The Auwahi dry forest on leeward Haleakalā volcano, Maui, is among the most tree-diverse forests of the Hawaiian archipelago. However, due to the combined effects of non-native species and human impacts including wildfire, less than 10% of the historic geographic extent of dry forest remained intact by the 1980s. In response to extensive forest degradation, in 1997, regional biologists and ranchers collaborated to initiate experimental ecological restoration at Auwahi forest. The chronological sequence of restoration actions included 1) ungulate exclusion, 2) herbicidal control of the dominant invasive African Kikuyu grass (Pennisetum clandestinum), and 3) dense planting of native shrub seedlings. Fifteen years later, the cover of native shrubs had increased dramatically (3.1% to 81.9%) replacing non-native grass cover (75.4% to 3.3%).
To better understand the ecohydrological impacts of forest restoration, sap flow data (Granier) was collected for remnant native tree species Myrsine lessertiana (Ml) and Nestegis sandwicensis (Ns), both inside and outside of ecological restoration area. In addition, meteorological, soil moisture, and biometric stand data was collected. The following hypotheses were evaluated: 1) soil water availability will be higher in restored native forest compared to unrestored, non-native pasture, and 2) forest transpiration will be higher within the restored forest boundary.
Results/Conclusions
Soil moisture was significantly (paired t-test P<0.001) greater in the restored native forest (VWC=0.24 ± 0.003) compared to unrestored non-native pasture (VWC=0.21 ± 0.005) during the wet season, with greater variability during the dry season (forest VWC=0.16 ± 0.003; unrestored pasture VWC=0.12 ± 0.002). Sap flow (g*cm-2d-1) was higher in unrestored non-native pastures (Ns=41.7 ± 2.83; Ml=43.58 ± 2.20) compared to the restored forest (Ns=29.16 ± 2.26; Ml=22.32 ± 1.70 for), and was driven by solar radiation and wind speed rather than atmospheric vapor pressure deficit (vpd) or soil moisture.
While the data supported hypothesis 1), lower tree transpiration in the restored forest was unexpected and contrary to previous studies, suggesting increased soil water availability in this forest ecosystem does not necessarily lead to increased transpiration. However, higher soil moisture in the restored native understory at Auwahi may influence the dramatic increase in recruitment of native tree seedlings observed over time. Kikuyu grass is a highly invasive species that negatively impacts tropical ecosystems worldwide. These results improve our understanding of the ecohydrological consequences of Kikuyu grass invasion of tropical dry forest ecosystems and suggest that ecological restoration of non-native pastures may prove to have significant watershed implications.