Get reliable rangeland science

Rangeland ecologists have long worked at a community level and increased attention to the role of biodiversity on ecosystem function has placed greater importance on how we measure and analyze diversity. Conventional measurements such as species richness and diversity indices are simple to calculate and make intuitive sense, but they often fail to account for useful information in species composition data. Multivariate procedures take species identities into account and can describe dissimilarities along environmental gradients, and their use throughout ecology is increasing. We systematically review 153 papers published between 1967 and 2016 that report species composition and/or diversity data in the Society for Range Management research journals (Journal of Range Management and Rangeland Ecology and Management) and delineate four categories of community analysis along a gradient of Very Weak to Very Strong, depending on the level of information on species identity and environmental/management gradients incorporated into the analysis. We identify a trend in the frequency of multivariate techniques; although their use has historically been limited by computational power and access to proprietary software, the Internet has connected an ecology-minded community that has developed numerous analytical options built around free, open-source software. To illustrate the ease-of-use and explanatory power of multivariate analysis and demonstrate potential pitfalls of using diversity indices to describe community composition along environmental gradients, we calculate and compare diversity indices and ordinations using an example dataset in the open-source R statistical environment.