2018 ESA Annual Meeting (August 5 -- 10)

COS 15-4 - Precipitation variability at multiple time scales drives phenological variation in seasonally dry tropical forests

Monday, August 6, 2018: 2:30 PM
R05, New Orleans Ernest N. Morial Convention Center
Naomi B. Schwartz, Ecology, Evolution, and Behavior, University of Minnesota, St Paul, MN, Xue Feng, Civil, Environmental, and Geo- Engineering, University of Minnesota and Jennifer S. Powers, Department of Ecology, Evolution, and Behavior, University of Minnesota, St. Paul, MN
Background/Question/Methods

Temporal variation in water availability is an important environmental filter in seasonally dry tropical forests (SDTF), characterized by distinct wet and dry seasons. Drought deciduousness is an effective strategy to cope with seasonal water shortages that maximizes carbon gain but avoids dry season water stress. Alternatively, evergreen trees endure seasonal water stress by relying on deep roots and more conservative water use. These contrasting strategies coexist in SDTF, though their relative dominance differs across the biome. While prior studies have assessed the importance of mean annual rainfall and dry season intensity in driving this variation, the global drivers of deciduousness are not well understood. In particular, the timing, contrast, number, and length of wet vs. dry seasons vary across SDTF, as does the magnitude of year-to-year variation in these factors, with likely effects on phenology. Furthermore, the soil template may modulate the relationship between rainfall and phenology through water storage or variation in nutrient availability.

In this study, we used remotely sensed data on phenology across SDTF to ask:

1) How do inter- and intra-annual variability in rainfall affect the degree of deciduousness across global SDTF?

2) How does the soil template modulate the relationship between rainfall and phenology?

Results/Conclusions

We used the correlation between rainfall and satellite-derived photosynthesis at monthly timescales as a proxy for the degree of drought deciduousness, with ecosystems with a higher proportion of deciduous trees showing tighter coupling between monthly rainfall and productivity. Though average rainfall conditions (total rainfall, length and contrast of wet vs. dry seasons) explained a large proportion of the variation in phenology across SDTF, inter-annual variability in these predictors generally had more explanatory power than mean conditions. In general, deciduousness was lower in SDTF with higher inter-annual variability in the amount, seasonality, and timing of rainfall. The soil template modulated the relationships between rainfall regimes and leaf habit, with higher proportions of deciduous trees on soils with low water-holding capacity and higher proportions of evergreen trees on low-nutrient soils. These results suggest that considering inter-annual variability in rainfall seasonality, in addition to average conditions, may be key to understanding SDTF phenology. Phenology in future climates will depend on not just on changes in rainfall amount and seasonality, but also on shifts in the magnitude of inter-annual variability and how those shifts are filtered by soil characteristics.