The rare species problem: the challenge of complete forest demographic inference

Most species in most forests are rare, with few stems for a given species, and therefore few observations that we can use to infer their dynamics. Rare species are an integral part of the community and can strongly contribute to forest structure and function, and yet we know little about the drivers of their growth and death rates because there are so few of them. Over the next year, I will work with Claire Fortuneland members of the AMAP lab to rebuild the demography of temperate and tropical forests from information derived from individual trees. We will group rare species about which we know litt le according to rules derived from species we know a lot about. Machine learning, Bayesian statistics, traits, phylogeny, and common sense will all be required for this adventure. The result should be a dynamics, updateable pipeline that will produce comprehensive insights into the total demography of global forest inventories.