Exploring the use of rocky shore biotic assemblages as indicators of environmental condition

Abstract

A first evaluation was made to use macroalgae to discriminate between seven rocky shores around the Maltese Islands with different water quality conditions. Selection of sites was based on the results of long-term monitoring of levels of nutrients, chlorophyll a and water transparency as indicators of environmental quality (Axiak, 2004). Site locations included the Grand Harbour, which is subject to episodes of eutrophication, and a reference site in Gozo, considered amongst the most pristine available locally, with the other sites falling between these extremes. All sites had the same substratum type and comparable slopes but were subject to different degrees of anthropic impact. The study focused on the macroalgal communities from the mediolittoral to uppermost levels of upper infralittoral zone, since these assemblages could be easily accessed from the shore. Five replicate 0.5 m x 0.05 m quadrats were placed at random in each of the upper infralittoral, lower mediolittoral, upper mediolittoral and supralittoral zones at each site and the percentage cover of each species of macroalga and of macroinvertebrates was recorded in spring, summer and winter each year in 2003 and 2004. A total of 112 macroalgal and 32 macroinvertebrate species were recorded during the whole study period from all zones of all sites considered collectively. The highest percentage of macroalgal taxa were recorded from the upper infralittoral zone followed by the lower mediolittoral, upper mediolittoral and supralittoral zones. The highest percentage of macroinvertebrate taxa was recorded from the lower mediolittoral zone, followed by the upper infralittoral, upper mediolittoral and supralittoral zones. The results obtained showed that distinct macroalgal assemblages inhabited the selected sites and that the assemblage at the reference site was the most dissimilar compared to the rest. Macroinvertebrates did not discriminate between sites. The Ecological Evaluation Index (EEI) of Orfanidis et al. (2001) discriminated among sites, however, two sites, St. Angelo and Qbajjar, gave anomalous results. The same EE] values were obtained for each site in 2003 and in 2004, except for one site, where a ‘good’ result was obtained in 2003 and a ‘moderate’ one in 2004. Marsascala-J, Manoel Island and St. Angelo formed one cluster of ‘good’ Ecological Status sites, Qbajjar and Marsascala-W formed another cluster of ‘moderate’ Ecological Status sites, whereas St. Paul’s Bay (‘high’ Ecological Status) and Birzebbuga (‘low’ Ecological Status) were separate from the rest. The Rhodophyta/Phaeophyta Mean Ratio Index (R/P rt. I.) proposed by Giaccone and Catra (2004) was consistent in the classification of each site in 2003 and 2004. According to this index, Birzebbuga and St. Angelo had a ‘bad’ Ecological Status, Marsascala-J and -W had a ‘low’ Ecological Status while Manoel Island, St. Paul’s Bay and Qbajjar had a ‘high’ Ecological Status. The two indices considered classified the seven sites differently, agreeing only on St. Paul’s Bay, which both classified as ‘high’. Cluster analysis and Non-Metric Multidimensional Scaling (nMDS) ordination plot carried out on the upper infralittoral algal data on ratio of Rhodophyta to Phaeophyta resulted in three clusters, corroborating the results of R/P rt. I. The BENTHOS methodology, based upon Detrended Correspondence Analysis (DCA), discriminated sites having a different water quality, corroborating nMDS and Principal Component Analysis (PCA) results. Correlations between environmental variables and Axis 1 of the DCA plot were not significant even though the Pearson correlation coefficients were considerably different from zero. This is probably due to the small sample size of seven sites. A list of only 39 species was able to place the seven sites along an environmental gradient. When datasets for separate years was used, sites from the same location grouped near each other, indicating community composition stability from year to year. ANOSIM revealed that the interannual variation in the macroalgal assemblage cover of Chlorophyta, Phaeophyta and Rhodophyta, vegetation layers, functional group, and the number of annual and perennial species, was not significant for the upper infralittoral zone, however, a significant difference between sites as well as between seasons occurred in this zone. The SIMPER procedure showed that species within the functional groups ‘foliose’, ‘crustose’ and ‘corticated terete’ mostly contributed to the patterns observed in the cluster and nMDS analyses. On the basis of these multivariate analyses, it was concluded that functional groups did discriminate between sites. Seasonal variations in macroalgal and macroinvertebrate taxa and percentage cover were investigated by computing various diversity measures. The RELATE test for cyclicity, showed that macroalgal species followed a cyclical pattern of variation in all sites in all four zones studied. The Average Taxonomic Distinctness (AvTD) and Variation in Taxonomic Distinctness (VarTD) showed a significant strong negative Pearson’s correlation between them in the case of lower mediolittoral macroalgae and upper mediolittoral and a non-significant value for the upper infralittoral macroalgae. When the five Taxonomic Distinctness indices were plotted for the mean upper infralittoral percentage cover by species for macroalgae, none showed a progressive decrease in value when plotted against the seven sites that were positioned along a water quality gradient. Analyses on different levels of taxonomic resolution showed that the level of ‘Family’ retained sufficient information to detect the impacts of water quality on the benthic assemblages. The results confirm that the type of transformation used introduces major changes in the multivariate pattern of the samples and to a greater extent than the taxonomic level at which organisms are identified. The cost/benefit ratio was calculated to identify the taxonomic level with the least loss of information and the least taxonomic effort. The results obtained seemed to suggest that identification of macroalgae to the genus level is the most cost-effective. The overall result of the present study was that the percentage cover and taxonomic composition to at least genus level of the upper infralittoral assemblage of macroalgae on Maltese rocky bottoms can be used to discriminate between sites along a water quality gradient, but that rocky shore macroinvertebrates are not suitable for this purpose. The EEI, R/P rt. I. and BENTHOS methodologies proposed for use to meet the obligations of the European Union’s Water Framework Directive were not concordant in their assignment of Ecological Status when applied to the Maltese rocky shores studied. The BENTHOS methodology was found to be better than the other two indices in assigning an Ecological Status to a particular site. The EEL, R/P rt. I. and BENTHOS methodology all showed variation with season.