P t
u H!
.3 H 4.
i

{tn

I
v1,

.5 .. _ 3.. « ._. .._..
m If , he.
, w a
u. u “Imm- . . 3
A . . , Lfihflflhr x
r. , ‘ .5 .230

,. .x 9.1.; ‘ , _
N If, . x
E a; .

m
I a

.21; ,
— . .5» Ti
Win: in. w?- ,
. Erin". 2. m1»? . ,
.\ .nuuQJ-till . . ‘ . . ‘
r. : . ,
....a&.. ‘ J . ‘ . an ...
.53 . . . , \V. . 1%
I. fi‘nndvfiflmfi. 35.; flank-Hun. I I

o. s gt! 6 .. .tL
Ill-103313....“ mm

:0-.-

. 0.
5.. 1.. IS. sh“

NE... A a 18-.
.r...

.31 dbfiiflhfiu. 8.1:. .
.213. I. .tllivisir
EDI]! ADI 4’n.l¢.§
unitarubu an... p93" 0»..-
1.5...kfi... :2 1.9.,

.ltlvll
I15: Sit-IA
.ythlifiitl.ili.l.fitll I-III
Irv . .~ |||~$5 iniho I:
§Ii\'

 

 

 

 

xvi-11:31.? .39":
52133;! .1.
.Ibi x

y.
. 1 IEIDIKPIP
: ..:§I...Lin

. .II:v.....LC....
.00:le O‘I‘s...1v.
.an-A_0vv0.|,v.: I . .7..|v
..b.l.u...‘.. .2

 

. .. ,5 c v ‘ .51.». in;

‘Ili.0\.bu|‘l.| I. l ..‘ ‘

a.» ‘- 1dr»..- ...t.».}ctuvcu.. . t vbokrl?
V ‘ firth-unit

 

1 l "7
”717::

 

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

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

5/08 K:/Pro§lAcc&Pres/ClRC/Dat90ue.indd

TIME DOMAIN REFLECTOMETRY AND TURF IRRIGATION MODELING

Charles E. Kome

A DISSERTATION

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

DOCTOR OF PHILOSOPHY

Department of Crop and Soil Sciences

1996

ABSTRACT

TIME DOMAIN REFLECTOMETRY AND TURF IRRIGATION MODELING.
By

Charles E. Kome

Hydrologic models are widely used in crop water management. However, water
balance simulations in turf ecosystems are very rare. A three-year study (1992-1994) was
conducted at the Hancock Turf Research Center to evaluate the use of Time domain
reflectometry (TDR) based turf irrigation scheduling compared to two modified
evapotranspiration (ET) models: i) Soil Conservation Service SCHEDULER (SCS-
SCHEDULER) and ii) Systems Approach to Land Use Sustainability (SALUS). Both models
required weather, soil, and turf-specific input data files with ET and other hydrologic
components as output. Volumetric moisture content (VMC) was monitored using
horizontally installed TDR probes at 2.5, 7.5, 12.5 and 20 cm depths on established fairway
turfs. Im'gation plots (11 x 11 m) were split and randomly seeded to either annual bluegrass
(Poa annua L. var. reptans) or Penncross creeping bentgrass (Agrostis palustris Huds.)
grown on an Owosso sandy loam soil (fine Typic mesic Hapludalt). The irrigation treatments
were: apply 2.5 cm of water only upon the appearance of stress (STR), apply 2.54 mm daily
(DLY), and maintain the soil at field capacity (F C) using TDR. Simulations began on 1 May
and ended on 30 September each year. Volumetric moisture content data measured by TDR

for up to 60 days each summer were used to validate the SALUS model. SCHEDULER was

used to evaluate available water content and excess water under these three irrigation
regimes. Annual bluegrass and creeping bentgrass biomass and inter-species competition,
and quality ratings were also measured. Irrigation scheduling based on TDR was more
conservative than from weather-based estimates. Because 1992, 1993 and 1994 were wet
years, the stress treatment received no irrigation but quality ratings were also lower for the
stress treatment. Bentgrass biomass and quality ratings were significantly higher than for
annual bluegrass. Differences among irrigation treatments were significant mostly during dry
periods. Excess water as predicted by the SCS-SCHEDULER model and drainage as
predicted by the SALUS model ranked, DLY > F C > STR for all years. Average VMC for
the rooting depth exceeded the 0.28 cm3 cm'3 field capacity level for more than 50% of the
time for the irrigated treatments resulting in overwatering for the irrigated treatments.

Volumetric moisture content by depth and irrigation treatment as measured by TDR for most

sampling dates were within the range of VMC as predicted by the SALUS model.

ACKNOWLEDGMENTS

I wish to express my sincere gratitude to Dr. Paul Rieke for his guidance, patience and
direction through the course of my graduate program. I like to thank all members of my
guidance committee, Drs. V. F. Bralts, B.G. Ellis, Joe Vargas, and Francis Pierce for their
intellectual contribution throughout my graduate program at Michigan State University.

I want to thank Dr. V. F. Bralts and Dr. W. Shayya for providing the SCS-
SCHEDULER and Dr. J. T. Ritchie for providing the SALUS model and for many hours of
computer programing that made it possible to complete this project.

A special note of thanks goes to the O. J. Noer Foundation and the Michigan Turf
Foundation, the American Geological Institute, the Crop and Soil Science Department, the A
College of Agriculture, and the Graduate School for their financial and moral support.

I would like to thank all my fiiends and fellow students who made adult education
bearable. Finally, I owe a depth of gratitude to my parents and family for their moral and

financial support through these challenging years.

iv

Table of Contents

Page
List of Tables ........................................................ vii
List of Figures ....................................................... ix
CHAPTER 1. Introduction .............................................. 1
CHAPTER 2. Literature Review .......................................... 4
Literature Cited .......................................... 27
CHAPTER 3. Moisture Depletion on Fairway Turfs Under Three Irrigation Regimes
Using Time Domain Reflectometry ...................... 36
Abstract ................................................ 36
Introduction ............................................. 37
Materials and Methods ..................................... 3 8
Results and Discussion ..................................... 44
Conclusions ............................................. 57
Literature Cited .......................................... 59
CHAPTER 4. Annual Bluegrass and Creeping Bentgrass Competition Under Variable
Irrigation Treatments ................................ 61
Abstract ................................................ 61
Introduction ............................................. 62
Materials and Methods ..................................... 64
Results and Discussion ..................................... 67
Conclusions ............................................. 72
Literature Cited .......................................... 73
CHAPTER 5. Bihourly Moisture Depletion Patterns in Fairway Turfs ............. 75
Abstract ................................................ 75
Introduction ............................................. 76
Materials and Methods ..................................... 79
Results and Discussion ..................................... 81
Conclusions ............................................. 96
Literature Cited .......................................... 98
CHAPTER 6. Adaptation of a New Water Balance Model for Turfgrass Under
Variable Irrigation regimes ........................... 100
Abstract ............................................... 100
Introduction ............................................ 101
Materials and Methods .................................... 103
Results and Discussion ..................................... 108
Conclusions ............................................ 123
Literature Cited ......................................... 125

Table of Contents Cont’d.

P_ag§
CHAPTER 7. Evaluation of Turfgrass Irrigation Scheduling Using the Soil
Conservation Service SCHEDULER Model .............. 126
Abstract ............................................... 126
Introduction ............................................ 127
Materials and Methods .................................... 131
Results and Discussion .................................... 135
Conclusions ............................................ 147
Literature Cited ......................................... 150
APPENDICES
Solar Radiation for 1992, 1993 and 1994 ...................... 152
Minimum and Maximum Temperature for 1992, 1993 and 1994 ..... 153
Rainfall for 1992, 1993, and 1994 ............................ 154
Relative Humidity for 1992, 1993 and 1994 .................... 155
Wind Velocity for 1992, 1993, and 1994 ...................... 156

Table 3.3.

Table 4.1.

Table 4.2.

Table 4.3.

Table 4.4.
Table 4.5.

Table 5.1.

Table 5.2

Table 5.3.

Table 6.1.

Table 6.2.

Table 6.3.

Table 6.4.

List of Tables

Page
. Descriptive statistics for consecutive TDR measurements .............. 43
. Comparison of volumetric soil moisture content by the
gravimetric method (GMC) and by TDR for difi‘erent
depths. ................................................. 43
Regression equation coefficients for moisture depletion for
the stress treatment during a dry-down cycle in 1992. ............. 50
Root mass density for annual bluegrass and creeping bent-
grass by irrigation treatment and by depth 1993 and 1994. .......... 68
Root mass density for annual bluegrass and creeping bent-
grass by irrigation treatment, 1993 and 1994. .................... 69
Clipping weight means by irrigation treatment for 1993
and 1994. ............................................... 70
Quality ratings by irrigation treatment and species .................... 7O
Species composition by irrigation treatment ......................... 71
Analysis of variance for the test of hypotheses for between-
(among-) subject effects for bihourly VMC ..................... 84
Repeated-measures analysis of variance for the test of
hypotheses for VMC among subject efl‘ects ..................... 85
Comparison of soil moisture depletion (mm) by irrigation
treatment and daily ET from weather station at the
Hancock Turf Research Center, Michigan State University. ......... 93
Soil physical properties for SALUS model ........................ 104
Seasonal water infiltration by irrigation treatment and year
based on 152 day season as predicted by SALUS ................. 110
Seasonal drainage by irrigation treatment and year based on
152 day season as predicted by SALUS. ...................... 111

Seasonal soil evaporation by irrigation treatment and year
based on 152 day season as predicted by SALUS ................. 112

vii

List of Tables Cont’d.
Table Page

Table 6.5. Seasonal transpiration by irrigation treatment and year based on
152 day season as predicted by SALUS. ....................... 114

Table 7.1. Seasonal crop coefiicients, rooting depth, and management
allowable depletion model for SCHEDULER model ............. 134

Table 7.2. Farm (tum-specific input data for the SCS-SCHEDULER
simulation ............................................... 135

Table 7.3. Seasonal rainfall, minimum and maximum temperature and
means for 1992, 1993, and 1994. ............................ 136

Table 7.4. Seasonal cumulative potential ET (ETP), turf ET (ET), and
adjusted seasonal ET(ET,), by year and irrigation treatment
as predicted by the SCS-SCHEDULER. ....................... 138

Table 7.5. Seasonal excess rainfall and irrigation, by year and by irrigation
treatment as predicted by the SCS-SCHEDULER model. ......... 145

Table 7.6. Frequency of excess rainfall and irrigation by irrigation treatment
as predicted by the SCS-SCHEDULER model ................... 146

viii

Fig.

Fig.

Fig.

Fig.

Fig.

Fig.

Fig.

Fig.

Fig.

Fig.

3.4.

3.5.

3.6.

3.7.

3.8.

5.1.

5.2.

5.3.

5.4.

5.5.

List of Figures

ix

Page
Spatial arrangement of TDR probes per plot for the different
depths ................................................. 40
Volumetric moisture content (VMC) means for the different
irrigation treatments by depth, 1992 ............................ 45
. Volumetric moisture content (VMC) for the different irrigation
treatments by depth, 1993 ................................... 47
Volumetric moisture content (VMC) means for the different
irrigation treatments by depth, 1994. .......................... 49
Moisture depletion patterns across species for the difi‘erent
irrigation treatments by depth, 1993 ............................ 5]
Moisture depletion means for the different depths by irrigation
treatment. ............................................... 53
Volumetric moisture content (VMC) means versus turf quality
means, 1992. ............................................ 55
Moisture depletion by depth for a dry down period in 1992. .............. 56
Bihourly minimum temperature (TMIN) maximum temperature
(TMAX) and solar radiation (SRAD) for13 August 1993,
16 August 1994 and 13 September 1994. ....................... 82
Bihourly relative humidity (RH), and wind velocity (WV) for
13 August 1993, 16 August 1994 and 13 September 1994 .......... 83
Bihourly moisture depletion patterns for annual bluegrass
Penncross creeping bentgrass by irrigation treatment
August 13 1993. ......................................... 87
Bihourly moisture depletion patterns for annual bluegrass
Penncross creeping bentgrass by irrigation treatment
August 16 1994. .......................................... 89
Bihourly moisture depletion patterns for annual bluegrass
Penncross creeping bentgrass by irrigation treatment
September 13 1994. ....................................... 91

List of Figures Cont’d.

Figure

Fig. 6.1 Simulated seasonal infiltration and drainage for the 2.5 mm daily
(DLY); 2.5 mm every other day (EOD); field capacity ( FC);

and stress treatments (STR) for 1992 .....................

Fig. 6.2. Cumulative infiltration (a) and cumulative drainage (b) for daily
(DLY), 2.5 mm every other day (EOD) field capacity (F C),
and stress (STR) treatments, and (c) cumulative infiltration

and drainage for the daily treatment, 1992 .................

Fig. 6.3. Daily soil evaporation (Es) and transpiration (Et), (a) Cumulative
soil evaporation (Cum Es) transpiration (Cum. Et) and
evpotranspiration (Cum Ep), (b) cumulative potential
evaporation for the daily (DLY), 2.5 mm every other
day (EOD), field capacity (FC) and stress (STR)

treatments (c), 1992 ..................................

Fig. 6.4. Measured and simulated volumetric moisture content by depth

for the 2.5 mm daily treatment, 1992 .....................

Fig. 6.5 Measured and simulated volumetric moisture content by depth

for the field capacity treatment, 1992 .....................

Fig. 6.6. Measured and simulated volumetric moisture content by depth

for the stress treatment, 1992 ...........................

Fig. 7.1 Available water content (AWC), irrigation and rainfall for the 2.5
mm daily, field capacity and stress treatments in relation
to the drained upper limit (DUL) and management allowable

depletion (MAD), 1992 ................................

Fig. 7.2. Available water content (AWC), irrigation and rainfall for the 2.5
mm daily, field capacity and stress treatments in relation
to the drained upper limit (DUL) and management allowable

depletion (MAD), 1993. ...............................

Fig. 7.3. Available water content (AWC), irrigation and rainfall for the 2.5
mm daily, field capacity and stress treatments in relation
to the drained upper limit (DUL) and management allowable

depletion (MAD), 1994. ...............................

l?
a)
(D

114

116

118

119

120

121

139

141

143

CHAPTER 1

Introduction

It is estimated that about 70% of the earth is covered by water, most of which is sea
water. In spite of its apparent abundance, water remains the most limiting factor for crop
production in the world. ' There has been a considerable decline in access to quality water due
to an increase in population and industrial growth. Recreational and institutional use also
contribute toward the contamination of surface and ground waters decreasing potential water
resources.

The natural response to water shortages is to explore new sources. Eventually, water
may have to be transported longer distances as local sources are depleted or contaminated.
Future generations will need to deal more efficiently with the high cost Of pumping,
purification, conveyance, and redistribution in the allocation of this important, but limited
natural resource.

The quest for maximum yield and high-quality crops has resulted in a significant
increase in the amount of land under irrigation. Jensen (1974) predicted a threefold increase
in irrigated acreage in the US. from 1939 to 1980. Increase in irrigation puts a strain on
water allocation and distribution, both locally and worldwide. The growing percentage of
potable water used for golf course and home lawn irrigation is a strong reason for

reevaluating urban irrigation practices and requirements (F eldhake et al., 1983). In California

2

and Texas, 95% and 55%, respectively, of water resources are used for irrigation (Beard,
1973). In Michigan alone, more than 300 million cubic meters of water are used for
conventional crop irrigation annually (GEM Report, 1982). Although this accounts for less
than 3% of Michigan's water resources, restrictions on water use in Detroit and many cities
across the US. are examples of the need to reevaluate water allocation efficiencies,
particularly in urban areas.

Evaluation of application uniformity is one component in detemrining irrigation
eficiency. Low application eficiencies reported in the literature for hose-move and sprinkler
systems used for irrigation of home lawns (Henggeler, 1993) suggest that turf irrigation
scheduling techniques must be improved to increase water use efficiencies. Ideally, such
improvement should integrate weather based ET as well as soil moisture depletion estimates.

Soil-based irrigation scheduling requires routine detemrination of soil moisture
content. Gravimetric soil moisture determination is time consuming, destructive, and labor
intensive, and requires a 24-hour waiting period. Widespread use Of the neutron probe in
research (Simpson and Meyer, 1987) and to a limited extent in conventional irrigation
scheduling, is not amenable to the shallow rooting depths encountered in fine turf
management (Snyder et al., 1984). Adaptation of time domain reflectometry (TDR) to
volumetric soil moisture determination in the early 19805, coupled with innovations in
computer and electronic technology, provided avenues for automated soil moisture data
retrieval (Baker and Allmaras, 1990) which can be used for turf irrigation scheduling and
modeling.

Although irrigation practices may not necessarily coincide with the recommendations,

modeling provides a means of evaluating such recommendations without the need for ad hoc

3

experimentation. TO do this, a working model that has been evaluated under various
conditions and at difi‘erent locations must be used. Such a model must be accurate with the
minimum input data requirement (weather data, management records, and soil- and plant-
specific data) possible, user fiiendly, and efficient with respect to computation time (Ritchie,
1991).

In spite of widespread use of simulation models in crop research, the application of
models in turfgrass research has been very limited. There is a need to employ hydrologic
models to evaluate potential economic and environmental impact of turf irrigation
recommendations in rural and urban communities. In this study, an irrigation scheduling and
evaluation model (SCS-SCHEDULER) (Shayya and Bralts, 1988) and the new water balance
routine of the Systems Approach to Land Use Sustainability model (SALUS) (Ritchie, 1994)
were adapted for turf irrigation management and evaluation. Adaptation and evaluation of
both models with respect to soil water dynamics under different turf irrigation scenarios was

the focus of this study.

CHAPTER 2

Literature Review

Introduction

America's passion for green and well-manicured turf is so compelling that most
homeowners and turf managers invest time, effort, and money for the upkeep of their turf.
One aspect of turf culture that has come under close environmental scrutiny is turf irrigation.
Irrigation improves the aesthetic and functional value of turfgrass (Beard, 1973; Beard and
Green, 1994). Unfortunately, irrigation practices employed by turf managers and home-
owners may result in overwatering due to lack of adequate methods for assessing turf
irrigation needs.

A significant proportion of urban water is used for turf irrigation. Tovey et al. (1969)
stated that water was the second largest expense in turf management, after labor cost. This
is particularly true when water is obtained from treated municipal sources. Fok and
Murabayashi, (1978) estimated that turf irrigation accounted for 33% to 70% of total water
use in some large cities in the US. In some states, domestic and commercial turf irrigation
consumes more water than any other single crop (Hengeller, 1993). Beard (1973) predicted
that the amount of water used in turf irrigation would double by the turn of the century.
According to these estimates, additional water resources will be needed to meet this growing

demand.

5

Water use regulations for turf irrigation are no longer restricted to arid regions.
Aronson et al. (1987a) concluded that turfgrass culture must be directed toward practices that
lower turf water requirement as the competition for water increases. Biran et al. (1981)
suggested that homeowners and turf managers may have to accept lower quality turf as a
means of reducing annual water consumption for turfgrass irrigation. Also, the use of effluent
water for turf irrigation is gaining acceptance across the US. (Feldhake et al., 1983).

Efficient water management strategies are necessary to sustain the availability of water
for turf irrigation (Watson, 1980). The challenge for turf managers and researchers is to
strike a delicate balance between water use for the maintenance Of quality turf and other
competing uses of water in light of scarce water resources and/or environmental concerns.

A considerable amount oftirne and effort has been given to the evaluation of turfgrass
water requirements (Marsh et al., 1980; Kneebone and Pepper, 1982; Johns et al., 1983; Fry
et al., 1989; Aronson et al., 1987). These studies range from evaluations of
evapotranspiration (ET) rates within and among species to studies on the effects of various
cultural practices on water-use rates (Beard, 1982; Beard, 1985). Danielson et al. (1981)
reported high correlations between ET and turf quality. That cool-season grasses have higher
ET rates than warm-season grasses has been amply documented (Marsh et al., 1980; Biran
et al., 1981; Kneebone and Pepper, 1982; Kim and Beard, 1988). Turfgrass ET in urban
environments is site-specific and can vary considerably fi'om regional climate predictions
(F eldhake et al., 1983). This calls for a concerted effort to search for site-specific turf
irrigation scheduling techniques.

Although Fry and Butler (1989) and Beard (1973) reported no significant difference

in ET rates between annual bluegrass and creeping bentgrass, Salaiz et al. (1991) observed

6

high variability in water use rates among creeping bentgrass cultivars. Shearrnan (1986,
1989) evaluated ET rates of cultivars of Kentucky bluegrass and perennial ryegrass. The
results showed interspecific as well as intraspecific differences in ET. These findings suggest
the potential for selecting and/or improving cultivars with lower ET rates for water
conservation purposes.

Several factors influence water use by turf. The amount of supplemental irrigation
needed by turf is a function of the area (A) to be irrigated, the evaporative demand (ET), the
efficiency of the irrigation system (EF), the degree of shading (S), the soil type (ST), the time
of day (T) when the water is applied, and the desired quality (Q) of the turfgrass (Aurasteh,

1983). He expressed the irrigation need (I) mathematically as:

1: f(A, ET EF, 5, ST, T, Q)

Water uptake by plants can be affected by the physical properties of the soil. Soil
compaction affects soil aeration (Carrow, 1980), bulk density (Boufford and Carrow, 1980),
and soil strength (Voorhees et al., 197 8). In a greenhouse experiment, Agnew and Carrow
(1985) studied the effect of long-tenn compaction and soil water status on Kentucky
bluegrass rooting response and water use. They found that compaction increased root mass
in the top 5 cm of the soil. Also, the interaction between compaction and water stress
significantly increased root porosity, with a corresponding increase in water uptake under low
soil oxygen conditions. Soil compaction may thus reduce water uptake, with a significant

decline in turf quality and carbohydrate reserves (Carrow, 1980).

7

Roots and Plant Water Uptake

Roots play a vital role in plant water and nutrient uptake. Letey et al. (1966) cited
root permeability and surface area as two important parameters in the quantification of plant
water uptake. Although root characteristics are controlled genetically, environmental
influences can afl‘ect root distribution within the soil profile (Koski, 1983).

Peacock and Dudeck (1984) studied the effects of irrigation intervals on St. Augustine
grass (Stenotaphrum secundatum L.) rooting. They reported 42% and 37% decreases in
root-length density for the first and second years of their study, respectively, with high-
frequency irrigation. They observed more than 90% of the roots within the top 30 cm. Doss
et al. (1962) observed a decrease in rooting depth of five warm-season forage species with
an increase in soil moisture content. Bennett and Doss (1960) explained an Observed increase
in rooting depth of cool-season forage species under moisture stress as a drought-avoidance
mechanism. Rooting patterns for most turfgrasses are likewise affected by the frequency and
amount of irrigation (Beard, 1973).

Root distribution within the soil profile is an important parameter in quantifying plant
water uptake. A good correlation between root-length density and water uptake makes this
measurement a valuable tool for modeling plant water uptake (Ritchie and Amato, 1990;
Hanks, 1992). Rooting depth is an indicator of the soil volume explored by the roots in terms
of resource capture (Barber, 1984; Amon, 1992). The root-shoot mass ratio provides an
estimate of the partitioning of assimilates in response to environmental conditions. Beard
(1973) reported seasonal variations in the partitioning of assimilates to the shoots and roots.
Reduced carbohydrate reserves during periods of water or temperature stress may alter

assimilate partitioning between roots and shoots (Valoras et al., 1966). Furthermore, water

8

and nutrient uptake, growth rate, and turf quality may be afi‘ected by extreme temperatures

or water stress (Schmidt & Snyder, 1984).

Water Use and Turf Quality

In turfgrass management, dry matter accumulation and yield are not meaningful
measures of response. Aesthetic appeal, turf quality, and stress tolerance of turfgrass are of
greater interest to homeowners and turf managers. Limited research has been conducted to
quantify, in detail, the relationship between soil water content and turf quality (F airbone,
1982). Feldhake et al., 1983 showed a linear relation between ET and turfgrass quality.
Whereas limited water availability correlates positively with lower turf quality, high moisture
content does not necessarily correlate well with high turf quality (Danielson et al., 1981).
High moisture content can result in water-logged soils with poor oxygen diffusion rates,
which may cause reducing conditions over extended periods with potential reduction in turf
quality. Despite the qualitative and subjective nature of turfgrass quality ratings, Aurasteh
(1983) suggested that correlations between turf quality ratings and plant-available water could

be used to improve turf irrigation scheduling.

Field Capacity and Irrigation Scheduling
Field capacity is a controversial, but useful concept. It is defined as the soil moisture
content after gravitational water has drained (Hall and Heaven, 1970). Like the wilting point,
field capacity also depends on soil texture (Taylor and Ashcroft, 1972), with moderations
from temperature and previous wetting history (Cuenca, 1989). A soil attains field capacity

when drainage fiom that soil, which was previously at or near saturation, has decreased

9

markedly. This is generally assumed to occur 24 to 48 hours afier wetting, depending on soil
type (Cassel and Nielsen, 1986).

Although field capacity is used as the upper limit of available soil moisture in most
crop models, it is important to note that most soils do not drain to a fixed field capacity, nor
is this soil characteristic maintained indefinitely (Hanks, 1992). Also, the use of field capacity
as the upper limit may be a conservative estimate because some gravitational water may be
available for plant growth (Hanks and Hill, 1980; Ritchie and Amato, 1990). In irrigation
scheduling, total plant available water is defined as the difl‘erence between FC and the
permanent wilting point. Plant available water at various moisture contents can be determined
as a fraction of total available water. This has practical applications in the determination of

irrigation frequency (11,) based on historic average ET.

: Plant available water (mm)

I
f' Average ET (mm day")

 

Irrigation scheduling requires the determination of ET (Cuenca and Amegee, 1987).
Conventionally, this determination is based on one of many ET models using climatological
data Some irrigation schedulers often ignore soil moisture content evaluations. However,
such measurements provide valuable information on the soil hydrologic processes, such as
infiltration, leaching, and evapotranspiration, as well as irrigation scheduling.

Spatial and temporal variations of soil moisture storage require routine monitoring of
soil moisture in order to reduce the incidence of water stress to turfgrass. According to Amer
et al. (1994), progress in the characterization of soil moisture in space and time depends on

the development of cost-effective measurement tools.

10

Water storage in soils depends on the fraction of the total pore space occupied by
water or air and is expressed as a mass or a volume fraction or percentage of the dry soil.
Volumetric soil moisture content has practical application as rainfall and irrigation are
expressed in terms of equivalent depth of water (DJ per unit area. A unique advantage of this
measurement is that it provides a basis for quantifying equivalent depths for a given amount
of rainfall or irrigation (Hanks, 1992).

Until recently, repeated soil moisture determination has been impeded by the lack of
rapid, economical, safe, and accurate techniques for VMC determination. For cool-season
grasses grown on sand mixes other than US. Golf Association (U SGA) greens, low-volume,
high-frequency irrigation is necessary to compensate for low water retention capacities, in
order to maintain high-quality turf. Methods of soil moisture determination include
gravimetric (Hanks and Ashcroft, 1980), neutron attenuation (Belcher et al. 1950), electrical
resistance (Kirkham and Taylor, 1950; Gardner and Kirkham, 1952), tensiometer (Hagan and

Haise, 1967; Hillel, 1980), and TDR (Topp et al. 1980) techniques.

Gravimetric Soil Moisture Determination
The gravimetric method of determining soil moisture content is the standard against
which all other methods are compared. It measures the mass of water per unit mass of bulk
soil after drying to constant weight at 105°C (Black, 1965). This method requires the
collection of several samples and at least a 24-hour drying period. Although the method does
not require expensive equipment, it is labor intensive and destructive, and thus it fails to allow
for repetitive measurements on the same site. Subsequent determinations are thus subject to

spatial variability.

ll

Tensiometers

Tensiometers have been used successfully in irrigation scheduling, with considerable
savings in terms of water use. For example, Snyder et al. (1984) reported significant
reductions in turf water use with tensiometers. Automated irrigation systems have been
designed in which tensiometers are used to activate the irrigation system (Amon, 1992).
These savings, nonetheless, are accompanied by an escalation in costs. Another drawback
to using tensiometers is that they require frequent servicing and must be removed from the
field during the winter months in cold regions. Today, the use of tensiometers for irrigation
scheduling is limited to high-value crops.

Soil moisture potential reflects plant response to water stress better than does
volumetric water content (Hagan and Haise, 1967). However, the effective moisture range
covered by tensiometers is limited from saturation to about 80 kPa. The water potential is
converted to soil moisture content by using the soil moisture characteristic curve. Because
this curve is assumed to be unique for each soil type, several curves may be needed for a
single field. Although this range may cover about 90% of the available moisture in sandy soils
(Amon, 1992), it encompasses less than 50% of the moisture range for clay soils (Hillel,
1980; Amon, 1992). At tensions beyond 80 kPa, air penetrates into the ceramic cup and the
instrument becomes dysfunctional (Amon, 1992).

Tensiometers measure a small soil volume and are thus subject to problems associated
with spatial variability. Richards and Marsh (1961) proposed installing tensiometers at
different depths to monitor soil moisture depletion for the purpose of irrigation scheduling

and suggested using several tensiometers to account for spatial variability in soils.

12

Neutron Probe

The neutron probe consists of a source of fast neutrons, usually a mixture of
americium or radium with beryllium, that emits alpha particles, and a slow neutron detector.
Once the housing is removed, the neutrons collide with various atomic nuclei in the soil
(Belcher et al., 1950). The nuclei of elements with low atomic weight, especially hydrogen,
thermalize the alpha particles. Because water is the main source of hydrogen in soils, the
number ofthermalized neutrons, as recorded on the sealer, is correlated to the volumetric soil
moisture content.

Soil moisture determination with the neutron probe is independent Of soil salinity
under field conditions (Arnon, 1992). Improvements in portability and reductions in
electronic and radioactive hazards have increased the use of the neutron probe as a tool for
irrigation scheduling (Gear et al., 1977). Application of the neutron probe in turf irrigation
was reported by Snyder et al. (1984). The dependence of the spatial resolution on the water
content (van Bavel, 1962) and the inaccuracies in predicting near-surface soil moisture
content (Snyder et al. 1984) make the neutron probe a less attractive alternative in turf
irrigation scheduling. Also, such disadvantages as the high initial and maintenance costs, and
potential radioactive hazard, have limited the use of the neutron probe primarily to research

(Amon, 1992).

Electrical Resistance Blocks
Soil moisture measurements with electrical resistance units involve a pair of electrodes
embedded in a porous material, usually gypsum, fiberglass, or nylon, buried in the soil. The

linear relationship between moisture content and electrical resistance is used to interpolate the

13

moisture content for various electrical resistance measurements. The need to calibrate for
each soil type, and the deterioration of gypsum blocks, make measuring soil moisture with
electrical resistance blocks highly labor intensive (Cassel, 1984).

The use of soil moisture sensors as tools for irrigation scheduling simplifies the
decisions about when to irrigate and how much water to apply. However, gravimetric
measurements, electrical resistance and gypsum blocks, and tensiometers in soil moisture
studies in laboratory experiments or in field trials have failed to demonstrate the same level
of versatility and convenience as time domain reflectometry (TDR) (Topp and Davis, 1985a).

The demonstrated success of TDR technology in soil moisture determination in
laboratory (1‘ opp et al., 1980) and in field experiments (Topp and Davis, 1985a & b; Carrow,
1991; Safl‘el, 1994) shows great promise for potential use in turf irrigation scheduling.
Although research on TDR has been primarily limited to laboratory experiments (Heirnovaara
et al., 1993) this technique can be used to maintain quality turf without compromising

environmental quality (Carrow, 1991, Saffel, 1994).

Timing Criteria for Soil-based In'igation
Irrigation timing is important, regardless of the system of employed. Soil moisture
monitoring indicates when the soil moisture content has attained a critical limit beyond which
plant growth and quality are affected (Jensen et al., 1971). Management allowable depletion
(MAD) is one of the most fi'equently used timing criteria for soil-moisturebased irrigation
scheduling (Stegrnan, 1985; Cuenca, 1989; Campbell and Campbell, 1982). Doorenbos and
Kassam (1979) suggested a 50% allowable depletion, with adjustments of up to 30%, based

on evaporative demand, for most crops and soils. Establishing such a threshold provides

l4

timing criteria for irrigation scheduling. With low-volume, high frequency irrigation, the
allowable depletion approach becomes less critical because irrigation rates and amounts are
held within the infiltration rate and storage capacity of the root zone (Stegrnan, 1950).
However, environmental concerns and water shortages are also compelling considerations for
more eflicient water management strategies.

Using TDR Topp and Davis (1985a) described soil-based irrigation scheduling as a
simple network whose major limitation has been soil moisture monitoring and interpretation.
The TDR method has been proven effective for soil moisture determination in field crops and
to a limited extent in turf soils. The remainder of this discussion focuses on the origin,

development, and agricultural applications of TDR.

Time Domain Reflectometry

The process of sending high-frequency electromagnetic signals and analyzing the
waveforms through different media is called time domain reflectometry (TDR) (Campbell,
1990). Telephone and electric companies have used TDR extensively for cable testing, to
locate cable ends and defects (Dalton, 1992). In the last decade, agricultural applications of
TDR have received much attention.

Time domain reflectometry originated fi'om attempts to predict molecular structures
from the frequency dependence of the dielectric constants of organic molecules (Fellner-
Feldegg, 1969). Studies on the application of TDR for determining dielectric properties of
soil were first reported by Davis and Chudobiak (1975). In 1977, Davis et al. reported the

electromagnetic detection of soil water content using TDR.

15

Credit for the development of agricultural applications of TDR has been given to
Topp and his co-workers. Topp et al. (1980) proposed an empirical polynomial relationship
between volumetric soil moisture content (VMC) and the apparent dielectric constant of soil
(6). They assumed that the dielectric constant of soil, for all practical purposes, was
insensitive to variations in bulk density, temperature, salinity, and mineral composition (Topp
et al., 1980; Baker and Almaras, 1990), eliminating the need for site-specific calibration.

Since its discovery in the early 19803, agricultural applications of TDR have been
extensively evaluated (Topp et al., 1980, 1980; Topp and Davis, 1985a, b & c; Dalton and
van Genuchten, 1986; Baker and Allamras, 1990; Wraith and Baker, 1991). These
researchers confirmed that TDR determination of soil moisture was not significantly affected
by bulk density, temperature, salinity, or mineral composition Of the soil. This finding,
strengthened the argument for a single empirically derived calibration curve for most soils.
Recent research, however, has demonstrated the dependence of the soil dielectric constant
on temperature and bulk density (Roth et al. 1990).

Knight (1993) investigated the effect of probe geometry, and Zegelin et al. (1989)
studied the effect of transmission line geometries on measurements of VMC by TDR. Roth
et al. (1990) resolved the composite dielectric constant of soil into real and imaginary
components. Their work also showed that the composite dielectric constant of soil can be
resolved into the contributions of the water, soil, and air fractions, while Paterson and Smith
(1981) concluded that TDR discriminates between frozen and free water.

An advantages of TDR is the ease of automation for remote VMC data retrieval
(Baker and Allmaras, 1990; Heimovaara and Bouten, 1990; Herkelrath and Hamburg, 1986).

Hook et al. (1992) used remote diodes to improve soil water determination by TDR,

16

particularly in saline or layered soils. They recovered the effective amplitude of problem
waves with significant reduction in background noise through the amplification of highly
attenuated waves. Their work showed that reliable VMC measurements could be obtained
even at locations of up to 100 m from the TDR unit. For TDR applications involving several
pairs of TDR probes, multiplexing has become the method of choice for automated data
retrieval (Baker and Allmaras, 1990).

Using an impedance-matching transformer (balun) to reduce the mismatch at the
connection between the probes and the transmission lines is a common practice that improves
the resolution of the TDR trace (Biran et al., 1981). Bonnell et al. (1991) compared probes
with and without baluns. They concluded that the TDR wave resolution for probes with a
balun was superior to those without. Hook et a1. (1992) used a three-diode probe to obtain
better wave resolutions compared to a two-wire probe. Comparing multiple-wire probes to
the standard two—wire probes Zegelin et al. (1989) concluded that three-wire probes provided
better resolution than two-wire probes but that probes with four or more wires did not further

improve TDR resolution.

Application of TDR to Soil Water Determination
The permittivity (dielectric constant) of a medium is a measure of the polarization of
the medium when subjected to an electric field (Ledieu et al., 1986). TDR measures the
transit time of generated electromagnetic waves along a probe pair embedded in the soil. The
velocity of a wave along the probe is determined from the travel time, based on the reflections
at the beginning and end of the probes. The propagation velocity of a wave depends on the

dielectric constant within the medium of propagation (Topp et al., 1980; Ledieu et al., 1986).

17

A pulse generator produces a time-step impulse whose reflections are monitored by
an oscilloscope. The dissipation of the signal amplitude varies with the bulk soil electrical
conductivity (Ledieu et al., 1986). The pulse travel time is proportional to the soil apparent
dielectric constant, which is proportional to the soil water content (Topp et al., 1980; Dalton,
1992). At the end of the probe, the electromagnetic wave is reflected back to the TDR unit.
The elapsed time fi'om the emission to the reception of the reflected pulse is a firnction of the
probe length and the dielectric constant of the soil.

Soil moisture determination using TDR is thus based on a step-pulse transit time along
the length of the TDR probes (Topp et al., 1980). This measurement has been extensively
correlated to gravimetrically detemrined volumetric soil moisture content measurements. The
pulse velocity is calculated from the transit time over twice the length of the probe. The
volumetric soil moisture content is related to the distance between inflection points.

A critical step in TDR application to VMC determination is the identification of the
inflection points. The first inflection point, which is the junction of the probe and the
transmission line, is unaffected by the soil water content or salinity (Dalton et al., 1984). The
second inflection point, which occurs at the end of the probe, may be very weakly defined in
problem waves due to dissipation of the voltage in highly dispersive media (saline soils) or
noises in the circuitry (Dalton, 1992).

Tangents at the inflection points are used to determine the distance between
reflections. When manually obtained, such measurements are subject to error (Dasberg and
Nadler, 1987). Keng and Topp (1983) identified the detection of the inflection point as a
major source of error in VMC determination by TDR. Computer algorithms provide better

estimates of the inflection points and travel times and, therefore, more precise VMC

18

measurements. However, computer-determined inflection points may not necessarily
discriminate between good and problem waves (Keng and Topp, 1983; Yanuka et al., 1988).
The derivation of the relationship between soil dielectric constant and volumetric moisture
content, as presented by Topp et al. (1980), is as follows. The pulse velocity, v, of an

electromagnetic wave in a medium of dielectric constant, e, is given by:

where e is apparent soil dielectric constant and c is speed of light. The wave velocity back

and forth the probe length, l is given by:

where t is transit time which ranges from 2 to 3 ns. Rearranging the above equations results

in:

The pulse travel time is proportional to the soil dielectric constant, and the dissipation in the
wave amplitude is proportional to the bulk soil electrical conductivity (Bonnell et al., 1991).
Volumetric moisture content is estimated from the soil dielectric constant, according to the
empirical relationships developed by Topp et al. (1980) and Ledieu et al. (1986). Only the

equation of Topp et al. (1980) is presented here:

19

WC = - 0.053 + 0.02926 - 0.0005562 + 0.000004363

The TDR trace shows the inflections of the pulse as the impedance varies due to
mismatch or changes in capacitance around the transmission lines. The signal energy is
dissipated as the wave travels along the cable (Zeglin et al., 1992). The magnitude of the loss
depends on the type of wire. A twisted-pair wire generally incurs higher losses than coaxial

cable (Pierce, 1992, personal communication).

TDR Set-up

A principal component of the TDR is a metallic cable tester, which is employed
principally to locate faults in metal cables used in the communications industry (Dalton,
1992). The TDR system consists of a high-speed oscilloscope connected to coaxial cables
that are linked to a set of probes embedded in the soil. For automatic data retrieval, the
oscilloscope is firrther connected to a computer or data logger.

Probes usually are rrrade of nonmagnetic materials whose conductivity is greater than
that of the medium of interest. In soil studies, stainless steel normally is used (Dalton, 1992).

Ramo et al. (1965) proposed the following relationship for approximating probe impedance:

120

z=—12£
o (6)0(d)

where Z, is probe impedance, 6 is soil dielectric constant, sis probe separation distance in

centimeters, and d is probe diameter in centimeters. The wave propagation velocity is

20

affected if the cable or probe impedance does not match the TDR output impedance of 50 Q.

Baluns are thus used to reduce this mismatch (Spaan and Baker, 1993).

Effect of Probe Length Spacing and Diameter
The accuracy of the TDR reading depends, in part, on the probe length, which is a
function of the reflection coefficient (p), the dielectric constant (e), and the bulk soil electrical
conductivity. Dalton et al. (1984) and Dalton and van Genuchten (1986) proposed equations

for predicting the maximum probe length (I) for reflection coefficients of at least 10%:

I = [In (gin/Er I (1209)
R

where V, and V, are the amplitudes of the transmitted and reflected waves, respectively, and

p is the reflection coeficient determined from:

 

p .. "(i-I) "0)
20 + Zr)
04) o)

where Z,” is the apparent fiequency of the first reflection and Z,“ is the apparent frequency
of the next reflection (Dalton, 1992).

Highly conductive soils require shorter probe lengths than soils with low bulk soil
electrical conductivity. Short rod lengths induce greater errors than long rods in the
identification of inflection points, whereas long rods reduce resolution as voltage decays

exponentially along the probe length (Dalton, 1992).

21

Yanuka et al. (1988) addressed the problems of frequency loss by attenuation. They

proposed the following equation for calculating the apparent fi'equency (Z,):

__ Z:
Z 0 — E

where Z, is frequency in the air, Z, is apparent fiequency in the medium of interest, and 6 is
dielectric constant. Whereas probe diameter has no significant influence on the TDR
readings, the probe spacing significantly affects soil moisture readings (Knight, 1993). On
the basis of spatial resolution of the TDR a 50 mm spacing has been recommended as the

optimal spacing distance for most soil research (Topp et al. 1980; Dalton, 1992).

Spatial Resolution of TDR

The spatial resolution of the TDR is another concern in soil water determination.
Theoretical analysis (Topp et al., 1980) and experimental measurements (Dalton et al., 1984;
Knight et al., 1993) have suggested that TDR measures an integrated VMC average along the
probe length. Baker and Lascano (1989) suggested that the area of influence of TDR ranges
fiom elliptical to rectangular, with no significant variation in sensitivity along the length of the
wave guide. The approximate soil volume measured by TDR using the proposed configuring
is slightly larger than a cylinder of diameter wider than the probe separation distance.

Using a set of TDR probes with 3.2 mm in diameter, 50 mm spacing distance, and 300
mm long, Baker and Lascano (1989) estimated an effective spatial area of about 1000 mm2
along the length of the probe. Ledieu et al. (1986) stated that 94% of the electromagnetic

energy is restricted to a cylinder with a diameter twice the rod spacing; they also suggested

22

a 50 mm thickness as the maximum effective spacing between probes. Unlike the neutron
probe, spatial resolution of TDR is independent of moisture content (Amon, 1992).

Limited information is available on the spatial and temporal distribution of water in
surface soils under different irrigation regimes. Ritchie (1972) developed equations relating
soil moisture depletion patterns with time for a bare soil in order to predict water use rates
and irrigation needs. Time domain reflectometry provides as high accuracy, precision, and
spatial and temporal resolution as the neutron probe, without the need for site-specific
calibration or the potential risk of exposure to radiation (Wraith and Baker, 1991).

The spatial resolution of a soil moisture measurement technique depends on the
sampling volume. Variation in soil moisture content resulting from infiltration and soil
moisture extraction by roots contribute to differences in VMC over short distances (Baker
and Lascano, 1989). This variability may be accentuated by cracks, biopores, textural
layering, and fingering, as well as preferential routing by roots and water resulting in larger

sampling errors (Bouma et al., 1982).

Applications of TDR to Agricultural Research
The empirical relationship between soil water content and soil dielectric properties has
been applied successfully in many studies (Patterson and Smith, 1981; Dalton and van
Genutchten, 1986; Zeglin et al., 1989; Heimovaara and Bouten, 1990). Topp et al. (1983)
used TDR in field infiltration studies to monitor the advancement of a wetting fiont.
Volumetric water content and bulk soil salinity were determined simultaneously by Dalton

(1987) and TOpp et al. (1988).

23

Advantages of TDR for Soil Moisture Determination

Major weaknesses Of the gravimetric method of volumetric moisture determination
are the concurrent determination of bulk density and soil mass water content, and destructive
sampling. Errors inherent in standard soil moisture determination methods also raise
questions as to the accuracy of these methods, given the spatial and temporal variations in
bulk density.

Time domain reflectometry provides excellent spatial resolution (Hook et al., 1992).
Roth et al. (1990) showed that the uncertainty in water content for a broad range of soil types
did not exceed 0.013 cm’cm". Ledieu et al. (1986) reported an accuracy of within 1% VMC.
For all practical purposes, moisture measurements by TDR presents no radioactive risks, are
independent of soil type and salinity (Topp et al., 1980); thus require no calibration for most
mineral soils. The stainless steel probes used in TDR are not expensive. Also, the ease of
automation with multiplexing for remote data retrieval is a distinct advantage of this technique
(Baker and Allmaras, 1990; Hook et al., 1992).

As opposed to the neutron probe, TDR has been applied to near-surface
measurements of soil water without any loss in accuracy (T Opp et al., 1980, 1982, ; Topp and
Davis, 1981). In turf management, Carrow (1991) and Safl‘el (1994) successfully
demonstrated the potential use of TDR for measuring soil moisture depletion and irrigation
scheduling. Once the TDR probes are installed, the technique is nondestructive and can be
used for continuous monitoring of soil moisture content when automated.

A method of soil moisture determination that is fast, accurate, precise, dependable,
and affordable is desirable for soil-based irrigation scheduling and irrigation modeling.

Research on the use of such technology for the high-frequency, low-volume irrigation in turf

24

management is necessary to evaluate the benefits of turf irrigation modeling and eflicient
irrigation water management. The objective of the turf manager after all is to provide
optimum turf quality and reduce the potential for leaching of agricultural chemicals into

ground water.

Disadvantages of TDR for Soil Moisture Determination

Certain exceptions have been encountered in the use of Topp's empirical relationship.
In their investigation, Patterson and Smith (1981) and Baker et al. (1982) found that the small
difference between the dielectric constant of ice (3.2) and that of soil (2.5-3.5) may cause
difficulties in measuring water contents in thawing frozen soils. Topp et a1. (1980) and
Herkelrath et al. (1991) suggested that high soil organic matter content causes lower
dielectric values and lower VMC. Nadler et al. (1991) outlined difficulties encountered in the
interpretation of TDR signals in layered soils. These problems indicate the need to explore
further the capabilities of TDR for soil moisture determination in unusual soil situations.

Despite the widely reported ease of remote retrieval of data through multiplexing,
(Dasberg and Dalton, 1985) extensive field trials using this method have not been successfirl.
This has delayed the use of TDR in real-time irrigation programming. The attenuation of the
electromagnetic impulse by long cables, saline soils, and layered soils (Topp et al., 1985a) is
a problem that needs to be resolved. The air-gap effect (Anann, 1977; Topp et al., 1988),
variation of moisture content along the probe length, variability resulting from contact with
succulent roots and/or living organisms such as earthworms (Saffel, 1994), and the calibration
recommended for organic samples (Topp et al., 1980; Dalton, 1992) are some limitations of

TDR determination of VMC. Keng and Topp (1983) cited the accuracy in inflection-point

25

delineations as a major source of error in VMC detennination by TDR This was true for
both manually obtained or computer-derived inflection points.

In spite of the above-mentioned advantages of TDR as a practical tool for soil
moisture deterrrrination, there is a need to improve the reliability, precision, and accuracy of
TDR readings. Such improvements will come about through a better understanding of the
theory and limitations of the TDR. Spatial and temporal variations in soil properties may
cause variable wave forms without distinct inflection points, resulting in problem reflections
that make it difficult to obtain reliable VMC readings. Frequent breakdowns of soldered
junctions between stainless steel probes and cable wires require time and effort to repair and
may alter the integrity of the waves. New probe designs and construction reduce the chance
of breakdown as the junctions of the probes and the cable wires are mounted and molded in
an epoxy resin that reduces the chances of breakdown. As with the neutron probe, the
prohibitive initial cost of the TDR equipment is a deterrent to its practical use by turf
managers.

The general objectives of this study were to evaluate the use of TDR in turf irrigation
scheduling, evaluate the competition between annual bluegrass and creeping bentgrass fairway
turfs under three irrigation regimes and to evaluate soil moisture retention predictions from
two irrigation models to TDR estimates.

Chapter 3 addresses the adaptation of TDR for volumetric soil moisture content
determination and irrigation scheduling in fairway turfs. Chapter 4 deals with annual
bluegrass-creeping bentgrass competition under three irrigation regimes. Chapter 5 evaluates
bihourly moisture depletion patterns during daylight hours. In Chapter 6, the water balance

subroutine of the Systems Approach to Land Use Sustainability (SALUS) model was used

26

to evaluate hydrologic components for turfgrass ecosystems during the summers of 1992
through 1994 and in Chapter 7 an irrigation scheduling model (The SC S-SCHEDULER) was
used to evaluate seasonal water applications, excess water and excess rainfall and excess

irrigation for the three irrigation regimes.

Literature Cited

Agnew, ML, and RN. Carrow. 1985. Soil compaction and moisture stress
preconditioning in Kentucky bluegrass. 1. Soil aeration, water use and root response.
Agron J. 77:872-878.

Amer, SA, O.T. Keefer, M.A. Weltz, D.C. Goodrich, and LB. Bach. 1994. Soil moisture
sensor for continuous monitoring. Water Resource Bulletin.

Amon, I. 1992. Agriculture in dry lands, principles and practices. In Developments in
agricultural and managed-forest ecology. Elsevier, New York.

Arman, A0. 1977. Time-domain reflectometry air-gap problem for parallel wire transmission
lines. Report of Activities, Part B. Paper 77-1 B 59-62: Geol. Surv. of Canada, Ottawa.

Aronson, L.J., A]. Gold, RJ. Hull, and J .L. Cisar. 1987a. Evapotranspiration of cool-season
turfgrasses in the humid Northeast. Agron. J. 79:901-905.

Aronson, L.J., A.J. Gold, and R]. Hull. 1987b. Cool-season turfgrass response to drought
stress. Crop Sci. 27:1261-1266.

Augustin, 3.1., and GB. Snyder. 1984. Moisture sensor-controlled irrigation for maintaining
Bermudagrass turf. Agron. J. 76:848-850.

Aurasteh, MR. 1983. A model for estimating lawn grass water requirement considering
deficit irrigation shading and application efliciency. Ph.D. Diss. Utah State Univ,
Logan.

Baker, J.M., and RR. Allmaras, 1990. System for automating and multiplexing soil moisture
measurement by time-domain reflectometry. Soil Sci. Soc. Am. J. 54:
1-6.

Baker, J.M., and R]. Lascano. 1989. The spatial sensitivity of time-domain reflectometry.
Soil Sci. 1172378-383.

Baker, T.H.W., J.L. Davis, H.N. Hayhoe, and GO Topp. 1982. Locating the fi'ozen-

unfrozen interface in soils using time-domain reflectometry. Can. Geotech. J. 19:511-
517.

27

28

Barber, SA. 1984. Soil nutrient bioavailability. A mechanistic approach. John Wiley and
Sons, New York.

Beard, J .B. 1973. Turfgrass science and culture. Prentice-Hall, Englewood Clifl‘s, NJ.
Beard, J.B. 1982. Turf management for golf courses. Macmillan, New York.

Beard, J.B. 1985. An assessment of wateruse by turfgrasses. p. 45-60. In V.A. Gilbeault and
ST. Cockerham (ed) Turfgrass water conservation. Publ. 21405. Univ. of California,
Riverside.

Beard J .B. and KL. Green. 1994. The role of turfgrasses in environmental protection and
their benefits to humans. J. of Environ. Qual. 23 :(3) 452-460.

Belcher, D.J., TR Cuykendall, and HS. Sack. 1950. The measurement of soil moisture and
density by neutron and gamma ray scattering. Civil Aeron. Admin. Tech. Develop. Rep.
No. 127. Washington, DC.

Bennett, OE, and ED. Doss. 1960. Effect of soil moisture level of root distribution of cool-
season forage species. Agron. J. 52:204-207.

Biran, I., B. Brando, I. Bushkin-Harav, and E. Rawitz. 1981. Water cOnsumption and growth
rate of 11 turfgrasses as affected by mowing height, irrigation frequency, and soil
moisture. Agron. J. 75:85-90.

Black, CA 1965. Methods ofsoil analysis. Vol. 1. ASA-CSSA-SSSA Monograph. Madison,
WI.

Bonnell, RR, R S. Broughton, and F. Gomaa. 1991. The measurement of bulk soil salinity
using time domain reflectometry. Canadian. Soc. of Agric. Eng. Proc. 33:225-229.

Boufford, RW., and RN. Carrow. 1980. Effects of intense, short-term traffic on soil physical
properties and turfgrass growth. Trans. Kansas Acad. Sci. 32:78-85.

Bouma, J., C.F.M. Belmans, and L.W. Dekker. 1982. Water infiltration and redistribution in
a silt loam subsoil with vertical worm channels. Soil Sci. Soc. Am. J. 46:917-921.

Campbell 6.8, and MD. Campbell. 1982. Irrigation scheduling using soil moisture
measurements: Theory and practice. p. 25-42. In D. Hillel (ed.) Advances in irrigation,
Vol. 1. Academic Press, New York.

Campbell 6.8. 1990. Dielectric properties and influence of conductivity in soils at
one to fifty megahertz. Soil Sci. Soc. Am. 54:332-341.

29

Carrow, RN. 1980. Influence of soil compaction on three turfgrass species. Agron. J.
72:1038-1042.

Carrow, RN. 1991. Turfgrass water use, drought resistance and rooting patterns in the
Southeast. Technical Completion Report. ERC. 01-91. USDI/USGS.

Cassel, D.K. 1984. Irrigation scheduling. 11. Simple low-cost techniques. Crops Soils.
36:615-618.

Cassel, D.K., and DR. Nielsen. 1986. Field capacity and available water capacity.
In A. Klute (ed.) Methods of soil analysis. Part 1. Agronomy No. 9. Madison,
WI.

Cuenca, RH, and KY. Amegee. 1987. Analysis of evapotranspiration as a regionalised
variable. Adv. Irrig. 4: 180-220.

Cuenca, RH. 1989. Irrigation system design, an engineering approach. Prentice-
Hall, Englewood Cliffs, NJ.

Dalton, EN. 1987. Measurement of soil water content and electrical conductivity using tirne-
domain reflectometry. International conference on measurement of soil and plant water
status. I. Soils. Logan, UT.

Dalton, RN. 1992. Development of time-domain reflectometry for measuring soil water
content and bulk electrical conductivity. p. 143-167 . In Advances in measurement of soil
physical properties: Bringing theory into practice. SSSA Spec. Pub]. 30. SSSA,
Madison, WI.

Dalton, F.N., W.N. Herkelrath, D.S. Rawlings, and JD. Rhoades. 1984. Time domain
reflectometry simultaneous measurement of soil water content and electrical
conductivity with a single probe. Science (Wash, DC) 224:989-990.

Dalton, F.N., and MT. Van Genuchten. 1986. The time-domain reflectometry method for
measuring soil water content and salinity. Workshop on mechanisms of ion transport in
soils, Zurich, Switzerland, 20-23 May 1985. Geoderrna 38:237-250.

Danielson, RE, OM. Feldhake, and WE. Hart. 1981. Urban lawn irrigation and manage-
ment practices for water savings with minimum effect on lawn quality. Water resources
Res. Inst, Fort Collins, CO.

Dasberg, S., and RN. Dalton. 1985. Time domain reflectometry field measurements of soil
water content and electrical conductivity. Soil Sci. Soc. Am. J. 49:293-297.

30

Dasberg, S., and A. Nadler. 1987. Field sampling of soil water content and bulk electrical
conductivity with time-domain reflectometry. p. 99-102. In International conference on
measurement of soil and plant water status. I. Soils. Logan, UT.

Davis, J.L., and W.J. Chudobiak. 1975. In situ meter for measuring relative permittivity of
soils. Geol. Survey of Canada. Energy, Mines and Resources of Canada, Ottawa. Paper
75. 1:75-79.

Davis, J .L., G.C. Topp, and AP. Annan. 1977. Electromagnetic detection of soil water
content: Progress report 11. p. 96-109. In Remote sensing of soil moisture and
groundwater. Workshop Proceedings, Canadian Aeronautics and Space Institute,
Ottawa, Canada.

Doorenbos, J ., and W.O. Pruitt. 1977. Guidelines for predicting crop water requirements.
F AO Irrig. and Drain. Paper 24. Rome, Italy.

Doorenbos, J., and AH. Kassam. 1979. Yreld response to water. FAO irrigation and drainage
paper 33. FAQ, UN Rome.

Doss, B.D., D.A. Ashley, O.L. Bennett, and RM. Patterson. 1966. Interactions of soil
moisture, nitrogen, and clipping frequency on yield and nitrogen content of coastal
berrnudagrass. Agron. J. 57:539-542.

Doss, B.D., O.L. Bennett, DA Ashley, and HA. Weaver. 1962. Soil moisture regime effect
on yield and evapotranspiration from warm-season perennial forage species. Agron. J.
54:239-242.

Fairbome, ML. 1982. Water use by six grass species. Agron. J. 74:824-830.

Feldhake, C.M., R.E. Danielson, and JD. Butler. 1983. Turfgrass evapotranspiration. 1.
Factors influencing rate in urban environments. Agron. J. 75:824-830.

Fellner-Feldegg, H. 1969. The measurement of dielectrics in the time domain. J. of Phys.
Chem. 3:616-623.

Fok, Y.S., and ET. Murabayshi, 1978. Estimation of urban irrigation water use
and requirements. ASAE Technical Paper No. 78-2030. St. Joseph, MI.

Fry, J .D., and ID. Butler. 1989. Annual bluegrass and creeping bentgrass evapotranspiration
rates. Hortsci. 24:269-271.

Gardner, W.R., and D. Kirkharn. 1952. Determination of soil moisture by neutron
scattering. Soil Sci. 73:391-401.

31

Gear, RD., SA Dransfield, and M. D. Campbell (1977) Irrigation scheduling with neutron
probe. J. Irriga. Drain. Div., Am. Soc. Civ. Eng. 102:291-298.

Groundwater Education in Michigan (GEM). 1982. Western Michigan University,
Ypsilanti.

Hagan, H.M., HR Haise, and T.W. Edminister (ed.). 1967. Irrigation of Agricultural lands.
Agronomy 11. ASA-CSSA-SSSA, Madison, WI.

Hall, D.G.M., and F.W. Heaven. 1970. Comparison of measured and predicted soil moisture
deficits. J. Soil Sci. 30:225-237.

Hanks, R.J., and G.L. Aschcrofi. 1980. Applied soil physics: Soil water and temperature
applications. Springer-Verlag, New York.

Hanks, R.J., and R.W. Hill. 1980. Modeling crop response to irrigation in relation to soils,
climate and salinity. Pergarnon Press, New York.

Hanks, R.J. 1992. Applied soil physics. Soil water and temperature applications. Springer-
Verlag, New York.

Heimovaara, T.J., and W. Bouten. 1990. A computer-controlled 36-channel time domain
reflectometry system for monitoring soil water contents. Water Resources Research
26:2311-2316.

Heimovaara, T.J., J.I. Freijer, and W. Bouten. 1993. The application of TDR in laboratory
column experiments. Soil Techn. 6:261-272.

Hengeller J. 1993. Turf battles. Agric. Engr. ASAE, St Joseph, MI.

Herkelrath, W.N., and SP. Hamburg. 1986. Real-time, multi-site monitoring of soil moisture
with time-domain reflectometry. p. 157. In Agronomy Abstracts, ASA, Madison, WI.

Hillel, D. 1980. Applications in soil physics. Academic Press, New York.
Hook, W.R., NJ. Livingston, Z.J. Sun, and PB. Hook. 1992. Remote diode shorting
improves measurement of soil water by time domain reflectometry. Soil Sci. Soc. Am.

J. 56:1384-1391.

Jensen, ME. (ed). 1974. Consumptive use of water and irrigation requirements. ASCE, New
York.

Jensen, M.E., J.L. Wright, and BJ. Pratt. 1971. Estimating soil moisture depletion from
climate, crop and soil data. Trans. ASAE 14:954-959.

32

Johns, D., J.B. Beard, and C.H.M. van Bavel. 1983. Resistance to evapotranspiration from
a St. Augustinegrass turf canopy. Agron. J. 75:419-422.

Keng, J.C.W., and GO Topp. 1983. Measuring water content of soil columns in the
laboratory: A comparison of gamma ray attenuation and TDR techniques. Can. J. Soil
Sci. 63:37-43.

Kim, KS, and J .B. Beard. 1988. Comparative turfgrass evapotranspiration rates and
associated morphological characteristics. Crop Sci. 28:328-332.

Kincaid, D.E., E.G. Kruse, H.R Duke, and DP. Champion. 1978. Evapotranspiration
computed to estimate leaching fractions. ASAE 22:310-319.

Kirkham, D., and GS. Taylor. 1950. Some tests of a four-electrode prone for soil moisture
measurement. Soil Sci. Soc. Am. Proc. 34:536-537.

Kneebone, W.R, and IL. Pepper. 1982. Consumptive water use by sub-irrigated turfgrasses
under desert conditions. Agron. J. 74:419-423.

Knight, I. 1993. Sensitivity of time domain reflectometry measurements to lateral variations
in soil water content. Water Resour. Res. 28:2345-23 52.

Kopec, D.M., RC. Shearman, and TR Riordan. 1988. Evapotranspiration of tall fescue turf.
Hort. Science 23:300-301.

Koski, AJ. 1983. Seasonal rooting characteristics of five cool season turfgrasses. MS Thesis,
Ohio State Univ, Columbus.

Ledieu, J., P. de Ridder, P. deClerck, and S. Dautrebande. 1986. A method of measuring soil
moisture by time-domain reflectometry. J. Hydro]. (Amsterdam) 88:319-328.

Letey, J., Morgan, W.C., Richards, 8.1., and Valoras, N. 1966. Physical soil Amendments,
Soil compaction, irrigation, and wetting agents in turfgrass management 111. Effects on
oxygen diffusion rate and growth. Agr. J.

58:531-535.

Marsh, AW., RA Strohmarr, S. Spaulding, V. Younger, and V. Gibeault. 1980. Determining
the water requirements of various turfgrass species. Irrig. Assoc. Ann. Tech. Conf. Proc.
1:34-35.

Nadler, A., S. Dasberg, and I. Lapid. 1991. Time domain reflectometry measurements of
water content and electrical conductivity of layered soil columns. Soil Sci. Soc. Am. J.
55:93 8-943.

33

Patterson, DE, and M.W. Smith. 1981. The measurement of unfrozen water content by
time-domain reflectometry: Results fiom laboratory tests. Can. Geotech. J. 18: 13 1-144.

Peacock, CH, and AB. Dudeck. 1984. Physiological response of St. Augustinegrass
to irrigation scheduling. Agr. J. 76: 275-979.

Ramo, 8., JR. Whinnery, and T. Van Duzer. 1965. Fields and waves in communication
electronics. John Wiley and Sons, New York.

Richards, 8.1, and AW. Marsh. 1961. Irrigation based on soil moisture
measurements. Soil Sci. Soc. Am. Proc. 25:65-69.

Ritchie, J.T. 1972. Model for predicting evaporation fiom a row crop with incomplete cover.
Water Resour. Res. 8: 1204- 12 13.

Ritchie, J.T., and M. Amato. 1990. Field evaluation of plant extractable soil water for
irrigation scheduling. Acta Horticulturae 278:595-615.

Ritchie, J.T. 1991. Specifications of an ideal model for predicting crop yields. p. 97-122. In
RC. Muchow and IA. Bellamy (eds) Climatic risk in crop production: Models and
management for semi-arid tropics and subtropics. Proc. Int]. Symposium, St. Lucia
Brisbane, Queensland, Australia. CAB International, Willingford, UK.

Roth, K., R. Schulin, H. Fluhler, and W. Attinger. 1990. Calibration of time domain
reflectometry for water content measurement using a composite dielectric approach.
Water Resour. Res. 26:2267-2273.

Safell, MT 1994. Time domain reflectometry for turf irrigation programming. MS thesis,
Mich. State Univ., East Lansing.

Salaiz, T.A., RC. Shearman, T.P. Riordan, and E]. Kinbacher. 1991. Bentgrass cultivar
water use and rooting responses. Crop Sci. 31: 1331-1334.

Shayya, W.H., and V.F. Bralts. 1988. Irrigation scheduling sofiware "SCS-SCHEDULER."
Version 3.0. Department of Agricultural Engineering, Michigan State University, East
Lansing.

Shearman, RC. 1986. Kentucky bluegrass cultivar evapotranspiration rates. HortSci. 21:455-
457.

Shearman, RC. 1989. Perennial ryegrass cultivar evapotranspiration rates. HortSci. 24:767-
769.

34

Simpson, JR, and J .J . Meyer. 1987. Water content measurement comparing a TDR array to
neutron scattering. p. 111-1 14. In International conference on measurement of soil and
plant water status. I. Soils. Logan, UT.

Snyder, G.H., B.J. Augustin, and J.M. Davidson. 1984. Moisture sensor controlled irrigation
for reducing N leaching in berrnudagrass turf. Agron. J. 76:964-969.

Spaans, E.J.A., & Baker, J.M. 1993. Simple baluns in parallel probes for time-domain
reflectometry. Soil Sci. Soc. Am. J. 57:668-673.

Stegrnan, EC. 1985. Scheduling irrigation: Applied timing criteria. Adv. Irrig. 2: 1-30.

Taylor, SA and GE Ashcrofl. 1972. Physical edaphology. In The physics of Irrigated and
non-irrigated soils. San Francisco:

Topp, GO, and J.L. Davis. 1981. Detecting infiltration of water through soil cracks by time-
domain reflectometry. Geoderrna 26:13-23.

Topp, GO, and J.L. Davis. 1985a. Measurement of soil water content using time-domain
reflectometry (TDR): A field evaluation. Soil Sci. Soc. Amer. J. 49: 19-24.

Topp, GO, and J .L. Davis. 1985b. Comment on "Monitoring the unfiozen water content of
soil and snow using time domain reflectometry" by Mean Stein and Douglas L. Kane.
Water Resour. Res. 21: 1059-1060.

Topp, GO, and J.L. Davis. 1985c. Time-domain reflectometry (TDR) and its application to
irrigation scheduling. p. 107-127. In D. Hillel (ed.) Advances in irrigation. Academic
Press, New York.

Topp, G.D., J .L. Davis, and AP. Arman. 1980b. Electromagnetic determination of soil water
content: Measurements in coaxial transmission lines. Water Resour. Res. 16:574-5 82.

Topp, G.D., J .L. Davis, and AP. Arman. 1982a. Electromagnetic determination Of soil water
content using TDR I. Applications to wetting fronts and steep gradients. Soil Sci. Soc.
Amer. J. 46:672-678.

Topp, G.C., J .L. Davis, and AP. Annan. 1982b. Electromagnetic determination of soil water
content using TDR. 11. Evaluation of installation and configuration Of parallel
transmission lines. Soil Sci. Soc. Amer. J. 46:678-684.

Topp, G.C., J.L. Davis, and 1H. Chinnick. 1983. Using TDR water content measurements
for infiltration studies. p. 231-240. In Advances in infiltration. ASAE, St. Joseph, Mich.
Publ. no. 11-83.

35

Topp, G.C., M. Yanuka, W.D. Zebchuk, and SJ. Zegelin. 1988. The determination of
electrical conductivity using TDR: Soil and water experiments and coaxial lines. Water
Resour. Res. 24:945-952.

Tovey, R, 18. Spencer, and DC. Muckel. 1969. Turfgrass Evapotranspiration. Agron. J.
61:863-866.

Valoras, N., and J. Letey. 1966 Soil oxygen and water relationships to rice grth Soil Sci.
101:210-215.

van Bavel, CHM, and DO. Harris. 1962. Evapotranspiration rates from berrnudagrass and
corn at Raleigh, North Carolina. Agron. J. 54:319-322.

Voorhees, W.B., C.G. Sensit, and W.W. Nelson. 1978. Compaction and soil structure
modification by wheel traflic in the northern Corn Belt. Soil Sci. Soc. Amer. J. 43:344-
349.

Watson, J. R 1980. Turfgrass irrigation practices for maximum efficiency. Agron Abstracts,
Madison, WI.

Wraith, J .M., and J .M. Baker. 1991. High-resolution measurement of root water uptake using
automated time-domain reflectometry. Soil Sci. Soc. Am. J. 55:928-932.

Yanuka, M., G.C.Topp, S. Zegelin, and W.D. Zebchuk. 1988. Multiple reflection and
attenuation of TDR pulses: Theoretical considerations for applications to soil and water.
Water Resour. Res. 24:939-944.

Zegelin, SI, 1. White, and CF. Russell. 1992. A critique of the time domain reflectometry
technique for determining field soil water content. p. 187-207. In Advances in
measurements of soil physical properties: Bringing theory into practice. SSSA Spec.
Pub]. 30. SSSA, Madison, WI.

CHAPTER 3

Moisture Depletion on Fairway Turfs Under Three Irrigation Regimes
Using Time Domain Reflectometry

Abstract

Improvements in irrigation management by turf managers have focused mainly on
irrigation system performance with limited attention paid to the temporal and spatial
distribution of soil water following irrigation. The objectives of the study were: i) to evaluate
volumetric moisture content dynamics in turf root zones under three irrigation regimes and
ii) to adapt time domain reflectometry (TDR) for turf irrigation scheduling. Volumetric soil
moisture content was monitored using TDR at 0-5, 5-10, 10-15, and 15-25 cm depths under
three irrigation regimes during the summers of 1992, 1993, and 1994. The irrigation
treatments were: i) return soil to field capacity (F C); ii) apply 2.5 mm daily (DLY); and iii)
apply 25 mm only upon the appearance of wilt stress (STR). Penncross creeping bentgrass
(Agrosris palusrris L. Huds.) and annual bluegrass (Poa annua L. var. reptans) were
maintained on an Owosso sandy loam soil under fairway conditions. Moisture depletion
patterns for both species were measured daily except on rainy days. Applied irrigation
differed significantly by irrigation treatment but not during wet periods. Volumetric moisture
content by depth ranked 0-5 > 5-10 > 10-15 > 15-25 for the irrigated treatments and for the

stress treatment during wet periods. During dry periods, VMC for the stress treatment

36

37

ranked 15-25 > 10-15 > 5-10 > 0-5. Volumetric moisture content for the three irrigation
treatments ranked PC > DLY > STR during dry periods and DLY > FC > STR for wet
periods. Mean volumetric moisture content for the irrigated treatments were above field
capacity (0.28 cm3 cm") during wet periods while full recharge of the stress treatment
occurred only during heavy rainfall. Under such conditions, there is a potential for leaching
of agricultural chemicals into ground water. Time domain reflectometry proved to be a useful
tool for turf irrigation scheduling but the question of number of probes and locations on the
golf course or other turfs needs to be addressed. Regular monitoring of VMC by TDR could
reduce overwatering and minimize potential leaching of agricultural chemicals into ground

water.

Introduction

Research on the spatial and temporal dynamics of soil moisture has been impeded by the
lack of adequate techniques for soil moisture determination. Soil moisture data could provide
usefirl information on soil hydrological processes, irrigation scheduling and modeling (Ritchie
and Amato, 1990; Hanks, 1992). Repeated soil moisture measurements could be used to
monitor soil moisture storage and depletion for turf irrigation management. Topp et al (1980)
adapted time domain reflectometry (TDR) for determining volumetric moisture content, based
on soil dielectric properties, which can be used for irrigation programming. However, the use
of TDR in real-time turf irrigation scheduling has been limited to research trials (Carrow,

1991; Saffel, 1994).

38

Ideally, evapotranspiration (ET) models should be used to schedule irrigation. In
practice irrigation scheduling remains a challenge that depends on the experience and
convenience of turf managers. In urban areas, ET is site specific (Danielson et al., 1981) and
may vary considerably from regional predictions (F eldhake et al., 1983). A soil-based method
of irrigation programming is expected to provide better estimates of soil moisture depletion
by site than regional ET predictions (Carrow, 1991), thereby improving site-specific irrigation
scheduling.

This study evaluated TDR as a tool for repeated and rapid VMC determination for turf
irrigation scheduling using two cool-season fairway turfs. The objectives of this study were:
i) to evaluate moisture storage in turf root zones using TDR; and ii) to adapt TDR for turf

irrigation programming.

Materials and Methods

A three-year study was conducted at the Hancock Turfgrass Research Center at
Michigan State University. The soil type was a modified Owosso sandy loam (fine-loamy
mixed mesic T ypic Hapludalf). Volumetric soil moisture content for established annual
bluegrass (Poa annua L. var. reptans) and Penncross creeping bentgrass (Agrostis palustris
L. Huds.) turfs was evaluated under three irrigation regimes. The turf was mowed at 16 mm
height and maintained according to typical fairway practices recommended for cool-season
grasses in Michigan.

The irrigation treatments were: i) return the soil water content to field capacity (F C)
on days when TDR readings were taken; ii) apply 2.5 mm of water daily (DLY) (Vargas,

1994); and iii) a rainfed or stress treatment that received no irrigation (STR). Irrigation was

39

applied at 0300 h. All plots received additional irrigation following fertilization to minimize
phytotoxicity and during sprinkler evaluations . Time domain reflectometry readings were
taken between 0700 and 0900 h for up to 60 days during the summers of 1992, 1993, and
1994.

Irrigation blocks (11 x 11 m) were split in half to form subplots (5.5 x 11 m) randomly
seeded to creeping bentgrass or annual bluegrass, established in 1989 (Safl‘el, 1994). Each
plot had a buffer strip about 1.2 m wide. Because there were no significant difl‘erences in
VMC by species, VMC data for both species were averaged for analysis. A Rainbird pop-up
head with 21.5 L min‘1 average output was installed at each corner of the plot. Plots with the
same irrigation treatment were controlled by the same irrigation switch. Irrigation treatments
served as whole plots, whereas the turf species and depths of TDR installations served as the
split and strip plots, respectively. There were three replications for each treatment.

The TDR setup consisted of the following components: i) a time domain reflectometer
(1502C, Tektronix, Redmond, OR) that generates pulses, a sampler that receives the reflected
impulse, and an oscilloscope that displays the waveform; ii) a set of antenna or cable wires;
and iii) sets of stainless steel wave guides or probes. The cables transmit the pulses fi'om the
TDR to the stainless steel probes embedded in the soil (Campbell, 1990).

Stainless steel TDR probes, 200 mm long and 3.2 mm in diameter, were installed as
described by Saffel (1994). The center-center separation distance between probe installations
was 50 mm. Each subplot (species) had 11 pairs of stainless steel rods installed horizontally
at three equidistant locations 2 m from a switch box at the center of each plot (Fig. 3.1).

Impedance matching transformers (baluns) were connected to each rotary switch to reduce

4o

 

Bentgrass Annual bluegrass

probes
I

    

i
l
' probes, 4 depths
l
l
I

probes SWIICh iJOX

probes, 3 depths
4.

 

  

 

 

 

 

 

Fig. 3.1 Spatial arrangement of TDR probes per plot for the difl‘erent depths.

the mismatch between TDR connectors and cable wires (Topp et al., 1980; Biran et al., 1981;
Spaans and Baker, 1993). The tOp three probes were insralled horizontally and parallel to the
soil surface at 2.5, 7.5, and 12.5 cm depths, with three replications. The last pair of probes
also was installed parallel to the soil surface but vertically, one at 17.5 cm and the other at
22.5 cm depths. The probes at the 15-25 cm depth were replicated only two times due to the
limited number of positions on the rotary switches used to transmit TDR signals from one set
of probes to another. This resulted in a statistically unbalanced design.

The spatial arrangement of TDR probes for one plot is shown in Fig. 3.1. The soil

volume measured depends on probe configuration and the length of the TDR probes, based

41

on the elliptical sphere of influence of the TDR rods (Topp et al., 1980). The probe placement
allows for VMC measurements at four depths (2.5, 7.5, 12.5 cm for the first three probe pairs
and 17.5 and 22.5 cm for the last set of probes) with an approximate sphere of influence
down to the 25 cm depth. The relative area of soil moisture measurement along the length
of the probe pair is shown in the lower-right-hand comer of Fig. 3.1.

The waveform of a voltage pulse propagated through stainless steel probes embedded
in the soil and reflected back to the oscilloscope is analyzed to estimate the composite
dielectric constant (e) of the soil. The travel time (t) along a stainless steel probe of length
L was used to calculate e and VMC, using an empirically derived polynomial as described by

Topp et al. (1980):

VMC = - 0.053 + 0.02926 - 0.0005562 + 000000439363

where e = (ct/2L)2 and c = the speed of light.

In 1993, gravimetric moisture measurements were converted to VMC by multiplying by
the soil bulk density as a basis for comparison to the TDR measurements. Three samples
were taken per subplot, using a 3.2 cm diameter probe, and sectioned by depth corresponding
to the vertical sphere of influence of the installed TDR probes (0-5, 5-10, 10-15, and 15-25
cm). Because of the destructive nature of gravimetric sampling, samples were taken away
from the probes to avoid disturbing the probe installations. This was a potential source of
error but was necessary for long term use of TDR installations.

The weighted field capacity of the soil for the 0-25 cm depth (0.28 cm’cm") was

determined from TDR measurements two days afier a soaking rain and as a reference for

42

estimating irrigation need for the field capacity treatment. Runofl‘ was not measured but may
have accounted for some of the moisture loss.

The experimental design was a split-strip plot with three replications. All unbalanced
data were analyzed using SAS GLM and SNK mean-separation procedures; otherwise, PROC
ANOVA was used (SAS, 1994). The precision of the TDR readings were evaluated using
descriptive statistics (means, standard deviations and standard error, variance, and minimum

and maximum values).

Results and Discussion

Descriptive statistics for repeated TDR measurements for difi‘erent probe pairs on two
dates are presented in Table 3.1. The mean, mode, and median had the same value, and the
standard error and variance were very small for both dates. This confirms that TDR provides
reproducible soil moisture measurements in turf ecosystems.

Agreement between gravimetrically determined VMC and TDR measurements (R2 =
0.99), given anticipated spatial variability, shows that TDR provides reasonable moisture
estimates even under turfgrass conditions. However, TDR over predicted VMC at lower
moisture contents by up to 2.1 % (Table 3.2). Because TDR and VMC data were not
collected from the same location, expected spatial variability could explain some of the
differences between methods. The high R2, despite the expected spatial variability of VMC,
suggest that TDR is a dependable tool for soil moisture determination for the shallow rooting
depths encountered in cool-season turfs.

Soil moisture content was influenced mostly by precipitation and applied irrigation for

the irrigated treatments, moderated by other weather factors. Weather conditions for 1992

43

Table 3.1. Descriptive statistics for consecutive TDR measurements.

 

 

17 June 1993 23 Aug. 1994

Maximum 0.247 0.245
Minimum 0.234 0.225
Mean 0.241 0.238
Mode 0.241 0.238
Median 0.241 0.238

Std. error 0.0004 0.0004
Variance 0.0000 1 0.00002

n ' 43 108

 

Table 3.2. Comparison of volumetric soil moisture content means across all treatments
by the gravimetric method (GMC) and by TDR for different depths.

 

 

 

 

DeLth TDR GMC Difference
(cm) cm3 cm‘3
0-5 30.4aT 30.8a -0.4
5-10 27 .9b 26.4b 1.5
10-15 26.7c 24.6c 2.1
15-25 26.0c 23.9c 2.1

 

TMeans within columns followed by the same letter are not significantly different.
11 = 66 for 0-5, 5-10, and 10-15 depths, and n = 44 for 15-25 cm depth for TDR and n = 18
for GMC for each depth.

to 1994 are summarized in the appendix. Rainfall amounts and distribution were variable
within seasons and among years. In all years, most of the rainfall coincided with the peak of
the warm summer months. In 1992 and 1993, heavy rains occurred toward the second half

of the season. In 1994, rainfall amounts and frequency masked the effect of higher

44

temperatures and solar radiation on moisture depletion from mid-June to mid-July. The
highest single rainfall event of 70 mm occurred on day 166 in 1994. All three years were

relatively wet for East Lansing, MI.

Volumetric Moisture Content by Irrigation Treatment

The analysis of variance showed that VMC content was significantly affected by
irrigation treatment. However, high amounts of rainfall occasionally masked the differences
in VMC among treatments. Volumetric moisture content by depth for the different irrigation
treatments for 1992 are presented in Fig. 3.2. Volumetric moisture content for the irrigated
treatments fell within 0.25 and 0.35 cm3 cm‘3 for most sampling dates, while VMC for the
stress treatment ranged from a low of0. 17 cm3 cm‘3 during a prolonged dry period in 1992
to a maximum of 0.33 cm’cm" at the beginning of the study for the 0-5 cm depth (Fig. 3.2).

Soil moisture depletion ranges were highest for the stress treatment than for the irrigated
treatments particularly, the 0-5 cm depth. A critical period for the stress treatment was
between sampling day 30 and 40 in 1992 when VMC fell below cm3 cm". The high root
density (Chapter 4), the high organic matter content and proximity of the 0-5 cm depth to
fluctuating environmental conditions are possible explanations for this. During wet periods
VMC by depth ranked 0-5 > 5-10, > 10-15 > 15-25 for all treatments. During dry periods
however, the trend was reversed in the stress treatment (IS-25 >10-15 > 5-10 > 0-5).

Time series plots for VMC for 1993 are presented in Fig. 3.3. Moisture content for the
irrigated treatments fell between 0.24 and 0.34 cm3 cm'3 for most sampling days, while VMC

for the stress treatment ranged from 0.13 to 0.33 cm3 cm”. Again VMC by depth ranked 0-5

45

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

2.5 mm daily treatment
.1:
E
U
#1
I
3
2
§
0.10
- 0 10 20 30 40 50 60
Days of sampling
Field Capacity
1 0.35
0.30 -
:7
a5 0.15 .
n
i
8.
g 0.20 ..
‘
>
0.15 -
0.10 ,
0 10 20 30 40 50 60
Days of sampling
Stress treatment
0.35 \
03° " w . r ' ' r-r ------------------ R -----
.. "a. \\
E 0.25 N .
3 ‘ - - - a -----
E
J:- -
.0— 010 d -----------------
G
>
0.15 - ....................................................
...._ 0-5 cm _— 5-10 cm ..—. 10-15 cm —— 15-25 cm
0.10 . - -
0 10 20 30 40 50 60
Days of sampling

 

 

 

Fig. 3.2. Volumetric moisture content (VMC) means for the different irrigation treatments
by depth, 1992.

46

> 5-10 > 10-15 > 15-25 cm for the irrigated treatments while VMC by depth for the stress

treatment ranked 15 -25 > 10-15 > 5-10 > 0-5 cm during dry periods. The 0-5 cm depth for
the irrigated treatments had consistently higher VMC than other depths. Volumetric moisture
content trends by treatment also followed the same general pattern as in 1992. The 2.5 mm
daily and field capacity treatments showed only minor increases in VMC from rainfall for all
years. Conversely, the stress treatment showed significant gains in VMC following heavy
rains. This indicates that when soils are maintained at or above field capacity, most of the
rainfall is lost as drainage or runoff.

Volumetric moisture content trends for 1994 are presented in Fig. 3.4. Early in the
season in 1994 moisture readings were out of range due to faulty operation of equipment for
Thereafter, VMC levels for the irrigated treatments were above 0.24 cm3 cm'3 as in 1992 and
1993. Increase in VMC on day 40 in 1994 (Fig. 3.4) was highest for the stress treatment
while the irrigated treatments showed only marginal gains in VMC for all depths. This
indicates that both irrigation treatments failed to efficiently accommodate potential moisture
gains from rainfall. On most days, VMC for the bentgrass plots was consistently higher than
for the annual bluegrass plots although the differences were not significant. This suggests
I differential moisture extraction by depth, which may be explained by the differences in root
mass density between species.

Volumetric moisture content response to irrigation and rainfall varied significantly by
irrigation treatment primarily during dry periods. The field capacity (F C) and 2.5 mm daily
(DLY) treatments, respectively, received 234 and 382 mm of irrigation in 1992, 297 and 382

mm in 1993, and 298 and 382 mm in 1994. Applied irrigation ranked DLY > PC > STR for

47

 

2.5 mm daily treatment. 1993

 

 

 

O 10 20 30 4Q 50 60
Days of Sampling

 

 

 

Field Capacity Treatment. 1993

 

(cm3/cm 3)

 

 

 

0 10 20 3O 40 50 50 1
Days of Sampling i
|

 

Stress treatment 1993 l

 

 

 

o 10 20 30 40 50 63
Days of sampling

—.— 0-5 cm —.—— 5-10 cm .... 10-15 cm —— 15-25 cm

 

 

Fig. 3.3. Volumetric moisture content (VMC) for the different irrigation treatments by depth,
1993.

48

all years. The stress treatment received no irrigation for all years as 1992, 1993, and 1994
were very wet years. The stress treatment showed the highest moisture depletion for all
depths, as expected. Because no irrigation was applied to the stress treatment, water deficit

accumulated until it rained.

Volumetric Moisture Content by Depth

The analysis of variance showed that soil moisture content was significantly different by
soil depth. Because of the large amount of data involved, only data from 1993 were used
for the discussion that follows. Time series plots (Fig. 3.5) show the effect of irrigation
treatment on VMC by depth for the three irrigation treatments. The 0-5 cm depth had the
highest VMC for all irrigated treatments throughout the study. This implies that with light
and frequent irrigation, more water resides in the 0-5 cm depth. The delineation between the
0-5 cm depth and the other depths was most consistent in the 2.5 mm daily treatment. In this
treatment, the rankings for the three other depths were 5-10 > 10-15 > 15-25. On several
dates, there were significant differences in VMC among sampling depths and irrigation
treatment.

Volumetric moisture content for the irrigation treatments for a typical wet day in 1993
and for a dry day in 1992 are presented in Fig. 3.6a and Fig. 3.6b. Moisture content
decreased with depth for the irrigated treatments. Generally, differences between irrigation
treatments were more apparent during dry periods than in wet periods. For the stress
treatment, VMC followed the same pattern as in the irrigated treatments during very wet

periods. During dry periods VMC trends by depth ranked 15-25, > 10-15 > 5-10 > 0-5 cm.

49

 

1.5 .- Dally Treat-ant. 1m

—-— O-Sem —~— S-IOan —— 10-15esn —-— 15-15em

 

20 30 40 50 60
Days of sampling

0
n—
O

 

Field Capadly Treatment. 1994

men-taut (culls-J)
F. E E i E

is.“

 

 

us
->— 0-5e- —-‘- $10:- —-- 10-15e- + 1545:-
{an
> as
0 10 20 )0 50 60
TI-(d-yl)
Sue-Treatment. 1994

—O— 05c- -"—- 54h —— )C-ISe- --— 1525:.

 

 

 

 

Fig. 3.4. Volumetric moisture content (VMC) means for the different irrigation treatments
by "depth, 1994.

50

However, during wet periods, a reversal of the VMC sequence for the different depths was
observed (0-5 > 5-10 > 10-15 > 15-25) as for the irrigated treatments. Reversals in soil
moisture content for the different depths in the stress treatment occurred in all years when
VMC fell below the 0.20 and 0.25 cm3 cm‘3 range for the stress treatment as shown earlier
(Figs. 3.2, 3.3, and 3.4).

Upward water flux may contribute moisture to the 15-25 cm depth and subsequently to
that of other depths. The mechanisms operational under these conditions are not fully
understood and deserve further investigation. It has been suggested that hydraulic lift by
roots (Richards and Cardwell, 1987) and capillary rise during periods of high evaporative

demand may contribute water to the rootzone

Table 3.3. Regression equation coefficients for moisture depletion for the stress
treatment during a dry-down cycle in 1992.

 

 

 

 

Depth a b
0-5 cm 31.3 0.44 0.03 0.87
5-10 cm 30.2 0.38 0.02 0.87
10-15 cm 30.8 0.36 0.02 0.96
15-25 cm 27.5 0.06 0.01 0.91

 

Mean VMC by depth for the irrigated treatments were generally at or above field
capacity for most days. For the stress treatment, VMC dropped as low 0.13 cm3 cm‘3 in 1992,
0.14 cm’cm" 1993 and less than 0.10 cm’cm'3 in 1994 during peaks of dry down cycles. Most
water balance models estimate plant water uptake from root length density. Higher root mass

density fi'om bentgrass plots (Murphy et al., 1994; Saifel, 1994) suggests higher water uptake,

51

.89 £520: .58 E 2:05.59: :C:sw_._.__ :_o._o._.:_o 2: .8 530-7. 7.9.0.5: 2.52:. 22.220: 2:262 .m.m .wE

 

555.3% be anna—
cw cm 3. cm .5

:—

 

. is.- UL ll >JG r!" 11.

 

 

3.:

(ma/am) om

 

 

5.....an be oh;—

on cm 6.. an :N c— c

 

Uh ll >JD wen i

 

c'
(run/sun) DNA

 

 

 

 

 

 

 

 

 

 

 

 

 

.c
r 11111111111111111111111111111111111111 mm... mad
cw... 0...:
no: £19: :3 mN-m— no: .59.: :3 v4.2
u:=._E-v. be in: nearest 5.. aha:
cw cr- cv cm .3 c— c on on . cw on cu e— c
3.: c—é
m—é m—é
2.: m S... m
3 D
m~.= m. r-«é m.
{new r - - -.- -- -- - - (6.35m
3.... 1V -- - -- -- -- - -- - - on...
5.... e96

 

 

3.: .59: a: 26

 

no: 52...: :8 m...

 

 

52

assuming equally functional roots. Differences in VMC by depth between species occurred
on a few dates in all years. However, VMC means across all depths were not significantly
different by species as reported by Satfel, 1994. In earlier studies, Beard (1973), Fry and
Butler (1989) reported no significant difference in ET between annual bluegrass and creeping

bentgrass.

Regression Equations for Moisture Depletion During
a Dry-Down Period

Ritchie (1972) reported a linear relationship for moisture depletion fi'om a bare soil
plotted against the square root of time. Regression equations fitted to moisture depletion
data from the stress treatment during a 1992 drydown period could be used to predict when
the soil moisture content will attain an established threshold for irrigation. The general
equation for each depth was of the form:

Y = a - bX + cX2

where Y is the VMC, X is time in days, a is the y-intercept, and h and c are X coeflicients for
the exponential dry-down prediction equation, as shown in Table 3.3. Because the intercept
and coefficients of the dry-down curve may vary over time depending on evaporative demand
and rainfall, regular moisture readings are necessary for this method.

Seasonal VMC means and turf quality ratings for the different irrigation treatments for
1992 showed a linear relationship as presented in Fig. 3.7. Danielson et al. 1981; Aurasteh,
1983, also reported linear turf quality response to applied irrigation over certain moisture
ranges. From Fig. 3.7 it is evident that to maintain a turf quality rating of 7or higher, soil

moisture content must be greater than 0.27 cm3 cm". To maintain the turf at the minimum

53

 

 

 

 

 

 

 

 

 

 

0.40
0.35 — L
I '\'\
O 30 d ............. . .N\ .—
. .\-. “gm“ ...... g
to 0.25 A i‘~~..~§a
I §u+ .............. .1
8
m
E 0.20 —
O
E
g 0.15 — 3
{=2 0.10 - . - “'" DLY -
.5 -V—' STR
E
g 0.30 "‘ I
E at I...
” '\. ............... i
E, 0.25 — \.‘_ _____ i______i
O
>
0.20 -
./../§' ----- -v— ------ ""7",
i”
0.15 -
b
0.10 l l 1 T l
0 5 10 15 20 25 30

Depth of TDR installations, cm

Fig. 3.6. Moisture depletion means across species for the field capaciy, FC; 2.5 mm daily,
DLY; and stress, STR treatments for 20 Aug. 1993 (a wet day, a) and 25 Aug. 1992
(a dry day, b).

54

acceptable quality rating of 6, volumetric moisture content should be maintained at about
0.26 cm3 cm". This implies that if water conservation is the goal then lower quality turf must
be acceptable. Using a soil moisture sensor to predict when soil VMC attains a pre-
established critical level will lead to more efficiently irrigation scheduling with acceptable
reductions in turf quality. When moisture sensors become affordable repeated moisture
measurements may help conserve water and reduce drainage losses from irrigation in urban

areas as well as on golf courses.

Moisture Depletion and Irrigation Scheduling

Soil moisture sensors have been used for turf irrigation scheduling with significant water
savings (Augustin and Snyder, 1984). Volumetric soil moisture content trends by depth for
the TDR study were the same for all treatments during wet periods, i.e., 0-5 > 5-10 10-15 >
15 -25 cm. During drydown periods the stress treatment showed unique soil moisture
depletion trends by depth (Fig. 3.8) that could be used for irrigation scheduling. A transient
state (0.20 to 0.25 cm’ cm") was evident in the stress treatment during each prolonged
drydown period, where all soil depths had about the same soil moisture content
implying that there was very little net movement of moisture between soil depths of interest
over this moisture range. This could be an ideal level at which soil moisture should be
maintained to minimize drainage losses without subjecting the turfs to moisture stress. For
example if the 0.25 cm3 cm" moisture level is assigned as the setfull and the 0.20 cm3 cm"
the refill point as suggested by Topp and Davis, 1985, then irrigation is applied each time
VMC is depleted to the 0.20 cm3 cm'3 level and stopped when VMC attains the 0.25 cm3 cm”

level. The lower limit selected in this illustration is 0.03 to 0.05 cm3 cm'3 above the moisture

Turf quality ratings

55

 

10

 

 

 

 

 

 

 

 

 

 

 

 

 

22 24 26 28 30 32

Volumetric soil moisture content (crn3 cm‘3)

Fig. 3.7. Volumetric moisture content (VMC) means versus turf quality means, 1992.

levels at which Safi‘el (1994) reported visual signs of wilt stress, and 0.03 cm3 cm" below
the drained upper limit of 0.28 cm3 cm". Further moisture depletion below the above range,
however, led to difl‘erential moisture retention by depth but the observed trends were again
reversed with adequate rainfall (Figs.3.3, 3.4, 3.5 and 3.8). When moisture levels fell below
the 0.20 cm3 cm‘3 limit the 0-5 cm depth had the lowest VMC followed sequentially by the
5-10, 10-15, and 15-25 cm depths. The observed VMC trend is the opposite of that observed

during wet periods and coincides with root mass distribution patterns by depth.

56

.32 E veto; :33 be a macaw 208.3: 9.2.... 2: 8.. 53v .3 5:033 2:302 .w.n .wE

wEEEam mo 93D
ov on

cm

0—

 

Eo mm..m_ rill Eu 2.

 

:coz

 

2 '0' Eco—-m Ill

:5 mic if!

0—

 

 

 

 

 

 

 

zabom

 

 

 

mm

gum/gum iuaiuoo amisrour otnaumIo A

57

Two important decision faced by turf managers are: i) when to inigate, and ii) how
much to irrigate. Using a soil moisture sensor such as TDR simplifies the decision on how
much and when to irrigate by establishing two critical limits - the “setfirl” and "refill" points
(Campbell and Campbell, 1982; Topp and Davis , 1985). The refill point indicates when to
start irrigation and the setfiill point indicates when to stop. The establishment of these points
will depend on the quality of turf desired as well as other management considerations.

A relationship between VMC versus turf quality for a given soil type, turf species and
location may facilitate the selection of the setfull and refill points by selecting appropriate
VMC limits in relation to turf quality. Repeated VMC measurement provide information as
to when the established limit is attained for efficient irrigation used to schedule turf irrigation

based on established soil moisture limits.

Conclusions

The above discussion confirms that 'IDR measurements in turf soils show a high degree
of precision (standard error < 0.001) and accuracy (R2 = 0.99). Repeated monitoring of
VMC by TDR in turf root zones as a basis for irrigation planning can help reduce
overwatering and the potential for ground water contamination in turf ecosystems.

Volumetric soil moisture content trends for the irrigated treatments show average
moisture content levels at or above field capacity most of the time. Moisture depletion from
both irrigated treatments were over a narrow range while VMC ranges for STR were
particularly, for the 0-5 cm depth. The high VMC for the irrigated treatments in this study
suggest that irrigation scheduling in wet years should not be based on field capacity (100%

ET) or 2.5 mm daily as such recommendations fail to accommodate potential moisture gains

58

from rainfall. Conjunctive use of ET and soil moisture depletion data could improve turf
irrigation management and minimize drainage losses. Such a schedule would account for
contributions fiom upward water flow, particularly when the water table is shallow.

The ranking of soil moisture retention by irrigation treatment during wet periods was
DLY > FC > STR but during dry periods the ranking was PC > DLY > STR. Soil moisture
depletion by depth was similar for both species but differed by irrigation treatment depending
on rainfall amounts and frequency. Moisture retention by depth was in the order 0-5 > 5-10
> 10-15 > 15-25 cm for the irrigated treatments. For the stress treatment the trend was the
same as for the irrigated treatments during wet periods, but a complete reversal of the above
trend (15-25 > 10-15 > 5-10 > 0-5) was evident during peak dry periods. Although the 0-25
cm depth is usually considered as one depth in conventional cropping, the partitioning of this
depth is important in understanding soil moisture dynamics in shallow-rooted turf.
Regression equations developed from dry down cycle showed that short term irrigation
forecasting is possible from repeated VMC measurements.

A method of scheduling irrigation that is accurate, site specific, rapid, dependable, and
affordable is most desirable for the turf industry. Time domain reflectometry accurately
measures a known spatial volume. The possibility of automation and accuracy at shallow
depths are advantages of TDR compared to the neutron probe that may benefit the turf
industry. In addition to saving time and labor, widespread adaptation of TDR in irrigation
programming will result in more efficient use of limited water resources and reduce drainage
losses and the potential for ground water contamination in turf ecosystems. However, the
high initial cost of TDR is prohibitive at this time. Secondly, the question of placement and

number of probes to be used for specific sites need to be addressed.

Literature Cited

Augustin, B.J., and 6H. Snyder. 1984. Moisture sensor-controlled irrigation for maintaining
Bermudagrass turf. Agron. J. 76:848-850.

Aurasteh, MR. 1983. A model for estimating lawn grass water requirement considering
deficit irrigation shading and application efficiency. PhD. Diss. Utah State Univ.,
Logan.

Beard, J .B. 1973. Turfgrass Science and Culture. Prentice-Hall, Englewood Cliffs, NJ.

Biran, 1., B. Brando, I. Bushkin-Harav, and E. Rawitz. 1981. Water consumption and growth
rate of 11 turfgrasses as affected by mowing height, irrigation frequency, and soil
moisture. Agron. J. 75:85-90.

Campbell, 6.8., and MD. Campbell. 1982. Irrigation scheduling using soil moisture
measurements: theory and practice. p. 25-42. In D. Hillel (ed.) Advances in
irrigation, Vol. 1. Academic Press, New York.

Campbell, 0.8. 1990. Dielectric properties and influence of conductivity in soils at l to 50
megahertz. Soil Sci. Soc. Am. 54:332-341.

Carrow, RN. 1991. Turfgrass water use, drought resistance and rooting patterns in the
Southeast. Technical Completion Report. ERC. 01-91. USDI/USGS.

Danielson, RE, OM. Feldhake, and WE. Hart. 1981. Urban lawn irrigation and manage-
ment practices for water savings with minimum effect on lawn quality. Water resources
Res. Inst, Fort Collins, CO.

F eldhake, C.M., R.E. Danielson, and J .D. Butler. 1983. Turfgrass evapotranspiration. I.
Factors influencing rate in urban environments. Agron. J. 75:824-830.

Fry, J .D., and J .D. Butler. 1989. Annual bluegrass and creeping bentgrass evapotranspiration
rates. Hortscience. 24:269-271.

Hanks R.J. 1992. Applied soil physics: Soil water and temperature applications. Springer-
Verlag. New York.

Murphy, J.A., M.G. Hendricks, P.E. Rieke, A.J.M. Smucker, and BE. Branham. 1994.

turfgrass root systems evaluated using minirhizotron and video recording method.
Agron. J. 86:247-250.

59

6O

Richards, J .H. and MM. Caldwell. 1987. Hydraulic lifi: substantial nocturnal water transport
between soil layers by Artemisia tridentata roots Oecologia 73:486-489.

Ritchie, J.T. 1972. A model for predicting evapotranspiration fi'om a row crop with
incomplete cover. Water Resour. Res. 8: 1204-1213.

Ritchie, J .T., and M. Amato. 1990. Field evaluation of plant extractable soil water for
irrigation scheduling. Acta Horticulturae 278:595-615.

Safell, MT. 1994. Time domain reflectometry for turf irrigation programming. MS thesis,
Mich. State Univ., East Lansing.

SAS Institute. 1994. SAS user's guide: Statistics. SAS Inst., Cary, NC.

Spaans, E.J.A., and J.M. Baker. 1993. Simple baluns in parallel probes for time domain
reflectometry. Soil Sci. Soc. Am. J. 571668-673.

Tektronix. 1987. Tektronix metallic TDR's for cable testing. Tektronix, Redmond, OR.
Topp, G.C., and J .L. Davis. 1985c. Time-domain reflectometry (TDR) and its application to
irrigation scheduling. p. 107-127. In D. Hillel (ed.) Advances in irrigation. Academic

Press, New York.

Topp, G.D., J .L. Davis, and AP. Annan. 1980. Electromagnetic determination of soil water
content: Measurements in coaxial transmission lines. Water Resour. Res. 16:574-582.

Vargas, J .M., Jr. 1994. Management of turfgrass diseases. Lewis Publishers, Ann Arbor, MI.

CHAPTER 4

Annual Bluegrass and Creeping Bentgrass Competition
Under Variable Irrigation Treatments

Abstract

Interspecific competition between annual bluegrass (Poa annua L. var. reptans) and
creeping bentgrass (Agrastispalusrrr’s Huds. L.) has gained much attention over the years as
efforts to control annual bluegrass economically have not been very successfiil. It is widely
agreed that annual bluegrass is favored over creeping bentgrass in compacted soils especially
when over irrigated. A 3-year study (1992-1994) was conducted at the Hancock Turfgrass
Research Center to evaluate the interspecific competition between annual bluegrass and
Penncross creeping bentgrass fairway turfs under three irrigation regimes: i) apply 2.5 mm
daily irrespective of rainfall; ii) return soil to field capacity based on daily TDR estimated
moisture deficits; and iii) apply 25 mm only upon the appearance of wilting stress. Irrigation
blocks, established in 1989 with dimensions (11 x 11 m), were split (11 x 5.5 m) and seeded
at random as pure stands of annual bluegrass or creeping bentgrass. The turfs were mowed
at a height of 16 mm and maintained according to standard practices for cool-season fairway
turfs in Michigan. Volumetric soil moisture content (VMC) was measured daily between 700

and 900 h except on rainy days using horizontally installed time domain reflectometry (TDR)

61

62

probes at 0-5, 5-10, 10-15, and 15-25 cm depths. The TDR readings averaged over all depths
were used to determine irrigation need for the field capacity treatment (F C). Turf biomass,
quality ratings and interspecific competition were evaluated. Root mass densities were
greatest in the 0-5 cm depth, ranging fi'om 63-73 % of total root mass with an average of
69.7% for 1993 and 1994. The two year average for the percent root mass by depth for the
5-10, 10-15, and 15-25 cm depths were 15.9, 9.5 and 4.4% respectively. Quality ratings for
bentgrass were significantly higher than for annual bluegrass. Annual bluegrass/creeping
bentgrass species composition was not significantly affected by irrigation treatment. High
bentgrass percentages on both annual bluegrass and creeping bentgrass plots was strong

evidence of superior bentgrass competition over annual bluegrass under our conditions.

Introduction

Annual bluegrass (Poa annua L var. reptans) is a highly competitive species on closely
mowed inigated golf course fairways (Beard et al., 1978). The invasive nature of this weed
is favored by: i) the proliferation of persistently viable seeds; ii) strong fibrous roots; iii) lack
of a dormancy mechanism and iv) compacted soils (Beard, 1973; Beard et al., 1978;
Gaussoin, 1988).

Proper shot control on fairways requires optimum turfgrass uniformity, density,
smoothness, and resiliency (Beard, 1982). These factors depend, in part, on soil moisture
content. Irrigation of fairway turf should be based on evaporative demand, rooting depth, soil

moisture depletion, topography, and traffic intensity (Beard, 1982). Ideally, just enough

63

water should be applied to meet plant requirements. In practice, turf irrigation continues to
be a challenge for many turf managers, sometimes resulting in overwatering even as water
shortages are reported in some parts of the US.

Both annual bluegrass and creeping bentgrass show striking similarities in adapted
growing conditions and cultural requirements such as cutting height, mowing frequency, and
a wide range of tolerance of fertilization requirements (Beard, 1973; Vargas, 1994).
However, different pH, irrigation, and N fertilization requirements are traditionally
recommended for each species (Beard, 1982; Vargas, 1994). High N application rates
(Engel, 1974), high soil moisture content (Beard, 1982), and soil compaction (Wilson and
Latham, 1969) also enhance annual bluegrass encroachment into bentgrass turf.

Various cultural factors influence population dynamics between the desired species and
annual bluegrass. Phosphorus fertilization rates (up to 86 Kg ha") have been shown to
enhance annual bluegrass encroachment into Penncross creeping bentgrass whereas doubling
the above P rate produced better bentgrass turf with less annual bluegrass (6085 et al., 1975).

Cultural practices such as clipping removal which removes seedheads can significantly
reduce annual bluegrass encroachment into bentgrass plots (Gaussoin et al., 1985) while
frequent irrigation generally favors annual bluegrass competition over the desired species
(Mahdi and Stoutmeyer, 1953; Younger, 1959; Beard et al., 1978). To the contrary, Safl‘el
(1994) found that irrigation fi'equency had no significant effect on annual bluegrass/creeping
bentgrass population dynamics.

Effective chemical annual bluegrass control has been reported by Gaul and Christians
(1988). However, annual bluegrass control remains a permanent feature in turfgrass

maintenance programs for many golf courses, with ever-increasing costs. The high cost of

64

annual bluegrass control implies that cultural practices that discourage its competition with
desirable species deserve firrther study. As an alternative, some turf scientists have suggested
the use of annual bluegrass as a desired turfgrass species (Vargas, 1994). Development of
improved annual bluegrass selections also has been reported (White, 1989).For most turf
managers and researchers annual bluegrass control remains a major challenge.

With the growing competition for water amid declining water resources, the study of
annual bluegrass competition under variable irrigation deserves renewed attention. This study
evaluated the effect of three irrigation regimes on the plant responses and interspecific

competition between annual bluegrass and creeping bentgrass under fairway conditions.

Materials and Methods

A 3-year study was conducted at the Hancock Turfgrass Research Center to evaluate
interspecific competition between established annual bluegrass and creeping bentgrass under
three irrigation regimes. The soil type was a modified Owosso sandy loam (fine-loamy mixed
mesic Typic Hapludalt). Irrigation plots were established in 1989 (Safl‘el, 1994).

The irrigation treatments were: i) apply 2.5 mm daily; ii) apply irrigation daily to return
the soil to field capacity, based on moisture depletion as measured by time domain
reflectometry (TDR); and iii) apply 25 mm only upon the appearance of wilting stress.
Irrigation treatments served as whole plots with dimensions 11 x 11 m. Each plot was split
(5.5 x 11 m) and seeded at random to either annual bluegrass or creeping bentgrass. Turfs
were mowed three times a week at a cutting height of 16 mm and fertilized with phosphorus
and potassium based on soil test with normal N recommendations for cool season fairway

turfs in Michigan.

65

A rotary pop-up Rainbird Maxipaw sprinkler was located at each corner of the plots,
with an average flow rate of 21.5 L h". The operating pressure of the sprinklers was 276
kPa. The radius of throw was 11 m with head to head coverage. Irrigation was scheduled
at 03 00 each day to ensure minimum wind effects on application and distribution
uniforrnities.

Time domain reflectometry (TDR) was used to measure volumetric soil moisture
content (VMC) at four depths between 0700 and 0900 h except on days when rainfall
interrupted data collection. The TDR data were used to schedule irrigation for the field
capacity treatment and to provide a basis for comparing soil moisture retention and turfgrass
performance under a range of irrigation scenarios practiced in Michigan.

The stainless steel TDR probes, 20 cm in length and 3.2 mm in diameter were installed
horizontally, with a 5 cm center-to-center spacing at 2.5, 7.5, and 12.5 cm depths in three
replicates. A fourth set of probes was installed with one at 17 .5 cm and the other at 22.5 cm,
replicated only twice due to the limited number of positions on the rotary switch used to
direct the TDR signal from probe to probe. The arrangement of the probes provided
moisture content estimates for the 0-5, 5-10, 10-15, and 15-25 cm depth, respectively. The
probes were soldered to coaxial cables buried underground and connected to a switch box at
the center of each plot.

Composite soil dielectric constant (6) data for each treatment, replicate, depth and
species (198 data points per day) as determined by TDR, were converted directly into VMC

according to the equation (Topp et al., 1980):

WC = 0.053 + 0.02926 - 0.00055(—:2 + 0.000004363.

66

Data collected were stored according to probe number by plot and species for each date.

Root mass was determined using a 1.8 cm diameter probe. Three cores about 40 cm
long were taken at random from each split-plot. The thatch layer was cut off and the
remainder of each core was separated into subsarnples corresponding to the depth increments
of the sphere of influence of the probes: 0-5, 5-10, 10-15, and 15-25 cm. The roots were
separated from the soil using a hydro-pneumatic elutriator (Smucker et al., 1982). Washed
samples were rinsed with 5% alcohol and oven dried at 60°C for 24 h and weighed.

Interspecific competition was determined twice each year (14 July and 16 September
1992, 26 July and 10 September 1993, and 24 May and 29 July 1994) expressed as percent
bentgrass on each split-plot. The predominant species at cross hairs of a 1 x 1 m grid on an
entire split-plot was used to quantify the population dynamics of annual bluegrass and
creeping bentgrass under the study conditions.

Clipping weights were taken three times during the growing season in 1993 and four
times in 1994 using a mower with a 0.45 m swath for a 5 m distance. The fresh clippings
were oven dried at 60°C for 24 hours and weighed.

Quality ratings were taken three times during the summer of 1993 and 10 times in 1994
for each species by irrigation treatment. Ratings were from 1 to 9, 9 representing ideal turf
conditions and 6 as the minimum acceptable turf quality.

The experimental design was a split-strip plot design with three replications. The
irrigation treatment served as whole plots. The two turf species were the split, and the strip
was the depth. This design allowed for the evaluation of soil moisture retention by depth and

by species under the three irrigation regimes and all possible interactions. The data were

67

analyzed using the Statistical Analysis System (SAS) general linear models with the Student

Newman Keuls (SNK) mean separation procedure (SAS Institute, Inc., 1994).

Results and Discussion

The analysis of variance showed that irrigation treatment (IRR), turf species (SPE),
depth (DEP) and the interactions IRR“ SPE and IRR*DEP significantly affected root mass
density. Root mass distribution by depth for the different irrigation treatments is presented
in Table 4.1. Root mass density in the 0-5 cm depth was significantly higher than in all other
depths. The data indicated a range of 63 to 73% of the root mass within the 0-5 cm depth
for the three sampling dates. The 5-10 cm depth had about 1.5 times more roots than the 10-
15 cm depth and more than three times the root mass density of the 15-25 cm depth averaged
across species.

The highest root mass density was recorded in early June, 1994 in the 0-5 cm depth.
Samples taken in late July and mid-August for this depth were generally lower with the
exception of 27 July 1994 in the stress treatment. This agrees with the conclusion of Koski
et al. (1983), who reported reduction in root mass for cool-season grasses from heat and
water stress during the hot summer months. The root mass densities for the 0-5 cm depth for
STR were higher than for FC on all sampling dates but higher than DLY only on 27 July
1994. Root mass density for STR were highest in the 5-10 cm depth for all sampling dates.

Root mass density averaged across depths for annual bluegrass and creeping bentgrass
are presented in Table 4.2. Bentgrass had significantly higher root mass densities than
annual bluegrass for all sampling dates for the DLY and FC. For STR root mass density for

bentgrass was significantly higher than for annual bluegrass only in 1993. The lack of

68

significant difference between species in 1994, may be explained by heavy rainfall that
coincided with the peak of the summer heat

Clipping weights by irrigation treatment are shown in Table 4.3. The irrigated
treatments had higher clipping weights than the stress treatment, particularly during dry

periods. Even when the difi‘erences were not significant, the stress plots had consistently.

Table 4.1. Mean root mass density averaged across species by irrigation treatment, 1993

 

 

 

 

 

 

and 1994.
ug. 1993 6 June 1994 27 July 1994
Kg rn'3
0-5 cm 26.4a1' 33.6a 21.9a
5-10 cm 5% 5.7b 5.9b
DLY 10-15 cm 3.8bc 3.5bc 3.6bc
15-25 cm 1.3d 1.5d 1.1d
0-5 cm 21.2a 28.3a 20.7a
FC 5-10 cm 4.8b 4.7b 4.3b
10-15 cm 3% 4.6b 4.2b
15-25 cm 1.5c 1.2c 2.0c
0-5 cm 25.6a 30.2a 29.0a
5-10 cm 7. lb 7.0b 8.4b
STR
10-15cm 3.3c 3.3c 2.8c
15-25 cm 2.0c 2.3c 2.00

 

T Means within columns by irrigation treatment followed by the same letter are not
significantly different. LSD at = 0.05. DLY, F C, and STR and 2.5 mm daily, field capacity,
and stress treatments, respectively.

lower clipping weights than did the irrigated treatments. Similar results were reported by

Saffel, 1992 suggesting that turfs on the stress plots may not firlly recover from moisture

69
stress induced within the growing season. Differences in clipping weight between the 2.5 mm
daily and the field capacity treatments were not significant at the 5% probability level. But
differences in clipping weight between the FC and DLY were significant at lower probability
levels only during dry periods (15 July and 17 Aug. 1993). Differences in clipping weight
between species were not significant. This was probably due to the high percentages of
bentgrass encroachment on annual bluegrass plots or an inherent similarity in ET for both

species as reported by Beard, 1973; and Saffel, 1994.

Table 4.2. Root mass density averaged across depths for annual bluegrass and creeping
bentgrass by irrigation treatment, 1993 and 1994.

 

 

 

 

 

Irrigation S . Date
Treatment p ecres
18 Au&1993 6 June 1994 27 July 1994
Kg m"

Bent 9.1a? 12.7a 11.2a
DLY

Poa 8.7b 9.5b 5.0b
FC Bent 8.2a 12.7a 10.1a

Poa 3% 6% 5.5b

Bent 8.6a 10.6a 10.8a
STR Poa 5.8b 10.7a 10.2a

 

’r Means within columns by irrigation treatment followed by the same letter are not
significantly different. LSD at = 0.05. DLY, F C, and STR are 2.5 mm daily, field capacity,
and stress irrigation treatments, respectively.

Quality ratings for the different irrigation treatments and turf species for 1993 and 1994
are presented in Table 4.4. Quality ratings for the daily treatment (DLY) were consistently

higher than for the stress treatment (STR). However, differences between DLY and FC were

70
significant only 2 of 7 rating dates. Quality ratings for creeping bentgrass were significantly

higher compared to those for annual bluegrass on all rating dates

Table 4.3. Clipping weight means across species by irrigation treatment for 1993 and 1994.

 

 

 

 

 

Irrigation Date

Treatmem 18 July 1993 24 Aug. 1993 18 Sept. 1993 20 June 1994
s m"

DLY 15.7211 12.1a 13.0.a 15%

FC 12.8a 11.9a 12.1a 12.7b

STR 6% 8.2b 7.5b 6.0c

 

1' Means within columns by date followed by the same letter are not significantly difl‘erent.
LSD at = 0.05. DLY, FC, and STR are 2.5 mm daily, field capacity, and stress irrigation

treatments, respectively.

Table 4.4. Turfgrass quality ratings by irrigation treatment and species.

 

Date
6/20/93 7/15/93 8/17/93 9/1/93 9/15/93 6/2/94 8/22/94

 

Treatment

DLY 8.2a? 7.8a 7.8a 8.9a 8.8a 8.7a 9.0a
FC 8.0ab 7.3b 7.7a 8.3ab 7.6ab 8.8ab 8.8ab
STR 7% 5.0c 6.2b 7.3b 7.3b 8.3b 8.6b
5m

Bent 8.2a 7.3a 7.6a 8.6a 8.1a 8.4a 8.5a
Poa 7.3b 6.3b 6.8b 7.7b 7.7b 8.3b 7.6b

 

1' Means within columns for each treatment followed by the same letter are not significantly
different. LSD at = 0.05. DLY, FC, and STR are 2.5 mm daily, field capacity, and stress
irrigation treatments, respectively.

71
The percent bentgrass in both bentgrass and annual bluegrass plots is given in Table 4.5. The
percent bentgrass on annual bluegrass plots ranged from a low of 52% in 1992 on annual
bluegrass stress plots to over 81% for the daily irrigation treatment in 1994. High bentgrass
counts on both annual bluegrass and creeping bentgrass plots was strong evidence of superior
bentgrass competition under our study conditions. The analysis of variance showed that
irrigation had no significant effect on annual bluegrass and creeping bentgrass population
dynamics under our conditions. This finding is consistent with that of Saffel (1992) but
contradicts the popularly held view that higher irrigation favors annual bluegrass competition
It must be noted that there was limited trafl'rc on this

over bentgrass (Younger, 1959).

experimental site compared to that on golf course fairways.

Table 4.5. Percent creeping bentgrass in annual bluegrass and creeping bentgrass plots as
affected by irrigation treatment.

 

 

 

 

 

 

 

Irrigation . 1992 1993 1994
Treatment Species 7/14 9/16 7/26 9/10 5/24 7/29
% Composition

DLY Bent“ 93.0a1' 93.1a 97.0a 95.3a 81.4a 91.6a
Poa 65.3b 64.3b 79.3b 78.5b 73.0b 81.9a
Bent 89.7a 91 .0a 96.7a 97.0a 78.2a 87.5a

PC Poa 74.7b 76.3b 79.7b 79.3b 70.1b 76.4b
Bent 85.7a 94.0a 95.0a 97.0a 87 .9a 89.8a

STR Poa 62.0b 52.0b 75.2b 65.3b 70.4b 75.0b

 

T Means within columns for each treatment followed by the same letter are not significantly
different. LSD at = 0.05. DLY, FC, and STR are 2.5 mm daily, field capacity, and stress
irrigation treatments, respectively. "' Plots were established in 1989 as pure stands.

Saffel, 1992.

72

Conclusions

The analysis of variation showed that irrigation treatment did not significantly affect the
percent composition of annual bluegrass and Penncross creeping bentgrass. Consistently high
bentgrass percentages in annual bluegrass and bentgrass plots was strong evidence of superior
bentgrass competition under our study conditions.

Penncross creeping bentgrass had superior quality than annual bluegrass on most
sampling dates. With the exception of 20 June 1994, there was no significant difference in
quality ratings between the 2.5 mm daily and the field capacity treatment. Quality ratings for
both irrigated treatments were significantly higher than for the stress treatment.

Root mass densities for the creeping bentgrass was significantly higher than for annual
bluegrass in both irrigated treatments for all sampling dates. About 63 to 70% of the root
mass averaged across species was within the 0-5 cm depth. Although root data was collected
for two years, root mass densities for early June were higher than those sampled later in the

summer.

Literature Cited
Beard, J .B. 1973. Turfgrass Science and Culture. Prentice Hall, Englewood Cliffs, NJ.

Beard, J.B., P.E. Rieke, A.J. Turgeon, and J.M. Vargas Jr. 1978. Annual bluegrass (Poa
‘ annua L) description, adaptation, culture and control. Res. Rpt. 352. Michigan State
Univ. Agric. Exp. Sta., East Lansing, MI.

Beard, J .B. 1982. Turf management for golf courses. Macmillan, New York.

Engel, RE. 1974. Influence of nitrogen fertilization on species dominance in turfgrass
mixtures. p. 104-111. In Proc. 2nd Int. Turfgrass Res. Conf, Blacksburg, VA, June 1973.
Amer. Soc. Agron. Crop Sci., Madison, WI.

Gaul, MC, and NE. Christians. 1988. Selective control of annual bluegrass in cool season
turfs femarimol and chlorsufirlfirron. Agron. J. 80:120-125.

Gaussoin, RE, RE. Branham, and PE. Rieke. 1985. Cultural factors in annual bluegrass and
creeping bentgrass competition. p. 35-37. In Proc. 55th Annual Michigan Turfgrass Conf.,
Lansing, MI, 21-23 Jan. 1985. Michigan State Univ., East Lansing, MI.

Gaussoin, RE. 1988. Species dominance in mixed creeping bentgrass annual bluegrass.
Ph.D. dissertation, Michigan State Univ., East Lansing.

Goss, R.L., S.E. Brauen, and SP. Orton. 1975. The effects of N, P, K and S on Poa annua
L. on bentgrass putting greens. J. Sport Turf Res. Inst. 51:74-82.

Koski, AJ. 1983. Seasonal rooting characteristics of five cool-season turfgrasses. MS thesis,
Ohio State Univ., Columbus.

Mahdi, Z., and VT. Stoutmeyer. 1953. A method of measurement of populations in dense
turf. Agron. J. 45:514-515.

Safiel, M.T., P.E. Rieke, and TA. Nikolai. 1992. Time domain reflectometry and irrigation
of turf. 62nd Ann. Mich. Turfgrass Conf. Proc. Vol. 21.

Saffel, MT. 1994. Time domain reflectometry and irrigation scheduling. MS Thesis,
Michigan State Univ., East Lansing.

Smucker, A.J.M., S.L. McBurney, and AK. Sritvastava. 1982. Quantitative separation of

roots from compacted soil profiles by hydronpneumatic elutriation system. Agron. J.
74:500-503.SAS Institute. 1994.

73

74

SAS user's guide: Statistics. SAS Inst. Inc., Cary, NC.

Topp, G.C., J .L. Davis, and AP. Arman. 1980. Electromagnetic determination of soil water
content using TDR: Applications to wetting fronts and steep gradients. Soil Sci. Soc. Am.
J. 46:672-678.

Vargas, J .M., Jr. 1994. Management of turfgrass diseases. Lewis Publishers, Ann Arbor, MI.

White, DB. 1989. Progress in developing improved annual bluegrass. 59'II Annual Michigan
Turfgrass Conference Proceedings. Michigan.

Wilson, 06., and J.M. Latham, Jr. 1969. Golf fairways, tees, and roughs. p. 584-602. In
A.A. Hanson and F .V. Juska (ed.) Turfgrass science, Agronomy No 14. Madison, WI.

Younger, V.B., A.W. Marsh, R.A. Shoham, V. A. Gilbeault, and S. Spaulding. 1981.
Water use and turf quality of warm-season and cool-season turfgrasses. p. 257-259. In
Proc. 4th Intl. Turfgrass Res. Conf.

CHAPTER 5

Bihourly Moisture Depletion Patterns in Fairway Turfs

Abstract

Syringing during hot afiemoons is a common practice in fine turf management.
However, timing criteria for syringing are mostly based on visual observations. Time domain
reflectometry (TDR) was used to monitor soil moisture depletion in bihourly time steps at
four depths under three irrigation regimes for established annual bluegrass and creeping
bentgrass fairway turfs. The objectives were: i) to establish the time of the day when soils
moisture depletion is highest under fairway turfs; and ii) compare the summation of bihourly
moisture depletion during daylight hours to daily adjusted ET. The irrigation treatments
were: i) apply 2.5 mm daily; ii) apply 25 mm upon the appearance of wilting stress; and iii)
return the soil to field capacity daily based on soil moisture depletion as measured by TDR.
This study was conducted at the Hancock Turfgrass Research Center at Michigan State
University to evaluate soil water dynamics in fairway turfs during daylight hours on three
different dates with contrasting evaporative demand in 1993 and 1994. Bihourly moisture
depletion by species and irrigation treatment were compared to weather based ET. Moisture
depletion varied by turf species, soil depth, irrigation treatment and time. Moisture depletion

patterns were also dependent on initial moisture content and evaporative demand. The

75

76

greatest change in moisture content was from the 0-5 cm depth for the field capacity and
2.5 mm daily irrigation treatments between 0900 and 1300 h confirming that syringing in the
early afiemoon may reduce moisture and temperature stress. Moisture depletion data as
measured by TDR for the irrigation treatments were generally more conservative than the
adjusted Penman estimate fiom a weather station on site. This suggests that potential water
savings could be made from soil moisture depletion-based turf irrigation scheduling.

However, agreement between moisture depletion and Penman estimates was poor.

Introduction

The competition for water and declining water resources call for more efficient water
management strategies in turf as in conventional agriculture. Improved water conservation
may delay the search for alternative water sources and reduce possible contamination of
surface and ground waters. One method of conserving water is through the development of
environmentally sound and eflicient irrigation scheduling programs.

Accurate application rates and timing are critical management decisions in turf
irrigation management. Several factors must be considered in developing timing criteria for
golf course irrigation scheduling. Among these are the prevention of wilting stress and
reduction of disease incidence (Vargas, 1994) and leaching of agricultural chemicals into
ground water (Augustin and Snyder, 1984). Another factor is interference with play or
cultural activities (Beard, 1973).

Few golf courses have automatic weather stations that provide site-specific

evapotranspiration (ET) data. Although such technology facilitates the decisions on how

77

much water to apply, it does not provide answers to the question of when to irrigate.
Furthermore, it is difficult for turf managers to justify the purchase of such expensive
equipment. For the most part, irrigation scheduling on golf courses is based on visual
observations of wilting stress, foot printing or as a security measure, light daily applications
(Vargas, 1994). Emcient irrigation scheduling remains a major challenge for most turf
managers as they strive to maintain a delicate balance between growing quality turf and
environmental concerns.

Syringing is commonly used to moderate canopy temperature during hot days.
However, studies on the effect of syringing on canopy temperature show variable results.
Dufi‘an Beard, (1966) reported a l to 2 °C reduction in temperature lasting 2 h following the
application of 6.4 mm of irrigation at 1200 h, while Hawes (1965) concluded that application
of 3 mm of irrigation between 1130 and 1500 h resulted in wow temperature reductions
between 0.8 to 4 °C for up to 600 8. On the other hand, Dipola (1984) found no significant
difference in canopy temperature 1 h after syringing regardless of timing or syringing amounts
over a certain range. Ideally, irrigation should be applied prior to the appearance visual signs
of stress (Beard, 1982). Soil moisture depletion monitoring could improve timing criteria for
turf irrigation (Carrow, 1991) and possibly for syringing on hot days.

Despite good correlation between mini-lysimeters and weather based ET estimates
(F eldhake et al., 1984) mini-lysimeters have been criticized for having soil-container interfaces
that interrupt upward and lateral water flow and for not being representative of field turf
conditions. Weather-based ET estimates fail to account for soil moisture status even when

adjustments are made by multiplying the potential ET by a turf or crop coefl'rcient as

78

suggested by Carrow (1991). Comparing various modifications of the Penman equation,
Allen (1986) reported 10-25% error compared to large weighing lysimeters.

Although models have been developed that estimate moisture depletion in hourly or
shorter time steps (Ritchie, 1990), field methods for soil moisture determination needed to
verify such models usually require longer waiting periods and thus fail to serve as effective
tools for model validation. For the gravimetric method, at least 24 h are needed to obtain
soil moisture data (Hanks, 1992). In addition to destructive sampling, this method is also
subject to errors from spatial variability (Hillel, 1980).

While neutron scattering requires considerably shorter waiting periods, the problems
of exposure to radiation and the need for routine calibration remain (Hillel, 1980). Also, the
neutron probe is not amenable to the shallow rooting depths encountered in turf due to errors
fi'om neutrons escaping to the surface (Snyder et al., 1984). Finally, the dependence of the
spatial resolution of the neutron probe on the degree of wetness is of concern to some
scientists (Arnon, 1992).

Moisture depletion measurements in soils integrate plant, soil and atmospheric
interactions into a single measurement. Kirsch (1993) used neutron scattering and gravimetric
measurements to locate the zero flux plane and estimate ET with errors of up to 153 %. He
acknowledged that techniques such as TDR may yield more reliable ET estimates.

In sand-based soils with low moisture retention, low-volume, high-fiequency irrigation
is required to supply adequate amounts of water for quality turf maintenance. This implies
that methods for rapid and accurate moisture determination are needed to provide adequate
and timely supplies of water to the turf, in order to reduce overwatering and the incidence of

wilting stress.

79

In the last decade, time domain reflectometry (TDR) has become an acceptable
method for soil moisture determination. This method is accurate, fast, and precise, and does
not involve radioactive risks associated with the neutron probe or gamma-attenuation
techniques (Topp et al., 1980; Dalton, 1992). Despite the potential benefits of TDR and
other soil moisture sensors, their use in real-time turf irrigation scheduling and modeling has
not been firlly explored.

The literature is replete with articles on the theoretical development and potential
agricultural applications of TDR. However, few studies have addressed the use of TDR in
real-time irrigation scheduling for field crops (Topp and Davis, 1985) or for turfgrass
irrigation (Carrow, 1991; Saffel, 1994). The use of TDR in turf irrigation scheduling may
result in more efficient water application rates and timing on greens and fairways and
eventually, on home lawns, when the technology becomes affordable.

This study evaluated bihourly moisture depletion patterns in annual bluegrass (Poa
amua L. var. reptans) or creeping bentgrass (Agrostr's palustris Huds. L.) fairway turfs under
three irrigation regimes. The specific objectives were: i) to monitor moisture depletion
patterns of annual bluegrass and Penncross creeping bentgrass fairway turfs under three

irrigation regimes and ii) to compare Penman ET to daily moisture depletion.

Materials and Methods
This study was conducted at the Hancock Turfgrass Research Center at East Lansing,
MI. The soil type was an Owosso sandy loam (fine-loamy, mixed Typic, Hapludalf) with an
average field capacity of 0.28 cm3 cm". Bihourly volumetric moisture content was determined

by TDR from 0700 to 2100 h on 13 Aug. 1993 and 16 Aug. 1994, and fi'om 0700 to 1900

80

on 13 Sept. 1994. From long-tenn weather forecasts, all three dates were expected to be
sunny, with high temperatures and high evaporative demand. In 1993, 198 readings were
taken bihourly, whereas only 66 readings were taken in 1994 due to some dysfirnctional
probes.

Pairs of TDR stainless steel rods 3.2 mm in diameter and 200 mm long, with 50 mm
center-to-center spacing, were installed at 2.5, 7.5, and 12.5 cm from the soil surface in three
replicates (Topp et al., 1980; Saffel, 1994). A fourth set of probes of similar configuration
was also installed, with one rod at 17.5 cm and the other at 22.5 cm in only two replicates due
to restrictions on the number of positions on the rotary switch used to move between sets of
probes. These probe pairs measured VMC in the 0-5, 5-10, 10-15, and 15-25 cm depths
respectively.

Three irrigation treatments were evaluated on Penncross creeping bentgrass or annual
bluegrass turfs. The irrigation treatments were: i) apply 2.5 mm daily (Vargas, 1994); ii)
return the soil to field capacity (0.28 cm3 cm") daily based on TDR readings; and iii) apply
25 mm only upon the appearance of wilting stress. Irrigation treatments served as whole
plots with dimensions 11 x 11 m in three replicates. Each irrigation plot was split in half (11
x 5.5 m) and planted at random to pure stands of either annual bluegrass (Poa annua L. var.
reptans) and Penncross creeping bentgrass (Agrostis palustris L. Huds.) established in 1989
(Saffel, 1994). Both species were maintained according to standard management practices
for cool-season fairway turfs in Michigan with a cutting height of 16 mm.

Evapotranspiration data for each date were recorded fi'om a Rainbird automatic
weather station at the site 2 m above the ground. Soil moisture depletion (mm) for each

depth was calculated from the equation of Arya et al. (1975):

81

(VMCn ,1 - VMC“) * depth of soil layer (m)

where n is initial VMC reading and n+1 represents successive VMC readings.

The experimental layout was a 3*2*4 factorial split-strip design with three replicates.
Time domain reflectometry installations at four depths served as the strip. Bihourly VMC
data were analyzed using repeated-measures analysis of the general linear model and Student

Neuman-Keuls (SNK ) mean separation procedures (SAS Institute, Inc., 1994).

Results and Discussion

Solar radiation, minimum and maximum temperatures values for the three study dates
are given in Fig. 5.1. Temperatures from 0600 to 1700 h were similar for 13 Aug. 1993 and
16 Aug. 1994. However, minimum and maximum temperatures on 13 Sept. 1994 started
about seven degrees lower than on the other days. Temperatures increased rapidly from 07 00
to 1100 h, with similar slopes on all dates. After 1100 h, the rate of increase declined and
temperatures remained fairly constant until 1900 when maximum temperature peaked again
for all dates, although this was more pronounced on 13 Aug. 1993. Solar radiation trends
were about the same on all days, except that changes between 1600 and 1700 h were more
drastic on 13 Sept. 1994 (Fig. 5.1). This date also had shorter daylight hours.

Bihourly relative humidity and wind velocity for study dates are presented in Fig. 5.2.
Relative humidity ranged fi'om 85% (between 0600 and 0900 h) to a low of 60% at 1600 h
on 13 Aug. 1993 and 16 Aug. 1994 but rose back to 85% and 75%, respectively. Relative
humidity on 13 Sept. 1994 was significantly lower than on other dates, with a maximum of

less than 60% early in the morning to a low of 25% at 1800 h.

82

 

Temperature, C
N
Lit

 

   

Solar radiation, MJ/n12/l1r

  

 

. f

..00
2111315171921

A

 

 

10

 

' 161112 1471611; ‘21)
Time, hours

—---TMIN -.-.TMAX—v—SRAD

Fig. 5.1. Bihourly minimum temperature (TMIN), maximum temperature (TMAX) and
solar radiation (SRAD), for 13 August 1993, 16 August 1994 and 13 September
1994.
Wind velocities for both dates in August were comparable and were much higher than
those on 16 Sept. 1994. Turfgrass leaves were wet early in the morning either from dew,

guttation or irrigation. By 1100, most of the moisture had either infiltrated into the soil or

evaporated as solar radiation and temperature increased.

Relative humidity, %

83

 

 

 

 

 

100 g l _ 34.0
i l " "- ."x.
z I _/ ... 1
80 .L‘r-\.-------., ....... ' ... \ \
l '\ 'l L l: 'I\ 1‘ - / 'H 13.0
i . .A,’ \ ,' an ' m
60 l I-\\" i 'x I! " \x‘jf\ 'F-I {-1 i E
7"”: ' """" """"" "“12" ““““ >2
1 1 \\ "- ll 1 \ I l ' E
T '\ * 1' ' ‘ I T20 8
i I l .i '. I! 1 : >
4011\‘yflil """""""""" r-"r'jfi- ''''''' '_;°1 E
.l '- .1. l I, 1‘13: 3
l 5 ll ' -' 1 110
1- I 'l ' I ~ 1 '
20 Th """"""""" 1r ------------- ’t-jr’r-""""'"";"i
i J. 8-13—93 1; 3-16-94 ~. Ex 9-13-94 1 i
l \.
0 l i’ \- 11L
7 '9' ‘11'13'1‘5‘17'19’2'1'8 ‘11)‘12‘14'16'18‘20'7 193 1117113 171921
Time,hours
_...WN_._RH

Fig. 5.2. Bihourly relative humidity (RH) and wind velocity (WV) for 13 August 1993,
16 August 1994 and 13 September 1994

The analysis of variance (Table 5.1) showed that irrigation treatment (TRT) and the

TRT’DEP (sampling depth) interaction significantly affected VMC for all sampling dates

while species (SPE) and sampling depth (DEP) were significant only on the first two dates.

The efi‘ect of species on VMC was significant on 13 Aug. 1993 and 16 Aug. 1994, but not

84

at all intervals on 13 Sept. 1994. The significance of the TRT‘DEP interaction was due to

inherent moisture-retention capacities for the various soil depths as shown earlier (Fig.3 .6).

Table 5.1. Analysis of variance for the test of hypotheses for between- (among-)
subject effects for bihourly volumetric moisture content.

 

 

 

 

13 Aug. 1993 16 Aug. 1994 13 Sept. 1994
Source

df MS df MS df MS
REP 2 0.003 2 0.001 2 0.0003
TRT 2 2013*" 2 0029*“ 2 0.584’”
REP’TRT 4 0.003 4 0.003“ 4 0.007
SPE 1 0.046‘" 1 0.005‘I 1 0.004
TRT’SPE 2 0.013“ 2 0.001 2 0.011
REP’SPE*(TRT) 6 0.004 6 0.003 6 0.002
DEP 3 0.015" 3 0022*" 3 0.004
TRT'DEP 6 0043‘" 6 0.003‘ 6 0.016“
SPE‘DEP 3 0.001 3 0.001 3 0.001
TRT’SPE‘DEP 6 0.002 6 0.002 6 0.002
Error 162 0.003 30 0.001 30 0.003

 

*,**,***Significant at the P = .05, .01, and .001 levels, respectively.
REP, DEP and SPE are replicate, depth and species respectively.

The repeated measures procedure (SAS Institute Inc., 1994) was used to analyze the
data since multiple observations were taken per experimental unit at different times. The
analysis of variance for repeated measures showed a time dependent variation of VMC during
the course of the day as expected (Table 5.2). Variations in solar radiation, temperature,

relative humidity and wind speed with time resulted in variable moisture redistribution within

85

the soil profile for each sampling date. The TIME*DEP and TIME*TRT interactions were

significant (p > 0.01) on all sampling dates.

Table 5.2. Repeated-measures analysis of variance for the test of hypotheses
for VMC among subject effects.

 

 

 

13 Aug. 1993 16 Aug. 1994 13 Sept. 1994
Source '
df MS df MS df MS

TIME 7 0.0030"I 7 0.0019“"""I 7 0.0009“"'
TIME*REP 14 0.0001 14 0.0017 14 0.0001"
TIME‘TRT 14 0.0001 14 0.0002 14 0.0001
T1ME*REP*TRT 28 0.0001 28 0.0002 24 0.0001
TIME*SPE 7 0.0001 7 0.0001 7 0.0001
TlME‘TRT‘SPE 14 0.0002‘I 14 0.0003 14 0.0001
TIME*REP"‘SPE*(TRT) 42 0.0001 42 0.0002 42 0.0001
TIME‘DEP 21 0.0003*** 21 0.0004" 21 0.0001"I
TIME‘TRT‘DEP 42 0002* 42 0.003" 42 00001"
TIME‘SPE‘DEP 21 0.0039 21 0.0001 21 0.0000
TIME*TRT"SPE*DEP 42 0.0043 42 0.0002 42 0.0000
Error (TIME) 1134 0.0135 210 0.0002 180 0.0000

 

*,",***Significant at the P = .05, .01, and .001 levels, respectively.
REP, DEP and SPE are replicate, depth and species respectively.

Bihourly volumetric moisture content (VMC) changes, however, were not greater
than the measurement error for TDR and thus fail to provide conclusive evidence of soil
water movement. Except for the stress treatment on 16 Aug. 1994, all maximum VMC
values for the irrigation treatments were recorded at 0900 h for each date. All minimum

VMC values were recorded at 2100 h on 13 Aug. 1993. On 16 Aug. 1994 and 13 Sept.

86

1994, minimum VMC values were recorded either at 1700 or at 1900 h for all treatments.
There was no significant difference between the 2.5 mm daily and the field capacity
treatments, both of which were significantly higher than the stress treatment.

Initial VMC by depth for 13 Aug. 1993 (Fig. 5.3) and 16 Aug. 1994 (Fig. 5.4) ranked
0-5 > 5-10 > 10-15 > 15-25 cm depths. Graphs for moisture depletion with time for the 5-10
and 10-15 cm depths for field capacity treatment on 13 Aug. 1993 (Fig. 5.3) were about the
same and followed the same moisture depletion pattern throughout the period of data
collection. However, variations in VMC trends occurred with time for most other depths and
treatments on both dates. For example, both annual bluegrass and creeping bentgrass
showed a decrease in VMC for the 0-5 cm depth from 1700 to 1900 h, with a corresponding
increase in VMC for the 15-25 cm depth for the stress treatment.

Moisture depletion from the 2.5 mm daily treatment on 13 Aug. 1993 showed
contrasting trends between species and among depths (Fig. 5.3). At 0700 h, VMC by depth
was in the order 0-5 > 5-10 > 10-15>15-25 cm, as expected since water wets the surface
during irrigation. By 0900 h, the 0-5 and 5-10 cm depth for the bluegrass species had the
same VMC. The greatest moisture depletion in the bluegrass plots was evident between 0700
and 0900 h for the 0-5 cm depth. While the high root density for this depth may explain
greater water depletion there was no logical explanation for the timing since temperature and
solar radiation were still increasing. Similar depletion for the bentgrass species was between
0900 and 1300 h (Fig. 5.3) suggesting a difference in timing between annual bluegrass and
creeping bentgrass response to changes in environmental conditions.

Generally, the 0-5 cm depth for the irrigated treatments had significantly higher

moisture content than all other depths. This may have been due to higher organic matter

Moisture content (VMC). cm3/cm3

Moisture content (VMC), cm3/cm3

Moisture content (VMC), cm3/cm3

040

w-r-ur- v- ._..I 7'

 

0.35 l

2.5 mm daily, bent

 

 

O35

 

 

Field capacity, bent I

 

 

0.25

 

 

0.20 a

0.15 4

Stress. bent

 

 

 

 

F

a- 0-5cm -.- S-lOCm .... lO-lScm «— lS-25cm

ig. 5.3.

87

Moisture content (V MC), cm 3/cm3

Moisture content (VMC), cmB/cm3

Moisture content (VMC), cm3/cm3

0.35

025

 

l 2.5 mm daily, poa

................................

 

 

 

Field capacity, poa

 

 

 

 

 

 

Stress, poa

 

 

7 9 11 .13 13 17 1b 21
Time (hours)

+ 0-5cm + 5-10cm —.- 10-15cm -- 15-25cm

Bihourly moisture depletion patterns for annual bluegrass and Penncross

creeping bentgrass by irrigation treatment 13 August 1993.

 

88

content or it may indicate that TDR detects water in root tissues (about 70% of the root
mass was in the 0-5 cm depth). Moisture depletion from the irrigated treatments was higher
than from the stress treatment because VMC was lower for the stress treatment. The 15-25
cm depth for the stress treatment had the highest VMC levels until 1300 h. Whereas this
trend persisted for the bluegrass species, VMC for the bentgrass species (0-5 cm depth)
increased rapidly between 1100 and 1300 h.

Symmetrical moisture depletion and accretion patterns between depths suggest
preferential as well as differential moisture uptake and redistribution by species for the
different depths. The 15-25 cm depth for the bentgrass species showed the greatest increase
in soil moisture content for both species on 13 Aug. 1993 in the stress treatment (Fig. 5.3).
Moisture depletion for the bentgrass species showed similar patterns for all depths except the
15-25 cm depth (Fig. 5.3). Between 1300 and 1500 h, VMC for the 0-5 and 5-10 cm depths
increased slightly. This increase may have been due to hydraulic lifting of water by the roots,
coinciding with partial closing of the stomata (Jensen and Taylor, 1971). From 1700 h to the
end of the day most depths showed only a minor depletion in moisture as solar radiation and
temperature had dropped substantially.

On 13 Aug. 1993, volumetric moisture content means across treatments and depths
for the bentgrass species were significantly higher than for the annual bluegrass for all
sampling time intervals (Fig. 5.3) but not on the other two dates. All treatments showed an
increase in VMC fi'om 0700 to 0900 h except the stress treatment on 16 Aug. 1994 and the
field capacity and stress treatments on 13 Sept. 1994. Since no irrigation was applied during
this time course experiment significant increases in VMC suggest net water movement from

one depth to another.

Moisture content (VMC) curl/cm)

p
as
o

.0
or
on

Moisture content (VMC) cm3/cm3
p o
N (.0
cn o

.0
N
0

Moisture content (V MC) cm3/cm3
.0
or
tn

.0
N
o

 

 

2.5 mm daily, bent

 

 

 

7 -9 11 1315 1719 21

 

 

Field capacity, bent

 

 

 

7 9111315171921

 

.0
T5

P
N
or

 

Stress, bent

 

 

 

7 8111315171921

Time, hours

a-O-S crn + S-IOcm *10-15 cm... 15-25 cm

89

Moisture content (V MC) cm3/cm3

Moisture content (VMC) cm3/cm3

Moisture content (VMC) cm3/cm3

0.40

 

0.35 1

 

2.5 mm daily, poa

 

 

 

 

 

 

 

 

 

0'20 7 9 11 13 15 17 19 21
0.40
Field capacity, poa
0.35-
0.30‘ /\—/\\—-—.
\A- — \W‘
M
0.25- \——‘\—_\/
°'2° 7 9 11 13 15 17 19 21
0.35

 

0.30 -

.°
N
or

 

.0
N
o

 

 

 

 

19111315171921

Time, hours

+0.56!!! +5-10cm *IO-ISm+ 15-25ern

Fig. 5.4. Bihourly moisture depletion patterns for annual bluegrass and Penncross
creeping bentgrass by irrigation treatment, 16 August 1994.

90

The mechanisms governing upward flow observed in this study were not investigated but
deserve fiirther study. Upward flow of water may be due to hydraulic lift by roots (Murphy
et a1. 1994), capillary rise (Hillel, 1980) or vapor diffusion if there is a high temperature
gradient. Beard (1973) stated that dew may contribute as much as 0.3 to 0.4 mm a night
under favorable conditions. While dew formation or guttation may explain the increase in
VMC in the late afiemoons, it is difficult to discard their contribution to increases in VMC
during early morning hours. For the most part, changes in VMC were again within the
margin of error of TDR.

Although moisture changes in soil varied by depth, species, and irrigation treatment,
there was no clear-cut trend for moisture depletion by irrigation treatment, species, or depth.
Symmetry between moisture accretion and depletion patterns between the 0-5 and 5-10 cm
depths for annual bluegrass in the stress treatment suggests a net flux of water from one depth
to another over certainn time intervals (Fig. 5.3). For the bentgrass species, the less
pronounced symmetry between the 0-5 and the 15-25 cm depths fiom 0700 and 1300 h may
be the effect of deeper roots used in water uptake causing a net flux of water from the 15-25
cm depth to the 0-5 cm depth. Between 1700 and 2100 h, VMC was fairly constant for both
species. The greatest moisture depletion was from the 0-5 cm depth of the daily treatment
between 0900 to 1300 h for the bluegrass species on August 16, 1994. Moisture content for
the 0-5 cm depth for the bentgrass species decreased steadily till 1700 h. (Fig. 5.4).

Moisture depletion patterns for the 2.5 mm daily treatment on 16 Aug. 1994 were
characterized by many interactions among depths (Fig. 5.4). The increase in VMC observed
in the bluegrass species for the 2.5 mm daily treatment for all depths from 0700 to 0900 h was

seen only in the 15-25 cm depth for the bentgrass species.

 

 

2.5mmdaily,poa
7 9 11 13 15 17 19

 

 

 

0 7 4 1. 5
3 2. 2. 2. 1. 1.
0 0 0 0 0 0
AmEu3Euv E8..OU 057.22 252:...

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
_
.
w

5

2

0
.5363 Eacou 8:302 030...... >

O

>

91

 

19

T7

   

13 15

2.5 mm daily, bent

1'1

 

 

.p———--~--m—-pc—-

 

0.30
0.20

5
1
o

A

n

Time (hours)

Time (hours)

7 .

 

 

 

 

 

 

. .
_ _ m.
. .
_ .
. _ 7
_ _ 4|
. _
I . .
w _ _ «.1
Wu _ . F.
.1 . . W
a _ _ h
w 3(
. . at C
d . .
m _ . .m
F . - v1 T
_ . 1
. .
_ fi _
. . .9
_ .
. .
. _
. . 7
w w
5 o 5 0
3 3. 2. 2.
0 0 0 0
AMEQMEUV “COEOU 953.02 330530"
9
4|
7
(I

15

Field capacity, bent
1 3

 

 

 

1
1
9
7
5 2 9 6 3 0
3. 3. 2. 2 2. 2.
O O O 0 0 0
2.53.5. 32.80 2:33: 030520.»

 

   

 

 

 

. _
. 9
1
.
_
. 7
. 4|
.
_ .
5
_ 1 m
. u
. to
W . mm (\
9 - m
m _ 1
s 1 T
1
9
_
_ .
. . 7
a .
5 O 5 0
2 2. J 1..
0 O 0 0
3.5353 2.2.50 8:382 032...: >

 

13 15 17 19
Time (hours)

Stress. Beat

11

  

 

7

 

 

 

.
_
n
0
2.
0
E9388 2.2.25 0530.2 250...... >

 

0.25

o
1
0

A

M

Fig. 5.5. Bihourly moisture depletion patterns for annual bluegrass Penncross creeping
bentgrass by irrigation treatment 13 Septemberr 1994.

92

Initial moisture contents for the 0-5 cm depth for the field capacity treatment for both
species were about the same at 0700 h and increased until 0900, but decreased between 0900
and 1100 h (Fig. 5.4). Although the largest decrease in VMC for the 15-25 cm depth
occurred between 0700 and 0900 for bentgrass, the major decrease at this depth for the
bluegrass species was between 1700 and 1900. As temperature and solar radiation had
dropped significantly this decrease could either be due to rapid uptake of water following the
opening of the stomata or due to TDR error as the computer battery was running low.

The stress treatment showed the most variation in VMC on 16 Aug. 1994. Initial
VMC for the 0-5 cm depth for both species was about the same at 0700 h. For both species,
there was no increase in moisture content in this treatment between 0700 and 0900 h on this
date. However, VMC for the bluegrass species decreased until 1100 h, but the decrease in
bentgrass species lasted until 0900 h. Symmetrical moisture depletion patterns were more
obvious in the bluegrass species than in the bentgrass species. The deeper rooting depth of
bentgrass compared to annual bluegrass may have contributed to more uniform water uptake
along the soil profile. The largest increase in VMC was observed in the bentgrass species at
1700 h for all depths; only the 10-15 and 15-25 cm depths showed a corresponding increase
between 1500 and 1900 h.

On 13 Sept. 1994, the 2.5 mm daily treatment showed similar moisture distribution
patterns with respect to initial moisture content by depth at 0700 and 1300 h (Fig. 5.5).
Whereas maximum depletion for the bluegrass species occurred at 1700 h, maximum
depletion for the bentgrass species occurred at 1500 h. The large decrease in VMC observed

toward the end of the day may have been due to TDR error.

93

On 13 Sept. 1994, cloudy skies and shorter day length accounted for low evaporative
demand. The 15-25 cm depth had the lowest VMC, whereas the 0-5 cm depth had the
highest VMC levels for both species in this treatment. However, by the end of the day,
moisture levels for the 5-10, 10-15, and 15-25 cm depths were about the same (Fig. 5.5).
This suggests differential moisture extraction patterns by depth for both species. Different
rooting habits for annual bluegrass and creeping bentgrass during the heat of the summer
(Koski, 1983; Murphy et al., 1994) may be a possible explanation for this. Also, Saffel (1994)

suggested differential water uptake by depth for annual bluegrass.

Table 5.3. Comparison of soil moisture depletion (mm) across species by irrigation treatment
and daily adjusted ET from a weather station at the Hancock Turf Research Center,

 

 

 

 

 

Michigan State University.
Irrigation Treatment
Date DLY F C STR ET (adjusted)
mm
13 Aug. 1993 3.0 (46) 3.0 (46) 2.0 (64) 5.6
16 Aug. 1994 3.5 (29) 3.8 (22) 3.3 (32) 4.9
13 Sept. 1994 3.3 (32) 2.8 (42) 2.3 (52) 4.8

 

T Values in parentheses reflect potential percent water savings for each irrigation
treatment by date compared to the adjusted ET from a Rainbird weather station on
site

On 13 Sept. 1994, all four depths for the annual bluegrass species in the 2.5 mm daily
treatment had different initial moisture content (Fig. 5.5). Initial VMC for the bentgrass
species showed only two distinct moisture levels. Only minor depletions were observed for

the bentgrass species. This is probably due to the high soil moisture content in this treatment

on this date. For the bentgrass species, the 5-10 and 10-15 cm depths had identical moisture

94

depletion patterns throughout the day, increasing only from 0700 to 0900 h. Moisture
depletion occurred mostly in early afiemoon. Differences in moisture depletion patterns
between species may be attributed to preferential moisture uptake by species and/or
difl‘erential rooting patterns by depth (Safi‘el, 1994). This suggests that moisture depletion
in response to evaporative demand is moderated by turf species and hence the need for
site/cop-specific turf coefficients.

On 13 Sept. 1994, VMC differences by depth for the bentgrass species were not as
great as those for the bluegrass species. For both species, moisture content for the 0-5 and
5-10 cm depths increased from 0700 to 0900 h. The 15-25 cm depth showed very little
variation in the bluegrass plots, whereas all other depths showed a gradual decrease in VMC
through the course of the study.

Initial water content for the bentgrass species was higher than for the bluegrass
species in the 2.5 mm daily treatment. Changes in VMC for the bluegrass species observed
in the 2.5 mm daily and field capacity treatments were lowest between 1100 and 1300 h (Fig.
5.5). In particular, the 2.5 mm daily treatment showed only minimal gains or losses in soil
moisture content for the depths of interest over this interval, even though this coincided with
peak values for solar radiation, minimum relative humidity and high temperatures, conditions
that favor high evaporative demand. Possible explanations for these observations are: i) the
stomata closed to the extent that plant water uptake was about the same as the
evapotranspiration thus maintaining a steady VMC, and ii) the shallow root system of annual
bluegrass may not be as emcient in water uptake as that of creeping bentgrass.

There was no distinct pattern to the data with respect to irrigation treatment or

species. This suggest that the different species show variable response to different levels of

95

evaporative demand. This variation may be based on amount of available water, root
distribution within the profile, and time of the day. While bentgrass species have been shown
to have higher and deeper root mass than annual bluegrass (Murphy et al., 1994; Saffel,
1994). It must be cautioned that higher root mass may not necessarily imply higher water and
nutrient uptake efliciency.

Moisture depletion as measured by TDR ranked FC > DLY > STR for all sampling
dates because dates selected for bihourly moisture depletion studies coincided with dry
periods. Although the Penman ET recorded on August 13, 1993 was highest, the moisture
depletion data were highest on August 16, 1994 for all treatments. This illustrates the
discrepancies in comparisons between ET and soil moisture depletion and underscores the
complex nature of the interactions between plant species and soil under different evaporative
demand. It also illustrates the need for conjunctive use of weather based ET and soil moisture
depletion measurements in irrigation scheduling.

Moisture depletion data shown on (Table 5.3) were more conservative than the
adjusted ET from the modified Penman equation for all dates and treatments probably due to
upward water movement. The percent difference between soil moisture depletion and
weather based ET (shown in parenthesis in Table 5.3) indicate potential water savings of up
to 46 % for the irrigated treatments from soil moisture depletion based irrigation scheduling
compared to the adjusted Penman ET. These results show less variation than those of Kirsch
(1993) and confirm his prediction that instruments with better resolution like TDR could

improve assessment of plant water use.

Conclusions

The above results show that TDR could be used as a tool for soil moisture monitoring
and irrigation scheduling even in turf ecosystems. However, bihourly VMC changes were
within the margin of error of TDR measurements and thus fail to provide conclusive
information on moisture depletion patterns.

Evidence of upward flow at different times under different conditions, even under the
shallow rooting depths for turf species suggest that water drained below the root zone may
again become available to the plant following moisture redistribution. Moisture depletion
measurements such as these reflect actual VMC changes in response to variable irrigation
treatments, plant species and evaporative demand.

Time domain reflectometry estimates of moisture depletion were more conservative
than weather based ET estimates even after adjustments have been made. Water savings of
up to 46 % could be achieved from moisture depletion based irrigation scheduling compared
to weather based estimates. Despite widespread efforts to equate moisture depletion to ET,
it should be noted that weather based ET estimates fail to account for upward flow of water
to replenish water extracted from the root zone. This may explain in part the difl‘erence
between ET and TDR estimates. When it becomes affordable, widespread use of this soil
moisture sensor could reduce overwatering and the potential for leaching of agricultural
chemicals into ground water.

Moisture depletion patterns were highly variable between annual bluegrass and
creeping bentgrass during the course of the day although daily means of VMC by species

show no significant differences. Bihourly moisture patterns show three major trends: they

96

97

upward water movement, water from guttation or infiltration of dew during morning hours.

Moisture depletion from transpiration or soil evaporation in response to evaporative
demand also contributed to observed differences. Stomatal regulation during hot periods may
account for reduced water loss by turf leaves as turfgrass strive to maintain osmotic balance
in the tissues under conditions of high evaporative demand. Generally, the greatest moisture
depletion occurred between 0900 and 1300 h suggesting that syringing after this period may
be optimum timing to minimize heat and moisture stress. Periods that show no moisture

change may reflect steady states between moisture loss and gains.

Literature Cited

Allen, RG. 1986. A Penman for All Seasons: Amer. Soc. of Civ. Eng, J. of the Irrig. and
Drainage Division. 112:348-368.

Arnon I. 1992. Agriculture in dry lands, principles and practices. In development in
agricultural and managed-forest ecology. Elsevier, New York.

Augustin, B.J., and GB. Snyder. 1984. Moisture sensor-controlled irrigation for maintaining
bermudagrass turf. Agron. J. 76:848-850.

Beard, J .B. 1973. Turfgrass Science and Culture. Prentice-Hall, Englewood Clifl‘s, NJ.

Beard. J. D. 1982. Turf management for golf courses. Macmillan, New York.

Carrow, RN. 1991. Turfgrass water use, drought resistance and rooting patterns in the
Southeast. Tech. Compl. Report USDI/USGS Project 02. ERC 01-91. Univ. of Georgia,
Athens, GA.

Dalton, EN. 1992. Development of time domain reflectometry for measuring soil water
content and bulk electrical conductivity. In Advances in measurement of soil physical

properties: Bringing theory into practice. SSSA Spec. Publ. no. 30.

Dipola, J. M. 1984. Syringing effects on canopy temperatures of bentgrass greens. Agron.
J. 76:951-953.

Dufl; DI, and J .B. Beard. 1966. Effect of air movement and syringing on the microclimate
of bentgrass turf. Agron. J. 58:495-497.

Feldhake, C. M. R. E. Danielson, and J .D. Butler. 1984. Turfgrass evapotranspiration. 11.
Responses to deficit irrigation. Agron. J. 6:85-89.

Jensen, RD, and SA. Taylor. 1971. Effect of temperature on water transport through plants.
Plant Physiol. 36:639-642.

Hawes, D. T. 1965. Studies of the grth of Poa annua as affected by soil temperature, and
observations of soil temperature under putting green turf. M.S. thesis Cornell Univ.,
Ithaca, NY.

Hillel, D. 1980. Applications of soil physics. Academic Press, New York.

98

99

Kirsch, SW. 1993. Field test of a soil-based measure of evapotranspiration. Soil Sci. 93 :204-
207.

Koski, AJ. 1983. Seasonal rooting characteristics of five cool season turfgrasses. MS thesis,
Ohio State Univ., Columbus, OH.

Murphy, J.A., M.G. Hendricks, P.E. Rieke, A.J.M. Smucker, and BE. Branham. 1994.
turfgrass root systems evaluated using minirhizotron and video recording method.

Agron. J. 86:247-250.

Saffel, MT. 1994. Time domain reflectometry for turfgrass irrigation scheduling. MS Thesis,
Michigan State Univ., East Lansing, MI.

SAS Institute Inc. 1994. SAS user's guide: Statistics 1994 ed. SAS Inst. Inc. Cary NC.

Snyder, G.H., B.J. Augustin, and J .M. Davidson. 1984. Moisture sensor-controlled irrigation
for reducing N leaching in bermudagrass. Turf. Agr. J. 762964-969.

Topp, G.C., and J .L. Davis. 1985. Measurement of soil water content using time domain
reflectometry (TDR): A field evaluation. Soil Sci. Amer. J. 49:19-24.

Topp, G.C., J .L. Davis, and AP. Arman. 1980. Electromagnetic determination of soil water
content: Measurements in coaxial transmission lines. Water Resour. Res. 16:574-5 82.

Vargas, J .M. 1994. Management of turfgrass diseases. Lewis Publishers, Ann Arbor, MI.

CHAPTER 6

Adaptation of a New Water Balance Model for Turfgrass
Under Variable Irrigation Regimes

Abstract

The water balance routine of the System Approach to Land Use Sustainability (SALUS)
model was modified to simulate water balance in fairway turfs under three irrigation regimes
at the Hancock Turf Research Center at Michigan State University. The irrigation treatments
were: i) apply 2.5 mm daily; ii) maintain soil at field capacity; and iii) apply 2.5 mm only
upon the appearance of wilting stress. Model components included subroutines for estimating
infiltration, drainage, soil evaporation, transpiration, evapotranspiration, and soil volumetric
moisture content (VMC) fiom soil, plant, and weather input files. Model modifications
included the assumption of a fixed leaf area index and root length density through the growing
season. Field measurements of VMC by depth were obtained for up to 60 days each year
using time domain reflectometry (TDR) during the summers of 1992 through 1994.
Evapotranspiration estimates were within the expected ranges for the site. The TDR
measurements fell within the range of simulated VMC values. However, agreement between
model simulations and field observations ranged from 0.12 < R2 > 0.29 for the irrigated

treatments. The best agreement between observed and simulated VMC was with the stress

100

101

treatment for all depths (0.39 < R2 > 0.47) suggesting that poor application uniformity may
account for some of the variability. More turf-specific data would be needed to improve

agreement between field VMC observations and model output.

Introduction

Turfgrasses continually are subjected to biochemical and physical adjustments in response
to changing environmental conditions or managerial inputs. These changes result either from
natural inputs such as rainfall, solar radiation, wind speed, temperature, and relative humidity
or from managerial inputs such as traffic, irrigation, fertilizers, pesticides, mowing, and other
cultural practices. For example, turfgrasses grown on sandy soils can lose most of the
available soil water over a very short period.

Available water content (AWC) is widely used as the basis for scheduling irrigation, yet
inexpensive, rapid, accurate, repetitive, and noninvasive methods of soil moisture
determination are still lacking. Accurate VMC measurements are needed to improve
irrigation scheduling for quality turf maintenance without compromising environmental
quality.

Improvements in wata balance models and computer simulation techniques can be used
to improve water management strategies in urban agriculture. The need for predictive models
such as the System Approach to Sustainable Land Use (SALUS) for soil water balance
studies (Ritchie, 1991) stems fiom the lack of fast, safe, accurate, and affordable methods for
determining volumetric moisture content (VMC). Such models are invaluable in estimating
soil moisture status within the root zone for irrigation scheduling and for reducing the

leaching of agricultural chemicals into ground water.

102
The SALUS model integrates soil, plant, and climatological inputs using either physically
based or empirically derived relationships between system components to calculate the desired
outputs. This study focuses mainly on water inputs, storage, and outputs as influenced by
environmental conditions, plant factors, and managerial inputs, conveniently described by the

hydrologic balance equation (Ritchie, 1981):
WC, = VMC,_, + P, + I, - R, — D, — ET,

where VMC, is the volumetric moisture content at time t, P. is precipitation, L is irrigation,
R, is runoff, Dt is drainage, and ET‘ is evapotranspiration for a given time interval.

Operational estimates of the range of AWC are based on laboratory determination of soil
water content. Ritchie and Amato (1990) summarized some of the criticisms of this
approach: i) drainage water may be available to plants; ii) 1500 kPa may not represent the
lowest potential of moisture extraction by plants; iii) moisture extraction is highly dependent
on root density; and iv) the effects of spatial and temporal variability of soil moisture are
largely ignored.

The water balance models of Richardson and Ritchie (1973) and Skaggs (1978) were
among the earliest dynamic water balance models. Innovations in computer technology have
increased the number of user-fiiendly models. Despite widespread applications of water
balance models in crop studies, their use in the study of turfgrass ecosystems has been limited.
Ritchie et al. (1991) cited the lack of accommodation for spatial and temporal variability in
the factors used to predict plant performance as a deficiency in the usefulness of such models

for agrotechnology transfer.

103

Time domain reflectometry provides the convenience of rapid and accurate moisture
determination in space and time (Topp and Davis, 1980; Dalton, 1992). Carrow (1991) and
Saffel (1994) demonstrated the successful use of TDR for VMC determination within
turfgrass root zones. This could provide a database for the verification of water balance
models in turfgrass ecosystems.

The objectives of this study were: i) to simulate the water balance components of a
turfgrass ecosystem, using the new water balance subroutine of the SALUS model; ii) to
compare volumetric moisture content under different irrigation regimes; and iii) to compare

modeled volumetric soil moisture content to TDR data measured in the field.

Materials and Methods

This study was conducted at the Hancock Turfgrass Research Center at Michigan State
University. Three irrigation treatments provided contrasting moisture inputs for established
annual bluegrass (Poa annua, L. var. reptans) and Penncross creeping bentgrass (Agrostis
palustris Huds. L.) fairway turfs mowed three times a week at a cutting height of 16 mm.
Irrigation treatments were: i) return the soil to field capacity (FC); ii) apply 2.5 mm daily
(DLY) (Vargas, 1994); and iii) apply 25 mm only upon the appearance of wilting stress
(STR). There were three replications for each treatment. A fourth irrigation management
scenario, the effects of applying 2.5 mm every other day (EOD), also was simulated. There
were no corresponding field observations for this treatment.

Irrigation plots had a slope of about 1.5%, with dimensions 11 m x 11 m. Each plot was
split (11 x 5.5) and randomly seeded to annual bluegrass or creeping bentgrass. Four pop-up

Rainbird irrigation heads were located at each comer of the plots, with average flow rate of

104

21.5 L per minute. Irrigation clocks were set to apply water at 03 00 h under the assumption

that low wind velocities at that time would be less likely to reduce application uniformity.

Table 6.1. Soil physical properties for SALUS model.

 

Depth LL DUL SAT INISW RWCON KSMAC KSMTX
(cm) cm3 cm'3 cm day-l

 

2.0 0.12 0.33 0.38 0.33 0.25 40.0 3.1
5.0 0.11 0.29 0.32 0.29 0.25 53.0 3.0
8.0 0.11 0.28 0.32 0.28 0.25 55.0 2.9
11.0 0.10 0.27 0.32 0.27 0.25 57.5 2.8
14.0 0.10 0.27 0.32 0.27 0.15 75.5 2.5
17.0 0.10 0.27 0.32 0.27 0.12 74.1 2.3
20.0 0.10 0.27 0.32 0.27 0.11 80.0 2.0
23.0 0.10 0.27 0.32 0.27 0.11 80.0 1.8

 

LL is drained lower limit, DUL is drained upper limit, SAT is moisture

content at saturation, INTSW is initial soil moisture content, RWCON is

root constant, KSMAC is saturated macropore hydraulic conductivity,

and KSMTX is matrix hydraulic conductivity.

Field volumetric soil moisture was determined by time domain reflectometry (TDR).
Pairs of stainless steel probes, 3.2 mm in diameter and 20 cm long, were placed horizontally
at 2.5, 7.5, and 12.5 cm for the first three pairs of sensors and at 17.5 and 22.5 cm for the
fourth pair of sensors. These represent depths of 0-5, 5-10,10-15 and 15-25 cm. The depths
did not correspond to the soil-depth increments in the SALUS model as shown in Table 6.1.

Weighted VMC averages from the water balance simulations were calculated to correspond

to the field installations and were used in validating the model.

105

The soil type was a modified Owosso sandy loam (Fine mixed mesic Typic Hapludalt).
Initial soil hydraulic properties were taken fi'om taxonomic properties of an Owosso sandy
loam, as described in the Ingharn County Soil Survey. Weighted means of the volumetric soil
moisture content by depth from the model were calculated to yield VMC values that
correspond to the depth increments of the TDR installations. The soil file included the wilting
point or drained lower limit (LL), the drained upper limit (DUL) or field capacity, moisture
content at saturation (SAT), and the macropore (KSMAC) and soil matrix (KSMT X)
hydraulic conductivities for the different soil depths, as shown in Table 6.1. The drained
upper limit (DUL, 0.28 cm3 cm”) was determined fiom TDR field observations 48 h following
a soaking rain. Hall and Heaven (1970) used similar procedure to estimate field capacity, in-
situ, in early spring using a neutron probe. This value agreed with the gravimetrically

determined value of 0.281 cm3 cm" (Saffel, 1994).

Model Description

Functional models like SALUS have modest input requirements, minimum computational
time, making them more user-fiiendly. The model contains subroutines for estimating
infiltration, drainage, upward-flow, moisture redistribution, potential soil evaporation,
transpiration and evapotranspiration, and root water uptake. The model requires a weather
file, a soil file, and a plant data file. A detailed outline of the input files was provided by
Ritchie and Baer (1994, personal communication).

Infiltration amounts were assumed equivalent to irrigation and/or rainfall assuming

negligible runoff. The model assumes that drainage occurs only when soil moisture content

106

in a given depth exceeds the DUL. Daily drainage amounts were estimated from the equation

(Ritchie, 1981):

drainage = SWON (VMC, - DULI) * DIAYR, (

where SWON is a unitless drainage coeflicient that varies between 0 and l; VMCI and DUL,
are daily volumetric moisture content and the drained upper limit, respectively; and DLAYRl
is depth of the soil layer.

Potential evapotranspiration was calculated according to the equation of (Ritchie, 1994
personal communications). This equation dilfers fiom the Penman equation in that it does not
include a wind firnction or vapor pressure deficit term. An advantage of the SALUS model
is tha it partitions evapotranspiration (E,) into soil evaporation (E,) and transpiration (Ep)
on the basis of the leaf area index. Potential soil evaporation (E,) was estimated from
potential evapotranspiration as a function of the leaf area index (LAI) according to the

equation (Ritchie, 1972):

Es ____ Ep—(OA LA!) (

Upward water flow was calculated using Richards (1931) equation. The new water balance

model was written in FORTRAN. Modifications in the FORTRAN code were made by J. L.

Ritchie (personal communication).

Model Inputs
Weather data from a Rainbird weather station 2 m high and 30 m from the plots were

used in soil-water-balance simulations. Weather data sets for these simulations included daily

107
solar radiation (MJ m'2 day"), maximum and minimum temperatures (°C), and rainfall (mm).
For the irrigated treatments, irrigation amounts were added to rainfall. These data were used
to estimate infiltration, drainage, soil evaporation and turf transpiration. Data files were set

in arrays specified for the SALUS model (Ritchie, Personal communication).

Plant Properties

Plant variables used for simulations were LAI and root length density (RLD) for the four
depth increments. An efl‘ective root length density of 5 cm cm'1 was assumed for the 0-5 cm
depth. Root length density for the 5-10, 10-15, and 15-25 cm depths were calculated based
on the percent root mass distribution by depth (1.1 cm cm", 0.55 cm cm'1 and 0.45 cm cm'1
respectively). These percmtages were derived from average root mass ratios for four depths
from three sampling dates presented in Chapter 4. These estimates fell within the range of
values observed by Murphy et al. (1994) in minirhizotron studies.

Leaf area index is needed in practically all ET models (Ritchie and Amato, 1990).
Instruments for accurate leaf area determination for closely mowed turf are not yet practical.
Leafarea indices for fairway turfs are not easy to measure because of the small leaf area for
closely mowed turfs and the folding habits of some species such as annual bluegrass. For a
fairway turf mowed three times a week at a cutting height of 16 mm, a constant LAI of 3.5

was assumed (Ritchie personal communications)

Model Modifications
Simulations began on May 1 each year and ended on September 30 (152 days annually).

The model was modified by; 1) setting a fixed leaf area index, and ii) allowing for user defined

108

root length density. This provides greater flexibility in simulation options. Output files were
also modified to generate only data relevant for the water balance study. Descriptive statistics
was used to analyze the output data. Volumetric moisture content as measured by TDR were
compared to simulated VMC by regression analysis.

The SALUS model assumes homogeneity of each soil depth. Water movement fiom
irrigation or rainfall into lower depths occurs only after the previous depth is saturated. The
model also uses a two-domain saturated hydraulic conductivity: matrix and macropore

hydraulic conductivities.

Results and Discussion

Weather for the 1992, 1993 and 1994 seasons are presented in the appendix. In 1992,
rainfall was spread more evenly than in 1993 or 1994. In 1993, most of the rainfall occurred
in the latter halfof the season. In 1994, heavy and frequent rains occurred between days 163
and 23S, coinciding with the peak of hot summer conditions.

Model output of interest include infiltration, drainage, VMC, and ET components.
Infiltration amounts generated by the model were the sum of rainfall and/or irrigation for the
different treatments. The 2.5 mm daily treatment had the highest total water input as
irrigation was applied daily regardless of rainfall. As a result, cumulative seasonal infiltration
was also highest in the 2.5 mm daily treatment for all years. The water input per application
tended to be highest for the FC which had lower irrigation frequency than the DLY treatment
but had higher amounts of applied water per irrigation event. The stress treatment had the

lowest mean seasonal infiltration, as expected.

109

Mean seasonal infiltration ranked DLY > FC > EOD > STR (Table 6.2). Seasonal
water application for the FC, EOD and STR treatments were 0.80, 0.75 and 0.50 of the
amounts for the DLY treatment in 1992. Similar results were observed for other years (1993;
0.89, 0.76, 0.51 and 1994; 0.91, 0.73, and 0.53 respectively for the FC, EOD and STR).
This suggest that during wet years application of 2.5 mm daily supplies more water than is
required to maintain the soil at field capacity. Futhermore, maintaining soil moisture levels
‘ at field capacity fails to accommodate potential moisture gains from rainfall, hence increasing
water losses.

Cumulative infiltration by irrigation treatment for 1992 is given in Fig. 6.2a. The data for
all years are shown in Table 6.2. Cumulative infiltration of more than 700 mm for the DLY
treatment was nearly double that for STR (357 mm) in 1992 but only 1.3 times compared to
F C and BOD. Cumulative infiltration ranked DLY > FC > EOD > STR for all years.

Daily drainage for 1992 are presented in Fig. 6.1. Seasonal values for all years and
treatments are presented in Table 6.3. Drainage fi'om the stress treatment was minimal
compared to the irrigated treatments. Overall daily and seasonal drainage by irrigation
treatment also ranked DLY > FC > EOD >STR. Variation in drainage losses was also
dependent on rainfall amounts and distribution. For a water balance model in which the depth
of interest is only 25 cm, drainage losses are expected to be high due to low storage capacity
of the shallow soil depth under consideration.

This was particularly true when 2.5 mm of irrigation is applied daily or when the soil is
returned to field capacity daily in years with above average rainfall. Drainage in this
modelwas estimated below the 100 cm depth hence the low drainage values reported in Table

6.3 Daily application of 2.5 mm of irrigation resulted in 1.4 times more drainage losses than

110

Table 6.2. Seasonal water application by irrigation treatment and by year based on 152

 

 

 

 

 

 

 

day season.
Irrigation Treatment
Year . 2.5 mm . .
2.5 mm Daily Alternate Days Field Capacrty Stress
mm

1992
Mean 4.7:t5.2 3,515.6 3.8i6.5 2.3 21:55
Total 717.4 536.1 573.6 357
1993
Mean 4.8 :t 5.0 3.7 :1: 5.4 4.3 i 6.7 2.5 i 5.4
Sum 740.9 563.9 658.7 384.3
1994
Mean 5 i 5.9 3.7 :1: 6.7 4.4 d: 6.0 2.6 d: 6.3
Sum 757.9 556.2 689.3 404.2

 

EOD, 1.2 to 1.3 more drainage loses than FC and more than 2 times more drainage compared

to STR.

Cumulative drainage for the different treatments for 1992 are presented in Fig. 6.2b.

Drainage amounts were low because drainage was estimated below the 1 m depth, four times

deeper than the sphere of influence of TDR installations. Estimates for the 25 cm depth are

expected to be much higher suggesting poor resource capture by shallow rooted cool season

turfs. Maintaining the soil at or near field capacity does not allow maximum utilization of

rainfallHowever, there was evidence of upward capillary presented in Chapter 5.

111

Table 6.3. Seasonal drainage by irrigation treatment and by year based on 152 day season
as predicted by SALUS.

 

 

 

 

 

 

 

 

Irrigation Treatment
Year 2.5 mm Daily A] 1122;113:1111.” Field Capacity Stress
mm

199;
Mean 0.4 :1: 0.3 0.3 i 0.3 0.3 :1: 0.5 0.2 :1: 0.3
Range 2.8 2.7 3.5 2.6
Total 64.9 47.0 50.9 29.3
l_9_93
Mean 0.4 :1: 0.3 0.3 :1: 0.3 0.4 :1: 0.6 0.2 :t 0.3
Range 2.3 2.1 3.5 1.8
Sum 67.4 49.3 58.3 31.6
19%
Mean 0.5 :t 0.4 0.3 :1: 0.3 0.4 :1: 0.5 0.2 :t 0.3
Range 2.6 2.5 3.4 2.3
Total 68.5 49.2 57.8 33.4

 

Daily soil evaporation, (E,) and plant transpiration (E) for 1992 are presented in Fig.

6.3a. Daily E, soil as predicted by the SALUS model ranged from 0.02 to 1.4 mm but

seasonal E, means ranged from 0.8 to 1.0 mm for all years as presented in Table 6.4. It is

worth noting that E, was strictly a fimction of the leaf area index. Based on an assumed fixed

leaf area index of 3.5, assumed in the calculation of E, the more or less constant value is

expected. This implies that soil evaporation is independent of soil moisture conditions, once

112

Table 6.4. Seasonal soil evaporation by irrigation treatment and by year based on 152 day

 

 

 

 

 

 

season as predicted by SALUS.
Irrigation Treatment
Year . 2.5 mm . .
2.5 mm Daily Alternate Days F reld Capacrty Stress
mm

1992
Mean 0.9 :h 0.3 0.9 :h 0.3 0.9 $0.3 0.8 :h 0.3
Range 1.3 1.3 1.9 1.3
Sum 129.4 129.3 126.4 126.4
1993
Mean 0.8 i 0.3 0.8 :h 0.3 0.9 d: 0.5 0.8 :1: 0.3
Range 1.4 1.3 2 1.3
Total 125.2 125.2 141 123.3
1994
Mean 0.9 $0.3 0.8 10.4 1 :1: 0.3 0.9 d: 0.3
Range 1.3 1.8 1.6 1.3
Sum 136.7 120.3 156.6 132.4

 

again emphasizing the need for soil moisture-based irrigation scheduling. Further research
would be needed to obtain accurate seasonal variation of LAI for closely mowed turf. This
may improve predicted values of the different hydrologic components.

Cumulative seasonal values for soil evaporation, plant transpiration and potential
evapotranspiration (E,) are presented in figure 6.3b. The highest and lowest transpiration and
soil evaporation values were recorded between day of the year (DOY) 140 and 160. Overall,

soil evaporation contributed up to 20% of daily potential ET. It must be noted that the

113

Table 6.5. Seasonal transpiration by irrigation treatment and by year based on 152 day

 

 

 

 

 

season as predicted by SALUS.
Irrigation Treatment
Year . 2.5 mm . .
2.5 mm Daily Alternate Days Field Capacrty Stress
mm

1992
Mean 3.4:t1.3 3.321: 1.2 3.5:t1.2 3.3:1:1.3
Range 5.1 5.1 7.3 5.3
Total 511.2 509.5 531 514.2
1993
Mean 3.2113 3.2i1.3 3.6120 3.3i 1.6
Range 6 5.4 9 5.7
Total 494.1 494.4 554.4 495.9
1994
Mean 3.6i1.3 3111.6 4.1112 3.6il.3
Range 5.4 7.7 6.1 5.5
Total 540.2 474.3 617.9 544.3

 

quantification of E, was based on an assumed rather than measured leaf area index (LAI)

given the difficulties involved in measuring LAI for closely mowed turfs.

Mean seasonal transpiration values for all years and treatments are shown in Table 6.5.

Mean transpiration values among irrigation treatments were fairly constant in 1992 and 1993,

but not in 1994. The highest mean seasonal transpiration values from the simulations were

for the field capacity treatment, whereas those fiom the 2.5 mm daily and the stress treatment

114

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

>.500 b
a 1
E400......-.....................] ............................
E V 1
............. j ........... 1----1 ..................... .
320.0 ......... -_--l ........... Hunt ________ --I ......... I
1 l1 ‘1 1
_§ 'f """"" """ 1. ' ' ' """"
g 0.0 .. A. 1.1‘ ’ 17' 1‘1.s.1.c..."1.1..1.1.1 1‘ '1 A)u1:J1.1.°°.1.1a.3.1.1‘-.1.111‘L'R
40.0 a . L, g - ,
120 110 160 I10 200 220 210 260 210
DayoftheYear
$00
" c
3.2 ......................... - ..... l .......................
E
,.. ------- ;---l ------------ (ml -----------------------
€200 ....... fi““,’i“""""““;“ ................ I ..........
f 1
110.1. ....... 1.1---H----L ....... l --- 3% ---*--.{LI-- ---1
‘ ‘: , \ ,1 ,
é .r L:- ’1.*.av-”".M7J- Jill-«1&3 NJ "1‘ K."- i
g + rc.'92 *INF‘.L-_ *0»:
--100- A f ‘ .
120 m 110 1 200 o 210 210 210
DayoftheYear
530.0. d
a . l
3100 .....................................................
E. . l, 1,
1""1 ------------ , ----------- 1,. ----- 1,
1 l! 1 I
.- I 1' Y: 1
so. ........................................ -- .........
1 °. 1' 'i l l l " 1'
8100-11. ....... ,----l!,--f ....... ii --- -- .. ----;-- 3---.1 .....
l 6 l;
.§10£45X7‘_r’+1 \M 0 F1 A ,lll‘ ‘10
1% 51192 *M *0“
”40.0 3 .
no no 160 200 210 210 210
DayoftheYear

 

 

 

Fig. 6.1 Seasonal infiltration and drainage for the 2.5 mm daily (DLY); 2.5 mm every other
day (BOD); Field capacity ( F C); and Stress treatments (STR) for 1992.

115

were about the same.

Transpiration estimates were based on weather data at the site. Mean transpiration rates
were about the same, regardless of irrigation treatment, for all years. Because the stress plots
received no irrigation, one would expect less moisture depletion fi'om this treatment and
hence lower transpiration values. This provides a good argument for conjunctive use of soil
moisture monitoring and conventional ET methods for irrigation scheduling.

Cumulative transpiration was similar for all irrigation treatments for 1992 (Fig. 6.30).
The lack of separation of the different treatments in the cumulative transpiration plots from
year to year is evidence that irrigation treatments did not significantly affect transpiration in
this model. This is again obvious because unlike rainfall, the effect of irrigation on ET is not
accounted for in weather data used in ET calculations.

Volumetric moisture content was the most important variable in this study because TDR
observations in the field could be compared to simulated VMC. Simulation outputs for VMC
for the difl‘erent depths for the 2.5 mm daily treatment and TDR measurements are presented
in Fig. 6.4. Both measured and simulated VMC ranked 0-5 > 5-10 2 10-15 > 15-25 cm
depths with values ranging from 0.24 cm3 cm'3 for the 15-25 cm depth to a high of 0.33 cm3
cm'3 for the 0-5 cm depth. Although the 5-10 cm depth was consistently higher in VMC than
the 10-15 cm depth, the differences were not always significant. Measured VMC values were
often lower than simulated values with increasing depth.

Measured and simulated VMC response to rainfall and im'gation for the field capacity
treatment are presented in Fig. 6.5 The field capacity treatment showed higher variability

than the 2.5 mm daily treatment in all years. Both the predicted and measured VMC show

116

 

 

 

Cumulative Infiltration. an

 

 

 

 

 

 

 

 

 

 

 

 

 

 

.. I992 --— DLY —.‘.— BOD —— FC _..... STR
-2oo.0 L - - 4. , - -
120 140 160 180 200 220 240 260 280
Day of the Year
200.0
.. b
100.0..-----------------; ............................. (7.-
E ’
I 0
5120.0 .l ...........................................
C
.s l
5 .00 a ............................................ m .-
3 v
i 10.0 - ..........................
s l
p 0.0 ‘r
l ”91 *nLv..._500_ rc ._....sn
40.0 : f ; 1
no 140 100 no 1:0 110 :10 :10

1“
Day of the Year

 

 

 

 

 

 

 

 

 

Cum. infiltration & drainage, .111

DLY. '92 —.—- Cum. INF. — CI".- DEN.

 

120 140 160 180 200 220 240 260 280
Day of the Year

 

 

 

 

Fig. 6.2. Cumulative infiltration (a), and cumulative drainage (b) for the daily (DLY), 2.5 mm
every other day (BOD), field capacity (F C), and stress (STR) treatments and cumulative
infiltration and drainage (c) for the daily treatment , 1992.

117

a very narrow range over time. Volumetric moisture content, by depth, followed the pattern
observed in the 2.5 mm daily treatment. Maximum VMC values during rainy periods for the
0-5 cm depth seemed to coincide for both irrigated treatments. The field capacity treatment
showed greater moisture depletions for the 15-25 cm depth than did the 2.5 mm daily
treatment.

The stress treatment (Fig. 6.6) showed the highest variation in VMC for all years. There
was better agreement between simulated VMC and field observations for the stress than other
treatments in both years, 0.39 < R2 < 47 compared to 0.12 < R < 0.29 for the irrigated
treatments. For the irrigated treatments, VMC for the 0-5 cm depth was comparably high
during wet periods. Ritchie and Amato (1990) stated that organic matter content increases
the drained upper limit by 23% for each percentage increase in organic matter content.
Accumulation of organic matter at the surface may thus explain higher VMC values observed
in the 0-5 cm depth. However, during extended dry periods, the 0-5 cm depth for the stress
treatment was also the driest depth. The highly dynamic nature of this depth, as exemplified
in the stress treatment, was due to its proximity to the changing environmental conditions,
high root density, and organic matter content.

Three prevalent philosophies that have been used in validation of models are: i)
comparison of measured versus modeled values; ii) expert opinion, and iii) use of an existing
model (Manetsch and Park, 1993). This new water balance model is being developed and
verified for various field crops. Field observations of VMC as measured by TDR under
fairway turfs for 1992 were compared to model predictions. Field measurements using TDR
fell with the range of the SALUS model VMC for all treatments and for all years. However,

R2 values (up to 0.29 for the irrigated treatments and 0.47 for the stress indicate that much

l 113

 

 

 

 

 

 

Evaporation and transpiration. nun/day

120 140 160 180 200 220 240 260 280

 

 

 

 

 

 

 

 

 

Day of the Year
800.0
E r b
E 6000
D.- . a
in:
'3 .
a 400.0 -
a 0
:5
0 200.0 -
.2.
3 1
g 0.0 -
U T 1992 ._.. Cum. Ba ._._ Cum. Bt -- Cum Bp
-200.0 . . 4 r .
120 140 160 180 200 220 240 260 280
Day of the Year

 

 

 

 

 

 

7
[.—
‘3
5
'r.
2
E
8.
«i
.1
I
I
5 I :
l 1992 --DLY_.—r:00_ rc _.._STR l
400.0 J _. . . . . . .
120 140 160 110 200 220 240 260 230
Day of theYear

 

 

 

F ig.6.3. Daily soil evaporation (Es) and transpiration (Es) (a); cumulative soil evaporation
(Cum Es), transpiration (Cum. Bt), and evapotranspiration (Cum. Ep) (b); and cumulative
potential evapotranspiration (ET) for the daily (DLY) 2.5 mm every other day (BOD),
field capacity (FC) and stress (STR) treatments (c), 1992.

119

Volumetric moisture content, cm3/cm3

 

 

 

 

 

 

 

 

 

 

 

 

 

0.40. j
% 05cm '
0.35 ----------------------------------------------------
T ~. 0 I l. l
0,0 ‘ A /\ / fl 0 ll. l. ~le M .l l. L. A— Hg.“- IflLr-a j
_ ._~.—---a‘— —————— o.’—----——-.—{ o- .4 ...-. -‘L-
T/ Wuu\% V “W" '17-" “exec-J 1;»;wa ' 1
‘ - n - |
0.25-f- ---------------------------- ?--------~----—--~----—--.f
0.20-K ----------------------------------------------------- 1'
l
i l
0.l5—‘- ----------------------------------------------------- J
‘ t
010’ l
120 140 I60 180 200 220 240 260 280
040
i 5-10cm l
0.35-— ------------------------------------------------------
l l
.l l
0.3.. -------- A---~“.—--«x ------ , -Q-r .-i--r-----.'---.-L;--.rr-.-.---f
; /\ A I l . \ ,1 . .- ', f ’l 1- ‘ at 7“ 7’ -. - -’ - . , .- 3-4- " It . "1‘ ‘
- / k/V IV, ‘V' \J \4 v! NV ‘Wv ‘J - L? '- . \;. --\ I \J‘ '
011' - --- ' '5‘ - 'va- l
""""""""""""""""""""""" d"‘ """~""’""""“"‘
i l
T l
0.20—- ------------------------------------------------------
7 1
o,;5._....-----------..-----------_--..-----..---------..---__-__.1
9
0.10' i
120 140 160 180 200 120 240 260 280
010
.' lO-lfcm I
0.35—-----------------—------------------------—-----------‘
; i
030- ........ A-7-a'r"--N_79 ..... p-50—Lfi-7op—ok—g-———J\— -;-LL--rq-P—
" I ‘ ' " " l‘. _ —‘\ I
° / i“ x \a \V \. V \' V \\_’~—— ¥ - i‘ . V‘c“ — -* y-v'i \—
0,25 ............................. a---- ...... -.------l":' .....
0.20—l---------~—------------—--------—------------------—-7
l :
0.15 ...................................................... . l
i I
7 :
0.10
l20 NO 160 180 200 220 240 260 280
‘1 t
l
040‘ 1
4, lS-25cm I
0.33-7---------—----—---------------------—-----------------‘
l i
0.30— -------- 4----'\---—-—--~-:‘-- -'--- I.---i
l kf 'l /\ 4’ l I. ‘\ [\f' ’3 l ' IV. {-1 J\_,:.4f\- _[¥-.‘ ' H ~7\. ‘\ j ‘
T \— \a\/ g...\ \_.._.\-‘V" & 3:};AP V I
025...../.¥.’§ ....................... .9- FKJi‘. ...... .3? -'-‘-—-1:
0‘207 ---------------------------------------------------- J
i .. Measured —. Simulated ?
0.15-’- ---------------------------------------------------- J
orol :
HO HO l60 l80 200 220 240 260 280
Day of the year

 

 

 

Fig. 6.4. Measured and simulated volumetric moisture content by depth for 2.5 mm daily
treatment (DLY), 1992.

120

Volumetric moisture content, cm3/cm3

(I40
i 0-5 cm

(130 j. DNA. JJNAJ" .

 

 

    

 

 

I20 140 I60 ISO 200 220 240 260 280

 

 

040 ,i

j- S-HJGn
035 . . .m,,.._. .-.m u- H.

030 t

025 i- rNN—VVJNW/AV‘ “f .V- ~ . PH? WWW)!"

 

0l5

 

0l0 *r 44 .
lZO I40 160 ISO 200 220 240 260 280

 

 

040 7
lO-lScm
035 ._.. H n . ,”.- ... _ W.. h . . ..-. . , ,,.. . h-.--n--mmm--MH. ”W-N-“ .u.mmnmn.

   

030 R i x I, Til/TIA; .3 “a

, I.
,rz‘w-.-~M,v" Awl~.J..R.,.. .‘R R x. t W

r—o—

025

0.20 4-

0.15 T

 

o to - - ,
r 20 l 40 I60 130 200 220 210 260 280

 

 

 

l .
i WA/iWAW/xvjm ft . J \xwa .1” A Vi V'—
l

- Measured — Simulated

 

 

i A A
v v

0.10 f . - a ~ -
120 I40 I60 180 200 220 2‘0 260 280
‘ Day of the year

 

 

 

Fig. 6.5. Measured and simulated volumetric moisture content by depth for the field capacity
treatment (FC), 1992.

121

Volumetric moisture content. cm3/cm3

 

0..)

  

o-Scm
035 .,

0.30

 

 

 

 

 

 

 

 

l

030 -— ,. I ,_ ,_ \
0.25 J- WfVAVAJN; \ , F‘~.‘I‘J\ , [A

 

 

 

 

 

 

0.15 .,_, < 15-25 cm

_. A .5. ;, \ ‘ ‘ YR

‘ ‘J 1 . \g. r-"_‘ _ . .
0.25 <— ~ [5. flfx/ \f‘J \ J VV R. - t“, "7*“ ‘4 W.Wix
\ \ J \ '

x
4;“

~l

- ‘ Measured -—-I -- ISimulated

 

 

 

DO “0 I60 "0 220 2‘0 260 280

200
Day of the year

 

 

 

Fig. 6.6. Measured and simulated volumetric moisture content by depth for the stress (STR)
treatment, 1992.

122

of the variation in soil moisture content is not explained by the model. This low correlation
for the irrigated treatments could explained in terms of poor irrigation application uniformity,
and the spatial variability of VMC in the field.

Simulated VMC and TDR measurements showed reversals in VMC trends above and
below the 0.20 to 0.25 cm3 cm" range for this soil type. Over this range all soil depths
showed only minor differences in VMC for the stress treatment and it is reasonable to assume
that there is no net moisture flux among depths in spite of evapotranspiration losses. From
an environmental standpoint, this could be an ideal level to maintain soil moisture content if
soil moisture based irrigation scheduling is employed. Two advantages of this are: i) the
above moisture level guarantees at least the minimum acceptable turf quality rating (6); ii)
potential water savings could result from lower irrigation rates, with greater accommodation
for water inputs from rainfall.

Simulated versus the measured VMC for 1992 are presented in Figs. 6.4, 6.5, and 6.6.
Volumetric moisture content for the irrigation treatments at four different depths were
compared to the simulated values for 1992 and 1993. For the 5-10, 10-15, and 15-25 cm
depths, the model overpredicted soil moisture levels for all treatments in turfgrass ecosystems.

The assumption of constant soil hydraulic properties for both years, when soil hydraulic
properties are indeed dynamic may also account for differences between the model and field
data. In addition, poor irrigation application efficiency may contribute to the observed
differences. Model predictions of volumetric moisture content were based on the
instantaneous VMC following rainfall or irrigation. The time lag between irrigation
application (0300 h) and TDR measurements (0700 to 0900 h) may explain some of the

variability between predicted VMC and TDR readings.

Conclusions

The modified water balance routine of the SALUS model provided reasonable estimates
of hydrologic components for turf ecosystems. Time domain reflectometry data fell within
the range of simulated VMC. The best agreement was with the 0-5 cm depths and the stress
treatment. The highest R2 values between simulated and TDR VMC were for the stress
treatment and the O~5 cm depths for all years. Model estimates for the 0-5 cm depth provided
the best approximation of field conditions for all treatments in all years. Although simulated
VMC were within the range of TDR measurements, R2 values were low. More turf specific
data would be needed to improve correlations between field observations and simulation

output.

Mean seasonal infiltration and drainage ranked DLY > PC > EOD > STR. The low
drainage amounts may not imply less leaching from the 25 cm depth as drainage was
calculated below the 100 m depth as for field crops. Daily soil evaporation contributed about
20% of total ET based on assumed leaf area index of 3.5. Transpiration values were within
the expected range for East Lansing MI. Contrary to expectation, seasonal cumulative
transpiration was not different for the various irrigation.

The stress treatment showed the greatest variation in VMC for all years. While changes
in VMC for the irrigated treatments were over a very narrow margin, volumetric moisture
content by depth ranked 0-5 > 5-10 >10-15 >15-25 cm at all times for the irrigated
treatments. This trend was also true for the stress treatment during wet periods but during dry
periods, the trend was reversed (15 -25 > 10-15 > 5-10 > 0-5 cm). These reversals were
evident in both the simulated and measured VMC. For a model based on limited weather

input the SALUS model provides accurate ET estimates and VMC estimates by depth for a

123

124
turfgrass ecosystem. Once all the data is arranged, it takes less than 5 seconds to execute.
This is a high level of efficiency with respect to computation time.

Because the model was initially designed for crops the model stmcture did not allow for
adjustments in the soil data input by depth. Future modifications in programming may
hopefiilly resolve this weakness. With technological advancement more accurate leaf area
indices for closely mowed turf will improve the partitioning of soil evaporation and
transpiration. Overall this model could serve as a management tool for improving turf

irrigation management.

Literature Cited

Carrow, NR. 1991. Turfgrass water use, drought resistance and rooting patterns in the
Southeast. ERC 01-91. Department of Agronomy, Griffin, GA.

Dalton, RN. 1992. Development of time domain reflectometry method for measuring soil
water content and salinity.

Hall, D.G.M., and F .W. Heaven. 1970. Comparison of measured and predicted soil moisture
deficits. J. Soil Sci. 30:225-237.

Manetsch, T.J., and G.L. Park. 1993. The systems approach as a problem solving
methodology. In Systems analysis and simulations with applications to economic and
social systems. Michigan State University, East Lansing, MI.

Murphy, J.A., M.G. Hendricks, P.E. Rieke, A.J.M. Smucker, and BE. Branham. 1994.
turfgrass root systems evaluated using minirhizotron and video recording method.
Agron. J. 862247-250.

Richardson, C .W., and J.T. Ritchie. 1973. Soil water balance for small watersheds. Trans.
ASAE 16:72-77

Ritchie, J.T. 1972. A model for predicting evapotranspiration from a row crop with
incomplete cover. Water Resour. Res. 8:1204-1213.

Ritchie, J.T. 1981. Water dynamics in soil-plant-atmosphere system. Plant and Soil 58:327-
338.

Ritchie, J.T. 1991. Specifications of an ideal model for predicting crop yield. In Muchow and
Bellamy (ed.) Climatic risk in crop production: Model and management for the semiarid
tropics and subtropics. CAB International.

Ritchie, J.T., and M. Amato. 1990. Field evaluation of plant extractable soil water for
irrigation scheduling. Acta Hort. 2782595-615.

Saffel, MT 1994. Time domain reflectometry for turf irrigation scheduling. MS. Thesis,
Michigan State University, East Lansing, MI.

Skaggs, RW. 1978. A water management model for shallow water table soils. Report N. 134.
Water Resources Research Institute, Univ. of North Carolina, Raleigh, NC.

Topp, G. C., and J. L. Davis. 1980. Electromagnetic determination of soil water content:
measurements in coaxial transmission lines. water Resour. Res. 162574-582.

Vargas, J. M. 1994. Management of turfgrass diseases. Lewis Publishers, Ann Arbor,
MI.

125

CHAPTER 7

An Evaluation of Turfgrass Irrigation Scheduling Using the
Soil Conservation Service (SCS) SCHEDULER Model

Abstract

Modeling has become an important tool in crop research. The use of models in
turfgrass studies, however, has been very limited. A three-year study (1992-1994) was
conducted at the Hancock Turfgrass Research Center to simulate evapotranspiration,
available water and drainage losses (excess water) from turfgrass plots under three irrigation
regimes using the Soil Conservation Service (SCS)-SCHEDULER model (Shayya and Bralts
1994). Irrigation treatments were: i) apply 25 mm upon the appearance of wilt stress (STR),
ii) apply 2.5 mm daily (DLY), and iii) maintain soil at field capacity daily (F C) based on time
domain reflectometry (TDR) measurement. Predicted daily potential evapotranspiration
(ETP) values during 152 day season over a three year period ranged from 0.5 to 7.6 mm.
Predicted turf ET (ET) for the stress treatment was lower than the actual ET for all years,
suggesting that soil moisture levels for the stress treatment did not meet the optimum amounts
and frequency needed for optimum turf growth. The ranking for amounts of applied
irrigation by treatment was DLY > FC > STR during wet periods and FC > DLY > STR
during dry periods. In all years seasonal cumulative rainfall exceeded the cumulative potential

ET, but poor rainfall distribution necessitated supplemental irrigation. Poor rainfall

126

127

distribution resulted in inefficient use of rainfall water, with a high potential for drainage
losses. Excess rainfall and im'gation ranked DLY > FC > STR. During wet periods, the
excess water from DLY treatment was higher than from the field capacity and stress
treatments. Simulated available water content did not agree well with field measurements.
These disparities suggest that firrther modifications are necessary to improve the agreement
between field observations and model output. Additional site—specific seasonal rooting depth
data may be necessary to improve model AWC and TDR observations. Since, the SCS-
SCHEDULER model does not account for upward water flow, predicted daily moisture

depletion values were much larger than normal.

Introduction

Considerable water savings can be achieved in turf management through eflicient
irrigation scheduling based on soil moisture depletion measurements (Augustin et al., 1984;
Snyder et al., 1984; Carrow, 1991). Economical irrigation requires the determination of
actual evapotranspiration. The multiplicity of ET models and the lack of agreement among
methods (Penman, 1948; Jensen and Haise, 1963; Doorenbos and Pruitt, 1977; Wright, 1981)
makes it diflicult to choose anyone model. Operational estimates for most ET models are
based on the assumptions of Penman (1948) or some modification of them. The basic
assumptions of Penman's definition of potential ET (ETP) do not match the conditions
encountered in turf culture. Thus, adjustments are necessary to acCommodate variations in
climate and model specifications.

Two important considerations in crop ET calculation are; crop coefficients needed to

estimate actual ET; and plant height that influences the aerodynamic component or wind

128

fimction of the Penman model. The cutting heights recommended for turf maintenance range
from less than 4 mm on greens to more than 80 mm on some low maintenance turfs. This
cutting height range is well below the specifications of Doorenbos and Pruitt (197 7) or Jensen
and Haise (1963). The lack of agreement among ET models by location, even after
adjustments have been made to obtain realistic ET values, suggests a higher level of
empiricism into an otherwise physical model. Crop coemcients are used to adjust ET, to
actual turf ET (ET, ). Mini-lysimeters are used to estimate actual or turf ET (ET,) and hence
the turf coefficient (K, ) (Carrow, 1991) from the ratio:
Et, / ET, = K,

Turfgrasses show a wide degree of genetic diversity and adaptive radiation reflected
in the variability in water use rates. Evapotranspiration rates have been shown to vary by
species (Tovey et al., 1969; Biran et al., 1981; Kneebone and Pepper, 1982; Kim and Beard,
1988) and even by cultivar (Biran et al., 1981; Shearman, 1986; Kopec et al., 1988). Johns
et al. (1983) developed a method of classifying turfgrass ET that assigns the following ratings:
low for ET values less than 4 to 4.9 mm per day, medium for ET values from 5 to 7.5, and
high for ET values above 8 mm.

Ideally just enough water should be applied to wet the rooting zone. Suggested
recommendations for turf irrigation include: application of water to replace daily ET or some
fi'action thereof; irrigation using timers to apply fixed amount of water on regular basis for
example 2.5 mm daily or 2.5 mm every other day (Vargas, 1994); or irrigation upon the
appearance of wilting stress. Although some recommendations are successfirl in maintaining
turf quality, there has been no quantitative evaluation with respect to total water use or excess

water for each of the above irrigation scenarios.

129

An initial step in the development of moisture-depletion-based irrigation scheduling
is the establishment of the "setful" point or field capacity (also called the drained upper limit,
DUL) and the refill point based on the management allowable depletion (MAD) (Campbell
and Campbell, 1982), where MAD ranges from 30-50% of available soil water (AW) . The
basic assumption is that when the soil moisture is between (DUL) and MAD, moisture stress
is less likely to occur. Below the MAD, turf growth and quality could be adversely affected.

Conventional methods for soil moisture determination include the gravimetric method,
resistance blocks, neutron and gamma attenuation techniques. These methods are time
consuming, labor intensive, involve radioactive sources or destructive soil sampling. Spatial
variability from changing sampling sites during successive measurements also introduces
errors into these soil moisture-evaluation methods. Despite good accuracy (Simpson and
Meyer, 1987), the neutron probe is not amenable to the shallow depths encountered in cool-
season turf management (Augustin and Snyder, 1983). However, there is still a need to
develop affordable and dependable soil moisture sensors for evaluating turf irrigation
requirements in addition to ET estimates.

In recent years, time domain reflectometry (TDR) has been widely applied in
detennining soil moisture and salinity (Topp et al., 1980). Time domain reflectometry uses
the relationship between the composite soil dielectric constant and volumetric moisture
content (VMC) proposed by Topp et al. (1980). Compared with other methods of
determining VMC, TDR is adaptable for automation and multiplexing (Wraith and Baker,
1991), and remote retrieval of soil moisture data on a continual basis by a single TDR (Baker
and Allmaras, 1990). Good agreement between simulated values and TDR estimates is

necessary to establish confidence in the modified model for turf irrigation scheduling. One

130

such model, the Soil Conservation Service (SC S)-SCHEDULER (version 3) was developed
by Shayya and Bralts (1994) at Michigan State University. A previous version of this model
has been described as one of the most accurate irrigation scheduling and evaluation software
for field crops (Allan, 1991).

The SCS-SCHEDULER model uses the F AO-modified Penman equation to calculate
potential ET (ETP), actual ET (ET,), and cumulative ET. Moisture depletion within the
rooting zone is calculated using the following method (Hillel, 1980):

rainfall + irrigation - (change in soil moisture content + drainage + runoff) = ET

Estimates of excess rainfall and irrigation are important for improving irrigation application
rates and reducing the frequency of drainage losses. Good agreement between model output
and field data would indicate that the model may serve as a management tool (Ritchie, 1991)
for irrigation managers. The SCS-SCHEDULER has a short-term forecasting routine based
on historic or actual data that predicts irrigation scheduling for up to 60 days, for a given
location. It also. Like other models, it allows for ad hoc experimentation without the labor,
delays, and costs associated with field experimentation (Ritchie, 1991).

Although considerable turf irrigation research has been reported in the literature
(Danielson et al., 1981; Aronson et al., 1987), the data are not reported in a form that can
easily be incorporated into models. With increased use of computers in turf irrigation
programming and research, modeling will no longer be restricted to researchers and
consultants. Innovations in technology may eventually reduce the cost of soil moisture
sensors. Conjunctive use of soil moisture sensors such as TDR with computers models will

bring turf irrigation management another step into the computer age.

131

Model Description

The SCS-SCHEDULER (Shayya and Bralts, 1994) is an irrigation scheduling and
evaluation package. This model provides a viable option for interactive crop irrigation
scheduling and evaluation for periods of up to 60 days. The input requirements include local
crop, soil, and weather data. The model computes the water balance within the rooting zone
or depth of interest and suggests irrigation scheduling for optimum crop management.
Evapotranspiration, rainfall, irrigation, and deep percolation are used to calculate excess
water (excess rain + irrigation). The model quantifies various irrigation scheduling scenarios
and their effects on plant-available water.

The objectives of this study were to: i) adapt the SCS-SCHEDULER model for use
in turf irrigation; and ii) compare ET, soil moisture depletion and excess water under three

turf irrigation regimes.

Materials and Methods

Three irrigation treatments were applied on an annual bluegrass (Poa annua var.
reptans L) and a Penncross creeping bentgrass (Agrostis palustris Huds. L.) fairway turfs
established in 1989 at the Hancock Turfgrass Research Center at Michigan State University
(Saffel, 1994). The soil type was a modified Owosso sandy loam (fine-loamy mixed mesic
Typic Hapludalf). The turfs were mowed three times a week at a cutting height of 16 mm and
were maintained according to standard practices for cool-season fairway turfs in Michigan.
Because there were no significant differences in water use by species, the data were pooled

across species.

132

The irrigation treatments in this study represented a range of prevalent irrigation
practices in Michigan: i) apply 25 mm only upon the appearance of wilt stress; ii) apply 2.5
mm daily (Vargas, 1994); and iii) return the soil to field capacity based on TDR-measured
moisture depletion.

Weather data for the summers of 1992, 1993, and 1994 were down-loaded from a
Rainbird Maxipaw TM station located at the site. The weather data were converted into
arrays compatible for use in the SCS-SCHEDULER by a routine in Quick Basic written for
this study by Walid Shayya, visiting Professor and V. F Bralts Professor of Agricultural
Engineering at MSU. This model uses the F AO-modified Penman equation (Penman, 1948)
to calculate ET. The calculated ET serves as the basis for estimating soil moisture depletion,
and excess irrigation and rainfall, and scheduling irrigation for various field crops.

Irrigation treatments served as whole plots (1 l m by 11 m) with split plots (5.5 x 11
m) seeded randomly to either annual bluegrass or creeping bentgrass. Pop-up irrigation
heads were located at each corner of the plot. Irrigation was applied at 0300 h and TDR
readings were taken between 0700 and 0900 h. The application and distribution uniforrnities
were greater than 85% (Saffel, 1994). However, lower values were obtained during windy
conditions early in the morning.

The model inputs included crop, weather, and soil data files. Seasonal turf
coeficients, rooting depth, and management allowable depletion (MAD) used for the SCS-
SCHEDULER simulations are presented in Table 7.1. The values were selected from the
literature (Feldhake et al., 1983; Aronson et al., 1987; Kim and Beard, 1988), where
available, or estimated from established values for field crops in the SOS-SCHEDULER data

base.

133

7.1. Seasonal crop coefi'rcients, rooting depth, and management allowable
depletion used for the SCHEDULER simulations.

 

 

Minimum Available
Growing Season Rooting Depth Crop Coefficient
Water Before Irrigation

(°/o) cm K11 %
0 15.2 0.45 50
10 20.3 0.50 50
20 21.6 0.60 50
30 12.7 0.70 50
40 16.5 0.75 50
50 14.0 0.75 50
60 15.2 0.75 50
70 17.8 0.70 50
80 17.8 0.60 50
90 16.5 0.50 50
100 15.2 0.40 50

 

1‘ Initial K, values were obtained fiom Kneebone and Pepper, (1982). The rooting depth
variation with time was adapted from Koski (1983).

Data used to develop the crop curve (turf curve), root zone expansion curve used in the
simulations were adapted from the rooting-depth patterns of cool-season grasses in response
to temperature stress during the warm summer months (Koski, 1983). Turf coefficients (K,)
ranging 0.50 to 0.85 have been reported in the literature (Kneebone and Pepper, 1982).
However, the moisture depletion rates using the above values were so high that simulated
AWC was much lower than values obtained by TDR or gravimetric methods. The K, values

were lowered progressively until there was agreement between simulated AWC and field

134

measurements. The coetficients presented in Table 7 .1 were reduced sequentially to obtain

the best fit to the field data.

Table 7.2. Farm (turf)«specific input data for SCS-SCHEDULER simulations.

 

Plant type

Starting date

Ending date

Duration of simulation

Degree days or % season

Management allowable depletion (MAD)

Available water content by depth

Allowable excess water above field capacity

Available water-holding capacity by depth 0-5 cm
5-10 cm

10-15 cm
15-25 cm

Turf
May 1
October 1
152
% season
50%
100%
105%
0.155 cm3 cm :1
0.145 cm3cm

0.135 cm3cm
0.125 cm3cm

 

TThe available water-holding capacity of each soil depth was assumed to be half of the

volumetric soil moisture content, expressed as equivalent depths.
moisture content for all depths rs 0. 28 cm3 cm‘3

The weighted volumetric

Simulations began on May 1 (DOY 122 for 1992, and DOY 121 for 1993 and 1994) of each

year and ended on October 1, resulting in 152 days of simulation each year. The default

MAD value of 50% of soil-available water content was used for all simulations (Table 7.1).

This value, commonly used in most irrigation models, indicates that irrigation is suggested

when half the available water content is depleted (Campbell and Campbell, 1982). Below

135

50% available water lefi in the soil, the plants expend higher amounts of energy per unit of
water uptake, hence the model calculates an adjusted ET (ET,). Also, the default allowable
excess water (105%) of field capacity was utilized for all simulations. Above this value, the
water is lost either through drainage or as runofl‘ and labeled excess water. Excess water
amounts and fi'equency estimates were based on the amount and frequency with which soil
moisture content exceeded the allowable excess water in the rooting depth (Table 7.2).
The plot used in this study had a slope of approximately 1.5%, so runoff would be
limited unless the rainfall intensity was high. In this modeL water lost by percolation or runoff

was considered excess water.

Results and Discussion

Seasonal rainfall, temperatures and means are presented in Table 7.3. Rainfall
maxima and means increased fiom 1992 through 1994. For example rainfall in 1992 and 1993
were 0.75 and 0.87 times that for 1994. Mean maximum and rrrinimum temperatures showed
less variation compared to rainfall. The range within minimum temperatures and maximum
temperatures were strikingly similar for all years.

The lowest and highest potential evapotranspiration (ET?) values for 1992 were 0.5
mm on May 31 and 6.8 mm on June 15, respectively. In 1993, a minimum ETP of 0.5 mm was
recorded on September 3, whereas a maximum of 7.4 mm occurred on both 16 and 17 July.
Potential ET ranged from a minimum of 1.0 mm on 1 May, 1994, to a maximum of 7.6 mm
on 22 June is not understood. Potential evapotranspiration is a weather-dependent variable

estimated strictly from atmospheric variables. Predicted cumulative ETn and ET, for all

treatments (Table 7.4) ranked 1993 > 1994 >1992. The order for ET, was 1993 > 1994 >

1992.

Table 7.3. Seasonal rainfall totals, daily seasonal means and maximum rainfall events;
seasonal lowest, highest and mean maximum temperatures; seasonal lowest, highest and
mean minimum temperatures for 1992, 1993, and 1994.

 

 

 

 

 

 

 

Year
Weather Factor
1992 1993 1994

mm
Mobil
Maximum 38.9 52.6 70.0
Mean&SD 2716.6 3.2:t8.1 3.73: 10.1
Total 421 493 565
Max. Temperature °C
Maximum 34.4 33.9 38.0
Minimum 9.4 10.0 10.4
Mean & SD 23.1 :t 4.6 24.0 d: 5.2 24.8 :1: 5.7
Range 25.0 23.9 27.6
Min. Temperature °C
Minimum -1.7 0.6 -0.1
Maximum 20.6 23 .9 22.1
Mean& SD 9.8i5.7 11.9i5.4 11.6i5.7
Range 22.2 23.3 22.2

 

137

Adjusted ET (ET,) values for the stress treatment were substantially lower in 1992
than for the irrigated treatments but not in 1993 and 1994. This indicates that the stress
treatment supplied less than adequate amounts of water to turfs in 1992 which had the lower
rainfall (Table 7 .4). The ET, values were remarkably consistent within years for all
treatments. With the exception of the stress treatment in 1994, ET, was consistent for all
irrigation treatments for each year. Variation in evapotranspiration fi'om year to year may thus
be attributed more to weather factors other than temperature since year to year variations in
temperature were not statistically significant.

A weakness of ET models is that they fail to include the soil water status or soil water
dynamics. This underscores the importance of incorporating soil moisture data into efficient
irrigation scheduling practices. Rainfall and irrigation amounts and simulated available water
content (AWC) for 1992, 1993 and 1994 are presented in Figs. 7.1, 7.2 and 7 .3, respectively.
The figures show the variation in AWC with time in relation to the drained upper limit
(DUL) and the management allowable depletion (MAD).

Rainfall and irrigation amounts and available water content for the 2.5 mm daily, field
capacity, and stress treatments for 1992 are presented in Fig. 7.1. Rainfall amounts and
frequencies were highly variable from year to year. The variability of rainfall amounts and
frequencies, as evident in the available water graphs for the field capacity and stress
treatments, dictated the irrigation requirement. Available water content for the irrigated
treatments was above the MAD for the entire season. This indicates that both irrigated
treatments supplied adequate or more than adequate amounts of water for plant needs.

The field capacity treatment deviated slightly from the 0.28 cm3 cm“3 average field
capacity value during days with successively high ET. There were four distinct dry-down
periods when AWC for the stress treatment fell below the MAD limit in 1992 (around day

140, 163, 170 to 180 and 235 in 1992). Quality ratings for the stress treatment were

138

Table 7.4. Seasonal cumulative potential (ET,), turf (ET; ), and adjusted seasonal ET
(ET,), by year and irrigation treatment as predicted by the SCHEDULER model.

 

 

 

 

 

Irrigation Year
Treatment 1992 1993 1994
ET, 526 716 696
Field Capacity ET, 328 452 436
ET, 324 448 436
ET, 520 716 696
2.5 mm Daily ET, 326 457 439
ET, 324 452 439
ET, 526 714 696
Stress ET, 328 456 429
ET, 269 334 418

 

significantly lower than for the irrigated treatments (see Chapter 4). With the increase in
rainfall amounts and frequency in the second half of the season, the AWC for STR stayed
between the MAD and the drained upper limit (DUL) most of the time. Occasionally, the
water content exceeded the 105% AWC limit, with the excess water lost as drainage or as
runoff. The available water content for the 2.5 mm daily treatment in 1992 (Fig. 7 . 1) shows
that VMC was at or above field capacity from July 10 to the end of the 1992 season.
Compared to the stress treatment, the 2.5 mm daily treatment had two distinct dry-down

periods at about days 160 to 170 and 180 to 195 but AWC fell never fell below MAD.

139

 

 

 

 

Irrigation, 2.5 mm daily

I IIIIIIIIIIIIIIIIIII IIIIII

 

l
Irrigation, mm

I‘I’I'IIIIIIIIIIIIIIIII'IIIIIIIIII'IIIIIIII IIIIIIIII IIII IIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIII 0

 

 

lO

 

Irrigation, field capacity

IIIIIIIII II III IIIIIIILIIIIIISLIm I IIIIIIIIIIII III III 0

1
LA
Irrigation, mm

 

 

 

 

Available water content (A WC) mm
IQ
O
l
5 '

 

 

 

 

50-1
40—
30..
20-I
10—
0- 6055
Seasonal rainfall E
-40___:
I I II I I' '
IIIIIIII‘IIJJI!I IIIIII 09¢

 

 

l
100 120 140

160

N
00..
O

I
I
180 200 220 240 260
Day of the year

Figure 7.1. Available water content (AWC), irrigation and rainfall for the 2.5 mm daily,
field capacity and stress treatments in relation to the drained upper limit (DUL) and
management allowable depletion (MAD), 1992.

140

The field capacity treatment maintained the soil moisture level at about field capacity
more consistently than DLY and received less seasonal irrigation. The amounts of water
applied per irrigation for the field capacity treatment were greater than for the 2.5 mm daily
treatment but irrigation was applied less frequently. This indicates that DLY supplied more
water than F C and hence more than adequate water for turf growth.

In 1993 (Figure 7.2), the 2.5 mm daily treatment supplied more than adequate water
for turf grth early in the season and toward the end of the season. Mid-season application
of 2.5 mm daily was lower than the evaporative demand. In no instance, however, was the
AWC below the MAD for this treatment. The deviation of the AWC from the DUL for the
field capacity treatment was less than for the 2.5 mm daily treatment. The field capacity
treatment maintained AWC at or near field capacity, as expected. There were several days in
1993 when the AWC was below the MAD for the stress treatment. When AWC is less than
the MAD, there is a potential for plants to be subjected to moisture stress, particularly on
days with high ET. Actual turf evapotranspiration (ET, ) is adjusted accordingly to reflect the
potential for moisture stress. Poor rainfall distribution thus causes a need for supplemental
irrigation to maintain quality turf as stated by Aronson et al., (1987) even when seasonal
rainfall exceeded cumulative turf ET.

The figures showing rainfall, irrigation, and available water content for 1994
are presented in Fig. 7.3. The 1994 season began with a dry period from day 130 to 163.
The heaviest rainfall (70 mm) for the entire study occurred on day 163. Rainfall was more
evenly spread from day 163 through the rest of the season. Soil water content for both
irrigated treatments oscillated about the DUL line depending on rainfall or irrigation. Soil
water content for the stress treatment was below the DUL line between day 140 to 162, 207

to 217, and 245 to 258.

Available water content (AWC) mm

60

 

50—
40—
304

20—

 

 

 

60

 

50‘
4o-
30—
20.,

 

 

60

H
B.

"8
:3.
O

3::
:31
a
O-
8

‘U
D)
O
H.

Q
1

 

 

 

so,
40.—
3o—
20—
101

 

Seasonal rainfall

 

 

 

100

 

i iLL i i l J 11] l r l i L l. r; I lit
T T I l l T
60 180 200 220 240 260

Day of the year

I
l

120 280

.II.
1
140 1

III
our—4H

1
O Ur

I l 1
O N -b O\
O O O

OM

Irrigation, mm

HD—l
Om

Irrigation, mm

Rainfall, mm

Figure 7.2. Available water content (AWC), irrigation and rainfall for the 2.5 mm daily,
field capacity and stress treatments in relation to the drained upper limit (DUL) and
management allowable depletion (MAD), 1993.

142

For the last third of the season, AWC by irrigation treatment ranked DLY > FC >
STR for all years due to frequent rainfall (Figs. 7.1 through 7.3). In a dry year, one would
expect the trend to be F C > DLY > STR. In terms of irrigation scheduling, the results
showed that during wet periods, the field capacity and 2.5 mm daily treatments both supplied
more than adequate water for quality turf maintenance. Turf quality ratings for both species
for the stress treatment were significantly lower than for the irrigated treatments (Chapter 4).
Excess water by year for three irrigation treatments are given in Table 7 .5. Excess
water lost as drainage or runoff reduces water use efficiency, having environmental and
economic implications. Excess water resulted from applied irrigation and/or rainfall when
AWC is near or above the field capacity. Excess water trends were highly variable from year
to year between irrigated and stress treatments due to variation in rainfall. Rainfall accounted
for all the excess water from the stress and most of the excess water from the field capacity
treatments. The amount of water applied per application for the field capacity treatment
varied fiom day to day, as expected. By contrast, the 2.5 mm daily treatment was obviously
constant. As expected, the least amount of excess water was from the stress treatment for
all years. Although considerable water savings were evident from the stress treatment, this

treatment subjects turfgrass to cycles of stress, with potential reduction in turf quality.
Because excess water lost from the stress treatment was due to rainfall, the amounts
were much lower than for the irrigated treatments (Table 7.5). Excess water from irrigation
for DLY ranged from 29 to 47, times that for FC, while the combined excess from rainfall and
irrigation ranged from 1.2 to 1.3 times that for FC. For the stress treatment excess water
from rainfall ranged from 0.1 in 1992 to a peak of 0.34 times that of F C in 1993. The excess
rainfall amounts ranked DLY > PC > STR. This indicates a rather low efficiency with

respect to water-resource capture for shallow-rooted, cool-season turfs. However, some of

Available water content (AWC) n n

143

60

 

50
40
3O

 

20

#

 

OM

Irrigation, mm

Irrigation, 2.5 mm daily

1

1

OUrt-d

" ill I “HI

 

50—
40—
30—

 

20 _,

p—s
K)!

Irrigation, mm

 

T
Lht—I
O

 

O

 

6O
50 —
40 -
30 —
20 — I
10 _.

 

 

1

Seasonal rainfall

1
ON-fiGOO

1

0000

Rainfall, mm

 

 

 

llL .-
I I

120 140

l
l

160

 

 

III. III II,JIII I II LL

1 80 200 220 240 260
Day of the year

100 280

Figure 7.3. Available water content (AWC), irrigation and rainfall for the 2.5 mm daily,
field capacity and stress treatments in relation to the drained upper lrnut (DUL) and

management allowable depletion (MAD), 1994.

144
the excess water could reenter the root zone through upward water flow. Irrigation to field
capacity or irrigation during wet periods are thus not efiicient water management strategies.

The fi'equencies of excess rainfall and irrigation by treatment are presented in Table
7.6. The frequency of excess irrigation as predicted by the SCHEDULER model showed
greater variation than that of excess rainfall for all years. The 2.5 mm daily treatment had the
highest fiequency of excess water applied for all three years. A maximum combined excess
frequency of 104 occurred in the 2.5 mm daily treatment in 1992, whereas the minimum
combined excess frequency (26) occurred in the field capacity treatment in 1994. The stress
irrigation treatment had fewer excess events, as expected. The excess frequency of rainfall
for all treatments was consistent from year to year for the irrigation treatments.

The number of days with excess rainfall was not very different between the 2.5 mm
daily and the field capacity treatments (Table 7.6). However, the number of days with excess
irrigation was drastically different. The 2.5 mm daily treatment had excess water loss from
irrigation. This suggests that for wet years, the application of adequate amounts of water to
replenish the soil to field capacity would result in water savings over daily application of 2.5
mm. Still losses from rainfall remained high because the soil was maintained at about field
capacity. During dry years, however, the 2.5 mm daily treatment subjected the turfs to
moisture stress (Saffel, 1994).

A major objective of turf irrigation scheduling is to supply adequate water for quality
turf grth without compromising environmental quality. Both irrigation and rainfall play
a vital role in the amounts of excess water (drainage losses) and the potential leaching of
nutrients and other agricultural chemicals. Table 7 .7 shows the means and ranges of excess
water amounts by irrigation treatment for the different years. The highest excess water event

of about 64 mm of water occurred on June 14, 1994, after 70 mm of rainfall. Up to 91% of

145

Table 7. 5. Seasonal excess rainfall and irrigation, by year and by irrigation treatment as
predicted by the SCS SCHEDULER model.

L

Irrigation Treatment

 

 

 

 

Beggar“ Year , , ,

Freld Capacrty 2.5 mm Darly Stress

mm

Rainfall 1992 363 386 28
Irrigation 2.6 123 0
Total excess 395 509 28
Rainfall 318 292 108
Irrigation 1993 1.8 86 0
Total excess 320 378 108
Rainfall 419 396 135
Irrigation 1994 3 . 1 90 0
Total excess 419 486 135

 

the rainfall was not stored in the root zone for both irrigated treatments. A minimum leaching
amount of 0.3 mm was recorded for the stress and field capacity treatments in all years but
only in 1992 and 1993 for the field capacity treatment. A distinct advantage of the SCS-
SCI-IEDULER is that it partitions rainfall and irrigation amounts into the fraction stored in
the soil depth of interest, the amount lost by evapotranspiration, and the excess water loss by

runoff and percolation. Although leaching is not estimated by the model, the solubility of

146
various agricultural chemicals, in the excess water, could provide usefirl estimates of the

leaching potential of various chemicals of interest.

Table 7.6. Frequency of excess rainfall and irrigation, by irrigation treatment as predicted
by the SC S-SCHEDULER model.

 

 

 

 

 

Irrigation Treatment
Year Type
Field Capacity 2.55 mm Daily Stress
Rainfall 26 27 13
1992 Irrigation 0 77 0
Combined 26 104 13
Rainfall 27 22 11
1993 Irrigation l 46 0
Combined 28 68 11
Rainfall 26 24 12
Irrigation 0 55 0
1994
Combined 26 79 12

 

The high frequencies of excess water from the 2.5 mm daily and field capacity
treatments suggest that irrigation scheduling should utilize a more conservative approach than
returning the soil to field capacity daily, or applying 2.5 mm daily and when possible should

include soil moisture monitoring as a feed back system.

147

Table 7.7. Means and ranges of excess water amounts by irrigation treatment as predicted
by the SC S-SCHEDULER model.

 

Irrigation Treatment

 

 

 

 

Year Excess
Field Capacity 2.5 mm Daily Stress
mm
Maximum 39.1 44.2 31.9
1992 Minimum 0.3 0.3 0.3
Mean 2.6 3.3 1.2
Maximum 41.4 50.3 34.8
1993 Minimum 0.3 0.3 0.3
Mean 1.5 2.5 0.7
Maximum 64.0 60.2 34.4
1994 Minimum 0. 8 0.3 3 .0
Mean 2.7 3.2 0.9
Conclusions

Evapotranspiration estimates from the model were within the ranges of ET for the
location. The available water content measured by TDR did not agree with simulated values.
This may be due to faulty assumptions in adapting the model for turfgrass. More research
would be needed to improve agreement between model prediction and field data.

Moisture depletion estimates from the SCS-SCHEDULER account for seasonal
variations in rooting depth. This is important because rooting patterns of cool-season grasses

are affected by root senescence during periods of high temperatures and other weather

148
conditions in response to differential partitioning of assimilates.The model provides
information on water use that would otherwise be unavailable or require time to calculate.
Most important, it provides a method for site-specific irrigation scheduling and evaluation.

Benefits from simulations in terms of time, money, and labor savings have been
documented (Hanks and Hill, 1980; Ritchie, 1991). Good simulation models serve as a usefiJl
basis for irrigation management and decision making. A method of irrigation scheduling that
is accurate, site- and crop-specific, rapid, and dependable can be used to rapidly build a data
base for model validation purposes. At this time the data base for turf modeling is not as
extensive as in conventional agriculture. Model development calls for an extensive data base
with quantitative information on the development of turfgrasses.

Although much research has been done in turf, the results are not reported in a form
that can be readily applied in simulations. Information on parameters such as leaf area index,
root length density, leaf extension rate and turf coefficients for different locations are not
readily available. The collection of such data requires several years of research. To
compensate for these shortcomings, assumptions are made which quite ofien oversimplify
various processes and at the same time introduce errors in the analysis. In this study, a crop
irrigation simulation model (the SCS SCHEDULER) was adapted for use in turfgrass
research. Conventionally, rooting depths for row crops exceed those for cool season turfs.
Turf coefficients for row crops are different from those encountered in fine turf management.
These may lead to higher drainage amounts and higher than normal moisture depletion rates
under turf ecosystems than with other crops. Hopefirlly, this study will generate interest in
further development of turf specific models that could serve as useful tools for research and

decision making in turf irrigation.

Literature cited

Allan, SC. 1991. Irrigation scheduling for ground water protection. In W.F. Rigger (ed.)
Irrigation and drainage. American Society of Civil Engineers, New York.

Aronson, L.J., AJ. Gold, R.J. Hull, and J.L. Cisar. 1987. Evapotranspiration of cool-season
turfgrasses in the humid Northeast. Agron. J. 792901-905. ,

Augustin, B.J., and G.H. Snyder. 1984. Moisture sensor-controlled irrigation for maintaining
bermudagrass turf. Agron. J. 76:848-850.

Baker, J.M., and RR Allmaras. 1990. System for automating and multiplexing soil moisture
measurement by time-domain reflectometry. Soil Sci. Soc. Am. J. 54:
1-6.

Biran, 1., B. Brando, I. Bushkin-Harav, and E. Rawitz. 1981. Water consumption and grth
rate of 11 turfgrasses as affected by mowing height, irrigation frequency, and soil
moisture. Agron. J. 75285-90.

Campbell, GS, and MD. Campbell. 1982. Irrigation scheduling using soil moisture
measurements. p. 25-42. In Advances in irrigation. Vol. 1.

Carrow, RN. 1991. Turfgrass water use, drought resistance and rooting patterns in the
Southeast. Tech. Compl. Report USDI/USGS Project 02. University of Georgia, Athens,
GA.

Danielson, RE, OM. Feldhake, and WE. Hart. 1981. Urban lawn irrigation and
management practices for water savings with minimum effect on lawn quality. Colorado
Water Resources Res. Inst., Fort Collins, CO.

Doorenbos, J ., and W0. Pruitt. 1977. Guidelines for predicting crop water requirements.
FAO Irrigation and Drainage Paper no. 24. United Nations, Rome, Italy.

Feldhake, C.M., RE. Danielson, and J .D. Butler. 1983. Turfgrass evapotranspiration. Factors
influencing rate in urban environments. Agron. J. 75:824-830.

Hanks, R.J., and R.W. Hill. 1980. Modeling crop response to irrigation in relation to soils,
climate and salinity. Pergamon Press, New York.

Hillel, D. 1980. Applications of soil physics. Academic Press, New York.

Jensen, ME, and HR. Haise. 1963. Estimating evapotranspiration from solar radiation.
Proc. Am. Soc. Civ. Eng. J. Irrig. and Drain. Div. 89:15-41.

Johns, D., J .B. Beard, and C.H.M. van Bavel. 1983. Resistance to evapotranspiration fi'om
a St. Augustinegrass turf canopy. Agron. J. 752419-422.

149

150

Kim, KS, and J.B. Beard. 1988. Comparative turfgrass evapotranspiration rates and
associated plant morphological characteristics. Crop Sci. 28:328-331.

Kneebone, W.R., and IL. Pepper. 1982. Consumptive water use by subirrigated turfgrasses
under desert conditions. Agron. J. 742419-423.

Kopec, D.M., RC. Shearman, and TR Riordan. 1988. Evapotranspiration of tall fescue turf.
HortSci. 232300-301.

Koski, AJ. 1983. Seasonal rooting characteristics of five cool-season turfgrasses. M.S.
thesis, Ohio State Univ., Columbus.

Penman, H.L. 1948. The dependence of transpiration on weather and soil conditions. J. of
Soil Sci. 1:74-89.

Ritchie, J.T. 1991. Wheat phasic development. In R. J. Hanks and J.T. Ritchie (ed.)
Modelling plant and soil systems. ASA Monograph no. 31. ASA, Madison, WI.

Saffel, MT. 1994. Time domain reflectometry and turfgrass irrigation scheduling. MS.
Thesis, Michigan State Univ., East Lansing, MI.

Shayya, W.H., and VP. Bralts. 1994. SCS-SCHEDULER. Version 3. Department of
Agricultural Engineering, Michigan State Univ., East Lansing, MI.

Shearman, RC. 1986. Kentucky bluegrass cultivar evapotranspiration rates. HortSci. 21 :455-
457.

Simpson, JR, and J .J . Meyer. 1987. Water content measurement comparing a TDR array to
neutron scattering. p. 111-114. In International conference on measurement of soil and
plant water status. I. Soils. Logan, UT.

Snyder, G.H., B.J. Augustin, and J .M. Davidson. 1984. Moisture sensor-controlled irrigation
for reducing N leaching in bermudagrass turf. Agr. J. 76:964-969.

Topp, G.C., J.L. Davis, and AP. Annan. 1980. Electromagnetic determination of soil water
content: Measurements in coaxial transmission lines. Water Resour. Res. 16:574-5 82.

Tovey, R, J .8. Spencer, and DC. Muckel. 1969. Turfgrass ET. Agron. J. 61 :863-866.
Vargas, J.M. 1994. Management of turfgrass diseases. Lewis Publishers, Ann Arbor, MI.

Wraith, J. M., and J. M. Baker. 1991. High -resolution measurement of root water uptake
using automated time-domain reflectometry. Soil Sci. Soc. Am. J. 552928-932.

Wright, J.L. 1981. Crop coefficients for estimates of daily crop evapotranspiration.
Proc. 1981 Irrig. Scheduling Conf., Chicago, IL, 14-15 Dec. 1981. ASAE, St.
Joseph, MI.

151

APPENDICES

152

 

 

Solar Radiaton. 1992

 

 

 

 

 

V.
.
.flm _
tic}! i or.
RHIIIII. I I.
lilr
slit/.1? .
qul.1t41IInHHH$rlr
It”?!
\rHulIiolifIilr-ril i“
\aIIIIIIIJIi . II
\xllrmUIHHUII. .
Ilir _. .
. inilolli.
Vlnwihlul « ..ti
0 o o 0
3 2 I
95.5% .533... 53.

268

247

226

205

184

163

142

INJY

 

1993

 

 

iii
83:141.

 

 

 

 

 

.IIIHIIHI I” L
Iiuflfldrrr
QnOIIlHHbr t .
It IIIILIil'l. II I
HIV. . .
P . .
. . .
.II. IV
I.
d J
0 o 0 0
3 2 1

2.5525 5.3.3. 55

268

247

226

205

184

163

142

121

INDY

 

1994

 

 

_ [If
. - We
. .
.nu.\s .
_ IIIILIUI.
VII...
'nlhfll'L. IU‘ -
Irrhurri _
rnnrfl . . .

I IIIHUIHI
. lrrul .
\JIWI. .
lollil. _ .

1.111.111.1111;
.
.

 

 

 

25:55 3:33: 39a

 

0

268

163 184 205 226
INDY

142

121

 

 

Fig. Al. Seasonal solar radiation at the Hancock Turf Research Center, East Lansing, M1

for 1992, 1993 and 1994.

153

 

 

Temperature (0C)

 

 

 

 

 

120 140 160 180 200 220 240 260 280
Day of the year '

 

 

JS-I

I:

Tamper awe (0C)

 

 

 

 

 

120 140 160 180 200 220 240 260 280

 

 

Temperature (0C)

 

 

 

 

 

 

120 I40 160 180 200 220
Day of the year

 

 

 

Fig. A2. Seasonal maximum and minimum temperatures at the Hancock Turf Research
Center, East Lansing, MI for 1992, 1993 and 1994.

154

70

 

60 -
50 -
1992
40 _

30-

 

20- . I
I

10- PL
0. III II

701

 

 

 

 

1993
50 a

40-
30-
20-

10- _
O- IIIIIIII In“ 11 i Jn

7o _
60 - 1994
so I
40 -
3o 1
20 _

“JIIII I. II IL I

120 140 160 180 200 220 240 260

:

Rainfall (mm)

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Day of the Year

Fig. A3. Seasonal rainfall at the Hancock Turf Research Center, East Lansing, M] for 1992,
1993 and 1994.

155

 

100

 

80-

 

601

 

Relative Humidity (7.)

 

 

 

120 140 150 rso zoo - 22b 240 260 280
. Day of the year, 1992

 

 

Relative Humidity ('/o)

 

 

 

1‘ I ................ l ,2 ............... 163' ............... 184 ............... 205 ............... 221$ ............... 2,7 .............. 268'"
Day of the year, 1993

 

 

 

Relative Humidity ('/o)

'.

:S‘
',O

I
I I
I
E
i
I

I I
I :1
I i
i i
I
l i
l ‘I ”:9”-
i t
I I
if) I

' bx r
: .421
I I

......................................................

 

 

 

 

120 140 160 180 200 220 240 260 280
Day of the year, 1994

 

 

 

Fig. A4. Seasonal relative humidity at the Hancock Turf Research Center, East Lansing, M]
for 1992, 1993 and 1994.

156

 

 

 

 

g 2 III bk Rf} I XVII/14 ‘MVJIX/EIALW/IAN’II/u I‘Ifl' - - -

 

O
a

 

 

 

 

wmvruruu)
N .
I
r E?
4"”?
N":=-~
1:1.“
"z
I a.
r ..__..
z I
11""- 7'
.3}.
.‘
. Ci"
I
2'37"?”
r, r
I i
r t
I
I
I
'2‘: .
. "u i
r r
r
(i?
e<.g—'t
X1. ‘
., ' :
:,E~ .
.23 ‘
x I
:
I
I
a-< ,
<" :
,«7
,._.. ”5“.»-
\ i
1.1:». r
Kr. I
5.1%" i
I
I

 

 

 

 

 

 

 

0120 140 160 180 Day of “2.0:". 1993 220 2‘0 250 280 i
10

8 .......................................................
\ ..............
i I. I III -
3‘ IIIJIIIIII “““ III """""""""" I A """"" I """
’ I- _.I LIE;I4,...IIJIIVI,I.I:-III..I.- III: II-._,-,._~,.‘.III

II‘ ’II III IPIIII III; III

0120 140 160 180 220 240 260 230

200
Day of the year. 1994

 

 

 

Fig. A5. Seasonal wind velocity at the Hancock Turf Research Center, East Lansing, MI for
1992, 1993 and 1994.

   
  

flWWII/111Mflfllflfliflflilsflmfllfiflfiijit

0264