Water, heat and crop growth

In this paper attention has mainly been paid to the effects of the physical environmental conditions on plant development. As local research object, part of the economically important horticultural region of Geestmerambacht was taken, situated in the Netherlands in the northern part of the province of North-Holland. This area is partly covered with heavy sea clay soils under insufficient water management, resulting in a one sided cropping pattern. This together with a fragmented parcelling and a poor accessibility makes cultivation of profitable intensively managed crops difficult.

To a large extent the results of a farmer's efforts to get higher crop yields will be determined by the prevailing environmental conditions, i.e. by the existing complex of physical, chemical and biological factors. The possibilities of an efficient use of these factors are enlarged by our knowledge of their effects on agricultural production. Such knowledge is in first instance mostly gained from practical experience or field experiments often giving, however, only empirical information of local value.

To apply local results with success to analogous problems in other regions a fundamental treatment of research data is necessary, by which the occurring phenomena and processes can be analyzed more objectively.

In this paper attention has mainly been paid to the effects of the physical environmental conditions on plant development. As local research object, part of the economically important horticultural region of Geestmerambacht was taken situated in the Netherlands in the northern part of the province of North-Holland.

This area is partly covered with heavy sea clay soils under insufficient water management, resulting in a one sided cropping pattern. This together with a fragmented parcelling and a poor accessibility makes cultivation of profitable intensively managed crops difficult.

To investigate, in connection with an integrated land consolidation program for the region, the possibilities of improving the soil profile and the water management a groundwater level experimental field was laid out (Chapter II). On the experimental field ten different constant groundwater table depths were realized in three soil profiles, namely the original heavy clay soil and two improved clay soils: a sandy loam and a clay on sandy loam. Previous investigations on this field had the purpose of finding relationships between yield of a number of vegetable crops and mean optimum groundwater table depth during the growing season. Due to changes in weather conditions the results differed from year to year. It was therefore decided to apply a more fundamental treatment to the encountered problems by considering by means of mass and energy balance approaches what happens to the mass and energy fluxes in the soil-plant-atmosphere system.

Because the experimental field did not lend itself to water balance studies, a special type. of non-weighable lysimeter was developed in which the same water table could be maintained as in the surrounding field, including the fluctuations. Lysimeters were installed in the profiles of the 0.90 m and 1.20 m groundwater plots (Chapter II).

Before dealing with the influence of physical conditions on crop growth, attention was paid to the evaluation of the physical properties of the system by which the transport processes are determined. Chapter III deals with the factor water, Chapter IV with the factor heat.

In Chapter IIIA the transport of water in the unsaturated zone has been treated. To describe the state of water in the soil the energy concept has been used and the various terms involved have been discussed. The soil moisture retention curves were determined in the laboratory. Changes of soil moisture content with depth and time in the field were followed mainly by the gamma transmission method as well as by sampling with undisturbed soil cores. With the gamma transmission method an accuracy in the range of 0.5 vol. % could be obtained by expressing the mass densities of the soil components as an equivalent 'electron' density of water and by applying corrections for distance deviations and non-parallelism of the gamma access tubes. Because of large variations in the dry bulk density of the various layers, a relationship between moisture content, matric pressure and dry bulk density was developed, from which moisture retention curves at various dry bulk densities could be derived. This provided a method to be less dependent on the differences in dry bulk densities between the individual samples, especially in the gravimetric sampling procedure.

Hydraulic conductivity data were determined both from laboratory experiments as well as from measurements in the field in dry periods. The results compared favourably well with each other. Having evaluated the hydrological properties of the soil, relationships between flux, matric head and height above the groundwater table could be given. Also the influence of surface evaporation on drawdown of the groundwater table, assuming certain boundary conditions, was theoretically evaluated. It appeared that the calculated groundwater depths agreed reasonably well with groundwater depths found in similar soils in and zones. Moreover it could be shown that the evaporation flux decreased roughly with the square root of time, as reported in literature for different boundary conditions.

To obtain the amount of moisture available for evaporation over a certain growing period, the various quantities of moisture available in and coming available from below the root zone and from capillary rise from the groundwater table were calculated. The calculations were performed taking into account the admissable pressure at which soil water begins to limit plant growth. As appeared from literature studies by the present author, this value is for most vegetable crops about -0.4 bar (pF 2.6).

In Chapter IIIB the transport of water to the atmosphere has been treated. Evaporation was calculated by means of a combination method based on rather easily measurable meteorological as well as soil and crop factors. The calculations were compared with data obtained from water balance studies. For the aerodynamic resistance of the crop, values were calculated for various crop heights and wind velocities.

As net radiation is one of the most important terms determining evaporation from a surface, and direct measurements of net radiation are often not available, generally empirical formulae are used. Some of these formulae were used in this study and the calculation results were compared by means of linear regression techniques with measured net radiation data. It appeared that for calculated shortwave radiation high correlations were obtained with various formulae. For thermal radiation, however, calculations over short periods of a week gave poor results. Therefore a few expressions based on the high correlationship between net and shortwave radiation were derived for various lengths of time and analyzed by linear regression. The relation obtained on a 24-hour basis compared strikingly well with similarly computed relations found in WesternAustralia, Also equations computed on a daylight basis compared favourably well with literature data. It further appeared that for calculations of evaporation on a weekly basis, radiation data measured at relatively distant meteorological stations can sometimes be used with sufficient accuracy.

Reflection coefficients of various surfaces were determined from continuous measurements taken on all kind of days. A discussion was presented on the various methods determining this coefficient. Results of reflection coefficients at increasing fractions of soil covered for a number of vegetable crops were given. Except for spinach, they showed all a more or less similar pattern within a small band of data. It was shown that wetting the surface reduces it, both for bare soil as well as for a cropped surface.

Evaporation from bare soils with different groundwater tables was studied on a hourly basis by means of an energy balance approach on clear days in a rather wet and a rather dry period. Differences in the energy balance components between the various objects were established and calculated vapour pressures at the surface were compared with data obtained from sampling of the surface layer, giving a rather good agreement.

A discussion on the diffusion resistances encountered by evaporation from cropped surfaces has been presented. It appeared that the soil must be covered to about 70 to 80 % before constant evaporation rates are obtained. If the soil is rather wet, however, variation in the area of soil covered has not much effect.

Measurements of interception of red cabbage agreed with literature data concerning interception of grass. It was shown that interception is important in periods of reduced evaporation, as it then influences evaporation most.

An analysis of the transport resistance for liquid flow in the plant as well as an investigation on the geometry factor of the root system for red cabbage was presented. The variation of these factors with depth could be shown, and root extraction rates at different depths were calculated and compared with data obtained from water balance studies. The plant resistance data were in reasonable agreement with literature data. Because of a non-homogeneous and poor root development in the early stages of growth, the geometry data of the initial growing stages differed a factor ten from data found in literature. After root development increased with depth, geometry data decreased to values also reported for other crops.

From comparison of evaporation calculated by the combination method and evaporation obtained from water balance studies, satisfying results were obtained. The problems encountered in the first growing stages with seedling emergence and initial growth were emphasized. It appeared from the calculations that except for a few periods during the first growing stages only a small reduction in evaporation by water shortages under the prevailing circumstances was present.

Chapter IV deals with the flow of heat in soil. To evaluate the thermal properties of the soil, calculations were carried out on measurements taken in the laboratory as well as in the field. The various determination methods as found in literature were discussed. Thermal capacity was calculated by considering it as a function of volume fractions and specific heat capacities of the various components.

Calculations of thermal conductivity based on the assumption that soil particles can be considered as spheroids gave poor results, possibly because of a wrong choice of the depolarization factor. Measurements of thermal conductivity in the laboratory with the transient needle method yielded good results for sandy loam. For the shrinking and swelling clay, however, a large scatter in the data occurred because of change in dry bulk density which could hardly be corrected for. Therefore preference was given to measure thermal conductivity in the soil in situ.

Thermal diffusivity of the sandy loam could be deduced from laboratory data. Determination of thermal diffusivity in the field from the analysis of temperature measurements at various depths, by considering the amplitude and phase relationships with depth, failed, indicating that the thermal properties did vary with depth. To establish functional relationships of thermal diffusivity with depth, an electric analog as well as a numerical approach, both based on the principle that a discontinuous field can approach the continuous thermal field by expanding the partial differential equation into a set of finite difference equations, was used.

The scatter in the data obtained with the electric analog method was large due to various reasons like for example recording and reading errors from the temperature recorders, relatively widely spaced temperature observations at the larger depths and heterogeneity of the soils. With the numerical approach also disappointing results were obtained. Improvement of recording and reading errors by replacing measured temperature data by reconstructed smoothed temperatures computed from Fourier analysis, did not give better results. Improvement of the error inherent in discretizing the continuous temperature field to a discontinuous field by application of interpolation by means of spline functions, neither did. From extensive calculations on a possible disturbing influence of water transport in the vapour phase as induced by thermal gradients, it could be shown that this process could be neglected as a possible explanation of the scatter in the thermal diffusivity data.

The best method to determine thermal diffusivity proved to be deriving thermal capacity from soil sample data, measuring thermal conductivity with the transient needle method in situ, and estimating thermal diffusivity at each depth by considering the ratio of thermal conductivity and capacity. It was found that at higher groundwater levels thermal diffusivity is generally lower, especially in the top 0.10 to 0.15 m layer, indicating lower soil temperatures at higher groundwater levels. From temperature measurements it appeared that the mean daily temperatures of the plots with the higher groundwater tables were 1 to 2°C lower than the temperature of the plots with the deeper groundwater tables. With the same groundwater depth, clay proved to be warmer than sandy loam. It could be concluded that on the investigated soils the difference in groundwater level was playing a more important role than the difference in type of profile. The maxima and minima in the top soil were higher and the amplitudes decreased with depth faster in soils which had a deep groundwater table. The decrease was more marked in the clay than in the sandy loam soil.

In Chapter V the combined effects of water and heat on seedling emergence and crop production were treated. In the stage of germination and seedling emergence, the soil moisture content and soil temperature are the most important factors. In the stage from seedling emergence to maturity growth depends also on air temperature, but mainly on leaf area and net radiation.

The known effects of soil temperature and soil moisture content on germination and seedling emergence were reviewed. To relate temperature and emergence the heat sum concept was used. This approach is in practice frequently applied to schedule plantings, to predict maturity, to select crop varieties appropriate to different areas, etc. In such type of studies mostly the environmental air temperature is used. In the case of emergence of seeds it is advisable, however, to register the temperature in the direct environment of the seed, c.q. to measure soil temperature at sowing depth. The results of investigations on the influence of soil moisture content on germination and emergence as reported from laboratory experiments in literature, generally differ widely, mainly because of differences in applied experimental conditions as for example the use of a solution instead of soil as a germination medium.

The combined effect of soil temperature and moisture content on seedling emergence was studied with four different kinds of vegetable seeds in field experiments in a clay and a sandy loam profile, both with a shallow and a deep groundwater table. Various sowing dates were applied. It appeared that emergence was highly correlated with rainfall, and that because of the favourable hydrological properties of the soil seeds emerged earlier in sandy loam than in clay. It was found that on all sowing dates the sandy loam plots with the shallow groundwater table showed the highest emergence rate as well as the highest total emergence percentage. The mean heat sums required for 50% emergence were lower on sandy loam than on clay, and the heat sums of the shallow groundwater plots were lower than those of the deep groundwater plots.

The minimum temperatures for emergence of the various seeds were calculated using only those treatments in which no limitation of water could be expected. The effect of soil moisture on emergence could be evaluated by calculating the heat sums of all treatments, taking into account the minimum temperature for emergence. As indicator for the minimum moisture content required for emergence, the first five days after sowing were used, for which the average soil moisture content was calculated from sampling data and precipitation records.

It was found that the heat sum required for 50 % emergence increased sharply below a matric pressure of -0.49 bar (above pF 2.7) of the soil. Laboratory experiments with radish seed under controlled soil moisture and temperature conditions confirmed the results obtained from the field experiments that -0.49 bar (pF 2.7) is the critical value for emergence as regards the dry side. From laboratory experiments it could be shown that -0.098 bar (pF 2.0) is the critical value for emergence as regards the wet side. The time needed for emergence increases rapidly at matric pressure above -0.098 bar and below -0.49 bar. It appeared from these experiments that heat sums can give a relatively accurate prediction for emergence, if soil moisture content is taken into account.

For a fast and adequate seedling emergence both a high temperature and a sufficient moisture content are necessary. Under field conditions this combination is seldom reached because higher soil temperatures are generally related with lower moisture contents and deeper groundwater tables. To get out of this dilemma one can maintain a relatively deep groundwater table, which gives a relatively high temperature, and keep by means of sprinkler irrigation the sowing bed at the desired moisture content. This offers the additional advantage of keeping the temperature of the seed bed low in periods when soil temperature would exceed the optimum temperature for germination and emergence. It is to be noted that temperatures of 40°C and higher (which are far too high for germination and emergence of various seeds) in the top layer of clay soils in the Netherlands in mid summer are no exception.

From literature it is known that yields of spring crops decrease when sowing dates are later and that owing to adequate drainage, soils can be cultivated and sown approximately 5 to 14 days earlier. An additional advantage of drainage is the shortening of the germination and emergence period as a result of the higher soil temperatures. From calculations it was shown that due to drainage sometimes a 10-day gain in emergence can be obtained as compared with a shallowly drained soil. The effect of drainage (or any other measure) on seedling emergence will be the largest when soil temperatures in the range close to the minimum temperature for germination and emergence are increased.

If available water is limiting, plant production will be reduced, which especially holds true for vegetable crop production where production is more aimed at quality and fresh weight than at an increase in dry matter. To weigh the influence of various measures, as for instance change in groundwater table depth or soil profile, on the production of crops, potential (gross) production rates which could have been obtained under the prevailing weather conditions with an optimum water and nutrient supply were calculated for the crops red cabbage, dwarf French beans and celery. Taking into account effects of soil cover and diffusion resistances of the crop, maximum production rates for the prevailing environmental conditions were computed. These rates were compared with the dry matter production rates obtained from periodical harvests. The relationships between real and calculated production rates were analyzed by linear regression of measured on calculated production, from which reduction factors (alpha ph ) were derived. From various data reported in literature it appears that losses in dry matter production by respiratory processes is about 30 per cent, so that maximum dry matter production is about 0.7 (= alpha ph ) of the potential production.

For dwarf French beans an average alpha ph of 0.67 was found for the various objects, indicating that the environmental conditions were fairly optimum.

For red cabbage rather low alpha ph values were found: 0.51 to 0.56 for clay and clay on sandy loam respectively and 0.40 for sandy loam. The main reason for these low alpha ph values seemed to be nitrogen deficiency. This deficiency occurred mainly in dry periods on the clay profiles and in wet periods (in autumn) on the sandy loam profiles. Moreover air deficiency induced by heavy rains in the later stages of growth may have been of importance.

For celery also a low average alpha ph of 0.52 for the various objects was obtained, again mainly because of nitrogen deficiency. Part of the discrepancy could be explained by the applied calculation procedure.

A discussion on the various existing methods in determining the water use efficiency of a crop was presented. By considering total dry matter production and the ratio total evaporation over mean vapour pressure deficit of the three crops mentioned, it could be shown that the water use efficiency of dwarf French beans is highest, followed by celery and then red cabbage. It was found that under the prevailing environmental conditions other than water shortage, red cabbage on sandy loam was rather soon limited in its production. On the other profiles limitation of production did not set in.

The influence of groundwater table depth on production of crops has always been widely investigated in the Netherlands, especially on groundwater level experimental fields. Because of changes in weather conditions the relationship between yield and groundwater level differs from year to year and it seems of little use to make this comparison for years on end.

Because for a certain dry matter production a certain amount of water has to be evaporated, it is better to consider for a growing period the amount of water available for evaporation, partly determined by the amount of precipitation and partly by the amount of water that can be delivered by the soil. Taking into account the admissable pressure -0.4 bar (pF 2.6) at which soil water begins to limit growth of most vegetable crops, the amounts of water available in the soil at different lengths of growing period were calculated.

Fresh yields measured on all profiles of all groundwater plots were plotted against the total amounts of water available at each object for red cabbage, potatoes and lettuce. It appeared that maximum fresh yield productions were obtained at total amounts of available water of 365 mm for red cabbage, of 310 mm for potatoes (which agrees rather well with data derived from literature) and of 200 mm for lettuce.

Having obtained the optimum amounts of available water for maximum production of the crops, the maximum groundwater table depth, over a period in which the critical value of the total amount of rainfall was known from frequency distributions of rainfall depth, could be calculated for each crop and each profile. It was shown that per profile despite differences in crops, in lengths of growing periods and in frequency distributions of amount of rainfall, the deepest admissable water table depths did not differ very much. The reason for this is that the amounts of available water in the investigated profiles change considerably with slight changes in groundwater table depths around 0.90 m below surface. Generally speaking, the crops on the sandy loam admit deeper groundwater levels than on the clay and on the clay on sandy loam profile.

In a similar way also highest admissable groundwater table depths to ensure maximum production could be calculated. It was found that for early crops like potatoes and lettuce rather high groundwater tables can be admitted. For late crops like red cabbage deeper water tables are required. This is a logical consequence of the fact that early crops are in the Netherlands grown in periods with generally less rainfall than late crops are.

Finally optimum water table depths for maximum productions could be calculated, the procedure of which has been explained. The results showed that the optimum water table depth for red cabbage on clay is about 1.00 to 1.10 m, on sandy loam 1.10 to 1.20 m and on clay on sandy loam 0.90 to 1.00 m. For potatoes the optimum groundwater table depths are 0.90 m, 1.00 m and 0.90 m respectively; for lettuce 0.90 m, 1.00 m and 0.80 to 0.90 m respectively. It appeared that at deeper groundwater tables, the sandy loam carries less risk in production than the clay. This last profile in its turn carries less risk (for red cabbage much less risk) than the clay on sandy loam profile. It was shown that the latter is more susceptible for water table drawdowns below the optimum level than the other two profiles.

In general it is to be advised to maintain groundwater levels at depths that are somewhat lower than the optimum depth.

It is thought that with this approach there is a possibility to solve the problem of determining the total amount of water required for maximum crop production and, where relevant, to decide on the optimum groundwater table depth then necessary.