Distance-based Linear Models Research Articles

Spatial patterns of macrofaunal community structure in coastal lagoons of Uruguay

The spatio-temporal variability of macro benthic communities was investigated in the 5 main coastal lagoons of Uruguay, and faunal responses to environmental variables were evaluated. This is the first study that addresses spatial and temporal variability simultaneously for benthic subtidal communities of the coastal lagoons in Uruguay. Sampling was conducted in February, May, August, and November 2005, covering 3 spatial scales: lagoons, sites, and replicates. The lagoons were compared through environmental variables, using permutational multivariate analysis of variance and principal coordinates analysis. In addition, distance-based linear model analysis was performed to analyze which environmental variables best determined the macrobenthic community. The benthic community was characterized by a low number of species and high dominance. The most important difference in community structure occurred between Diario, a closed, eutrophic lagoon, and the other lagoons, which are intermittently connected with the ocean. Ocean−land gradients in community structure were observed within all lagoons. Greatest variability in community structure was observed at the largest spatial scale (among lagoons) and at the smallest scale (among replicates). Therefore, both largeand small-scale processes may shape the spatial patterns of these macrobenthic communities. The environmental variables that best explained the macrobenthic community structure were total suspended solids, conductivity, sediment organic matter, phytobenthic chlorophyll a, and macrophyte total biomass. Short-term temporal variability explained a small fraction of the total variation in comparison with spatial variation between lagoons.

Contrasting the species abundance, species density and diversity of seaweed assemblages in alternative states: Urchin density as a driver of biotic homogenization

Differences in seaweed assemblages' structure (species abundance, species density and diversity) were examined in two habitats, urchin barrens and upright seaweed beds on the Canarian Archipelago (eastern Atlantic Ocean) to demonstrate the key role of extreme density of the sea urchin Diadema africana in the homogenization of assemblages in shallow rocky reefs. Univariate and multivariate analyses were used to test for differences in seaweed the assemblages at multiple spatial scales, from sites (<10km apart) to islands (25–450km apart), based on samples collected from six islands. Distance-based linear model routine (DistLM) and distance-based redundancy analysis (dbRDA) were also applied to analyze and model relationships between seaweed assemblages and environmental variables in each habitat. The patterns of spatial variation in assemblage structure were different in urchin barrens compared to upright seaweed beds. In urchin barrens, spatial variation of seaweed assemblages differed between sites only, whereas in upright seaweed beds were observed differences between sites and islands. Sea urchin density and substrate roughness were the two factors determining assemblage structure in urchin barrens. In contrast, in upright seaweed beds, the major factors influencing assemblages were wave exposure, temperature and productivity. We conclude that potential biogeographic patterns of assemblage structure induced by oceanographic conditions were observed only in pristine areas with low urchin density.

Changes in Ecosystem Function Across Sedimentary Gradients in Estuaries

The input of terrestrial silt and clay (hereafter mud) into coastal environments can alter sediment grain size distribution affecting the structure and functioning of benthic communities. The relationship between sediment mud content and macrofaunal community structure has been well documented, but not the effects on ecosystem function. In 143 plots from the mid-intertidal sites in 9 estuaries, we measured sediment properties, macrofaunal community composition and fluxes of O2 and NH4+ across the sediment–water interface to derive process-based measures of ecosystem function across the sand–mud gradient. We observed reductions in measures of macrofaunal diversity and decreases in the maximum density of key bioturbating bivalves (Austrovenus stutchburyi and Macomona liliana) with increased mud content. Concurrently, the maximum rates of sediment oxygen consumption (SOC), NH4+ efflux (NH4+) and biomass standardized gross primary production (GPPChl-a) also decreased with increasing mud content. Environmental predictors explained 34–39% (P = 0.005–0.01) of the total variation in ecosystem function in distance-based linear models. After partitioning out the effect of mud, A. stutchburyi abundance was positively correlated and explained 25 and 23% (P = 0.0001) of the variation of SOC and NH4+, respectively. Also, mud content (negatively correlated) and temperature (positively correlated) explained 26% of variability in GPPChl-a (P = 0.0001). Our results highlight the importance of increased mud content and the associated reduction in the abundance of strongly interacting key species on the loss of ecosystem function in intertidal sand flats.

Oribatida (Acari) in grassy arable fallows are more affected by soil properties than habitat age and plant species

Oribatid mites are one of the numerically dominant arthropod groups in soils. They play an important role in soil food webs via regulating the decomposition of organic matter and propagating microorganisms within the soil. To our knowledge, the influence of different plant functional groups on oribatid mites has not been studied in abandoned farmland with undisturbed succession before. The density and assemblage structure of oribatid mites in nine grassy arable fallows relative to three habitat age classes (2–3, 6–8, 12–15 years) and three selected plant species (legume: Medicago sativa, forb: Taraxacum officinale, grass: Bromus sterilis) were investigated in soil associated with single plants.Mite density declined marginally not significant with habitat age because of high abundances of the ubiquitous species Tectocepheus velatus sarekensis and Punctoribates punctum in young and mid-aged fallows and their subsequent decline in old fallows. Oribatid mite density and species assemblage were not affected by plant species. Only P. punctum had significantly higher densities in B. sterilis samples than in T. officinale samples due to a higher amount of fine roots. Distance-based linear models revealed that 65% of the variation in mite assemblage was explained by soil properties, soil type, exposition and geographic position, while habitat age was of minor importance. Canonical correspondence analysis revealed that the mite assemblage was best explained by soil organic and microbial carbon, water content and pH.

Tools for bioindicator assessment in rivers: The importance of spatial scale, land use patterns and biotic integration

This study outlines an alternative, versatile and flexible procedure to the Assessment System for the Ecological Quality of Streams and Rivers throughout Europe using Benthic Macroinvertebrates (AQEM) protocol for selecting and assessing candidate bioindicators for Water Framework Directive (WFD) compliant monitoring programmes. Based on different forms of benthic macroinvertebrate data using relative abundance of family level taxonomic groups, metrics and traits collected during spring 2010 at 96 lotic sites across northern Portugal, the procedure employs components of top down and bottom up analytical processes and introduces the concept of niche breadth to assess biological quality element response to environmental and stressor parameters across different spatial levels. Random Forest classification revealed that fractal and non-fractal land use metrics at basin and local level were extremely important determinants of Water Framework Directive determination of “Good” ecological quality, particularly at the local scale. The amount of urbanization at the lower spatial level was a particularly important determinant of ecological quality, while the extent and type of forest (especially coniferous) was more important at higher, river basin scale. Distance-based linear models (DISTLM) and distance-based redundancy analysis (dbRDA) were used to determine associations between invertebrate data and non-redundant environmental predictors selected using the Akaike Information Criterion (AIC). Results revealed a stronger association between invertebrate traits and selected environmental predictors compared to the other types of invertebrate data, although some association between invertebrate relative abundance and eutrophication was detected. Principal Components Analyses (PCA) were run for the non-redundant sets of predictors for each macroinvertebrate data set to extract an environmental quality gradient along the first axis. Niche breadth, calculated for candidate indicators to avoid bias resulting from expert judgement, was distributed by rank along its respective PCA gradient. Five candidate indicators for each data type were selected for their preference for the most pristine sites and five were selected due to their close link with the most degraded streams. Candidate bioindicators for impacted sites tended to be stenobiotic in character, due to the impoverished structural and functional diversity associated with such conditions. Finally Partial Least Squares Regression was used to refine and validate selected candidate metrics, to produce a comprehensive final list of macroinvertebrate measures of ecological quality.

Influence of plantation of an exotic mangrove species, Sonneratia caseolaris (L.) Engl., on macrobenthic infaunal community in Futian Mangrove National Nature Reserve, China

Sonneratia caseolaris, a superior mangrove species characterized by rapid growth and proliferation, has been introduced to Futian Mangrove National Nature Reserve, Shenzhen, China for mangrove restoration since 1993. However, the effect of Sonneratia plantation on the macrobenthic infaunal community, a vital component of energy flow and nutrient recycling in mangrove ecosystems, remains obscure. The present study aimed to answer this question by determining the macrobenthic infaunal community, habitat characteristics and physico-chemical properties of sediment in the vegetated sites, dominated by different mangrove species (Kandelia obovata, Avicennia marina and S. caseolaris), and un-vegetated mudflat at four sampling times from August 2008 to September 2009. Results showed that the presence of Sonneratia can enhance the biodiversity of macrobenthic infauna in terms of Shannon–Weaver index (H′), compared to the mudflat, since August 2008 onwards. Multivariate analyses (CLUSTER and SIMPER) showed that the macrobenthic infaunal community in the Sonneratia site was different from all the other sites due to higher abundance of deposit and filter feeders, including Limnodriloides sp., Capitella capitata, Potamilla acuminata, Tharyx sp. and Ampharete arctica. The presence of Sonneratia, however, could not significantly alter the sediment properties. Spearman correlation analysis revealed that H′ was negatively correlated with cadmium concentration whereas distance-based linear model (DistLM) suggested that cadmium was a significant predictor variable correlating with the macrobenthic infaunal community. The higher cadmium concentration in August 2008 explained why H′ was lower and different macrobenthic infaunal communities were observed compared to the other sampling times. However, cadmium was the only significant predictor variable explaining 14.9% total variation, suggesting that the macrobenthic infaunal community was generally not affected by the sediment properties. Therefore, we concluded that Sonneratia plantation in the open mudflat can enhance the biodiversity of macrobenthic infauna in view of higher habitat complexity (e.g. presence of pneumatophores, below-ground vegetation and litter) on condition that the cadmium concentration was below the harmful level.

Controls of biological soil crust cover and composition shift with succession in sagebrush shrub-steppe

Successional stage may determine strength and causal direction of interactions among abiotic and biotic factors; e.g., species that facilitate the establishment of other species may later compete with them. We evaluated multivariate hypotheses about abiotic and biotic factors shaping biological soil crusts (BSCs) in early and late successional stages. We surveyed vegetation and BSC in the shrub-steppe ecosystem of the Columbia Basin. We analyzed the relationships with bryophyte and lichen covers using structural equation models, and analyzed the relationships with BSC composition using Indicator Species Analysis and distance-based linear models. Cover, indicator species, and composition varied with successional stage. Increasing elevation and bryophyte cover had higher lichen cover early in succession; these relationships were negative in the later successional stage. Lichen cover did not appear to impede B. tectorum cover, but B. tectorum appeared to strongly negatively affect lichen cover in both stages. Biological soil crust composition varied with bunchgrass cover in the early successional stage, but with elevation and B. tectorum cover later in succession. Our findings support the hypotheses that as succession progresses, the strength and direction of certain community interactions shift, and B. tectorum leads to reductions in biological soil crust cover regardless of successional stage.

Improving the “chain and tape” method: A combined topography index for marine fish ecology studies

The “chain and tape” method is used to quantify topographic complexity in reef ecology studies, consisting of the ratio of the linear distance between the start and end points of a chain moulded to the surface of the substrate to its stretched length, a measure known as the substrate rugosity (SR) index. This measure has several advantages in the field when compared to other methods, but some weaknesses have been pointed out. However, it is still one of the most frequently used topography measures in reef fish ecology. The present study proposes a combined topography index (CTI) that uses the “chain and tape” method in the field, with results that can match more complex methods, outperforming the traditional SR index. The CTI is structured as a weighted sum of 3 topographic features: SR index, number of corrugations (NC) and maximum vertical relief (MVR), where NC and MVR are given weight coefficients ranging from 0 to 1. In order to establish weight coefficients, fish assemblages were sampled at 6 training sites, representing a topographic complexity gradient. A series of candidate weight combinations were then selected so that CTI was optimally correlated with each one of several fish assemblage parameters. The list of possible solutions was then applied to nine previously published schematic profiles and compared with other methods. The final index was established as: CTI=(1−SR)+NC/25+MVR/25. Ultimately, the predictive performance of CTI and SR was tested by applying them to 11 new sites as predictors of species abundances in distance-based linear models. The CTI outperformed SR when added to 3 previously fitted variables (depth, percent cover of sand, percent cover of cobble), explaining 5.6% additional variation when using all species and 8.1% when using only cryptobenthic species, whereas SR showed no significant additional effects.

Short-term meso-scale variability of mesozooplankton communities in a coastal upwelling system (NW Spain)

The short-term, meso-scale variability of the mesozooplankton community present in the coastal upwelling system of the Ría de Vigo (NW Spain) has been analysed. Three well-defined communities were identified: coastal, frontal and oceanic, according to their holoplankton–meroplankton ratio, richness, and total abundance. These communities changed from summer to autumn due to a shift from downwelling to upwelling-favourable conditions coupled with taxa dependent changes in life strategies. Relationships between the resemblance matrix of mesozooplankton and the resemblance matrices of meteorologic, hydrographic and community-derived biotic variables were determined with distance-based linear models (DistLM, 18 variables), showing an increasing amount of explained variability of 6%, 16.1% and 54.5%, respectively. A simplified model revealed that the variability found in the resemblance matrix of mesozooplankton was mainly described by the holoplankton–meroplankton ratio, the total abundance, the influence of lunar cycles, the upwelling index and the richness; altogether accounting for 64% of the total variability. The largest variability of the mesozooplankton resemblance matrix (39.6%) is accounted by the holoplankton–meroplankton ratio, a simple index that describes appropriately the coastal–ocean gradient. The communities described herein kept their integrity in the studied upwelling and downwelling episodes in spite of the highly advective environment off the Ría de Vigo, presumably due to behavioural changes in the vertical position of the zooplankton.

A critical approach to the use of published data for baseline characterisation of marine fish assemblages: An exercise on Portuguese coastal waters

Policy-makers are becoming increasingly aware of the importance of coastal fish communities for the integrity of marine habitats. However, initial assessments are often done prior to any major funding and available published data is sometimes seen as a low-cost approach to this problem. The present work constituted an exercise on marine fish assemblage characterisation, by relying only on published data from mainland Portugal, covering soft substrate areas and near-shore rocky reefs. Regardless of the characterisation itself, this exercise aimed to critically approach these procedures, by pinpointing problems and highlighting achievements. Criteria were defined for the inclusion of published studies and included datasets were standardised in terms of species proportions, spatial and temporal scales. Species were assigned to ecological guilds and all analyses were performed separately for species and guild data. The influence of season, depth and latitude were analysed through multivariate analysis of variance using permutations and distance-based linear models and assemblages were characterised based on multivariate ordination and hierarchical classification methods. Significant differences were found between studies employing different methods in rocky reefs and outliers were not used in assemblage characterisation. Due to limitations in spatial and temporal detail in published data, only a rough, large scale characterisation of assemblage types was accomplished, with datasets divided in rocky reefs, shallow soft-substrate (0–20 m), intermediate soft-substrate (20–100 m) and deep soft substrate (100–200 m). No seasonal differences were detected, but a non-linear relationship with latitude was found in deep soft-substrate assemblages. During the exercise, spatial and temporal scales were forced by the quality of the data and some analyses failed to detect known patterns. Differences between methods, designs and types of data required assumptions to be made regarding data comparability. Overall, the exercise provided a hands-on insight into the use of published data for planning in a coastal management context, pinpointing data comparability, scale and information gaps as severe limitations of these approaches. In order to improve planning and design, some proposed measures were highlighted, such as the establishment of national databases and the use of pilot studies.

Individual variability in trophic position and diet of a marine omnivore is linked to kelp bed habitat

Omnivores play an important role in the routing and distribution of organic matter across food webs. We demonstrate a novel approach to quantifying the coincidence of landscape- scale nutritional gradients with niche breadth in terms of variability in trophic level and use of basal organic matter sources. We provide an example of the links between individual variability in resource use and habitat of a broad-spectrum omnivore, the red rock lobster Jasus edwardsii. Information on the co-occurrence of J. edwardsii with kelp bed habitats Ecklonia radiata and with their preferred prey Mytilus edulis galloprovinciallis were collected at 60 sites across Fiordland, southwest New Zealand. Analysis of distance-based linear models (DISTLM) indicated that the presence of mussels was the best predictor of lobster occurrence in the model set. At a subset of sites, we collected lobster muscle for stable isotope analysis and measured 3 demographic para- meters from the lobsters: relative abundance, sex and carapace lengths. We characterised habitats with surveys of common kelp and mussels. Using stable isotope signatures (δ 13 C and δ 15 N), we calculated individual-based estimates of trophic level and the mixture of organic matter sources, i.e. phytoplankton and macroalgae. Using DISTLM, lobster and mussel densities best explained variability in lobster diet. Variability in resource use was distinct inside and outside of kelp bed habitat. In kelp beds, lobsters fed at a higher average trophic level, with low variability among individuals in trophic level and use of organic matter sources. Outside kelp beds, individual vari- ability indicated broad trophic diversification. These patterns indicate a strong influence of the nutritional landscape at the scale of the metapopulation, which has important implications for understanding dietary influences on population structure.

Hypotheses of spatial stock structure in orange roughy Hoplostethus atlanticus inferred from diet, feeding, condition, and reproductive activity.

We evaluate hypotheses for meso-scale spatial structure in an orange roughy (Hoplostethus atlanticus) stock using samples collected during research trawl surveys off the east coast of New Zealand. Distance-based linear models and generalised additive models were used to identify the most significant biological, environmental, and temporal predictors of variability in diet, proportion of stomachs containing prey, standardised weight of prey, fish somatic weight, fish total weight, and reproductive activity. The diet was similar to that observed elsewhere, and varied with ontogeny, depth, and surface water temperature. Smaller sized and female orange roughy in warmer bottom water were most likely to contain food. Fish condition and reproductive activity were highest at distances more than 20 km from the summit of the hills. Trawl survey catches indicated greater orange roughy densities in hill strata, suggesting hill habitat was favoured. However, analyses of feeding, condition, and reproductive activity indicated hill fish were not superior, despite fish densities on hills being reduced by fishing which, in principle, should have reduced intra-specific competition for food and other resources. Hypotheses for this result include: (1) fish in relatively poor condition visit hills to feed and regain condition and then leave, or (2) commercial fishing has disturbed feeding aggregations and/or caused habitat damage, making fished hills less productive. Mature orange roughy were observed on both flat and hill habitat during periods outside of spawning, and if this spatial structure was persistent then a proportion of the total spawning stock biomass would remain unavailable to fisheries targeting hills. Orange roughy stock assessments informed only by data from hills may well be misleading.

Spatial and temporal variation in denitrification and in the denitrifier community in a boreal lake

We investigated the spatial and temporal variation in denitrification rates (isotopepairing technique, IPT) and in the denitrifier community (examination of gene nirK by denaturinggradient gel electrophoresis [DGGE] of microbial DNA) in the sediments of a boreal, clear-water, eutrophic lake using samples collected from shallow littoral, deep littoral and shallow profundal sediments during early summer, mid-summer, autumn and winter. The measured denitrification rates (44 to 613 μmol N m−2 d−1) are among the lowest ever reported from lacustrine sediments. Denitri fi cation rates varied both spatially and temporally, being highest in the profundal zone during midsummer and in the littoral zones during winter. Correlation analyses indicated that these variations were due to variations in the concentrations of nitrate and oxygen in the water overlying the sediment. The structure of the denitrifier community was temporally extremely stable and differed only slightly between the sites. Distance-based linear model (DISTLM) analysis indicated that the observed variation was probably due mainly to variations in the content of organic matter, and in the porosity, of the sediment. The structure of the denitrifier community and the denitrifying activities were uncoupled.

The influence of ontogeny and environment on the diet of lookdown dory, Cyttus traversi

Research on trophic relationships is an important component in the move towards an ecosystem approach to fisheries management. The diet of lookdown dory, Cyttus traversi, was determined from examination of stomach contents of 239 specimens of 13.9–50.7 cm total length (TL), sampled at depths of 220–744m on Chatham Rise, New Zealand. The diet was characterised by the macrourid fish, Coelorinchus oliverianus, the natant decapods, Lipkius holthuisi, Campylonotus rathbunae, Sergestes spp., Notopandalus magnoculus, and Pasiphaea spp., and the galatheid, Munida gracilis. Multivariate analyses using distance-based linear models, non-parametric multi-dimensional scaling and similarity percentages indicated the best predictors of diet variability were depth and fish length. Diet changes associated with depth were most apparent in the crustacean prey. Changes in the diet with ontogeny suggested four ontogenetic trophic units: pelagic (<13.9 cm TL), juveniles (13.9–25.8 cm TL; benthopelagic invertebrate feeders), maturing females and males (26–39.5 cm; benthopelagic omnivores) and mature females (>40 cm TL; benthopelagic piscivores/omnivores).

Diet and dietary variation of New Zealand hoki Macruronus novaezelandiae

The diet of hoki was determined from examination of stomach contents of 1992 fish of 26–112 cm total length (TL) sampled at depths of 209–904m on Chatham Rise, New Zealand, from summer research trawl surveys and seasonal commercial fishing trawls, during 2004–2008. Prey was predominantly euphausiids, mesopelagic fishes and natant decapods. Multivariate analyses using distance-based linear models, non-parametric multi-dimensional scaling and similarity percentages indicated that the best predictors of diet variability were the position of the fish in relation to the subtropical front (STF), fish size and longitude. Pasiphaeids were more important to the north of the STF, and sternoptychid fishes and euphausiids more important in the STF convergence area. Euphausiids and sternoptychid fishes were important for smaller hoki (26–55 cm TL), myctophid fishes and natant decapods for larger hoki, and macrourids for the largest hoki (>84 cm TL). The longitudinal effect was characterised by pasiphaeids, euphausiids and sternoptychids to the west, and myctophids in the centre of Chatham Rise. Feeding activity was analysed using generalised additive models, and was found to vary with time of day, sample source (research or commercial), longitude, hoki size and depth. The variability in diet suggested hoki forage opportunistically within their preferred habitat and biological limits.

Diet of Two Large Sympatric Teleosts, the Ling (Genypterus blacodes) and Hake (Merluccius australis)

Ling and hake are tertiary consumers, and as a result both may have an important structuring role in marine communities. The diets of 2064 ling and 913 hake from Chatham Rise, New Zealand, were determined from examination of stomach contents. Ling was a benthic generalist, and hake a demersal piscivore. The diet of ling was characterised by benthic crustaceans, mainly Munida gracilis and Metanephrops challengeri, and demersal fishes, mainly Macrourids and scavenged offal from fishing vessels. The diet of hake was characterised by teleost fishes, mainly macrourids and merlucciids. Multivariate analyses using distance-based linear models found the most important predictors of diet variability were depth, fish length, and vessel type (whether the sample was collected from a commercial or research vessel) for ling, and fish length and vessel type for hake. There was no interspecific predation between ling and hake, and resource competition was largely restricted to macrourid prey, although the dominant macrourid species predated by ling and hake were different. Cluster analysis of average diet of intraspecific groups of ling and hake confirmed the persistent diet separation. Although size is a central factor in determining ecological processes, similar sized ling and hake had distinctly different foraging ecology, and therefore could influence the ecosystem in different ways, and be unequally affected by ecosystem fluctuations.

Different food preferences in four sympatric deep-sea Macrourid fishes

Feeding habits of the four most abundant deep-sea demersal trawl-caught macrourids on Chatham Rise, New Zealand, were examined from stomach contents. Two species were predominantly benthic foragers: Coelorinchus bollonsi on infaunal and epifaunal polychaetes, and C. aspercephalus on epifaunal crustaceans; and two species were predominantly mesopelagic foragers; C. oliverianus on calanoid copepods, and Lepidorhynchus denticulatus on mesopelagic crustaceans. The most important predictors of diet variability were identified using distance-based linear models and included areal predictors in all four species, fish size in C. aspercephalus, C. bollonsi and L. denticulatus, and sample year in C. bollonsi. Cluster analyses showed that the diets of C. aspercephalus and C. bollonsi were most distinct. There was a greater interspecific similarity in diet in the spatial and fish size subgroups of C. oliverianus and L. denticulatus, than at the species level. Failing to account for areal, temporal, and ontogenetic variability in diets may bias evaluations of resource competition.

Biodiversity of epibenthic community in the inshore waters of southeast coast of India

AbstractThe epifaunal assemblage was evaluated at three different depths (5, 15 and 25 m) in the inshore waters of Pazhayar, Parangipettai and Cuddalore apart from 5 m depth near SIPCOT covering totally 10 stations (11°21′ N to 11°42′ N; 79°46′49″ E to 79°52′34″ E) in the southeast coast of India, Bay of Bengal. The occurrence of as many as 112 species belonging to 6 groups was recorded. Among these, gastropods constituted the largest component (42.85%) with 48 species. Bivalves came next with the percentage contribution of 24.10% with 27 species. Crustaceans and polychaetes contributed with 16.96% (19 species) and 6.25% (7 species), respectively. Others contributed with 9.82% (11 species). The maximum number of species was recorded in Cuddalore transect (65 species) and the minimum in SIPCOT (20 species). The maximum abundance was recorded in SIPCOT (1363 ind./haul) but the diversity was found to be lower here than in the other stations. The bivalve, Scapharca inaequivalvis was abundant only in this station whereas the large sized polychaetes, Epidiopatra hupferiana monroi and Sternaspis scutata were found only in the Pazhayar transect during all the seasons. Multivariate analyses were done to define assemblages. The total number of species estimated by various extrapolators varied from 112 species to 169.73 species. The cluster analysis revealed the gradual change in species composition with increase in depth. In the principal coordinate analysis (PCO) the first two axes explained 49.8% of the total variability in the case of biota and 98.3% of the total variability in the case of environmental variables. The distance based linear model (DISTLM) was used to find out the relationship between the abundance of epifauna and environmental variables. Depth explained about 23.7% of the variability while temperature and pH explained 17.5 and 10.6%, respectively. The best solution suggested all the nine environmental variables to combinedly explain about 92.49% of the total variability

Feeding habits and niche separation among the deep-sea chimaeroid fishes Harriotta raleighana, Hydrolagus bemisi and Hydrolagus novaezealandiae

The diets of Harriotta raleighana, Hydrolagus bemisi and Hydrolagus novaezealandiae were determined from examination of digestive tract contents from specimens sampled at depths of 219 to 876 m on the submarine ridge, Chatham Rise, New Zealand. All 3 species fed predominantly on benthic epifauna and infauna. Similarity percentages indicated the diet of H. raleighana was characterised by polychaetes and molluscs, H. bemisi by crabs and polychaetes, and H. novaezealandiae by crabs and sea urchins. Multivariate analyses using distance-based linear models found the most important predictors of diet variability were a categorical predictor for the subtropical front (STF) and bottom temperature. Cluster analysis of species subgroups indicated there was a greater similarity in diet between H. raleighana and H. bemisi, and between H. bemisi and H. novaezealandiae, in similar locations, than between subgroups of H. bemisi in different locations. Generalised additive models indicated that depth was the most important spatial predictor of research trawl survey catch rates, and depth separation was most pronounced between H. raleighana and H. novaezealandiae. Other significant predictors of diet variability included longitude, and fish length or weight. In all 3 species the diet of smaller fish included more polychaetes, small crustaceans and gastropods, and the diet of larger fish contained more decapod crustaceans. The primary factor determining niche separation among the 3 species appears to be depth and distribution rather than dietary specialisation.