Multivariate and univariate asymmetrical analyses in environmental impact assessment: a case study of Mediterranean subtidal sessile assemblages

Abstract

Characterizing the potential effects of human activities on natural systems is a central problem in applied ecology. This requires the development of analytical procedures able to separate human perturbation from natural spatio-temporal variability displayed by most populations. Beyond-BACI experimental designs provide a framework to address these issues but, to date, their use is limited to the analysis of human impacts on the abundance of single species or other univariate measures. Here, we describe in detail an asymmetrical design that included 1 impact location (I) and a set of 3 controls (Cs), sampled at a hierarchy of spatial scales 4 times over a period of 15 mo. We focused on shallow subtidal assemblages of sessile organisms exposed to sewage discharge along a stretch of coast in southern Italy. The purpose of this paper is to illustrate (1) the comparison of variance components for the assessment of impacts and (2) the use of recently developed multivariate methods (distance-based premutational MANOVA) in the analysis of multivariate species data in response to a complex asymmetrical design. Results indicated that temporal changes in the whole assemblage at I were distinct from those occurring at Cs, and that the nature of this difference (although not its size) was fairly consistent through time. A suite of taxa was identified as important in characterizing the differences found between I and Cs. Some algae (e.g. Colpomenia sinuosa, Gelidium sp. and Pterocladiella sp.), in particular, occurred uniquely at I. Univariate analyses indicated significant Time × I-v-Cs interactions for several taxa, and significantly smaller spatial variation at the scale of quadrats at I compared to Cs. In contrast, the small-scale spatial variation in the number of taxa was significantly greater at I than at Cs. The findings of this study have important implications for future multivariate and univariate analyses in environmental impact assessment.

Effects of sewage pollution on the structure of rocky shore macroinvertebrate assemblages

“…The same partitioning was performed for the residual variability for observations within Imp (Res Imp) within Rs (Res Rs). Appropriate denominators for F ratios were identified from expected mean squares and tests were constructed following the logic of asymmetrical design (see Terlizzi et al, 2005b). All analyses were based on Bray-Curtis similarity of square-root transformed data, and each term in the analysis was tested by 4,999 random permutations of appropriate units.…”

Influence of a breakwater on nearby rocky intertidal community structure

“…Generally, the factor location was partitioned into two components: one contrasting the impacted and control locations (I vs. C's) and the other contrasting control locations among each other (C's). This was also true for the time  location interaction and for the residuals associated with variability among quadrats, respectively within I and C's (see Terlizzi et al, 2005 for a detailed description of these procedures). Prior to analysis, Cochran's test was used to check for problems of heteroscedascity and data were transformed where appropriate (Underwood, 1997).…”

“…However, beyond-BACI designs can be modified and applied to situations where no before data exist and have been widely used (e.g. Chapman et al, 1995;Terlizzi et al, 2005). These designs allow the determination of consistent differences between the disturbed and several control locations although it is generally not possible to attribute causation to a particular event (see Terlizzi et al, 2005;Glasby, 1997 for discussion).…”

“…was adapted from Terlizzi et al (2005) and is presented in Table 1. Generally, the factor location was partitioned into two components: one contrasting the impacted and control locations (I vs. C's) and the other contrasting control locations among each other (C's).…”

Recovery stages over long-term monitoring of the intertidal vegetation in the ‘Abra de Bilbao’ area and on the adjacent coast (N. Spain)

“…Centroids, as arithmetic averages, were calculated using these principal coordinates. The Euclidean distance between each pair of centroids was then calculated and used as the input distance matrix for the nMDS (Terlizzi et al, 2005). The indicator species that contribute most to the multivariate patterns detected were determined by means of the SIMPER routine (Clarke & Warwick, 2001).…”