Are richness patterns of common and rare species equally well explained by environmental variables?

Jack Lennon, Colin M. Beale, Catherine, L. Reid, Martin Kent, Robin, J. Pakeman

Research output: Contribution to journalArticlepeer-review

80 Citations (Scopus)


We investigated relationships between richness patterns of rare and common grassland species and environmental factors, focussing on comparing the degree to which the richness patterns of rare and common species are determined by simple environmental variables. Using data collected in the Machair grassland of the Outer Hebrides of Scotland, we fitted spatial regression models using a suite of grazing, soil physicochemical and microtopographic covariates, to nested sub-assemblages of vascular and non-vascular species ranked according to rarity. As expected, we found that common species drive richness patterns, but rare vascular species had significantly stronger affinity for high richness areas. After correcting for the prevalence of individual species distributions, we found differences between common and rare species in 1) the amount of variation explained: richness patterns of common species were better summarised by simple environmental variables, 2) the associations of environmental variables with richness showed systematic trends between common and rare species with coefficient sign reversal for several factors, and 3) richness associations with rare environments: richness patterns of rare vascular species significantly matched rare environments but those of non-vascular species did not. Richness patterns of rare species, at least in this system, may be intrinsically less predictable than those of common species.
Original languageEnglish
Pages (from-to)529-539
Number of pages11
Issue number4
Early online date14 Dec 2010
Publication statusPublished - Aug 2011

ASJC Scopus subject areas

  • Ecology, Evolution, Behavior and Systematics


Dive into the research topics of 'Are richness patterns of common and rare species equally well explained by environmental variables?'. Together they form a unique fingerprint.

Cite this