Ideal dominance distributions: a test using red-winged blackbirds (Agelaius phoeniceus)

Summary The ideal dominance distribution model predicts that competition between individuals of a species for territories will result in socially dominant individuals acquiring territories in higher quality habitat than their subordinates. Although the dispersion and relative reproductive success of male red-winged blackbirds (Agelaius phoeniceus) across habitats in eastern Ontario appears to conform to the ideal dominance distribution model, data from a study of three captive groups comprised of males from both high (marsh) and low (upland field) quality habitats failed to support the prediction that males from marsh habitat are dominant to those from upland habitat. Contrary to the prediction males from uplands were generally dominant to males from marshes. We found a significant positive correlation between dominance and both increased epaulet size and increased body size. Controlling for these positive effects, upland males remained generally dominant to marsh males. Measurements of independent samples of males from both habitats indicated that the overall distribution of males does not conform to ideal dominance. We suggest that the strong between-year territory fidelity shown by male red-winged blackbirds and chance events when they initially acquire territories may contribute to this lack of conformity.     

Experimentally induced polygyny in a monogamous bird species: prothonotary warblers and the polygyny threshold

Summary Hypotheses regarding evolution of polygyny were tested using prothonotary warblers (Protonotaria citrea), through experimental manipulations of habitat quality. Two experiments were performed. Experiment 1 involved supplying nest boxes in different breeding habitats: flooded riparian areas (high food availability) and adjacent dry bottomland (low food availability). Nine bigamous matings were induced in this experiment, all occurring in flooded habitats even though monogamous mating opportunities existed in dry areas. Costs of polygyny for secondary females in flooded habitat were similar to costs incurred by monogamous females in dry habitat. Thus, a polygyny threshold apparently existed, possibly based on differences in food abundance in the two habitats. However, male quality covaried with habitat quality, as males in flooded areas were older and larger than males in dry habitat. Secondary females may have chosen polygyny in the best breeding situation, a combination of male and territory quality. In Experiment 2 nest box density was varied within flooded habitat to allow males to monopolize different numbers of nest-sites. Monogamous females settled earlier on territories with a large number of nest-sites, and males that defended many nest-sites were more likely to become polygynous. Male physical characteristics were not related to occurrence of polygyny in flooded habitat. Limitation of suitable nest-holes ultimately constrains occurrence of polygyny in prothonotary warblers.     

Satellites and sneakers: opportunistic male breeding tactics in pupfish (Cyprinodon pecosensis)

Summary Manipulations of population density, availability of oviposition substrate, and size of breeding habitat affected the proportion of pupfish (Cyprinodon pecosensis) males adopting territorial and satellite breeding tactics. Satellites occupied stations above breeding substrates defended by territorial males. Both territorial and satellite males developed a bright blue breeding coloration. A third breeding tactic, sneak-spawning, was occasionally observed. Sneak-spawning males retained the cryptic female coloration and occasionally spawned on territories. The lower reproductive success of satellites and sneak-spawning males suggested that both are conditional breeding tactics adopted by competitively inferior males. Satellites were present in all treatments that favored a territorial breeding system, but were absent when the breeding system was a dominance hierarchy. Satellites were associated with territorial males that had higher reproductive success and larger territories than males without satellites. Satellites functioned as parasites rather than as mutualists, since they disrupted spawnings and stole copulations from territorial males. The occurrence and frequency of the conditional breeding behaviours in pupfish represent facultative responses of males to changes in the intensity of competition for breeding sites and females.     

Seasonal variation in territorial response,and other aspects of the ecology of the Australian temperate pomacentrid fish Parma microlepis

The Australian temperate pomacentrid fish Parma microlepis Gunther maintains permanent, interspecifically defended territories. Breeding occurs in late spring (October to December), and juveniles after a pre-settling period of 4 to 6 weeks take up territories in the adult habitat. P. microlepis feed in their territories on benthic algae and invertebrates. Density of fish is significantly correlated with the availability of suitable shelter. There is a variation in the territorial response directed at different species, and a significant correlation exists between intensity of response to a species and its similarity to P. microlepis in use of living space. No significant correlation exists between intensity of response and dietary similarity. P. microlepis defends larger areas in the breeding than in the non-breeding season against those species which represent a possible threat to its breeding success. It shows a reduced tendency to attack juvenile conspecifics at the time of year when they would be settling from the plankton. Responses to species not likely to interfere with breeding success do not vary through the year. P. microlepis is thus seen to have complex territorial responses, specific to the species of intruder and time of year. In this it is quite similar to other territorial pomacentrids (chiefly tropical) for which information exists.     

Influence of territoriality on adult density in two rockfishes of the genus Sebastes

Nonterritorial Sebastes carnatus and S. chrysomelas existed, along with territorial individuals, at 3 tagging sites off southern California, USA, which were monitored for nearly 1 yr. To test the hypothesis that territoriality affected adult density in these species, territorial fish were removed and the subsequent utilization of vacated territories by other fish was monitored. Intrusion of other fish into vacated territories increased significantly in 90% of the removals. Other fish colonized both the feeding and sheltering parts of the vacated territories, indicating that the previous owners had successfully defended both parts of their territories. Many of the colonizers had already possessed territories; they expanded their territories or moved into presumably better havitat. Several previously-nonterritorial fish also moved into vacated areas, and at least some of them appeared to establish territories. These fish, then, had previously been capable of establishing territories, but were prevented from doing so by resident territory holders. Thus territoriality, rather than such other factors as predation or low recruitment, limited the number of territorial fish at each site. However, territorial fish did not inhibit the settlement of larval recruits, and the relative mortality rates of older territorial vs nonterritorial fish were not determined. Thus the question of whether territoriality was a major factor controlling total density remains unresolved.     

On the possible contribution of mixed species flocks to species richness in neotropical avifaunas

Summary Mixed species foraging flocks are a dominant component of the infra-structure of avian communities in neotropical forests. In Amazonia, these flocks consist of pairs of 10–20 species, many of which are permanently associated with mixed flocks. At least half of these flocking species maintain territories that correspond exactly to the flock home range. Small individuals that participate as permanent members of the flocks must adopt the large home range of the larger nucleus species. Therefore, the densities of smaller species are dependent on the availability and density of flocks rather than the availability of food resources. Single pairs of 4 small flocking species with individual body masses of 8 g occupied exclusive territories of 8–12 ha. These were the same exact territories that were defended by at least 6 other flocking species with individual body masses of up to 37 g. Because of their attachment to flocks with large territories, small species are expected to under-utilize available food resources. The under-utilization of food resources is expected to allow smaller species to coexist with greater niche overlap resulting in increased species richness. This hypothesis was tested by quantifying foraging niche in terms of foraging height, foraging maneuver, and prey substrate; and using these values in addition to body mass and bill size (length, depth and width) to determine relative niche overlap between large versus small species pairs.Smaller species had greater foraging overlap than large flocking species and particularly the three smallest species of the genus Myrmotherula; longipennis, axillaris and menetriesii had very high overlap (average foraging niche overlap for the 3 species=0.83±0.12 compared with 0.12±0.19 for all flocking species), similar body sizes (body masses differing by no more then 8%) and similar bill morphologies (maximum ratio in length=1.08, width=1.07, and depth=1.06). These results are consistent with the hypothesis that small species participating in Amazonian mixed flocks can coexist with greater niche overlap because their density is flock dependent rather than resource dependent.     

Coupling ecology and GIS to evaluate efficacy of marine protected areas in Hawaii.

In order to properly determine the efficacy of marine protected areas (MPAs), a seascape perspective that integrates ecosystem elements at the appropriate ecological scale is necessary. Over the past four decades, Hawaii has developed a system of 11 Marine Life Conservation Districts (MLCDs) to conserve and replenish marine resources around the state. Initially established to provide opportunities for public interaction with the marine environment, these MLCDs vary in size, habitat quality, and management regimes, providing an excellent opportunity to test hypotheses concerning MPA design and function using multiple discrete sampling units. Digital benthic habitat maps for all MLCDs and adjacent habitats were used to evaluate the efficacy of existing MLCDs using a spatially explicit stratified random sampling design. Analysis of benthic cover validated the a priori classification of habitat types and provided justification for using these habitat strata to conduct stratified random sampling and analyses of fish habitat utilization patterns. Results showed that a number of fish assemblage characteristics (e.g., species richness, biomass, diversity) vary among habitat types, but were significantly higher in MLCDs compared with adjacent fished areas across all habitat types. Overall fish biomass was 2.6 times greater in the MLCDs compared to open areas. In addition, apex predators and other species were more abundant and larger in the MLCDs, illustrating the effectiveness of these closures in conserving fish populations within their boundaries. Habitat type, protected area size, and level of protection from fishing were all important determinates of MLCD effectiveness with respect to their associated fish assemblages. Although size of these protected areas was positively correlated with a number of fish assemblage characteristics, all appear too small to have any measurable influence on the adjacent fished areas. These protected areas were not designed for biodiversity conservation or fisheries enhancement yet still provide varying degrees of protection for fish populations within their boundaries. Implementing this type of biogeographic process, using remote sensing technology and sampling across the range of habitats present within the seascape, provides a robust evaluation of existing MPAs and can help to define ecologically relevant boundaries for future MPA design in a range of locations.     

Intrasexual competition,territoriality and acoustic communication in male strawberry poison frogs (<Emphasis Type="Italic">Oophaga pumilio</Emphasis>)

In many species male reproductive success is limited by access to females. Territoriality is one behavioural strategy which helps to acquire females. In the present study, we investigated the correlation between territory size and (1) female availability and (2) rate of intrusion by conspecific males in strawberry poison frogs, Oophaga pumilio. Males defended smaller territories in areas with a high female density and high rate of intrusion by conspecific males. Only males with high body condition values were able to establish territories in areas of high female density probably due to better fighting abilities. Moreover, dominant calling frequency was lower during agonistic interactions. Because only males with high body condition values were able to produce very low dominant frequencies, the acoustic properties appear to be an honest signal containing information about fighting abilities. Thus, the negative correlation between dominant frequency and mating success of males found in previous studies seems to be in part the result of intrasexual competition between males.     

Territorial behavior of the black and yellow rockfish and gopher rockfish (Scorpaenidae,Sebastes)

Tagged black-and-yellow rockfish Sebastes chrysomelas and gopher rockfish S. carnatus were observed in situ for 1 yr at 3 study sites off California, USA in order to determine the nature of their home ranges and territories. Three basic patterns of space use were identified: (1) a shelter hole located within a larger area of activity (in which the fish presumably fed, since its size increased with size of fish and decreased with prey density); (2) discontinuous home ranges, in which the shelter holes and feeding areas were separte; and (3) dispersed home ranges with little or no exclusive area. Fish in the first category were regarded as territorial, since their home ranges were defended and relatively exclusive. Because individuals did not patrol borders, however, territories often overlapped at their peripheries. In addition, some territory holders appeared to tolerate the sheltering of other fish within their territories, as long as these intruders left the territory during periods of activity. Some fish with discontinuous home ranges (commuters) may have been territorial, but others may have been forced to occupy suboptimal feeding areas. Fish in the third category (floaters) were regarded as nonterritorial; their activities were apparently limited to areas where they could escape the aggression of territory holders. The varying patterns appeared to develop through defense and accommodation during the turnover of individuals, with each fish securing shelter and feeding areas as these became available.     

Correlates of male mating success in two species of Anthidium bees (Hymenoptera : Megachilidae)

Summary Field observations were made on the mating behavior of two congeneric species of solitary bees, Anthidium porterae in an arid grassland and A. palliventre in a coastal sand dune habitat. Males of both species exhibited resource defence polygyny and defended hostplants to gain access to females foraging for nectar and pollen. The mating frequencies of marked and measured resident (territorial) males were monitored during periods of continuous observation, following which measurements of territory size and floral resources were obtained. Mating success of A. palliventre males was strongly influenced by territory characteristics: Males that defended small areas with a few rich hostplant patches mated more often than males that held larger territories containing many hostplant patches of low floral density. Large males generally held high-quality sites and thus had a mating advantage over smaller individuals. In A. porterae, on the other hand, male mating success was unrelated to any measure of territory quality. Copulation frequency and male size were positively correlated, however, apparently due to the increased ability of large males to seize and hold females for mating. The two species also differed in the incidence of non-territorial, sneaky males. While absent in A. palliventre, sneaky males accounted for 12% of all mating observed in A. porterae. Males of A. porterae that displayed sneaky tactics mated, on average, as often as resident males. Offprint requests to: E.M. Villalobos     

Age-specific social dominance affects habitat use by breeding American redstarts (Setophaga ruticilla): a removal experiment

Summary By removing older males from their breeding territories, we tested the hypothesis that age-related dominance behavior influenced the pattern of habitat selection by breeding American redstarts Setophaga ruticilla (Aves: Parulinae). Fifteen older male redstarts removed in five experimental replicates during three breeding seasons were replaced by ten yearling and five older males; no redstart males of either age colonized the control sites during the same time periods, although two yearlings disappeared. Significantly more yearling males (67%, n=9) colonized the vacated areas than were present in the redstart population at large (26.8%, n=209). We reject the alternative hypothesis that yearling male redstarts occupy different habitats from older males because of age-related (innate) habitat preferences. Redstarts that colonized the territories made vacant by our removals (i.e., floaters) were a behaviorally heterogeneous group of animals. The presence of both yearling and older male floaters indicates that suitable habitat is limiting for this species and that intraspecific competitive interactions are important in habitat distribution, and potentially in population regulation.     

Why is the coral-reef fish Valenciennea strigata (Gobiidae) monogamous?

Parental-investment theory predicts that monogamy should be rare, and empirical evidence supports this prediction. Monogamy has generally been explained by either (1) a need for biparental care, or (2) a uniform distribution of limiting resources. By contrast, monogamy has evolved in several coral-reef fishes without biparental care, and many coral-reef fishes may not be limited by resources. Monogamy in these fishes might be explained by either (3) a low population density that favors mate fidelity, or (4) an abundance of resources that allows all males to breed and thus lowers the cost-benefit ratio for females that mate guard. We tested predictions of these hypotheses in the coral-reef fish, Valenciennea strigata, except that (1) biparental care had previously been rejected. We found no evidence of resource limitation (2a): food densities within territories did not differ from the density of food in unused habitat adjacent to territories; potential nest sites also existed in this unused habitat. Similarly, (2b) if resources limit the population, territory defense might require the coordinated efforts of a pair; however, widowed fish maintained their territories and (3) remated rapidly. Finally, (4) all males maintained a nest, both sexes enforced monogamy by mate guarding, and females benefited from guarding a high-quality (large) mate. Females mated to large males fed more than when they paired with small males. The results support the hypothesis that abundant and uniformly distributed resources can lead to monogamy by lowering the cost of guarding a mate when mate guarding provides benefits. Received: 25 November 1997 / Accepted after revision: 29 December 1997     

Lek structure and territoriality in the chryxus arctic butterfly,Oeneis chryxus (Satyridae)

Summary The territorial and mating systems of the satyrid butterfly Oeneis chryxus were studied for 3 years on individually marked populations in algonquin Provincial Park, Ontario. Males defended territories in a dry open forest clearing by means of chases and spiral flights with conspecifies. Defended areas did not contain food or oviposition sites required by females. Territories were not uniformly distributed over the study area; the highest density of territories was clustered over bare ground covered with sand and small stones rather than areas of vegetation. This pattern of distribution of territories occurred during each year of the study. The patterns of occupancy of territories showed a high degree of consistency from one day to the next, some males defending a given territory for as many as 11 consecutive days. Removal experiments showed that a surplus of non-territorial males did not exist. Vacant territories in the high-density area were absorbed by neighboring males; those in the low-density area remained empty. Male size was not correlated with territory position in the study are. Wing patterning (i.e. size of spots) differed between males occupying territories on sandy/stony versus vegetated areas. The territorial and mating systems of O. chryxus resemble lek polygyny, and are more similar to some vertebrate lek systems than to hilltopping or landmark territoriality.     

Estimating total abundance of a large temperate-reef fish using visual strip-transects

Total abundance estimates for the large, common, reef fish Cheilodactylus spectabilis (Hutton) were obtained for a marine reserve and adjacent section of coast in north-eastern New Zealand during 1985. Visual strip-transects were used to estimate abundance and size structure in both areas. The accuracy, precision and cost efficiency of five transect sizes (500, 375, 250, 100, 75 m²) were examined over three times per day (dawn, midday and dusk), by simulating transects over mapped C. spectabilis populations. Two transect sizes showed similarly high efficiency. The smaller of the two (20x5 m) was chosen for the survey because of the general advantages attributable to small sampling units. Biases related to strip-transect size are discussed. Preliminary sampling indicated that C. spectabilis was distributed heterogeneously, and that density was habitat-related. An optimal stratified-random design was employed in both locations, to obtain total abundance and size-structure estimates. This reduced the between-habitat source of variability in density. The total number of sampling units used was governed by the time available. The resulting total abundance estimates obtained were 18 338±2 886 (95% confidence limit) for the 5 km marine reserve, compared to 3 987±1 117 for an adjacent, heavily fished 4 km section of coast. When corrected for total area and habitat area sampled, this represented a 2.3-fold difference in abundance. If sampling had been designed to detect an arbitrary 10% difference in abundance within each habitat, an infeasible 440 h of sampling would have been required. Size-frequency distributions of C. spectabilis at the reserve had a larger model size class than distributions from the adjacent area. The data suggest that reserve status is causal in these differing abundance and size structure estimates.     

Ecological bases of philopatry and cooperation in Ethiopian wolves

Ecological reasons for philopatry and cooperation are frequently invoked when kin selection is an insufficient explanation. The Ethiopian wolf (Canis simensis) is a specialised rodent hunter that forms family groups with cooperative breeding but also lives as monogamous pairs in suboptimal areas. Given the apparent absence of fitness gains to helpers from cooperative breeding, we set out to explore the benefits accrued by communal territorial defence measured as the acquisition and retention of habitats with more and most preferred rodent prey. Pairs defended relatively large territories to encompass critical amounts of key habitats within a matrix poor in rodents. Groups in optimal areas had relatively small territories and were expansionist, such that wolves in larger packs benefited per capita from increased good-quality foraging habitat. The fitness benefits of philopatry became evident after a rabies epizootic, when philopatry and expansionism prevailed in under-saturated conditions, until large groups split or provided dispersers that established locally. This study shows that high concentrations of prey can shift the balance of costs and benefits towards group living and cooperation in long-lived territorial carnivores, in so far as this dictates immediate rewards accrued from a given increment in territory size, namely greater foraging area per animal, leading to group enlargement and eventual inheritance of breeding space.     

Geographic variation in social organization of Galápagos mockingbirds: ecological correlates of group territoriality and cooperative breeding

Summary To investigate ecological influences on cooperative social organization, I studied the four allopatric species of mockingbirds (Nesomimus spp.) endemic to the Galápagos archipelago on four islands. On three small, low and arid islands (Genovesa, Champion and Española), mockingbird territories filled all terrestrial habitat, mean group size varied from 4.5 to 14.2 adults, maximum group size ranged from seven to 24 birds, and 70–100% of groups contained more than two birds. San Cristóbal is larger and higher, and it supports a broader range of habitats. At one highland and two coastal sites on this island, mockingbirds did not hold territories in all available habitats, group size averaged 2.2 adults, only 25% of groups were larger than two, and none included more than three adults. Adults dispersed into vacant habitat to establish new territories only on San Cristóbal. Helping behavior has not yet been observed on San Cristóbal, but it occurs on the other three islands. These results support the hypothesis that social groups and cooperative breeding are maintained where limited availability of preferred habitat constrains dispersal. The mechanism relaxing habitat saturation on San Cristóbal, however, remains undetermined. Predation by introduced rats and cats may reduce survival and indirectly reduce group size; these predators are absent from Genovesa, Champion and Española. Differences in food supplies could also affect interand intra-island variation in population density. Variation in social organization among arid coastal sites on the four islands, and similarity between climatically different sites on San Cristóbal, suggest that climatic conditions are less important as determinants of dispersal and breeding. Skews in adult sex ratios also fail to account for inter-island variation in sociality. Although they live in a climatically variable environment, territorial behavior and the physical limits of suitable habitat have an overriding influence on cooperative social organization in Galápagos mockingbirds.     

Interspecific territoriality and competition for food between the reef fishes Forsterygion varium and Pseudolabrus celidotus

Long-term behavioural studies were made of the blennioid fish Forsterygion varium and the wrasse Pseudolabrus celidotus in the Cape Rodney-Okakari Point Marine Reserve in New Zealand. Both male and female F. varium defended small 1 to 2 m² territories from the wrasse all year round, althoughl aggressive encounters were most frequent from December to April. Only Large P. celidotus (≧180 mm standard length) were observed to dominate the blenny in interactions. Defence of nests, which were found only on the territories of males between May and October, did not explain the distribution of F. varium aggression between the sexes and seasons. An extensive overlap in the categories and size of prey eaten suggested that protection of food was a major function of F. varium territoriality. F. varium was most frequently aggressive toward P. celidotus of an intermediate size (110 to 160 mm SL) and exhibited greatest overlap in diet with wrasses in this size lange. This dnected the study toward examining the effect of F. varium on the feeding activity of P. celidotus. Within a shallow, Leterogeneous reef habitat, feeding pressure by P. celidotus (bites/are/time) was inversely related to the density of F. varium. Removal of F. varium from small 5x5 m quadrats resulted in large increases in wrasse feeding pressure, suggesting that the blenny influences the foraging of wrasses within their home ranges. When F. varium were removed from the entire home ranges of five 1+yr (110 to 130 mm SL) P. celidotus, these fish exhibited relative increases in both feeding rate and foraging time. The results suggest that P. celidotus grow through a period during which individuals compete with F. varium for food, adding support to the contention that territoriality in F. varium functions to preserve its food supply.     

Correlates of female choice in resource-defending antelope

Summary Where males offer their mates assistance in rearing offspring as well as access to defended resources, female mate choice should be influenced by both male phenotype and resource quality. In contrast, where there is no paternal care the relative importance of choice for male and territory traits is less well understood. We looked at female distribution across male territories in order to assess mate choice criteria in puku Kobus vardoni, and topi Damaliscus lunatus, two antelope where males defend resources but are not involved in parental care. In both species female distribution was correlated with male phenotype as well as the quality of forage and risk of predation on different territories. Male and territory characteristics were intercorrelated, but statistical analysis revealed that male traits, forage quality and predation risk were all significant, independent predictors of female choice in both study species. This is the first demonstration that female choice in mammals might be simultaneously influenced by male phenotype as well as the quality of defended resources.Correspondence to: A. Balmford     

Correlations of food distribution and patch size with agonistic interactions in female vervets (Chlorocebus aethiops) and patas monkeys (Erythrocebus patas) living in simple habitats

Food distribution is hypothesized to be important in determining the nature of female relationships within social groups of primates. When food limits female reproductive success, spatially clumped foods are expected to produce strong, linear dominance hierarchies within groups, whereas more spatially dispersed foods are expected to produce weaker or non-existent dominance hierarchies. The association between food distribution and competitive relationships presumably occurs because clumped foods are usurpable but dispersed foods are not. We examined the spatial distribution of food patches (trees) and patch size relative to feeding behavior and agonistic interactions in vervets and patas monkeys, two closely related and sympatric species that nonetheless differ in the strength of the female dominance hierarchy. Food patches of both patas monkeys and vervets were small in size and randomly distributed in Acacia drepanolobium habitat. In contrast, in A. xanthophloea woodland, the habitat type that was exclusively used by vervets, food patches were larger and more spatially clumped. These similarities and differences between and within species were correlated with similarities and differences in the strength and linearity of their dominance hierarchies. Patas monkeys and vervets in A. drepanolobium habitat had dominance hierarchies that were weakly defined because there were relatively few agonistic interactions between females. By contrast, in A. xanthophloea habitat, vervets had a stronger, linear dominance hierarchy characterized by a higher rate of agonistic interactions over food. The covariation of agonistic interactions with patch size is discussed in relation to depletion time, another characteristic that may covary with food distribution, and resource renewal rate, an important determinant of agonistic interactions in insectivorous birds, fishes, insects, and mammals. Received: 18 February 2000 / Revised: 5 September 2000 / Accepted: 26 September 2000     

Prey selection by the hogchoker, Trinectes maculatus (Pisces: Soleidae), along summer salinity gradients in Chesapeake Bay, USA

We investigated feeding by the hogchoker, Trinectes maculatus (Bloch and Schneider), in freshwater, oligohaline, mesohaline, and polyhaline regions of Chesapeake Bay, USA, and examined prey selection in relation to food availability. Otter trawling for fish and Van Veen grab sampling for benthic macrofauna occurred in July and August 1992 and August and September 1993. Hogchokers exhibited both opportunistic and selective feeding patterns along the estuarine salinity gradient in four tributaries (Potomac, Rappahannock, York, and James Rivers) and in the mainstem Chesapeake Bay. Major prey taxa included annelids, arthropods, and tellinid siphons. In polyhaline habitat, polychaetes dominated both the benthos and gut contents numerically and gravimetrically. On the other hand, oligochaetes were numerically dominant in freshwater/oligohaline areas but were rarely eaten, perhaps because of their burial depth. Arthropods (mostly amphipods) occurred at most salinities, were common in gut contents in low-salinity areas, and were replaced as prey by larger proportions of polychaetes in polyhaline regimes. Although hogchokers ate tellinid siphons, they rarely consumed whole bivalves or gastropods. These diet patterns (and especially the importance of siphon nipping) are similar to those of juvenile or small flatfish elsewhere in Europe, Africa, and North America. A size–salinity relationship for hogchokers occurred along the summer salinity gradient, with smaller fish predominating upstream and larger fish downstream. It was not clear from our data if variation in diet composition reflected changes in prey composition along the salinity gradient rather than changes in fish size. Received: 14 June 1997 / Accepted: 27 June 1997