Effect of Irrigation and Variety on Potato Water Uptake: Study at Cambridge Farm, Notas de estudo de Engenharia Agronômica

Baixe Effect of Irrigation and Variety on Potato Water Uptake: Study at Cambridge Farm e outras Notas de estudo em PDF para Engenharia Agronômica, somente na Docsity! Water uptake in the potato (Solanum tuberosum) crop M. A. STALHAM* AND E. J . ALLEN Cambridge University Farm, Huntingdon Road, Cambridge, CB3 0LH, UK (Revised MS received 15 September 2004) SUMMARY Experiments were conducted on sandy loam soils at Cambridge University Farm over the period 1989–99 to examine the effects of irrigation regime and variety on water uptake (WU) in potatoes. Unirrigated crops extracted water from considerable distances ahead of the rooting front but fre- quently watered crops took up water from depths shallower than the current depth of rooting. There was an increase in the extraction of soil water at depth if crops were irrigated less frequently at moderate (i.e. 40 mm) soil moisture deficits (SMD). The SMD measured at different positions across the ridge always differed and the relationship changed during the season. This is of concern since most reports on water use in potatoes are based on a single measurement position for the neutron probe in the centre of the ridge and this location over-estimates crop water use. Crops grown on the flat had a more uniform extraction of soil water across the row width than crops grown in ridges but there was no evidence that having one part of the rooting system drier than another affected overall crop water use. Once rooting systems were established to considerable depth, WU continued from deeper roots even though upper horizons were periodically re-wetted by irrigation. For this reason, it proved impossible to relate WU to rooting density in specific horizons over the course of the season. Only early in the season did the proportion of total WU correspond reasonably closely with the proportion of total root length in each horizon. It appeared that the pattern and extent of soil drying created by a crop changes the horizons where water is extracted at different growth stages and the relative rooting density in a particular horizon is not a good indicator of the potential to take up water from that depth. Although rooting density decreased rapidly with increasing depth, roots deeper in the profile contributed a considerable component of total crop water requirement irrespective of the water status of horizons closer to the soil surface. A series of close relationships were established between the ratio of actual : potential evapo- transpiration and SMD for different daily evaporative rates. These showed that there was a limiting deficit at which the ratio of actual : potential evapotranspiration decreased rapidly with increasing SMD and this limiting deficit was inversely related to daily evapotranspiration rate. However, even at small SMDs, as daily evapotranspiration rate increased there was a significant, slow decrease in actual : potential evapotranspiration ratio. In order to maintain potential evapotranspiration rates in conditions of extreme demand in the UK (e.g. 5–7 mm/day), crops need to be maintained at<25 mm deficit but allowable deficits can be increased as demand moderates. INTRODUCTION Water is extracted from the soil by roots and their depth and distribution are key factors influencing the accessibility of water, its use to satisfy the demand created on the canopy by the atmosphere and hence yields. Gregory & Simmonds (1992) stated that the apparent drought sensitivity of potatoes may be caused by the limited ability of the root system to convey water. They speculated that the low values for below-ground conductance of water in the plant were due primarily to the relatively small total root length per unit area of soil (TRL) rather than an inherently small conductance per unit length of root. However, Stalham & Allen (2001) showed that, in the absence of compaction, TRL in irrigated potatoes was typi- cally 14–15 km/m2, which compared favourably with, or exceeded, other spring-sown crops including sugar beet. However, the range in published maximal TRL for potatoes is large, from 1.6 to 24.1 km/m2 (Stalham & Allen 2001), which may be more a consequence of soil structural conditions than the inherent inability * To whom all correspondence should be addressed. Email : m.stalham@farm.cam.ac.uk Journal of Agricultural Science (2004), 142, 373–393. f 2004 Cambridge University Press 373 DOI: 10.1017/S0021859604004551 Printed in the United Kingdom of the potato plant to produce a large, dense rooting system per se, since potatoes have been shown to be sensitive to compaction (Boone et al. 1985; Van Oijen et al. 1995; Rosenfeld 1997). Soil water extraction by plants has been described at many levels of complexity in order to calculate the rate of water uptake (WU) from soil as demanded by the atmosphere. In most models, many of which were reviewed by Molz (1981), the rate of WU depends mostly on the root length per unit volume of soil (RLD, cm/cm3), or per unit area of soil surface (TRL, km/m2). Sometimes a distinction has been made between ‘active’ and ‘non-active’ roots (Nimah & Hanks 1973), although which roots are functionally capable of taking up water is still poorly understood. When water supply to the roots is not limited, crops with closed canopies should transpire water at a rate corresponding closely to potential evapotranspiration (ET). If the soil water content is similar throughout a homogeneous soil profile, it could be assumed that WU will be distributed within the soil profile as some function of RLD, with the depth integral of WU equalling the actual transpiration rate. However, experimental evidence from field crops has indicated that there are environmental conditions where RLD and TRL are not well correlated with either WU or transpiration rate, such as when some parts of the deeper profile have a plentiful supply of water and short lengths of root in these layers are able to supply the crop adequately, whilst more densely rooted layers rapidly dry out (Gregory et al. 1978; Sharp & Davies 1985). In irrigated crops, soil water is replenished at intervals and if the interval between irrigations is lengthy, surface horizons, where rooting is usually dense, may have a low water content and most probably a low relative rate of WU per unit length of root just prior to the soil being replenished by water (Klepper et al. 1973; Brown et al. 1987). Interpretations from rain-fed environments are in- adequate to improve the efficiency of soil water usage in irrigated crops where water is more readily avail- able and there is a dearth of measurements in the potato crop. It is also necessary to establish whether the efficiency of the root system with respect to WU is the same at every depth and whether it alters during the course of the crop’s life. Although ageing of roots may affect their ability to take up water, Taylor & Klepper (1975) indicated that cotton roots at all depths were equally effective in taking up water when compared at equivalent soil and plant water poten- tials. Taylor & Klepper (1973) also showed that maize roots at depth took up water at a faster rate than those near the surface. They suggested that this resulted from the deeper roots being younger, less crowded and located in wetter soil. Asfary et al. (1983), working with potatoes, showed that inflow rates of water were at a maximum c. 6 weeks after emergence and decreased thereafter, with inflow rates below 30 cm depth being c. six times greater than those above 30 cm. However, this would suggest that many potato crops would have access to little water for the greater part of their life, particularly if shallow compaction restricted rooting depth. It seems feasible to suggest that if water is available in the topsoil, plants will preferentially use water higher in the profile, reducing the ‘value’ of deeper roots. This is obviously very pertinent in irrigated potato crops which will have several wetting/drying cycles in most seasons. The effect of varying evaporative demand on the distribution of WU with respect to the relative wetness of different horizons must also be considered. The ratio of potential to actual evapotranspiration depends on both the soil moisture deficit (SMD) within the profile as a whole, or within individual horizons, and on the rate of ET (Denmead & Shaw 1962; Bailey & Spackman 1996). The purpose of thoroughly examining the most important aspects of crop root growth which influence crop response to meeting ET under varying environmental conditions is to ascertain the likely contribution of soil water to plant needs. This will improve understanding of the dynamic changes in the WU potential that occur throughout the life of the crop, thereby permitting more efficient use of both soil and irrigation as water supplies to the growing crop. In order to study the relationships between rooting characteristics and the patterns of water availability in irrigated potato crops which normally fluctuate between wet and dry soil for a considerable period of their life, the neutron probe (NP) was used to measure the WU of contrasting crops in a number of exper- iments conducted at Cambridge University Farm over the period 1989–1999. MATERIALS AND METHODS General The current paper reports measurements of WU from a series of experiments conducted at Cambridge University Farm (CUF) on aMilton Series soil (Anon. 1983) over the period 1989–99. Some experiments were subjected to natural rainfall (Expts 3, 5 and 8; Tables 1 and 2) but the rest were grown under per- manent polythene rainshelters (16r8 m, Polybuild Ltd). Temperatures were increased under the rain- shelters compared with ambient but the cladding polythene ended 30–50 cm above the ground at the sides of the shelter and the ends had large openings so air flow through the shelters was good. Global radi- ation under the rainshelters was reduced by c. 23% (Stalham 1989) but crops appeared normal compared with crops grown outside the shelters. Soil cultivations for all experiments involved ploughing, spring tining and rotavating or power harrowing to 20–30 cm 374 M. A. STALHAM AND E. J. ALLEN pits. At 50% plant emergence pits were dug by hand using a spade across two harvest rows and maximum rooting depth recorded under both rows. The pits were enlarged both vertically and down the length of the plot using a spade or JCB digger, leaving two discard plants between each successive digging. Measurements were taken from the top of the ridge or the soil surface for flat plantings. Neutron probe measurements Throughout the course of the experimental series, a Soil Moisture Probe Type IH II (Didcot Instrument Co. Ltd) was used to measure changes in soil water content (SWC). Aluminium access tubes of 45 mm external diameter were gently hammered into 40 mm diameter holes made using a gouge corer (Eijkelkamp) attached to a percussion hammer (Atlas Copco), until the top of the tube was between 10 and 20 cm above the soil surface. The access tubes were installed at crop emergence mid-way between two plants in the row centre approximately 1 m in from the end of the plot, with a single tube per plot. This spatial instal- lation assumes that WU and infiltration are uniform across the width of the module used. The NP measures SWC within a radius of 15 cm in wet soils to 30 cm in dry soils (Bell 1987) and therefore a single access tube will not measure the water content on widely-spaced row crops such as potatoes. In order to ensure that the NP measurements represented the whole row-width, four access tubes were installed in each plot in Expt 8. In the ridge plots, the tubes were installed in the ridge centre (RC); one-third of the distance between ridge centre and furrow centre (RF); two-thirds of the distance between ridge centre and furrow centre (FR); and in the furrow centre (FC). In the flat plots, the tubes were installed in the equivalent positions to the ridge plot : the RC tube being in the row centre ; the FC tube midway between rows, whilst the RF and FR tubes were one-third and two-thirds the distance respectively from the RC to the FC tube (Gaze et al. 2002). Access tubes were in- stalled in the two central rows of each plot to avoid the excessive soil disturbance created by installing four tubes in the same ridge. A portable gantry spanning four rows was used which enabled the NP readings to be taken without damage to the soil surface or crop near the access tubes. In earlier ex- periments, a 1r0.25 m board was used to spread the weight of the operator on the soil when approaching the access tube. Single readings of 16 s duration were taken at 10 cm intervals down the tubes to 90–150 cm depth relative to the top of the ridge in the ridge plots, or to the soil surface in the flat plots. An horizon- based integration was used to calculate the water content of the profile down to the maximum depth of measurement. For the data presented in this paper, measurements were taken immediately prior to irrigation events and at the same time of day when- ever possible. RESULTS The NP was used for measuring changes in soil water content in all the experiments but the results must be interpreted with care, since Gaze et al. (2002) showed that the NP was inconsistent in measuring known irrigation or rainfall input, even when multiple access tubes were used to ensure the entire row width was sampled. They found that the NP was unable to detect all the water applied to the soil, particularly where the water was largely confined close to the soil surface. Replicated measurements of the change in SMD in the field experiment were precise for a given event and treatment but were not accurate when compared against the input measured in raingauges. It was concluded, therefore, that the NP could not be used reliably to measure changes in soil water storage immediately following irrigation or substantial rain. For periods when there were minimal inputs of water, there was a closer correlation between changes in SMD measured by the NP and those predicted by a modified Penman–Monteith equation than after sub- stantial inputs of water. However, the frequency of NP measurements taken in most of the reported experiments was such that many periods of minimal or zero water input could be used to determine the location of WU and the changes in soil water content can be regarded as accurate within the measurement zone of the NP. Water uptake and rooting depth It is important to establish whether the maximum depth of rooting coincides with the maximum depth of measured water uptake and whether the drying front moves downwards at the same rate as roots. Experiment 2 showed that there were differences be- tween irrigation treatments in the time lapse between emergence and measuring water uptake in each con- secutive 10 cm horizon down the profile (Fig. 1). In Dry-Dry crops, the average rate of rooting (interp- olated from measurements taken at emergence and 80 days after emergence (DAE)) was 1.35 cm/day, whilst WU data from the NP suggested that the rate of downward movement of the drying front to 110 cm (the maximum depth of recording) was 1.30 cm/day. Similarly, in Wet-Dry plots which were irrigated from planting until 44 DAE, the downward movement of the uptake front (1.16 cm/day) was the same as the rate of vertical root growth (1.15 cm/day), but there was a lag before WU was measured in each horizon compared with Dry-Dry crops. However, in the Dry- Wet crops, the rate of rooting (1.11 cm/day) was fas- ter than the measured increase in the depth of drying front (0.93 cm/day) taken over the first 80 DAE. Water uptake in the potato crop 377 Additionally, the rate of increase in depth of water extraction followed the same pattern as Dry-Dry crops until irrigation was started 44 DAE. From this point, the downward progression of the drying front slowed compared with unirrigated crops, indicating clearly that soil moisture conditions directly affect the distribution of WU. In Expts 6 and 7, there were more measurements of rooting depth throughout the season, so it was possible to examine the relationship between rooting depth and depth of WU more closely. Figures 2 and 3 show the fitted linear relationships of depth of WU and rooting depth against time after emergence. Some of the changes in soil water content at shallow depths early in the season were due to evaporation from bare soil rather than root uptake. In Expt 6, rates of rooting across all irrigation treatments were slow (c. 0.81 cm/day) owing to shallow compaction. Over the first 25–35 DAE rooting depth was deeper than depth of extraction (Fig. 2). However, following this period, the drying front extended faster than the rooting front, so thatwater was extracted fromdeeper than themaxi- mum depth of rooting. The rate at which the drying front progressed down the profile was fastest for unirrigated crops (W1), slower for crops irrigated according to the CUF schedule (W2) and slowest for the crops maintained at an SMD of 10–25 mm (W6; Fig. 2). Since the rate of root penetration was the same for these three irrigation treatments, the unirrigated crop was extracting water from further ahead of the rooting front than either irrigated crop, and, therefore, it can be assumed that the frequently irrigated crops did not fully utilize the lower part of the rooting system since they could extract sufficient water from themoist soil in superficial horizons. Later in the season, crops grown under all irrigation regimes were extracting water from 90 cm, even with frequent irrigation being applied to the soil surface as found in Expt 2. In Expt 7, it was found that in unirrigated crops the drying front also moved faster down the profile than the rooting front, so that by 62 DAE Dry crops were extracting water from c. 25 cm deeper than the measured depth of rooting (Fig. 3). However, in the CUF crops maintained at an SMD of c. 25–48 mm, the drying front extended at the same rate as rooting depth, but for most of the season was c. 17 cm shal- lower than the maximum depth of rooting. Spatial variability of measurement of water uptake In order to understand WU more effectively, it is necessary to establish the uniformity of uptake within 0 20 40 60 80 100 120 0 20 40 60 80 Days after emergence D ep th o f w at er e xt ra ct io n (c m ) Fig. 1. Time taken for neutron probe to register water extraction at different depths in Expt 2. Dry-Dry (&) ; Dry-Wet (%) ; Wet-Dry (m). Arrow marks change between Dry-Wet and Wet-Dry irrigation regimes. (a) 0 20 40 60 80 100 D ep th o f m ea su re d w at er up ta ke /r oo tin g de pt h (c m ) (b) 0 20 40 60 80 100 D ep th o f m ea su re d w at er up ta ke /r oo tin g de pt h (c m ) (c) 0 20 40 60 80 100 0 20 40 60 80 Days after emergence D ep th o f m ea su re d w at er up ta ke /r oo tin g de pt h (c m ) Fig. 2. Relationship between depth of measured water uptake and rooting depth for three contrasting irrigation regimes in Expt 6. (a) W1; (b) W5; (c) W6. Equations of water uptake (—&—) regressions: (a) y=1.47x+3.7, R2=0.98; (b) y=1.30x+3.5, R2=0.99; (c) y=1.22x+2.8, R2=0.99. Equations of rooting depth (- - -%- - -) re- gressions: (a) y=0.88x+18.1, R2=0.99; (b) y=0.79x+21.9, R2=0.99; (c) y=0.80x+18.8, R2=0.99. 378 M. A. STALHAM AND E. J. ALLEN each horizon. To this end, individual access tube locations in Expt 8 were used to test (a) if there were differences in water inputs and/or use between pos- itions and (b) if it was possible to predict the mean SMD from measurements taken at a single position (in particular the RC position since this is the most common position measured). It is important that any relationships derived between a single position and the mean apply throughout the season if they are to be of practical use. In both unirrigated and irrigated crops, the overall mean SMD from the four tube positions measured in Expt 8 was the same for ridge and flat profiles (Fig. 4). This is significant in that cropping profile had no effect on overall water use, whether the soil was dried out considerably or kept much wetter. There was no evidence of an inherently poorer capture of water following irrigation on ridges compared with a flat profile. However, in unirrigated crops, the RC tube always recorded a greater SMD than other positions and this difference increased when dry soils were wetted with rainfall later in the season. There was a larger difference between the SMD at RC than the other positions under ridge cropping than under flat profiles. When irrigation was applied, the absolute and relative differences in SMD between tube pos- itions were dynamic throughout the season, i.e. there was no way of predicting the mean SMD from any single tube position, or any combination of two tubes, since the relationship between the SMD measured at different positions altered during the season (Fig. 5). Differences in water use between the access tube positions were best examined for periods when there was no irrigation and minimal rainfall and the crop had full ground cover. There were two periods (beginning 15/16 June and 6/7 July) when the crop had full ground cover and the patterns of water use across the rows could be studied without undue in- terference from irrigation or rainfall inputs. However, even for these periods, there was a complicating irrigation input which meant data could not be com- pared between all irrigation treatments over the same number of days. On both occasions, treatments W3 and W4 were irrigated the day before regular weekly NP readings. Pre- and post-irrigation readings were taken for the cropped plots in these treatments but data for W1 and W2 were not taken until the next day, when the whole experiment was monitored. To compare drying patterns, data for W1 and W2 were those collected from the regular weekly monitoring (15/16–22/23 June, 6/7 July–13/14 July) and for W3 and W4 were from the beginning of the period until pre-irrigation readings (15/16–21 June, 6/7–12 July). It has been assumed that the effect of a single extra day’s water use would have minimal influence on the patterns of water use between flat and ridge profiles and these can, therefore, be compared. Comparison between irrigation treatments, however, could not be made. Also, for the period beginning 15 June, water use was calculated only over the top 50 cm to exclude slow drainage from the lower depths over the period. Changes in water storage were consistent for dif- ferent positions in the bare soil plots but not in the cropped treatments (Table 3). The different water use across rows for the cropped treatments was, there- fore, a consequence of crop water extraction and not evaporation from the soil surface. For both periods there was no significant difference between ridge and flat profiles in the mean change in water stored in the profile (Table 4). However, the pattern of water use across the rows differed between ridge and flat, and the pattern was different for the two periods. For the period beginning 15/16 June, the change in water stored at the RC position was the same in ridge and flat profiles (Table 4). The change in water stored at the RF, FR and FC positions with respect to the RC position was significantly less in the flat profiles than in the ridge, with the difference increasing with in- creasing distance from the RC position. This suggests that the crop extracted water more uniformly across the row-width in the flat profiles than in the ridged plots. A different pattern of water use was observed for the period beginning 6/7 July (Table 4). Water use (a) 0 20 40 60 80 100 120 D ep th o f m ea su re d w at er up ta ke /r oo tin g de pt h (c m ) (b) 0 20 40 60 80 100 120 0 20 40 60 80 100 Days after emergence D ep th o f m ea su re d w at er up ta ke /r oo tin g de pt h (c m ) Fig. 3. Relationship between depth of measured water uptake and rooting depth for two contrasting irrigation regimes in Expt 7. (a) Cara, Dry; (b) Cara, CUF. Equations of water uptake (—&—) regressions: (a) y=1.93x+5.9, R2=0.98; (b) y=1.04x+13.4, R2=0.98. Equations of rooting depth (- - -%- - -) regressions: (a) y=1.12x+26.4, R2=0.98; (b) y=1.01x+30.0, R2=0.97. Water uptake in the potato crop 379 − 5 0 5 10 15 20 25 30 35 40 45 50 RC RF FR FC RC RF FR FC SM D ( m m ) in 1 0 cm d ep th in cr em en ts (a) Begin End RC RF FR FC RC RF FR FC 10 20 30 40 50 60 (b) Begin End Depth (cm) −5 0 5 10 15 20 25 30 35 40 45 50 RC RF FR FC RC RF FR FC SM D ( m m ) in 1 0 cm d ep th in cr em en ts (c) Begin End -5 0 5 10 15 20 25 30 35 40 45 50 RC RF FR FC RC RF FR FC 10 20 30 40 50 60 (d) Begin End Depth (cm) Fig. 6. For legend see opposite page. 382 M. A. STALHAM AND E. J. ALLEN water content, whilst at 70 cm and below uptake continued at a rapid rate. By contrast, the summer of 1991 was cooler and wetter than 1990. June had plentiful rain (98 mm) which meant that unirrigated crops in Expt 5 achieved complete canopy cover. However, July, August and September were drier than average but nearly half of the rain that fell over this period occurred on 31 July and 1 August (44 mm) and it failed to rain on 68 days during July–September. As a consequence, SMDs began to increase rapidly in unirrigated crops from the beginning of July, eventually reaching c. 80 mm in both Cara and Desiree (Fig. 9). All crops took up water from 80, 90 and 100 cm depths, with the ‘ lag phase’ of WU from deeper horizons following in irrigated as well as unirrigated crops (Fig. 10). Water uptake during August was occurring in irrigated crops simul- taneously at depths below 80 cm and in superficial horizons which were being periodically replenished with irrigation. Efficiency of water extraction from different horizons under contrasting irrigation regimes TheWet crops in Expt 1 were maintained at an overall SMD of <10 mm, yet water was still extracted from as deep as 90 cm. Owing to the frequent irrigation of these wet treatments, superficial horizons were (a) −17 −13 −9 −5 −1 01 May 22 May 12 Jun 03 Jul 24 Jul 14 Aug 04 Sep SM D ( m m ) (b) −17 −13 −9 −5 −1 SM D ( m m ) Fig. 7. Horizon soil moisture deficits in (a) unirrigated Cara and (b) irrigated Cara in Expt 3. 10 cm (–&–); 20 cm (–%–); 30 cm (–m–); 40 cm (–
–); 50 cm (–$–); 60 cm (–#–); 70 cm (- -%- -) ; 80 cm (- -
- -) ; 90 cm (- -#- -) depths. −100 −80 − 60 − 40 −20 0 01 May 22 May 12 Jun 03 Jul 24 Jul 14 Aug 04 Sep SM D ( m m ) Fig. 8. Total soil moisture deficits in Expt 3. Unirrigated Cara (&) ; Unirrigated Desiree (%) ; Irrigated Cara (m) ; Irrigated Desiree (
). −90 −80 −70 −60 −50 − 40 −30 −20 −10 0 20 May 17 Jun 15 Jul 12 Aug 09 Sep 07 Oct SM D ( m m ) Fig. 9. Total soil moisture deficits in Expt 5. Unirrigated Cara (&) ; Unirrigated Desiree (%) ; Irrigated Cara (m) ; Irrigated Desiree (
). Fig. 6. Nominal soil moisture deficits (SMD) measured in 10 cm increments down the soil profile at four positions across the row width for cropped ridge and flat profiles at the beginning and end of two 5–7 day periods in Expt 8. (a) ridge, period beginning 15 June; (b) flat, period beginning 15 June; (c) ridge, period beginning 6 July; (d) flat, period beginning 6 July. Data are means of all irrigation treatments. Water uptake in the potato crop 383 maintained at, or over, field capacity for much of the season, nevertheless the drying front progressed downwards at a rate comparable with the rate of rooting. The results from this experiment also show that on homogeneous soils crops can extract similar quantities of water from each horizon, with irrigation regime changing the magnitude of the amounts extracted compared with unirrigated crops (Table 5). Superficial horizons did not need to be exhausted before significant WU was observed in deeper horizons. These results, therefore, support those obtained from the uncovered Expts 3 and 5. Where the profile had markedly dissimilar top- and subsoils, with moderately water-retentive topsoils but very stony sand subsoils (Expt 2), less water was ex- tracted at depths below the ploughed layer than in shallow horizons but crops irrigated for the first 44 DAE and then forced to exist on soil reserves could still exhaust deeper horizons as completely as crops grown without any water (Table 6). Relationship between water uptake and rooting density Figure 11 shows the comparison between the pro- portion of TRL in each horizon and the proportion of total WU contributed by each horizon for three sample periods in Expt 7 for Dry and Wet Cara. Changes over time in proportional WU from a particular horizon did not follow the changes in RLD in each horizon. At the earliest sampling, when the soil was wet in most horizons except those closest to the surface (18–25 DAE), the proportion of total WU at a particular depth over a 7-day period corre- sponded reasonably closely with the proportion of TRL in the horizon. As the soil was dried out in surface horizons, root activity with respect to WU decreased unless the horizon was replenished with irrigation. Dry crops were forced to exist on soil water alone and as the season progressed the proportional contribution to uptake by lower horizons bore no resemblance to the rooting density (see specific root (a) − 20 −16 −12 −8 −4 0 4 20 May 17 Jun 15 Jul 12 Aug 09 Sep 07 Oct SM D ( m m ) (b) −20 −16 −12 −8 − 4 0 4 SM D ( m m ) Fig. 10. Horizon soil moisture deficits in (a) unirrigated Cara and (b) irrigated Cara in Expt 5. 10 cm (–&–); 20 cm (–%–); 30 cm (–m–); 40 cm (–
–); 50 cm (–$–); 60 cm (–#–); 70 cm (- -&- -) ; 80 cm (- -%- -) ; 90 cm (- -m- -) ; 100 cm (- -
- -) depths. Table 5. Effect of irrigation regime on maximum soil moisture deficit (mm) in different horizons in Expt 1 Horizon (cm) Irrigation regime S.E. (6 D.F.)Dry Moist Wet 0–10 4.5 4.3 1.4 0.59 10–20 10.9 8.9 3.3 0.59 20–30 12.1 8.0 5.7 0.67 30–40 12.5 7.9 6.1 1.01 40–50 13.1 7.1 6.3 0.95 50–60 12.3 6.0 5.0 0.91 60–70 11.9 5.9 4.4 1.33 70–80 10.0 6.1 4.2 1.24 80–90 7.0 6.0 4.4 1.23 Table 6. Effect of irrigation regime on maximum soil moisture deficit (mm) in different horizons in Expt 2 Horizon (cm) Irrigation regime S.E. (6 D.F.)Dry-Dry Dry-Wet Wet-Dry 0–10 5.5 4.5 4.6 0.34 10–20 14.7 11.6 13.8 0.74 20–30 12.8 10.5 13.0 0.26 30–40 10.1 9.3 10.8 0.49 40–50 8.7 5.7 7.9 0.83 50–60 7.7 4.5 6.3 0.95 60–70 6.6 3.7 6.0 0.63 70–80 6.0 2.7 6.0 0.52 80–90 6.4 3.2 6.3 1.00 90–100 6.1 3.0 6.3 0.93 100–110 5.8 2.9 6.1 0.91 384 M. A. STALHAM AND E. J. ALLEN (1–3 days) during which irrigation was not applied to avoid confounding the measurement of soil water with large amounts of water entering the soil (see Gaze et al. 2002). The water use of the crop was calculated for these periods and compared with the PE for the crop. Potential ET was estimated from Kc *ET0, where Kc is a function of ground cover, crop height and stomatal conductance and ET0 is Penman–Monteith reference crop (grass) ET. Initially, only crops with full ground cover were compared but this eliminated some useful data when the canopy was expanding and subjected to high atmospheric evaporative demand, so subsequently all the short measurement periods of soil water content were in- cluded in the analysis. These periods started at c. 40% ground cover in most experiments which eliminated the first 3–4 weeks after emergence. The ratio AE:PE was plotted against the SMD at the start of the measurement period rather than the mean SMD for the period (Fig. 12). The ratio of AE:PE was rela- tively unaffected by increasing SMD up to c. 40 mm (remaining close to 1.0) but then decreased with further increase in the SMD at the start of the measurement period. During the phase when the AE:PE ratio was decreasing, the reduction in AE was greater for Estima thanCara and suggested a cessation of transpiration at a lower SMD (80 cf. 100 mm). Figure 13 shows the AE:PE ratio in Expt 7 during the 1–3 day measurement periods in relation to the average daily ET0 and SMD during the period. In the Dry Cara plots, as the soil water reserves were depleted, AE:PE dropped gradually during May. On 7–8 June, ET0 increased dramatically to an average of 5.65 mm/day resulting in a drop in AE:PE from 0.74 to 0.56 (Fig. 13a). However, over the following 3 days ET0 decreased to 3.05 mm/day and the AE:PE ratio increased back to 0.71. There was a similar, but longer, period of high ET0 in early July (average 4.76 mm/day for a 7 day period) which steadily reduced the AE:PE ratio. However, as earlier, the subsequent decrease in ET0 demand over the next 7 days (2.63 mm/day) resulted in a significant recovery in AE:PE ratio. Clearly, even when plants were under severe water stress (the SMD at the beginning of July was 78 mm), a significant decrease in the evapor- ative demand can allow their root systems to access Table 8. Effect of variety and irrigation regime on rate of water uptake per unit length of root (r10x4 cm3/cm/ day) in different horizons during three periods in Expt 7. (a) 18–25 DAE, (b) 34–41 DAE, (c) 76–83 DAE. (Horizons with negative values became wetter during period studied) Depth (cm) Variety Irrigation regime# Estima Cara S.E. (10 D.F.)Dry CUF Wet Dry CUF Wet (a) 0–10 14.9 41.5 28.2 12.2 28.3 22.1 6.07 10–20 13.0 20.8 15.6 14.5 18.6 13.8 2.39 20–30 6.8 10.3 8.7 9.2 13.0 7.8 2.33 30–40 0.5 9.2 12.4 3.9 5.8 14.1 2.75 40–50 27.3 0 30.6 26.0 0 21.6 11.74 DWU* 0.99 2.08 1.91 1.14 1.82 1.64 0.245 (b) 0–10 0.6 35.1 33.1 5.6 29.9 29.9 2.63 10–20 5.0 14.0 11.4 4.8 10.6 9.2 2.32 20–30 11.2 14.2 11.2 4.8 11.5 9.3 2.54 30–40 6.3 14.0 25.0 4.1 12.9 20.2 1.78 40–60 12.1 19.5 23.8 11.3 14.7 34.4 6.31 60–80 15.7 11.1 10.2 10.3 16.2 0 4.32 DWU* 0.78 2.38 2.49 0.70 2.22 2.28 0.358 (c) 0–10 x8.0 46.3 17.7 x0.2 54.8 45.9 4.50 10–20 x1.1 14.7 11.4 1.6 17.5 14.9 2.97 20–30 x0.9 20.9 20.6 2.0 15.1 15.5 2.10 30–40 x4.5 28.7 22.6 0.4 22.9 22.0 3.02 40–60 x3.2 x4.1 10.9 2.4 6.5 6.5 2.99 60–80 1.5 x2.2 5.4 7.2 3.7 0 2.35 80–100 30.0 x6.0 50.0 80.0 12.0 178.0 56.30 DWU* 0.01 2.61 2.31 0.85 3.27 3.27 0.466 * DWU=daily water use (mm/day). # See Table 1 for details. Water uptake in the potato crop 387 sufficient water in the soil to meet the greater pro- portion of the reduced demand. Similar, temporary, alterations in AE:PE ratio occurred in both irrigated treatments during late June and early July when ET0 averaged 4.5 mm for a 2 week period followed by a cooler period when ET0 decreased and AE:PE ratio recovered (Fig. 13b, c). The data for Estima (not shown) were similar in terms of response of AE:PE ratio to fluctuating ET0 demand. The effect of varying ET demand on the canopy on AE:PE ratio was further analysed by plotting the AE:PE ratio against SMD for different reference crop ET0, 1–2, 2–3, 3–4, 4–5, 5–6 and 6–7 mm. It was felt that using reference crop ET0 rather than the potato crop ET would permit comparison between crops with partial ground cover and those with full ground cover. Using a split-line approach, linear regressions were fitted to the data using the Penman (1970) principle of an abrupt change in the ratio of AE:PE equating to the limiting SMD. In these results, this was the split point for two lines of statistically different slope. Individual analyses were conducted for each variety in each experiment and for all ET0 values close relationships were found which allowed the limiting deficit to be established and the slopes of the lines either side of the limiting deficit (Table 9). Figure 14 presents the data from Table 9 in a visual form that is easier to interpret but for Cara only to avoid excessive duplication of data. The results show a number of important features. First, the AE:PE ratio was close to 1.0 when the SMD was close to field capacity or zero SMD (Fig. 14). Second, there was a sudden change in the slopes of the lines which indicated the limiting SMD. This limiting SMD decreased as the daily ET0 demand increased and was slightly lower for Estima than for Cara, but not significantly so (Table 9). Third, prior to the noticeable change in the relation- ship between AE:PE and SMD at the limiting SMD, the AE:PE ratio was decreasing as the soil became drier even at low SMDs, and became more steeply negative as ET0 increased. This differs from the ap- proach of Penman (1970) and French & Legg (1979) (a) 0.0 0.2 0.4 0.6 0.8 1.0 R at io A E /P E (b) 0.0 0.2 0.4 0.6 0.8 1.0 0 20 40 60 80 100 Soil moisture deficit (mm) R at io A E /P E Fig. 12. Relationship between the ratio of actual (AE): potential (PE) evapotranspiration and soil moisture deficit in (a) Cara and (b) Estima in Expt 7. (a) 0.0 0.2 0.4 0.6 0.8 1.0 R at io A E : PE SM D ( ×1 0 −2 , m m ) 0 1 2 3 4 5 6 7 M ea n E T 0 (m m /d ay ) (b) 0.0 0.2 0.4 0.6 0.8 1.0 R at io A E : PE SM D ( ×1 0 −2 , m m ) 0 1 2 3 4 5 6 7 M ea n E T 0 (m m /d ay ) (c) 0.0 0.2 0.4 0.6 0.8 1.0 17 May 14 Jun 12 Jul 9 Aug 6 Sep R at io A E : PE SM D ( ×1 0 −2 , m m ) 0 1 2 3 4 5 6 7 M ea n E T 0 (m m /d ay ) Fig. 13. Ratio of actual (AE):potential (PE) evapo- transpiration, mean ET0 during the measurement period and soil moisture deficit (SMD) in Cara in Expt 7. (a) Dry; (b) CUF; (c) Wet. AE:PE ratio (&) ; ET0 (%) ; SMD (—). 388 M. A. STALHAM AND E. J. ALLEN who surmised that crops function at potential (i.e. AE=PE) until the limiting SMD is reached. Clearly, the results from the current study differ from this conclusion. Fourth, the rate of decrease in AE:PE as SMD increased beyond the limiting SMD was steeper at low ET0 than at high ET0, and all lines converged to a point (96–98 mm in Cara and 76–79 mm in Estima) where the AE:PE ratio was zero. When combining data from Cara and Estima over Expts 4 and 7, the same type of close relationships were found but as a result of small variation in texture and stone content limiting SMDs were similar. Other varieties also showed similar close relationships (Table 10). DISCUSSION The results presented provide considerable insight into root growth and water uptake in a range of potato crops which have implications for commercial and experimental purposes. All crops rooted to a considerable depth and the variation in final depth was associated with soil conditions. The significance of soil conditions throughout the entire profile cannot be over-emphasized for the contribution of the deepest roots under all water regimes was considerable and much greater than expected. The maximum depths of extraction were considerable (90–120 cm) and these abstraction depths were reached rapidly, typically 55–75 DAE and therefore well before the onset of senescence in maincrop varieties grown in the UK. Roots in deep horizons were found to be capable of taking up water simultaneously with those in the surface horizons irrespective of the soil water content in more superficial horizons, although there was sometimes a lag phase between roots reaching a horizon and then extracting water from it. Crops kept unirrigated for large parts of the season nearly always had a deeper maximum rooting depth than irrigated crops but were considerably sparser in terms of RLD (Stalham & Allen 2001). In such crops, it appeared that roots could extract water from considerable dis- tances ahead of their tips and therefore using maxi- mum rooting depth to assess the depth of water extraction may be an underestimate, especially later in the season. Unless irrigation is excessive and waterlogging or anaerobiosis occurs, soil water status has a much greater effect on the depth of water ex- traction than on maximum depth of rooting. Durrant et al. (1973) observed that depth of water extraction in potatoes was related to, but could be 10–15 cm shallower than, rooting depth, and in the reported experiments many frequently irrigated crops kept 0.0 0.2 0.4 0.6 0.8 1.0 0 20 40 60 80 100 Soil moisture deficit (mm) R at io A E /P E Fig. 14. Relationship between ratio of actual (AE):potential (PE) evapotranspiration and soil moisture deficit for varying daily ET0 in Cara in Expt 7. ET0 (mm): 1–2 (&) ; 2–3 (%) ; 3–4 (m) ; 4–5 (
) ; 5–6 ($) ; 6–7 (#). Table 9. Limiting soil moisture deficit (SMD) and slope of linear regressions between AE:PE and SMD before and after limiting SMD in (a) Cara and (b) Estima in Expt 7 ET0 (mm/day) Limiting SMD S.E.* Slope before limit S.E.* R2 Slope after limit S.E.* R2 (a) 1–2 62.6 2.86 x0.0017 0.00013 0.83 x0.0222 0.00111 0.91 2–3 49.4 2.19 x0.0011 0.00007 0.87 x0.0197 0.00081 0.94 3–4 42.2 2.13 x0.0017 0.00010 0.89 x0.0168 0.00067 0.96 4–5 34.0 1.74 x0.0020 0.00013 0.84 x0.0150 0.00074 0.89 5–6 27.9 1.67 x0.0055 0.00028 0.91 x0.0121 0.00053 0.93 6–7 25.5 1.49 – – – x0.0097 0.00040 0.96 (b) 1–2 59.2 2.90 x0.0015 0.00010 0.82 x0.0294 0.00121 0.94 2–3 45.6 2.21 x0.0015 0.00009 0.84 x0.0286 0.00120 0.97 3–4 36.6 1.85 x0.0038 0.00022 0.85 x0.0230 0.00099 0.97 4–5 34.4 1.92 x0.0061 0.00033 0.88 x0.0203 0.00084 0.94 5–6 26.4 1.49 x0.0081 0.00045 0.87 x0.0170 0.00090 0.90 6–7 23.7 1.40 – – – x0.0168 0.00071 0.95 * S.E.s have variable D.F. Water uptake in the potato crop 389 The fit of the linear regressions of AE:PE versus SMD prior to the limiting SMD was close but poorer than the fits of the lines subsequent to the limiting SMD. This was probably in part because some juvenile crops with undeveloped rooting systems were measured which would have been less capable of extracting soil water particularly at high demand and would affect the relationship between AE:PE ratio and SMD. However, some of these crops had in- complete canopy covers during their expansion phase and therefore would have had a lower daily demand for water which the rooting system could have sup- plied more completely. Further examination of the data for all crops with incomplete canopies showed a cluster of points in unirrigated crops in Expt 1 that had lower AE:PE ratios (0.57–0.70) than expected for the SMD (18–31 mm). The crops maintained at an SMD of 9–15 mm had AE:PE ratios over the same period of 0.89–0.96 in comparison. The daily ET0 in this 12-day period was c. 4.6 mm but frequent measurement of SMD began 10 days after emergence when the canopies were small (c. 20% ground cover) and depth of water extraction shallow. Such extreme demand during May is rare but it does show that young plants can come under water stress even at small SMDs when the rooting system is small. For all irrigation scheduling, it is important to recognize the importance of commencing irrigation just prior to the limiting SMD being reached so that the field can be completely irrigated before plants commence closing their stomata and begin to wilt. Further, when ET demand is extreme, growers have to irrigate to satisfy the demand on the crop canopy and reduce the SMD to a point where roots can function at a lower suction potential. In order to maintain potential evapotranspiration rates in con- ditions of extreme demand that occur infrequently in the UK (e.g. 5–7 mm/day), crops would need to be maintained at SMDs <25 mm but SMDs can be increased when the demand is less extreme. Owing to the frequency of intense or prolonged rainfall events in the UK, irrigation and overall water use efficiency is improved by maintaining higher SMDs since there is a greater capacity for accommodating rainfall and preventing the drainage loss which often occurs on soils maintained at small SMDs. The authors wish to acknowledge financial support from the Ministry of Agriculture, Fisheries and Food (now DEFRA) in the form of a studentship for postgraduate study, Cambridge University Potato Growers Research Association and more recently the British Potato Council (formerly the Potato Marketing Board). Special thanks should be extended to members of the team at CUF for their help. REFERENCES ANON. (1983). Soils of England and Wales. Sheet 4 Eastern England. Soil Survey of England and Wales. Southampton: Ordnance Survey. ASFARY, A. F., WILD, A. & HARRIS, P. M. (1983). Growth, mineral nutrition and water use by potato crops. Journal of Agricultural Science, Cambridge 100, 87–101. BAILEY, R. J. & SPACKMAN, E. (1996). A model for estimating soil moisture changes as an aid to irrigation scheduling and crop water-use studies: I. Operational details and description. Soil Use and Management 12, 122–128. BAILEY, R. J., GROVES, S. J. & SPACKMAN, E. (1996). A model for estimating soil moisture changes as an aid to irrigation scheduling and crops water-use studies: II. Field test of the model. Soil Use and Management 12, 129–133. BELL, J. P. (1987). Neutron Probe Practice. Report 19 (Third Edn). Wallingford, Oxfordshire: Institute of Hydrology. BOONE, F. R., DE SMET, L. A. H. & VAN LOON, C. D. (1985). The effect of a ploughpan in marine loam soils on potato growth. 1. Physical properties and rooting patterns. Potato Research 28, 295–314. BROWN, K. F., MESSEM, A. B., DUNHAM, R. J. & BISCOE, P. V. (1987). Effect of drought on growth and water use of sugar beet. Journal of Agricultural Science, Cambridge 109, 421–435. DAVIES, W. J. & ZHANG, J. (1991). Root signals and the regulation of growth and development of plants in drying soil. Annual Review of Plant Physiology and Plant Molecular Biology 42, 55–76. DENMEAD, O. T. & SHAW, R. H. (1962). Availability of soil water to plants as affected by soil moisture content and meteorological conditions. Agronomy Journal 54, 384–390. DRY, P. R. & LOVEYS, B. R. (1999). Grapevine shoot growth and stomatal conductance are reduced when part of the root system is dried. Vitis 38, 151–156. DRY, P. R., LOVEYS, B. R. & DURING, H. (2000a). Partial drying of the rootzone of grape. I. Transient changes in shoot growth and gas exchange. Vitis 39, 3–7. DRY, P. R., LOVEYS, B. R. & DURING, H. (2000b). Partial drying of the rootzone of grape. II. Changes in the pattern of root development. Vitis 39, 9–12. DURRANT, M. J., LOVE, B. J. G, MESSEM, A. B. & DRAYCOTT, A. P. (1973). Growth of crop roots in relation to soil moisture extraction. Annals of Applied Biology 74, 387–394. FRENCH, B. K. & LEGG, B. J. (1979). Rothamsted irrigation, 1964–76. Journal of Agricultural Science, Cambridge 92, 15–37. FRENCH, B. K., LONG, I. F. & PENMAN, H. L. (1973). Water use by farm crops. II. Spring wheat, barley, potatoes (1969); potatoes, beans and kale (1968). In Rothamsted Report for 1972, Part 2, pp. 43–61. GAZE, S. R., STALHAM, M. A. & ALLEN, E. J. (2002). Accuracy of the neutron probe for measuring changes in soil water storage under potatoes. Journal of Agricultural Science, Cambridge 138, 135–152. 392 M. A. STALHAM AND E. J. ALLEN GREEN, S. R. & CLOTHIER, B. E. (1995). Root water uptake by kiwifruit vines following partial wetting of the root zone. Plant and Soil 173, 317–328. GREGORY, P. J., MCGOWAN, M. & BISCOE, P. V. (1978). Water relations of winter wheat. 2. Soil water relations. Journal of Agricultural Science, Cambridge 91, 103–116. GREGORY, P. J. & SIMMONDS, L. P. (1992). Water relations and growth of potatoes. In The Potato Crop: The Scientific Basis for Improvement (Ed. P. M. Harris), pp. 214–246. London: Chapman & Hall. HAMER, P. J. C., CARR, M. K. V. &WRIGHT, E. (1994). Crop production and water use. III. The development and validation of a water-use model for potatoes. Journal of Agricultural Science, Cambridge 123, 299–311. JEFFERIES, R. A & MACKERRON, D. K. L. (1987). Aspects of the physiological basis of cultivar differences in yield of potato under droughted and irrigated conditions. Potato Research 30, 201–217. KLEPPER, B., TAYLOR, H. M., HUCK, M. G. & FISCUS, E. L. (1973). Water relations and growth of cotton in drying soil. Agronomy Journal 65, 307–310. LONG, I. F. & FRENCH, B. K. (1967). Measurement of soil moisture in the field by neutron moderation. Journal of Soil Science 18, 149–166. MOLZ, F. J. (1981). Simulation of plant-water uptake. In Modeling Wastewater Renovation (Ed. I. K. Iskander), p. 69. New York: John Wiley & Sons. NIMAH, M. N. & HANKS, R. J. (1973). Model for estimating soil water, plant, and atmospheric interrelations. I. Description and sensitivity. Soil Science Society of America Proceedings 37, 522–527. PAYNE, R. W., LANE, P. W., DIGBY, P. G. N., HARDING, S. A., LEECH, P. K., MORGAN, G. W., TODD, A. D., THOMPSON, R., TUNNICLIFFE WILSON, G., WELHAM, S. J. & WHITE, R. P. (1993). Genstat TM 5 Release 3 Reference Manual. Oxford: Clarendon Press. PENMAN, H. L. (1970). Woburn irrigation, 1960–1968. IV. Design and interpretation. Journal of Agricultural Science, Cambridge 75, 69–73. PRESTT, A. J. (1983). Soil management and the water use of potatoes. Ph.D. thesis, National College of Agricultural Engineering, Cranfield Institute of Technology, Silsoe, Bedfordshire. RAMADAN, M. H. A. E. (1986). Modelling for irrigation scheduling with particular reference to the potato crop. Ph.D. thesis, University of London. ROSENFELD, A. B. (1997). Effects of nitrogen and soil con- ditions on growth, development and yield in potatoes. Ph.D. thesis, University of Cambridge. SHARP, R. E. & DAVIES, W. J. (1985). Root-growth and water-uptake by maize plants in drying soil. Journal of Experimental Botany 36, 1441–1456. SINGH, G., BROWN, D. M., BARR, A. G. & JUNG, R. (1993). Modelling soil water status for irrigation scheduling in potatoes. II. Validation. Agricultural Water Management 23, 343–358. STALHAM, M. A. (1989). Growth and water use in the potato variety Record on contrasting sites. Ph.D. thesis, University of Cambridge. STALHAM, M. A. & ALLEN, E. J. (2001). Effect of variety, ir- rigation regime and planting date on depth, rate, duration and density of root growth in the potato (Solanum tuber- osum) crop. Journal of Agricultural Science, Cambridge 137, 251–270. STOLL, M., LOVEYS, B. & DRY, P. (2000). Hormonal changes induced by partial rootzone drying of irrigated grapevine. Journal of Experimental Botany 51, 1627–1634. TAYLOR, H. M. & KLEPPER, B. (1973). Rooting density and water extraction patterns for corn (Zea mays L.). Agronomy Journal 65, 965–968. TAYLOR, H. M. & KLEPPER, B. (1975). Water uptake by cotton root systems: an examination of assumptions of the single root model. Soil Science 120, 57–67. VAN OIJEN, M., DE RUIJTER, F. J. & VAN HAREN, R. J. F. (1995). Analysis of the effects of potato cyst nematodes (Globodera pallida) on growth, physiology and yield of potato cultivars in field plots at the three levels of soil compaction. Annals of Applied Biology 127, 499–520. Water uptake in the potato crop 393