1

journal o f Animai |he taxonomic distinctness of coastal bottom-dwelling E c o lo g y 1 9 9 9 , 68,769 782 fish communities of the North-east Atlantic

STUART I. ROGERS*, K. ROBERT CLARKEI and JOHN D. REYNOLDSI *CEFAS, Lowestoft Laboratory, Pakefield Road, Lowestoft, NR33 OHT, UK; '\CCM S, Plymouth Marine Laboratory, Prospect Place, West Hoe, Plymouth, P LI 3DH, UK; and\School o f Biological Sciences, University o f East Anglia, Norwich, N R4 7TJ, UK VLIZ (vzw)

VLAAMS INSTITUUT VOOR DE 2 E ÍSummary f-LANDERS MARINE INSTITUTE O ostende Bol ' 1. New techniques for identifying the average taxonom ic range of species assem blages were applied to an extensive dataset of bottom -dw elling fish in the coastal waters of NW Europe. These taxonom ic distinctness indices provided m uch greater resolution than traditional diversity indices as they incorporated inform ation on taxonom ic relationships into an index which m easures species dom inance. U nlike standard m ea­ sures of species richness and diversity, the m ean value of these statistics is independent of sam pling effort, and this allows objective com parisons to be m ade between sam ples from studies where sam pling effort is not standardized. 2. A reduction in the average taxonom ic range between the fauna of western waters of the U K and that of the southern N orth Sea was consistent w ith the general decline in species richness observed between these regions, and suggests that these two factors m ay be spatially positively correlated. Indices calculated for individual sam ples of fish on a local scale, how ever, did not all fit this trend. 3. M uch of the variability in taxonom ic diversity within the coastal waters of NW Europe was caused by the variable geographical distribution of the elasm obranchs. O f all the families which com prise the fish com m unities, this group has life-history characteristics w hich m ake it m ost susceptible to im pact by com m ercial traw l fisheries. 4. The use of taxonom ic distinctness measures provided additional insights, of rel­ evance to biodiversity assessm ent, suggesting that they m ight usefully be applied to other aquatic and terrestrial fauna.

Key-worcls: taxonom ic diversity, groundfish, com m unity structure, fishing impact, N W E u r o p e .

Journal o f Anim al Ecology (1999) 68, 769-782

explain patterns on a smaller provincial or regional Introduction scale, but it has become increasingly necessary to at Explaining the large-scale patterns in species dis­ least understand the relative role of natural and tribution and abundance of both marine and ter­ anthropogenic perturbation. restrial environments requires inform ation on habitat The response of benthic m arine assemblages to dis­ extent and complexity (M agurran 1988; Rosenzweig turbance is thought to be more easily detected at 1995), the rates and effects of disturbance and levels higher taxonomic levels, because of the confounding of productivity (Connell 1978; Sousa 1979; Petraitis, influence of abiotic factors such as water depth and Latham & Niesenbaum 1989; Baltz 1991), and also the tem perature which will affect small-scale diversity at im pact of environm ental gradients on species richness the species level (W arwick & Clarke 1993). This hier­ (Pianka 1966; Ekm an 1967; Nelson 1994). Rarely are archical response to stress, first proposed by Pearson there sufficient environm ental data to be able to fully & Rosenberg (1978), exploits the fact that various phyla differ in their sensitivity to disturbance. Benthic environments which have been perturbed are gen­ Correspondence: D r S. I. Rogers, CEFAS, Lowestoft Lab­ © 1999 British oratory, Pakefield Road, Lowestoft, NR33 OHT, UK. Fax: erally kept in an early successional stage with low Ecological Society 01502513865. E-mail: [email protected] . species diversity, and often consist of species which

769 7 7 0 are closely related, while unperturbed benthic com­ by differing sample sizes, making meaningful com­ T a x o n o m ie m unities in a late successional stage often consist of a parisons impossible (Rogers e ta l. 1999). In addition, distinctness offish wider range of taxonomically distinct species (W ar­ the presence of a statistical testing fram ework for the communities wick & Clarke 1995). It is not clear, however, whether A+ index enables a comparison to be made between all communities which have a small num ber of species an observed taxonomic distinctness measure and its should, by definition, have a m ore limited taxonom ic expected range of variation. range than those with many species. Assuming that These measures of taxonomic distinctness have taxonomic diversity translates into ecological diver­ been applied to macrobenthos samples from the Eko- sity, then the taxonomic range of an assemblage may fisk oil-field in the North Sea (Warwick & Clarke be crucial in maintaining ecosystem stability during 1995). This study identified a high degree of sensitivity natural or anthropogenic perturbations (Hughes to subtle environm ental im pacts at this site, although 1994;Tilman 1996). these patterns have not been supported by data from Obscuring these patterns are the underlying bio­ other N orth Sea oil fields (Somerfield, Olsgard & Carr geographic processes, such as rates of spéciation, 1997). Taxonomic distinctness measures have also immigration and extinction, which determine total been applied to literature data on marine benthic species richness of a province (Rosenzweig 1995). The nematodes from various intertidal/subtidal and coas- resulting species diversity gradients are largely deter­ tal/estuarine sites in the U K , and also coastal habitats mined by the influence of environmental factors on in Chile (W arwick & Clarke 1998). These analyses rates of diversity generation and extinction. For m ar­ dem onstrate a decrease in taxonomic distinctness A+ ine organisms, the species richness between the tropics at polluted sites, such as the U K ’s Clyde Estuary and and polar regions generally, although not always, Liverpool Bay, com pared with taxonom ic distinctness shows consistently decreasing trends (Findley & Find­ at ‘clean’ sites in the Exe Estuary, the Scilly Isles and ley 1985; Pauly 1994) but see also Gee & W arwick the Northum berland coast. (1 9 9 6 ). The majority of published studies in which these These observations influence the way we interpret indices have been applied have used data from soft- species diversity trends in m arine environments, yet sediment macro-and meiobenthic faunas, yet some the conventional indices of community diversity, of the most extensive marine datasets describe the which use only the relative abundance of species, do temporal and spatial abundance of fish. In the only not describe the degree of taxonomic relatedness of known application of these indices to vertebrate popu­ those species. Assemblages with the same species rich­ latio n s, Haii & Greenstreet (1998) show that, for a ness may either comprise species which are closely long time-series dataset for bottom -dwelling fish from related to one another taxonomically, or they may the northern North Sea, trends in taxonomic dis­ be more distantly related (Ludwig & Reynolds 1988; tinctness and diversity were identical to those shown W arwick & Clarke 1995). To quantify such changes by conventional indices. Their observed positive cor­ in taxonomic relatedness, W arwick & Clarke (1995) relation between A* and Hills N1 and N2 diversity and Clarke & W arwick (1998b) have defined three indices suggests that assemblages that are more biodiversity indices, which quantify the taxonomic diverse (species-rich) always contain species that have diversity and taxonomic distinctness of a faunal a wider average taxonomic range than assemblages assemblage using the path lengths between fish gro­ which are less diverse, but this has yet to be tested at uped by their taxonomic relationships. They show a range of spatial scales. that taxonomic diversity (A) is a natural extension In contrast to these large-scale patterns in species of Simpson diversity, incorporating information on diversity, little is known about the processes which taxonom ic relationships within a sam ple into an index structure the diversity of smaller regional environ­ measuring species dominance. Taxonomic dis­ ments. In this paper we re-examine an extensive data­ tinctness (A*) is a form of taxonomic diversity that set describing the relative abundance of bottom -dwell­ limits the influence of patterns in species dominance ing fish in the coastal waters of NW Europe (Rogers by dividing A by a form of Simpson diversity, thus e t a l. 1998,1999) to test whether these new taxonomic constructing a measure which more nearly reflects diversity indices can m irror changes in species diver­ pure taxonom ic relatedness. A third index, taxonom ic sity at different scales, ranging from the entire study distinctness (A); considers only the special case where area to individual sample sites. Using a comprehensive abundance inform ation is not available or is ignored taxonomy of the bottom-dwelling fish found in the (i.e. presence/absence data). Clarke & Warwick N orth-east Atlantic, we first calculate indices of taxo­ (1998b) show that these statistics are largely inde­ nom ic diversity using fish relative abundance, and also pendent of sampling effort, which makes their use using simple presence and absence data. W e then test attractive in spatially extensive or long time-series whether these taxonom ic diversity indices correspond studies where total sampling effort in different areas to know n patterns of species diversity, at intermediate © 1999 British or at different times is rarely standardized. Standard and at small spatial scales. W e also describe how local Ecological Society Journal o f Animal diversity indices such as species richness, Shannons’ factors which may structure the community, such as Ecology, 68, 769-782 index and evenness measures can be highly influenced the effect of man made disturbance on sensitive taxa, 7 7 1 can influence the diversity of the entire demersal w h e re x¡ denotes the abundance of the ith of s sp ecies S. I. Rogers, fa u n a . o b s e r v e d , n ( — £,x,) is the total num ber of individuals K.R. Clarke & in the sam ple and co,yi s the weight given to the path J.D. Reynolds Materials and methods length linking species i a n d j in the taxonomy. Taxo­ nomic diversity (A) can be thought of as the average path length between any two randomly chosen in d i­ SURVEYS v id u a ls from the sample (including individuals of the A series of international beam trawl surveys of north­ same species), whereas taxonom ic distinctness (A*) is east Atlantic coastal waters have sampled the abun­ the average path length between two random ly chosen dance of coastal (10-120 m depth) bottom-dwelling individuals, conditional on them being from different species since the late 1980s, using sampling techniques species. W hen only presence/absence data are used described in detail by Rogers e ta l. (1998). Different (i.e . x ¡ = 1 for all species present) both equations beam trawls were used during the surveys, depending reduce to an even simpler form of taxonomic dis­ on the ability of the different vessels to deploy them, tinctness, A+, which can be thought of as the average and the varying nature of the seabed in different parts path length between any two random ly chosen sp ec ie s of the region. All samples were collected during the present in the sample. All three indices were calculated third quarter of the year, at which time m ost fish were for the groundfish data, including an additional m odi­ dispersed and not undergoing migrations to or from fied form of eqn 2 in which the abundances were spawning grounds. In general, fishing stations sam­ square root transformed, to downweight the con­ pled were at fixed positions, except in D utch and G er­ tribution to the index of the numerically dominant m an surveys w here the station positions were stratified sp ec ie s. by International Council for the Exploration of the Following W arwick & Clarke (1995), the simplest Sea (ICES) rectangle (1 degree of longitude x 0-5 form of path length weighting was adopted for the degree latitude) and selected annually on a pseudo­ 13 taxonom ic levels (Table 1), namely: co = 0 for two random basis. In the North Sea, all stations were individuals of the same species; co — 1 for two indi­ grouped by quarter ICES rectangle, and the centre of viduals in the same subgenus but of different species; this unit was used as the nominal fixed station co — 2 for species in the same genus but different position. The duration of each tow was either 15 or subgenera; co = 3 for species in the same tribe but 30 m in. A t each station the fin-fish catch was identified different genera, etc. W here finer subdivisions do not to species where possible, measured, and their num ­ exist within some groups of taxa, the path length is bers recorded (Rogers e ta l. 1998). M ean catch rates defined as that for the first existing link. For example, of each species or taxa (raised to the num ber per 8 m if a particular genus does not possess subgenera then beam trawl per hour tow) at each fixed station position all links between its constituent species carry a weight between 6 m to 35 m depth, and for the period 1990— o f co = 2. 96, were calculated. Only those fish species which were There is a degree of arbitrariness about a constant classified as bottom-dwelling species, and were there­ path length of one unit between each taxonomic level fore well sampled by beam trawls, were used in analy­ in the hierarchy; for exam ple, if a wholly unused level ses of community structure (Rogers e ta l. 1998). The is inserted, the relative distinctness values for the set of ICES Divisions surveyed were subdivided into nine samples will change. W here there is a comprehensive coastal sectors (Fig. 1). regional list for a group of species, it is possible to refine the weighting to reflect the reduced number of

TAXONOMIC DIVERSITY INDICES representatives at each taxonomic level. An alter­ native weighting assessed here ({w0}, Table 1) sets the A taxonom y of bottom-dwelling fish species was used path length from one group to the next coarsest classi­ to determine their relatedness (Fig. 2). W e compiled a fication as proportional to the percentage by which composite taxonomy based primarily on Nelson taxon richness decreases. The path lengths are then (1994), with additional information for Chon- summed and the final path lengths scaled so that the drichthyes based on M cEachran & M iyake ( 1990) and total weight linking two species that belong to differ­ M cEachran, Dunn & Miyake (1996). In addition to ent subdivisions (the top category of the hierarchy) is the five taxonom ic levels shown in Fig. 2, we included set equal to 100. It can be seen from Table 1 that subgenera, tribes, subfamilies, superfamilies, sub­ this alternative weighting achieves the desired balance. orders, series, superorders, and subdivisions where For example, from genus (s3 = 72) to tribe (s4 = 67) possible. Three biodiversity indices defined by Clarke now adds only 3-6 to the path length whereas that & W arwick (1998b) were then calculated from the from suborder (s8 = 33) to order (s9 = 14) adds 10'5, bottom-dwelling fish abundances: and addition or deletion of an unused category will

A = ['£Zi

58°

Scotland

North 56° Denmark Sea

5 0 m

German Bight

Ireland G ermany

Netherlands England 52° 5 0 m

B ris to l D o v e r C h a n n e l S tra it B elgium

C h a n n e l

C e ltic S e a

F r a n c e

Fig. 1. The shelf seas of continental NW Europe, showing the 50 m depth contour and the nine coastal sectors used in these analyses: (1) = Bristol Channel; (2) = Western Irish Sea; (3) = Eastern Irish Sea; (4) = Western Channel; (5) = North-eastern Channel; (6) = South-eastern Channel; (7) = South-western North Sea; (8) = South-eastern North Sea; (9) = East Central N o rth Sea.

ence/absence, not only can distinctness A+ still be From its construction, the taxonomic distinctness calculated, and com pared across samples of different index A* should be independent of the Simpson spec­ size, but a significance test can also be carried out. ies diversity index, i.e. they must be quantifying This tests for the departure of A,„+, the distinctness different aspects of diversity. It is thus interesting to measure for any sample of m species, from the overall show how these two community attributes are cor­ value A + for a ‘global’ species list for that region. The related across the available, spatially dispersed, sets test is based on the theoretical mean and variance of of samples. Simple scatter plots are used to examine A,„+, values obtained by random sam pling of m species th is . (without replacement) from the total list of s species (Clarke and Warwick 1998b). Although the theor­ Results etical mean remains constant, the variance naturally increases as m decreases, and so the approxim ate 95% The taxonomic diversity of the untransformed fish confidence intervals take the form of a ‘funnel’. The abundance matrix showed similar patterns for the values of A+ for any particular set of samples can then three diversity indices A, A* and A+ (Fig. 3). Although be related to this confidence ‘funnel’, to gauge the m ost coastal sectors showed similar values, the East­ extent to which their taxonom ic distinctness falls sig­ ern Central and South-eastern North Sea, and the nificantly below (or above) that expected. Assuming a W estern Irish Sea had lower m ean values of A (taxo­ 0 1999 British Ecological Society null hypothesis that each sample is a random selection nom ic diversity) than others. M ean values of A* (taxo­ Journal of Animal from the total species list, all values of A+ should fall nomic distinctness) m aintained the same general pat­ Ecology, 68, 769-782 within the confidence funnel. tern (Fig. 3b). The use of square root transformed 7 7 3 S. I. Rogers, CLASS ORDER SUBORDER FAMILY GENUS K.R. Clarke &

J.D. Reynolds Carcharhiniformes Scyliorhinus I' —"Mustelus »—'■"Galeorhinus Chondrichthyes Squaliformes Squalus Torpediniformes Torpedo R ajiform es — 'Raja Anguilliformes Anguilla Conger — Raniceps Coryphaenoides Ciliata Enchelyopus G ad ifo rm es EGaidropsarus — Merluccius ■Molva 'Gadus Melanogrammus •Merlangius ■Trisopterus Pollachius Lophiiformes — '— Lophius Mugiliformes^ — ■Chelon Z eifo rm es - Z eus Capros .Gasterosteus Gasterosteiformes Syngnathus Entelurus E Hippocampus -Trigla -Aspitrigla ■Eutrigla -Trigloporus Scorpaeniformes i Myoxocephalus ■ Taurulus Agonus Cyclopterus Teleostei Liparis — - Dicentrarchus Spondyliosoma Mullus Cepola Labrus Ctenolabrus E Crenilabrus Zoarces Perciformes Chirolophis ■Lumpenus ■Pholis Echiichthys Trachinus •Blennius •Diplecogaster •Callionymus Gobius •Lesuerigobius •Scophthalmus Lepidorhombus Phrynorhombus Zeugopterus Arnoglossus Hippoglossoides Hippoglossus Pleuronectiformes Glyptocephalus Limanda Microstomus Platichthys Pleuronectes Buglossidium Microchirus Pegusa

Tetraodontiformes Solea Mola Fig.2. The taxonomy compiled for this analysis, simplified to show five from a total of 13 levels of classification.

© 1999 British Ecological Society Journal o f Animal Ecology, 68, 769-782 7 7 4 Table 1. The 13 levels of classification (k ) used in the bottom- (a) dwelling fish taxonomy, with the number of each type {s*}, T a x o n o m ie the ‘standard’ weighting scheme {co*}, and an alternative distinctness o f fish version {co*0} utilising an accumulation of the proportional communities decrease in taxon richness values { 5 *}. M ore precisely, fo r a A specific k , co* is the weighted path length between species belonging to differing taxon group k but the same group k + 1

k Level Sk cu* cok° (b) 1 Species 93 1 1-3 2 Sub-genus 89 2 6-9 3 G enus 72 3 8-9 A' 4 T ribe 67 4 12-5 5 Sub-fam ily 59 5 2 1 4 6 F am ily 41 6 22-9 7 Super-family 39 7 2 7 4 8 S u b -o rd er 33 8 4 4 4 11 9 O rd er 14 9 54-9 (0 ) 10 Series 9 10 61 4 11 Super-order 7 11 65-6 10 12 Sub-division 6 12 85-3 13 C lass 2 13 100 A’(/) JB

( d ) abundance data for the calculation of A*, to rh

downweight the contributions of dom inant species, 10 again identified the Eastern Central and Southeastern A + North Sea as taxonomically depauperate, and gen­ r+i rh erated sim ilar values for all other coastal sectors (Fig. 3c). Using only presence/absence data, the pattern in

taxonom ic distinctness A+ between sectors was m ain­ 8 9 tained for the Eastern Central and South-eastern North Sea, but the South-eastern Channel also had West UK Channel North Sea low values of A+ (Fig. 3d). M ean values of A + for all Fig. 3. For the nine coastal sectors in Fig. 1, the taxonomic sectors except the Bristol Channel fell below that of diversity A (a) and the taxonomic distinctness A* (b) of the untransformed fish abundance matrix, and the taxonomic the theoretical mean, indicating that m ost of the indi­ distinctness A* of square root transformed abundance data vidual sectors had lower average taxonomic range (c), and the taxonomic distinctness A+ of presence/absence than the entire survey area. data (d). For each sector the mean value with 95% confidence Comparison of taxonomic diversity A and taxo­ limits is shown. The theoretical mean value of A+ calculated nomic distinctness A* values determined from fish by random sampling of species from the total species list is marked on Fig.3d. samples from all fishing station positions (Fig. 4a), showed an underlying positive correlation, but with considerable variability in this relationship to the lower right hand side of the figure. Here, som e samples which were highly dom inated and so had low values of A+ calculated with and without the modified weigh­ taxonom ic diversity A, represented assemblages which ting shows that they are highly correlated (Fig. 4d). were taxonomically broad (i.e. high values of A*). There was no relationship between the total num ber Note that the converse situation, i.e. taxonomically of species in individual samples, and their taxonomic restricted assemblages with high A, cannot occur. The distinctness (A+) calculated using presence/absence effect of data transform ation on the calculation of A* data (Fig. 5). M any of these A+ values, however, were was limited, and the correlation between A* calculated lower than the theoretical mean determined by ran­ using transformed and untransformed data remained dom sampling from the entire bottom-dwelling fish positive (Fig. 4b). In contrast, the apparently poor fauna, a pattern already evident from Fig. 3d. correlation between A+ and A* clearly showed the A lthough m ost sample values fell within the 95% con­ influence of species abundances when calculating fidence limit funnel, a proportion had significantly taxonom ic distinctness A* (Fig. 4c). Finally, although lower values of A+ than the theoretical mean. © 1999 British These values of A+ averaged over a quarter ICES Ecological Society the use o f a m ore refined weighting to reflect the quan­ Journal o f Animal titative reduction in taxon richness on moving up the rectangle showed consistent trends in taxonomic dis­ Ecology, 68, 769-782 hierarchy seems logical, in fact the correlation between tinctness from the Eastern Central N orth Sea to the 11 11 7 7 5 (a) (b) S. I. Rogers, 10

K.R. Clarke & 9 J.D. Reynolds 10 8

7 A* x xxx- A 6 (T)

5 XX

4 X

X 3 X y X \ X 2 X X

1

0 10 11 12 5 6 9 10 11 12 A* A* 12 (c) ( d )

11 X' 70

i A+ XXX' X (co;; A+10 60

X X

50 #

40 10 11 12 10 11 12 A* A+ Fig.4. A comparison using all catch data between (a) taxonomic diversity A and taxonomic distinctness A*, (b) the root transformed form of A* against ordinary A*, (c) taxonomic distinctness A+ against ordinary A*, and (d) A+ using a>k° w eights against A+ using ordinary cok weights.

coastal waters west of the UK (Fig. 6). The Eastern family Gobiidae), Pleuronectiformes (flatfishes) and Central and South-eastern N orth Sea, and South-east­ Scorpaeniformes (gurnards, Triglidae). These groups ern Channel, supported assemblages in which average contained almost all the species that were identified taxonomic distinctness was less than the theoretical by Rogers etal. (1999) as numerically most abundant mean. Assemblages close to, and occasionally greater in the entire coastal demersal fauna, and which typi­ than the theoretical mean, occurred in the South-west­ fied coastal sectors or discriminated between them ern N orth Sea, along the south coast of England, in (Table 3). Only two elasmobranch orders were rep­ the Bristol Channel and in some parts of the western resented here, and these only by a few species. From Irish Sea (Fig. 6). a total of 14 orders of fish, only 10 were represented It is helpful when interpreting these regional differ­ in the N orth Sea, although the taxonom ic distinctness ences in taxonom ic distinctness to identify which fish of the South-western N orth Sea was increased by the taxa have contributed most to patterns in A.+ The presence there of greater numbers of more distantly num ber of species belonging to each of the 14 orders related cartilaginous species (Carcharhiniformes, of bottom -dwelling fish in the dataset were identified Rajiformes and Squaliformes) (Table 2). The bottom- (Table 2). In the Eastern Central and South-eastern dwelling fish fauna of the Channel was broadly similar N orth Sea, which were the least taxonom ically distinct to that of the N orth Sea in terms of the relative dis­ areas of all those surveyed, m ost species belonged to tribution of species am ongst phyla, but the increased © 1999 British Ecological Society the orders Gadiform es (cod-like fishes), Perciformes abundance of cartilaginous species (Rajiformes) and Journal o f Animal (bottom-dwelling or benthic fish particularly drag- species from the order Perciformes, together with Ecology, 68, 769-782 onets of the family Callionymidae and gobies of the members of a more southerly fauna (e.g. Zeiformes, < O 7 7 6 Z e u s fa b e r (L.)) increased the taxonom ic distinctness T a x o n o m ie of the region. This pattern continued into the western distinctness o f fish waters of the UK, where additional representatives communities of the southern fauna, for example the electric ray (Torpedinidae), mullet species (Mugiliformes), and sun fish (Tetradontiformes) were found. The most 10 diverse fauna, representing 14 orders of cartilaginous and bony fishes, was found in the Bristol Channel (T a b le 2 ).

9

Discussion

8 W e have found clear spatial patterns in the taxonomic 0 10 20 3 0 40 Number of species diversity of bottom-dwelling fish assemblages in the Fig. 5. The departure from the theoretical mean taxonomic coastal waters of NW Europe, at both regional and distinctness A+, and 95% confidence funnel, of all individual local scales. The reduction in average taxonom ic range samples of bottom-dwelling fish species calculated using between the western waters of the UK and the sou­ presence/absence data. All values of A'1' should fall within thern N orth Sea m irrors similar patterns in the num­ the confidence funnel assuming a null hypothesis that each ber of taxa represented, and the relative abundance of sample contains species randomly selected from the total species list. the species which com prise them (Fig. 6; Rogers etal. 1999). A lthough this suggests that taxonom ic diversity and traditional species diversity may indeed be posi­ tively correlated within this limited latitudinal range,

© 1999 British Fig.6.Taxonomic distinctness A+ for samples grouped by quarter ICES rectangles. Filled circles show areas with mean values Ecological Society of A+ above the theoretical mean, and open circles identify those regions with mean values below the theoretical mean. Journal o f Animal Increasing bubble size reflects the magnitude of departure from the mean. E cology, 68, 769-782 Ill Table 2. The mean number of species of demersal fish that belong to the same order, in each of the nine sectors sampled S. I. Rogers, Secto r K.R. Clarke & O rder 1 2 3 4 5 6 7 8 9 J.D. Reynolds R ajiform es 3.1 2.0 3.5 3.1 2.4 1.3 1.5 0.2 0.1 Torpediniformes 0.1 0.2 0.2 0.6 0 0 0 0 0 Squaliformes 0.1 0 0 0 0 0 0.1 0 0 Carcharhiniformes 2.1 3.2 3.8 1.8 1.0 0.5 1.3 0.1 0.1 Anguilliformes 0.4 0.8 1.2 1.1 0.2 0 0.3 0.2 0.1 Gadiiformes 4.7 0.2 0.5 0.3 2.3 1.8 3.6 2.7 2.9 Lophiiformes 0.6 0 0 0 0 0 0 0 0.1 Gasterosteiformes 0.4 0.3 0.5 0.2 0.3 0.5 0.2 0.3 0.1 Perciformes 3.5 4.4 7.2 5.3 3.8 4.2 2.3 2.6 1.9 Pleuronectiformes 8.1 1.1 1.8 1.6 6.5 7.6 6.8 7.4 7.3 Scorpaeniformes 2.8 0 0 0.1 3.1 3.5 3.5 3.1 3.0 Z eiform es 0.4 0 0 0 0.3 0.2 0 0 0 Tetradontiformes 0.1 0 0 0 0 0 0 0 0 Mugiliformes 0.1 1.7 2.2 2.5 0 0 0 0 0

as suggested by Haii & Greenstreet (1998) for the and flatfish (Pauly 1994) - there are some exceptions. N orthern N orth Sea, the techniques used to calculate For example, the deep w ater fish fauna of the rise and these two types of index quantify different aspects of abyssal regions of the open oceans are m ore influenced biodiversity. This observation m ust therefore be a real by the tem perature difference throughout their depth effect rather than an artefact of the methodology. range than by the latitude, and so for this group, Records of samples which were taxonomically broad species diversity declines w ith increasing depth (W hite but had highly dom inated species abundances (i.e. low 1987). In the North-east Atlantic, the warm water taxonom ic diversity) (Fig. 4a), highlight the additional fauna of the M editerranean and Lusitanean regions degree of resolution that can be achieved when cal­ reaches its northern boundary on the continental shelf culating these new indices on individual rather than to the west of the British Isles, close to the English regionally aggregated samples. It also shows that cor­ Channel (Ekm an 1967). In addition to the native fish relations of the type described by Haii & Greenstreet fauna, there are also representatives of the m ore sou­ (1998) m ay only apply at relatively large geographical therly Lusitanean fauna, particularly those from the sc a le s. order Perciformes, such as the Sparidae, M ullidae and These new indices incorporate inform ation on taxo­ Labridae, but also species from the families Zeidae, nomic relationships into diversity measures, and also Molidae, and Mugillidae (Table 3). In most of the support a statistical testing framework. They are N orth Sea, there are two species groupings delimited therefore a valuable additional tool for interpreting by the boundary between stratified and mixed water assemblage structure. It is reasonable to assume that masses approxim ately at the 40 m depth contour, and the observed patterns in taxonomic diversity and dis­ in much of the region, gadoids (cod, Gadus morhua tinctness are: (i) the inevitable consequence of the (L.), whiting, Merlangius merlangus (L.) and pleuro- regional patterns in natural processes, or (ii) a result nectids (dab, Limanda limanda (L.)) predominate of natural or man-made disturbance at a local scale, ( D a a n e ta l. 1 9 9 0 ). or (iii) a com bination of both. The role of local environmental disturbance and productivity is particularly im portant in the southern N orth Sea, which, unlike other regions that we exam­ NATURAL PROCESSES ined, is relatively shallow and supports an extensive A lthough not universally true, m any taxa show a clear and homogeneous sandy substrate. The benthic fauna decrease in their species diversity from the tropics to of this region is dom inated by annelids, molluscs and the poles (Pianka 1966; Ekman 1967). One expla­ crustaceans, and it is maintained in an early stage of nation for this trend is that the tropics are larger in succession partly by the influence of storm events and area than temperate or tundra zones, and experience tides. In addition, the nutrient-rich fresh water from relatively high and uniform annual mean tem­ the Rivers Rhine and Elbe which flow into the Germ an peratures (Rosenzweig 1992). In marine and fresh­ Bight, have created extensive eutrophication in w ater systems, the largest num ber of fish species occur inshore waters which has contributed to these patterns in the tropics with progressively fewer towards polar (Hickel, M angelsdorf & Berg 1993). The absence of C 1999 British areas (Nelson 1994). W hile this rule can be applied extensive reef-forming colonial invertebrates, and the Ecological Society Journal o f Animal universally to groups that occupy the continental varied fauna which can become associated with these Ecology, 68, 769-782 shelf - e.g. coral reef fishes (Findley & Findley 1985) biogenic structures, has reduced the range of niches 778 Table 3. The bottom-dwelling fish taxa caught during the beam trawl surveys, with their taxonomic hierarchy. Those species shown by Rogers et al. (1998a) to comprise 0.4% or more of the total numerical abundance of the bottom-dwelling fish catch T a x o n o m ie are marked by *. Amongst this group, those which they additionally found to be consistent in typifying fish communities distinctness offish within sectors, or consistent in discriminating between them, are shown by ** communities CLASS ORDER FAMILY

Chondrichthys R ajiform es RAJIDAE b lo n d e ray Raja brachyura thornback ray Raja clavata* small-eyed ray Raja microocellata spotted ray Raja montagui starry ray Raja radiata cu ck o o ray Raja naevus undulate ray Raja undulata Torpediniformes TORPEDINIDAE marbled electric ray Torpedo marmorata electric ray Torpedo nobiliana Squaliformes SQUALIDAE sp u rd o g Squalus acanthias Carcharhiniformes SCYLIORHINIDAE lesser spotted dogfish Scyliorhinus canicula* nurse hound Scyliorhinus stellaris CARCHARIIDAE to p e shark Galeorhinus galeus TRIAKIDAE smooth hound Mustelus mustelus starry smooth hound Mustelus asterias T eleostei Anguilliformes ANGUILLIDAE eel Anguilla anguilla CONGRIDAE C o n g er eel Conger conger G ad ifo rm es GADIDAE cod Gadus morhua* h ad d o ck Melanogrammus aeglefinus w hiting Merlangius merlangus* p o o r cod Trisopterus minutus* bib Trisopterus luscus* pollack Pollachius pollachius common ling Molva molva PHYCIDAE five-bearded rockling Ciliata mustela four-bearded rockling Enchelyopus cimbrius shore rockling Gaidropsarus mediterraneus three-bearded rockling Gaidropsarus vulgaris rockling species Gaidropsarus spp. RANICIPITIDAE lesser forkbeard Raniceps raninus MACROURIDAE roundhead rat-tail Coryphaenoides rypestris MERLUCCIIDAE hake Merluccius merluccius Lophiiformes LOPHIIDAE angler Lophius piscatorius Gasterosteiformes GASTEROSTEIDAE stickleback Gasterosteus aculeatus SYNGNATHIDAE greater pipefish Syngnathus acus Nilsson’s pipefish Syngnathus rostellatus snake pipefish Entelurus aequoreus sea horse Hippocampus ramulosus short snouted seahorse Hippocampus hippocampus CAPROIDAE boar-fish Capros aper Perciformes SPARIDAE black sea-bream Spondyliosoma cantharus MULLIDAE red m ullet Mullus surmuletus CEPOLIDAE red bandfish Cepola rubescens PERCICHTHYIDAE sea bass Dicentrarchus labrax bass family Dicentrarchus spp. LABRIDAE cuckoo wrasse Labrus mixtus ballan wrasse Labrus bergylta rock cook Centrolabrus exoletus goldsinny Ctenolabrus rupestris corkwing wrasse Crenilabrus melops ZOARCIDAE eelp o u t Zoarces viviparus PHOLIDAE butterfish Pholis gunellus STICHAEIDAE Yarrels blenny Chirolophis ascanii snake blenny Lumpenus lampretaeformis TRACHINIDAE greater weever Trachinus draco lesser weever Echiichthys vipera** BLENNIIDAE butterfly blenny Blennius ocellaris tompot blenny Blennius gattorugine GOBIESOCIDAE two-spotted clingfish Diplecogaster bimaculata clingfish family Gobiesocidae © 1999 British CALLIONYMIIDAE dragonet family Callionymiidae** Ecological Society Journal o f Animal Ecology, 68, 769-782 779 Table 3.— continued. S. I. Rogers, CLASS ORDER FAMILY K.R. Clarke &

J.D. Reynolds GOBIIDAE rock goby Gobius paganellus black g o b y Gobius niger Stevens goby Gobius gasteveni Fries’s goby Lesuerigobius friesii goby family G obiidae* Pleuronectiformes SOLEIDAE sole Solea solea** solenette Buglossidium luteum* san d sole Pegusa lascaris thick back sole Microchirus variegatus* PLEURONECTIDAE plaice Pleuronectes platessa** flo u n d er Platichthys flesus d a b Limanda limanda** lem on sole Microstomus kitt* w itch Glyptocephalus cynoglossus long rough dab Hippoglossoides platessoides' h alib u t Hippoglossus hippoglossus SCOPHTHALMIDAE brill Scophthalmus rhombus tu rb o t Scophthalmus maximus m egrim Lepidorhombus whiffiagonis Ekstroms topknot Phrynorhombus regius Norwegian topknot Phrynorhombus norvegicus to p k n o t Zeugopteius punctatus BOTHIDAE scaldfish Arnoglossus laterna* Imperial scaldfish Arnoglossus imperialis Scorpaeniformes TRIGLIDAE tub gurnard Trigla lucerna* red gurnard Aspitrigla cuculus* grey gurnard Eutriglia gurnardus* streaked gurnard Trigloporus lastoviza COTTIDAE b u ll-ro u t Myoxocephalus scorpius sea scorpion Taurulus bubalis AGONIDAE h o o k -n o se Agonus cataphractus** CYCLOPTERIDAE lump sucker Cyclopterus lumpus LIPARIDAE sea snail Liparis liparis sea snail genus Liparis spp. Zeiform es ZEIDAE Jo h n d o ry Z e u s fa b e r Tetradontiformes MOLIDAE sunfish M o la m ola Mugiliformes MUGILIDAE thick lipped mullet Chelon labrosus grey mullet family m ugilidae

1 23456789

© 1999 British West UK Channel North Sea Ecological Society Fig. 7. Taxonomic distinctness A+ of the nine coastal sectors using presence/absence data (solid bars) (shown also in Fig. 3d), Journal o f Animal showing the effect of removing all elasmobranch species (tinted bars), and all flatfish (pleuronectidae) (open bars) on the Ecology, 68, 769-782 calculation of the index. 7 8 0 which the region can support. This in turn has limited long-term effects of bottom trawling throughout the T a x o n o m ie the range of fish species which can use these environ­ area (Quero & Cendrero 1996). O f all the families distinctness o f fish ments for food and shelter, such as the herbivorous which com prise the total fish assemblage therefore the communities mullets (M ugillidae) (Daan e ta l. 1990). elasmobranchs are the most likely to show effects of com m ercial activity, if it is present. To what extent do the elasmobranchs influence HUMAN IMPACTS the spatial patterns in taxonom ic distinctness that we The N orth Sea has also experienced extensive m anipu­ identified in Fig. 6? To investigate this further, values lation of its fish communities by commercial fishing of A+ were recalculated for each sector, but using a activity, and the extent of this disturbance, and its revised dataset in which all elasmobranchs, i.e. mem­ potential impact on the bottom-dwelling fish assem­ bers of the Chondrichthyes (Table 3)) were assumed blage, deserves particular attention. As one may to be absent from catches. The result of this analysis expect from a consistently high level o f fishing activity, describes the taxonomic distinctness of the bottom- the most obvious effects of the removal of fish from dwelling, non-elasmobranch fishes, and should the higher trophic levels have been on the temporal exclude the potentially confounding effects of hum an changes in size structure of communities (Pope 1989; activity on this group to reveal the underlying bio­ Greenstreet & Haii 1996; Rice & Gislason 1996; Rice geographic pattern. Differences in mean A+ between & Kronlund 1997). The reduction in the abundance sectors using this revised dataset were less significant of larger fish (Pope & Knights 1982; Rice & Gislason (F statistic = 8-4) than when previously calculated 1996; Jennings, Reynolds & Mills 1998) has caused using the full dataset (F = 42-9) (Fig. 7). In particular, changes in the dominance of the smaller target and mean taxonomic distinctness in all sectors of the non-target components of the assemblage, as well as N orth Sea appeared only marginally less than those the species com position of the catch. A lthough several west of the UK (Fig. 7). This is not simply a scale studies undertaken in tropical waters show that locally effect in which a reduced dataset leads to reduced intense fishing activity can cause reductions in species discrimination. W hen the calculations were repeated richness, there are few examples of this effect in tem­ but with all the flatfish (Pleuronectiform es, a similar­ perate waters (Greenstreet & Haii 1996; Rice & Gis­ sized group) excluded from the dataset instead of the lason 1996); and see review by (Jennings & Kaiser elasmobranchs, the effect was barely detectable (Fig. 1998). Is there any evidence that anthropogenic 7). M ean values of A+ retained the same general activity has contributed to the lower taxonom ic range relationships between sectors, and the differences of populations in the southern N orth Sea, compared between sector means were still relatively large to elsewhere on the continental shelf? (F=31.7). This analysis confirms that the elasmo­ There are some examples of commercial groundfish branchs have a m ajor influence on the taxonomic dis­ species in the North-east Atlantic disappearing from tinctness of the North-east Atlantic fauna, and that large parts of their former range (Heessen & Daan when they are artifically excluded from the samples, 1996; Rijnsdorp e ta l. 1996; W alker & Heessen 1996). the remaining species have a m ore uniform taxonomic Perhaps one of the best known involves the common range throughout the region. In such a widely dis­ s k a te , R a ja b a tis (L.), (Brander 1981). The com­ tributed bottom-dwelling group, this effect is surpris­ bination of late age at maturity (11 years) large ing. One possible explanation is that the adverse maximum size (>2m ) and a low potential rate of effects of high fishing m ortality have influenced the population increase, are thought to have been respon­ distribution and abundance of some elasmobranch sible for its decline. A similar decline has recently species to create an artificial pattern. Alternatively, it been documented for the barndoor skate Raja laevis may result from different levels of fishing activity (Mitchill) (Casey & Myers 1998). The presence of within coastal waters. Further analysis of time-series these life-history characteristics was also directly changes in the distribution and abundance of elasmo­ linked to decreasing trends in abundance of 18 inten­ branchs, their life-history characteristics and their pat­ sively exploited fish stocks, after accounting for terns of exploitation, will be necessary to confirm the differences in their fishing mortality (Jennings e ta l. extent of the role of high fishing effort in the decline 1998). The elasmobranchs have some or all of these of some of these species. characteristics, and dogfish, and several species of ray In the event of perturbations caused by either (Squaliformes and Carcharhiniformes), which gen­ hum an im pact or changing environmental conditions, erally have a widespread distribution throughout the Tilman (1996) claims that the taxonomic range of North-east Atlantic, are exploited by commercial an assemblage will be im portant for maintaining the fisheries (Holden 1974; Quero & Cendrero 1996; ICES stability of the ecosystem. In some m arine invertebrate 1997; W alker, Howlett & M illner 1997). In the Bay of communities there is considerable redundancy at the Biscay, the long-term decline in abundance of three species level, such that similar patterns in structure 0 1999 British species o f ray (R a ja b a tis (L .), Raja brachyura (L a f o n t) can be identified at a higher taxonomic level. This Ecological Society Journal o f Animal a n d Raja clavata (L.)), and other species of large suggests that there may also be several species which Ecology, 68, 769-782 bodied carcharhiniform sharks, was attributed to the perform the same functional role (Ferraro & Cole 781 1992; Clarke & W arwick 1998a). M arine fish com­ tinctness and diversity measures: responses in marine fish communities. Marine Ecology Progress in Series, 1 6 6 ,2 2 7 - S.I. Rogers, munities with low levels of species redundancy and 229. K.R. Clarke & where few species fulfil key functional roles, for exam ­ Heessen, H.J.L. & Daan, N. (1996) Long-term trends in ten J.D. Reynolds ple some coral reef ecosystems, may be least able to non-target N orth Sea fish species. IC ES Journal o f Marine withstand environmental change (Hughes 1994; Jen­ Science , 53, 1060-1078. nings & Kaiser 1998). These indices of taxonomic Hickel, W., Mangelsdorf, P. & Berg, J. (1993) The human y I impact in the German Bight: eutrophication during three diversity and distinctness can therefore provide valu­ decades (1962-91). Helgoländer Meeresunters, AÍ, 2 4 3 - able inform ation on the extent to which fish assem­ 263. blages are able to resist change, and this will become Holden, M.J. (1974) Problems in the rational exploitation of increasingly relevant in our interpretation of changes elasmobranch populations and some suggested solutions. in assemblage structure. Sea Fisheries Research (ed. F.R. Harden Jones), pp. 117- 137. Elek Science, London. Hughes, T.P. (1994) Catastrophes, phase shifts and large- scale degradation of a Caribbean reef. Science, 265, 1547— Acknowledgements 1551.ICES. (1997) Report o f the Study Group on Elasmo­ branch Fishes IC E S C M 1997/G: 2. The authors wish to thank Uii Damm (Germany), Jennings, S. & Kaiser, M .J. (1998) The effects of fishing on Adriaan Rijnsdorp (The Netherlands), Willy Vanhee marine ecosystems. Advances in Marine Biology, 34, 2 0 1 - (Belgium), and their colleagues, for making these 352. international data available for this analysis. Nicholas Jennings, S., Reynolds, J.D. & Mills, S.D. (1998) Life history Dulvy and Suzanne Mills kindly helped to compile correlates of responses to fisheries exploitation. Pro­ ceedings o f the Royal Society of, London, 265, 333-339. the taxonomy. This work was supported by M OU(A) Ludwig, J.A. & Reynolds, J.F. (1988) Statistical Ecology: a of the M inistry of Agriculture, Fisheries and Food, ' w ' • /wj\ Primer on Methods and Computing. John Wiley and Sons, and the participation of K.R.C. was funded jointly by New York. M A FF (AE1113) and the NERC M arine Biodiversity M agurran, A.E. (1988) Ecological Diversity and its Measure­ programme at CCM S Plymouth M arine Laboratory. m ent. Chapman & Haii. London. McEachran, J.D. & Miyake, T. (1990) Phylogenetic inter­ W e thank Simon Greenstreet and an anonymous ref­ relationships of skates: a working hypothesis (Chon- eree for helpful comments. drichthyes, Rajoidei). Elasmobranchs as Living Resources: Advances in the Biology, Ecology, Systematics, and the Status o f the Fisheries (eds H.L. Pratt, S.H. Gruber & T. References Taniuchi), pp. 285-304. NOAA Technical Report NMFS 90. Baltz, D.M . (1991) Introduced fishes in marine systems and McEachran, J.D., Dunn, K.A. & Miyake, T. (1996) Inter­ inland seas. Biological Conservation, 56, 151-177. relationships of the Batoid fishes (Chondrichthyes: Bato­ Brander, K. (1981) Disappearance of common skate Raia idea). Interrelationships o f Fishes (eds M.L.J. Stiassny, batis from the Irish Sea. N ature, 290,48-49. L.R. Parenti & G.D. Johnston), pp. 63-84. Academic Casey, J.M . & Myers, R.A. (1998) Near extinction of a large, Press, San Diego. widely distributed fish. Science, 281, 690-692. Nelson, J.S. (1994) Fishes o f the World, 3rd edn. John Wiley Clarke, K.R. & Warwick, R.M. (1998a) Quantifying struc­ & Sons, New York. tural redundancy in ecological communities. Oecologia, Pauly, D. (1994) A framework for latitudinal comparisons 113,278-289. of flatfish recruitment. Netherlands Journal of Sea Clarke, K.R. & Warwick, R.M. (1998b) A taxonomic dis­ Research, 32, 107-118. tinctness index and its statistical properties. Jo u rn a l o f Pearson, T.H. & Rosenberg, R. (1978) Macrobenthic suc­ Applied Ecology, 35, 523-531. cession in relation to organic enrichment and pollution of Connell, J.H. (1978) Diversity in tropical rain forests. Y the marine environment. Oceanography and Marine Science, 199, 1302-1310. Biology, Annual Review, 16,229-311. Daan, N., Bromley, P.J., Hislop, J.R.G. & Nielsen, N.A. Petraitis, P.S., Latham, R.E. & Niesenbaum, R.A. (1989) (1990) Ecology of North Sea fish. Netherlands Journal o f w The maintenance of species diversity by disturbance. Quar­ Sea Research, 26,343-386. terly Review o f Biology, 64,393-418. Ekman, S. (1967) Zoogeography o f the Sea. Sidgwick and Pianka, E.R. (1966) Latitudinal gradients in species diversity: Jackson, London Y a review of concepts. American Naturalist, 100, 33-46. Ferraro, S.P. & Cole, F.A. (1992) Taxonomic level sufficient Pope, J.G. & Knights, B.J. (1982) Comparison of the length for assessing a moderate impact on macrobenthic com­ distributions of combined catches of all demersal fishes in y munities in Puget Sound, Washington, USA. C anadian surveys in the North Sea and at Faroe Bank. Canadian Journal o f Fisheries and Aquatic Science, 49, 1184-1188. Special Publication o f Fisheries and Aquatic Sciences, Vol Findley, J.S. & Findley, M.T. (1985) A search for pattern in 59 (ed. M.C. Mercer), pp. 116-118. Government of butterfly fish communities. American Naturalist, 126,8 0 0 - Canada, Ottawa. 816. Pope, J.G. ( 1989) Fisheries research and management for the Gee, J.M . & Warwick, R.M. (1996) A study of global bio­ North Sea: the next hundred years. Dana, 8, 33-43. diversity patterns in the marine motile fauna of hard sub­ Quero, J.-C. & Cendrero, O. (1996) Incidence de la pêche s tra ta . Journal o f the Marine Biological Association, UK , sur la biodiversité ichtyologique marine: Le bassin d’Ar- 76,177-184. cachon et le plateau continental sud Gascogne. C ybium, Greenstreet, S.P.R. & Haii, S.J. (1996) Fishing and the 20, 323-356. © 1999 British ground-fish assemblage structure in the north-western Rice, J. & Gislason, H. (1996) Patterns of change in the Ecological Society ^ N orth Sea: an analysis of long-term and spatial trends. size spectra of numbers and diversity of the North Sea Journal o f Animal Journal o f Animal Ecology, 65, 577-598. assemblage, as reflected in surveys and models. IC E S Jour­ Ecology, 68, 769-782 Haii, S.J. & Greenstreet, S.P.R. (1998) Taxonomic dis­ nal o f Marine Science, 53, 1214-1225. C 782 Rice, J.C. & Kronlund, A.R. (1997) Community analysis fields: the nonequilibrium maintenance of species diversity. v and flatfish: diagnostic patterns, processes, and inference. Ecology, 60, 1225-1239. T a x o n o m ie Journal o f Sea Research, 37,301-320. Tilman, D. (1996) Biodiversity: population versus ecosystem distinctness o f fish Rijnsdorp, A.D., Van Leewen, P.I., Daan, N. & Heessen, stability. E cology, 77,350-363. communities H.J.L. (1996) Changes in abundance of demersal fish spec- Walker, P., Howlett, G. & Millner, R. (1997) Distribution, )0 ies in the North Sea between 1906 and 1909 and 1990-95. movement and stock structure of three ray species in the X ICES Journal o f Marine Science, 53, 1054-1062. North Sea and eastern Channel. ICES Journal o f Marine Rogers, S.I., Maxwell, D., Rijnsdorp, A.D., Damm, U. & Science, 54, 797-808. Vanhee, W. (1999) Fishing effects in northeast Atlantic Walker, P.A. & Heessen, H.J.L. (1996) Long-term changes Shelf Seas: patterns in fishing effort, diversity and com- in ray populations in the North Sea. ICES Journal of munity structure. IV. Can comparisons of species diversity Marine Science, 53, 1085-1093. be used to assess human impacts on coastal demersal fish Warwick, R.M. &Clarke, K.R. (1993) Comparing the sever­ faunas in the Northeast Atlantic? Fisheries Research, 40, ity of disturbance: a meta-analysis of marine macrobenthic 135-152. community data. Marine Ecology Progress in Series, 92, Rogers, S.I., Rijnsdorp, A.D., Damm, U. & Vanhee, W. 221-231. (1998) Demersal fish populations in the coastal waters of Warwick, R.M. & Clarke, K.R. (1995) New ‘biodiversity’sity’ ¿ y the UK and continental NW Europe from beam trawl measures reveal a decrease in taxonomic distinctness with y with f ? * ' survey data collected from 1990 to 1995. Journal o f Sea increasing stress. Marine Ecology Progress in Series, 129, Research, 37, 79-102. 301-305. Rosenzweig, M.L. (1992) Species diversity gradients: we Warwick, R.M. & Clarke, K.R. (1998) Taxonomic dis- know more and less than we thought. Journal ofMammo- tinctness and environmental assessment. Journal o f logy, 73,715-730. Applied Ecology, 35, 532-543. ^ Rosenzweig, M.L. (1995) Species Diversity in Space and White, B.N. (1987) Oceanic anoxic events and allopatric )' • Tim e. Cambridge University Press, Cambridge. spéciation in the deep sea. Biological Oceanography, 5, Somerfield, P.J., Olsgard, F. & Carr, K.R. (1997) A further 243-259. examination of two new taxonomic distinctness measures. Marine Ecology Progress in Series, 154,303-306. Sousa, W.P. (1979) Disturbance in marine intertidal boulder Received 22 June 1998; revision received 4 November 1998

£ 1999 British Ecological Society Journal o f Animal E cology, 68, 769-782