Testing the suitability of a morphological monitoring approach for identifying temporal variability in a temperate sponge assemblage

Abstract

Sponges are a dominant component of benthic assemblages in hard substratum environments across the world, but despite the importance of sponges, they are generally poorly represented in most marine monitoring programmes. There is considerable need to develop effective monitoring tools to monitor changes in sponge assemblages. Morphological monitoring has been proposed as a suitable method to monitor sponges and while morphological monitoring has already taken place at Skomer Marine Reserve (MNR), Wales, here we investigate whether species level and morphological level data sets are correlated with respect to temporal variation. Furthermore, we examine the environmental factors that correlate with the patterns of temporal variability. Both species and morphological data sets revealed significant seasonal changes and spatial variation in sponge assemblages; these data sets were highly correlated and explained by a number of environmental factors. We conclude that morphological monitoring of sponge assemblages may represents a cost-effective method for assessing temporal and spatial variation in sponges, where full species level monitoring is not possible, as patterns identified from morphological data were a suitable surrogate of species-level data.

Introduction

Sponges are a dominant component of hard substratum benthic marine communities (Alvarez et al., 1990, Bell and Barnes, 2000, Bell and Smith, 2004, Dayton, 1978, Diaz et al., 1993, Hooper et al., 2002, Leys et al., 2004, Økland and Økland, 1996). However, sponges are generally poorly represented in most large-scale marine monitoring programmes across the world because they are considered a difficult group to work with because: they are taxonomically challenging and are hard to identify easily in the field (Rützler, 1978, Wulff, 2001); and because they are difficult to quantify, since sponge abundance can be represented by the number of patches present, volume or area occupied (e.g. see Wulff 2001). For these reasons, there is considerable interest in developing rapid assessment tools for monitoring sponges (see Bell et al. 2006).

Due to the difficulties associated with including sponges in monitoring programmes, morphological monitoring has been proposed as a surrogate for identifying patterns of sponge biodiversity (Bell, 2007a, Bell, 2007b, Bell and Barnes, 2001, Bell and Barnes, 2002, Bell et al., 2006). A morphological approach focuses on counting the number of different morphologies present, rather than measuring the abundance of different species. Such surrogate approaches can be a time/cost effective method to estimate the patterns of biodiversity within an area instead of collecting species level data. While the collection of morphological data almost certainly results in lower resolution data than collecting species-level data it offers the potential of including sponges in monitoring programmes where otherwise they might be ignored. While sponge morphological monitoring has been successfully applied to a number of locations, and is able to consistently identify spatial patterns identified for species-level data sets (Bell and Barnes, 2001, Bell and Barnes, 2002), there has still been no formal assessment of whether temporal patterns in morphological diversity correlate with temporal variation in species level data.

The majority of studies examining variability in sponge assemblages have examined spatial variation, usually focusing on only a single geographic location at a single sampling interval (e.g. Barnes and Bell, 2002, Berman and Bell, 2010, Cleary et al., 2005, Duckworth et al., 2008, Janussen and Tendal, 2007, Picton and Goodwin, 2007) or a specific genus (Duckworth and Battershill, 2001, Klautau and Valentine, 2003), family (Plotkin 2004) or functional role (Lemloh et al. 2009). Temporal variation has mostly been explored at the species level (for some examples see De Caralt et al., 2008, Garrabou and Zabala, 2001, Koopmans and Wijffels, 2008, Mcmurray et al., 2008), with fewer studies examining temporal variation in entire sponge assemblages over more than a year (but see Bell et al., 2006, Carballo et al., 2008, Pansini and Pronzato, 1990, Roberts et al., 2006, Sara, 1970, Wulff, 2006). Furthermore, the results from these studies of temporal variation are somewhat variable. For example, research in the Mediterranean has shown slow growth within sponge assemblages with no clear seasonality, with the abundance of different sponge species remaining similar throughout an eight year study (Sara, 1970, Pansini and Pronzato, 1990). Similar results showing little overall change have also been reported from sites in the Atlantic (Fowler and Laffoley, 1993, Hiscock, 1994) and Caribbean (Hughes 1996). In fact, even though the overall abundance of sponge individuals and percentage cover remained relatively constant at Caribbean sites, the actual individual sponges were very variable suggesting recruitment and mortality were in equilibrium (Hughes 1996). However, more recent work in the Atlantic, Caribbean and Pacific suggests that more rapid changes in sponge assemblages are common and recruitment and mortality do not always occur in equilibrium (at least over the temporal scales examined). Therefore overall sponge assemblage composition can show dramatic changes over relatively short temporal scales (Bell et al., 2006, Carballo et al., 2008, Wulff, 2006).

The sponge assemblages at Skomer Marine National Reserve (MNR; Pembrokeshire, Wales, UK) have been morphologically monitored over a ten year period prior to this study and although changes have been identified between some years in the morphologies present and their abundance, there was evidence the sponge assemblages could recover rapidly within a year to their original assemblage composition (Bell et al. 2006). In this previous work, no strong correlation was found between any of the measured environmental variables and the inter-annual changes in the sponge assemblage, therefore changes were attributed to natural variation or short-term impacts that could not be identified (Bell et al. 2006). Although this previous study highlighted the potential for morphological monitoring of sponges, there were some limitations. Firstly, there was a lack of independence and replication of the sampling units (as the study was part of an existing programme), therefore it was not possible to draw statistically supported conclusions about changes in sponge abundance or specific morphological types over time. Secondly, the monitoring was annual and therefore seasonal changes were not quantified. Finally, unlike for spatial variation, there has been no formal assessment of whether temporal variability in sponge morphologies correlates with species level data.

This study describes the re-establishment of the sponge-monitoring programme at Skomer Island MNR to address these earlier criticisms and further examines the potential for sponge morphological data to be used as a surrogate for monitoring temporal changes in sponge assemblages. The aims of this study were to: (1) measure the patterns of seasonal and inter-annual variability in the sponge assemblages based on morphological and species level data; (2) determine if the patterns of temporal variation in sponge species and morphological assemblage data are correlated; and (3) identify which environmental factors best explain the temporal variation in these sponge assemblages based on both morphological and species-level data.