Anjasjostrom and Ryan Stephenson

Competition between the comber (Serranuscabrilla) and painted comber (Serranusscriba) at STARESO, Corsica, France.

AnjaSjostrom and Ryan Stephenson

Abstract
Interspecies competition is occurring between two Serranids; Serranusscribaand Serranuscabrilla at STARESO in the bay of Calvi, Corsica. Normally partitioned by depth, the two species overlap in resource and habitat use in this area. Our goal was to identify differences and similarities between the two species, determine the extent to which their habitats and resources overlap, and establish how they are partitioning them to avoid constant competition. 37 fish specimen were examined in the lab for stomach content and jaw morphology. Available habitat and selective usage by the two species was established by uniform point contact (UPC) and opportunistic surveys noting size, depth, substrate and cover habitat of 290 S. scribaand 39 S. cabrilla were taken on SCUBA. Our data show the two species consuming highly similar prey items, and having little difference in jaw morphology. The two species are partitioning substrate, cover and depth within the study site. We conclude that competition is occurring at STARESO, and that S. scribaand S. cabrillaare partitioning habitat to avoid resource overlap.
Introduction

Serranusscriba, the painted comber,and Serranuscabrilla, the comber, are two species of grouper that coexist in the nearshore rocky reefs off Corsica, France. Though similar in trophic level and morphology, both species tend to segregate by depth making co-occurrence uncommon. According to Dr. Pierre Lejeune (pers., Comm), the research director at STARESO, S. scribahave had a long and stable population in the area whereas S. cabrillawas present in the area until 1995, when the population became severely depleted. This deep population of S. cabrillaused to be found in 50-80m depths. In 2000, S. cabrillabegan to reappear in the area. Serranusscribaare found in depths of 0-30m, whereas S. cabrillaoccupy depths of 20-500m (LythgoeLythgoe, 1975). At STARESO, S. cabrillacan be seen in shallower depths, and overlapping with the local population of S. scriba. Despite the species normal depth segregation, the two species share similar habitat and dietary preferences, indicating high potential for interspecies competition (Fasola et al., 1999).

Serranusscribaand S. cabrilla are small fish of the family Serranidae and range across the Mediterranean and Eastern Atlantic and are true hermaphrodites. Serranusscriba lives over seagrass beds and rocky reefs and is territorial and solitary. Serranuscabrillais the less abundant of the two and is also a solitary species occupying sand bottoms and rocky reefs (Guidetti, 2000). We know that the diets of the two species are similar (LythgoeLythgoe, 1975) but we expect that there will be small scale changes in the diet due to competition. Interspecific competition between such sympatric species has been well documented in rocky reefs off the California coast (Holbrook & Schmitt 1988; Larson, 1980) but has not been addressed for these two species. We believe that both species may be competing for resources and space. In a study in California (Larson, 1980), one species of rockfish was found to have established a more pronounced dominance in the overlapping habitats than the subordinate rockfish. This may be present in the competition occurring between S. scribaand S. cabrilla.

With limited background information on comparative ecology between S. scribaand S. cabrilla, the purpose of this study was to a) determine whether both species are partitioning resources, and b) whether both species are partitioning habitat. To answer these questions, we conducted diet analysis and morphological comparisons in the lab and used SCUBA to determine small scale spatial habitat use in the field. We expect that dietary composition will vary between the two Serranids species, and that they will characteristically differ in their utilization of the Posidoniaand rocky reef assemblages for foraging and shelter at STARESO.
Methods and Materials
Species Description
Apparent habitat and diet similarity between S. scriba and S. cabrilla present a valuable opportunity to examine resource and habitat partitioning of sympatric species. Previous work on Mediterranean rocky slope fish assemblages indicates the species partition their habitat by depth (Fasola et al 1997) and generally S. scriba utilizes Posidoniaoceanicameadows more as habitat than and S. cabrilla. This makes STARESO an ideal study site, with meadows extending past the depths that stratify the two species, forcing S. cabrilla up into the shallows dominated by S. scriba. Both species are relatively common at STARESO, where they utilize nearshore rocky reefs as habitat. Their diets consist mainly of small crustaceans and other fish (Fasola et al., 1999). Both species have relatively small home ranges and are simultaneous hermaphrodites, breeding in the summer.
Site description
We gathered all observational data on SCUBA at STARESO field station, Calvi, Corsica in the northwest Mediterranean Sea (N 42.581795, E 8.725128). We collected data north and south of Stareso harbor in Posidoniaoceanica meadows and rocky reef slopes between 3 and 14 meters (Figure 1).
Lab Methods
Resource partitioning is occurring between Serranusscriba and Serranuscabrilla.
Gut Content Analysis
To examine if diet composition differs between S. scribaand S. cabrilla, we collected individuals of both species for gut content analysis by spearfishing. We emptied the contents of the stomach onto a petri dish marked with a twenty-five point grid. We counted all contents in a standardized way. At each of the grid’s intersections, prey was identified as fish, crustacean, mollusk or cephalopod. By tallying all the points for each prey type, the percent composition that each prey type represented was determined.
Jaw Measurements
In order to test whether jaw morphology is dissimilar in S.scribaand S.cabrillawe took measurements to provide an illustration of potential prey differences between the two species. Standard length and width of each specimen was taken prior to gut removal and analysis. To test for differences in jaw morphology between the two serranidspp. vertical and horizontal gape measurements were taken. All measurements were recorded in millimeters.

Field Methods
Serranusscriba and Serranuscabrilla are partitioning habitats and depth in the mediterranean rocky reefs and posidonia fields.
UPC and Focal data surveys
To examine habitats of both species, we conducted uniform point contact (UPC) and opportunistic individual focal surveys to quantify comparative habitat features and resource use due to competition. We conducted surveys to the north and south of STARESO harbor. To characterize substrate habitat type, UPC counts were taken along 30m transects placed in areas of habitats known to be occupied by Serranus spp. We used the following substrate categories: sand, cobble (1-10cm), boulder (rocks larger than 10cm unattached to substrate), bedrock, and other. Cover layers were grouped in the following manner: turf=less than 3cm, bushy=greater than 3cm, foliose, erect coralline, encrusting coralline, gunk=film layer less than 1cm, sessile invert, other or none. To gather available cover and substrate data for the northern portion of the study area we used nine point quadrats (0.5m x 0.5m)off a permanent transect line. At 5 meter intervals we ran 30 meter transects off of the permanent line and placed the quadrat down for sampling at 5 meter intervals on either side of the transect tape. We counted substrate and cover under each of the quadrat’s nine points.

To quantify how both species were associating with the different habitats we conducted opportunistic focal data surveys each S.scribaor S. cabrillaencountered. Each fish was visually sized to the nearest centimeter and the depth it was sighted at recorded in meters. We recorded the focal data of each fish by their associated substrate and cover layer as defined by the UPC categories. Surveys were conducted from October 22th until October 30th 2012.

Statistical Analysis
Gut content was characterized by the number of diet items in each category and standardized to total length of the fish. Data were analyzed for similarity (Bray-Curtis distances) in PRIMER. Horizontal and vertical gapes were divided by the standard length of each specimen to standardize gape size for each fish. We then used these standardized gapes to analyse data for similarity (Bray Curtis distances) in PRIMER to determine similarity in jaw morphology of S. scribaand S. cabrilla. Chi square tests were run to compare the availability of each substrate and cover type to the observed use by S.scribaand S.cabrilla. ANOVA tests were run to analyze the available depth range and quantify any depth stratification between the two species.

Results
We must accept the null hypothesis that the diets of the two species do not differ (Figure 3, p<0.564). Many of the replicates from the analyses ate the exact same food type as another replicate, which gives the appearance of a lesser number of fish. We must accept the null hypothesis that there is no difference in the maximum vertical and horizontal gape between the two species (Figure 3, p0.964).S.cabrillaexhibits a strong positive association with bedrock, utilizing it significantly more than the available quantity (Figure 4, p<0.001). They showed a strong negative association with the relatively abundant Posidonia, with only two out thirty-nine individuals utilizing it as habitat (Figure 5, p<0.001). Negative associations with boulders reflect the high abundance of boulders in the survey area and the low abundance of S. cabrilla. S.scribashows a positive association with bedrock and cobble (Figure 6, p<0.001).Both results support the hypothesis that species distribution will differ by substrate type (Figure7, p<0.001). S. scriba is using Posidonia for habitat with double the presence it occurs in the study site. Depth stratification between the species revealed S. cabrillausing consistently deeper depths than S. scriba (Figure 8, ANOVA pooled variance, p<0.006). We removed Posidoniaas a variable, as only one S.cabrillaindividual was sighted over this cover, as a potential confounding factor and ran the ANOVA again. This generated an even more statistically significant p<0.001.

Discussion

In order for competition to be occurring between the fishthere are three important factors to consider: diet composition, habitat and depth. If they overlap in diet and habitat, they must partition the resources or the space or they would fight constantly. In territorial fish like S. scriba, defense of resources other than eggs from other species indicates the presence of competition (Grant 1997, Kirchshofer 1954). We detected no dietary difference between the two species in the study site. Serranuscabrilla’s ability to break down the chitinase of invertebrate shells exceeds that of S. scriba, but this difference did not reflect on the diets of the specimens we collected (Benmouna et al. 1986). The similarity in jaw morphologies indicates no adaptation to specific prey type. While more accurate assessments of jaw morphology require factors other than maximum gape, the species feed in a similar fashion (Viladiu et al. 1999). We hoped to increase the sample size of the fish collected to reveal smaller scale dietary differences, but this might prove biologically insignificant.

Use of habitat is quite different. Serranuscabrillaassociated most commonly with bedrock whereas boulders were the most prevalent substrate in the study area (Figure 5, P=0.001). Serranusscribaselected bedrock and cobble the most of the available substrates (Figure 6, P=0.001). In terms of cover data, S. scribawere consistently using Posidoniaoceanica, whereas the S. cabrillawere avoiding it. This proves interesting in two regards: one is that they are distinctly partitioning habitat both in terms of substrate and cover availability. The second is while the S. scribaat times overlap the rocky reef substrate, the S. cabrillapopulation are avoiding any association with the Posidoniaoceanica(Figure 7, P=0.001). This further solidifies S. scriba acting as the more dominant species in the area, and occupying a substantial portion of the available habitat while the S. cabrillaare limited to subset of that available habitat (Larson, 1980). Although Serranuscabrillaare commonly found over seagrass, we see very different habitat use at STARESO (Francour, 1997).

The presence of more S. cabrillawith increasing depths was consistent with literature pertaining to normal depth stratification between the two Serranids(Fasola et al., 1997; LythgoeLythgoe, 1976). S. cabrillawas consistently deeper than S. scriba, and even more dynamically so when Posidoniawas removed from the ANOVA test. S.cabrillaalso exhibited a higher standard deviation in depth towards deeper habitat, whereas the S.scribadeviated minimally from their shallower depth ranges. In a survey of rocky reef fish assemblages in the Mediterranean, the abundance of S. scribadeclines as the abundance of S. cabrilla increases along the depth gradient of 25m (Fasola et al. 1997). Given that our findings indicate overlap in diet and avoidance of habitat and depth overlap, the species interactions are quite similar to the black and yellow and Gopher rockfish of temperate reefs in California. A socially dominant species S. scriba/black and yellow rockfish forces the other S. cabrilla/gopher rockfish into a subset of the available habitat (Larson, 1980). In the 1980’s there was a stable population of S. cabrillain the area which declined significantly and became nearly non existent in 1995-2000 . The population is only now re-establishing itself in the area, while the S. scribahave had a consistent and stable presence in the area for this entire period. It is also noted that there is a population of S. scribafound at depths between 50-80m, but this deep population disappeared (P. Lejeune. pers. Comm). This gradual transition of the S. cabrillaback into an overlapping community as well as the unique situation in the bay of Calvi drive the Serranidsinto inevitable ecologically competitive circumstances. Given the morphological similarity of the sympatric species and prey preferences, they must partition either resources or habitat and depth, the latter of which we see to be true.
Another interesting visible characteristic of competition is territoriality, of which we did have several clear encounters. All aggressive displays between the two Serranidswitnessed resulted in S.scribachasing S.cabrillafrom the area of dispute. While we were not able to quantify this from a statistical standpoint as we did not take data on these incidences, we can register this as another important component to the ongoing interspecific competition between the species, and take this as a potential further indication of the S.scribadominance in this particular set of ecological circumstances.
We were limited on S.cabrilladata, due to the mere infrequency with which we encountered individuals. More collected specimen to analyze gut contents as well as more field observations would have allowed a clearer depiction of the resource and habitat preferences of the two species. Future studies on the deeper population of S.cabrillaand the depth ranges of S.scribawould give a more complete ecological picture of the two species and their interactions. There is still very minimal research on the S.cabrillaand even less on the interaction of ecological niches of the Serranus spp.
Acknowledgements
We would like to thank Pete Raimondi for his statistical genius, Gaby Keeler for all her help with transects and fish spearing, GiacomoBernardi for his incredible fish knowledge and the TA’s for fishwacking, diving and moral support. Also, the Bio 159 class of 2012.
References
Allen, Larry Glenn., Daniel J. Pondella, and Michael H. Horn. "Chapter 17: Competition." Ecology of Marine Fishes: California and Adjacent Waters. Berkeley: University of California, 2006. N. pag. Print.
Benmouna H., M.F. Jaspar-Versali, C. Toussaint, and Ch. Jeuniaux, 1986, A Comparative Study of Chitinase Activity in Digestive Tract of Serranuscabrilla and Serranusscriba, Biochemical Systematics and Ecology, Vol. 14 (4), p. 435-437.
Canova L., M. Fasola,, F. Foschi, O. Novelli and M. Bressan, 1996, Resource Use by a Mediterranean Rocky Slope Fish Assemblage, Marine Ecology, vol. 18 (1), p. 51-66.
Casinos A., C. Viladiu, P. Vandewalle, and J.W.M. Osse, 1999, Suction Feeding Strategies of Two Species of Mediterranean Serranidae (Serranuscabrillaand Serranusscriba), Netherlands Journal of Zoology, Vol. 49 (2), p. 81-95.
Francour, P., 1997. Fish assemblages of Posidoniaoceanica beds at Port-Cros (France, NW Mediterranean): assessment of composition and long-term fluctuations by visual census. P.S.Z.N.I., Marine Ecology, 18: 157-173.
Grant, J.W.A. 1997. Territoriality, p. 81–103. In: Behavioral ecology of teleost fishes. J.-G.J. Godin (ed.). Oxford University Press, Oxford, UK.
Kirchshofer, R. 1954.Okologie und revierverhailtnissebeimschriftbarsch
(Serranusscribacuv.).Osterreich. Zool. Z. 5:329-349.

Larson R. J., 1980, Territorial Behavior of the Black and Yellow Rockfish and Gopher Rockfish (Scorpaenidae, Sebastes), Marine Biology, Vol. 58, p. 111-122.

Lythgoe, John, and G. I. Lythgoe. Fishes of the Sea: The Coastal Waters of the British Isles, Northern Europe and the Mediterranean. New York: Anchor/Doubleday, 1975. Print.