No abstract

Summary The pattern of a few abundant species and many rarer species is a defining characteristic of communities worldwide. These abundant species are often referred to as dominant species. Yet, despite their importance, the term dominant species is poorly defined and often used to convey different information by different authors. Based on a review of historical and contemporary definitions we develop a synthetic definition of dominant species. This definition incorporates the relative local abundance of a species, its ubiquity across the landscape, and its impact on community and ecosystem properties. A meta‐analysis of removal studies shows that the loss of species identified as dominant by authors can significantly impact ecosystem functioning and community structure. We recommend two metrics that can be used jointly to identify dominant species in a given community and provide a roadmap for future avenues of research on dominant species. In our review, we make the case that the identity and effects of dominant species on their environments are key to linking patterns of diversity to ecosystem function, including predicting impacts of species loss and other aspects of global change on ecosystems.

Abstract. The impact of non-native plant invasions on ecosystems has been controversial because obvious local effects have not yet led to the global extinction of any native plant species on continents and large islands. We suggest that extinction is not the appropriate measure of impact on ecosystem function and present evidence that non-native plant invasions or the replacement of native plants with non-native ornamentals results in significant bottom-up reductions of energy available for local food webs. Using replicated common gardens we compared Lepidoptera species richness and abundance on native plants, non-native congeners of those natives, and non-native species with no close relatives in the study area. Non-native plants supported significantly fewer caterpillars of significantly fewer specialist and generalist species even when the non-natives were close relatives of native host plants. However, the effect size was smaller in the latter category indicating phylogenetic similarity to local natives may positively impact herbivory. Cluster analysis revealed that a non-native plant congener often supports a lepidopteran community that is a subset of the similar, but more diverse community found on its native congener. The proportion of the Lepidoptera community consisting of specialist species was about five times larger across native species within sites compared to non-native plant species. In addition, species accumulation trajectories suggested that in a fully sampled community the differences between the Lepidoptera supported by native and non-native plants may be even greater than presented here.

Aim Efforts to evaluate the impact of non‐native plants on ecosystems rarely consider the role of such plants in sustaining local food webs of arthropods, or whether the effect is similar across arthropod feeding guilds and life stages. We assess this response by comparing arthropod herbivores on native plants (species with evolutionary histories within local food webs) and non‐native plants (species without such histories). We further compare the impact of non‐native plants that are congeners of local native species with those of non‐natives with no close local relatives. Location Suburban landscapes, Mid‐Atlantic United States (Delaware, Pennsylvania). Methods We manipulated the composition of the first trophic level by planting four large common gardens of 50 woody plant species replicated in time and space. After an establishment year, we sampled herbivorous arthropods supported by each plant species and classified them by herbivore species, feeding guild and life stage. Results We found that (1) native plants support larger and more diverse herbivore communities, (2) arthropods with chewing mouthparts and immature herbivores were more sensitive to plant origin than arthropods with piercing‐sucking mouthparts and adult herbivores, (3) arthropods laid more eggs on native plants, (4) internal feeders were rare on all non‐native plants and (5) the reduction in herbivore populations on non‐native plants was smaller, but still significant, if species had a close native relative. Main conclusions Novel ecosystems comprised of plant species with no evolutionary history with local members of higher trophic levels may reduce the diversity and complexity of local food webs. Given the rate at which novel ecosystems are replacing coevolved plant communities worldwide, this result has global implications for the conservation of biodiversity. Using more native plants within human‐dominated landscapes may help maintain the integrity of arthropod communities, as well as the populations of animals that consume arthropods.