The relative influence of local to regional drivers of variation in reef fishes

1 Running headline: spatial and temporal drivers of fishes The relative influence of local to regional drivers of variation in reef fishes F. TUYA*†, T. WERNBERG‡§ AND M. S. THOMSEN¥ *BIOGES, Marine Sciences Faculty, Universidad de Las Palmas de Gran Canaria, 35017 Las Palmas de G.C., Canary Islands, Spain, ‡School of Plant Biology, UWA Oceans Institute, University of Western Australia, Fairway, Crawley 6009 WA, Australia, §Australian Institute of Marine Science, Fairway, Crawley 6009 WA, Australia and ¥Department of Marine Ecology, National Environmental Research Institution, P.O. Box 358, DK-4000 Roskilde, Denmark †Author to whom correspondence should be addressed. Tel.: +34 928457456; fax: +34 928452900; email: [email protected] 2 In this study, fishes and habitat attributes were quantified, four times over 1 year, on three reefs within four regions encompassing a c. 6º latitudinal gradient across south-western Australia. The variability observed was partitioned at these spatio-temporal scales in relation to reef fish variables, and the influence of environmental drivers quantified at local scales, i.e. at the scale of reefs (the number of small and large topographic elements, the cover of kelp, fucalean and red algae, depth and wave exposure), and at the scale of regions (mean and maximum nutrient concentrations and mean sea water temperature) with regard to the total abundance, species density, species diversity and the multivariate structure of reef fishes. Variation in reef fish species density and diversity was significant at the regional scale, whereas variation in the total abundance and assemblage structure of fishes was also significant at local scales. Spatial variation was greater than temporal variation in all cases. A systematic and gradual species turnover in assemblage structure was observed between adjacent regions across the latitudinal gradient. The cover of red algae within larger patches of brown macroalgae (a biological attribute of the reef) and the number of large topographic elements (a structural attribute of the reef) were correlated with variation observed at local scales, while seawater temperature correlated with variation at the scale of regions. In conclusion, conservation efforts on reef fishes need to incorporate processes operating at regional scales with processes that shape local reef fish communities at local scales. Key words: spatial patterns; variability; reef fish assemblages; diversity; macroecology; Australia. 3 INTRODUCTION Identifying the scales of spatial and temporal variation relevant to patterns of distribution and abundance of species is a first step in the process of understanding the forces that govern community structure (Underwood et al., 2000). Advances in sampling techniques, coupled with the refinement and computational simplicity of statistical routines, are increasingly enabling the analysis of spatio-temporal patterns, from small to large scales, and have enhanced the ability to explore a broad range of natural phenomena with increasing sophistication (Fraschetti et al., 2005). Reef-associated fish assemblages respond to changes in environmental conditions with fluctuations in abundance at different spatial and temporal scales (Connell & Kingsford, 1998; Anderson & Millar, 2004; García-Charton et al., 2004; Malcolm et al., 2007). Understanding the processes underlying such variation is a central aim of fish ecologists, and requires identifying the patterns of distribution and abundance of fishes over ecologically and biogeographically relevant spatial and temporal scales. Few studies (e.g. Anderson & Millar, 2004; Malcolm et al., 2007) however, have examined spatial variability in fish communities over scales > 100 km using hierarchically nested designs which allow a complete partitioning of variation. Indeed, the extent to which the amount of small-scale variability (from metres to > 10 km but < 100 km) compares to large-scale variation (> 100 km) affecting patterns in reef fish abundance, diversity and assemblage structure remains largely unexplored. Even less understood is the extent to which any spatial pattern remains consistent through time. For example, seasonal patterns of fish recruitment can change between adjacent islands within an archipelago as a result of oceanographic differences (Tuya et al., 2006). At large spatial scales (> 100 km), variability in mesoscale oceanographic conditions can affect reef fishes (Leathwick et al., 2006; Wellenreuther et al., 2008), for example via control over larval dispersal and recruitment patterns of fishes that have pelagic larvae that 4 disperse over long distances (Kinlan et al., 2005). Indeed, patterns of abundance of reef fishes can follow different biogeographical models across their distribution ranges, e.g. ‛normal’ v. ‛ramped’ patterns (Tuya et al., 2008). At small spatial scales (from metres to >10 <100 km), on the other hand, the physical structure of the reef plays a key role in the organization of fish assemblages, affecting protection from predators and accessibility to food (e.g. Jones & Syms, 1998; Tuya et al., 2009), as well as local recruitment patterns (Tolimieri, 1995). In fact, the relationship between reef habitat and fish assemblages is becoming a major tool for the sustainable management of fisheries and marine park planning (Anderson & Millar, 2004; García-Charton et al., 2004). Studies examining the distribution and abundance of fishes in relation to the habitat structure are particularly common from tropical coral reefs (e.g. Luckhurst & Luckhurst, 1978; Newman & Williams, 1996; Friedlander & Parrish, 1998; Jones & Syms, 1998), but have been also conducted more recently from temperate reefs (e.g. Willis & Anderson, 2003; García-Charton et al., 2004; Ordines et al., 2005; Tuya et al., 2009). Local scale reef habitat attributes that can determine the distribution and abundance of reef fishes include: substratum type (e.g. Guidetti et al., 2004), vertical relief (e.g. Luckhurst & Luckhurst, 1978; Yoklavich et al., 2000) and the cover of different types of algal assemblages (e.g. Levin, 1994; Tuya et al., 2009). On temperate reefs, one of the main habitats is provided by laminarian and fucalean canopy-forming algae, with smaller foliose algae interspersed in the gaps (Connell & Irving, 2008). Fish assemblages associated with algal habitats are affected by the physical structure of the algal canopies (Ebeling & Laur, 1985; Anderson & Millar, 2004), which influence environmental conditions and the intensity of biological processes (e.g., predation pressure, Levin, 1994). Western Australia is one of 18 major centres of endemism of the world’s reefs (Roberts et al., 2002). The south-west region (Fig. 1) is a transition zone between the Damperian (tropical) and Flindersian (temperate) biogeographical provinces. The high species 5 diversity and endemism of demersal fishes (Williams et al., 2001) is largely attributed to a long period of isolation from other continents (c. 80 million years), the moderating influence of the warm Leeuwin Current over the past c. 50 million years, and the lack of mass extinctions associated with unfavorable conditions, such as glaciations, over the recent geological past. Rocky reefs are a predominant feature along more than 1600 km of coastline, where they run almost continuously from Shark Bay to Cape Leeuwin. On most of these reefs, the dominant alga is the canopy-forming kelp Ecklonia radiata that create habitats extending from metres to > 100 km, interspersed with frondose fucalean algae and smaller foliose red algae (Wernberg et al., 2003). South-western Australia, as a result of its latitudinal gradient in temperature (Smale & Wernberg, 2009; Wernberg et al., 2010), is an ideal candidate to disentangle the effect of different scales of spatial and temporal variation over reef fishes. In this study, a multi-scale sampling design was used to partition the variability observed at several spatial and temporal scales over reef fishes. This variability was correlated with a variety of environmental drivers operating at different scales. Specifically, it was assessed how patterns in the abundance, species density, species diversity and multivariate assemblage structure of reef fishes varies on multiple spatio-temporal scales, and quantified the possible influence of various environmental drivers was quantified. MATERIALS AND METHODS STUDY AREA AND SAMPLING DESIGN This study included four regions across south-western Australia: Hamelin Bay (HAM), Marmion (MAR), Jurien Bay (JUR) and Kalbarri (KAL) (Fig. 1). These regions are evenly spaced across c. 6º of latitude (c. 800 km, adjacent regions are c. 150-200 km apart), encompassing a temperature gradient of 3-4 ºC (Smale & Wernberg, 2009; Wernberg et al., 2010); each region encompasses a stretch of coast between 10 and 20 km. Within each region, 6 shallow-water reef fishes were surveyed at three sites (i.e. reefs, 1-20 km apart, 8-12 m depth). Sampling was repeated four times over the course of one year (October 2006, March, June, and October 2007); sampling times were randomly selected, with successive times separated by between 3 and 5 months. All reefs were outside areas under any protection (i.e. no-take areas). At each site and time, three replicate, randomly oriented, 25 x 5 m belt transects were sampled by the same SCUBA diver (F. Tuya), identifying and counting all adult and sub-adult fishes (moving bits) on the way out (Tuya et al., 2008), and the number of large (>1 m) and small (<1 m) topographic elements (hereafter large and small ‛drops’) of the rocky substrata (i.e. cracks, crevices, caves, holes per 125 m2) when rolling up the transect (non-moving bits). Another diver (T. Wernberg) visually estimated the percent cover of kelp, fucalean algae, and red algae, following standardized procedures for the study region (Wernberg et al., 2008). These environmental drivers were considered to operate at the scale of reefs. For each reef the following were recorded: the depth (using a SCUBA depth gauge), wave exposure (rank from 1 to 12; qualitatively based on >100 dives on each reef under various hydrodynamic and climatic conditions over the last 10 years), the mean and maximum nutrient concentrations [NOx analyses of water samples collected c. 1 m above the bottom, see Wernberg et al. (2010) for details], and the mean sea water temperature [Tidbit loggers, logging c. 5 cm above the reef, for details, see Smale & Wernberg (2009)]. Mean and maximum nutrient concentrations and mean sea water temperature do not vary significantly among reefs within regions, but among regions along the latitudinal gradient (Wernberg et al., 2010); therefore, these environmental drivers were consider to operate at the scale of regions. DATA ANALYSES Univariate attributes of the reef fish assemblage included fish abundance, species 7 density (number of species per area, sensu Gotelli & Colwell, 2001) and species diversity (Shannon-Wiener index). These univariate attributes were expressed per transect (125 m2). Permutation-based ANOVA was used to partition variation in reef fishes into spatial and temporal components. The ANOVA model incorporated, in all cases, the random factors: ‘Regions’, ‘Times’ (orthogonal to ‘Regions’), and ‘Sites’ (nested within ‘Regions’). ANOVAs were performed on the species density, diversity and total abundance of fishes, as well as abundance of the 20 most common species. Before analysis, Cochran’s test was used to test for homogeneity of variances. The α value was set at the most conservative 0.01 level, instead of the traditional 0.05 value, to reduce the probability of type I errors when variances remained heterogeneous despite transformation (Underwood, 1997). A partitioning of multivariate variability (all species with their corresponding abundances) was carried out, using the same model as outlined before, via a permutational multivariate ANOVA (Anderson 2001). Pair-wise comparisons between each pair of regions were executed when significant differences were detected at the regional scale. The contribution of each scale to explain the total variation in univariate and multivariate fish responses was estimated by calculating their variance components and expressing the outcomes as percentages of total explained variation (Graham & Edwards, 2001; Anderson & Millar, 2004). Multiple regression models, using the DISTLM routine (Anderson, 2001), tested whether variation in the environmental drivers operating at the scale of reefs (the number of small and large topographic elements and the cover of the three types of algal canopies, depth and wave exposure) and regions (mean and maximum nutrient concentrations and the mean seawater temperature) affected variation in the total abundance, species density, species diversity, and assemblage structure of reef fishes across the study area (times were pooled for each site) by fitting a linear model. To retain variables (i.e. environmental drivers in each case) with good explanatory power, the AIC routine was used as a selection criterion [the 8 smaller the value the better the model, Anderson & Legendre (1999)]. A Canonical Analysis of Principal coordinates (CAP, Anderson & Willis, 2003) was used, as a constrained ordination procedure, to visualize differences in the assemblage structure of reef fishes. Two separate canonical analyses were done: one to visualize differences among regions (pooling sampling times for each region), and the other to visualize differences among times (pooling regions for each time). Distance-based redundancy analysis (db-RDA, Legendre & Anderson, 1999) was used to visualize whether variation in the environmental drivers of each reef affected variation in the assemblage structure of reef fishes across the study area (times were pooled for each site). In all analysis, fish abundances were Log(x+1) transformed to down-weigh the influence of abundant small schooling fishes over larger, more solitary, fishes. Trends in serial correlation (seriation) in assemblage structure among regions were tested via the RELATE routine. All multivariate analyses were based on Bray-Curtis similarities. P values were calculated from 4999 permutations of the residuals under the reduced model. All data analyses were done on PRIMER & PERMANOVA 1.0.1 (Anderson et al., 2008). RESULTS PATTERNS IN TOTAL ABUNDANCE, SPECIES DENSITY AND DIVERSITY OF REEF FISHES A total of 6111 reef fishes, belonging to 47 taxa, were observed during the study. Twenty species dominated the assemblage in terms of abundance (c. 85% of total individuals), and were present in > 20% of the censuses (Fig. 2). Total abundance, species density, and species diversity of reef fishes were dominated by spatial rather than temporal variation (Table I, Fig. 3). Total abundance of individuals [Figs. 3(a) and 4(a)] changed from site to site within each region inconsistently among times [(Fig. 3(a), ‛Sites (Region) x Time’, ANOVA, 9 P = 0.024, Table I]. Indeed, the lowest levels of variation [‛Sites (Region) x Time’ and the residual term] accounted for almost 80% of the total variation. Significant variation in species density was found among regions [Figs. 3(b) and 4(b), ANOVA, P = 0.004, c. 23% of total explained variation, Table I], despite significant differences among sites within regions [Fig. 3(b), ANOVA, P = 0.024, 13.8% of total explained variation, Table I]. Species diversity only differed significantly among regions [Figs. 3(c) and 4(c), ANOVA, P = 0.011, 21.1% of total explained variation, Table I]. Pair-wise comparisons indicated that Kalbarri had lowest species density and diversity, but not total abundance, of reef fishes than the other three regions [Figs. 4(b) and 4(c), respectively]. Sampling time did not have any effects over these three attributes (total abundance, species density, and species diversity of reef fishes) on its own (Fig. 3, ANOVA, d.f. = 3, P > 0.07 in all cases, Table I), and generally accounted for a small proportion of variation (< 12.3%). In all cases, the residual (i.e. variation among replicate transects) accounted for a large amount of variation (30.6% - 51.5%, Table I). Patterns in the abundance of individual fish species (Fig. 5) were also dominated by spatial rather than temporal variation, i.e. only one (the western talma Chelmonops curiosus Kuiter 1986) of the 20 species showed a significant difference among sampling times. Twelve species varied significantly in abundances among regions across the latitudinal gradient (Fig. 5); ‛regions’ accounted for 10-40 % of total variation for these species. Again, the residual term accounted for the largest amount of variation (Fig. 5). PATTERNS IN THE ASSEMBLAGE STRUCTURE OF REEF FISHES Variation in reef fish assemblage structure was complex with almost 80% of variation attributable to combinations of temporal and spatial components (Table I). Despite small-scale spatial variability [i.e. the residual plus ‛Sites (Region) x Time’ and ‛Sites’, Fig. 3(d), Table I] which accounted for 66.1% of total variation in assemblage structure, spatial variation at the 16 large topographic elements, while variation at regional scales was driven by seawater temperature. As a result, conservation efforts on reef fishes need to incorporate large-scale processes operating at regional scales in conjunction with knowledge of processes that shape local reef fish assemblages at small scales. The resurgence of ‘macroecology’, that unifies biogeography and ecology, provides a sound conceptual framework to connect the effects of different scales on ecological patterns as a way to optimize the sustainable management of coastal fisheries and marine park planning. This study was funded by an ARC Discovery grant to TW. FT was supported by a postdoctoral research fellowship within the Spanish MCINN framework 'Plan de Formación y Perfeccionamiento' and the “Ramón y Cajal” programme during writing. References Anderson, M.J. (2001). Permutation tests for univariate and multivariate analysis of variance and regression. Canadian Journal of Fisheries and Aquatic Sciences 58, 626-639. Anderson, M.J. & Willis, T.J. (2003). Canonical analysis of principal coordinates: a useful method of constrained ordination for ecology. Ecology 84, 511 – 525. Anderson, M.J. & Millar R.B. (2004). Spatial variation and effects of habitat on temperate reef fish assemblages in northeastern New Zealand. Journal of Experimental Marine Biology and Ecology 305, 191-221. Anderson, M.J., Gorley, R.N. & Clarke, K.R. (2008). PERMANOVA+ for PRIMER: Guide to software and statistical methods. Plymouth, U.K.: PRIMER-E Ltd. 17 Ayvazian, S.G. & Hyndes, G.A. (1995). Surf-zone fish assemblages in south-western Australia: do adjacent nearshore habitats and the warm Leeuwin current influence the characteristics of the fish fauna? Marine Biology 122, 527-536. Bouchon-Navaro, Y., Bouchon, C., Louis, M. & Legendre, P. (2005). Biogeographic patterns of coastal fish assemblages in the West Indies. Journal of Experimental Marine Biology and Ecology 315, 31–47. Brown, J.H. (1995). Macroecology. Chicago: University of Chicago Press. Connell, S.D. & Jones, G.P. (1991). The influence of habitat complexity on postrecruitment processes in a temperate reef fish population. Journal of Experimental Marine Biology and Ecology 151, 271-294. Connell, S.D. & Kingsford, M.J. (1998). Spatial, temporal and habitat-related variation in the abundance of large predatory fish at One Tree Reef, Australia. Coral Reefs 17, 49-57. Connell, S.D. & Irving, A. (2008). Integrating ecology with biogeography using landscape characteristics: a case study of subtidal habitat across continental Australia. Journal of Biogeography 35, 1608-1621. Depczynski, M. & Bellwood, D.R. (2005). Wave energy and spatial variability in community structure of small cryptic coral reef fishes. Marine Ecology Progress Series 303, 283–293. Ebeling, A.W. & Laur, D.R. (1985). The influence of plant cover on surfperch abundance at 18 an offshore temperate reef. Environmental Biology of Fishes 12, 169-179. Edgar, G.H. (2008). Australian Marine Life. Sydney: New Holland. Fraschetti, S., Terlizzi, A. & Benedetti-Cecchi, L. (2005). Patterns of distribution of marine assemblages from rocky shores: evidence of relevant scales of variation. Marine Ecology Progress Series 296, 13-29. Friedlander, A.M. & Parrish, J.D. (1998). Temporal dynamics of fish communities on an exposed shoreline in Hawaii. Environmental Biology of Fishes 53, 1–18 García-Charton, J.A. & Pérez-Ruzafa, A. (1998). Correlation between habitat structure and rocky reef fish assemblages in SW Mediterranean. PSZNI Marine Ecology 19, 111–128. García-Charton, J.A., Pérez-Ruzafa, A., Sanchez-Jerez, P., Bayle-Sempere, J.T., Renones, O. & Moreno, D. (2004). Multi-scale spatial heterogeneity, habitat structure, and the effect of marine reserves on western Mediterranean rocky reef fish assemblages. Marine Biology 144, 161–182. Gotelli, N.J. & Colwell, R.K. (2001). Quantifying biodiversity: procedures and pitfalls in the measurement and comparison of species richness. Ecology Letters 4, 379-391. Graham, M.H. & Edwards, M.S. (2001). Statistical significance versus fit: estimating the importance of individual factors in ecological analysis of variance. Oikos 93, 505513. 19 Guidetti, P., Bianchi, C.N., Chiantore, M., Schiaparelli, S., Morri, C. & Cattaneo-Vietti, R. (2004). Living on the rocks: substrate mineralogy and the structure of subtidal rocky substrate communities in the Mediterranean Sea. Marine Ecology Progress Series 274, 57–68. Jones, G.P. & Syms, C. (1998). Distribution, habitat structure and the ecology of fishes on coral reefs. Australian Journal of Ecology 23, 286–297. Kinlan, B.P., Gaines, S.D. & Lester, S.E. (2005). Propagule dispersal and the scales of marine community process. Diversity and Distributions 11, 139-148. Leathwick, J.R., Elith, J., Francis, M.P., Hastie, T. & Taylor, P. (2006). Variation in demersal fish species richness in the oceans surrounding New Zealand: an analysis using boosted regression trees. Marine Ecology Progress Series 321, 267–281. Legendre, P. & Anderson, M.J. (1999). Distance-based redundancy analysis: testing multispecies responses in multifactorial ecological experiments. Ecological Monographs 69, 1–24. Levin, P.S. (1994). Small-scale variation in a temperate fish: the roles of macrophytes and food supply. Environmental Biology of Fishes 26, 29-37. Luckhurst, B.E. & Luckhurst, K. (1978). Analysis of the influence of substrate variables on coral reef fish communities. Marine Biology 49, 317–323. Malcolm, H.A., Gladstone, W., Lindfield, S., Wraith, J. & Lynch, T.P. (2007). Spatial and 20 temporal variation in reef fish assemblages of marine parks in New South Wales, Australia— baited video observations. Marine Ecology Progress Series 350, 277–290. Newman, S.J. & Williams, D.M. (1996). Spatial and temporal variation in assemblages of Lutjanidae, Lethrinidae and associated fish species among mid-continental shelf reefs in the central Great Barrier Reef. Marine and Freshwater Research 52, 843 – 851. Ordines, F., Moranta, J., Palmer, M., Lerycke, A., Suau, A., Morales-Nin, B. & Grau, A.M. (2005). Variations in a shallow rocky reef fish community at different spatial scales in the western Mediterranean Sea. Marine Ecology Progress Series 304, 221-233. Roberts, C.M., McClean, C.J., Veron, J.E.N., Hawkins, J.P., Allen, G.R., McAllister, D.E., Mittermeier, C.G., Schueler, F.W., Spalding, M., Wells, F., Vynne, C. & Werner, T.B. (2002). Marine biodiversity hotspots and conservation priorities for tropical reefs. Science 295, 12801284. Smale, D.A. & Wernberg, T. (2009). Satellite-derived SST data as a proxy for water temperature in nearshore benthic ecology. Marine Ecology Progress Series 387, 27-37. Taylor, M.S., Hellberg, M.E. (2003). Genetic evidence for local recruitment of pelagic larvae in a Caribbean reef fish. Science 299, 635-645. Thompson, A.A. & Mapstone, B.D. (1997). Observer effects and training in underwater visual surveys of reef fishes. Marine Ecology Progress Series 154, 53– 63. 21 Tolimieri, N. (1995). Effects of microhabitat characteristics on the settlement and recruitment of a coral reef fish at two spatial scales. Oecologia 102, 52–63. Tuya, F., Boyra, A., Sanchez-Jerez, P., Barbera, C. & Haroun, R.J. (2004). Relationships among fishes, the long-spined sea urchin Diadema antillarum and algae throughout the Canarian Archipelago. Marine Ecology Progress Series 278, 157-169. Tuya, F., Ortega-Borges, L., Sanchez-Jerez, P. & Haroun, R.J. (2006). Effect of fishing pressure on the spatio-temporal variability of the parrotfish, Sparisoma cretense (Pisces: Scaridae), across the Canarian Archipelago (eastern Atlantic). Fisheries Research 77, 24-33. Tuya, F., Wernberg, T. & Thomsen, M.S. (2008). Testing the 'abundant centre' hypothesis on endemic reef fishes in south-western Australia. Marine Ecology Progress Series 372, 225230. Tuya, F., Wernberg, T. & Thomsen, M.S. (2009). Habitat structure affect abundances of labrid fishes across temperate reefs in south-western Australia. Environmental Biology of Fishes 86, 311-319. Tuya, F., Vanderklift, M.A., Hyndes, G.A., Wernberg, T., Thomsen, M.S., Hanson, C. (2010). Proximity to reefs affects the balance between positive and negative effects on seagrass fauna. Marine Ecology Progress Series 405, 175-186. Underwood, A.J. (1997). Experiments in Ecology: their Logical Design and Interpretation Using Analysis of Variance. Cambridge: Cambridge University Press. 22 Underwood, A.J., Chapman M.G. & Connell, S.D. (2000). Observations in ecology: you can't make progress on processes without understanding the patterns. Journal of Experimental Marine Biology and Ecology 250, 97-115. Wellenreuther, M. & Connell, S.D. (2002). Response of predators to prey abundance: separating the effects of prey density and patch size. Journal of Experimental Marine Biology and Ecology 273, 61– 71. Wellenreuther, M., Syms, C. & Clements, K.D. (2008). Consistent spatial patterns across biogeographic gradients in temperate reef fishes. Ecography 31, 84-94. Wernberg, T., Kendrick, G.A. & Phillips, J.C. (2003). Regional differences in kelp-associated algal assemblages on temperate limestone reefs in south-western Australia. Diversity and Distributions 9, 427-441. Wernberg, T., Thomsen, M.S., Tuya, F., Kendrick, G.A., Staehr, P.A. & Toohey, B.D. (2010). Decreasing resilience of kelp beds along a latitudinal temperature gradient: potential implications for a warmer future. Ecology Letters 13, 685-694. Wernberg, T., Thomsen, M.S., Tuya, F. & Kendrick, G.A. (2011, in press). Biogenic habitat structure of seaweeds change along a latitudinal gradient in ocean temperature. Journal of Experimental Marine Biology and Ecology. Williams, A., Koslow, A.J., Last, P.R. (2001). Diversity, density and community structure of 23 the demersal fish fauna of the continental slope off Western Australia (20 to 35ºS). Marine Ecology Progress Series 212, 247-263. Willis, T.J. & Anderson, M.J. (2003). Structure of cryptic reef fish assemblages: relationships with habitat characteristics and predator density. Marine Ecology Progress Series 257, 209221. Willis, T.J., Badalamenti, F. & Milazzo, M. (2006). Diel variability in counts of reef fishes and its implications for monitoring. Journal of Experimental Marine Biology and Ecology 33, 108– 120. Yoklavich, M.M., Greene, H.G., Cailliet, G.M., Sullivan, D.E., Lea, R.N. & Love, M.S. (2000). Habitat associations of deep-water rockfishes in a submarine canyon: An example of a natural refuge. Fishery Bulletin 98, 625-641.