Identifying the multi‐scale spatial structure of plant community determinants of an important national resource. (19th March 2013)

• Record Type: Journal Article
• Authors: Lewis, Rob J.; Pakeman, Robin J.; Marrs, Rob H.; Podani, Janos
• Is Part Of: Journal of vegetation science. Volume 25:Number 1(2014:Jan.)
• Page Start: 184
• Page End: 197
• Publication Date: 2013-03-19
• Subjects: Plant ecology -- Periodicals; Plant communities -- Periodicals; Plant populations -- Periodicals; 581.7
• DOI: 10.1111/jvs.12071
• Languages: English
• ISSNs: 1100-9233

Abstract

<abstract abstract-type="main" xml:lang="en" id="jvs12071-abs-0001"> <title>Abstract</title> <sec id="jvs12071-sec-0001" sec-type="section"> <title>Questions</title> <p>What is the significance of climate and land‐use management as determinants of plant species composition of Scotland's soft coasts, and how are these determinants spatially scaled? What is the relative contribution of different community assembly processes in governing the coastal plant communities of an important national resource?</p> </sec> <sec id="jvs12071-sec-0002" sec-type="section"> <title>Location</title> <p>Scotland, UK.</p> </sec> <sec id="jvs12071-sec-0003" sec-type="section"> <title>Methods</title> <p>We used national‐scale survey data of Scotland's soft coasts, and a subset representative of machair grassland, a conservation priority habitat. Principal coordinates of neighbour matrices (PCNM), an eigenvector‐based method, was used to assess the spatial component of environmental determinants at multiple scales. The variation‐partitioning framework was applied to unravel the scale‐specific importance, relative to the study design (Broad > 50 km, Meso 10–50 km, small + fine 1.5–10.0 km) of each environmental predictor set.</p> </sec> <sec id="jvs12071-sec-0004" sec-type="section"> <title>Results</title> <p>Modelled environmental and spatial predictors captured ca. 20% of the variation for both response matrices. Management predictors captured significant proportions, identifying vegetation structure, proxies for grazing intensity and disturbance as important descriptors of patterns in species composition for both data sets. The spatial scale of management predictors was poorly captured by modelled PCNM variables, suggesting spatially dependent management variables operate at finer spatial scales than this study detected. Climate also captured significant, yet smaller fractions of variation compared to management. Potential evapotranspiration (PET) and humidity were identified as important climatic determinants of species composition operating entirely at the broad spatial scale. Pure spatial fractions across all scales were significant (<italic>P</italic> ≤ 0.001) for both data sets, alluding to unmeasured, spatially structured environmental variables such as soil chemistry and/or exposure at larger scales, and potentially biotic process such as seed dispersal at the finest detectable scale.</p> </sec> <sec id="jvs12071-sec-0005" sec-type="section"> <title>Conclusions</title> <p>The use of spatial PCNM variables within the variation‐partitioning framework is a valuable tool for dissecting scale‐specific importance of environmental determinants of species composition. This study reveals important climatic and management determinants of Scotland's soft coast and machair vegetation, and will help to better understand the relative scale at which they operate.</p> </sec> </abstract>