Sorghum water loss in relation to irrigation treatment

Aiming towards rational irrigation water management in a seasonally dry cli mate, sorghum water losses via evapotranspiration were studied during a two -year experiment in relation to irrigation treatments in Central Greece. Re lative to high irrigation (I-H), that provided the root depth with 458 mm o f water in 1994 and 512 mm in 1995, 56 and 64% of the water was supplied by the medium (I-M) and 34 and 46% by the low (I-L) treatments, respectively, during the two years. A fourth treatment (I-HA) was performed like (I-H) u ntil the end of anthesis, when irrigation stopped. Gravimetric soil moistur e was measured, biometric measurements were taken and all meteorological pa rameters required to estimate evapotranspiration by the Penman-Monteith equ ation were logged. A model estimating sorghum actual water loss was first run with the 1994 da ta. During the model-establishment year, it was found that (a) surface resi stance r(s), consisting of a canopy r(sc) and a soil r(ss) resistance actin g in parallel, was almost exclusively dependent on soil water shortage, (b) under the I-M and I-L irrigation treatments, the lowest possible (immediat ely after water application) canopy resistance r(sc)', higher than the (I-H ) minimum canopy resistance r(sc) (min) = 40 sm(-1), was irrigation-deficit dependent and (c) the r(ss) (min) was as high as 1200 sm(-1), common to al l treatments. The model established was then verified with the 1995 data and used to calc ulate the crop coefficient k(c) values for sorghum. The model, although ten ding to underestimate actual evapotranspiration by 4-10%, depending on the treatment, may be considered as reliable. The k(c) values calculated are co nsiderably higher than the k(c) values suggested for sorghum by the Food an d Agriculture Organization (FAO). Therefore, taking also into account that any additional mm of water supplied results in an increase of 0.052 t of dr y biomass per hectare, higher irrigation water applications could be recomm ended, although the low irrigation treatment made slightly better use of wa ter.