USING the SURFACE-FEATURES CREATED by BIOTURBATING ORGANISMS AS SURROGATES for MACROFAUNAL DIVERSITY and COMMUNITY STRUCTURE S Widdicombe, M Kendall, D Parry

USING THE SURFACE-FEATURES CREATED BY BIOTURBATING ORGANISMS AS SURROGATES FOR MACROFAUNAL DIVERSITY AND COMMUNITY STRUCTURE S Widdicombe, M Kendall, D Parry

To cite this version:

S Widdicombe, M Kendall, D Parry. USING THE SURFACE-FEATURES CREATED BY BIOTUR- BATING ORGANISMS AS SURROGATES FOR MACROFAUNAL DIVERSITY AND COMMU- NITY STRUCTURE. Vie et Milieu / Life & Environment, Observatoire Océanologique - Laboratoire Arago, 2003, pp.179-186. hal-03205257

HAL Id: hal-03205257 https://hal.sorbonne-universite.fr/hal-03205257 Submitted on 22 Apr 2021

HAL is a multi-disciplinary open access L’archive ouverte pluridisciplinaire HAL, est archive for the deposit and dissemination of sci- destinée au dépôt et à la diffusion de documents entific research documents, whether they are pub- scientifiques de niveau recherche, publiés ou non, lished or not. The documents may come from émanant des établissements d’enseignement et de teaching and research institutions in France or recherche français ou étrangers, des laboratoires abroad, or from public or private research centers. publics ou privés. VIE MILIEU, 2003, 53 (4) : 179-186

USING THE SURFACE-FEATURES CREATED BY BIOTURBATING ORGANISMS AS SURROGATES FOR MACROFAUNAL DIVERSITY AND COMMUNITY STRUCTURE

S. WIDDICOMBE*1,M.A.KENDALL1, D. M. PARRY 2

1 Plymouth Marine Laboratory, Prospect Place, West Hoe, Plymouth, PL1 3DH,UK 2 Institute of Marine Studies, University of Plymouth, Drakes Circus, Plymouth, PL4 8AA,UK * Corresponding author Email: [email protected]

SURFACE FEATURES ABSTRACT. – Relationships between the surface features created by infaunal bio- BENTHIC DIVERSITY turbators and the diversity and structure of a macrobenthic community were identi- BIOTURBATION fied at spatial scales of 30 to 100 cm for an area of shallow (10 m) subtidal NON-DESTRUCTIVE MONITORING ≥ MEGAFAUNA sediment at Jennycliff Bay (Plymouth, UK) in May 2000. At spatial scales 80 cm, MACROFAUNA macrofaunal heterogeneity, and diversity, was greater in areas that contained many different surface features compared with areas containing few features. This observation is field evidence to substantiate suggestions that the non-destructive quantification of conspicuous surface features is a useful surrogate of macrobenthic diversity. This surrogate may be most useful in assessing large scale impacts on sediment habitats such as faunistic changes over long distances, long-term temporal changes in a particular place or assessing differences between habitats or geographical locations.

STRUCTURES DE SURFACE RÉSUMÉ. – Des rapports entre les dispositifs de surface créés par des organismes DIVERSITÉ BENTHIQUE fouisseurs et la diversité et la structure d’une communauté macrobenthique ont été BIOTURBATION identifiés aux échelles spatiales de 30 à 100 cm pour un secteur (10 m) de sédiment SURVEILLANCE NON DESTRUCTIVE MÉGAFAUNE subtidal peu profond dans la baie de Jennycliff (Plymouth, R-U) en mai 2000. Aux MACROFAUNE échelles spatiales ≥ 80 cm, l’hétérogénéité de la macrofaune, et la diversité, sont plus élevées dans les secteurs qui contiennent de nombreuses structures extérieures différentes en comparaison avec les secteurs contenant peu de ces dispositifs. Cette observation est une démonstration in situ pour justifier des suggestions selon les- quelles la quantification non destructive des structures de surface remarquables est un substitut utile de la diversité macrobenthique. Ce substitut peut être particulière- ment utile lors de l’évaluation des impacts à grande échelle sur les habitats des sé- diments, tels que les changements faunistiques sur de longues distances, les changements temporels à long terme d’un site particulier ou lors de l’évaluation des différences entre habitats ou localisations géographiques.

INTRODUCTION solely a result of differences in the intensity and frequency of disturbance (Widdicombe et al. Bioturbation has a substantial influence on the 2000). Mesocosm studies have shown that the physical and chemical properties of the sediments; intensity of bioturbation and the identity (or burrowing and/or feeding activities can have a functional type) of the bioturbator responsible, strong influence on sediment granulometry, the act simultaneously on different elements of penetration of oxygen, nutrient flux and the fre- macrobenthic communities (Widdicombe & Austen quency of sediment disturbance (e.g. Widdicombe 1999, Widdicombe et al. 2000). Further evidence & Austen 1988, Rowden & Jones 1993, Nedwell & for the importance of sediment complexity in the Walker 1995). As the properties of the sediment maintenance of high diversity macro-infaunal com- change so too does its suitability for colonisation munities was provided by Thrush et al. (2001). and exploitation by secondary species. Bioturbating These authors suggested that between 74 and 86% organisms have been shown to impact on of the variance in macrobenthic diversity was ex- macrobenthic communities both directly through plained by habitat structure. Clearly bioturbation is physical interactions (e.g. predation) and indirectly an important process in producing heterogeneity or through changes in the sediment environment. complexity within the sediment habitat thus main- However, the scale of bioturbation impact is not taining levels of diversity. 180 S. WIDDICOMBE, M. A. KENDALL, D.M. PARRY

In the face of growing environmental pressures lected from quantitative remotely operated vehicle on coastal marine ecosystems the need to assess (ROV) observations. Such data have been validated and monitor changes in marine communities is previously against those derived from direct diver greater than ever. In particular, legislation de- observation (Parry et al. 2002) and have proved to signed to protect marine habitats and biodiversity be an accurate and effective representation of the (e.g. EC Habitats Directive 1992) has placed great biogenic landscape. Furthermore, spatial referenc- demands on those responsible for the management ing of ROV images allows mapping of megafaunal of coastal resources. However, traditional methods assemblages within benthic landscapes (Parry et al. of biodiversity assessment that involve the quanti- 2003). fication of entire communities are highly labour in- Through the intensive study of a relatively small tensive and therefore expensive (McIntyre 1983, area of bioturbated seabed, the current study aims Warwick 1993). If efficient monitoring and man- to establish whether a relationship exists between agement is to be achieved over large areas of the identity and abundance of biologically pro- coastal waters, techniques need to be developed duced surface features and the structure and diver- that allow rapid, cost effective assessment of com- sity of infaunal macrobenthic communities. The munity structure and diversity. Such methods do potential of surface features as an accurate surro- not need to deliver outputs with high levels of pre- gate of macrobenthic community structure and di- cision, it is sufficient that they produce information versity is discussed. that is “fit for purpose”. A number of authors (e.g. James et al. 1995, Kendall & Widdicombe 1999) have shown that when assessing the pattern of distribution of benthic assemblages, coarser sieve METHODS meshes or identification to higher taxa produce results that differ little from those produced using a complete survey design. Others have suggested the Sample area: The field sampling was conducted between use of surrogates, elements of the biota whose pres- 30th-31st May 2000 at Jennycliff Bay (50º21.0’N ence predicts biological properties of the whole as- 04º07.8’W), a sheltered bay within Plymouth Sound. semblage. For example, Oslgard & Somerfield This area is an extensive stretch of sandy mud (Marine (2000) suggested that analysis of a small subset of Biological Association 1957) with a water depth of ap- the polychaete fauna might be sufficient to predict proximately 10 m and generally low current speeds. Jen- the diversity of the whole assemblage. Similarly, nycliff Bay supports high megafaunal densities, Warwick & Light (2002) have investigated the use characterised by thalassinidean shrimp (Upogebia del- of molluscan death assemblages on seashores to taura and Callianassa subterranea), Goneplax rhomboi- predict the biodiversity of the adjacent offshore des and several large bivalve species (Parry 2002). fauna. Other studies have shown that it is not al- Patterns in macrobenthic community structure have been well studied at this site (Kendall & Widdicombe 1999). ways necessary to identify organisms to species level as the same patterns of community change Characterisation of biogenic landscape: Megafaunal as- may be observed for data at the family or genus semblage data extracted from remote observations of the level (Warwick 1988, Olsgard et al. 1997). seabed may be considered descriptors of the biologically-mediated landscape rather than an absolute estimate Even with a reduced sample processing effort, of megafaunal species abundance because different bur- classical analysis of benthic samples remains both rows have a variable number of surface openings. Never- costly and time consuming. As remote collection of theless, the morphology of surface features may be used benthic sediment is destructive it must be used with to infer the identity of species responsible for burrow great discretion in environmentally sensitive areas. construction (e.g. Nickell & Atkinson 1995). The defini- If the extent of bioturbation and the identity of the tion of megafaunal assemblage used in the present study bioturbators combine to influence local sediment encompassed surface dwelling megafauna and biogenic diversity, the acquisition of information on sediment features, such as burrow openings and mounds, bioturbators and their distribution might be used to because many soft-sediment megafaunal species bury deep in the sediment. To describe the biologically-me- predict patterns of infaunal benthic biodiversity. diated landscape, the abundance of both megafaunal in- Although the majority of benthic organisms have dividuals and surface features was used to construct some bioturbational impact, it seems intuitive that matrices of Bray–Curtis similarity. larger bodied, highly active species will have the To quantify the abundance and diversity of both me- greatest impact on sediment structure and associ- gabenthic and macrobenthic infauna a 3 × 3 m steel ated communities. These larger animals often cre- frame was deployed. This frame was subdivided into a ate characteristic surface features (Nash et al. grid of 50 × 50 cm and was fixed to the sea floor by em- 1984, Atkinson & Nash 1990, Atkinson et al. 1998, bedding the steel legs at each corner of the frame firmly Parry 2002) from which their identity can be de- into the sediment. The area within each of the 50 × 50 duced. cm cells was filmed using a Remotely Operated Vehicle (ROV) and the images recorded on video tape. The ROV To facilitate the identification and mapping of camera was fitted with the Automated Benthic Image bioturbators and surface features, data may be col- Scaling System (ABISS), a structured lighting array that SURFACE-FEATURES AND MACROFAUNAL DIVERSITY 181 was used for image scaling (Pilgrim et al. 2000). Benthic rowing megafauna detected in underwater images were Imager software (University of Plymouth, UK) (Pilgrim measured using Benthic Imager (Pilgrim et al. 2000), et al. 2000) was used to analyse the laser spot pattern but were only included in analyses if the diameter excee- and calculate camera orientation allowing accurate mea- ded10mm. surements (±5%) of features to be taken from each still The euclidean distance between each macrofaunal image. Still images were selected and captured from the core and each megafaunal surface-feature was calculated videotape such that the entire survey plot was represen- using MATLAB software (version 5.03, The MathWorks ted to allow measurement of all biogenic surface-featu- Inc.). Around each macrofauna core biogenic features res observed. Features counted included burrows, tubes were quantified with progressively larger circles of 30, and all surface dwelling megafauna. The identity of spe- 40, 50, 60, 70, 80, 90 and 100 cm diameter. These circles cies responsible for burrow construction was determined are subsequently referred to as virtual quadrats. Within where possible using the morphological characteristics each of the virtual quadrats data on the identity and described by Marrs et al. (1996) and by direct sampling abundance of surface features were obtained from analy- of observed features (Parry 2002) sis of the video images. Sampling macro-infauna: After the grid had been map- Multivariate data analyses followed the methods des- ped by the ROV, 61 macrofauna samples were taken, cribed by Clarke (1993) using the PRIMER version 5.0 one sample at the centre of each cell and a sample at software package (Clarke & Warwick 1994). Analysis √√ each intersection (Fig. 1). Macrofaunal samples were ta- was carried out on both untransformed and transfor- ken by divers using 10 cm diameter plastic cores med macrofauna data, using the Bray-Curtis similarity (0.008 m2 surface area) pushed into the sediment to a measure, to examine different components of the com- depth of 30 cm. These were carefully excavated and sea- munity. Analysis of untransformed data is sensitive to led inside plastic bags. On the surface, the sample bags changes in the abundance of the dominant species whilst √√ were opened and the water inside decanted over a analysis of transformed data detects effects on commu- 0.5 mm mesh. Each bag was filled with 10% buffered nity structure generally including changes in the abun- formalin and resealed. After 4 days, the samples were dance of the rarer species, without being unduly washed over a 0.5 mm mesh before all animals were ex- influenced by dominant, high abundance species. tracted. Animals were identified under a binocular mi- Pair-wise Bray-Curtis similarities (untransformed croscope to the lowest practical taxonomic level. In the data) between megafaunal surface-feature samples of the case of the cirratulid polychaetes, the separation of indi- same size were calculated and similarity matrices for viduals of the genera Monticellina / Tharyx, Caulerialla each of the eight sample sizes created. RELATE was and Chaetozone was normally possible; however when used to compare each of these matrices with the equiva- damaged the individual species of Monticellina / Tharyx lent matrix calculated from the infaunal abundance data. could not be reliably separated and hence have been For the smallest virtual quadrat size (30 cm diameter) all combined in all data analyses. Wherever possible, spe- macrofauna samples were used in the analysis. However, cies nomenclature followed Howson & Picton (1997). for larger quadrats an area of overlap existed when the quadrat was overlain neighbouring macrofauna samples. Relating surface features to macrobenthic community Consequently, for comparisons involving quadrats grea- structure: Megafauna have been defined operationally as ter than 30 cm in diameter smaller subsets of macrofau- those organisms large enough to be observed by a came- nal samples were used to ensure sample independence. ra (Grassle et al. 1975), yet the absolute dimensions of For quadrats of 40 cm and 50 cm diameter 2 separate an organism in an image depend upon image resolution. subsets of macrofauna cores were used (Fig. 2a), whilst In the present study, all epibenthic megafaunal indivi- for quadrats with a diameter greater than 50 cm a subset duals and biotic sediment structures associated with bur- of 9 macrofauna samples (B2, B6, B10, F2, F6, F10, J2, J6 and J10) was used (Fig. 2b). For each of the nine, non-overlapping, 1 m diameter surface-feature samples, a mean Bray-Curtis similarity value (√√ transformed data) was calculated from pair- wise analyses using all of the 5 macrofauna cores contained within the area of the surface-feature sample. To establish the extent to which the number of different types of surface-features in any given area predicted macrofaunal heterogeneity, mean Bray-Curtis similarity was plotted against surface feature diversity (number of different types of surface feature within a given area).

Relating surface feature diversity to macobenthic diversity: Univariate measures of diversity [number of species, number of individuals, Margalef species richness (d), Shannon-Wiener diversity (H’) and Pielou’s eveness (J’)] were calculated, using PRIMER, for each macrofauna sample and plotted against the diversity of the surrounding surface features. This was done for all sizes of virtual quadrat. In addition, for quadrats with a diameter of 1 m, the 5 macrobenthic samples contained within Fig 1. – Positions and labelling scheme for macro-infau- each one were pooled and the diversity of these pooled nal samples. samples was compared with the diversity of the surroun- 182 S. WIDDICOMBE, M. A. KENDALL, D.M. PARRY

Fig 2. – Macrofauna cores used (filled circles) for comparisons with a, virtual quadrats of 40 and 50 cm diameter (virtual quadrats of 50 cm diameter shown) and b, virtual quadrats of 60, 70, 80, 90 and 100 cm diameter (virtual quadrats of 100 cm diameter shown).

ding surface features. In all cases surface feature diversi- nated by polychaetes (73), molluscs (19) and crus- ty was the number of different types of surface feature taceans (19). The species rich community at within a given area. Jennycliff Bay was typical for sandy-mud sediments of unpolluted boreal regions of the northeast Atlantic. RESULTS

Relating surface features to macrobenthic Surface features observed community structure

A total of eleven different megafaunal biotic For small virtual quadrat sizes (less than 80 cm feature types were identified from the images diameter), patterns of similarity generated by sur- (Table I). Biogenic sediment features included face feature abundance data were different from thalassinidean mud shrimp burrow openings patterns generated using the macrofauna abun- (Upogebia deltaura and Callianassa subterranea) dance data (Table II). Significant correlations were and mounds (Callianassa subterranea), angular observed when the larger virtual quadrat sizes were crab burrows (Goneplax rhomboides), bivalve si- used. Changes in the relative abundance of the nu- phons and a variety of unknown surface openings merically dominant macrofauna (untransformed and feeding pits. Megafaunal species included the data) correlated to changes in the surface features infaunal bivalves Lutraria lutraria, Acanthocardia for virtual quadrat sizes of 90 cm and 1 m, whilst sp. and Unknown Bivalve 1, plus ophiuroids. macrobenthic community changes due to the rarer species (√√ transformed data) were correlated to Macrofauna community of study area changes in the surface features for virtual quadrat sizes of 80 cm, 90 cm and 1 m (Table II). This A total of 124 taxa were identified during this would suggest that changes in the identity and den- study with the community being numerically domi- sity of surface features over these larger, spatial SURFACE-FEATURES AND MACROFAUNAL DIVERSITY 183

Table I. – Description of the surface features observed.

scales mimic those changes observed in both the sity of the surrounding megabenthic surface fea- numerically dominant and the rarer macrofaunal tures. However, if 5 macrofauna cores, taken species. within each of the nine 1 m diameter surface fea- In areas with a large number of different surface ture quadrats, are pooled then a relationship be- features, the macrobenthic community exhibited tween macrobenthic diversity and that of surface more small scale (< 1 m) heterogeneity than in ar- features emerges. At this scale as the diversity of eas with a low diversity of surface features (Fig. 3). surface features increases so does that of This response was only observed in √√transformed macrobenthos (Fig. 3). During sampling it was ob- data and would suggest that this increased patchi- served that one quadrat contained a very extensive ness is due to changes in the identity and abun- Goneplax burrow. It was considered that the large dance of the rarer species. amount of disturbance associated with this feature would have masked the effects of the other features and caused lower than expected macrobenthic di- Relating surface feature diversity to versity. This sample was therefore omitted from macrobenthic diversity the analysis. The relationship between surface feature diversity and macrobenthic diversity gradually broke down as the different diversity indices used No relationship was observed between the diver- became increasingly more reliant on eveness in sity of individual macrofauna cores and the diver- preference to species richness. 184 S. WIDDICOMBE, M. A. KENDALL, D.M. PARRY

Table II. – Comparisons of surface feature community DISCUSSION structure (untransformed data) with macrobenthic community structure (untransformed and transformed data) at a range of spatial scales using RELATE (bold values indicate significant correlations). This study has shown that for the particular macrobenthic community observed, faunal heterogeneity, and diversity, was greater in areas that contained many different surface features compared with areas containing few features. This observation is field evidence to substantiate previous work that suggested macrofaunal diversity would be highest in areas populated by a high diversity of bioturbating species (Widdicombe et al. 2000, Widdicombe & Austen 1999). This assumption had previously been based on experimental evidence that showed different species of bioturbators, or types of bioturbation, generated characteristic associated communities of macrofauna. If such a mechanism was responsible for the relationships seen in the current study it would be expected that the strongest patterns should be visible in the identity and abundance of the rare species, as was the case. Only when macrofauna community data was analysed using moderately heavy transformation, thus reducing the impact of numerically dominant species and enhancing that of rare species, were any relationships observed. By using a particularly fine scale of spatial resolution it could be seen that, only when analysing areas of 80 cm in diameter and larger was the diversity and community structure of surface features related to that of the infauna. The reason the relationship is only seen for the largest size of surface feature quadrats is most probably related to the scale at which the surface features occur. A single feature may be in excess of 30 cm across and consequently the smaller virtual samples contained only a few surface features. This low number of surface features per quadrat would make macrobenthic – surface feature comparisons unre- alistic. It is clear that the processes involved in the creation of sediment surface features, particularly bioturbation, might be considered to set infaunal diversity at the scale of habitats rather than that of samples. This suggests the greatest value of using surface features as a surrogate for diversity would come when assessing large scale impacts on sediment habitats; faunistic changes over long distances, long-term temporal changes in a particular place or assessing differences between habitats or geographical locations. Whilst visual mapping of the sediment surface provides an enormous amount of information on the identity and distribution of many key bioturbating species there are some species (e.g. Fig 3. – Top, the relationship between megabenthic sur- polychaetes such as Nephtys sp., burrowing face feature diversity and infaunal homogeneity [as mea- echinoids and nuculacid bivalves) which do not sured by mean Bray-Curtis similarity] (± 95% create conspicuous surface features. These “invisi- confidence intervals). Bottom, relationship between sur- ble” species have been shown to have a significant face feature diversity and the total number of macrofau- impact on the diversity and structure of nal species from 5 pooled samples (R=0.6697). macrobenthic communities (Widdicombe et al. SURFACE-FEATURES AND MACROFAUNAL DIVERSITY 185

2000) and their inclusion in any measure of bioturbators should be examined and the differen- bioturbator diversity may well provide a more ro- tial effects of such species on the macrofauna need bust surrogate for diversity than surface features to be considered when interpreting the data. Addi- alone. In the past it has been impossible to map tionally, to be a truly valuable assessment tool, the these species non-destructively at broad spatial relationship must be shown to exist in other geo- scales. However, many of these animals have a di- graphic locations and in different soft-sediment rect impact on aspects of sediment structure e.g. al- habitats. tering the depth of oxygen penetration, consolidat- ACKNOWLEDGEMENTS. – The work contained in this ing / destabilising the sediment fabric. A number of paper was funded by a grant from UK DEFRA (project techniques are now available that allow such no AE1137) and by the UK Natural Environment Re- changes in sediment quality to be observed. Firstly, search Council through the Plymouth Marine Laboratory acoustic imaging techniques (e.g. RoxAnnTM and research program (Aim 3). We would like to thank Dr M QTC ViewTM) can provide data on the “hardness” Austen, Dr M Burrows, Dr L Nickell, Dr P Somerfield, or “softness” of sediments by measuring the extent Dr D Pilgrim’ Rob Wood, and the Crew aboard RV to which sound waves penetrate below the sedi- Squilla and RV Sepia, all of whom provided valuable ment surface. Secondly, vertical cross sections of contributions to aspects of this study. the sediment, down to a depth of 20 cm, can now be made using Sediment Profile Imagery (Rhoads & Germano 1982, Smith et al. 2003) providing information on the depth of the oxic layer and reveal- REFERENCES ing sub-surface bioturbatory structures. It may therefore be possible to develop an efficient, non- destructive method of monitoring biodiversity by Atkinson RJA, Nash RDM 1990. Some preliminary ob- using an integrated approach of visual observations servations on the burrows of Callianassa subterranea of the sediment surface and vertical profiles from the West Coast of Scotland. JNatHist24: 403- 413. through the sediment in addition to fine scale Atkinson RJA, Froglia C, Arneri E, Antolini B 1998. acoustic assessment of sediment parameters. Observations on the burrows and burrowing beha- viour of Brachynotus gemmellari and the burrows of several other species occurring on the Squilla grounds Whilst the results from the current study are off Ancona, central Adriatic. Sci Mar 62: 91-100. clearly encouraging for the use of conspicuous sur- Clarke KR 1993. Non-parametric multivariate analyses face features as surrogates for macrobenthic diver- of changes in community structure. Aust J Ecol 18: sity, a number of issues should be addressed. Due 117-143. to the enormous amount of effort required to sam- Clarke KR, Warwick RM 1994. Changes in marine com- pleevenasmallareawiththedegreeofspatialres- munities: an approach to statistical analysis and inter- olution achieved here, the current study was con- pretation. Natural Environment Research Council, ducted over a very small spatial scale. Swindon. Consequently, the findings in this paper can not EC Habitats Directive 1992. The Council Directive on the conservation of the natural habitats and of wild necessarily be extrapolated to all soft-sediment fauna and flora (92/43/EEC of 21 May 1992) communities at all places or times. It is also true Grassle JF, Sanders HL, Hessler RR, Rowe GT, McLel- that the ability to discriminate between the similar lan T 1975. Pattern and zonation: a study of the ba- surface features generated by functionally different thyal megafauna using the research submersible organisms must be improved. For example, the cur- Alvin. Deep Sea Res 22: 457-481. rent study identified two features described as James RJ, Lincon Smith MP, Fairweather PG 1995. thalassinidean openings and thalassinidean Sieve mesh size and taxonomic resolution needed to mounds, both of these features were generated by describe natural variation of marine macrofauna. Mar two different species of thalassinidean shrimp; Ecol Prog Ser 118: 187-198. Callianassa subterranea and Upogebia deltura. Howson CM, Picton B 1997. The species directory of Whilst both these species are burrow-forming the marine fauna and flora of the British Isles and sur- decapods the nature of their burrows, their feeding rounding seas. Ulster Museum and the Maine Conser- mode and consequently their impact on the associ- vation Society, Belfast and Ross on Wye. ated macrofauna are different. Consequently, the Kendall MA,Widdicombe S 1999. Small scale patterns in the structure of macrofaunal assemblages of shal- inability to differentiate between the features gen- low soft sediment. J Exp Mar Biol Ecol 237: 127-140. erated by these two species will have had an impact Marine Biological Society 1957. Plymouth Marine Fau- on the findings presented in this paper. The poten- na. tial use of non-destructive video surveys of con- Marrs SJ, Atkinson RJA, Smith CJ, Hills JM 1996. Cali- spicuous surface features as a tool for monitoring bration of the towed underwater TV technique for use the diversity of macro-infaunal communities high- in stock assessment of Nephrops norvegicus.Final lighted here must therefore be tested more widely Report to the European Commission 94/069: 155. before being realised. In particular, other areas McIntyre AD 1983. What happened to biological effects with differing assemblages of megafaunal monitoring ? Mar Pollut Bull 15: 391-392. 186 S. WIDDICOMBE, M. A. KENDALL, D.M. PARRY

Nash RDM, Chapman CJ, Atkinson RJA, Morgan PJ Rowden AA, Jones MB1993. Critical evaluation of sedi- 1984. Observations on the burrows and burrowing be- ment turnover estimates for Callianassidae (Decapo- haviour of Calocaris macandreae (Crustacea: Deca- da: Thalassinidea). J Exp Mar Biol Ecol 173: 265- poda: Thalassinoidea). J Zool Soc Lond 202: 425- 272. 439. Smith CJ, Rumohr H, Karakassis I, Papadopoulou KN Nedwell DB, Walker TR 1995. Sediment – water fluxes 2003. Analysing the impact of bottom trawls on sedi- of nutrients in an Antarctic coastal environment: mentary seabed with sediment profile imagery. JExp Influences of bioturbation. Polar Biol 15: 57-64. Mar Biol Ecol 285-286: 479-496. Nickell LA, Atkinson RJA 1995. Functional morphology Thrush SF, Hewitt JE, Funnell GA, Cummings VJ, Ellis and trophic modes of three thalassinidean shrimp spe- J, Schultz D, Talley D, Norkko A 2001. Fishing dis- cies, and a new approach to the classification of tha- turbance and marine biodiversity: the role of habitat lassinidean burrow morphology. Mar Ecol Prog Ser structure in simple soft-sediment systems. Mar Ecol 128: 181-197. Prog Ser 223: 277-286. Olsgard F, Somerfield PJ, Carr MR 1997. Relationships between taxonomic resolution and data transforma- Warwick RM 1988. The level of taxonomic discrimina- tions in analyses of a macrobenthic community along tion required to detect pollution effects on marine an established pollution gradient. Mar Ecol Prog Ser benthic communities. Mar Pollut Bull 19: 259-268. 149: 173-181. Warwick RM 1993. Environmental impact studies on Olsgard F, Somerfield PJ 2000. Surrogates in marine marine communities: pragmatical considerations. benthic investigations – which taxonomic unit to tar- Aust J Ecol 18: 63-80. get ? J Aqua Stress Rec 7: 25-42. Warwick RM, Light J 2002. Death assemblages of mol- Parry DM 2002. Quantification of marine megafaunal luscs on St Martin’s Flats, Isles of Scilly: a surrogate distribution patterns using a remotely operated ve- for regional biodiversity ? Biodiv Conserv 11: 99- hicle. PhD Thesis, Univ Plymouth, 233p. 112. Parry DM, Kendall MA, Pilgrim DA, Jones MB 2003. Widdicombe S, Austen MC 1998. Experimental evi- Identification of patch structure within marine ben- dence for the role of Brissopsis lyrifera (Forbes, thic landscapes using a Remotely Operated Vehicle. J 1841) as a critical species in the maintenance of ben- Exp Mar Biol Ecol 285-286: 497-511. thic diversity and the modification of sediment che- ParryDM,NickellLA,KendallMA,BurrowsMT,Pil- mistry. J Exp Mar Biol Ecol 228(2): 241-255. grim DA, Jones MB 2002. Comparison of abundance Widdicombe S, Austen MC 1999. Mesocosm investiga- and spatial distribution of burrowing megafauna from tion into the effects of bioturbation on the diversity diver and Remotely Operated Vehicle observations. and structure of a subtidal macrobenthic community. MarEcoProgSer244: 89-93. Mar Ecol Prog Ser 189: 181-193. Pilgrim DA, Parry DM, Jones MB, Kendall MA 2000. Widdicombe S, Austen MC, Kendall MA, Warwick RM, ROV image scaling with laser spot patterns. Under- Jones MB 2000. Bioturbation as a mechanism for set- water Technology 24: 93-103. ting and maintaining levels of diversity in subtidal Rhoads DC, Germano JD 1982. Characterisation of or- macrobenthic communities. Hydrobiologia 440: 369- ganism–sediment relations on the muddy sea floor 377. using sediment profile imaging: an efficient method of Remote Ecological Monitoring of the Seafloor Reçu le 18 août 2003; received August 18, 2003 (REMOTS System). Mar Ecol Prog Ser 8: 115-128. Accepté le 23 octobre 2003; accepted October 23, 2003