Water Use and Drought Responses of Cool-Season Turfgrasses

Evapotranspiration (ET) rates of four turf grass species were compared in Rhode Island to aid in the selection of grasses with lower water requirements. ET was measured under well-watered conditions using weighing lysimeters placed into field plots of mature turf. Measurements were obtained regularly from July to September in 1984 and 1985. Average daily ET ranged from 0.23 to 0.41 cm of water/day for: Paa pratensis L. cvs. 'Baron' and 'Enmundi', Lolium perenne L. cv. 'Yorktown II', Festuca rubra var. commutata Gaud. cv. 'Jamestown', and Festuca ovina var. duriuscula L. Koch cv. 'Tournament'. Significant differences in ET rates were found between species. Kentucky bluegrass and perennial ryegrass transpired more than the fescues. Potential ET was computed using the modified Penman equation and the pan evaporation methods. Crop coefficients (KCs) were calculated to determine the predictive consistency of the methods. Seasonal KCs based on the Penman equation ranged from 0.88 to 1.09. KCs based on pan evaporation showed more variability, ranging from 0.86 to 1.35.

The response of the same turf grasses to moisture stress was investigated. Six lysimeters of each species and six well-watered control lysimeters were included in a greenhouse study; four lysimeters of each were used in a field study. The relationship between water loss due to ET and soil water potential was determined using tensiometers and electrical resistance blocks installed in separate lysimeters.
ET rates of all species remained unaffected by decreasing soil water potential until it reached -0.6 to -0.8 bars, after which ET rates declined. This decline corresponded to a decline in turf quality, growth rate, and relative leaf water content. Leaf water potential decreased 50-75% when soil water potential declined to -0.8 bars but did not continue to decrease when soil water potential became more nega-  Table 1   Table 2   Table 3   Table 4   Table 5   Table 6   Table 7   Table 8 LIST of TABLES Mean daily and seasonal ET rates of 5

INTRODUCTION
Turf grass maintenance can require the use of much irrigation water, even in the humid northeastern U.S.

3
As competition for water use increases, turfgrass culture must be directed toward practices that will lower water requirements.
Transpiration accounts for most of the water lost from a dense turfgrass canopy (1). It has been previously established that transpiration rate varies among turfgrass species (1,2,5,14,16).  cm. This is probably due, in part, to the greater humidity and cloudiness which occurred in July of 1984. The general conclusion to be drawn from these data for the two seasons is that Kentucky bluegrass cv. When soil water is readily available, turfgrass water use is usually assumed to be governed primarily by conditions external to the plant (5,7,11,20). Many  Mean values are based only on days that ET measurements were obtained . Table 4.
Average canopy density (l e aves / cm 2 ) and range i n leaf index (LAI) of f ive cool-season turfgrasses in 1984.  l/1 rt s on turfgrass ET under well-watered conditions re po C oncluded that ET is a function of meteorologic have conditions and the extent of vegetative cover (5,8,9). 16 However, the significant differences in ET between species under identical climatic conditions and cultural practices found in this study indicate that water use may also be under genetic control. This will have further implications in plant breeding for maximum water use efficiency.

Species
Differences in canopy density is one of several plant characteristics expected to influence water use rate. Increased density causes increased boundary layer resistance to convective air flow within the canopy (10).
This resistance results in a reduced saturation vapor deficit surrounding the plants in a turf stand, thereby reducing the evaporative demand which drives ET (13).
Canopy density measured on the turfgrass stands in the 1984 lysimeters are presented in Table 4. Canopy density is inversely related to water use rate for the five grasses. Those which have the greatest transpiration rates, Kentucky bluegrass and perennial ryegrass, have lower leaf densities than the fescues, which were found to use less water. The size of the leaves and thus their potential for reducing convective air flow will also vary. Kentucky bluegrass leaves range from 2 _ 4 mm wide, perennial ryegrass leaves range from 2 -5 mm wide, and the fescue leaves range from 0.5 -1 mm in width (1). The resulting differences in leaf area index (Table 4) between species will influence their water use rates by altering the boundary layer resistance of the canopy and well as alter the transpiring surface area.

II. Potential evapotranspiration:
Two predictive methods were assessed in this study for their ability to consistently estimate turf water use in southern New England.

A. The modified Penman equation method:
Average seasonal crop coefficients (KCs), These values indicate a consistent relationship between ET predicted by the equation and actual ET rates of the five grasses.
When the individual KCs are grouped and analyzed on a biweekly basis (Table  6 ) ) , more variation in the Penman methods' predictive ability is revealed.
In 1984, there is a general trend for over-prediction in July to under-prediction in September. This trend is rs ed in the 1985 biweekly analysis. The KCs for all reve · es range from 0.72 to 1.23. Given the average ET speci rate of 3.6 mm/day found in this study, this variation represents roughly five to ten mm of water transpired over a two week period, which is negligable in the context of an irrigation scheduling program.
It is concluded that the modified Penman equation   Average seasonal crop coefficients (KCs) for five cool-season tur f g r a sses in 198 L1 and 1985, based on the pan evaporation met hod.
--·-  well-watered through out the experiment, to compare ET rates of the drought stressed grasses with paten-tial ET rates when water was not limited.
The grasses were subjected to two successive drought stress periods. The first stress period was continued until the grasses showed visible signs of stress (quality scores below 6.5), after which they were allowed to recuperate under well-watered conditions for three weeks. Quality scores were recorded every three days.
Scores range from a perfect score of 9, representing dense, green, turgid grass cover, to a low of 1 when the grass appears dead. A score of 6.5 or above was considered acceptable turf quality. For the purposes of this study, drought tolerance was defined as the ability of a turfgrass species to maintain acceptable quality while under drought stress.
The grass in each lysimeter was mowed to a height of 5 cm every three days. The clippings were harvested, and both wet weight and dry weight was measured. Leaf growth rate was monitored on a gram DW clippings/m 2 / day basis. The water content of the clippings (gm WWgm DW) was divided by the water content of the clippings from the same species at full turgor to provide a relative leaf water content (RLWC) index.
Leaf water potential of the grasses in the greenhouse experiment was measured using a pressure chamber Under non-stressed conditions, above -0.6 bars, hard fescue transpired less rapidly than the other three grasses. This is consistent with previous research into comparative water use rates of cool-season grasses under well-watered conditions (1,2,8,9) These findings are consistent with Beard's catagorization of drought resistance in cool-season turfgrasses. He rated hard fescue and red fescue "good", Kentucky bluegrass "medium", and perennial ryegrass "fair" in overall resistance to drought stress (1).
The same trend was evident when quality scores were analyzed (Figure 2). The visual qualtiy of all the grasses declined when soil water potential fell below -0.6 bars. Lowered quality scores have previously been reported for cool-season grasses exposed to drought stressed conditions (6) although no threshold soil water potential was correlated with this decline.
Turf quality and Clipping growth: Neither Kentucky bluegrass nor perennial ryegrass sustained acceptable turf quality (a score of 6.5 or above) under moisture stress. Red fescue again ranked intermediate, and hard fescue maintained acceptable  Quality ratings of four cool-season turfgrasses subject to declining soil water potential.

44
Reduced growth rate of turfgrass leaf tissue during drought stress has been cited repeatedly (1,4,6). A decline in leaf growth rate of all species occurred from the onset of the moisture stress test, before ET rate or quality declined (Figure 3) These values represent the water content of the clippings, which by definition must be at least partially turgid since no flaccid tissue was harvested using our system. RLWC is therefore considered to be a partial index of leaf turgidity, although clipping Yields must also be taken into account.

Field Drought Stress Test
The 1985 field drought stress test was conducted in late June -early July, during a period of frequent rainfall events. This presents numerous difficulties in interpreting the results. Data from the well-watered control lysimeters, located in a separate plot area which was not as well protected from precipitation events were not available for much of the stress period.
Relative ET rates are therefore difficult to calculate and cannot be compared to those computed from the greenhouse study.
In addition, since rain shelters were required to cover the lysimeters throughout much of the experiment, the lysimeters were not exposed to legitimate "field"    (1,2,13,32,34), and in some cases even between cultivars within a species.
In 1941, N.L. Partridge demonstrated substantial differences in ET rates between ten grass species under wellwatered conditions, although no statistical analysis of the data was indicated (32). Of the cool-season grasses common to both his study and the present research, Kentucky bluegrass was found to use consistently more water than the fescues. Feldhake (16) found Kentucky bluegrass to use 24 percent more water than bermudagrass. This difference is due primarily to the well-documented differences in water use between warm-and cool-season grasses (3,25,27 Climatic variability, particularly in humid regions, requires a relatively long time period to make reasonably accurate estimates using the pan method (45). The calibration of the pan to a given site is also mandatory (11). The pan evaporation method has a tendancy to lag climatic conditions due to the high specific heat of water. The relationship between soil water content and plant water use has been presented in a variety of models, in an attempt t .o determine the soil moisture content at which the actual transpiration rate falls below the potential rate, and whether this can be predicted for any soil-plant-weather combination.
Water moves through the soil to plant roots and through the plant to the transpiring leaves along a gradient of negative pressure (water potential).
Gardner (19) presents an equation for the flow of 60 water from the soil to the roots of a transpiring plant.
He concluded that the water potential gradient between soil and root needed to maintain a given transpiration rate is proportional to the rate of water uptake or the potential transpiration rate, and inversely proportional to the capillary conductivity of the soil.
As the soil dries, large suction (negative pressure) gradients develop between the root and the soil. To maintain transpiration in a drying soil where capillary conductivity is rapidly decreasing and the water potential of the root is decreasing correspondingly, the water potential of the leaves must decline even further to maintain the necessary suction gradient.
Decreased leaf water potential coincides with decreased turgor pressure, which leads to stomatal closure.
This reduces the permeability of the leaf surface to water flow and hence reduces transpiration rate.
The soil moisture content at which transpiration rates are reduced depends on a variety of interactive factors. Morphological and physiological attributes of the plant (rooting depth, stomatal density, etc.) play a major role. Meteorologic conditions also have an influence. Increased evaporative demand will cause leaf water potential to decline more rapidly, leading to a more rapid decline in turgor and transpiration rate (20).
Soil properties which influence the relationship between soil moisture content and soil water potential determine the the quantity of water "available" to the plant.
Numerous models have been proposed to describe the relationship between soil moisture and ET rates.
V~ihmeyer and Hendrickson (SO) proposed that ET rates remain unaffected by decTeasing soil moisture until the level of soil moisture approaches the wilting point, at which time ET rate falls rapidly. Thornthwaite (46) stated that ET rates will be half the maximum for the prevailing meteorolgic conditions at a soil moisture content of half the available water. Other models postulate that ET declines linearly with a decrease in soil water from field capacity to permanent wilting point (20). Still others, such as Pierce (35) proposed a logarithmic relationship between soil water and ET.
The limitations, and hence incompatibility, of these models rest l~r~~ly on their o~ission o~ var~ iabilty in soil properties and climatic conditions.
The former contributes significantly to soil water availability to plants, and the latter directly influences plant response to soil water, or potential water use rate by the plant.
Denmead and Shaw (9)  Wilting is caused by a decrease in leaf turgor associated with a reduction in leaf water content.

63
This occurs under drought stress conditions when transpiration rate exceeds the rate of soil water extraction by the roots. Relative leaf water content (RLWC) is the water content (on a percent basis) of leaf tissue relative to the water content of the tissue when turgid. This measurement has been positively correlated to leaf water potential, and has been used as a plant water status index (15,22,30).
A change in ,leaf color, ranging from grey to bluegreen, frequently accompanies wilting. This color change will lower the visual quality rating of turf-grasses. A limited degree of water stress has been shown to have no adverse effect on turf grass quality (10,29). Feldhake et al. (17)' in a study involving Kentucky bluegrass, found that an irrigation deficit of 27% will only decrease growth, whereas larger deficits cause quality to decline rapidly. A reduction in turfgrass growth resulting from drought stress has been reported (1,13).
Total leaf water potential has gained wide recognition as a measure of plant water status. Total leaf water potential results from the combined but opposite actions of pressure (turgor) potential (ljlp) and osmotic potential (~'Tl"). The relationship between these components as volume changes is schematically described in the so-called Hofler diagram (40). The pressure bomb technique developed by Scholander (42) is regarded as an accurate method for estimating leaf water potential (4,30,48). The lack of references to this method in turfgrass literature reflects the difficulty in adapting the technique to fine-leaves grass species.
The temperature of a turfgrass canopy is expected grasses. They found leaf temperature differences in the stressed and non-stressed C-3 grasses to be almost 5°C, and almost 8°C differences between stressed and non-stressed C-4 grasses. Feldhake et al. (17) found that turfgrass canopy temperature increases l.7°C for each ten percent decrease in irrigation regime up to a 70% decrease.