Fixed and variable components of evapotranspiration in a Mediterranean wild-olive - grass landscape mosaic

On the transpiration of wild olives under water-limited conditions in a heterogeneous ecosystem with shallow soil over fractured rock

“…As expected, Tt decreased drastically as soil dried in rainfed cultivated olive trees, reaching very low values (Ennajeh et al, 2008;Sofo et al, 2008); this is also commonly represented in ecohydrologic models (Karray et al, 2008;Kutílek and Nielsen, 1994;Laio et al, 2001;Nadezhdina et al, 2015;Rallo and Provenzano, 2013;Verhoef and Egea, 2014). We used sap flow sensors for Tt measurements of the Sardinian wild olive trees (Montaldo et al, 2020), and investigated its relationship with soil moisture; this relationship should be almost invariant according to Detto et al (2006), in contrast with the observations in rainfed cultivated olives noted above.…”

“…Sap flux was measured from August 30, 2011, to December 30, 2011, hereinafter called the "SF-period" (DOY = 243-364, 2011). We also used the recent (from August 22, 2014 to July 9, 2017) sap flux measurements of Montaldo et al (2020) to increase the data extension.…”

“…We compared these results with observations in cultivated rainfed olive trees in Tunisia (Ennajeh et al, 2008) and Italy (Sofo et al, 2008), where instead, as expected, β proportionally decreased with soil moisture below a limiting soil moisture (θl) reaching almost zero values (Figure 3). We then further compared β estimates with the recent observations of Montaldo et al (2020) at the same site in a dry period during 2015 (from 18 June to 19 July). β values were similar to those observed in 2011 (the minimum value of β was a little lower, equal to 0.52, although soil reached drier conditions), confirming the low sensitivity of β to θ (Figure 3).…”

“…Large differences between F fp,t T t SF and ETEC were observed during periods with higher θ (e.g., in September and from 9 November to 30 December of 2011), when soil evaporation was high, and during late fall and winter, when Tg also increased. We compared the differences between F fp,t T t SF and ETEC in the wet September 2011 with those of the dry September 2014 reported in Montaldo et al (2020). The potential evaporation was similar (mean PE = 3.75 mm/d in September 2011, and mean PE = 3.78 mm/d in September 2014), while the differences in precipitation (P = 79 mm in September 2011 and P = 8 mm in September 2014) impacted ETEC, which was almost five times greater in the wet September 2011 (mean ETEC = 1.8 mm/d) than in the dry September 2014 (mean ETEC = 0.4 mm/d) (Figure 4).…”

“…Indeed, as soil dries, the mean ecosystem ET becomes low and the Tt component becomes predominant, especially with low tree cover (e.g., < 40% of the mosaic), as is typical in these ecosystems (Baldocchi et al, 2004;Detto et al, 2006;Kurc and Small, 2004;Montaldo et al, 2008). In this sense, the use of sap flow sensors to estimate tree transpiration (Oren et al, 1998) is very attractive (Bovard et al, 2005;Cammalleri et al, 2013;Hassler et al, 2018;Liu et al, 2015;Montaldo et al, 2020;Oishi et al, 2008;Paço et al, 2009;Williams et al, 2004). At the same time, sap flux measurements may themselves be affected by errors associated with calibration (Köstner et al, 1998;Steppe et al, 2015).…”