Direct and indirect effects of deforestation and weather on the demography of an understory herb in Costa Rica

Luis A. Arias-Medellin and Helene H. Wagner, Ecology and Evolutionary Biology, University of Toronto, Mississauga, ON, Canada, Adam S. Hadley, Forest Biodiversity Research Network, Corvallis, OR, Adam S. Hadley, Department of Natural Resources and Energy Development, Government of New Brunswick, New Brunswick, NB, Canada, Matthew G. Betts, Forest Ecosystems and Society, Oregon State University, Corvallis, OR

Background/Question/Methods: Habitat loss and fragmentation can directly affect plant populations by decreasing plant growth, survival and reproduction rates or indirectly by changes in animal movement behavior, which can affect pollination and seed dispersal. Using demographic analysis, we assessed whether direct effects of habitat loss and fragmentation have a higher impact on the understory herb Heliconia tortuosa than indirect effects. In 2013, 2015 and 2017, we counted, tagged and recorded the number of shoots, number of inflorescences and bracts of H. tortuosa individuals inside demographic plots (N=18) differing in forest patch size, forest amount and elevation. Pollination was incorporated by including estimates of fruit and seed set as a function of patch size. The proportion of seeds lost to the matrix was estimated from a movement behavior model parameterized using data from the main seed disperser of H. tortuosa. Seedling recruitment estimates were obtained from field experiments. We performed an elasticity analysis to understand the importance of demographic vital rates for H. tortuosa populations. We ran two Life Table Response Experiments (LTRE) to understand which demographic vital rates explained the observed differences between sampled periods and between the two populations with the lowest and highest population growth rate (λ) within each period.
Results/Conclusions: Demographic vital rates depended mostly on plant size, except for survival, which varied between sampled periods. The growth rate of all populations was lower in 2013-2015 (mean = 0.63) than in 2015-2017 (mean = 0.95). Elasticity analysis showed that survival, growth and seedling establishment had a strong influence on population growth rate, whereas pollination and seed dispersal had a limited effect. The LTREs showed that survival was largely responsible for the observed differences in λ between populations sampled in 2013-2015 and 2015-2017 and between populations with the highest and lowest λ within the same period. Our results suggest that the interannual variation in demography due to climatic effects is far greater than the demographic consequences of forest fragmentation.