions of the 1993 gowing season.

With the exception of the 1994 gowing season under the irrigated conditions, the response
curvesofannual appliedeae steep. ResponseofbeancroptoTSPunderall three fertilizer
regimes was consistart with the fatilizer amount applied. The exception was with the results
of the 1994 gowing season where the rainfed experiments resulted in higher TDM
production than did the irrigated plots. As indicated earlier, such results may have been due
to the attack by rodents which resulted in lower plant population at harvest maturity. In
gareraL the responses of bean to MPR were inconsistent. For examme, the response curve
under rainfed conditions in 1993, and that under irrigation in 1994 show that a yield plateau
was not reached. It was also interesting to observe that there was a progessive decline in
TDM production with higher rates of MPR for rainfed and irrigated conditions under residual
P (Figure 5.2).

The results fiom this were unexpected. As the extracted soil P increased, the crop
yield (TDM) decreased. Such phenomenon of decreasing crop yield with increasing available
P needs more investigation. It should be pointed out, however, that yields recorded in
residual plots at irrigated site could have been influenced by flooding effects that took place
in the previous season. The floods might have contaminated the plots by moving the P
treatments fi'om one plot to another (soil carryover). These results show that there was no

clear pattern of response of MPR under residual and annually applied fertilizer P.

 

194

3000

 

(a) Residual P

2000 2

  
 

 

 

 

1000

 

3000 ~

 

 

2000 -

 

 

1 000

 

40003
3000 3
20003
1000

 

 

Total dry matter production (kg ha")

 

:1
i

 

   

 

 

0 20 40 60 80 100
Rate of P applied (kg P l‘a ")

TDM rainfed 1993
TDM rainfed 1994
TDM irrigated 1993
TDM irrigated 1994

45..

Figure 5.1. Comparison of responses of bean crop to residual, flesh, and annual application
of difl'erent levels of TSP under rainfed and irrigated conditions.

 

195

 

3000 , r - i . r
(1) Residual P

    

 

 

 

 

 

 

 

 

 

 

Total dry matter production (kg ha")

 

 

 

 

 

0 50 1 00 1 50 200
Rate of P applied (kg P lla ")

O TDM rainfed 1993
I TDM rainfed 1994
A TDM irrigated 1993
v

TDM irrigated 1994

Figure 5.2. Comparison of responses of bean crop to residual, fresh, and annual application
of difi‘erent levels of MPR under rainfed and irrigated conditions.

196

SUMMARY AND CONCLUSIONS

Generally, annual applied P was more efi‘ective than fresh and residual P in TDM
production. Significant difi‘erences between the residual efi‘ects of MPR and TSP sources
were evident using orthogonal comparisons. At 44 and 88 kg P ha" residual agonomic
elfectivaress of MPR was lower than that of TSP 12 months afier application under rainfed
and irrigated conditions. With fresh applied P, TSP had sigrificant effects in most
comparisons while MPR performed better than TSP only at 22 kg P ha" under irrigated
conditions. From orthogonal contrasts and RAE values, it is concluded that MPR could be
used to a better advantage for bean crop gown under irrigated conditions than rainfed
conditions.

The unexpected, inconsistent behavior of response to residual MPR one year after
application in the two experiments provides a sound basis for filrther research on residual P
relationship with between response of bean crop to MPR, soil properties and crop
management. Bean seed yields reported fi'om residual P were higher than those usually
recorded by most farmers who usually practice annual fertilizer P application. Farmers' low
yields usually average 600 kg/ha (Silbemagel and Teri, 1990) and could be due to late
planting, use of low yielding varieties, poor soil fertility, water stress, vermin, pest, disease,

and roaming livestock attack which are common field hazards.

 

197
RECOMMENDATIONS FOR FUTURE RESEARCH

The present short-term results show inconsistences when rainfed and irrigated
experiments are compared. This suggests that residual P efl‘ectiveness on bean yield should
be continued so as to obtain information on the long-term fate of TSP and MPR when applied
to the acid soils in Tanzania This should include the idartification of form (s) of P availability
to plants relative to seasons alter fertilizer(s) application. Knowledge fi'om efi‘ects of long-
terrn P fertilization may help in understanding how such a practice may influence the

sustainability of Tanzania's agiculture.

BIBLIOGRAPHY

Adu-Gyamfi, J .J., K Fujita, and S. Ogata. 1990. Phosphorous fractions in relation to gowth
in pigeon pea (Cajanas cajan (L.) Millsp.) at various levels of P supply. Soil Sci.
Plant Nutr. 36:531-543.

Anderson, D.L., WR Kussow, and RB. Corey. 1985. Phosphate rock dissolution in soil:
indications fi'om plant gowth studies. Soil Sci. Soc. Am. J. 492918-925.

Bolland, M.D.A, RJ. Gilkes, and FF. D’Antuono. 1988. The effectiveness or rock
phosphate fatilizers in Australian agiculture: a review. Aust. J. Expl. Agric. 282655-
688.

Barrow, NJ. 1980. Evaluation and utilization of residual phosphorous in soils. p. 333-359.
In F.E. Khasawneh E.C. Sample, and B.J. Kamprath (eds) The role of phosphorous
in agiculture. ASA, CSSA, SSSA Madison, WI.

Black, CA 1993. Soil fertility evaluation and control. Lewis Publishers. Boca Raton. FL.

Brady, NC. 1990. The nature and properties of soils. 10th ed. Macmillan Publ. Co. New
York, NY.

Bromfield, AR, IR Hannock, and DE Debenham. 1981. Efi’ects of rock phosphate and
elanaltal-S on yield and uptake of maize in Westan Karya Expl. Agic. 17:383-387.

Chiar, S.H, L.L. Hammond, and LA. Leon 1987. Long-term reactions of phosphate rocks
with an Oxisol in Colombia. Soil Sci. 1442257-265.

Choudhary, M, TR Peck, L.E. Paul, and L.D. Bailey. 1994. Long-term comparison of rock

phosphate with superphosphate on corn yield in two cereal-legume rotations. Can. J.
Plant Sci. 74:303-310.

198

199

Devine, RJ., D. Gunary, and S.Larsen. 1968. Availability of phosphate as affected by
duration of fertilizer contact with soil. J. Agic. Sci. (Cambridge). 71 :3 59-364.

Fageria, N.K., V.C. Baligar, and CA Jones. 1991. Growth and mineral nutrition of field
crops. Marcel Dekker, Inc. New York, NY.

Hiatt, AJ., and H F. 1958. Zinc levels in relation to zinc content and gowth of corn. Agon.
J. 50:22-24.

Juo, AS.R., and B.G. Ellis. 1968. Chemical and physical properties of iron and aluminum
phosphates and their relation to phosphorous availability. Soil Sci. Soc. Amer. Proc.
322216-221.

Kalkafi, U., B. Bar-Yosef, R. Rosenbergh, and G. Sposito. 1988. Phosphorous adsorption
by kaolinite and montimorillinite. 11. Organic anion competition. Soil Sci. Soc. Am.
J. 52:1565-1589.

Khasawneh, FE, and EC. Doll. 1978. The use of phosphate rock for direct application to
soils. Adv. Agon. 302159-206.

Le Mare, PH. 1991. Rock phosphates in agriculture. Expl. Agic. 27:413-422.

Lindsay, W.L., and EC. Moreno. 1960. Phosphate equilibria in soils. Soil Sc. Soc. Amer.
Proc. 242177-182.

Marandu, W.Y.F., AP.O. Uriyo, and H0. Mongi. 1975. Residual effects on NPK fertilizer
application on yield of maize and pasture on ferrisol at Morogoro. E. Afiic. Agic.
For. J. 402400-407.

Moshi, AO., A Wild, and DJ. Grealland. 1974. Effects of organic matter on the charge and
phosphorous adsorption characteristics of Kikuyu red clay fiom Kenya. Geoderma.
1 12275-285 '

Mowo, J.G. and RW. Gama. 1988. Effect of application of MPR and TSP on seed cotton
yield and soil properties in acid soils of western cotton gowing areas. p. 56-58. In
F.B.S. Kaihura, J. Floor, and ST. Ikera (eds) Proc. of Phosphate Meeting. Tanga,
Tanzania, Aug. 1-3, 1988. Miscellaneous Publications. M9. Mlingano Agicultural
Research Institute. Tanga, Tanzania.

Michigan State University. 1993. User's guide to MSTAT-C. A software for desigt and
analysis of agonomic research experiments. Mich. State Univ., East Lansing, MI.

 

200

Piazynski, G.M., J.T. Sims, and GP. Vance. 1994. Soils and environmental quality. Lewis
Publ. Boca Raton, FL.

Piggott, T.J. 1986. Vegetable crops. p. 148-189. In DJ. Reuter and J.B. Robinson (eds)
Plant analysis: An interpretation manual. Inkata Press. Melbourne, Australia.

Robinson, 18., J.K Syers, and NS. Bolan. 1992. Importance of proton supply and calcium-
sink size in the dissolution of phosphate rock materials of difi‘erent reactivity in soil.
J. Soil Sci. 43:447-459.

Sale, P.W.G., and AU. Mokwunye. 1993. Use of phosphate rocks in the tropics. Fert. Res.
35:33-45.

Sample, E.C., RJ. Sopa, and G.J. Racz. 1980. Reactions of phosphate fertilizers in soils. p.
263-310. In E.J. Khasawneh, E.C. Sample, and B.J. Kamprath (eds) The role of
phosphorous in agiculture. ASA, CSSA, SSSA Madison, WI.

Scaife, A 1968. The effect of cassava fallow and various manurial treatments on cotton at
Ukiriguru, Tanzania. E Afiic. Agic. For. J. 33:231-236.

Shibles, RM, and CR Weber. 1965. Leaf area solar radiation interception and dry matter
production by soybeans. Cr0p Sci. 52575-578.

Schuman, G.E., RG. Spomer, and RF. Piest. 1973. Phosphorous losses fi'om four
agricultural watersheds in Missouri Valley. Soil Sci. Soc. Am. Proc. 37:424-427.

Silbemagel, M.J. and J .M. Teri. 1,990. Tanzania/Washington State University/Sokoine
University of Agriculture: Breeding beans (Phaseolus vulgaris L.) for diseases and
determination of socio-economic impact of smaller-farm families. p.56-59. In
Bean/CRSP (MSU) Proc. of International Research Meeting of Bean/Cowpea
Collaborative Research Support Progam, April 39- May 3, 1990. Mich. State Univ.,
East Lansing, MI.

 

APPENDICES

201

Appendix la. Soil profile description at the rainfed experimental site.

 

Order:

Oxisol soils with pedogenic horizons that are mixtures principally of
kaolinite hydrated oxides and quartz, and are low in weatherable
minerals.

Suborder: Orthox

0-20cm:

20 - 80 cm:

80 - 100 cm:

Reddish brown (2.5 YR3/2), moist, dark reddish brown (2.5
YR3/4); sand loam, weak medium ganular structure, sligthly
hard when dry, fine and medium roots; traces of mica and quartz
fragnents.

Dark red (2.5 YR 3/6), moist; dark red (2.5 YR 3/6), dry; sandy
clay; weak medium structureless, sofi when dry; very few fine
roots; small amount of quartz fragnents; difiirse boundary.

Red (2.5 YR 4/6), moist, clay loam; structureless, very friable
when moist;.traces of roots, some Fe-Mn concentrations.

The watertable was below 100 cm at the time of sampling. The
profile was well drained with permiability classification 0.40
according to Ritchie et al. (1990).

The dominant clay was earlier determined as of Kaolinite type
by Kesseba et al. (1972).

 

202

Appendix 1b. Soil profile description at the irrigated experimental site.

 

Order:

Suborder:

0 - 20 cm:

20 - 40 cm:

40 - 80 cm:

80 - 100 cm:

Alfisol: Gray to brown surface horizon.

Udalf: Seasonally saturated with water; when drained beans,
maize gow well. Moderately slapping.

Dusty red (2.5 YR 3/2) moist, dark gayish brown (10 YR 3/2);
sand clay loam, weak medium ganular structure; nonstick when
dry; plastic when wet; coarse roots, fine roots, Fe concentrations,
small amounts of yellowish (10 YR 5/6) and reddish (2.5 YR
4/6) material scattered throughout the horizon, clear smooth
boundary.

Dark geyish brown (10 YR 3/2); sand clay loam, structureless;
slightly stick and slightly plastic when wet; abundant fine roots;
Fe-concentrations; few fine faint diflirse mottlings; gadual

, smooth boundary.

Dark geyish brown (10 YR 4/2), sandy loam, structureless;
abundant Fe-concentrations, meddium distinct mottling, some
pebbles visible; gadual smooth boundary.

Dark geyish brown (10 YR 4/2) moist, learn structureless;
abundant Fe-concentrations, medium distinct mottling.

The ground water-table was below 100 cm depth at the time of
sampling. However, during the rainy season it came up to 50
cm

Drainage is moderately slow with permiability class (SWON) of
0.05 according to Ritchie et a1. (1990).

Both Kaolinite and Montmorillonite (Smectite) types in varying
quantities as determined by Kesseba er al. (1972).

 

 

203

Appendix 2a. Daily weather data for the 1993 gowing season at Sokoine Unversity of
Agiculture, Morogoro.

 

Day of Temperature °C Solar Rainfall (mm)
Year Month Day Maxim. Minim. Radiation Experiment
(MJ m") Rainfed Irrigated

 

 

93091 April 01 31.8 20.4 20.3 0.6 0.0
93092 April 02 31.6 20.2 21.3 0.0 0.0
93093 April 03 30.0 20.9 12.5 5.0 0.0
93094 April 04 29.5 19.1 12.9 0.0 0.3
93095 April 05 30.5 20.7 17.1 0.3 28.9
93096 April 06 31.0 19.6 19.1 28.9 21.9
93097 April 07 31.4 21.0 18.1 21.9 8.3
93098 April 08 29.0 20.2 14.4 8.3 0.0
93099 April 09 30.0 21.0 12.9 0.0 0.1
93100 April 10 30.5 21.0 15.5 0.1 26.4
93101 April 11 30.4 19.6 19.0 26.4 0.0
93102 April 12 31.0 21.4 16.7 2.7 0.0
93103 April 13 31.0 20.9 18.7 8.5 8.5
93104 April 14 28.7 21.6 10.6 8.3 8.3
93105 April 15 28.8 21.0 10.6 36.1 36.1
93106 April 16 29.0 20.4 17.6 2.6 2.1
93107 April 17 28.5 20.6 14.3 4.6 4.6
93108 April 18 30.2 20.0 21.2 12.5 12.5
93109 April 19 30.0 20.5 15.3 33.2 33.2
93110 April 20 29.5 20.1 14.7 26.1 26.1
93111 April 21 27.0 20.5 8.8 19.2 19.2
93112 April 22 29.2 20.8 16.0 17.1 17.1
93113 April 23 27.7 21.0 10.8 13.6 13.6
93114 April 24 29.0 21.1 12.6 15.6 5.0
93115 April 25 28.5 21.7 10.6 18.8 18.8
93116 April 26 29.3 21.3 12.6 10.6 10.6
93117 April 27 31.0 21.5 19.9 0.0 0.0
93118 April 28 30.5 20.6 14.2 3.3 3.3
93119 April 29 30.7 20.3 17.2 13.4 0.0
93120 April 30. 31.0 21.4 16.7 0.0 0.0

Mean values 29.9 20.7 15.2 13.5 9.9

204

 

Appendix 2a. (cont’d).

93121 May 01 31.5 20.8 17.1 3.9 3.9
93122 May 02 32.0 21.1 15.7 7.0 7.0
93123 May 03 30.2 21.1 14.3 5.0 5.0
93124 May 04 28.0 20.6 11.6 8.9 8.9
93125 May 05 30.0 21.0 12.3 2.5 0.6
93126 May 06 31.8 21.6 16.1 0.0 0.0
93127 May 07 31.1 19.9 21.1 0.0 0.0

93128 May 08 29.2 19.9 17.7 0.0 0.0

93129 May 09 29.0 21.2 11.9 24.4 24.4
93130 May 10 31.1 20.0 16.2 0.0 0.0

93131 May 11 30.3 21.0 14.3 26.5 26.5
93132 May 12 28.6 21.2 14.3 0.0 0.0

93133 May 13 26.4 18.4 11.8 0.0 0.0

93134 May 14 27.0 19.5 9.8 0.0 0.0

93135 May 15 26.2 20.2 9.8 0.5 0.5

93136 May 16 27.0 19.0 8.8 0.1 0.1

93137 May 17 29.0 17.0 11.4 0.0 0.0
93138 May 18 28.0 16.0 18.8 0.0 0.0
93139 May 19 27.5 15.6 9.7 2.1 2.1

93140 May 20 28.5 19.4 15.3 2.7 2.7
93141 May 21 25.7 19.7 19.5 1.4 1.4
93142 May 22 29.5 20.1 9.0 7.5 7.5

93143 May 23 28.0 19.2 16.2 1.8 1.8

93144 May 24 28.4 20.5 15.5 3.0 0.0
93145 May 25 30.4 19.9 17.9 3.5 2.0
93146 May 26 28.5 19.5 17.4 3.0 0.0
93147 May 27 30.1 19.0 13.4 3.0 0.0
93148 May 28 28.9 19.9 17.6 2.0 0.0
93149 May 29 30.0 17.5 16.9 0.0 0.0
93150 May 30 29.0 17.1 20.4 2.9 2.4
93151 May 31 29.0 .19.5 16.2 4.4 4.4

Mean values 28.8 19.6 14.7 5.5 2.0

Appendix 2a. (cont’d).

205

 

93152 June
93153 June
93154 June
93155 June
93156 June
93157 June
93158 June
93159 June
93160 June
93161 June
93162 June
93163 June
93164 June
93165 June
93166 June
93167 June
93168 June
93169 June
93170 June
93171 June
93172 June
93173 June
93174 June
93175 June
93176 June
93177 June
93178 June
93179 June
93180 June
93181 June

01
02
O3
O4
05
O6
07
08
09
10
1 l
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30

Mean values

30.0
29.2
28.0
28.5
29.2
25.5
27.3
28.4
28.4
27.4
27.5
24.8
27.4
27.0
27.2
26.1
26.8
26.0
26.5
27.0
28.5
28.0
28.5
28.0
29.0
27.6
26.5
26.4
25.5
26.6

27.3

18.5
16.8
16.2
17.0
17.0
16.3
14.7
14.5
14.5
14.6
13.5
13.5
16.0
14.0
18.6
16.1
17.4
16.0
16.0
17.1
15.8
16.1
16.0
17.5
16.5
13.5
14.0
18.0
17.8
17.3

16.1

17.0
18.2
13.2
13.4
16.0
11.1
11.8
19.5
18.6
14.8
19.8
10.1
18.6

8.8
10.2
14.2
11.9
12.9
20.8
15.8
15.4
15.8
17.1
21.7
17.1
13.3
12.9
12.9
11.9
14.9

14.7

5.4
13.7
0.0
0.0
0.0
0.0
0.0
0.0
0.0
1.0
0.0
0.0
1.4
0.0
2.0
0.0
0.0
0.0
0.0
0.0
1.2
0.0
0.0
0.0
0.0
0.0
2.5
5.6
0.0
0.0

4.1

0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
1.0
0.0
0.0
1.4
0.0
2.0
0.0
0.0
0.0
0.0
0.0
1.2
0.0
0.0
0.0
0.0
0.0
2.5
5.6
0.0
0.0

0.4

Appendix 2a. (cont’d).

206

 

93182 July
93183 July
93184 July
93185 July
93186 July
93187 July
93188 July
93189 July
93190 July
93191 July
93192 July
93193 July
93194 July
94195 July
93196 July
93197 July
93198 July
93199 July
93200 July
93201 July
93202 July
93203 July
93204 July
93205 July
93206 July
93207 July
93208 July
93209 July
93210 July
93211 July
93212 July

01
02
03
04
05
06
07
08
09
10
l 1
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31

Mean values

26.0
27.0
28.0
27.4
25.5
27.5
28.3
26.0
26.5
28.0
28.4
24.5
27.5
25.0
28.7
27.0
24.5
26.6
25.9
25.5
23.0
25.0
26.0
26.0
25.4
26.0
28.5
28.0
28.5
28.8
32.2

27.3

11.6
15.1
13.5
13.8
15.8
16.7
16.3
17.2
17.5
16.2
15.2
16.3
17.6
14.9
16.5
17.4
17.5
18.1
17.3
16.9
15.5
16.7
14.9
10.0
11.2
13.1
12.0
14.6
16.1
16.7
16.8

20.5

16.2
14.1
18.8
19.5
16.6
16.6
11.5
16.2
17.9
13.5
13.3
16.2
19.0

8.0
19.5
10.4
19.9
11.7

9.5
12.7
11.9
13.4

5.9

9.8
17.4
17.1
15.9
13.9
19.0
13.9
15.3

15.1

0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
1.3
0.0
0.0
0.7
1.4
1.8
0.5
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0

0.1

0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
1.3
0.0
0.0
0.7
1.4
1.8
0.5
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0

0.1

Appendix 2a. (cont’d).

207

 

93213 August 01
93214 August 02
93215 August 03
93216 August 04
93217 August 05
93218 August 06
93219 August 07
93220 August 08
93221 August 09
93222 August 10
93223 August 11
93224 August 12
93225 August 13
93226 August 14
93227 August 15
93228 August 16
93229 August 17
93230 August 18
93231 August 19
93232 August 20
93233 August 21
93234 August 22
93235 August 23
93236 August 24
93237 August 25
93238 August 26
93239 August 27
93240 August 28
93241 August 29
93242 August 30

Mean values

27.1
27.9
26.7
25.7
21.0
25.5
25.4
26.6
26.4
28.8
27.4
29.0
28.2
28.5
25.4
27.5
27.2
27.4
28.3
26.6
27.2
27.6
28.6
28.5
27.5
27.5
29.4
29.0
25.4
29.0

27.3

17.3
15.2
15.0
18.5
16.9
16.4
15.3
14.6
15.2
16.9
17.1
13.5
14.7
15.8
16.3
12.0
16.7
14.6
15.0
14.5
14.7
15.4
15.8
17.3
17.1
18.7
16.4
17.3
17.0

. 18.5

20.5

14.6
16.7
14.4
12.6

5.2
11.3
12.7
19.3
11.6
15.9
22.6
22.3
18.1
15.8
16.2
16.1
14.3
15.3
17.8
14.5
16.0
15.3
14.2
13.2
10.8
20.8
18.6
10.0
12.0
15.5

15.1

0.0
0.0
0.0
0.0
0.0
1.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.2
0.0
0.0
1.2
0.0

0.1

0.0
0.0
0.0
0.0
0.8
1.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.2
0.0
0.0
1.2
0.0

0.1

208

Appendix 2b. Daily weather data for the 1994 growing season at Sokoine Unversity of
Agriculture, Morogoro.

 

 

Day of Temperature °C Solar Rainfall (mm)
Year Month Day Maxim. Minim. Radiation Experiment

(MJ m") Rainfed Irrigated
94091 April 01 30.5 21.4 19.5 0.0 0.0
94092 April 02 32.0 17.1 16.3 0.0 0.0
94093 April 03 32.0 18.8 22.0 0.0 0.0
94094 April 04 30.0 19.9 12.5 0.0 0.3
94095 April 05 28.0 20.5 11.8 1.3 1.3
94096 April 06 28.0 20.3 10.6 31.9 31.9
94097 April 07 26.4 20.3 9.7 11.1 11.1
94098 April 08 29.5 18.1 13.2 6.9 6.9
94199 April 09 29.0 20.0 14.7 24.4 2.7
94100 April 10 27.5 20.5 10.9 24.2 8.3
94101 April 11 27.1 20.0 10.6 4.1 4.1
94102 April 12 29.4 19.8 15.0 15.0 0.0
94103 April 13 29.8 18.0 17.9 1.0 1.0
94104 April 14 30.1 19.2 15.7 0.0 0.0
94105 April 15 30.5 20.8 20.6 7.1 7.1
94106 April 16 30.5 20.4 12.7 2.6 0.0
94107 April 17 30.5 19.5 20.1 0.0 0.0
94108 April 18 33.3 20.0 17.8 0.0 0.0
94119 April 19 31.5 18.3 17.5 5.9 5.9
94110 April 20 28.5 20.2 12.5 7.2 4.6
94111 April 21 26.5 20.2 7.8 16.0 0.3
94112 April 22 31.0 20.6 19.1 14.0 14.0
94113 April 23 ’30.0 19.4 15.6 1.2 1.2
94114 April 24 27.0 20.3 9.0 26.0 18.0
94115 April 25 29.0 19.6 18.1 26.5 5.4
94116 April 26 29.0 18.8 11.7 35.0 0.8
94117 April 27 27.6 19.5 14.4 5.9 0.0
94118 April 28 28.4 17.0 14.5 0.0 0.0
94119 April 29 26.5 17.5 11.0 13.4 11.1
94120 April 30 23.5 19.6 13.7 33.0 33.2
Mean values 29.0 19.5 14.2 10.6 5.6

209

 

Appendix 2b. (cont’d).

94121 May 01 25.2 19.2 9.2 0.5 0.5
94122 May 02 24.5 19.8 6.8 8.0 5.4
94123 May 03 27.5 17.5 13.7 17.2 17.2
94124 May 04 28.5 19.5 14.4 16.1 16.1
94125 May 05 30.3 20.5 13.3 0.6 0.6
94126May 06 30.7 19.8 18.1 0.6 . 0.6
94127 May 07 29.5 19.4 19.7 0.8 0.8
94128 May 08 27.0 19.5 10.6 0.0 0.0
94129 May 09 26.5 19.6 11.6 0.0 0.0
94130 May 10 27.3 18.8 11.1 0.0 0.0
94131May 11 25.5 20.0 7.5 15.0 7.4
94132 May 12 25.5 19.0 6.7 -8.4 8.4
94133 May 13 29.2 19.8 11.8 9.6 9.6
94134 May 14 27.5 19.4 11.0 0.0 5.7
94135 May 15 28.4 19.8 14.8 18.5 4.9
94136 May‘ 16 29.5 19.5 14.1 0.0 0.0
94137 May 17 28.9 18.9 14.5 0.8 0.8
94138 May 18 29.2 19.0 16.0 1.1 1.1
94139May 19- 25.6 18.8 9.8 3.7 3.7
94140 May 20 28.6 19.0 16.2 6.9 3.6
94141 May 21 29.5 19.1 16.2 1.9 0.0
94142 May 22 27.8 18.7 14.3 0.5 0.0
94143 May 23 29.5 17.6 19.5 0.0 0.2
94144 May 24 28.6 17.1 18.8 0.0 0.0
94145 May 25 29.0 17.5 16.5 0.5 0.0
94146 May 26 26.0 19.0 11.2 0.0 0.0
94147 May 27 26.5 16.0 11.4 0.0 0.0
94148 May 28 26.0 15.1 11.1 0.0 0.0
94149 May 29 26.0 16.2 11.0 0.0 0.0
94150 May 30 25.0 16.0 6.4 1.9 1.4
94151May 31 26.8 17.5 11.1 0.0 0.0

Mean values 27.6 18.6 12.9 5.9 2.9

Appendix 2b. (cont’d).

210

 

94152 June
94153 June
94154 June
94155 June
94156 June
94157 June
94158 June
94159 June
94460 June
94161 June
94162 June
94163 June
94164 June
94165 June
94166 June
94167 June
94168 June
94169 June
94170 June
94171 June
94172 June
94173 June
94174 June
94175 June
94176 June
94177 June
94178 June
94179 June
94180 June
94181 June

Mean values

01
02
03

05
06
07
08
09
10
1 1
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30

26.8
28.0
28.5
28.6
29.0
29.5
29.3
28.5
27.5
27.2
27.5
26.5
28.3
27.7
26.2
27.0
28.5
27.0
25.0
26.6
25.5
23.0
28.5
28.8
28.5
28.6
29.0
27.8
27.0
28.5

27.6

13.2
12.6
14.5
13.9
15.7
14.8
14.1
15.4
15.0
14.0
12.4
11.2
11.0
13.9
16.3
14.4
13.0
13.2
13.8

14.1 i

13.7
16.1
16.2
16.3
14.2
13.2
17.0
17.0

16.6

12.5

14.3

18.6
17.0
19.0
19.4
17.1
21.2
20.7
15.5
14.3
15.9
19.5
18.9
17.1
16.2
11.5
15.2
18.7
15.4
12.6
15.0
13.2
5.8
17.1
17.5
16.9
14.0
17.5
14.3
12.9
27.2

16.4

0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
1.2
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.2
6.6
0.0
0.0
0.0
0.2
0.0
0.0
0.0
0.0

0.3

0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
1.2
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.2
6.6
0.0
0.0
0.0
0.2
0.0
0.0
0.0'
0.0

0.3

21 1
Appendix 2b. (cont’d).

 

94182 July 01 29.0 13.9 17.8 4.7 4.7
94183 July 02 24.4 18.2 6.1 0.0 0.0
94184 July 03 27.4 17.1 12.9 0.2 0.2
94185 July 04 27.5 16.1 13.8 13.6 13.6
94186 July 05 28.0 18.0 16.5 1.1 1.1
94187 July 06 27.0 16.5 12.9 17.2 17.2
94188 July 07 25.5 17.4 10.6 0.0 0.0
94189 July 08 27.4 14.5 17.1 0.0 0.0
94190 July 09 28.0 14.5 17.4 0.0 0.0
94191 July 10 29.3 13.5 18.7 0.0 0.0
94192 July 11 29.2 15.1 16.9 0.0 0.0
94493 July 12 28.1 15.0 19.0 0.0 0.0
94194 July 13 25.0 16.1 10.8 0.0 0.0
94195 July 14 27.0 11.4 19.1 0.0 0.0
94196 July 15 26.5 10.2 19.0 0.0 0.0
94197 July 16 28.4 11.1 21.0 0.0 0.0
94198 July 17 27.0 10.8 16.2 0.0 0.0
94199 July 18 27.5 14.0 16.2 0.0 0.0
94200 July 19 27.6 13.7 18.3 0.0 0.0
94201 July 20 28.0 16.1 12.7 0.0 0.0
94202 July 21 27.0 17.5 15.9 0.0 0.0
94203 July 22 27.9 16.0 13.4 0.0 0.0
94204 July 23 29.0 17.4 10.3 0.0 0.0
94205 July 24 24.9 19.2 16.1 0.0 0.0
94206 July 25 28.7 14.8 16.1 0.0 0.0
94207 July 26 30.0 15.5 18.6 0.0 0.0
94208 July 27 30.0 16.2 19.3 0.0 0.0
94209 July 28 27.6 15.0 ’ 18.7 0.0 0.0
94210 July 29 27.0 16.0 18.2 0.0 0.0
94211 July 30 27.7 14.6 16.2 0.0 0.0
94212 July 31 26.2 16.8 7.0 0.0 0.0

Mean values 27.5 15.2 16.6 1.2 1.2

Appendix 2b. (cont’d).

212

 

94213 August 01
94214 August 02
94215 August 03
94216 August 04
94217 August 05
94218 August 06
94219 August 07
94220 August 08
94221 August 09
94222 August 10
94223 August 11
94224 August 12
94225 August 13
94226 August 14
94227 August 15
94228 August 16
94229 August 17
94230 August 18
94231 August 19
94232 August 20
94233 August 21
94234 August 22
94435 August 23
94236 August 24
94237 August 25
94238 August 26
94239 August 27
94240 August 28
94241 August 29
94242 August 30

Mean values

27.5
26.5
24.6
25.9
28.0
28.0
27.0
27.5
29.0
30.3
28.2
28.5
30.0
29.0
28.3
28.2
30.0
29.0
30.5
24.5
27.5
28.6
28.8
27.5
28.6
28.5
28.2
28.6
29.0
29.5

28.1

16.0
15.1
16.1
16.6
13.1
14.6
15.4
15.5
17.0
14.1
15.0
17.4
16.0
14.7
15.5
16.8
16.0
14.8
17.6
17.2
15.6

14.1.

15.2
16.8
15.1
17.4
16.6
15.8

- 16.1

15.2

15.8

18.4
11.0
8.6
12.5
19.3
22.9
14.2
18.1
17.9
18.1
17.7
14.0
19.3
19.9
13.4
16.8
17.1
17.8
16.9
6.3
16.2
19.9
19.5
15.7
16.2
12.4
16.0
16.2
18.1
18.5

16.0

0.0
8.4
0.5
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.5
0.5
0.0
0.0
4.7
0.0
2.7
1.0
0.4
0.0
0.0
0.0

0.6

0.0
8.4
0.5
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.5
0.5
0.0
0.0
4.7
0.0
2.7
1.0
0.4
0.0
0.0
0.0

0.6

213

Appendix 3. Developmental stages of determinate bean (Phaseolus vulgaris L.) used in the

 

 

study‘.
Days After Panting Growth Stage Growth Stage
(DAP) Code Descriptions
7-14 V2 Crop emergence and establishment. Leaf edges
no longer touch.

14 - 18 V3 Two to three nodes on the main stem including
the primary leaf node. Two to three true leaves
(trifoliolate leaves).

26 - 30 V. Vegetative growth terminates in a flower
cluster. 50% of plants with four nodes on main
stem.

28 - 32 Rl 50% of plants per plot with at least on open
flower on any node (Flowering stage).

30 - 34 R2 50% of plants per plot with at least one flower
at node immediately below the upper most node
with a completely unrolled leaf.

35 - 40 R. 50% of plants per plot with a pod 2.0 cm long
at one of the four upper most nodes with a
completely unrolled leaf. Seeds not discernible
by feel. (Early pod development).

42 - 46 R5 50% of plants per plot with seeds beginning

to develop at one of the four uppermostnodes

'with a completely unrolled leaf. Seeds

discernible by feel.

214
Appendix 3. (cont’d).

 

50 - 55 R5 50% of plants per plot containing filll size
seeds at one of the four uppermost nodes
with completely unrolled leaf. Pods about
10cm long at first flower position. (Seed
filling).

56 - 60 R7 Oldest pods with largest seeds. Pods
developing over the whole plant. Most pods
at full length. Point of maximum production

reached.

63 - 67 R. 50% of plants per plot with leaves yellowing
over half of the plant. Some pods drying.
Physiological maturity.

65 - 77 R9 More than 80% pods yellowing Lessthan15%

of leaves still green in color. Harvest maturity.

 

’ Growth Stages Identified Visually According to:
Fehr et al. (1971), and Nuland and Schwartz (1989).

215

Appendix 48. Irrigation schedules for the 1993 season.

 

 

DOY DAP Ammount of Water Crop Growth
Applied (mm) StageI
145 - - Land preparation (V o) .
146 - 3 1.5 -
147 , - 27 .8 Planting date (VP).
152 4 33.8 --
154 6 50.0 -
157 9 - 50% Crop emergence (V1).
161 13 30.0 95% Crop emergence (V2).
162 14 27.5 First trifoliolate leaf (V 3)
168 20 21.1 -
174 26 21.1 -
180 32 28.2 -
181 33 31.5 First flower in 50% of the plants (R1).
182 34 20.6 -
187 39 21.1 50% of the plants with pods of 2.0 cm long
' (R4)-
191 43 21.1 -
195 47 24.5 -
198 51 24. 5 -
204 56 20.8 -
208 60 21.7 >50% of the plants with mature pods
(Rs).
214 67 20.0 90% of plants with mature pods (R,,). End
of irrigation process.
218 77 - 90% of pods browning, 90% of leaves

yellowing maturity, and harvest time (R,,).

 

Total amount of water applied (mm) 476.8

DOY = Day of the year.

DAP = Days after planting.

‘According to Fehr et al. (1971), and Nuland and Schwartz (1989).

216

Appendix 4b. Irrigation schedules for the 1994 season.

 

 

DOY DAP Ammount of water Crop growth
Applied (mm) stage1

120. - - Land preparation (V0).

128 - - Planting date (VP).

136 8 40.5 50% Crop emergence (V1).

140 12 - 95% Crop emergence (V1).

142 14 33.7 First trifoliolate leaf (V 3).

150 22 3 1.8 -

157 29 , 36.3 First flower in 50% of the plants (R1).

159 3 1 28.4 -

168 42 40.3 50 % of plant are with pods of 2.0 cm long
(Rd-

173 45 32.4 -

177 49 26.9 -

183 . 55 10.3 -

187 59 19.3 >50% of plants with filll size beans (R6).

193 65 13.9 . -

200 72 . >90% plants with mature pods. End of
irrigation process CR.) '

207 76 - >90% of the pods browning, 90% of leaves

yellowing maturity, and harvest time (R).

 

Total amount of water applied (mm) 313.80

DOY = Day of the year.

DAP = Days after planting.

1According to Ferh et al. (1971), and Nuland and Schwartz (1989).

217

Appendix 5a. Plant population (plants m") at V3 and R from fresh applied P fertilizer
regimes.

 

 

 

 

 

 

 

 

Treatments ------1993 Season ---—-—1994 Season
V: R» V: R
Rainfed Experiment
0 kg P ha" 22.67 a' 17.33 a 18.67 a 13.67 a
22 kgP ha" TSP 22.67 a 18.00 a 22.69 a 15.33 a
22 kg? ha" MPR 22.33 a 17.33 a 21.43 a 16.33 a
44 kgP ha" TSP 21.33 a 17.67 a 19.54 a 14.67 a
88 kgP ha'l TSP 21.67 a 17.33 a 21.67 a 16.67 a
88 kgP ha" MPR 21.33 a 18.00 a 22.00 a 16.00 a
176 kgP ha'l MPR 21.00 a 19.00 a 21.33 a 14.33 a
MEAN 21.86 17.81 21.03 14.33
CV (%) 12.57 6.60 6.26 7.28
Irrigated Experiment
0 kg P ha‘l 20.00 a 19.00 a 21.00 a 18.00 ab
22 kg? ha‘1 TSP 20.67 a 18.33 a 21.33 a 19.67 ab
22 kgP ha" MPR 20.33 a 17.67 a 21.67 a 20.67 a
44 kgP ha'l TSP 19.67 a 18.00 a 18.00 a 16.67 ab
88 kgP ha“1 TSP 20.00 a 19.67 a 21.67 a 16.33 b
88 kgP ha'l MPR 20.00 a 19.67 a 20.33 a 18.00 ab
176 kg P ha" MPR 20.00 a 19.00 a 22.00 a 19.00 ab
MEAN 20.10 18.76 20.89 18.33
CV (%) 11.58 13.89 13.61 11.71

 

’Means within the column for a given season and variable not followed by the same letter

are significantly difi'erent at 0.05 level using Duncan’s Multiple Range Test.

Appendix 5b. Plant population (plants m") at V3 and R, from annual applied P fertilizer

218

 

 

 

 

 

 

 

regimes.
Treatments --l993 Season --l994 Season
V; R V: R9
Rainfed Experiment

0 kg P ha" 24.00 a ' 20.33 a 22.00 a 12.67 a
22 kg? ha" TSP 22.33 a 19.00 a 20.33 a 16.33 a
22 kgP ha’l MPR 23.33 a 22.00 a 22.33 a 14.00 a
44 kgP ha“ TSP 24.00 a 18.67 a 22.00 a 17.00 a
88 kgP ha" TSP 22.33 a 19.00 a 22.00 a 15.67 a
88 kgP ha" MPR 22.00 a 20.00 a 20.00 a 16.33 a
176 kg P ha‘l MPR 22.67 a 18.33 a 21.33 a 16.67 a
MEAN 22.81 19.57 21.42 15.52
CV (%) 6.70 11.30 10.43 18.86

Irrigated Experiment

0 kg P ha" 22.00 a 12.67 a 21.67 a 16.67 a
22 kgP ha'l TSP 20.33 a 16.33 a 21.00 a 16.00 a
22 kgP ha" MPR 22.33 a 14.00 a 24.33 a 16.67 a
44 kgP ha‘I TSP 22.00 a 17.00 a 21.67 a 17.00 a
88 kgP ha'l TSP 22.00 a 15.67 a 24.33 a 17.67 a
88 kgP ha-l MPR 20.00 a 16.33 a 21.67 a 17.00 a
176 kg Pha’l MPR 21.33 a 16.67 a 22.67 a 17.33 a
MEAN 21.42 15.52 22.48 16.91
CV(%) 8.64 9.22 8.20 12.45

 

’ Means within the column for a given season and variable not followed by the same letter

are significantly different at 0.05 level using Duncan’s Multiple Range Test.

"71111111111117111111“