Leaf wetness in dry beans under semi-arid conditions

Agricultural and Forest Meteorology, 48 (1989) 149-162

149

Elsevier Science Publishers B.V., Amsterdam - - Printed in The Netherlands

LEAF WETNESSIN CONDITIONS*

DRY BEANS UNDER SEMI-ARID

A. WEISS 1, D.L. LUKENS 2, J.M. NORMAN ~ and J.R. STEADMAN 4

1Center for Agricultural Meteorology and Climatology, University of Nebraska, Lincoln, NE 68583-0728 (U.S.A.) 2Rocky Mountain Forest and Range Experiment Station, Fort Collins, CO 80526-2098 (U.S.A.) 3Department of Soil Science, 1525 Observatory Drive, University o[ Wisconsin, Madison, WI 53706 (U.S.A.) 4Department of Plant Pathology, University of Nebraska, Lincoln, NE 68583-0722 (U.S.A.) (Received September 16, 1988; revision accepted February 4, 1989)

ABSTRACT Weiss, A., Lukens, D.L., Norman, J.M. and Steadman, J.R., 1989. Leaf wetness in dry beans under semi-arid conditions. Agric. For. Meteorol., 48: 149-162. Leaf wetness measurements were made in and at the top of canopies of dry edible bean (Phaseolus vulgaris L., cvs. A55 and Midnight) for two growing seasons. Two similar approaches were used for this measurement; a cotton cloth (acting as an artificial leaf) and a real leaf were placed on a grid network of independent, adjacent fine wires. This configuration acted as a variable resistor, infinite resistance when dry and decreasing resistance as the leaf became wet, as part of an electronic circuit with an output signal proportional to the degree of wetting. Atmospheric humidity was measured in and above the canopies using fine wire psychrometers. Difficulties encountered in making representative measurements of leaf wetness duration inside dry edible bean canopies were due to problems of representative location and the heliotropic nature of bean leaves. After sunrise, this latter phenomenon could randomly distribute water from the uppermost leaves to locations inside the canopy. Measurements made with the cotton cloth at the top of the canopy had the greatest duration of leaf wetness. This sensor and measurement location are ideally suited for making disease predictions where duration of leaf wetness is an important climatological input. In addition, predictions of leaf wetness duration from the comprehensive, mechanistic plant environment model, Cupid, are included.

INTRODUCTION

Leaf wetness can regulate infection by fungal or bacterial pathogens (e.g., Palti, 1981; Steadman, 1983; Royle and Butler, 1986). The term "leaf wetness" encompasses all forms of liquid water on leaf surfaces whether caused by radiative cooling (dew), precipitation or irrigation. Correlative data on leaf wet*Published as paper No. 8020, Journal Series, Nebraska Agricultural Research Division.

0168-1923/89/$03.50

© 1989 Elsevier Science Publishers B.V.

150 ness from measurements made outside the canopy or by hygrothermographs inside the canopy have been used with moderate success to predict disease outbreaks and guide preventative measures (Wallin, 1967; Nutter et al., 1983 ). However, there is little accurate field data available from semi-arid environments on leaf wetness within the plant canopy. In attempting to quantify the microclimate of contrasting canopies of dry edible bean (Phaseolus vulgaris L.), Weiss et al. (1980) were not able to accurately measure leaf wetness and atmospheric moisture within the plant canopy. This early lack of success provided the stimulus for the development of a sensor to measure leaf wetness (Weiss and Lukens, 1981; Weiss and Hagen, 1983). Other sensors to measure leaf wetness have since been described by Sutton et al. (1984), Gillespie and Howard (1985) and Barthakur (1985). A detailed review of earlier leaf wetness sensors has been given by Schnelle et al. (1963). The objective of the field experiment was to determine the optimum location in a dry bean canopy for a representative measurement of leaf wetness that can be used for disease prediction schemes. When comparing the duration of leaf wetness as measured by different sensors, assuming there are no spurious signals, the sensor which yields the longest duration is most sensitive to this climatological parameter. Successful prediction of the duration of leaf wetness can be made from simple empirical models based on a given crop, location and environmental conditions. However, these models do not work well when predictions are made under conditions which differ from the original data set. These poor predictions occur because of the many factors involved in dew formation and evaporation, such as soil thermal and water properties, soil water content, radiation and temperature conditions on previous days, leaf size, leaf area density, canopy height, the vertical distribution of leaf area, wind speed, atmospheric humidity and temperature, and night-time cloudiness. The objective of the numerical experiment was to compare field measurements of leaf wetness with predictions from a comprehensive, mechanistic plant environment model, Cupid (Norman, 1979; Norman and Campbell, 1983 ). Input data requirements for Cupid are given in Norman and Campbell (1983). MATERIALSAND METHODS

Field experiment The 1984 and 1985 field data were collected at the Agricultural Meteorology Laboratory of the University of Nebraska Panhandle Research and Extension Center (41 ° 51' N, 103 ° 41' W; 1225 m above msl) located ~ 9 km northwest of Scottsbluff, NE. In 1984, six cultivars or breeding lines of dry beans (Midnight, Tacaragua, A55, Florida 72, Black Turtle Soup-Selection 3 and Great North-

151

ern Tara) were chosen for their range of upright to prostrate canopy architectures. On 6 June, the cultivars were planted in n o r t h - s o u t h rows with a row spacing of 0.76 m. There were three sections of beans each 27.4 X 15.2 m in size with two plots of 13.7 X 15.2 m per section. Six rows on the east and west sides of the field were considered border rows. No measurements were taken in these rows, although they received the same management (irrigation, pesticides, etc. ) as the remainder of the plots. Two black-seeded, indeterminate cultivars (A55 and Midnight) were chosen in 1984 for their upright canopy architecture. On 11 J u n e 1985, the two cultivars were planted in n o r t h - s o u t h rows with a row spacing of 0.56 m. Each plot was 24.4 m wide and 18.3 m long. Microclimate measurements were made near the center of each plot. P l a n t samples were taken far enough away from the sensors so as not to affect the microclimate and also away from the plot borders as was done in 1984. Two methods were used to measure leaf wetness. The first method used a cotton cloth, acting as an artificial leaf on a grid network of fine wires (Weiss and Lukens, 1981). This grid network consisted of two stainless steel wires supported on a frame so that adjacent wires were independent. The sensor behaved as a variable resistor, infinite resistance when dry and decreasing resistance as the cotton cloth became wet, in an electronic circuit that produced an output signal proportional to the degree of wetting. In 1984, these sensors were located 0.10 m above the soil line at the plant stem in the vicinity of the foliage at that height. Once the sensors within the canopy were operational, it became evident that these sensors were not yielding reliable data on the actual duration of leaf wetness. While the cotton cloth was very sensitive to the onset of dew, it also indicated dew when visual observations of leaves indicated that they had dried. An actual leaf was substituted for the cotton cloth on each sensor (Fig. 1 ). An additional sensor was placed at the top of the A55 canopy. The petiole of the leaf was placed in a sealed vial of distilled water. This vial was attached to the handle of the leaf wetness sensor and was refilled as needed by injecting distilled water into the vial via the soft plastic cap. Using this procedure, leaves were kept alive for a period of 2-7 days. We did not investigate the use of nutrient solutions in place of distilled water or keeping the excised leaves in an artifically high humidity environment for a day prior to their use on the leaf wetness sensor. This latter procedure would allow the leaves to become fully turgid before field use. From the experiences of data collection in 1984, in the following year (1985) leaf wetness was measured by attaching a grid network of fine wires to an intact bean leaf at 0.1 and 0.3 m above the base of the plant, and at the top of the plant canopy. In addition, leaf wetness was measured at the top of the canopy with a cotton cloth on a grid network of fine wires. Sensors at the top of the canopy were adjusted in response to changes in canopy height.

152

Fig. 1. Leaf wetness sensor with a n actual leaf on grid wires located at the top of the A55 dry bean canopy.

Whether using a cotton cloth or a leaf on the grid network of fine wires, each sensor was individually calibrated. This calibration was accomplished at midday by thoroughly wetting the leaf or cloth with distilled water and adjusting the output voltage so that all sensors had the same output. Whenever a leaf or cotton cloth had to be changed, this calibration procedure was repeated. The grid network of fine wires with attached leaves were all inclined at a slight angle so water would not form a pool on the leaf. A naturally ventilated thermocouple psychrometer, similar to that described by Mitchell et al. (1973), was constructed using 0.025-mm diameter welded copper-constantan thermocouples. A small container of distilled water was kept above the dry- and wet-bulb thermocouples. This psychrometer differed from the design of Mitchell et al. (1973) by having the feeder wick to the wet-bulb thermocouple in the container of distilled water go through a piece of Tygon tubing which in turn was friction fitted into a small acrylic tube. A small hole was drilled and tapped in the acrylic tube and fitted with a plastic screw. This screw, by constricting the Tygon tubing, controlled the flow of water to the wick. The thermocouple wires were not protected and, on several occasions, were damaged by insects or fluttering leaves. In 1984, psychrometers were also placed in each plot at 0.10 m above the

153 base of the plant in the center of the furrow. An additional psychrometer was placed at 2 m above the base of the plant in the A55 plot. Soil temperatures were measured in each plot at three positions on the same side of the furrow at 0.10-m depth by three thermocouples of equal resistance connected in parallel. In 1985, psychrometers were placed at 0.1 and 0.3 m above the base of the plant in each plot, and at 2 m above the base of the plant in the Midnight plot. Soil temperatures, the mean of three thermocouples as in 1984, were measured at 0.04 m in the center of the furrow and by the base of the plant. Data were collected with a Campbell Scientific CR5 data logger and recorded on tape for later analysis. Data for the 24-h periods beginning at noon on 27 August-1 September 1984 and 22 August-2 September 1985 will be discussed. No pesticides were applied during these periods. The final irrigation in 1984 was made on 27 August from 15.30 to 20.00 h and in 1985 starting at 16.45 h on 26 August and ending at 08.00 h on 27 August.

Numerical experiment The 1984 field measurements of leaf wetness duration were compared to predictions by Cupid. Cupid differentiates between causes of leaf wetness; i.e., it is possible that no dew (condensation) will be predicted, but leaf wetness is indicated either from precipitation or sprinkler irrigation. Cupid calculates dew from a solution of the leaf energy balance for leaves in various leaf angle classes and canopy layers (Norman, 1979). If the vapor pressure of the air is greater than the saturation vapor pressure at the temperature of the leaf, then dew is assumed to form on both sides of the leaf and water begins to be accumulated. Clearly, the vertical profiles of air temperature, vapor pressure, net radiation and boundary layer resistance are critical to the formation of dew. When dew occurs, the stomatal resistance is set to zero. The boundary layer resistance is a function of the wind speed profile in the canopy, the latter being calculated by the method of Landsberg and James (1971). The depth of water on the leaf is either increased or decreased during each time step calculation until the leaf is again dry. RESULTS AND DISCUSSION

Field experiment On 18 June 1984, a severe thunderstorm (47 mm of precipitation in a 2-h period) caused damage to the beans and flooding in the field. Four days later, another thunderstorm (12 mm of precipitation in 1 h) passed over the area with similar, although not so devastating, results. Where necessary, beans were

154

replanted on 26 June. Only the Midnight and A55 plots developed a uniform canopy. On 21 August 1984, the Midnight and A55 plots had similar geometries: leaf area indices of 2.2 and 2.1, heights of 0.61 and 0.70 m, heights at the widest part of the row of 0.28 and 0.37 m, and width at the widest part of the row of 0.76 and 0.71 m, respectively. Canopies with exposed soil surfaces between distinct rows were considered as "open" while in "closed" canopies the soil surface was not uniformly visible and distinct rows were difficult to discern. Both Midnight and A55 plots were considered "open" canopies. Changing the row spacing, from 0.76 m in 1984 to 0.56 m in 1985, had a major influence on the geometry of the plant. On 26 August 1985, the leaf area indices of the A55 and Midnight were 4.0 and 5.2, respectively. The plant heights, heights at the widest part of the row and widths at the widest part of the row were: 0.62 and 0.58 m, 0.27 and 0.27 m, and 0.52 and 0.60 m for the A55 and Midnight plots, respectively. The A55 canopy was "open" while the Midnight canopy was "closed". Table 1 shows the duration of leaf wetness in the Midnight and A55 plots and at the top of the A55 plot in 1984. Data for these time periods, considered representative of a mature canopy, were selected because there were no major instrument failures in any of the plots. For the first four time periods (27/2830/31 August), there was a distinct pattern; short durations of leaf wetness (dew) were observed at 0.10 m in the Midnight plot compared to the A55 plot and there was good agreement between measurements in and at the top of the A55 plot. Rain (3 mm) fell between 19.00 and 22.00 h on 31 August and similar TABLE1 Total duration from noon to noon of measured leaf wetness (hours and tenths) in dry bean canopies (Midnight and A55) at Scottsbluff, NE in August and September 1984. Predictions from Cupid of the duration of leaf wetness in the A55 canopy for the same time periods as the field measurements at the highest (0.48-0.70 m) and lowest (0.10-0.31 m) canopy levels Height (m)

A55 0.1 0.7 Midnight 0.1 Predictions 0.10-0.31 0.48-0.70

Date 27/28

28/29

29/30

30/31

31/1

12.7 10.2

6.8 5.0

7.5 6.8

2.2 0.7

14.3 12.2

0.7

1.5

1.5

-

13.8

7 4

7 7

12 8

0 0

12 12

155

durations of leaf wetness were observed in all locations during the period 31 August-1 September. Table 2 illustrates the distribution of leaf wetness in the A55 and Midnight canopies from 23/24 August to 1/2 September 1985. After the second day of data collection, the leaf wetness sensor at 0.3 m in the Midnight plot was not responding as well to leaf wetness as were the other sensors. Less than 1 mm of rain fell between the periods 15.00-16.00 h on 28 August and 17.00-18.00 h on 1 September 1985. A general pattern of leaf wetness duration emerges: duration increases as height in the canopy increases and, at the top of the canopy, the grid network using a cotton cloth measures a greater duration of wetness than the grid network using a leaf. This pattern of leaf wetness duration can be explained by the following. When small droplets of water form on the cotton cloth, the water is readily absorbed by the cotton and the electronic circuitry "senses" the presence of moisture. When an actual leaf is attached to the grid wires, there must be a continuity of water between the adjacent wires before moisture is detected. Depending on the rate of dew formation, it is possible in the early stages of this process that discrete droplets form and do not provide a continuous path between adjacent grid wires. In this situation, moisture is present on the leaf, but is not detected. This design problem can be remedied by developing grid networks with a smaller spacing between adjacent wires for use on actual leaves. During the latter part of the experiment, there were several days when the TABLE 2 Total duration from noon to noon of leaf wetness (hours a n d t e n t h s ) in dry bean canopies (Midn i g h t a n d A55) at Scottsbluff, NE, in August a n d S e p t e m b e r 1985. T h e m e a s u r e m e n t s at 0.1 and 0.3 m, a n d at the top of canopy ( T O C - L ) were made with a n actual leaf on a grid network of fine wires while a second m e a s u r e m e n t at the top of the canopy (TOC-C) was made with a cotton cloth on a grid network of fine wires Height

Date

Total

23/24

24/25

25/26

26/27

27/28

28/29

29/30

30/31

31/01

01/02

A55 0.1 0.3 TOC-L TOC-C

4.8 7.3 5.5 11.0

5.5 6.8 8.2 11.5

2.5 3.2 6.7 10.0

0.3 0.2 4.0 8.8

2.8 4.5 11.3 14.2

2.0 7.5 8.6 14.0

3.0 3.7 11.8 13.0

0.7 0.0 4.3 3.5

0.0 0.0 0.0 0.0

0.0 10.5 15.0 16.0

21.6 43.7 75.4 102.0

Midnight 0.1 0.3 TOC-L TOC-C

5.0 3.2 8.8 11.5

6.0 4.3 6.0 10.7

4.3 0.0 3.0 8.3

1.5 0.0 1.2 5.5

7.3 0.0 10.0 12.8

2.2 0.2 10.3 12.3

4.0 0.7 10.0 8.7

2.0 0.0 4.3 2.2

0.0 0.0 0.0 0.0

9.7 0.0 16.8 14.5

42.0 8.4 70.4 86.5

156 duration of leaf wetness, as measured by the leaf on the grid network, exceeded measurements made by the cotton cloth on the grid network. These occurrences may be attributed to the placement of the sensors at the "top of the canopy". The sensors were placed in the uppermost 0.05 m of the canopy and, as the plants grew and the leaves responded to sunlight, it is possible t h a t a microenvironment was created in the vicinity of the cotton cloth which differed from t h a t around the leaf attached to the grid network of fine wires. From a disease forecasting perspective, these differences in leaf wetness duration are small. Figure 2 illustrates hourly dew point depression (hourly average difference between the air and dew point temperature) at 0.10 m in the Midnight and A55 plots, and at 2 m above the A55 plot on 28/29 August 1984. There is excellent agreement between the dew point depression measured in both plots. The dew point depression at 2 m, although greater, follows the same trend as occurred within the canopies. Two periods of dew were recorded at 0.10 m in the A55 plot during this 24h period (19.30-20.50 and 02.00-06.50) and they closely correspond to the times when the dew point depression was ~<2 ° C. During the first and second dew periods, the average hourly wind speed recorded at a nearby automated weather station ( < 4 0 0 m from the field site) was ~0.8 and 0.72 m s -1, respectively. Before the first dew period and after the second dew period, the wind speed was > 1.5 m s-1, while between the periods the average wind speed was 1.7 m s-1. These greater wind speeds increased the turbulent mixing of the air, preventing the leaves from cooling to below the dew point temperature of the air. 28129 AUGUST1984 10 CM MIDNIGHT 10 CM A55 70 CM A55 24 0""

2O

~-~

16

t'~ft

I--

2M

•

"

~

12

4 o

. . . .

13

15

17

''~'~' , ,

19 21

23

1

' ~ / 3

5

.....

7

9

11

TIME (MST) Fig. 2. Hourly averages of dew point depression at 0.10 m above the soil line at the plant stem in the Midnight and A55 plots, and at 2 m above the soil line at the plant stem in the A55 plot on 28/29 August 1984. Also indicated are periods of leaf wetness at different levels (0.1 m in the Midnight cultivar, and 0.1 and 0.7 m in the A55 cultivar) measuredby an excisedactual leaf on a grid network of fine wires.

157

MIDNIGHT

=81~,uousT1 ~

COTTON CLOTH LEAF

-

--

2O

o 13

15

17

19

21

23

1

3

5

7

9

11

TIME ( M S T ) Fig. 3. Hourly averages of dew point depression at 0.10 and 0.30 m above the soil line at the plant stem in the Midnight plot, and at 2 m above the soil line at the plant stem on 28/29 August 1985.

Also indicatedare the periods of leafwetnessat the top of the canopyas measuredby the cotton cloth and an actual leafon a grid networkof fine wires. Figure 3, the hourly dew point depression from the Midnight plot for 28/29 August 1985, illustrates the same major feature found in Fig. 2; when the dew point depression increased, the sensors at the top of the canopy dried off and vice versa. Wind speeds representative of the former conditions at 21.00 and 01.00 h were 5.1 and 2.1 m s - 1, respectively, while wind speeds associated with periods of small dew point depression and leaf wetness conditions at 19.00 and 00.00 h were 1.2 and 1.0 m s-1, respectively. In addition, the variation and magnitude of the dew point depression at 0.1 and 0.3 m were almost the same. These similarities of dew point depression at the two heights are typical of other 24-h periods, with the exception that the magnitudes measured at the two heights were not always in such close agreement. The data from Tables 1 and 2 illustrate the difficulty in making representative measurements of leaf wetness duration inside the canopy as compared with measurements made at the top of the canopy. Sensor placement is critical in making these within-canopy measurements. Data in Fig. 2 indicate that dew point depressions in both the Midnight and A55 canopies were similar; yet the sensors indicated a leaf wetness duration ~ 6 times greater in the A55 canopy than in the Midnight canopy (Table 1 ). However, when leaf wetness was caused by precipitation (31 August-1 September 1984) rather than dew formation, as in the previous example, both sensors indicated near equal durations of leaf wetness (Table 1). Measurements of the duration of leaf wetness by the cotton cloth on the grid network of fine wires caused by rain compared favorably to visual observations on leaves (Weiss and Lukens, 1981 ). On the days that rain fell in 1985 (28/29 August and 1/2 September in Table 2), a general pattern of the duration of

158

leaf wetness occurred in both canopies. In the A55 canopy, the lowest durations were at the 0.1-m height while leaves at the 0.3-m and top of the canopy levels had durations of near equal magnitude, and the cotton cloth on the grid network of fine wires had the highest duration of leaf wetness. In the Midnight canopy, durations of leaf wetness at the top of the canopy were almost equal, while at the 0.1-m height durations of wetness were much lower. Another difficulty in making representative measurements of leaf wetness duration inside a dry edible bean canopy is attempting to account for the heliotropic behavior of the leaves. At night, the blade of the leaf is pointed toward the soil and if the proper conditions are present (clear skies, adequate soil moisture and near calm wind conditions), dew will form. After sunrise, the blade of the leaf points toward the sun (Fig. 4) and any water collected on the surface of the blade runs down the petiole to be randomly dispersed within the plant canopy. Once inside a "closed" canopy, the liquid water will remain for longer periods of time than in an "open" canopy where sunlight and wind have a greater potential to heat plant surfaces and the air, and thus evaporate any free moisture. Indirect evidence of the influence of "open" and "closed" canopies on leaf wetness duration comes from Fuller et al. (1984). They used a cultivar of dry edible bean (GN Tara) which normally has a "closed", sprawly-type architecture and is highly susceptible to white mold disease (Sclerotinia sclerotiorum Lib. de Bary). When this canopy was modified so it would have several values of "openness", the most "open" canopy had the least infection, i.e., the shortest duration of leaf wetness.

Numerical experiment The model predictions and field measurements of leaf wetness duration as a function of height are in excellent agreement (Table 1 ). Cupid assumes a flat soil surface and when comparisons are made against a crop grown in an undulating row/furrow geometry, the height reference is not clear; this causes uncertainty about predictions of dew at the lowest level in the canopy. Thus, the lowest layer in the model predictions is from 0.10 to 0.31 m. For the first 2 days, measurements and predictions follow the same trend, a greater duration of leaf wetness in the lower part of ~he canopy compared to the upper part. On the third day, the measured and predicted durations of leaf wetness at both levels in the canopy are almost equal. The same pattern of the greater duration of leaf wetness measured in the lower part of the canopy continued on the fourth day. Cupid predicted no dew formation at either level. It rained (3 m m ) on the evening of 31 August, resulting in the increase in leaf wetness duration from the previous day with the model predictions showing equal durations at both levels in the canopy, while measurements indicated a greater duration at the lower height.

159

Fig. 4. (a) Leaves of the dry bean cv. M i d n i g h t during the day pointing toward the sun. (b) Leaves of the dry bean cv. M i d n i g h t just at sunrise. Most of the leaves are pointing toward the soil surface while a few leaves are responding to the sun.

160 TABLE 3 The presence of dew measured at the top of the canopy compared to predictions of dew in the top layer of the canopy for selected hours on 27/28 August 1984 (measured hourly wind speeds and relative humidities, and calculated total canopy amounts of leaf wetness per ground area are also included) Hour

Dew at top of canopy

Wind speed (m s -1 )

Relative humidity

Dew at top layer of canopy

Total canopy amount (mm/ground area)

17 18 19 20 21 22 23 24 1 2 3 4 5 6 7 8

N Y Y Y Y Y Y N N Y Y Y Y Y N N

1.8 0.9 1.0 0.9 0.7 0.8 1.2 1.3 1.3 0.7 0.6 0.7 0.6 0.8 1.6 1.8

28 43 58 59 65 54 46 52 51 67 67 65 65 61 54 51

N N N N Y Y Y N N Y Y Y Y Y N N

0 0 0.01 0.03 0.07 0.11 0.10 0.10 0.09 0.14 0.18 0.22 0.27 0.24 0.09 0

T a b l e 3 s h o w s w h e t h e r dew f o r m e d (Yes or N o ) , w i n d s p e e d a n d r e l a t i v e h u m i d i t y m e a s u r e d 2 m a b o v e t h e b a s e o f t h e p l a n t s in t h e A55 c a n o p y a n d m o d e l p r e d i c t i o n s of dew a m o u n t s ( m i l l i m e t e r s , b a s e d o n g r o u n d a r e a ) for a n h o u r p e r i o d e n d i n g a t t h e t i m e i n d i c a t e d for 2 7 / 2 8 August. T h e m e a s u r e d t i m e s of dew o c c u r r e d w h e n t h e dew p o i n t d e p r e s s i o n w a s ~<2 ° C a t 0.10 m a b o v e t h e b a s e of t h e p l a n t a n d t h e w i n d s p e e d w a s ~< 1 m s - 1 a t 2 m a b o v e t h e b a s e of t h e p l a n t . T h e r e l a t i v e h u m i d i t y n e v e r e x c e e d e d 67% d u r i n g t h i s t i m e . A t t h e e n d of i n d i v i d u a l dew p e r i o d s (23.00 a n d 06.00 h ) , t h e signals f r o m t h e l e a f w e t n e s s s e n s o r s w e r e v e r y low a n d i n d i c a t e d w e t n e s s for o n l y 30 m i n of e a c h hour. A t t h e t i m e t h e leaves w e r e dry, t h e a m o u n t s of dew p r e d i c t e d b y C u p i d also decreased. W h e n a t m o s p h e r i c h u m i d i t y is r e l a t i v e l y low, w h i c h o c c u r s u n d e r s e m i - a r i d c o n d i t i o n s s u c h as t h o s e o c c u r r i n g in t h e field e x p e r i m e n t , t h e soil is t h e p r i m a r y s o u r c e o f w a t e r for dew f o r m a t i o n . C o n s i d e r i n g t h e r e w e r e no c a l i b r a t i o n s m a d e to Cupid, t h e m e a s u r e d a n d p r e d i c t e d v a l u e s of l e a f w e t n e s s d u r a t i o n w e r e in e x c e l l e n t a g r e e m e n t . CONCLUSIONS P r e v i o u s l y , it w a s f o u n d t h a t a c c u r a t e m e a s u r e m e n t s of l e a f w e t n e s s in alfalfa (Medicago sativa L. ) c o u l d be m a d e inside t h e c a n o p y w i t h a c o t t o n c l o t h

161 on a grid network of independent adjacent wires (Weiss and Lukens, 1981; Weiss and Hagen, 1983). This sensor and measurement location did not give representative measurements of leaf wetness in a canopy of dry edible bean, a row crop. Instead, it was found that representative measurements of leaf wetness could be made at the top of the canopy with the cotton cloth on the grid network of fine wires. This result has been extended to make successful predictions of Cercospora leaf spot (Cercospora beticola Sacc. ) in sugar beet (Beta vulgaris L.) (Weiss and Kerr, 1989). Rarely do plant diseases occur uniformly over a field (Hughes, 1988). The development of a disease at a focus depends on the classical disease triangle: a virulent pathogen, a susceptible host and the proper environment (Agrios, 1988). In addition, these events must occur in the proper sequence, e.g., spores may be present on a leaf but leaf temperature may be too low or the duration of leaf wetness may be insufficient for spores to germinate and infect a leaf; or the opposite may be true, environmental conditions are conducive to germination and infection, but spores are not present. Thus it is difficult to make generalizations about measurement location and duration of leaf wetness for specific pathogens in field crops. Prediction schemes, which can vary from simple "take action, don't take action, wait and see" choices to complex models that predict yield loss, should be developed for plant pathogens of economic importance. There is another measurement technique that has been reported to be more accurate than the cotton cloth on a grid network of fine wires (Barthakur, 1987 ). However, the cotton cloth sensor is well suited for practical applications of disease prediction where cost, durability and ease of operation are important factors. Field experiments and computer modeling are complementary approaches to understanding. Predictions of leaf wetness from the comprehensive, mechanistic plant environment model, Cupid, were in excellent agreement with field measurements. With this agreement, one can have confidence, for example, in the simulated amount of dew formed per unit of ground surface area (Table 3 ) which is very difficult to measure under field conditions. ACKNOWLEDGMENTS Drs. N.J. Rosenberg, S.B. Verma and K.G. Hubbard provided valuable comments during the preparation of this manuscript.

REFERENCES Agrios, G.N., 1988. Plant Pathology,3rd Edn. AcademicPress, New York,803 pp. Barthakur, N.N., 1985. A comparativestudy of radiometricand electronicleaf wetnesssensors. Agric.For. Meteorol.,36: 83-90.

162 Barthakur, N.N., 1987. Comparison of beta-ray and electronic leaf wetness sensors in a dew chamber. Int. J. Biometeorol., 31: 57-63. Fuller, P.A., Steadman, J.R. and Coyne, D.P., 1984. Enhancement of white mold avoidance and yield in dry beans by canopy elevation. Hort. Sci., 19: 78-79. Gillespie, T.J. and Howard, A.M., 1985. Cylindrical sensors for surface wetness duration. Preprint volume 17th Conference of Agricultural and Forest Meteorology and Seventh Conference of Biometeorology and Aerobiology, 21-24 May 1985, at Scottsdale, AZ. American Meteorology Society. Hughes, G., 1988. Spatial heterogeneity in crop loss assessment models. Phytopathology, 78: 883884. Landsberg, J.J. and James, G.B., 1971. Wind profiles in plant canopies: Studies on an analytical model. J. Appl. Ecol., 8: 729-741. Mitchell, S.T., Kimball, B.A. and Ehrler, W.L., 1973, A miniature gravity-fed thermocouple psychrometer. Agron. J., 65: 238-239. Norman, J.M., 1979. Modeling the complete crop canopy, pp. 249-277. In: B.J. Barfield and J.F. Gerber (Editors), Modification of the Aerial Environment of Crops. ASAE, St. Joseph, MI, 538 pp. Norman, J.M. and Campbell, G., 1983. Application of a plant-environment model to problems in irrigation. In: D. Hillel (Editor), Advanced in Irrigation, Vol. 2. Academic Press, New York, pp. 155-188. Nutter, F.W., Cole, H., Jr. and Schein, R.D., 1983. Disease forecasting system for warm weather Pythium blight of turfgrass. Plant Dis., 67:1126-1128. Palti, J., 1981. Cultural Practices and Infectious Crop Diseases. Springer-Verlag, Berlin, 243 pp. Royle, D.J. and Butler, D.R., 1986. Epidemiological significance of liquid water in crop canopies and its role in disease forecasting. Pages 139-154. In: P.G. Ayres and L. Boddy (Editors), Water, Fungi and Plants. Cambridge University Press, Cambridge, 413 pp. Schnelle, F., Smith, L.P. and Wallin, J.R., 1963. Report on instruments recording the leaf wetness period. WMO Tech. Note No. 55. Steadman, J.R., 1983. White mold - a serious yield limiting disease of bean. Plant Dis., 67: 346350. Sutton, J.C., Gillespie, T.J. and Hildebrand, P.D., 1984. Monitoring weather factors in relation to plant disease. Plant Dis., 68: 78-84. Wallin, J.R., 1967. Agrometeorological aspects of dew. Agric. Meteorol., 4: 85-102. Weiss, A. and Lukens, D.L., 1981. An electronic circuit for the detection of leaf wetness and a comparison of two sensors under field conditions. Plant Dis., 65: 41-43. Weiss, A. and Hagen, A.F., 1983. Further experiments on the measurement of leaf wetness. Agric. Meteorol., 29: 207-212. Weiss, A. and Kerr, E.D., 1989. Evaluating the use of pest management information by growers: An example using Cercospora leaf spot of sugar beet. Appl. Agric. Res., in press. Weiss, A., Hipps, L.E., Blad, B.L. and Steadman, J.R., 1980. Comparison of within-canopy microclimate and white mold disease (Sclerotinia sclerotiorum) development in dry edible beans as influenced by canopy structure and irrigation. Agric. Meteorol., 22:11-21.