17:15 〜 19:15
[AHW28-P01] Drought-induced decline in transpiration in a dry deciduous forest in Cambodia
キーワード：乾燥落葉林、蒸散、土壌乾燥、エルニーニョ
Although tropical dry forests in the Indochina Peninsula have been greatly affected by human activities, some of these natural forests remain in Cambodia. To conserve these valuable forests under climate changing environment, quantitative evaluations of their water usage are necessary. When El Niño occurred, extreme droughts were observed in tropical regions. Indeed, decreases in rainfall during the strong El Niño in 2015-2016 were reported in Cambodia (Kabeya et al., 2021). Dry deciduous forests, representing one major vegetation type in Cambodia, are exposed to higher probabilities of water limitation due to the extreme drought because they overlay areas with relatively shallow soils (Ohnuki et al., 2008). However, there are very few studies revealing the long-term trends in forest transpiration with respect to drought. Here, we measured forest transpiration by sap flow techniques in a dry deciduous forest in Cambodia spanning a total of seven years (2010-2014, 2018, 2020), and analyzed the effects of drought on forest transpiration.
Forest transpiration was evaluated by measurements of sap flux densities with the thermal dissipation method (Granier, 1985). We measured sap flux densities for five trees of Shorea obtusa, three trees of Dipterocarpus tuberculatus, three trees of Terminalia alata, a tree of Xylia xylocarpa. Leaf area index (LAI, m2 m-2) was estimated considering the ratio of photosynthetic photon flux density above forest floor to that above the forest canopy (Iida et al., 2016). Volumetric soil water content was measured at depths of 10, 20, 40 and 60 cm with the time domain reflectometry method, and the weighted mean value up to the depth of 60 cm (SWC, m3 m-3) was obtained. Relative extractable soil water (REW) was calculated as REW = (SWC – SWCmin)/(SWCmax – SWCmin), where SWCmax and SWCmin are maximum and minimum values of SWC during the measurement period, respectively. Responses of canopy conductance (Gs) to environmental variables, that is, vapor pressure deficit (VPD), solar radiation (S), REW and LAI, were evaluated by a multiplicative-type function (e.g., Granier and Breda, 1996) as Gs = Gsref·f(VPD)·f(S)·f(REW)·f(LAI), where Gsref is the reference value of conductance at VPD = 1 (kPa).
Year-to-year differences in REW were observed over seven years. Especially for 2020, a low REW of <0.1 was found at late May, at which REW of around 0.9 was recorded in 2017. This drought was probably caused by a relatively weak El Niño from early to the end of 2019, during which the Southern Oscillation Index (SOI) released by the Australian Government Bureau of Meteorology was around -10. Kabeya et al. (2021) judged the occurrence of El Niño when SOI was less than -10. Forest transpiration and LAI were smaller at that time compared with other years. When a multiplicative-type function was fitted for pooled dataset of whole period, the effect of drought on transpiration was underestimated. Models fitted for each-year dataset reasonably resulted in suitable transpiration calculations with the best performance. The most effective parameter was Gsref·f(VPD), which should be derived from each-year dataset. On the other hand, reasonable calculations were obtained in the case that f(S), f(REW) and f(LAI) were fitted for the pooled dataset. Therefore, the decreases in REW and LAI did not explain the decline in transpiration completely, and the changes in stomatal response to VPD were more essential to understand the fluctuation of transpiration under drought conditions. Long-term data of transpiration indicating changes in stomatal response to varying climatic forcings could contribute to the conservation of these rare natural tropical dry deciduous forests on the Indochina Peninsula.
References: Granier (1985) Ann. Sci. For., 42, 193-200. Granier and Breda (1996) Ann. Sci. For., 53, 537-546. Iida et al. (2016) Ecohydrol., 9, 472-486. Kabeya et al. (2021) JARQ, 55, 177-190. Ohnuki et al. (2008) Hydrol. Process., 22, 1272-1280.
Forest transpiration was evaluated by measurements of sap flux densities with the thermal dissipation method (Granier, 1985). We measured sap flux densities for five trees of Shorea obtusa, three trees of Dipterocarpus tuberculatus, three trees of Terminalia alata, a tree of Xylia xylocarpa. Leaf area index (LAI, m2 m-2) was estimated considering the ratio of photosynthetic photon flux density above forest floor to that above the forest canopy (Iida et al., 2016). Volumetric soil water content was measured at depths of 10, 20, 40 and 60 cm with the time domain reflectometry method, and the weighted mean value up to the depth of 60 cm (SWC, m3 m-3) was obtained. Relative extractable soil water (REW) was calculated as REW = (SWC – SWCmin)/(SWCmax – SWCmin), where SWCmax and SWCmin are maximum and minimum values of SWC during the measurement period, respectively. Responses of canopy conductance (Gs) to environmental variables, that is, vapor pressure deficit (VPD), solar radiation (S), REW and LAI, were evaluated by a multiplicative-type function (e.g., Granier and Breda, 1996) as Gs = Gsref·f(VPD)·f(S)·f(REW)·f(LAI), where Gsref is the reference value of conductance at VPD = 1 (kPa).
Year-to-year differences in REW were observed over seven years. Especially for 2020, a low REW of <0.1 was found at late May, at which REW of around 0.9 was recorded in 2017. This drought was probably caused by a relatively weak El Niño from early to the end of 2019, during which the Southern Oscillation Index (SOI) released by the Australian Government Bureau of Meteorology was around -10. Kabeya et al. (2021) judged the occurrence of El Niño when SOI was less than -10. Forest transpiration and LAI were smaller at that time compared with other years. When a multiplicative-type function was fitted for pooled dataset of whole period, the effect of drought on transpiration was underestimated. Models fitted for each-year dataset reasonably resulted in suitable transpiration calculations with the best performance. The most effective parameter was Gsref·f(VPD), which should be derived from each-year dataset. On the other hand, reasonable calculations were obtained in the case that f(S), f(REW) and f(LAI) were fitted for the pooled dataset. Therefore, the decreases in REW and LAI did not explain the decline in transpiration completely, and the changes in stomatal response to VPD were more essential to understand the fluctuation of transpiration under drought conditions. Long-term data of transpiration indicating changes in stomatal response to varying climatic forcings could contribute to the conservation of these rare natural tropical dry deciduous forests on the Indochina Peninsula.
References: Granier (1985) Ann. Sci. For., 42, 193-200. Granier and Breda (1996) Ann. Sci. For., 53, 537-546. Iida et al. (2016) Ecohydrol., 9, 472-486. Kabeya et al. (2021) JARQ, 55, 177-190. Ohnuki et al. (2008) Hydrol. Process., 22, 1272-1280.