Diagnosing restoration trajectories using demographic modeling and modern coexistence theory

Lina Aoyama and Lauren M. Hallett, Environmental Studies Program and Biology Department, University of Oregon, Eugene, OR, Benjamin Gilbert, Ecology and Evolutionary Biology, University of Toronto, Toronto, ON, Canada, Lauren G. Shoemaker, Botany Department, University of Wyoming, Laramie, WY, Akasha Faist, Department of Animal and Range Sciences, New Mexico State University, Las Cruces, NM, Sharon Kay Collinge, University of Colorado, Nancy Shackelford, nstitute of Arctic and Alpine Research, Colorado University, Boulder, CO, Vicky Temperton, Leuphana University of Lüneburg, Germany, Oscar Godoy, INMAR, Universidad de Cadiz, Puerto Real, Spain, Loralee Larios, Botany and Plant Sciences, University of California-Riverside, Riverside, CA, Gyorgy Barabas, Theoretical Biology, Linkoping University, Linkoping, Sweden, Emma Ladouceur, The German Centre for Integrative Biodiversity Research (iDiv), Leipzig, Germany, Margaret Mayfield, School of Biological Sciences, The University of Queensland, Brisbane, Australia

Background/Question/Methods
Understanding what modulates population dynamics is fundamental for identifying mechanisms driving successful ecological restoration. Assessing the level of restoration success is often achieved through the comparisons of target species abundance in both restored and associated reference populations. However, abundance in the first few years may be a poor predictor of long-term success because 1) populations established early through seed additions may not be temporally stable, and 2) environmental variation affects population fluctuations over time. Demography-based population models coupled with principles from Modern Coexistence Theory (MCT) offer a potentially powerful way to diagnose restoration trajectories. Here, we assessed the restoration trajectories of an endangered plant species (Lasthenia conjugens) in California vernal pools. To test whether per capita intrinsic growth rate is a better metric than abundance to assess long-term population trends, we modeled the population dynamics of L. conjugens in both reference and restored communities leveraging 16 years (1999-2015) of annual census data. To understand what contributed to its population level change and when to take management actions, we then analyzed the year-to-year environmental and competitive effects on L. conjugens, and simulated the effects of active management (0, 50, 75% exotic grass removal) on population dynamics of L. conjugens.
Results/Conclusions
The mean annual per capita intrinsic growth rates of L. conjugens estimated from demographic models were better predictors of its long-term population trend than its observed mean annual abundances. We observed rainfall variability stabilizing L. conjugens in reference communities between 2005 and 2008, as predicted by key mechanisms from MCT. This stability, however, was overridden by litter accumulation which was destabilizing when exotic grasses dominated and persisted since 2008. Overall, the environmental effects on L. conjugens positively correlated with the competitive effects, indicating that the good years for L. conjugens were also good years for its competitors. The partitioning of competitive effects revealed that exotic grasses were a source of concern. Thus, we focused on exotic grass removal as a management action to increase population growth of L. conjugens. From simulations, we found that the effect of exotic grass competition on L. conjugens was non-linear, and that exotic grass should be removed in years favorable to L. conjugens. Identifying what facilitates successful restoration and desirable restoration trajectories is projected to be an increasingly important pursuit as climate change. Our approach shows promise to improve the assessment of restoration success over time and indicate a pathway to desired restoration outcomes.