Elsevier

Marine Pollution Bulletin

Volume 58, Issue 12, December 2009, Pages 1850-1859
Marine Pollution Bulletin

Taxonomic sufficiency in the detection of natural and human-induced changes in marine assemblages: A comparison of habitats and taxonomic groups

https://doi.org/10.1016/j.marpolbul.2009.07.018Get rights and content

Abstract

Taxonomic Sufficiency (TS) is a promising analysis technique, particularly in light of the current need for rapid and reliable procedures in marine impact assessment and monitoring. However, generalizations are still difficult and there are few studies comparing the effectiveness of TS under different environmental settings. The present study investigates whether reduced taxonomy can be used to detect natural and human-driven patterns of variation in mollusk and polychaete assemblages from subtidal soft and hard bottoms in the Mediterranean. Results showed that, unlike in polychaetes, mollusk families represent effective taxonomic surrogates across a range of environmental contexts. These findings suggest that the mechanisms behind TS in mollusks could act homogeneously across habitats and environmental conditions. In contrast, multiple factors could interact to determine the robustness of polychaetes to taxonomic aggregation. This study highlights the need to go beyond the current pragmatism in this field of work and focus on the reasons underlying TS effectiveness in order to provide a general framework on the application of taxonomic surrogates in marine systems.

Introduction

As anthropogenic disturbances in marine systems have increased rapidly and dramatically worldwide in recent decades (Halpern et al., 2008), the development of fast and cost-effective procedures for impact assessment and monitoring has become a pressing issue for marine ecologists (Warwick, 1993, Bell et al., 2006). In this context, the concept of Taxonomic Sufficiency (TS; Ellis, 1985) has been receiving increasing attention in recent years. The basic concept of TS is that anthropogenic impacts can be detected using coarse taxonomic resolutions without a significant loss of information, thus avoiding long and difficult precise taxonomic identifications and saving time and costs (Olsgard and Somerfield, 2000). This concept is based on the principle that the effects of an environmental stress could propagate up through the taxonomic hierarchy as its magnitude increases (Pearson and Rosenberg, 1978, Ferraro and Cole, 1990), causing noticeable variation in assemblage structure above the species level. It has also been suggested that higher taxa may reflect the effect of human disturbance even better than species, with the latter being more sensitive to the confounding influence of natural environmental variation (Warwick, 1988a, Warwick, 1988b, Vanderklift et al., 1996).

Although TS has also been adopted in conservation studies in marine systems (Vanderklift et al., 1998, Bates et al., 2007), it has mainly been applied in the assessment of environmental impacts. When lumping species abundance into higher taxonomic ranks, the effects of anthropogenic disturbance on assemblage structure have been clearly detected at the genus or family level (e.g., Gomez Gesteira et al., 2003) and, in some cases, even at the phylum level (e.g., Defeo and Lercari, 2004). However, several authors have pointed out that differences in assemblage structure among impacted and reference conditions are generally less evident when using a decreasing level of taxonomic resolution (Somerfield and Clarke, 1995, Olsgard et al., 1998), whereas the opposite has been rarely documented (but see Smith and Simpson, 1993).

Currently, it is commonly acknowledged that family-level identification may be sufficient to examine variation in marine assemblages exposed to environmental impacts in routine monitoring programs (e.g., Somerfield and Clarke, 1995, Olsgard et al., 1998, Gomez Gesteira et al., 2003, Thompson et al., 2003, Lampadariou et al., 2005). Family is also likely to be a good predictor of species-level variation related to natural environmental gradients (De Biasi et al., 2003, Dethier and Schoch, 2006). However, it is still difficult to make generalizations regarding predictors of variation. The use of taxonomic surrogates in investigating patterns of assemblage variation along natural gradients still remains substantially unexplored (Włodarska-Kowalczuk and Kędra, 2007). In impact assessment, the application of TS has focused on strong pollution gradients, mostly including oil and heavy metal pollution (see Dauvin et al., 2003 for a review). The use of coarser taxonomic resolution in the analysis of weak or intermediate human threats has largely been disregarded. Most studies have examined soft bottom benthic assemblages in the North Atlantic in relation to oil pollution gradients (Herman and Heip, 1988, Warwick, 1988a, Warwick, 1988b, Somerfield and Clarke, 1995, Olsgard et al., 1998, Olsgard and Somerfield, 2000). Fewer attempts have been made to test TS in other geographic contexts or in habitat types such as tropical areas (Torres Mendes et al., 2007) or rocky shores (Lasiak, 2003).

The major risk in the use of TS is that the potential loss of information related to the decreased taxonomic detail could be underestimated, especially in poorly known systems or when pilot studies on the application of this technique are not available (Chapman, 1998, Narayanaswamy et al., 2003, Quijón and Snelgrove, 2006). Reduced taxonomic information could restrict inferences about the causality of observed patterns (Defeo and Lercari, 2004), or could fail to highlight the subtle effects of environmental stress on assemblage structure. For instance, low taxonomic resolution could reduce differences in the multivariate structure of assemblages between impacted areas and controls (Somerfield and Clarke, 1995, Vanderklift et al., 1996, Pagola-Carte et al., 2002), and the loss of information could be even more severe when low taxonomic detail is coupled with strong data transformations (Olsgard et al., 1997).

The sufficient level of taxonomic resolution could be strongly context-dependent (Terlizzi et al., 2003a) and could change according to the bio-geographic background (Roy et al., 1996) or the habitat type (Chapman, 1998) or due to different relationships of abundance and redundancy among species. Since the degree of taxonomic relatedness of species differs among phyla (Warwick and Somerfield, 2008), the ability of higher taxa to reflect species-level variation could also vary depending on the type of organisms involved (Vanderklift et al., 1998, Hirst, 2006). To date, the application of TS to single taxonomic groups is scarce, even for those groups widely employed in marine environmental benthic monitoring such as mollusks (Terlizzi et al., 2005a, Torres Mendes et al., 2007) or polychaetes (Giangrande et al., 2005, Domínguez-Castanedo et al., 2007). TS studies have generally focused at the assemblage level, and have included different groups of organisms. However, changes in phylogenetic ranking among phyla make conclusions based on taxonomic relationships more stringent when applied to a single phylum than to all phyla combined (Ellingsen et al., 2005). Moreover, the success of taxonomic surrogates could rely on mechanisms such as the numerical correlation among species and higher taxa, potential relationships between functional and taxonomic diversity, and different responses to human disturbance, which could be strongly group-dependent. The application of TS to whole assemblages, although of great interest from a practical point of view, could hide basic explanations for the suitability of taxonomic surrogates. In this light, studies comparing variation in the taxonomic sufficient level among different organisms, habitats and/or environmental conditions would also be of great help to building a general conceptual framework for the application of TS. The responses to reduced taxonomic resolution for detecting natural and human-driven patterns of variation have been tested in mollusk and polychaete assemblages from subtidal soft and hard bottoms. We examined eight existing data sets from previous studies on the effects of human impacts or natural environmental gradients on distributional patterns of assemblages in the Mediterranean. Specifically, we compared the effects of taxonomic aggregation on the analysis of multivariate patterns of assemblage variation related to natural gradients and human impacts among considered habitats and organism types. This allowed us to assess whether (1) the degree of sufficient taxonomic resolution varies between organisms and/or habitats; and (2) impacts produce patterns of variation in assemblage structure that are more robust to TS than natural environmental gradients.

Section snippets

Datasets

A total of eight data sets from Mediterranean benthic assemblages under different ecological and biological settings were analyzed. Datasets were chosen because both groups showed significant differences in the structure of mollusk and polychaete assemblages at the species level between impacted and control locations or along environmental gradients, and because they referred exactly to the same sources of anthropogenic disturbance and natural gradients. A detailed description of the data sets

Results

Results of PERMANOVAs carried out at different levels of taxonomic resolution and data transformation on all data sets are summarized in Table 2. For ‘Molluscs LB’, PERMANOVA at the species level highlighted significant differences (P < 0.001) in mollusk assemblages at increasing distances from the offshore gas platform. Pair-wise comparisons revealed that assemblages differed among the platforms at all distances, a result that was similar to results observed for the whole benthic assemblage (see

Discussion

Our findings revealed that mollusk families, and even orders, are reliable proxies of the whole soft bottom benthic assemblage in assessing the effects of offshore gas platforms in the Mediterranean. Patterns of variation in polychaetes, in contrast, were not consistent with the general response of whole assemblages at both the species level and at higher taxonomic levels. Opposing results have been found in studies assessing the effects of offshore oil platforms on soft bottom benthic

Acknowledgements

C. Vaglio and G. Guarnieri provided invaluable assistance during fieldwork. D. Scuderi, D. Fiorentino, A.L. Delos and S. Felline assisted with the sample processing and sorting and with species identification. Thanks to J. Claudet for valuable comments on regression analyses. Financial support was provided by Eni S.p.A., PEET (Partnerships for Enhancing Expertise in Taxonomy), and MURST (COFIN and FIRB projects). The authors acknowledge the support by the MARBEF Network of Excellence “Marine

References (71)

  • J.L. Gomez Gesteira et al.

    Taxonomic level for assessing oil spill effects on soft-bottom sublittoral benthic communities

    Mar. Pollut. Bull.

    (2003)
  • P.M.J. Herman et al.

    On the use of meiofauna in ecological monitoring: who needs taxonomy?

    Mar. Pollut. Bull.

    (1988)
  • N. Lampadariou et al.

    Cost/benefit analysis of a benthic monitoring programme of organic benthic enrichment using different sampling and analysis methods

    Mar. Pollut. Bull.

    (2005)
  • D. Maurer

    The dark side of the taxonomic sufficiency TS

    Mar. Pollut. Bull.

    (2000)
  • S. Pagola-Carte et al.

    Concordance degrees in macrozoobenthic monitoring programmes using different sampling methods and taxonomic resolution levels

    Mar. Pollut. Bull.

    (2002)
  • A.R.G. Price et al.

    Ocean-scale patterns of ‘biodiversity’ of Atlantic asteroids determined from taxonomic distinctness and other measures

    Biol. J. Linn. Soc.

    (1999)
  • P.A. Quijón et al.

    The use of coarser taxonomic resolution in studies of predation on marine sedimentary fauna

    J. Exp. Mar. Biol. Ecol.

    (2006)
  • A. Terlizzi et al.

    Effects of offshore platforms on soft-bottom macro-benthic assemblages: a case study in a Mediterranean gas field

    Mar. Pollut. Bull.

    (2008)
  • B.W. Thompson et al.

    Cost-efficient methods for marine pollution monitoring at Casey Station, east Antarctica: the choice of sieve mesh-size and taxonomic resolution

    Mar. Pollut. Bull.

    (2003)
  • M.A. Vanderklift et al.

    Use of assemblages derived from different taxonomic levels to select areas for conserving marine biodiversity

    Biol. Conserv.

    (1998)
  • R.M. Warwick

    The level of taxonomic discrimination required to detect pollution effects on marine benthic communities

    Mar. Pollut. Bull.

    (1988)
  • R.M. Warwick et al.

    All animals are equal, but some animals are more equal than others

    J. Exp. Mar. Biol. Ecol.

    (2008)
  • H. Akaike

    A new look at the statistical model identification

    IEEE T. Automat. Contr.

    (1974)
  • R.R. Alexander et al.

    Influence of sediment grain size on the burrowing of bivalves: correlation with distribution and stratigraphic persistence of selected neogene clams

    Palaios

    (1993)
  • M.J. Anderson

    A new method for non-parametric multivariate analysis of variance

    Austral. Ecol.

    (2001)
  • M.J. Anderson et al.

    PERMANOVA+ for PRIMER: Guide to Software and Statistical Methods

    (2008)
  • M.J. Anderson et al.

    Permutation tests for multi-factorial analysis of variance

    J. Stat. Comput. Sim.

    (2003)
  • J.J. Bell et al.

    Morphological monitoring of subtidal sponge assemblages

    Mar. Ecol. Prog. Ser.

    (2006)
  • Y. Bertrand et al.

    Taxonomic surrogacy in biodiversity assessments, and the meaning of Linnaean ranks

    Syst. Biodivers.

    (2006)
  • S. Bevilacqua et al.

    The use of taxonomic distinctness indices in assessing patterns of biodiversity in modular organisms

    Mar. Ecol.

    (2009)
  • D.F. Boesch et al.

    Response to stress in marine benthic communities

  • R.E. Bomkamp et al.

    Role of food subsidies and habitat structure in influencing benthic communities of shell mounds at sites of existing and former offshore oil platforms

    Mar. Biol.

    (2004)
  • M.G. Chapman

    Relationships between spatial patterns of benthic assemblages in a mangrove forest using different levels of taxonomic resolution

    Mar. Ecol. Prog. Ser.

    (1998)
  • K.R. Clarke et al.

    Changes in Marine Communities: An Approach to Statistical Analysis and Interpretation

    (1994)
  • O. Defeo et al.

    Testing taxonomic resolution levels for ecological monitoring in sandy beach macrobenthic communities

    Aquatic Conserv. Mar. Freshw. Ecosyst.

    (2004)
  • Cited by (53)

    • Impact of cage farming of cobia (Rachycentron canadum) on the benthic macrofauna in a tropical region

      2019, Aquaculture
      Citation Excerpt :

      The use of higher taxonomic levels in studies that use benthic organisms as indicators of impact in marine environments have been thoroughly investigated (Warwick, 1988; Olsgard et al., 1998; Olsgard and Somerfield, 2000). Even higher levels of species, such as family, order and even phylum, may reflect the effects of anthropogenic environmental disturbances (Bevilacqua et al., 2009). Thus, long and imprecise taxonomic identifications that cost time and money should be avoided (Olsgard and Somerfield, 2000).

    • The efficiency of Taxonomic Sufficiency for identification of spatial patterns at different scales in transitional waters

      2019, Marine Environmental Research
      Citation Excerpt :

      Consequently, transitional waters should be particularly suitable for TS, since they are highly fluctuating environments, with macrobenthic community showing high levels of functional redundancy (e.g. Mistri et al., 2001; Munari et al., 2005; Pitacco et al., 2018b). As already found elsewhere (e.g. Olsgard et al., 1998; Anderson et al., 2005; Włodarska-Kowalczuk and Kędra, 2007; Bevilacqua et al., 2009), this study showed that the choice of data transformation could influence the results of analyses as well as the level of taxonomic aggregation, or even more. For the analyses at local scale (differences among sites within lagoons), in most lagoons a reduction of similarity between matrices with different taxonomic levels was observed with increasing strength of data transformation, and the same was observed also at regional scale (differences among lagoons) for matrices at levels of order, class and phylum.

    View all citing articles on Scopus
    View full text