.ier4 .3355?
a. .V.Ҥu.pn. F .. r

.. 11...:
V win

‘

an: ,
my,

‘

aw

. Emma;
.4... 4.3.2,
.. w .
2!. m . u
.fifi
.. . .
w!
(d
.156...
..

1-1
. .

. _. ‘ x
. vim; .. ”mm Luv
, .mV‘awiél _.......u.. .uwfimmsfw
. fig w, % é.
V. . .

. . I a
“no?! 2‘

.
vn.um....(.. . . u .
$5me 3. .1. 1
$1ng .53 .x van.

2&5”. .

 

 

 

..‘

 

3A.. . .w n 3..
WW“, 3mg.

m?-
lfiu O :I
1) .. .

s
it. .. ... to .
”summuiumxmvu..m&
#5 V 3 x,

93.9%.
Jaw... rd...
t . ’93 .

 

 

THESIS
I,

fan

This is to certify that the

thesis entitled

Carbon, Nitrogen and Phosphorus Sequestration With Soil
Aggregates of Maiie and Tree Based Agroforestry Systems

presented by
James M. Kinyangi

has been accepted towards fulfillment
of the requirements for

M-S. degree in crop and Soil Sciences

WQWM

wll' ’
jOl’ professor

 

Date June 7, 2000

 

0-7639 MS U is an Affirmative Action/Equal Opportunity Institution

 

 

LIBRARY

Michigan State

University

 

 

PLACE IN RETURN Box to remove this checkout from your record.
TO AVOID FINES return on or before date due.
MAY BE RECALLED with earlier due date if requested.

 

DATE DUE

DATE DUE

DATE DUE

 

lJiZLlo9502cins

 

 

 

 

 

 

 

 

 

 

 

 

 

 

11m W.“

CARBON, NITROGEN AND PHOSPHORUS SEQUESTRATION WITHIN SOIL
AGGREGATES OF MAIZE AND TREE-BASED AGROFORESTRY SYSTEMS

By

James M. Kinyangi

A THESIS

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

MASTER OF SCIENCE

Crop and Soil Sciences

2000

ABSTRACT

CARBON, NITROGEN AND PHOSPHORUS SEQUESTRATION WITHIN SOIL
AGGREGATES OF MAIZE AND TREE-BASED AGROFORESTRY SYSTEMS

By

James M Kinyangi

Direct contributions by tree crops to the nutrient capital and nutrient cycling on
small farm holdings in Sub Saharan Africa have been demonstrated for several
soil types. In many ecosystems, carbon appears to be the most limiting factor
controlling microbial biomass-mediated nutrient turnover. Soil aggregate
sequestration of carbon (C), nitrogen (N) and phosphorus (P) was investigated
on a P deficient Oxisol in Western Kenya, during four crop cycles of maize and
tree rotations (1997—1999) in an agroforestry management system. Concentric
layers of soil aggregrates (4-8 mm across) were removed using stainless steel
meso-SAE chambers, developed for separating concentric layers from interior
regions by gentle erosion of aggregate surfaces. Anion exchange resins (Cl’),
converted to bicarbonate form, were used to evaluate plant available phosphorus
supplied by layers removed from soil aggregates. Carbon contributions by tree
(03, 8‘30 '26%o) and maize (C4, 5‘30 40%) crops to specific concentric layers
within soil aggregates sampled at the 0-15 cm depth were estimated by changes
in their delta ‘30 signatures. Total C and N were analyzed by the dry combustion
technique. Exterior to interior ratios (Ele), used to calculate relative locations of

plant available P within soil aggregates, showed maize crop removal of plant

available P from the exterior surface layers of soil aggregates when applied
fertilizer P was insufficient to meet plant demand. A positive linear relationship
(r2 = 0.66) was observed between a 9-month change in EIR with July 1999 maize
grain yield, suggesting that greater P depletion resulted in increases in maize
grain yield of the tree-crop rotation system. Poor plant growth in the continuous
maize cropping system resulted in 48% more accumulations of plant available P
to surface layers of soil aggregates. Shifting plant inputs of C between 03 and C4
sources revealed a pronounced seasonal influx of C4 carbon accumulations onto
surface layers of soil aggregates. The contemporary maize crop contributed as
much as 50% of the current carbon onto surface layers in July 1998 (field study)
and August 1999 (pot experiment). However, this influx was not apparent when
whole aggregate measurements were taken. During the same time period, C
contents in surface layers of soil aggregates, 4-8 mm across, declined nearly
23% when fertilizer P was applied to planted tree crops. These results suggest
that uniform separation and analyses of soil aggregate layers should be
integrated into current research methods to further enhance our understanding of
the rapid and short term changes in C and P associated with management
modifications in low input crop production systems of sustainable tropical

agriculture.

Copyright by
James M. Kinyangi
2000

This work is dedicated to my wife Millicent, our children, Sandra, Valessa,
Barbara, my parents, brothers and sisters.

ACKNOWLEDGMENTS

I would like to thank my advising committee, Dr. A. J. M. Smucker, Dr. Fl.
R. Harwood and Dr. R. J. Buresh for guidance and support through this program.
Other non-committee support from Dr. Pedro Sanchez and Dr. Cheryl Palm is
acknowledged. Further support from the CS. Mott Chair for Sustainable
Agriculture, Michigan State University and USAID through the lntemational
Center for Research in Agroforestry, Nairobi, Kenya is greatly appreciated. I am
grateful for support from Dr. Bashir Jama, Obadah Kyunguti, John Mailu and all
other colleagues at KEFRI/KARI/ICRAF research station, Maseno, who helped

with the field sampling and data collection in Western Kenya in 1997 to 1999.

vi

TABLE OF CONTENTS

INTRODUCTION .................................................................................................. 1

CHAPTER 1: MANAGEMENT MODIFICATIONS OF CARBON AND NITROGEN
DEPOSITS IN SOIL AGGREGATES OF TREE AND MAIZE BASED SYSTEMS
IN WESTERN KENYA

INTRODUCTION ................................................................................................ 8
MATERIALS AND METHODS 11
Site description ........................................................................................ 11
Experimental design ................................................................................ 11
Soil sampling ........................................................................................... 12
Aggregate dry sieving .............................................................................. 13
Carbon and nitrogen analyses ................................................................. 14
Statistical analyses .................................................................................. 15
RESULTS AND DISCUSSION ........................................................................... 15
Aggregate fraction sequestration of carbon ............................................. 17
Aggregate fraction sequestration of nitrogen ........................................... 18
Influence of tree crop rotation on carbon and nitrogen distributions ........ 18
Carbon to nitrogen ratios ......................................................................... 19
Yields ................................................. 19
CONCLUSIONS ................................................................................................. 19
REFERENCES ................................................................................................... 20

CHAPTER 2: CONCENTRIC SEQUESTRATION OF PHOSPHORUS IN SOIL
AGGREGATES OF MAIZE AND TREE BASED SYSTEMS IN WESTERN
KENYA

INTRODUCTION ................................................................................................ 37
MATERIALS AND METHODS ............................................................................ 40
Site description ........................................................................................ 40
Soil sampling ........................................................................................... 41
Aggregate erosion ................................................................................... 42
Modified resin extractable phosphorus .................................................... 43
Exterior to Interior ratio (EIR) ................................................................... 46
Statistical analyses .................................................................................. 47
RESULTS AND DISCUSSION ........................................................................... 48
Plant available phosphorus in whole aggregates ..................................... 48
Comparisons of EIR to gradients in plant available P .............................. 49
CONCLUSIONS ................................................................................................. 51
REFERENCES ................................................................................................... 53

vii

CHAPTER 3: ESTIMATING RECENT CROP CARBON INPUT INTO
CONCENTRIC LAYERS OF SOIL AGGREGATES FROM MAIZE AND TREE
BASED SYSTEMS

INTRODUCTION ................................................................................................ 68
MATERIALS AND METHODS ............................................................................ 71
Experimental design ................................................................................ 72
Soil sampling ........................................................................................... 73
Pot experiment ........................................................................................ 74
Aggregate erosion ................................................................................... 75
Carbon analyses ...................................................................................... 76
Statistical analyses .................................................................................. 78
RESULTS AND DISCUSSION ........................................................................... 78
5‘30 of plant material ............................................................................... 78
Total carbon ............................................................................................. 78
Sources of crop derived carbon ............................................................... 79
CONCLUSIONS ................................................................................................. 82
REFERENCES ................................................................................................... 83
CHAPTER 4: SUMMARY AND CONCLUSIONS ............................................... 97

viii

LIST OF TABLES

Table 1.1 Analysis of variance for the effect of cropping system and
aggregate size on carbon and nitrogen dynamics of a Kenyan Oxisol, 0-15
cm depth, 1998. (n = 4)

Table 1.2 Soil carbon distribution within aggregate size fractions of a
Kenyan Oxisol after 10 months of tree-crop rotation, 0-150m depth, 1998.

Table 1.3 Soil nitrogen deposits within soil aggregates of a Kenyan Oxisol
following 10 months of tree fallow-crop rotation, 0-150m depth, 1998.

Table 1.4 Cropping system influence on sequestration of soil carbon of a
Kenyan Oxisol, 0-15cm depth, 1998.

Table 1.5 Cropping system influence on sequestration of soil nitrogen of a
Kenyan Oxisol, 0-15cm depth, 1998.

Table 1.6 Changes in the C:N ratio of aggregate size fractions, 0-15 cm
depth, 1998.

Table 1.7 Maize cropping season and the change in the ratio of carbon
to nitrogen, 0-15cm depth, 1998.

Table 1.8 Single degree of freedom contrasts for the differences in means
of yields, 1998.

Table 2.1 Increasing soil: resin ratio and changes in extraction of plant
available phosphorus of 3 soils.

Table 3.1 Maize (C4) and Crotalaria (C3) contributions to surface layers
and interior regions of soil aggregates (4-8mm across), during the field
study, (0-150m depth).

Table 3.2 Maize (C4) and Crotalaria (Cs) contributions to surface layers
and interior regions of soil aggregates (4-8mm across), during a 5-week
pot study, 1999.

Table 3.3 Effect of fertilizer application on above and below ground
biomass of maize grown in 7 liter pots during a 5 week study, 1999.

Page

24

25

26

27

28

29

30

31

58

88

89

90

LIST OF FIGURES

Page
Fig. 1.1 Mean weight diameter changes of dry sieved soil aggregates as
affected by maize-tree crop rotation on a Kenyan Oxisol (0-15cm), 1998. 32

Fig 1.2 Carbon mass in soil aggregate size fractions of a Kenyan Oxisol,
0-150m depth, 1998. 33

Fig. 1.3 Nitrogen mass in soil aggregate size fractions of a Kenyan Oxisol,
0-15cm depth, 1998. 34

Fig. 1.4 Pre-season changes in the inorganic-N pool, with (a) or without (b)
phosphorus fertilization and their relationship to August 1998 maize grain

yield. Data are for continuous maize and plots converted to rotation with
Crotalaria grahamiana, a leguminous tree. Changes in N are measured at

the onset of the maize cropping season in February-March 1998. 35

Fig. 1.5 Maize grain and tree biomass yield of three cropping systems at
Osita Farm, Western Kenya, 1998. 36

Fig. 2.1 Diagrammatic illustration of meso Soil Aggregate Erosion (SAE)
chamber. , 59

Fig. 2.2 Time versus progressive removal of concentric layers of aggre-

gates, 4-8mm across (mean = 6mm), from continuous maize and Crota-

laria rotation plots of a Kenyan Oxisol, 1999. Each region/layer is one third

of the original mass of the aggregate (n = 12). 60

Fig. 2.3 Effect of sand removal on resin extractable phosphorus of a Ken-

yan Oxisol. The two extraction regimes are representative of the convent-

ional resin P method (2.59 soil, 1g resin) and a more recent method (0.125

9 soil, 0.6259 resin) developed to extract P from layers of soil aggregates.
(Bars are standard error, n = 24). 61

Fig. 2.4 Influence of resin load and soil weight on resin extractable phos-
phorus. The resin is chloride base converted to bicarbonate form. The soil
is shaken for 16 hrs at 175 reciprocations min". 62

Fig. 2.5 Seasonal flux of resin extractable phosphorus in whole aggre-

gates of 3 management systems without P fertilization on a Kenyan

Oxisol. Crop sequence represents (a) Tree crop rotation and (b) contin-

uous maize plots. Error bars are standard errors for the difference in

means. 63

Fig. 2.6 influence of post fallow C concentrations of aggregates on sub-
sequent pre-season changes in resin extractable phosphorus. Changes
of resin 0 in continuous maize plots (control) is subtracted from fallow
rotation plots (n = 4).

Fig. 2.7 Cropping system effect on seasonal location of plant available
phosphorus in concentric layers of soil aggregates (4-8mm across), of a
Kenyan Oxisol, 0-15cm depth. Crop sequence represents (a) Tree crop
rotation and (b) continuous maize plots. Error bars are standard errors
for the difference in means (n = 4).

Fig. 2.8 Relationship between change in the ratio (Jul-99-Sep 98) of
exterior to interior resin extractable phosphorus and July 1999 grain yield
(n = 4).

Fig. 2.9 Fertilizer management and gradients in P location in aggregates,
4-8mm across (0-15cm depth) of a Kenyan Oxisol, 1999 (n = 4).

Fig. 3.1 Diagrammatic illustration of meso Soil Aggregate Erosion (SAE)
chamber.

Fig. 3.2 Maize grain yield (Aug-98) as influenced by the EIR of total
carbon within layers of soil aggregates (4-8mm across) for a Kenyan
Oxisol. Aggregates were sampled in February 1998, from 0-15cm depth
of fertilized (annual 50kg P ha") continuous maize and Crotalaria tree
plots of an agroforestry management system. Regression excludes
square data point (n = 4).

Fig. 3.3 Total carbon contributions onto surface layers and interior
regions of soil aggregates (4-8mm across) for a 5-week study in 7 liter
containers. Bars are standard errors (n = 5, maize; n = 3, no maize).

Fig. 3.4 Effect of crop rotation on seasonal variation in maize (C4-source,
'10.6%o, n = 5 and Crotalaria Ca-source, 26.1%, n = 4) carbon contribut-
ions onto surface layers and interior regions of soil aggregates (4-8mm
across, 0-150m depth), of a Kenyan Oxisol. Reference soil was sampled
from a bush fallow (9-12 years since conversion from agriculture) at Osita
Farm in July 1999. Bars are std. Errors for the difference in means (n = 4).

Fig. 3.5 Crop rotation and fertilization modifications of carbon deposition
onto surface layers and interior regions of soil aggregates (4-8mm across),
for a Kenyan Oxisol (n = 4).

Fig. 3.6 Maize carbon contributions onto surface layers and interior regions

of soil aggregates (4-8mm across) for a 5-week study in 7 liter containers.
Bars are standard errors (n = 5, maize; n = 3, no maize).

xi

65

66

67

91

92

93

94

95

96

Appendix A1. Reconstruction of a microtomographic image of a soil

aggregate (1.1 mm across) from Kenya, rendered by algorithms at the

APS. Light areas represent surface breaking and internal pores >63

microns, while black areas are the solid portions of the aggregate and
surrounding air. The light tubular area within the middle appears to be a

root induced macropore (RIM). The aggregate was sampled from an

Oxisol soil in July 1998. 101

xii

INTRODUCTION

Improved tree fallow systems where maize is interplanted with fast
growing trees such as Sesbania, Tephrosia, Crotalaria or fithonia promote
biomass production and preserve soil resources. Tree fallows are usually untilled
for periods from 8-10 months before the woody material is taken off the field
while the leaf litterfall is incorporated into the plow layer during land preparation
before the next food crop is planted. Tree-crop production systems include;

a) sequential systems in which trees and crops are either grown in
rotation, or maximum growth rates of the tree and crop components
occur at different points in time, where competition for growth
resources is absent or minimal (Rao et al., 1997), and

b) simultaneous systems in which trees and crops are grown together.
These are fast becoming an important part of cropping systems in
western Kenya

Direct contributions by tree crops to soil nutrient deposition and availability
warrant more attention as nutrients from tree crops are added to nutrients
already in the soil. For example, the return of maize stover to cropland and the
transfer of nutrients from trees to the soil through pruning, leaf litter or
decomposition of roots in agroforestry systems, contribute to the nutrient capitals
of nitrogen (N), phosphorus (P) and other essential elements in the soil. These
nutrients become available to successive crops during a time scale ranging from

1-10yrs.

Soil fertility benefits by tree fallows include;

a) maintenance or increased levels of soil organic matter (SOM),

b) uptake of nutrients below the food crop rooting zone,

c) increased water storage and infiltration resulting in lower nutrient

losses,

d) reduced losses of soils and associated nutrients by erosion,

e) improved soil aeration

f) reduced soil acidity and

9) improved soil biological activity.

Most of these soil processes have been researched under simultaneous
agroforestry systems such as hedgerow intercropping (Juo et al., 1995; Kang et
al., 1997) than for sequential systems such as planted tree fallows (Sanchez,

1 995).

This report summarizes the carbon, nitrogen and phosphorus dynamics
associated with soil aggregates sampled from three maize-tree-cropping
systems. In the first study, carbon and nitrogen sequestration by Crotalaria
grahamiana, (nitrogen fixing) and Tithonia diversifolia (non-nitrogen fixing) trees
in rotation with maize in a low input cropping system was investigated.
Objectives of this first study included:

1) Determination of carbon and nitrogen deposits within aggregate size fractions

of 8-10 months planted tree fallow and maize crop rotations.

Hypothesis : Carbon accumulations in soil are a function of tree type and
rotation.
2) An evaluation of C and N dynamics within aggregates, when fertilizer P is

applied to maize and planted tree crops rather than maize crops only.

Hypothesis: Biomass production and carbon deposition within aggregates is

enhanced when fertilizer P is applied to planted maize and tree crops.

Short duration fallows (<14 months), in which crop production is omitted
for one season may have some potential in areas with high population where
cropping is often continuous, such as western Kenya (Franzel., 1998; Buresh
and Tian., 1998). The potential is greatest when returns to cropping are low, the
opportunity cost of labor is high and the tree provides valuable products such as
fuel wood (ICRAF, 1997; Swinkels et al., 1997; Buresh and Tian., 1998). It is
financially more attractive to apply phosphorus inputs during the cropping phase
of tree fallow systems as was shown with sesbania in western Kenya (ICRAF,
1997; Jama et al., 1998). Little is known, however, about the effects of short-term
fallows on soil P and supply of plant available P to subsequent crops.

Accumulation of carbon, nitrogen, phosphorus and sulfur in SOM is
dependent on P content of the soil parent material (Walker., 1965) but today few
studies indicate that higher P results in more carbon and nitrogen accumulation.
In Australia and New Zealand, fertilizer P increased the fertility of managed
tillage systems (Paul and Clark). Soil organisms contribute to the renewal of the

P pool both by their solubilizing effects on labile phosphorus (Nziguheba et al,

1998) and by their mineralization of organic phosphorus. Soil organisms and
plant roots participate in solubilization of soil P through the production of C02 and
organic acids (Smithson et al, 1999; Jones et al, 1999). The low level of P in soil
solution and necessity for its frequent renewal suggests that the transfer of SOM
P to inorganic P is primarily, but not exclusively, microbially mediated.
Mycorrhiza fungi play an important role in P transfers within soil root
interactions. Three mechanisms have been proposed for this; I) mycorrhizal
hyphae contribute to solubilization of mineral phosphate by their production of
C02 and organic acids, ii) mycorrhizal hyhpae extend soil exploration over and
beyond the soil exploration by the plant roots. Hyphae penetrate SOM particles
and macro aggregates more thoroughly than do roots, iii) hyphae may take—up P
at lower levels of concentration in solution than roots.
In a second study, supply of plant available phosphorus was determined in
concentric layers of soil aggregates, 4-8mm across.
The Objective was:
To evaluate the physical locations of recently applied plant available phosphorus
within concentric layers of soil macro-aggregates and its relationship to maize

yields.

Hypothesis: Soil aggregate P sequestration affects short-term plant available
phosphorus.
The application of a stable carbon isotope as an intrinsic marker for tracer

studies has recently provided a new impetus for root and soil C studies. During

photosynthesis, C4 plants discriminate less against 13002 than 03 plants. The
result of this fractionation of stable mrbon isotopes during photosynthesis is a
characteristic carbon isotope ratio of delta ‘30 signature in the plant tissue.
These isotopic signatures serve as effective markers for the occurrence of Ca
and C4 photosynthesis (Dzurec et al, 1985). Crotalaria, a N-fixing fallow crop, has
a 03 metabolism, which generates delta ‘30 signatures of approximately —26 %o.
Maize has a delta 13C signature of -10 it». The trend would be to observe
increased deposits of recent carbon at the surface layers of soil aggregates
decreasing towards the interior. While using stable carbon isotopes to study two
tillage systems, Smucker et al (1997) observed that twice as much carbon,
derived from the contemporary crop, was deposited in the outer layers than the
interior regions of soil aggregates sampled from 0—0.05m soil depth.

A third study was designed to investigate crop C deposits within
aggregates derived from rotations of 8-10 months which had been planted to tree
fallow (Ca metabolism) and continuous maize cropping (C4 metabolism) systems.
The Objective was:

To determine Crotalaria grahamiana (C3) and maize (Zea mays L.) (C4 ) carbon

contributions to external surfaces and interior regions of soil aggregates.

Hypothesis: Stable Carbon isotope signatures following a switch from continuous
maize cropping to planted tree fallows in an agroforestry management system
would demonstrate that recent organic inputs have a higher proportion of. carbon

derived from the contemporary crap.

REFERENCES

Buresh R J and G Tian, 1997 Soil Improvement by Trees in Sub.Saharan Africa.
. Agroforestry Systems. 38:51 -76.

Dzurec R S, T W Boutton, M M Caldwell, and B N Smith. 1985. Carbon Isotope
ratios of soil organic matter and their use in assessing community
composition changes in Curlew Valley, Utah. Oecologia 66: 17-24.

Franzel S, 1999. Socio-economic factors affecting adoption potential of improved
tree fallows in Africa. Agroforestry Systems. 47:305-321.

ICRAF, 1998, Annual report 1997, lntemational Ctr for Res. In Agrofor., Nairobi,
Kenya.

Jama B A., R A Swinkels and R J Buresh. 1997. Agronomic and economic
evaluation of organic and Inorganic sources of phosphorus for maize.
Agron. J 89:597-604.

Jones D L and J Farrar, 1999. Phosphorus mobilization by root exudates in the

rhizosphere: fact or fiction? Agroforestry Forum. Vol 9 No. 4.
Juo A S R, A Dabiri and K Franzluebbers, 1995. Acidification of a kaolinitic Alfisol

under continuous cropping with nitrogen fertilization in west Africa. Plant

and Soil 17:245-253.

Kang B T, F K Salako, J O Akobundu, J L Pleysier and J N Chianu, 1997. Soil
Use Management 1 3:130-135.

Linquist, B.A., P.W. Singleton, KG. Cassman, and K Keane. 1996. Residual
fertilizer phosphorous and long term fertilization strategies for an Ultisol.
Plant Soil 184: 47-55.

Nziguheba G., C A Palm, R J Buresh and P C Smithson, 1998. Soil Phosphorus
fractions and sorption as affected by organic and inorganic sources. Plant
and Soil 198:159—168.

Paul E A and F E Clark, 1989. Soil Microbiology and Biochemistry. Academic
Press, San Diego, CA.

Rao M R, P K Nair C K Ong, 1997. Biophysical interactions in tropical
agroforestry systems. Agroforestry Systems 38: 3-50.

Sanchez P A, 1995. The science of Agroforestry. Agroforestry Systems 30:5-55.

Smithson P C, 1999. Interactions of organic materials with phosphate rocks and triple
Superphosphate. Agroforestry forum. Vol. 9 No. 4.

Smucker A J M., D Santos and Y Kavdir, 1997. Concentric layering of carbon,
nitrogen and clay minerals within soil aggregates from tilled and non-tilled
agroecosystems and natural grasslands. Eastern Canadian Soil Structure
Workshop , Ottawa, Canada. pp 129-141.

Walker T W, 1965. The significance of phosphorus in pedogenesis. In
Experimental Pedology (E A Hallsworth and D V Crawford, eds.) pp295-
31 5.

Chapter 1

MANAGEMENT MODIFICATIONS OF CARBON AND NITROGEN DEPOSITS,
IN SOIL AGGREGATES OF TREE AND MAIZE BASED SYSTEMS IN
WESTERN KENYA

INTRODUCTION

Carbon (C) sequestration in soil appears to occur primarily in associations
with the structural configurations of soil aggregates. Nearly 90% of soil organic
matter has been reported to be located within soil aggregates (Angers and
Chenu, 1997, Jastrow et al, 1996). Exudates and rhizosphere deposits of
growing roots and hyphae are rapidly turned over by microbial populations,
resulting in the deposition of polysaccharides in rhizosphere soils. As roots and
hyphae die, organic substrates are deposited throughout the micro aggregate
structures of the soil where the microbial activity reduces these substrates to
particulate organic matter, resulting in further deposition of polysaccharides
(Jastrow and Miller, 1997). The concept of an aggregate hierarchy in soils, for
which organic matter is the major binding agent, is widely believed to occur
through the organization of primary particles, cemented together to form micro
aggregates up to 0.25mm in diameter. These micro aggregates are then bonded
together into macro aggregates greater than 0.25mm (T isdall and Oades, 1982;
Oades, 1984).

In a recent review, Angers and Chenu, (1997) propose that soil organic
matter (SOM) content in aggregate fractions results from, 1) the incorporation of

SOM into different fractions, 2) specific turnover rates of carbon (C) which

depend upon chemical and physical configurations within aggregates, and 3) the
turnover rate of the fractions. These mechanisms by which SOM becomes
incorporated into soil aggregates are poorly understood. Perhaps more research
is necessary to quantify the physical dissolution times of natural soil aggregates,
their stability and associated processes by which SOM is incorporated into
aggregates.

Loss of soil organic carbon resulting from long term continuous cultivation
comes chiefly from inter-aggregate organic material that binds micro aggregates
into macro aggregates (Elliot, 1986). This fraction of organic matter is
characterized by Paul, (1984), and Voroney et al (1981) as less highly
processed, aliphatic, stabilized slow or decomposable fractions.

In agricultural ecosystems, the portion of total organic carbon in the micro
aggregate fraction is highest under non-cultivated systems and will decline with
time under arable crops (Christensen, 1992). Agroforestry practices have an
important potential to increase the soil content of sequestered carbon in tropical
regions (Greenland, 1995). However, the magnitude of increase that can be
attained depends on several factors, including nitrogen (N) and phosphorus (P)
availability, and the sulfur (S) status of the soil. Management systems integrating
shifting tree fallow and food crop rotations with high C inputs and reduced tillage
should therefore favor carbon storage directly by reducing aggregate breakdown
and indirectly by enhancing the SOM mediated stable aggregation. Carbon and
N deposition within aggregates would be a function of total biomass added by the

cropping system. Fallow tree crops produce large quantities of leaf litter and root

litter that is incorporated back into the soil as inputs in the cropping system. Non-
N fixing trees of genus Senna and Tithonia accumulate as much N in their leaves
as N-fixing legumes (Szott et al, 1991; Gachengo, 1996; ICRAF, 1998) and may
compete as species alternatives on P limiting sites where N fertilizers can be
added.

Little is known however, about the effects of short-term fallows on soil
aggregation and C and N sequestration. We hypothesized that sequestration of
C and N occurs in soil aggregates when tree-crop rotation with leguminous N
fixing trees rather than maize monoculture is practiced. Additionally, non-N fixing
tree fallow crops, growing on phosphorus depleted soil may posses equal C and
N sequestration abilities within the same tree to maize crop rotation season, thus
expanding farmer options for using tree fallow species to regenerate soil fertility.
This study was conducted to a) determine carbon and nitrogen deposits within
aggregate size fractions of 8-10 months planted tree fallow and maize crop
rotations, b) assess nitrogen fixing Crotalaria grahmiana and non-nitrogen fixing
Tithonia diversifolia (Hemsley A) Gray, as suitable tree species for rotation.
These species have been proposed for adoption by farmers in a continuous
maize cropping system with limited C and N inputs (ICRAF, 1998), and c) to
compare C and N dynamics within aggregates, when fertilizer P is applied to

maize and planted tree crops rather than sole maize crops.

10

MATERIALS AND METHODS

Site description

The study site is located at Osita farm in Vihiga District, Western Kenya
(0° 06’ N, 34° 34’ E) at an elevation of 1420m. Mean annual rainfall is 1800mm
(split between 2 cropping seasons, February-August and September to January).
Often, the first cropping season will receive 1000-1200mm of the annual rainfall.
The soil type is a very fine lsohyperthermic Kandiudalfic Eutrudox (USDA, 1996).
It has few chemical or physical barriers to rooting in the top 4m. Air dried soil for
the top 15cm had a pH of 5.1 (1 :2.5 soilzwater suspension), organic carbon 179
C kg", ex. acidity 0.4 cmolc kg", calcium 3.4 cmolc kg", (1.0 M KCL extractable),
ex magnesium cmolc kg" 1.4, ex potassium cmolc kg" 0.12, phosphorus 2 mg
kg", (bicarbonate-EDTA extractable). Clay, silt and sand textural contents were

46%, 28% and 26%, respectively.

Experimental design

The experiment consisted of a factorial combination of three cropping
systems (maize-maize, Crotalaria-maize, and Tithonia-maize) and two rates of P
application (no P and 50kg P ha"). Phosphorus was applied in March 1997as
broadcast and incorporated into the top 0-15cm depth at 50kg P ha" (as Triple
Super Phosphate, 46% P205) The tree crop was relay planted into maize during
March 1997 and April 1998. Maize (HE-512, Kenya Seed Co, Kitale, Kenya), was
planted at 2 seeds per hill, at a spacing of 0.75m between rows and 0.25m

between plants, and later thinned to 1 plant per hill to achieve a plant population

11

of 53 000 plants ha". One row of tree crop was planted for every row of maize
resulting in the same tree plant population. Maize crop was harvested for grain
yield in August 1998 and July 1999 in the tree-crop rotation plots. Two additional
maize crops were harvested in January 1998 and 1999 (continuous maize plots).
All maize crops were assessed for grain yields. The tree crop was harvested in
February 1998 and 1999. All residue and litter fall was weighed and returned to
the plots, while the woody components were exported from the plots for fuel
wood. All above ground maize crop biomass was removed since it was common
practice for farmers to harvest the stovers and utilize them as livestock feed. We
only considered the residue returned as it had a direct effect on the fertility status
of the soil. Leaf litter and pods were harvested and oven dried at 60 ° C for dry

matter and plant N, P and K determination (Allen and Parkinson, 1975).

Soil sampling

Soils were sampled for C and N determination from an ongoing
experiment initiated in March 1997 to investigate phosphorus (P) replenishment
with fallow tree crops in western Kenya. Soil samples were collected from
different experimental plots on 13 Feb. 1998, which was designated to be the
end of the tree fallow and the beginning of the food-cropping season. A second
soil sample was collected on 7 Jul 1998, at the physiological maturity stage for
the maize crop. Soils were collected (0-15cm depth) with cores (10cm diameter,
15cm height) gently driven into the ground to minimize aggregate disruption.

Each soil sample was a composite of 9 locations per plot. The soil was loaded

12

onto trays secured with rubber strapping and transported to the station, where

they were air-dried and stored in a cool, dry and dark room.

Aggregate dry sieving

During both sampling dates, equal amounts (1 kg) of initial air-dried soil
(moisture content 209 H20 kg" soil) were prepared for dry sieving by removing
rock, root and other debris. Samples were manually shaken through a nest of
sieves (20-30cm diameter), separating them into eight size fractions (>8mm, 4-
8mm, 2-4mm, 1-2mm, 0.5-1 mm, 0.25-0.5mm, 0.053-0.25mm, 0.038-0.053mm
and <0.038mm). The fraction retained on each sieve was weighed to estimate
the mass of each size fraction of soil aggregates. These values were summed to
determine each aggregate size fraction as a percentage of the total soil mass.
The <0.038mm fraction was considered small (<0.5% of the fractions by mass)
and in some cases was missing from the sieving process. Any sample collected
was retained, but no further analysis was conducted for this fraction. Sub-
samples were obtained from these fractions for fine grinding and weighing at 5
decimal places for total carbon and nitrogen determinations. Angers and Giroux,
(1996) evaluated C and N contents among wet and dry sieved aggregate
fractions and found trends in C and N distribution among wet but not dry sieved
aggregates. Therefore, we explored the differences in C and N across time for
dry sieved aggregate fractions between February 1998 (undisturbed soil sample
at the end of tree crop season) and July 1998 (end of 1st residual maize crop).

Seasonal differences were a function of physical and chemical disruptions of

13

aggregates due to tillage, rainfall, maize crop activity and applied fertilizer
materials. This provided an in situ opportunity to observe changes in dry
aggregate mean weight diameter (MWD) as proposed by Kemper and Rosenau,

(1986). MWD was calculated using the equation;

MWD = Z "i =1 XIWI
Where:
Xi = mean diameter of size fraction/size class midpoint;

WI = proportion of total sample retained on the sieve.

Carbon and nitrogen analyses

Carbon and nitrogen were analyzed by the dry combustion technique
using a CHN analyzer (Carlo Erba, Milan, Italy). After fine grinding, a soil sample
(20-40mg) is weighed and loaded into pressed tin cups (8mm x 5mm). It is then
introduced to the CHN analyzer via an auto sampler. The sample was combusted
at 1000° C in a stream of 02 gas, oxidizing all forms of C to CO2 Nitrogen oxides
are reduced to N2 gas. These analyses provided carbon and nitrogen
concentrations in the soil sample. We converted these percentages into actual
amounts of C and N by mass for each aggregate size fraction and sub-sample in
order to estimate the quantities of C and N sequestered by whole aggregates for
each size fraction. Amounts of C or N in each size fraction were calculated by the

equations (1) and (2).

14

C mass (mg) = [(%C * aggregate size fraction mass (mg)]. (1)

N mass (mg) = [%N * aggregate size fraction mass (mg)]. (2)

Statistical analysis

Statistical analyses were performed with Statistical Analysis System (SAS
Institute, 1999). Analysis of variance using the Proc GLM option was run for 3
crop factors at 2 phosphorus levels within 8 aggregate size fractions for %
carbon, % nitrogen, mass of carbon, mass of nitrogen and C: N ratio. Seasonal
changes associated with these components were also examined. In addition,
hypotheses were decomposed into single degree of freedom contrast questions
in order to investigate the pair wise cropping system effect on biomass and
nutrient yields. The slice option was used to compare significant interaction data
for the effects of crop at P levels and P levels for cropping factors. The objective
is to gain insight into the interaction structure without physically splitting the data
and yet retaining the power of the overall analysis. In the absence of a P

interaction, data were pooled in order to increase the power of the F-tests.
RESULTS AND DISCUSSION
Carbon and nitrogen distribution in soil aggregates sampled from 0-15cm

soil depths followed a seasonal pattern. Initial P fertilization in March 1997 did

not significantly affect C and N weight or concentrations of aggregate size

15

fractions in February and July 1998. There were no significant effects of cropping
system on % C, % N, mass of C and mass of N in February 1998 (Table 1).
During this time, lack of significant differences may be influenced by the dry
season (December to February). In July 1998, significant differences in % C, %
N, mass of C and mass of N were observed both across aggregate size fractions
and among crop rotations. Changes in aggregate mean weight diameter by dry
sieving showed a breakdown of large macro aggregates (>4mm) into smaller
macro aggregates (<4mm) (Fig 1.). During the same time period, there was little
change associated with aggregates <2mm. Soil aggregates are sensitive to
management practices (Six et al., 1998). Upon cultivation, macro aggregates are
expected to break down (Tisdall and Oades, 1982) into micro aggregates <
0.053mm (Six-et al, 1998). Large macro aggregates are less stable. They are
probably bound together by less highly processed aliphatic and more stabilized
or slowly decomposable fractions of OM (van Veen and Paul., 1981 ). However,
SOM may not be the principal binding agent in Oxisols as iron and aluminium
oxides contribute to the aggregation process (Oades and Waters, 1991 ). Kay et
al (1988) suggested a model to propose a rapid decline of aggregate stability
during conversion of a native old meadow to cultivation. Angers, (1992) reported
little change in a single plow pass but recorded further declines in aggregation
with repeated cropping Baldock et al. (1987) found that changes in aggregation
following modifications in the cropping system were not always associated with

changes in total organic carbon. In this study, changes in aggregate stability may

16

have resulted from seasonal influences determined by tillage, crop root activity

and wetting and drying cycles in the field.

Aggregate fraction sequestration of carbon

Combined values of aggregates 4mm and larger accounted for 72% of the
total C by mass in February 1998 but declined to 31% in July 1998 (Fig 2). The
proportion of C and N sequestered in aggregates 0.5-4mm was highest in July.
Perhaps physical disruption of aggregates across the season had resulted in re-
allocation of C among size fractions. Further, an increase in the mass of C was
observed in aggregates <4mm (Table 2). In contrast, carbon concentration
increased in the 4-8mm fraction but decreased in the 0.038-0.053mm fractions.
No change was observed for all other fractions. In the same period, there was no
significant change in the total C of the bulk soil. The highest decline in the mass
of C was observed in large macro aggregates >8mm (Table 2). Redistribution
among smaller fractions resulted in gains ranging between 700-3000 mg C kg".

Numerous variations in the carbon and soil organic matter associations
among aggregate sizes have been reported. (Angers and Giroux,, 1996 ). Elliot
(1986), and Cambardella and Elliot (1993) reported decreasing contents of C and
N with decreasing aggregate size. Earlier, Baldock et al., (1987) had observed
elevations in both C and carbohydrate content with decreasing aggregate size
while Gupta and Germida (1988) and Beare et al (1994) found no conclusive
trends in C concentrations among size fractions. It is probable that some of the

uncertainty associated with locations in SOM, C and N is related to the seasonal

17

migration of C and N (Fig 2). Among other factors, these seasonal trends would

likely depend on climate, soil type and management.

Aggregate fraction sequestration of nitrogen

Significant trends in nitrogen location (Table 3) and distributions (Fig 3)
between February to July 1998, were similar to those observed with carbon.
However, macro aggregates larger than 4mm lost 644mg N while aggregates
0.053mm-4mm gained 639mg N. Therefore, any changes that occurred in the
total nitrogen pool were insufficient to cause significant losses of N within

aggregate size fractions.

Influence of tree-crop rotation on carbon and nitrogen distributions
Significant differences of % C and % N among crop rotations before maize
planting in February 1998 appear too close to be of practical importance. These
differences did not reflect alterations in the mass of C and N. Crop rotations
significantly affected C and N deposits (Table 4 and 5) and the subsequent
changes in C and N, after maize planting in March 1998. During the growing
season, Crotalaria and Tithonia tree fallow rotation with maize increased the
mass of N in contrast to maize-maize which, caused a depletion (p < 0.10, data
not reported). Seasonal changes showed that, only tree-crop rotation increased

the mass of C and N .in July (p=0.08, data not shown)

18

Carbon to nitrogen ratios

Significant differences among aggregate size fractions and crop rotations
(Table 6 and 7) suggest that an aggregate by crop rotation interaction was the
most important process influencing decomposition of recent organic inputs. We
hypothesize that recent inputs associated with and responsible for binding large
macro aggregates are transferred to smaller fraction sizes during decomposition.
This process narrows the C: N ratio and precipitates an enrichment in N rather
than C of aggregate 0.5-2mm across. Six et al., (1998) observed but failed to
explain a N enrichment of the coarse Inter aggregate particulate organic matter

(iPOM) in slaked aggregates.

Yields

Maize crop productivity for the first post-fallow season determined by grain
yield was related to nitrogen availability (Fig 4). The larger quantities of crop
residues that were returned to the fallow plots (Table 8) might have supplied
sufficient plant nutrients to support yield levels 2-3 times higher than the maize
control (Fig 5). Phosphorus addition to tree crops significantly increased P in

tissue and the yield of P in biomass but not N or K (data not shown).

CONCLUSIONS

There seems to exist a time by management interaction, which controls

the allocation of carbon and nitrogen within soil aggregate size fractions. Early

19

evidence has been presented to propose concentric sequestration of surface
applied material (eg crop residue C and fertilizer N), with the most recent input
being located on external regions of aggregates (Smucker et al., 1999; Santos et
al 1997). Development of gradients of C would be expected to diminish with
decreasing aggregate size as more highly processed, aliphatic or stabilized SOM
(Paul, 1984) becomes incorporated into the humic fractions of soil. Past models
do not seem to be adequate in explaining some of the differences occurring
when SOM is incorporated into soil aggregates. This study suggests a migratory
pattern of carbon and nitrogen that appears to be controlled by soil aggregate
stability. Perhaps this pattern is a controlling mechanism in the breakdown and
reformation of soil aggregates. Fallow cropping with tree-crop rotations rather
than only maize does not seem to deplete C and N stored in soil aggregates, but
maintains high productivity. Further research. is needed to separate these control
mechanisms in order to understand management modifications of C and N

sequestration in low input agriculture.

REFERENCES

Angers D A, 1992. Changes in Soil aggregation and organic carbon under corn
and alfalfa. Soil Sci. Soc. Am. J. 56:1244-1249.

Angers D A, M A Bolinder, M R Carter, E G Gregorich, C F Drury, B C Liang, R P
Voroney, R R Simard,R G Donald, R P Beyaert, J Martel, 1997. Impact of

tillage practices on organic carbon and nitrogen storage in cool, humid

20

soils of eastern Canada Soil Tillage Res 41: (3-4) 191-201.

Angers D A and M Giroux, 1996. Recently Deposited Organic Matter in Water-
Stable Aggregates. Soil Sci. Soc. Am. J. 60:1547-1551.

Baldock, J.A., B.D. Kay and M. Schnitzer. 1987. lnfuence of cropping treament
on the monosaccharide content of the hydrolysates of a soil and its
agregate fractions. Can. J. Soil Sci. 67: 489-499.

Beare, M.H., M.L. Cabrera, P.F. Hendrix, and DC. Coleman. 1994.
Aggregate-protected and unprotected organic matter pools in conventional
and no-tillage soils, Soil Sci. Soc. Am. J. 58:787-795.

Cambardella. CA, and ET Elliott. 1993. Carbon and nitrogen distribution in
aggregates from Cultivated and native grassland soils, Soil. Sci. Am. J.
57:1071-1076.

Christensen 8 T 1992. Physiwl fractionation of soil organic matter in primary
particle sizes and density separates. Adv. Soil. Sci. 20:1 -90.

Elliot, ET. 1986. Aggregate structure and carbon, nitrogen, and phosphorous
in native and cultivated soils. Soil Sci. Soc. Am. J.50:627-633.

Gachengo, C.N., Palm, C.A., Jama, B. and 0thieno, C. 1999 Tithonia and
senna green manures and inorganic fertilizers as phosphorous
sources for maize in Western Kenya. Agroforestry Systems 44:21 -36.

Greenland D G, 1995. land use and soil carbon in different agroecological
zones. Pp 9-24. In: R Lal, J Kimble, F Levine and B A Stewart (eds), Soil
management and greenhouse effect. Advances in Soil Sc. Series. Lewis

Pubs, Boca Raton.

21

Gupta, V V S R and J J Germida, 1988. Distribution of microbial biomass
and its activity in different soil aggregate classes as affected by cultivation.
Soil Biol. Biochem. 20:777-786.

ICRAF, 1999, Annual report 1998, lntemational Ctr for Res. In Agrofor., Nairobi,
Kenya.
Jastrow, JD. 1996. Soil aggregate formation and the accrual of particulate and
mineral-associated organic matter. Soil Biol. Bio-chem. 28: 665-676.
Jastrow J D and R M Miller, 1997. Soil aggregate stabilization and sequestration:
Feedbacks through organomineral associations In; Soil Processes and the
carbon cycle. Ed Lal R, J M Kimble, R F Follet and B A Stewart 15: 207-
23.

Kay , B D, D A Angers, P H Groenvelt and J A Baldock. 1988. Quantifying the
influence of cropping history on soil structure. Can. J . Soil Sci. 68: 359-
368.

Kemper, WA, and PC. Rosenau. 1986. Aggregate stability and size
distribution. p. 425-462. In A. Klute (ed) Methods of soil analysis Part 1.
2"" ed. Agron.Monogr. 9. ASA and SSSA, Madison, WI.

Oades, J.M. 1984. Soil organic matter and structural stability: Mechanisms
and implications for management. Plan Soil 76:319-337.

Oades J M and A G Waters. 1991. Aggregate hierarchy in soils. Aust. J Soil
Research. 29:815-828.

Paul E A, and R P Voroney, 1984. Field interpretation of microbial biomass and

activity measurements. In: Current perspectives in microbial ecology, M J

22

Klug and C A Reddy, eds pp 509-514. Am. Soc. of Microbiology.

Parkinson, J.A, and SE. Allen. 1975. A wet oxidation procedure suitable
for the determination of nitrogen and mineral nutrients in biological
materials. Commun. Soil Sci. Plant Anal. 6:1-11.

SAS Institute, 1999. SAS/STAT Users Guide Vol. 2 6th ed. SA Inst, Cary, NC.

Santos D., S L S Murphy, H Taubner, A J M Smucker and R Horn, 1997. Uniform
separation of Concentric Layers from Soil Aggregates. Soil Sci. Societ
Amer. J. 61: 720-724.

Six J., E T Elliot, K Paustian and J W Doran. 1998. Aggregation and Soil Organic
Matter Accumulation in Cultivated and Native Grassland Soils. Soil Sci.
Soc. Am. J 62:1367-1377.

Smucker, AJ.M., C.J. Dell, J Kinyangi and JD. Jastrow. 1999. Quantitative
removal of concentric layers from soil aggregates by soil aggregate
erosion chambers. ASA-CSSA-SSSA, 91St Annual Meeting, Salt Lake
City, Utah, Oct.

Szott L T, C A Palm and P A Sanchez. 1991. Agroforestry in Acid Soils of humid
tropics. Adv. Agron. 45:275-301.

Tisdall, J.M., and J.M. Oades. 1982. Organic matter and water-stable
aggregates in soils. J. Soil Sci. 33:141-163.

Voroney R P, J A van Veen and E A Paul., 1981. Organic carbon dynamics in
grassland sooils2. Model validation and simulation of long term effects of

cultivation and rainfall erosion. Can. J. Sooil. Sci. 61, 211-224.

23

Eng—Eu.» no: .. a:

.28.. nausea 22.: 9 8.988 .26. 3... a... 3... .8851. 8:858... 8.2% .. =5 t .t.

 

 

 

 

a.” Qua ZN a... E 3 EN «.8. 3 3 $5
a... a: a: a: a: t... r... t... t... t 3 3339220
.5 t s t... t... t a: a: a: a: N no.6
......asT... is... .i . 11.3271. :18 I.
oz! 2 o z 0 can. 2 o z o “E efiaom
26 2.6
22. Duane".

 

do I = .32. .232. Eu» #3 62.5 5E8. a .0 SEE? coach...
can coated so on? 8323.. new £30? 95398 Co “onto 05 Lo.— oecatg ho gee-:2 .5. sea...

24

8.90m.— uSeouca 223E 8 9:233

2235. $5053.» «0: 2a .632 2:3 2: an 332.8 5.5.3 o :. Enos.

 

 

 

tau. 3:. nofid =3. ova." «modéncd
town” oumw ouaé 03.0 000.9 mudnnmod
near... tnooN cum... on; 39—. mdéud
nae—.0 achv emu—um UNN: om...~ 70.9
32.2. can: ohvfl cast" “Cr-um N.._.
gas ommmu ouaofi ocean tam. —. TN
come. 9.. unvocn «cod acuov akin mu?
50mm. DOONn 25a. _. saga Una. _. aA
13:: I. 1-3.5 I EAST. 1.3.5 I IAST- :55
Case. < beacon. .3302.
on? 83952

coarse =3

 

.82 .58: :82... 8.32 serous Lo «55:. 3
Lots 32.6 5.38. a Co 23.88.— ou.» 8383a 55.3 c2255»... :33“. =0» .N... 03:

25

.23.. nausea 22.2”. 9 8.988 2.26% 5:85:22

«e: 2a .382 oan 2: an 6952.0.— cEEOo a E Enos.

 

26

 

 

3. a 83 a 83 «3.38.:
on» 8» .83 .89 83... 3.33...
89 .8: .83 as 8.... 3.2.:
«a: 2.3 8:. oz. 3.3 E...
an." «S» 83 a: a; 3
e3 8: 83 82 33 7a
.2... BEN 83 88 «a... 3
2%. SE 2.3 82 at... an
as. 82. < 16.5 IE..- tea. 1 13...... :55
2:... basic... 5.6!:
on.» 2aa25u<

mass... :8 l

 

.32. £52. Eon Tc 60:32 3.20.323— 3: .0 23:2: 2. a35:2
.320 FEES. a no «Samoa? :3 55.3 8.3%.. couch... =0» .n... 032.

 

 

 

n E 3.0 new 3.0 8... am.—
soou a _..u . «02.. m w." 3.05.. 5205:.
«NR «Nu 2.3 5 _..~ 3.0.2. 5.5.0.20
Eva no.“ $3 _. _..~ 0505.050!

19.....- is..- .12.... 53..
2:... >523... coafio. 020
5&3 =0m

 

.82 .508 £83 .323 ages. a .0
cofiau =00 00 5:230:03 :0 00:03:... E80: 9.3020 .v... 030k

27

 

 

 

t. m::.: :u m::.: 8... 90..
:«N 0...: 08 00...: 00.0.2. 0.005....
SN 0...: SN 00...: 00.02.0t0520
«94. 5...: 0:~ 00...: 00.02.0305.
:05... .....\.T :35... as...
:2. E0200". 00:50. 00.6
$006.... =8

 

.003 £300 Eon ...: .0005 Size. 0 .0
0000.0... =00 .0 00305000000 :0 00:02.0. 830.3 00.0020 .0... 0.00...

28

00.0DWJ Dcuuaachn Eon—um Ow

00.0.0000 «00.050 2.0005090 .00 0.0 .030. 0.000 05 .3 0032.8 0.00.00 0 0. 0000....

.30.......:... 000000

00.00000 00.0.0 05 000.00 0000.050 0.... 00 00000.98 0. 0.00000 2 000 o 0... 0. 000000

 

 

 

noad. 00.2. 0a.»; nmodénod
02.. —... andw n5.3. mudémcd
can. 9.. on rd —. U092. mdéud
and. andw nude Ted
08..»- aN.N_. 00.2. N...
0:0. F. to». _. _. 30.0.. 7N
auto. 09—... 00.: cu?
0~N.o.. 0th. 2. 0a.: cA
.000.
Jag—E < 3.... E0200“. 00.000...
00.0 0.0ao..ma<

2.0

 

.000? .5000 .000 To 000.000.. 00.0 0.000.600 00 0.00.. 2.0 0... 0. 0000000 .0... 0.00...

29

8.50.. 00.00.20 0.000.“. 0. 00.0.0000 «00.0...0 2.0005003
. .00 0.0 .000. 0.000 0.... >0 0032.0. 0.00.00 0 0. 00002

 

 

 

09.0- 0.20.. 00.0.. 00.02. 0.005....
000.... 00.... 00.0.. 00.02. 000.820
0.0...- 00.0.. 00.0.. 00.02.00.02

.30.. E0200“. 0000.0. 00.0
2.0 < 2.0 2.0

 

.002. .5000 .000 «.0 £000.00 2 000.00
.0 000. 0... 0. 000000 0... 000 000000 00.0020 00.02 .0... 0.00...

30

00.0... 00000003 0000.88 0.20 0.000.... 000 2.20.0.0 0. 00.0... 00.0.0 .0 000000.000:
. .000...00.0 .00 - 00

3.000.. 00.00.20 0.000.... 0. 00.0.0000
..0>0. 00.0 000 .0... 300.9 .0 00000...00.0 00.0000 .. 000 s. .3.

 

 

. .80 0....” L000. .0003... 0.2. 00.0.2

.. on 00 .8532 2.0.8. c. 0. 0:0...

.. e 2 .8530. 30.8. s n. .5...

on 03 02:39. 2.0.8. c. z .5:

8.: 02.0 .8538 2.0.02 09.0

a: 38 080 0.2. 30520 3.. .23

............. P.0... 0...----------
.26. 00000....H5 00.02- 0.005.... 00.02 02.0.0. 00.0
- 2.0.0.000

 

02.0.0. 00.0

 

.QmQF flaw-0; 5.0 @260:— :_ w@OCO.¢t_U 00h mamflhucou 80300.: “—0 caumcu 2W2.” .w.—. gnflh

31

 

 

 

 

 

 

2.5
2.0 (a) Iain-lain ”m"
T DJuly
é
2,5 .
2.0 » (b) Crotalaria-Iain lFobm-ry
may

 

 

 

 

 

 

 

 

 

3 I 1'! I;
O

0 20-0 0
0000-010
0 030-0 003

W the fraction (mm)

Fig. 1.1. Mean weight diameter changes 0fdry sievedsoil aggragatasas
affochd by main-tree crop rotation on a Kenyan Oxisol (0-1Scm), 1998.

32

EEOA I

8.00.0 n.

EETN D

EN... 2

.00. 9.0.0 B

0.50.0.0"... n.

EEoN.0.nnc.0 E

520006.000... n.

.002. .5000 Eur. v.0 ._00.xO

00.33. 0 .0 000000.. 00.0 0.000.000 =00 0. 0000. 000.00 .ué .0...

200

 

000.00..

$0
$0 ..
$0N
$00
$00.
$00
$00
$05
$00
$00
$00..

“093314 .1!’ u! “W30

33

580A I

550.0 n.

550.0 D

550... D

55 9.06 B

5506.006 D

550060006 B

5500060006 D

600—. .5000 500 v.0 .3005

00003. 0 .0 000000.. 00.0 8000.000 =00 0. 0000. 0000...: 0.. .0...

0....

000.00..

 

$0

$0..
$00
$00
$00
$00

. $00

$05
$00
$00
$02.

009984 on: 0! uoflonm

7.0

6.0 .

5.0 -

4.0 - 0

3.0 . ' o

2.0 + 0 .

1.0 - 0 .

0.0 1 . . . . J
-10 0 10 20 30 40 50 60

 

(a) O Total-N

 

 

Maize grain yield T ha"1

 

 

7.0
6.0 - (b)
5.0 .
4.0 i- .
3.0 ~

Maize grain yield T ha"

 

 

0.0 J l l l J l

-10 0 10 20 30 40 50 60
A Inorganic-N kg ha"1

 

Fig. 1.4. Pro-season changes in the inorganic-N pool, with (a) or without
(b) phosphorus fertilization and their relationship to August 1998 maize
grain yield. Data are for continuous maize and plots converted to rotation
with Crotalaria grahamiana, a leguminous tree. Changes in N are
measured at the onset of the maize croppng season in February-March
1998.

35

 

.000. .0500. 0.0.00.5 .000". 0000 .0 0.00.9.0
05000.0 00.0. .0 0.0... 00080.0 00.. 000 0.0.0 00.02 .0... .0...

00.02.0505: 00.02.000.520 00.0.2. 00.05.

 

 

 

 

 
 

«000 000 0

.9. 32 3.00.. 0.0.0 I
.9. .30 3.00.. 00006.0 I

 

 

000m

000*

0000

0000

0000..

000m —.

000.;

(Pat: 6):) pram

36

Chapter 2

CONCENTRIC SEQUESTRATION OF PHOSPHORUS IN SOIL AGGREGATES
OF MAIZE AND TREE BASED SYSTEMS IN WESTERN KENYA

INTRODUCTION

Soil phosphorus stocks in Sub Saharan Africa have decreased as
increasing population has led to replacement of traditional systems of shifting
cultivation with shorter duration fallow systems and sedentary agriculture
(Buresh, et al, 1997). As traditional fallow systems are replaced by continuous
cropping, soil organic matter (SOM) declines leading to reduced supplies of plant
available P. Reduced SOM and lower P mineralization of organic phosphorus
(Po) leads to greater reliance upon external P inputs. Elimination of P deficiency
can enhance N2 fixation by legumes (Giller et al, 1997) and integrated use of P
fertilizers with organic materials to supply K and N can potentially enhance P
availability (Palm et al 1997). In many ecosystems, including small farm holders
in Africa, carbon (C) is the most limiting factor controlling microbial biomass and
nutrient turnover. Huffman et al, (1996) suggested a close linkage between
cycling of P through plants and soil biota is closely linked with C cycling. Others,
(Gressel et al, 1996) have considered P as a component of organic material
'quality' and a factor in determining decomposition rates. Gressel et al, (1996)
expressed concern that extraction based indices and phosphatase activity failed
to explain P availability. They further proposed that chemical linkages with C-

structure might be an important factor controlling P release; in agreement with

37

Tiessen et al, (1992) who found that NaOHPo declined at about the same rate as
C in a Brazilian soil. It is most probable that the availability of P is largely
determined by organic equilibria but that current methods for studying these
equilibria are poorly developed.

During the formation of soil aggregates, many components of the soil are
redistributed within the structural unit. Equilibration of developing aggregates with
the various biogeochemical activities result in the development of gradients
within individual soil aggregates (Horn, et al., 1994; Santos et al, 1997).
Aggregate erosion of concentric layers using stainless steel meso soil aggregate
erosion (SAE) chambers (Smucker et al, 1999), is a new method being
developed for separating multiple surface layers from soil aggregates, by gentle
erosion, provide samples for further analysis. Quantification of these separate
components could enhance our understanding of the fate of surface applied
materials such as crop residues, fertilizer granules and pesticides that are
controlled by the formation, breakdown and reformation processes of soil
aggregation. Standard P-sorption isotherm methods, which grind and vigorously
shake soil (Fox and Kamprath, 1970), appear to expose diffusion limited P-
sorption sites, overestimating the buffering capacity of field soil. Based on
sorption isotherms of an aggregated soil, Cassman et al, (1981) estimated and
recommended >500kg P ha‘1 for optimum yields but Cassman, (1993) reported
that 100% of maximum yield could be achieved by applying 100kg P ha". Other
procedures that extract labile P from throughout the aggregate may give

potentially erroneous values of available P since some of this P may be

38

unavailable to plants due to its physical sequestration within aggregates (Linquist
et al, 1997).

Despite extensive research on soil P reactions, little is known about the
effect of soil aggregation processes on these reactions. Nutrients commonly
occurring at low concentrations in soil solution, such as P are transferred to root
surfaces at insufficient rates to meet uptake requirements of plants. Root uptake
of P results in a decrease in their concentration in the soil solution surrounding
plant roots. This depletion generates a concentration gradient at the soil-root
interface. Linquist et al, (1997) demonstrated that P sorption is related to
aggregate size, soil buffering capacity and soil P release rates. They observed P
retention on surface regions of soil aggregates 0.188mm across, which they
described as mean reactive mass. Therefore, P flux between .the surface layer
and interior regions of aggregates appears to be a timedependent process
controlled primarily by diffusion. These concepts led to the hypothesis that P
sequestration within interior regions of soil aggregates adversely affects short-
term plant available phosphorus. The objectives of this study were to determine,
i) the physical location of recently applied phosphorus within soil aggregates and
ii) the effect of planted tree-crop rotations on short term plant available

phosphorus and its relationship to yields.

39

MATERIALS AND METHODS

Site description

The study site of this project is located at Osita Farm in Vrhiga District,
Western Kenya (00 06’ N, 34° 34’ E) at an elevation of 1420m. Mean annual
rainfall is 1800mm (split between 2 cropping seasons, February-August and
September to January). Often, the first cropping season will receive 1000-
1200mm of the annual rainfall. The soil type is a very fine lsohyperthermic
Kandiudalfic Eutrudox (USDA, 1996). It has few chemical or physical barriers to
rooting in the top 4m. Air dried soil for the top 15cm had a pH of 5.1 (1 :25
soilzwater suspension), organic carbon 179 C kg", ex. acidity 0.4 cmolc kg",
calcium 3.4 cmolc kg", (1.0 M KCL extractable), Ex magnesium 1.4 cmolc kg",
ex potassium cmolc kg" 0.12, phosphorus 2 mg kg", (bicarbonate-EDTA
extractable). Soil textural distributions were 46%, 28% and 26% for clay, silt and
sand respectively.

The experiment consisted of a factorial combination of three cropping
systems (maize-maize, Crotalaria-maize, and Tithonia-maize) and two rates of P
applimtion (no P and 50kg P ha"). Phosphorus was applied in March 1997as
broadcast and incorporated into the top 0-15cm depth at 50kg P ha" (as Triple
Super Phosphate, 46% P205). Maize (HE-512, Kenya Seed Co, Kitale, Kenya),
was planted at 2 seeds per hill, at a spacing of 0.75m between rows and 0.25m
between plants, and later thinned to 1 plant per hill to achieve a plant population

of 53,000 plants ha". The tree crop was relay planted into maize during March

40

1997 and April 1998. One row of tree crop was planted for every row of maize
with the same intra-row spacing, resulting in the same tree plant population.
Maize crop was harvested for grain yield in August 1998 and July 1999 in the
tree-crop rotation plots. Two additional maize crops were harvested in January
1998 and 1999 (continuous maize plots). All maize crops were assessed for
grain yields. All above ground maize crop biomass was removed since it was
common practice for farmers to harvest the stovers for livestock feed. Therefore,
only root C, N, and P were considered for maize trials. The tree crop was
harvested in February 1998 and 1999. All residue and litter fall was weighed and
returned to the plots (Kinyangi et al., unpublished data), while the woody
components were exported from the plots for fuel wood. Only the residues
returned were considered as directly affecting the fertility status of the soil. Leaf
litter and pods were harvested and oven dried at 60 ° C for dry matter and plant

N, P and K determination (Allen and Parkinson, 1975).

Soil sampling

Soil from the field trial was sampled on the following dates; 13 February
1 998 (end tree crop season but before tillage), 7 July 1998 (post tree crop
season and maturity of residual maize), 19 September 1998, (initiation of 2"d tree
crop season), 13 February 1999, (end of tree crop season 2 and return to maize
cropping) and 1 July 1999 (post tree crop 2 and maturity of maize). Soil samples
were collected (0-15cm depth) with sharpened cores (10cm diameter by 15cm

height) which were gently driven into the soil surface to minimize disruption of

41

aggregates. The soil was loaded onto trays secured with rubber strapping and
transported to the laboratory. Air-dried samples (1 kg) were manually shaken
through a nest of sieves, separating them into eight size fractions (>8mm, 4-
8mm, 2-4mm, 1-2mm, 0.5-1mm, 0.25-0.5mm, 0.053-0.25mm and 0.038-
0.053mm). We followed P dynamics during the tree crop season in the 4-8mm
aggregate size fraction. This size fraction contained 40% of the total carbon and
nitrogen, by mass, for all aggregates <8mm, (Kinyangi et al, unpublished data).
This fraction of aggregates contains the least depth of the reactive mass layer as

defined by Linquist et al, (1997).

Aggregate erosion

Soil aggregates (4-8mm across) collected from the experiment were
further processed by eroding individual aggregates (n = 15) into external,
transitional, transitional and internal regions of equal mass (0.15 9) using
stainless steel meso soil aggregate erosion (SAE) chambers (Santos et al.,
1997). Bases of the SAE chambers were fitted with screens with 356pm
openings. Individual aggregates were placed into a SAE chamber (2.5cm
diameter) with finely knurled walls, the top was sealed with a rubber stopper
secured with para film to prevent dust contamination and secured into glass
beaker assembly (100ml) using foam packing. The glass beaker assembly was
rotated on a variable speed rotary shaker (lnnova 2300, New Brunswick,
Canada). As the aggregate rolled around the inside of the knurled SAE chamber

the fine (<356pm) fraction fell through the sieve and was deposited in the lower

42

base (Fig 1). Additional protocols for uniformly removing equal quantities of 33%
from each concentric layer are described by Smucker et al, (1999). In short,
samples were weighed periodically to determine the time required too remove
each concentric layer/region from each soil aggregate fraction processed from
each treatment (Fig 2). The final eroded samples obtained were passed through
a 0.053mm sieve to exclude the sand fraction as most of the resin extractable

phosphorus (REP) varied with sand content ( (Fig 3).

Modified resin extractable phosphorus.

Plant-available P is a functional concept rather than a measurable quantity
(Tiessen and Moir, 1993). Further, it may be described as solution inorganic P
(Pa) which is continuously recharged by that P; which enters the solution during
the period which defines availability during the growing season. Exchange resins
have recently gained importance in identifying plant-available Pi. The resin
offsets the equilibrium between dissolved and soluble P; allowing more
precipitated forms to enter the P depleted solution. The P sorbed by the resin is
extracted by a weak acid and quantified spectrophotometrically. This pool has
often been equated to the isotopically exchangeable P since the resin does not
chemically modify the soil solution. Linquist et al, (1996) as well as Guo and Yost
(1998) have shown that extraction by Fe-impregnated paper can also
characterize this pool. However we chose to use the exchange resin approach
because it provided the best opportunity to determine available P in the presence

of a P sink similar to the root activity of growing plants.

43

A variety of methods have been developed and adopted using different
anionic forms, shaking times, ratios of soil:water: resin for enclosed bags or
beads in free suspension (Sibbesen, 1977; Sibbesen 1978; Tiessen, 1991,
Tiessen, 1993). Enclosed bags are commonly used because they offer the
advantage to regenerate the resin by washing through HCL baths. It seems
however that the decision regarding ratios of a resin-P to soil mass should be
determined by the objective of the research. A resin membrane (BDH No. 55164,
9‘62mm) extraction of 0.59 soil in 30ml water was described by Tiessen and
Moir, (1993). Maroko et al, (1999) used mesh bags (Dowex 1-X8, 1g converted
to bicarbonate form) to extract 2.59 soil in 40ml water. In a widely adopted
sequential extraction, Hedley et al, (1982) used resin enclosed in nylon bags
(Dowex 1-X8-50, 0.4g converted to bicarbonate form) to extract 0.5g soil in 30ml
of de-ionized water (DlW)., Agbenin et al, (1998) have also used the Hedley
(1982) fractionation to characterize available P.

We modified the resin in mesh bags approach in order to: a) establish a
minimum soil sample weight required to extract a detectable P level which could
be removed by the resin. This approach was based on the fact that soil eroded
from concentric layers of aggregates were small with a mass <200mg; b) vary the
resinzsoil ratio to achieve maximum sensitivity of extraction while reducing in-
sample variability and c) determine if the method could adequately characterize
available P in concentric layers of soil aggregates. We selected soils with a range

of resin extractable P and varied the ratio of resin: soil (1, 2, 5 and 10) at 3 soil

weights (0.59, 0.259 and 0.1259). The soil was collected from the experiment on
7 July 1998 in the following plots;

Maize-maize, no P applied, (Soil 1)

Maize-maize, 50kg P he" as initial application in March 1997, (Soil 2).

Maize- maize, 50kg P ha" with a repeated P application in March 1998,

(Soil 3).
P was extracted by shaking soil samples in 60ml bottles for 16 hrs at 175
reciprocations min" with a Dowex resin (1-X8,Cl', Sigma Chemicals, St. Louis,
MO) enclosed in mesh bags (5mm*4mm, 83pm opening) in 40ml of de-ionized
water (DlW). The resin was supersaturated with sodium bicarbonate (0.5M).
Phosphate ions adsorbed on resin were extracted by shaking the P-resin,
retained within the mesh bag, in 30ml bottles with 15ml of 0.1 M HCl extractant for
1hr to establish equilibrium. The extract is filtered through Whatman paper (no.
1). Phosphorus was determined spectrophotometrically (Murphy and Riley,
1962), on an automatic Lachat (Quickchem Systems, Milwaukee, WI).
Ammonium molybdate and potassium tartarate reacted with dilute solutions of
phosphorus to form antimony phospho molybdate complex. This complex was
reduced to an intensely blue colored complex by ascorbic acid. The color,
determined at 900nm, was proportional to the phosphorus concentration in the
sample. Minimum detectable limits of P were 0.01 ppm.

Varying the resinzsoil ratio extracted different amounts of P (Fig 4). An
increase in resin load relative to the soil weight resulted in more REP. The

extraction seemed to increase with the amount of available phosphorus. This

45

suggests that there is need for higher resin loading in P fertilized soils, as well as
soils undergoing rapid organic matter transformation that releases phosphorus.

In contrast, a lower resin load, as is common in most present extractions seems
to under-estimate available phosphorus. Significant differences in available P
with decreasing sample weight (Table 1), may be due to greater concentrations
of P in solution, since the volume of DIW we used remained constant at 40ml.
We concluded that the best P extraction occurred with 0.1259 of soil sample and
0.6259 of resin at a ratio of 5:1 (resinzsoil). Although within sample variability was
comparable across extraction ratios, 5 :1 offered the best opportunity to

maximize P extraction. from soils with both low and high P concentrations.

Exterior to interior ratio (EIR).

Ratios of the measured REP (ie., interior to exterior ratio - EIR) were used
to estimate the relative allocation of one or more P pools within concentric layers
of soil aggregates (Kinyangi et al, unpublished data). This provided an evaluation
of the physical location of P from recently applied fertilizer or P from plant residue
materials. These evaluations were used to compare the ElR relationship to crop
productivity for the crop rotations reported in this study. Additionally, changes in
the EIR were monitored over time to observe shifts within pools. Studying these
dynamics was intended to yield knowledge of residual P benefits with the intent
of predicting the need for fertilizer P replenishment, using estimates of a “critical

ratio”. The EIR was calculated by the following equations:

46

ElRt1= [layer (Xt1)llayer (Yt1)] (Eq. 1)
ElRt2= [layer (Xt2)/layer (Yt2)] (Eq. 2)
Change in EIR =[EIRt2-ElRt1] (Eq. 3)
Where:

ElRt1 is the ratio at time 1

Elth is the ratio at time 2

Xti is the exterior value of P (mg kg") at time 1

Yt1 is the interior value of P (mg kg") at time 1

th is the exterior value of P (mg kg") at time 2

Ytz is the interior value of P (mg kg") at time 2

Statistical analysis

Analysis of variance was performed using the Proc GLM option of the
Statistical Analysis System (SAS Institute, 1999). All significant tests reported,
refer to a = 0.05 unless otherwise stated. Individual hypotheses were
decomposed into single degree of freedom contrast questions in order to
investigate the individual cropping system or combined fallow effect on available
phosphorus. In the absence of a significant P interaction, data were pooled in

order to increase the power of the F tests.

47

RESULTS AND DISCUSSION

Plant available phosphorus in whole aggregates

Seasonal patterns in resin extractable P of whole aggregates between
March 1997 and February 1999 were influenced by the presence or absence of
maize and tree crops (Fig 5). Initial application of 50kg P ha" to planted maize
and trees in March 1997 increased REP of whole aggregates in February 1998
(data not shown). A modest addition of 50kg P ha" as triple super phosphate in
the 5th season during March of 1999 increased resin extractable P (REP) within
all three cropping systems (data not shown). However, these increases of
available P in whole aggregates across 5 cropping seasons were not significant
probably due to high variability associated with heterogeneity among the different
soil aggregates. More replications of samples, up to 3 times, in the laboratory did
not seem to effectively address this problem. Extractable P peaked at 12mg P kg'
1 in March 1998 but declined over the maize cropping season (March 1998-July
1998). A build-up in this P pool was observed in September 1998 at the initiation
of the 2"‘1 fallow season. The absorption of P by growing maize and tree crops
was observed in February to July 1998 and February to July 1999.

It is known that the interaction between biological (organic) and chemical
(inorganic) transformation determines P availability. Considerable research
(Hedley et al, 1982; Tiessen et al, 1984; Stewart & Tiessen, 1987; Beck &
Sanchez, 1994) has revealed the existence of P fractions of differing

extractability based on their functional status. Phosphorus can be partitioned into

48

readily available, reversibly available and sparingly available pools based on a)
accessibility to plant roots, b) reversible conversion into available P and 0) pools
that undergo long term transformations which however may still contribute to P
cycling in the short term. In this study, the appearance of a trend, which lacked
significant differences among management systems, may be an indication that
REP interacts with additional pools that we did not measure. Tiessen and Moir
(1993) described this concept as an activity where less soluble or slower pools of
P replenish available P at varying rates.

Previously, (Kinyangi'et al, unpublished data), we reported changes of C
and N in aggregate fractions. We then correlated the fraction with the largest shift
in the C: N ratio (1-2mm) and larger macro-aggregates (>8mm) with changes in
REP (Fig 6). There is a strong relationship between changes in C concentrations
of aggregates 1-2mm across and P turnover during the post fallow season. This
observation suggested that P availability was influenced by a negative change in
C concentrations. It is speculated that this negative change was probably a result
of C-mineralization. This would support the proposal by Tiessen et al (1992) who
reported declines in the NaOHPo organic pool, at about the same rate as C

mntents declined in a Brazilian soil.

Comparisons of EIR P to gradients in plant available P
EIRs of resin-extracted P from concentric layers of soil aggregates,
reported in this study, appeared to be a sensitive and early indicator of plant

responses to cropping and fertilizer management modifications. Wan and El

49

Swaify (1998) reported an enrichment ratio (ER) and changes in the ER to study
mechanisms responsible for enrichment of organic carbon and extractable P in
fine aggregates of an Oxisol. Elsewhere, carbon to nitrogen (C: N) and carbon to
phosphorus (C: P) ratios have been used to demonstrate mineralization-
immobilization phases of carbon, nitrogen and phosphorus cycling. Smucker et al
(1997) used clay mineralization ratio (CMR) to show changes associated with
accelerated mineral weathering in the external layers of soil aggregates from
multiple locations and ecosystems.

During this two year study, the EIR of available P in aggregates of the
Crotalaria-maize rotation declined steadily (Fig 7) suggesting that plant available
P at the surface of these aggregates was constantly being utilized. This
absorption further resulted in 2 to 3 fold increases in maize yields of the
Crotalaria rotation in comparison to maize monoculture (Fig 8). In contrast,
available P at the surface of aggregates in the continuous maize management
system appeared to accumulate. Nitrate accumulations in bulk soils of P deficient
sites in western Kenya has been reported where maize growth is severely limited
by P deficiency (Hatermink et al, 1996; Jama et al, 1997; Mekonen et al, 1997).
Phosphorus application on these sites would therefore be expected to increase
maize production. The accumulation of P and subsequent poor yield of maize
(Fig 8) may indicate the need for both N and P fertilization where non-N fixing
fallow crops, such as Tithonia, are integrated into the cropping system.

It has been recognized that plant available P is a functional concept rather

than a measureable quantity. Total plant available P includes all P taken up

50

during a specific cropping season or growth cycle period (Tiessen and Moir,
1993). Phosphorus fertilization appears to initially deposit P on aggregate
surfaces where it becomes accessible to plant roots exploring the soil matrix.
Linquist et al (1997) showed fertilizer P retention on aggregate surfaces and
used a mean reactive mass to characterize this region of retention in aggregates.
They linked declining productivity from residual crops to aggregate sequestration
of P where a time dependent diffusion process limits P availability. We observed
an equilibration in the EIR of available P in aggregates when P application was
sufficient to overcome plant extraction (Fig 9b). The reverse also occurred when
P application was insufficient to meet extraction needs (Fig 9a) resulting in higher
P concentrations in the interior regions than the exterior layers of soil aggregates

under Crotalaria-maize rotation.

CONCLUSIONS

Phosphorus fertilization of soils deposits much of the P on surfaces of soil
aggregates, which slowly diffuses into the interiors of aggregates. The mean-free
diffusion pathway and physical chemistry of each aggregate appears to control
this diffusion of available P to surrounding plant root systems. Additional
research should reveal how these spatial gradients at multiple locations within
soil aggregates will temporally influence plant available P and how consistent
these aggregate layered P gradients are with other P pools previously described

by Hedley et al., (1982).

51

Whether similar P gradients exist in smaller aggregate sizes remains to be
discovered. The high correlation between yield and EIR of soil aggregates (4-8
mm across), representing 23.7% of the total bulk soil, suggests that a cascading
effect of P availability may be occurring from interior regions of smaller
aggregates, followed by the interiors of larger aggregates. In the process, plant
available P is removed from surface layers of large aggregates. Soil tests that
crush bulk soil samples suggest these current measurements of labile P fractions
may be poor indicators of P availability to plant roots since these tests destroy
the original aggregate structure. The absence of detectable differences among
different management systems and P extracted from crushed whole aggregates,
in this study, further suggests the need for establishing a concentric layer-based
index of soil aggregates before more accurate identification of P availability to
plants can be achieved. Accumulations of plant available P at surfaces of larger
aggregates in Tithonia-maize and continuous maize rotations would not have
been apparent, using contemporary methods, as these gradients would have
been diluted during the extraction of labile soil P from macerated whole soil
samples. Separation of uniform concentric layers by meso SAE chambers
provides an integrated non-destructive technique that more accurately projects
plant available P-pools and gradients experienced by roots in situ. Therefore, a
ratio relating extractable P at the surfaces of aggregates to interior P can be
useful for determining requirements for P fertilization. Using this knowledge of P
processes, it is hypothesized that, P fertilization will be required when a decrease

in the EIR coincides with a decline in residual plant yield, where temporal

52

demands for other essential nutrients (N and K) have been satisfied. Further
research is needed such as rate response experiments, to evaluate the accuracy
of EIR values for predicting plant and fertilizer interactions for multiple soil types

for maximizing the efficiency of P fertilization in low input agricultural systems.
REFERENCES

Agbenin J O and J T Goladi, 1998. Dynamics of phosphorus fractions in a
savana Alfisol under continuous cultivation. Soil Use Management 14:59-
64.

Parkinson, JA, and SE. Allen. 1975. A wet oxidation procedure suitable
for the determination of nitrogen and mineral nutrients in biological
materials. Commun. Soil Sci. Plant Anal. 621-11. 1

Beck, MA. and PA. Sanchez. 1994. Soil phosphorous fraction dynamics
During 18 years of cultivation on a Typic Paleudult. Soil Science Society of
America Journal 58, 1424-1431.

Buresh R J., P C Smithson, and D T Hellums. 1997. Building Soil Phosphorus
capital in Africa. P. 111-149. In R J Buresh et al (ed) Replenishing Soil
Fertility in Africa. SSSA Spec. Publ. 51. SSSA, Madison, WI.

Cassman, K G, A 8 Whitney and R L Fox, 1981. Phosphorus requirements of
soybean and cowpea as affected by mode of N nutrition. Agron. J. 73:17-
22.

53

Cassman K G, P W Singleton and B A Linquist. 1993. Input/output analysis of
soybean response to phosphorus on an Ultisol. Field Crops Res. 34:23-
36.

Fox, R.L.‘, and E.J. Kamprath. 1970. Phosphate sorption isotherms for evaluating
the phosphate requirements of soils. Soil Sci. Soc. Am. Proc. 34:902-907.

Giller K E., G Cadisch, C Ehaliotis, E Adams W D Sakala, and P L Mafongoya..
Building Soil Nitrogen capital in Africa. pp. 151—192. In R J Buresh et al .
(ed) Replenishing Soil Fertility in Africa. SSSA Spec. Publ. 51. SSSA,
Madison, WI.

Gressel.N, McColl, J.G, Preston, C.M, Newman, RH. and Powers, R.F.(1996)
Linkages between phosphorous transformations and carbon
decomposition in a forest soil. Biogeochemistry 33. 97-123.

Guo F and R S Yost, 1998. Partitioning soil phosphorus into three discrete pools
of differing availability. Soil Sci. 163:822-833.

Hatennink A E, R J Buresh, B Jama and B H Janssen, 1996. Soil nitrate and
water dynamics in sesbania fallows, weed fallows and maize. Soil Sci.
Soc. Am. J. 60:568z574.

Hedley, M.J., Stewart, J.W.B. and B S Chauhan, 1982. Changes in inorganic
and organic soil phosphorous fractions induced by cultivation practices
and by laboratory incubations. Soil Science Society of America Journal 46:
970-976.

Horn R, W Stepniewski, T Wlodarczyk, G Walenzik and E F M Eckhardt. 1994

Denitrification rate and microbial distribution within homogeneous soil

aggregates. Int. Agrophys. 8:65-74.

Huffman S A, C V Cole and N A Scott 1996. Soil texture and residue addition
effects on soil phosphorus transformations. Soil. Sci. Soc. Am. J. 60:1095-
1 101.

Jama B A, R A Swinkels and R J Buresh. 1997. Agronomic and economic
evaluation of organic and Inorganic sources of phosphorus for maize.
Agron. J 892597-604.

Linquist, B.A., P.W. Singleton, K.G. Cassman, and K Keane. 1996. Residual
fertilizer phosphorous and long term fertilization strategies for an Ultisol.
Plant Soil 184 pp47-55.

Linquist B A, P W Singleton, R S Yost and K G Cassman, 1997. Aggregate size
effects on the sorption and release of phosphorus in an Ultisol. Soil. Sci.
Soc. Am. J. 61:160-166. I

Maroko J B, R J Buresh and P C Smithson, 1999. Soil phosphorus fractions in
unfertilized fallow-maize systems on two tropical soils. Soil Sci. Soc. Am.
J. 63:320-326.

Mekonen K, R J Buresh and B Jama, 1997. Root and inorganic nitrogen
distriburions in sesbania fallow and maize fields.Plant and Soil
188:3192327.

Palm C A, R J K Myers and S M Nandwa. 1997. Combined use of organic and
inorganic nutrient sources for soil fertility maintenance and replenishment.
P. 193—217. In R J Buresh et al (ed) Replenishing soil fertility in Africa

SSSA Spec. Publ. 51. SSSA, Madison, WI.

55

SAS Institute, 1999. SAS/STAT Users Guide Vol. 2 6th ed. SA lnst., Cary, NC.

Santos D., S L S Murphy, H Taubner, A J M Smucker and R Hor, 1997. Uniform
separation of Concentric Layers from Soil Aggregates. Soil Sci. Societ
Amer. J. 61: 720-724.

Sibbesen, E. 1977. A simple ion-exchange procedure for extracting plant
available elements from soil. Plant Soil 46:665-669.

Sibbesen, E 1978. An investigation of the anion exchange resin method for soil
phosphate extraction. Plant and Soil 50:305-321.

Smucker, A.J.M., C.J. Dell, J Kinyangi and JD. Jastrow. 1999. Quantitative
removal of concentric layers from soil aggregates by soil aggregate
erosion chambers. ASA-CSSA-SSSA, 91“ Annual Meeting, Salt Lake
City, Utah, Oct.

Stewart, J.W.B. and Tiessen, H. (1987) Dynamics of soil organic phosphorous.
Biogeochemistry 4: 41-60.

Tiessen, H., Stewart, J.W.B. & Cole, CV. 1984. Pathways of phosphorous
transformation in soils of differing pedogenesis. Soil Science Society of
America Journal 48, 853-858.

Tiessen, H.,l.H. Salcedo, and E.V.S.B.Sampaio. 1992. Nutrient and soil organic

matter dynamics under shifting cultivation in semi-arid northeastern
Brazil Agric. Ecosyst. Environ. 38:139-151.

Tiessen, H. and Moir, J.'O.(1983) (Characterisation of available P by sequential

extraction. In: Soil sampling and methods of analysis (Carter, M.R. (ed.)).

Lewis Publ. CRC Press, pp.75-86.

Wan Y and S A El Swaify, 1998. Sediment enrichment mechanisms of organic
carbon and phosphorus in a well aggregated Oxisol. Journal Environ.

Quality 27:132-136.

57

Table 2.1. Increasing resin:soil ratio and changes in extraction of plant available
phosphorus of 3 soils.

 

 

 

56113 5611 2 SOII 1
Resin:SoiI ratio 0.259 0.51 0.251 0.125; 0.253 0.1253_
——(ma kg")
5.9 5.2 9.6 13.1 24.6 25.6
5.9 7.3 10.5 15.6 27.4 26.2
5 nd nd 12.1 17.8 31.5 33.8
10 nd nd nd nd 29.0 32.8
5611 mean ext P (mg kg") 6.1 13.1 28.9
ANOVA
Soil weight ns *‘* **
Ratio * m m
Ratlo‘Soil weight * ** 1*

 

m, “ and * denotes significance at 0.001, 0.01 and 0.05 level tests according
to Fishers protected LSD”;

ns - not significant.
nd - not determined.

58

Abrasion
Chamber

Knurled
wall

 

 

 

Bottom View

Base

Fig 2.1 Diagramatic illustration of meso- Soil Aggregate Erosion (SAE)
chamber

59

88%

l
I

 

4— Interior —>

     

i
‘ . 1
1
i

 

77% -

Transitional
66%
55% -

—O—— Maize-Maize

a-a-a—a-adIa-I-a—a—Idl

 

 

 

Cumulative erosion as fraction of aggregate mass

 

 

44% --O - - Crotalaria-Maize
33% ~
Exterior
22%
11%
0% I, L 1 I l
O 100 200 300 400 500
Time (min)

Fig. 2.2 Time versus progressive removal of concentric layers
of aggregates, 4-8mm across (mean=6mm), from continuous
maize and Crotalaria rotation plots of a Kenyan Oxisol, 1999.
Each regionllayer is one third of the original mass of the
aggregate. n= 12.

60

 

15

.. l2.5g soil, 1g resin
0
E, 12 ’ EIO.1259 soil, 0.6259 resin
IL
0
3 9
(G
‘6
E
if 6
C
'17:
0
a:
3
0

 

 

 

 

 

 

<2mm crushed <0.053mm sieved
Soil sample treatment

Fig. 2.3. Effect of sand removal on resin extractable phosphorus of a
Kenyan Oxisol. The two extraction regimes are representative of the
conventional resin P method (2.59 soil, 1g resin) and a more recent
method (0.1259 soil, 0.6259 resin) developed to extract P from layers
of soil aggregates. Bars are standard error. n=24.

61

Resin extractable P (mg kg")

36

 

 

 

 

 

. 8 8
U ('3
27 - i O
24 b . 056111,o.1259
21 - O
.SOII1,0.259
18 - g
ASCII 2, 0.1259
15 -
12 .. ‘ ASCII2,0.259
9 '- 135611 3, 0.259
6 .-
ISOII3,0.59
3 ..
o I 1 + l l
o 2 4 6 8 1o 12

Resin : soil ratio

Fig. 2.4. Influence of resin load and soil weight on resin
extractable phosphorus. The resin is chloride base converted
to bicarbonate form. The soil is shaken for 16 hrs at 175
reciprocations min-1.

62

Resln extractable P mg kg"

 

18

 

 

 

 

 

(a) Maize Tree crop Maize
(b) Maize Maize Maize
15 -
12 - ----------
O
........ . - _
§ ' \ \ O , ’
9 " \ x I I
/
ez’ ‘\ I
\ l
i .
6 - - O— ~CrotoIaria-Malze
- - I - - Maize-Make
3 _ —O—'I1thonla-Malze
o l J l l l

Jan-08 Apr-06 Jul-06 Nov-98 Feb-99 May-99 Aug-99

Fig. 2.5. Seasonal flux of resin extractable phosphorus in
whole aggregates of 3 management systems without P
fertilization on a Kenyan Oxisol. Crop sequence represents (a)
Tree-crop rotation and (b) continuous maize plots. Error bars
are standard errors for the difference in means.

63

20

 

 

 

 

 

16 t o >8mm
:3 . O 1-2mm
§ 12 -
'5 O O
t a - r’=o.47
G
s
2‘.
a 4 " \
2.
a
3 o-
3; O
X
0 -4 -
.s o .
0
2 .3 - O C
.5
81
-12 l.
g 3:010
o
.15 ..
-20 J J l l I

-0.5 -0.3 -0.1 0.1 0.3 0.5 0.7
Change in % C of aggregate size fractions

Fig. 2.6. Influence of post fallow C concentrations of
aggregates on subsequent pre-season changes in resin
extractable phosphorus. Changes of resin P in continuous
maize plots (control) Is subtracted from fallow rotation plots.
":4

 

Ratio of resin extractable P (EIR)

2.5

2.0

1.5

1.0 '

 

 

 

 

(a) Maize Tree crop Maize
(b) Maize Maize Maize

 

 

 

    

—0-— Crotalaria-Maize

+ Maize-Maize

--O--TIthonIa-Maize

 

 

I I I

Jan-98 Apr-98 Jul-96 Nov-96 Feb-99 May-99 Aug-99

Fig. 2.7. Cropping system effect on seasonal location of plant
available phosphorus in concentric layers of soil aggregates
(4-8mm across), of a Kenyan Oxisol, 0-15cm depth. Crop
sequence represents (a) Tree-crop rotation and (b) continuous
maize plots. Error bars are standard errors for the difference
in means. n= 4.

65

Maize grain yleld (T ha")

8.0 5

 

 

   

 

 

 

I
I
7.0 - .
I O
Continuous Maize I
6.0 - Mania-Maize rotation I .
I
5.0 r- E
I
O
I
4.0 - '
3.0 P
2.0 - "
I
l
1.0 - . I Crotalaria-Maize
I
. I
0.0 - .
l
I
.100 l l l i l
2.0 1.5 1.0 0.6 0.0 -0.5 -1.0

Flg. 2.8. Relationship between change In the ratio (Jul 99-Sep 98) of
exterior to Interior resin extractable phosphorus and July 1999 grain
yield. 11 = 4.

66

2.5

 

 

I Sep-98
Low P fertilization
2.0 ~ EIJuI-99

 

 

 

(a)

 

Ratio of resin extractable P (EIR)

 

 

 

 

 

2.5

 

 

High P fertilization I Step-98
2_o DJuI-99

 

 

 

(b)

 

Ratio of resin extractable P (EIR)

 

 

 

 

 

 

 

Crotolaria Maize Tithonia

Maize Maize Maize

Fig. 2.9. Fertilizer management and gradients in P location in
aggregates, 4-8mm across (0-15cm) of a Kenyan Oxisol, 1999. n=4.

67

Chapter 3

ESTIMATING RECENT CROP CARBON INPUT INTO CONCENTRIC LAYERS
OF SOIL AGGREGATES FROM MAIZE AND TREE BASED SYSTEMS

INTRODUCTION

Decline in soil organic carbon (SOC) appears to occur when farming
practices do not return crop residues to the soil and increase soil disturbance,
thereby enhancing carbon (C) mineralization and 002 losses to the atmosphere
(Lal, 1989; Bajracharya et al, 1996). The rate and quality of carbon transfer
below ground, together with microbial activity are recognized as important
processes governing SOC dynamics (Jastrow and Miller, 1997). A frequent C
loss during conversion of land to agriculture, within small holder farms in Africa,
has recently been characterized as a three-step process (Woomer et al, 1997).
First, nutrients and C stocks held in plant biomass of natural vegetation are
mobilized during land conversion. Second, the soil resource is exploited for
several years of productive cropping, during which nutrient rich root residues
decompose and organic matter (OM) is mineralized. At this point, reduced plant
yields will often be associated with a declining level of soil organic matter (SOM).
Third, a lower level equilibrium of SOM is established, resulting from
mineralization of the more labile SOM fractions. Continued productivity then
becomes dependent on external inputs, largely consisting of organic materials

and/or inorganic fertilizer.

68

Current research approaches for studying C mineralization have been
limited to estimating long-term C balances based on whole soil tests and above
or below ground productivity of crop biomass. Woomer et al, (1997) summarized
data (2-30yrs in 0-40cm soil depth) on C loss due to cultivation in Eastern and
Southern Africa. The authors reported system losses ranging between 0.9 T of C
ha" yr" to 10 T of C ha" yr". These losses are similar to those observed over 6
years (10 T of C ha") by Tiessen et al., (1992) during thorn bush conversion to
agriculture in Brazil. Current short-term estimations of soil C dynamics have been
limited to soil emissions of C02. However, it has been difficult to partition and
interpret some of the results when the presence of growing plants contributes to
rhizosphere respiration (Rochette et al., 1992; Reicosky and Lindstrom, 1993;
Rochette et al., 1999). Some of these difficulties may be resolved by integrating
different methods for studying C such as root exclusion by fine mesh screens,
isotopic labeling and more recently minirhizotron evaluation of root activities
(Aiken and Smucker, 1996). Smucker et al, (1997) identified C gradients within
soil aggregates and these C gradients changed rapidly after 20 months of plant
root growth in the field and as short as a few weeks in the greenhouse,
suggesting that total C analyses of bulk soil samples has concealed very
important short term changes of soil C.

The application of a stable carbon isotope as an intrinsic marker for tracer
studies has recently provided new and exciting opportunities for soil C studies.
During photosynthesis, C4 plants discriminate less against 13002 than C3 plants.

The result of this isotopic discrimination, is a characteristic carbon isotope ratio in

69 ‘

the plant tissue, which serves as an effective marker for the occurrence of C3 and
C4 photosynthesis (Dzurec et al, 1985). The carbon isotopic composition of SOM
is comparable to that of the source plant material prior to humification (Schwartz
et al, 1986). The natural abundance of ‘30 has been used to evaluate vegetation
shifts, (Desjardins et al., 1996; Boutton et al., 1998), community composition,
(Mariotti et al., 1994; Boutton et al, 1999), plant and soil respiration,(Hogberg et
al., 1996; Nyberg et al., 1999; Rochette et al., 1999), SOC (Arrouays et al., 1995)
and SOM turnover, (Balesdent et al., 1993; Bemoux et al., 1998) and net primary
production (Jenkinson et al., 1999).

The location of carbon substrates within the aggregate structure appears
to be affected by source-sink relationships. Priesack and Priesack (1992)
observed that substrate supply at the aggregate surface induced higher microbial
activity and concluded that outer surfaces of aggregates are probably the regions
of highest microbial activity and biomass. Jones and Farrar, (1999) provided
estimates for the rate of C loss by maize (86,400 ug C 9" root dry weight d").
They showed that lactic acid and sugars account for the largest proportion of this
loss. Greater carbon or nitrogen deposition at aggregate surfaces appears to
support increased microbial activity in surface regions of aggregates. Augmented
microbial activities associated with the mineralization of C should produce more
essential nutrients for plant roots exploring aggregate surface layers. (Kinyangi et
al., Chapter 3) identified concentric layering and sequestration of soil phosphorus
(P) within soil aggregates and showed maize crop removal of plant available P

from the exterior surface layers of soil aggregates (4-8mm across) when applied

70

P was insufficient to meet plant demand. Ladd et al (1994) reported net losses in
SOM of an Australian Alfisol when P was added, during nine years of continuous

cultivation.

Objectives of this study were:

1) To determine Crotalaria grahamiana (C3) and maize (Zea mays L.) (C4 )
carbon contributions to external surfaces and interior regions of soil
aggregates.

2) To determine the effect of crop rotation and seasonal variations on recent C
deposits.

3) To assess the effect of P fertilization on shifts of C inputs into aggregate

layers in the presence or absence of a maize crop.

MATERIALS AND METHODS

Two experiments were setup, both in the field and in pots. Soil samples
were collected from 0 to 15cm depths of an ongoing field trial initiated in March
1997 to investigate P replenishment with tree crops. The study site is located at
Osita farm in Vihiga District, Western Kenya (0° 06’ N, 34°34: E) at an elevation
of 1420m. Mean annual rainfall is 1800mm (split between 2 cropping seasons,
February-August and September to January). Often, the first cropping season will
receive 1000-1200mm of the annual rainfall. The soil type is a very fine

lsohyperthermic Kandiudalfic Eutrudox (USDA, 1996). It has few chemical or

71

 

physical barriers to rooting in the top 4m. Air dried soil for the top 15cm had a pH
of 5.1 (122.5 soil:water suspension), organic carbon 179 C kg", ex. acidity 0.4
cmolc kg", calcium 3.4 cmolc kg", (KCL extractable), Ex magnesium 1.4 cmolc
kg", Ex potassium 0.12 cmolc kg", phosphorus 2 mg kg", (bicarbonate-EDTA

extractable). Clay, silt and sand contents were 45%, 28% and 45%, respectively.

Experimental design

The field experiment consisted of two cropping sequences of a maize
monoculture and maize crop in rotation with a tree crop (Crotalaria ). The tree
crop was relay planted into maize during March 1997 and April 1998. Maize (HB-
512, Kenya Seed Co, Kitale, Kenya), was planted at 2 seeds per hill, at a spacing
of 0.75m between rows and 0.25m between plants, and later thinned to 1 plant
per hill to achieve a plant population of 53,000 plants ha". One row of tree crop
was planted for every row of maize, at the same intra row spacing, resulting in
the same tree plant population. Maize crops were harvested for grain yields in
August 1998 and July 1999 in the tree-crop rotation plots. Two additional maize
crops were harvested in January 1998 and 1999 (continuous maize plots). All
maize crops were assessed for grain yields. The tree crops were harvested in
February 1998 and 1999. All fine residues and litter fall were weighed and
returned to the plots, while the woody components were exported from the plots
for fuel wood. All above ground maize crop biomass was removed since it was
common practice for farmers to harvest the corn stover for livestock feed. We

only considered the residue returned as it had a direct effect on the fertility status

72

of the soil. Leaf litter and pods were harvested and oven dried at 60 ° C for dry

matter yield determinations.

Soil sampling

Soil samples were taken from the field trials on the following dates; 13
February 1998 (end tree crop season, before tillage), 7 July 1998 (post tree crop
season and maturity of residual maize), 19 September 1998 (initiation of 2““ tree
crop season), 13 February 1999 (end of tree crop season 2 and return to maize
cropping) and 1 July 1999 (post tree crop 2 and maturity of maize). Soil samples
were collected at depths of 0-15cm by gently driving cores (10cm diameter x
15cm height) into the soil surface to minimize aggregate disruption. Soil cores
were loaded onto trays, secured with rubber bands and transported to the
laboratory. Air—dried samples (1 kg) were sieved by manually shaking through a
nest of sieves, separating them into eight size fractions (>8mm, 4-8mm, 2-4mm,
1-2mm, 0.5-1 mm, 0.25—0.5mm, 0.053-0.25mm and 0.038-0.053mm). Total and
delta 13C signatures of C were measured in the 4-8mm aggregate fraction which
contained 40% of the total carbon and nitrogen, by mass of all soil aggregates
<8mm, during the tree crop season, (Kinyangi et al, unpublished data).

In addition, soil from an adjacent bush fallow site was sampled in July
1999 to provide background reference data which was compared to changes in C
inputs between C3 and C4 cropping sequences. Smestad et al., (unpublished
data) characterized this reference site in July 1998, estimating the age of the

bush fallow to be 9-12 years, following its conversion from food crop agriculture.

73

Trees (Sesbania sesban L. Merr.), accounted for 97% (18.3 t ha") of the
overstorey standing biomass and 94% of the understorey standing biomass (6.8 t
ha") was contributed by Lantana camara L. and Justicr’a flava. Leaf and litter fall
was 2.7t ha". In total, the above ground biomass sequestered 275kg N ha" and
16kg P ha". Previously, the site had undergone a mixed cropping pattern of
maize, maize-bean, cassava (Manihot esculenta Crantz) or sweet potato

(Ipomea batatus) rotations, dispersed among mixtures of Ca and C4 weed fallows.

Pot experiment

The pot experiment was conducted for 5 weeks at the Maseno Research
Station under a shelter housing of translucent roofing sheet to allow light
penetration, with open sides provided to permit free flow of ambient air. The
experiment consisted of a factorial combination of 2 crop factors (Maize, No
Maize) at 2 fertilization levels (NPK and No NPK). Soil was sampled from
Crotalaria plots in February 1998, after leaf and litterfall had been incorporated
into the soil. These plant residues provided a substantial background of C3
organic input. However, since there was a previous history of 0., input, the initial
1"C signature of this soil in the pots represented a mixture of 03 and 0.. sources.
Bulked soil samples were sieved through 8mm to remove large rocks and plant
debris, was air dried, weighed and NPK fertilizer was thoroughly mixed with one-
half the soil treatments. Plastic pots (7|) were each filled with 4kg soil on a dry
weight basis. An equal weight of gravel (<2cm) was added to the bottom of each

pot to facilitate free drainage.

74

Fertilizer was applied as Triple Super Phosphate (46% P205), Urea
(46%N) and Muriate of Potash (60% K20) at 300kg P ha", 200kg N ha" and
100kg K ha". Five seeds of maize (HB-513, Kenya Seed Co, Kitale, Kenya),
were planted on 26 June 1999, in each pot receiving maize treatment. The pots
were watered and brought to field capacity every 3 days for the entire duration of
the experiment (5 weeks). Maize was thinned to 3 plants per pot on 11 July 1999.
An initial soil sample was collected from the bulk soil to represent conditions at
the start of the experiment. Additionally, soil samples for C analyses were
collected from 0-5cm depth at 2 weeks (11 July 1999) and the whole pot soil,

during maize harvest at 5 weeks (2 August 1999).

Aggregate erosion

Soil aggregates (4—8mm across), collected from the field and pot
experiments, were separated into concentric layers of external, transitional and
internal groupings of equal mass as described by Santos et al., (1997) by
separately eroding 15 individual aggregates in stainless steel soil aggregate
erosion (SAE) chambers. Individual aggregates were placed into a stainless steel
chamber (2.5cm diameter), with knurled walls and secured into glass beakers
(100ml) using foam packing, before mounting them onto a variable speed rotary
shaker (lnnova 2300, New Brunswick, Canada). The top was sealed with a
rubber stopper secured with para film to prevent losses. The erosion chamber
was fitted with a 356um screen at the base. As the aggregate rolled around the

internal knurled surface, the fine (<356pm) fraction fell through the sieve and was

75

deposited in the lower base (Fig. 1) which could be removed and the fine eroded
portion removed. Samples were weighed periodically to determine the time of
removal for each third of the three concentric layer/region of the aggregate
fraction sampled from each treatment. Each layer was eroded to yield 0.159 soil
using the meso soil aggregate erosion (SAE) chambers according to the protocol
developed by Smucker et al, (1999). The final eroded samples obtained, :1
including the crushed interior region were passed through a 0.053mm sieve to I
exclude the sand fraction. Further details of this protocol were discussed by

Smucker et al, (unpublished data). We concluded that exclusion of the sand

 

portion would provide a more accurate assessment of humic C fractions of SOM,
which are associated with the finer silt and clay mineral fractions of the soil.
While studying SOM dynamics, Balesdent et al., (1986) reported most of the
SOC is retained by the silt and clay particle sizes of soil measuring 0-50jum for
silty clay and sandy loam soils. An equivalent number of whole aggregates were
crushed and sieved through 0.053mm to compare extracted parameters from
peeled layers and whole aggregates. These samples were analyzed for total C

and the ratio of 13C/‘QC (613C) of soil, in concentric layers and whole aggregates.

Carbon analyses

Samples were further prepared by weighing 20-40m9 into pressed tin
cups (8mm x 5mm). Total C content was determined by automatic dry
combustion on a Carlo Erba NA 1500 elemental analyzer (Carlo Erba, Milan,

Italy), gas chromatograph. The ratio of 13C/‘2C (6 13C) of soil in concentric layers

76

of aggregates (4-8mm) was determined on a Europa Model 2020 mass
spectrometer (Europa Scientific, Crewe, UK). A finely ground and precisely
weighed sample was introduced to the ratio mass spectrometer via an auto
sampler. Carbon is converted to C02 and carried to the triple collector, isotope
ratio mass spectrometer. Isotopic compositions are expressed using the delta

notation:

51307.. = [(130/‘20 same/”0‘20 5mm) — 11*1000 [Eq. 1]

The laboratory reference was calibrated against the international standard
(Vienna-Pee Dee Belemnite). Multiple analyses of the reference samples,
included after every 12 soil samples yielded a standard deviation of <0.05%o.
We calculated the change in the fraction of C in concentric layers of soil

aggregates resulting from conversion to C3 or C. using the equation:

°/o C derived from maize =

[(8130 layer final - 6 13C layer initial) / (813C Crop maize - 6 1SC layer Mao] [Eq. 2]

or Crotalaria =

[(5130 layer final - 513C layer initial) / (O13C Crop crotalarja - 5130 layer initial” [Eq. 3]

77

 

ln' _. O‘I-._~..|_,

 

Statistical analyses

Analysis of variance was performed using the Proc GLM option of the
Statistical Analysis System (SAS Institute, 1999) to investigate crop and
fertilization effects on total C and shift in 5 13C of whole aggregates and
concentric layers of soil aggregates. All significant tests refer to a = 0.05 unless

othentvise stated.

RESULTS AND DISCUSSION

 

5 1"C of plant material

The difference in 5 13C values of Crotalaria (pods, -25.5 it» :1: 0.1, n =3: leaves
-26.7%o :1: 0.1, n =3) and maize (roots, -10.6 %o :1: 0.1, n =5, shoots, -10.8%o :t 0.2,
n =4) was —15.4%o. The 5 13C of roots was used to calculate (Eq. 2 and 3) the
contribution of 0.. carbon since all above ground biomass was harvested and

removed from the maize plots.

Total carbon

Total carbon in the cultivated maize plots (19.600319 C kg-1, n =16) was
nearly 60% lower than the non-cultivated bush fallow (3310.049 C kg", n =9).
There was no significant effect of an initial application of 50kg P ha", fertilizer
phosphorus to planted trees and maize on bulk soil carbon. Further, there was no
significant difference for total carbon in Febmary 1998 among crop rotations at

the experiment site (mean = 19.81059 C kg", n =8). In the field experiment, we

78

investigated changes in total carbon of whole aggregates and concentric layers
following two rotations of tree-crops with maize under moderate P fertilization
(annual 50kg P ha", 1997-1999). A background analysis of whole aggregate C
showed a significant increase (24.89 c kg" to 26.69 c kg", p = 0.05) for both
maize and tree—crop rotation in plots that did not receive P fertilizer (February
1998 to July 1999). At this point, the exterior to interior ratio (EIR) of C deposited
in concentric layers was positively correlated (r2=0.95, p = 0.0002) with the first
season of residual maize yield in August 1998 (Fig 2). During this period, we also
observed a significant rapid decline in total C of interior regions (23.89 C kg" to
18.59 c kg", p<0.0001), and surface layers (25.49 c Kg" to 19.59 c kg",
p<0.0001) under P fertilization. In the pot experiment, significant differences
(p=0.03) in total C of interior regions of aggregates were detected after only 2
weeks (Fig. 3), among fertilizer levels of maize and no maize (bare soil)
treatments. However, there was a highly significant trend of carbon accumulation
at 2 weeks within the interior region (p<0.0001) and exterior surface layer

(p<0.0001) (Fig. 3).

Sources of crop derived carbon

The delta 13C signature of SOM in the bush fallow, considered, as the
reference was -21%o. This value is more negative when compared to the 5 ‘30
signature recorded in the maize mono-culture plots (-17.5%o) of the field
experiment site in February 1998. This suggests that a 9-12 year conversion of

the cultivated site to bush fallow has resulted in an accumulation of 63 derived

79

carbon of the SOM of the non-cultivated fallow. Nyberg et al, 2000 studied the
decomposition of Sesbania green manure and C3 sugar on a C4 plant dominated
Alfisol (soil 5 ‘30, ranged from -15.9%o to -16.7%o), in eastern Kenya. Sixteen
samples of Sesbania leaves yielded a mean 5 ‘30 of —27.4%o. Using this value
and reference from the continuos maize plots to infer, we can calculate (Eq. 2)
the C3 contribution to restoration of SOM at the non-cultivated bush fallow site "I
since Smestad et al., (unpublished data) determined that Sesbania was the
dominant tree at this site. During this period, C3 crop C contribution to SOM was

35.4% of total c. '1

 

Shifting C inputs between C3 and C4 sources revealed a pronounced
seasonal, flux of C4 carbon deposits to surface layers of soil aggregates (Fig 4).
However, this flux was not apparent when whole aggregate measurements were
taken, as is common in soil analysis methods (Fig. 5). This distinct shift was
observed for both crop rotations in the presence of the residual maize crop (July
1998)

Seasonal variation in 513C signatures of maize and tree crops was also
investigated. Since there were no significant differences in the 513C signatures
across the tree crop season, any temporary maize contributions of C to the soil
appear to have been respired. At the end of each crop season the, 5 ‘30
signature was observed to return to within variability of the initial value by
February 1998. Estimates of the sources and proportions of C inputs into surface
layers and interior regions of soil aggregates are presented in Table 1. In the pot

experiment, we confirmed maize root depositions of C (p<0.0001) to surface

80

layers of aggregates but not interior regions of soil aggregates 4-8 mm across
(Table 2). In addition, we recorded a more negative shift in the 5 13C signatures in
fertilized and non-fertilized pots that had no maize growth (Fig 6). Maize biomass
yield did not explain this difference (Table 3). This observation seemed unusual
since there was no maize input. However, Balesdent, et al, (1987) explained this
phenomena to be the result of a normal isotope effect as they reported respired
CO: is nearly void of 13C resulting in the enrichment of ‘30 in microbial products
relative to their substrate. This enrichment is thought to be in the order of 2%».
Rochette et al, (1999) separated soil respiration into soil and plant components
using 13C of the C02 evolved on an historically C3 soil converted to maize. They
maintained control plots with no maize growth and attributed the difference to the
effect of maize. Maize carbon contribution to respiration was observed 40 days
and peaked 65-105 days after planting. In this study, samples in the field trial
were collected 98 days after planting, to coincide with the period of maximum
nutrient absorption. It appears that 100 days may be close to the maximum
period for seasonal carbon deposition by maize, whereas 35-40 days (as we
observed in the pots) may be the minimum time. Based on the observations of
Rochette et al, (1999), we propose that the maize “effect” may not last beyond
100-105 days after planting. Sanchez et al, (2000) evaluated and found that,
maize roots enhanced N mineralization in the presence of a large short-term pool
of nitrogen. An assessment of C mineralization, showed that a one year maize
crop did not reduce potential mineralizable carbon, when the opposite was true

for a one year wheat crop and bare control soil

81

CONCLUSIONS

Resource poor farmers will often be interested to maximize crop
productivity when land is converted from fallow to crop agriculture. This study
reveals that integrating tree crops with maize can build up soil organic carbon,
when farmers cannot afford fertilizer inputs. The marginal increase in carbon
under continuous maize CI‘OpS is a function of poor maize yields, hardly
exceeding 1 T ha" in farmers’ fields. Fertilizer applications to maize with or
without tree crops resulted in significant decreases of SOC. Although fertilizer
inputs are required to increase crop productivity, the magnitude of C depletion
can dramatically be amplified when crops receive initial fertilization. Further
research is needed to explore the extent of C-Ioss and the magnitude of maize C
and its effect on functional pools of nitrogen and phosphorus. A mid-season
influx of a large pool of soluble C is hypothesized to provide the energy required
by microorganisms to decompose organic matter. Sanchez et al, (2000) and in
part this paper, present preliminary evidence of a) the existence of this carbon
pool and b) an indirect link between N availability in the presence of maize roots
and crop productivity. On conversion to maize cultivation, these data suggest that
we may not expect tree crop carbon inputs to persist beyond a single cropping
season when residual maize is planted after the tree crop season. Longer term
tree crops or repeated seasonal rotation with trees as was done in this
experiment may potentially mitigate this loss of C and instead, enhance C

sequestration.

82

REFERENCES

Aiken RM. and A J M Smucker, 1996. Root system regulation of whole plant
growth. Ann.Rev. Phytopathol. 34:325-346.

Arrouays D Balesdent J, Mariotti A, Girardin C, 1995. Modeling Organic Carbon
turnover in cleared temperate forest soils converted to maize cropping by
using C-13 Natural-abundance measurement. Plant and soil 173: (2) 191.

Bajracharya, R M, R Lal and J M Kimble, 1996. Soil organic carbon distribution in
aggregates and primary particle fractions as influenced by erosion phases
and landscape position: In: Proceedings of carbon sequestration in soils,
An international symposium. July 22-26, Columbus, Ohio.

Balesdent J and Mariotti A, 1987. Natural 13C abundance as a tracer fro studies
of soil organic matter dynamics. Soil Biol. Biochem. 19, 25-30.

Balesdent J, B Girardin C, Mariotti A, 1993. Site -related delta-c-of tree leaves
and soil organic- matter in a temperate forest. Ecology 74: (6) 1713-1721.

Bemoux M, C C Cerri, C Neill , J F L De Moraes , 1998. The use of stable carbon
isotopes for estimating soil organic matter turnover rates. Geoderrna 82:
(1 -3) 43-58.

Boutton T W, S R Archer, J A Midwood, S F Zilzer, R Bol, 1998. Delta C- 13
values of soil organic carbon and their use in documenting vegetation

change in a subtropical savanna ecosytem. Geodema 82: (1-3) 5-41.

83

Boutton ‘IW, S R Archer, J A Midwood , 1999. Stable isotopes in ecosystem
science: Structure, Function and dynamics of a subtropical savanna.
Rapid communications in mass spectrometry 13: (13) 1277.

Desjardins T, Cameiro A, Mariotti A Chauveal A, Girardin C, 1996. Changes of
the forest-savanna boundary in brazilian arizonia during the holocene
revealed by stable isotopes ratios of soil organic carbon Oecologia 108(4)
749-756.

Dzurec R S, T W Boutton, M M Caldwell, and B N Smith. 1985. Carbon Isotope
ratios of soil organic matter and their use in assesing community
composition changes in Curlew Valley, Utah. Oecologia 66: 17-24.

Hogberg P, and A Ekblad, 1996. Substrate-induced respiration measured in situ
in a C-3-plant ecosystem using additions of C-4-sucrose.Soil Biology &
Biochemistry.28 (9) 1 13138.

Jastrow J D and R M Miller, 1997. Soil aggregate stabilization and sequestration:
Feedbacks through organomineral associations In; Soil Processes and the
carbon cycle. Ed Lal R, J M Kimble, R F Follet and B A Stewart 15: 207-
223.

Jenkinson DS, Meredith J, Kinyamario J l, Warren GP, Wong MTF, Harkness
DD, Bol R, Coleman K, 1999. Estimating net primary production from
measurements made on soil organic-matter. Ecology 80: (8) 2762-2773.

Jones D L and J Farrar, 1999. Phosphorus mobilization by root exudates in the

rhizosphere: fact or fiction? Agroforestry Forum. Vol 9 No. 4.

Ladd, J N M Amato Z Li-Kai and J E Schultz. 1994. Differential effects of rotation,
plant residue and nitrogen fertilizer on microbial biomass and organic
matter in an Australian Alfisol. Soil Biol. Biochem. 26:821-831.

Lal R 1989. Conservation tillage for sustainable agriculture: Tropics vs temperate
environments. Adv. Agron. 42: 1073-1 082.

Nyberg G A Ekblad, R J Buresh, P Hogberg, 2000. Respiration from C-3 Plant
green manure added to a C-4 plant carbon domination soil Plant and soil.
218(1-2) 83—89.

Marotti A, and P Schmitte, 1994. Forest savanna ecotone dynamics in India as
revealed by Carbon Isotope ratios of soil organic matter. Oecologia
97: (4) 475-480.

Priesack, E., and GM. Kisser-Priesack. 1993. Modelling diffusion and microbial
uptake of C-glucose in soil aggregates. Geoderrna 562561 -573.

Reicosky D C., and M J Lindstrom, 1993. Fall tillage method:Effect on short CO2
fluxes from soil. Agron. Journal 8521237-1243.

Rochette, P, R. L Desjardins, E G Gregorich, E Pattey and R Lessard. 1992. Soil
respiration in Barley (Hordeum vulgare L.) and fallow fields. Can. J. Soil
Sci. 722591 -603.

Rochette P, L B Flanagan and E G Gregorich, 1999. Separating soil respiration
into plant and soil components using analyses of natural Abundance of .
carbon-13. Soil Sci. America Journal 63:1207-1213.

SAS Institute, 1999. SAS/STAT Users Guide Vol. 2 6'“ ed. SAS Inst, Cary, NC.

85

Sanchez J E, E A Paul E A, T C \Mlson J P Smeenk and R R Hanlvood.
Cropping Effects on Nitrogen Mineralization, Agronomy Journal
(submitted).

Santos D., S L S Murphy, H Taubner, A J M Smucker and R Hor, 1997. Uniform
separation of Concentric Layers from Soil Aggregates. Soil Sci. Society
Amer. J. 61: 720-724.

Smucker A J M., D Santos and Y Kavdir, 1997. Concentric layering of carbon,
nitrogen and clay minerals within soil aggregates from tilled and non-tilled
agroecosystems and natural grasslands. 3ml Eastern Canadian Soil
Structure Workshop Ottawa, Canada.

Smucker A.J.M, C.J. Dell, J Kinyangi and JD. Jastrow. 1999. Quantitative
removal of concentric layers from soil aggregates by soil aggregate
erosion chambers. ASA-CSSA-SSSA, 91St Annual Meeting, Salt Lake
City, Utah, Oct.1999.

Schwartz 0, A. Mariotti, R. Lanfranchi and e. Guillet, 1986 ‘3 C/ 12c Ratios of soil
organic matter as indicators of vegetation changes in the Congo.
Geoderma, 39: 97-103.

Tiessen H l.H. Salcedo, and E.V.S.B.Sampaio, 1992. Nutrient and soil organic
matter dynamics under shifting cultivation in semi-arid northeastern Brazil
Agric. Ecosyst. Environ. 382139-151 .

Woomer P L., C A Palm J N Qureshi and J Kotto-Same. 1997. Carbon
Sequestration and Organic Resource Management in African Smallholder

Agriculture. In; Soil Processes and the carbon cycle. Ed Lal R, J M

86

Kimble, R F Follet and B A Stewart 12: 153-172.

87

Table 3.1. Maize (C4) and Crotalaria (C3) contributions onto surface
layers and Interior regions of soil aggregates (4-8mm across), during the

field study, (0-16crn depth).

 

 

 

 

 

 

 

Crop rotation Maize-Maize Crotalaria -Malze
+P -P +P -P
£- (16)

C-Source (July-98)

C3-C Exterior nd _ - - -
Interior nd - 8.1 16.2

C4-C Exterior nd 56.4 61.8 49.8
Interior nd 4.2 - -

CSource (September-98)

C4-C Exterior nd - nd -
Interior nd 8.8 nd 12.6

C-Source (February-99)

C4-C Exterior nd - nd -
Interior nd 13.0 nd 19.7

 

nd, not detemllned

-, not calculated when 13C shift was In the opposite direction relative to crop Input.
Initial dates for C-source determination were, February, July and September 1998.

88

Table 3.2. Maize (Cz-carbon) and Crotalaria (Cg-carbon) contributions onto

surface layers and interior regions of soil aggregates (4-8rnrn across),
during a 5-weekpot study, 1999.

 

 

Crop Maize No Maize
+IlIPK -NPK «l-NPK -NPK
——C (96)

2wks (C3-source) Exterior 14 22 11 20
1Interior - 12 11 10

1: t

5wks (C4-source) Exterior 23 49 - -
Interior - 17 - -

 

 

‘3, ’, Calculations of carbon contributions for these layers are excluded as pets
without plant growth did not receive maize carbon Input.
n = 5 (maize), n = 3 (no maize)

89

Table 3.3. Effect of fertilizer application on above and below ground
biomass of maize grown In 7 literpots dum a 5 weekstudy, 1999.

 

 

 

 

Shoots Roots Root:SI‘loot
+NPK -NPK +NPK -NPK +NPK ~NPK
—---0 pot"---—-—
22.5 3.5 9.5 3.0 0.4 0.9
(2.2) (0.3) (2.3) (0.8) (0.1) (0.2)

 

Figures In parenthesis are standard deviations of the means

Biomass ls expressed as dry matter pot"
n85, fertilized pots
n=3, non-fertilized pets

90

 

Abrasion

Chamber
9 ‘ * Knurled

wall

 

 

 

   

Bottom View

Fig 3.1 Diagramatic illustration of meso- Soil Aggregate Erosion (SAE)
chamber '

91

Maize grain yield (T ha")

5.0

 

0 e
40- 0:095
o
p = 0.0002 0
3.0 -

  
 

O Fertilized

 

 

2.0 - O Non-fertilized
C?
1.0 r .0 O
O
0.0 1 ‘ . ‘

 

0.86 0.96 1.06 1.16 1.26 1.36
Exterior to interior ratio (EIR) of total carbon

Fig. 3.2 Maize grain yield (Aug-98) as influenced by the EIR of
total carbon within layers of soil aggregates (4-8mm across) for a
Kenyan Oxisol. Aggregates were sampled in February 1998, from 0
15 cm depth of fertilized (annual 50 kg P ha" ) continuous maize
and Crotalaria tree plots of an agroforestry management system.
Regression excludes square data point. 11 = 4.

92

 

 

 

 

 

 

 

 

   

 

 

 

 

 

 

 

 

 

 

 

     

 

 

 

25
23 _ (a) Initial, 0 weeks DExterior
”'0 Ellnterior
8
0 21 -
o 1
- 9 _
.2
17 -
15 - 2 <
25
(b) 11 July-99, 2 m DExterlor
,. ”I Ellnterior
'2
0 21 -
0 {1f
- 19 ' ' 2
s
'— r ',
17 - .
15
25
(cl 12 August-99, 5weeks 05mm»
.. 23". Ellnterior
'2
u 21 t-
0
T 19 -
5
.—
17 .
15
+NPK -NPK +NPK -NPK
LIME. NOMAIZE

Fig. 3.3 Total carbon contributions to surface layers and interior regions
of soil aggregates (4-8mm across) for a 5 week study in 7 liter containers.
Bars are std. errors. 71 = 5 (maize), n = 3 (no maize)

93

513C (*0)

8136 (*0)

Maize crop Crotalaria crop
-22.5 - “ “
I .215 +exterlor
—0- interior
-20.0 -
—I— bush fallow
-17.5 -
-15.0 r
-12.5
-25.0
Maize crop Maize crop
-22.5 - ‘ “ -
I -21.5 + exterior
-0— Interior
-20.0 -
—I- bush fallow
-17.5 -
-16.7
-16.1
-15.0 -
(b)
_12.5 1 1 a L
Jan-98 Apr-98 Jul-98 Nov-98 Feb-99 May-99

Fig. 3.4 Effect of crop rotation on seasonal variation in maize (C.-
source, -10.6 (15.), n=5) and Crotalaria (Cs-source, -26.1 (1.), n-4)
carbon contributions to surface layers and interior regions of soil

 

 

 

 

 

 

 

 

 

 

 

 

 

 

aggregates (4-8mm across, 0-15cm soil depth), of a Kenyan

Oxisol. Reference soil was sampled form a bush fallow (9-12 years
since conversion from agriculture) at Osita Farm in July 1999.

Bars are std. enors for the difference in means. 71 I 4.

 

 

513C

813C

6130

 

 

 

 

 

 

 

 

 

.20 . (a) Maize (Cd-Maize (0,) lFeb-98
without P fertilization El Jul-98

-13 .

.15 .

.14 .

.12 4 —

-22

 

(c) Crotalaria (C,)-rnaize (C4) without lFeb-98
.20 . P fertil'mation CHM-98

LI.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

-22
(c) Crotalaria (Cd-maize (C‘) with P IFeb-98

,20 . fertilization (50 kg P ha") mum

-18 - ' ,

.16 .

.14 .

-12 . _
Interior Exterior Whole aggregate

Fig. 3.5 Crop rotation and fertilhetion modifications of carbon
deposition onto surface layers and interior regions of soil
aggregates (4-8mln across), for a Kenyan Oxisol. n = 4

 

 

l (a) Initial, 0 weeks I'mfio,
Dlnterlor

 

 

-18 . ,

813C
:3
or

 

 

   

 

 

 

 

.20 . (b)11JuIy-99,2weeks

813C
.'.
0

.14 .

 

 

   

 

 

 

 

 

 

-20 l (c) 12August-99, I'm
5weeks Ellnterior

 

513C

 

 

 

 

 

 

+NPK -NPK +NPK -NPK

 

 

MAIZE NOMAIZE

Fig. 3.6 Maize carbon contributions onto surface layers and interior regions
ofsoilaggregates(4-8macross)fora5weekstudyln7lltercontahiers.
Bars are std. errors. n = 5 (maize), n I 3 (no maize)

Chapter 4

SUMMARY AND CONCLUSIONS

Integration of tree fallow crops with food crop farming in low input cropping
systems provides a short term solution to managing carbon, nitrogen and
phosphorus stocks in soils within small holder farms in Africa. When the overall
goal is to sequester carbon in order to build the fertility status of the soils the use
of shrubby and woody perennials has been attractive to farmers due to the
abilities of the trees to fix large quantities of above and below ground biomass.
Tree-crop rotation with maize did not lead to short-term reductions in SOC.
Whereas the total biomass production for Crotalaria and Tithonia are
comparable, their contributions to soil fertility greatly contrasts. Crotalaria
returned significantly more biomass to the cropland (58% of the total) while
Tithonia returned significantly less biomass (11.3% of the total. Although the
tissue N P K concentrations for Tithonia are higher than Crotalaria, these
differences in biomass production resulted in lower nutrient inputs (kg ha")
recycled with Tithonia (36N, 4P and 36K) in contrast to Crotalaria (143N, 10P
and 67K). Overall, these nutrient inputs are low when one considers maize crop
requirements. It is estimated that a 2T ha" grain and 3T ha" stover maize crop
would remove approximately 66 kg N ha", 18 kg P ha" and 60kg K ha".

It can be concluded that, of the three systems studied, Crotalaria-maize
rotations are the most appropriate for fertility maintenance of nutrient-depleted

soils. However, Tithonia-maize rotations may still be important where fuel wood

97

 

is a critical requirement for households. T7thonia yielded twice the amount of
wood produced by Crotalaria. Supplemental inputs of inorganic P fertilizers
applied during tree-crop establishment greatly enhanced tree fixation of C.
Unfortunately, most of the biomass due to additional fertilizer is exported as
wood resulting in a net mining effect. This is further evidence that suggests C
losses were accelerated under modest fertilization. Greater N cycling by
Crotalaria-maize rotation suggests the need for higher P fertilization in order to
utilize these plant N inputs if yields were to be increased in this cropping system.
The lack of differences in plant available P among cropping systems for
the duration of the study raises doubts whether current analyses of bulk soils are
able to capture the effects of specific processes that appear to be mediated by
aggregate structure and function. The locations of plant available P from recent
applications of plant residues and fertilizer materials, and the resulting
relationships with grain yield suggest that these processes associated with
nutrient recycling are closely linked to the activities of growing plants. These
interactions appear to be rapid with high seasonal turnover. The low P status of
bulk soil and associated poor maize growth may erroneously suggest the need
for more P fertilization when accumulations of plant available P are occurring
within specific layers of soil aggregates. The large influx of maize carbon onto
surface layers of soil aggregates observed during maize growth in the third study
raises further questions on how plants acquire mineral nutrients. There is
speculation that these C compounds contributed by maize are primarily made up

of sugars and lactic acid (Jones et al, 1999) Perhaps this “maize effect” leads to

98

 

N immobilization in continuos cropping systems. In contrast, increased N inputs
from biological N fixation and residue return, further lead to adequate supplies of
available N in Crotalaria—maize rotations when the N is in excess of requirements
by microorganisms involved in C respiration. These micro-nitche and rapid
changes that seem to occur under tropical cropping conditions are concealed by
current analyses of bulk soils. Integrating concentric separations of soil

aggregate layers, using meso SAE chambers, into soil testing methods will E
greatly contribute to our understanding of short-term factors affecting I

management modifications of C, N and to maximize input utilization efficiencies

 

of P into low input sustainable agriculture systems of the tropics.

99

 

APPENDICES

APPENDIX A

 

 

Source: Albee P B, Stockrnan G C and Smucker A J M,
Advanced Photon Source at Argnne National Laboratory,
January, 2000.

Reconstruction of a microtomographic image of a soil
aggregate (1.1 mm across) from Kenya, rendered by
algorithms at the APS. Light areas represent surface
breaking and internal pores >63 microns, while black
areas are the solid portions of the aggregate and
surrounding air. The light tubular area within the middle
appears to be a root induced macropore (RIM). The
aggregate was sampled from an Oxisol soil in July 1998.

101

 

Hi I
llll

 

2