Abstract
BackgroundStudying the drivers and determinants of species, population and community spatial patterns is central to ecology. The observed structure of community assemblages is the result of deterministic abiotic (environmental constraints) and biotic factors (positive and negative species interactions), as well as stochastic colonization events (historical contingency). We analyzed the role of multi-scale spatial component of soil environmental variability in structuring earthworm assemblages in a gallery forest from the Colombian “Llanos”. We aimed to disentangle the spatial scales at which species assemblages are structured and determine whether these scales matched those expressed by soil environmental variables. We also tested the hypothesis of the “single tree effect” by exploring the spatial relationships between root-related variables and soil nutrient and physical variables in structuring earthworm assemblages. Multivariate ordination techniques and spatially explicit tools were used, namely cross-correlograms, Principal Coordinates of Neighbor Matrices (PCNM) and variation partitioning analyses.ResultsThe relationship between the spatial organization of earthworm assemblages and soil environmental parameters revealed explicitly multi-scale responses. The soil environmental variables that explained nested population structures across the multi-spatial scale gradient differed for earthworms and assemblages at the very-fine- (<10 m) to medium-scale (10–20 m). The root traits were correlated with areas of high soil nutrient contents at a depth of 0–5 cm. Information on the scales of PCNM variables was obtained using variogram modeling. Based on the size of the plot, the PCNM variables were arbitrarily allocated to medium (>30 m), fine (10–20 m) and very fine scales (<10 m). Variation partitioning analysis revealed that the soil environmental variability explained from less than 1% to as much as 48% of the observed earthworm spatial variation.ConclusionsA large proportion of the spatial variation did not depend on the soil environmental variability for certain species. This finding could indicate the influence of contagious biotic interactions, stochastic factors, or unmeasured relevant soil environmental variables.Electronic supplementary materialThe online version of this article (doi:10.1186/s12898-014-0026-4) contains supplementary material, which is available to authorized users.
Highlights
Studying the drivers and determinants of species, population and community spatial patterns is central to ecology
The Principal Coordinates of Neighbor Matrices (PCNM) approach is part of the distance-based Moran’s eigenvector map (MEM) analysis, which is included in the spatial eigenfunction family of tools [2,25,26] and is a powerful statistical method to model spatial structures at all scales; in other words, the environmental variability is linked to community structure on a multi-scale level [3,24] to obtain new ecological insights [21]
Patches and gaps of species assemblages The SADIE spatial index of aggregation (Ia) index and local vi and vj clustering indices were statistically significant for endogeic species and the group Martiodrilus, Glossodrilus and new genus 2 (1 anecic +2 endogeics), whereas only the vj index was significant for Andiodrilus, Aymara and new genus 1, i.e. one endogeic + two epigeics (Table 2)
Summary
Studying the drivers and determinants of species, population and community spatial patterns is central to ecology. Geostatistics [22] allows the assessment of the spatial distribution of soil environmental variability and soil organisms [22,23], but other powerful statistical tools are necessary to model spatial structures on various scales, such as principal coordinates of neighbor matrices (PCNM) [3,24,25]. The PCNM approach is part of the distance-based Moran’s eigenvector map (MEM) analysis, which is included in the spatial eigenfunction family of tools [2,25,26] and is a powerful statistical method to model spatial structures at all scales; in other words, the environmental variability is linked to community structure on a multi-scale level [3,24] to obtain new ecological insights [21]. It has been used to test and separate the niche from neutral mechanisms that influence the community structure [15], it may appear over-simplistic [27,28]
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
