Empirical Evidence that Declining Species Diversity May Alter the Performance of Terrestrial Ecosystems

Shahid Naeem, Lindsey J. Thompson, Sharon P. Lawler, John H. Lawton, Richard M. Woodfin

Abstract

We examined experimentally the association between species diversity and ecosystem processes in a series of terrestrial mesocosms. We developed and maintained 14 mesocosms whose biota were assembled from a single pool of plant and animal species and whose environmental conditions were identically controlled. Each community contained four trophic levels: primary producers (annual herbs), consumers (herbivorous molluscs and phloem sucking insects), secondary consumers (parasitoids) and decomposers (earthworms, Collembola and microbes). All mesocosms received the same diurnal pattern of light, temperature, relative humidity and water. The initial volume of soil, soil structure, composition, nutrient content and inocula of both soil microbes and nematodes were also identical among replicates. The only experimentally manipulated factor was the number of plant and animal species within each trophic level. High, medium and low diversity communities had nine, 15 or 31 plant and animal species, respectively. We measured five ecosystem processes as response variables in these mesocosms over the course of 206 days: (i) community respiration; (ii) productivity; (iii) decomposition; (iv) nutrient retention; and (v) water retention. The manipulation of diversity produced communities that differed significantly in their ecosystem processes. Our results provide the first evidence (obtained by a direct manipulation of diversity under controlled environmental conditions) that ecosystem processes may be affected by loss of diversity.