Figure 13: Two-weekly T for kiwi in 2003 and 2004 (Guimaraes, Portugal), accessed by Granier method with correction.
ETo was calculated according to Allen et al. (1998) with data from nearby standard meteorological station.Taking the terminology of the dual crop coefficient approach (Allen et al., 1998), where Kc (Kc = Kcb + Ke) , the average basal crop coefficient, Kcb, increased from 0.4 in April to 0.7 from May to August and finally to 0.9 from September on (details in Silva et al., 2007).
The average soil water evaporation coefficient (Ke) was 0.35 during July (2003) and 0.20 during August (2004). The resulting Kc for these periods was 0.9-1.0. The corresponded values suggested by Allen et al. (1998) are 1.05 for mid and for late season.
4.3 Peach Orchard
The experimental work for the last case-study - peach (Prunus persica [L.] Batsch, cultivar Silver King) - took place in a 60 ha orchard near Montijo, Portugal (latitude 38° 42'
1 Kc - crop coefficient; Ke - soil evaporation coefficient; Kcb describes plant transpiration while Ke describes soil evaporation.
According to predawn leaf water potential measurements, the plants were free of water restrictions at all times.
N, longitude 8° 48' W, elevation near 0), in the summers of 1998 and 1999. The region has cool, wet winters and hot, dry summers with average annual rainfall around 600 mm and mean air temperature around 16° C. The soil was sandy (an Arenosol, according to FAO classification). The trees were planted in 1996, at 5 x 2 m spacing and were drip irrigated (2000 emitters/ha with a flow rate of water of about 3.5 l/hour/emitter). Tallest branches reached 3 to 3.5 m. Ground cover, based on shadowed areas near solar noon, was 29% and the leaf area index was around 1.2 (1998) and 1.4 (1999). The orientation of rows was 13° NNE and dominating winds in the region blow between north and west directions.
EC data were collected between the 21st June and the 4th September in 1998 and between the 9th July and the 11th August for 1999. The sensors used were a 1-D sonic anemometer with a fine wire thermocouple and a krypton hygrometer (respectively, models CA27, 127 and KH20 from, Campbell Scientific, Inc. Logan, UT, USA). Rn and G were measured as described in Pa?o (2003) and Pa?o et al. (2006).
Sap flow was measured from June to September using six sensors each year in the irrigated plot, with 1 cm (1998) or 2 cm length (1999). Corrections were explicitly made for the influence of temperature natural gradients in the trunk and to take into account the radial profile of sap flux distribution.
Soil evaporation (Es) was measured using nine cylindrical microlysimeters, 15 cm of internal diameter and 12 cm height, built and used as described in Daamen et al. (1993). Five lysimeters were located on the row between two trees and the other four between rows, near the limit of the vertical projection of the canopy. The soil cores were taken from a different place every day and the lysimeters reinstalled. Es was calculated by a weighted average in relation to the area represented by each lysimeter (with respect to distance from emitter) and results were cumulated for daily values of soil evaporation. Long term Es was obtained with specifically developed models, where Es is estimated from ETo but using an adjustment which is a function of the available energy at the soil surface (Concei?ao, et al., 2004). Reference and crop evapotranspiration (ETo and ETc) were computed according to Allen et al. (1998), using the dual crop coefficient approach for ETc.
As before, transpiration measured by the Granier method (Tgr) showed an important underestimation (over 80%) when compared to T obtained with EC method and soil evaporation measurements (ETEC-Es). However, a strong correlation was found between Granier and EC data sets. The best relationships between ETEC-Es and Tgr were: 1998: ETec - Es = 1/(-2.19 + 3.77 exp (-Tgr) with r2 = 0.89, for Tgr values between 0.2 and 0.4 mm/day) and 1999: ETEC - Es = 0.75/ (-1.69 + 3.38 exp(-Tgr+0.08) with r2 = 0.95, for Tgr values roughly between 0.3 and 0.5 mm/day.
Differences between years may have been due to insufficient sampling or, eventually, to a possible misrepresentation of the shape of the radial profile of sap flow density, combined with the use of probes of different lengths. Regardless, the two relationships were respectively used to obtain long-term transpiration for the orchard under study for the two years which, combined with daily Es, provided ET from June to September. Mean daily ET during that period was close to 2 mm.
Crop coefficients, determined from ETec and ETo, varied between 0.3 and 1 (the highest values occurring in cloudy days), with a mean value close to 0.5. A previous study, developed in the same field site for a shorter period, in 1998 (Snyder et al., 2000), had shown already that the crop coefficient was around 0.5 on average. This value is much lower than the crop coefficient tabulated in Allen et al. (1998) for peach (0.9). An improved approach uses an adjustment for sparse vegetation and accounts for plant and soil contribution by using the dual crop coefficient method (Allen et al., 1998). The orchard had 1000 trees/ha, considered a medium density (Grappadelli and Sansavini, 1998), although it is higher than the mean density of Portuguese peach orchards (around 800 trees/ha). However, the low ground cover (29%), clearly justified the use of that adjustment for sparse vegetation. After such adjustment, the mean crop coefficient estimated for the period under study was close to 0.7.
This value is closer to the measured crop coefficient (^ 0.5) than the tabulated value, but still overestimates orchard water use. Daily ET calculated with the adjusted crop coefficient (ETc) was over 35% higher than the measured ET. Other details can be found in Pa?o et al. (2006).
In conclusion, even in cases where a specific relationship TEC and Tgr seems to be only of value for the specific conditions of the study, and independently of the importance of the underestimation, it was possible to obtain long term estimates using this approach. The condition is that both methods can be used simultaneously for a period of time sufficient to obtain the relationship EC vs. SF and lysimeters measurements can be performed to get Es.
Was this article helpful?