ABSTRACT: Contaminants may affect ecosystem functioning by reducing the fitness of organisms and these impacts may cascade through ecosystems, particularly if the sensitive organisms are also habitat-forming species. Understanding how sub-lethal effects of toxicants can affect the quality and functions of biogenic habitats is critical if we are to establish effective guidelines for protecting ecosystems. We carried out a global systematic review and meta-analysis critically evaluating contaminant effects on properties of habitat-formers linked to ecosystem functioning. We reviewed a total of 95 publications. However, 40% of publications initially captured by the literature search were identified as having flaws in experimental design and ~11% did not present results in an appropriate way and thus were excluded from the quantitative meta-analysis. We quantitatively reviewed 410 studies from 46 publications, of which 313 (~76%) were on plants and seaweeds, that is macro-algae, saltmarsh plants and seagrasses, 58 (~14%) studied corals and 39 (~10%) looked at toxicant impacts on bivalves, with 70% of those on mussels and the remaining studies on oysters. Response variables analysed were photosynthetic efficiency, amount of chlorophyll a (as a proxy for primary production) and growth of plants, seaweeds and corals as well as leaf area of plants. We also analysed filtration, growth and respiration rates of bivalves. Our meta-analysis found that chemical contaminants have a significant negative impact on most of the analysed functional variables, with the exception of the amount of chlorophyll a. Metals were the most widely harmful type of contaminant, significantly decreasing photosynthetic efficiency of kelps, leaf area of saltmarsh plants, growth of fucoids, corals and saltmarsh plants and the filtration rates of bivalves. Organic contaminants decreased the photosynthetic efficiency of seagrass, but had no significant effects on bivalve filtration. We did not find significant effects of polycyclic aromatic hydrocarbons on any of the analysed functional variables or habitat-forming taxa, but this could be due to the low number of studies available. A meta-regression revealed that relationships between concentrations of metal contaminants and the magnitude of functional responses varied with the type of metal and habitat-former. Increasing concentrations of contaminants significantly increased the negative effects on the photosynthetic efficiency of habitat-formers. There was, however, no apparent relationship between ecologically relevant concentrations of metals and effect sizes of photosynthetic efficiency of corals and seaweeds. A qualitative analysis of all relevant studies found slightly different patterns when compared to our quantitative analysis, emphasising the need for studies to meet critical inclusion criteria for meta-analyses. Our study highlights links between effects of contaminants at lower levels of organisation (i.e. at the biochemical and/or physiological level of individuals) and ecological, large-scale impacts, through effects on habitat-forming species. Contaminants can clearly reduce the functioning of many habitat-forming marine species. We therefore recommend the adoption of routine measures of functional endpoints in monitoring and conservation programs to complement structural measures.
