Sheephead, Urchins, and Kelp: the Effect of a No-Take

Marine Reserve on a Trophic Cascade


Patrick W. Robinson



Temperate kelp forests have frequently been the sites of study for ecological interactions. The study of trophic cascades is no exception, as evident by the wealth of relationships studied (see review by Pinnegar et. al. 2000). The sea urchin is the common species among most of the cascades, providing a critical link from the carnivorous predators to the primary producers. It has been shown that sea urchins are part of a dynamic relationship that determines the state of a kelp forest (Jones and Kain 1967). Urchin barrens and kelp-dominated regions are two possible stable states that urchins can create based on their grazing habits, which are in turn related to the availability of drift kelp (Harrold and Reed 1985). Sea urchin predators can also have a profound influence on this system. Several studies off the coast of Canada show that lobsters (Homarus americanus) can also control kelp abundance indirectly via urchin consumption (Mann and Breen 1972; Miller et. al. 1971). Mann and Breen (1976) go on to stress the importance of the lobster-urchin-kelp relationship because of the potentially devastating effects, given a perturbation, it can have on a kelp forest and the species that depend on it.

A different species of lobster (Panulirus interruptus) was also found to regulate urchin populations in southern California (Tegner and Levin 1983). However, this system is quite different due to the presence of a second predator: California sheephead (Semicossyphus pulcher), a large labrid that is also capable of controlling urchin populations (Tegner and Dayton 1981). Cowen (1983) helped to show the importance of the sheephead-urchin relationship by demonstrating that 36% of the sheephead studied showed evidence of recently consuming all or part of an urchin. Further proof of the influence sheephead have on urchins comes in the form of a behavioral adaptation. Black urchins (Centrostephanus coronatus) have developed a nocturnal feeding strategy in which the urchin remains close to a protective crevice to avoid predation by the diurnal sheephead (Nelson and Vance 1979). This behavior is notably different from that of Strongylocentrotus spp., in which individuals will subsist on drift kelp or actively graze living adult kelps (Harrold and Reed 1985). Despite this difference, C. coronatus can still have an impact on a kelp forest. Vance and Schmitt (1979) showed that C. coronatus prefers brown algae, of which Macrocystis pyrifera is a member, and can influence the number of juvenile kelps within its "halo" of grazing.

Because lobsters and sheephead can influence multiple trophic levels of a system, they can be thought of as keystone species. Sea otters (Enhydra lutris) are more commonly thought of as the important keystone species of the kelp forests on the western coast, but their absence, a remnant of the fur trade, precludes them from having a negative influence on the urchin population (Estes and Duggins 1995). When humans are included as the apex predators of the trophic cascade, a more complete understanding of the trophic interactions of this system results (fig. 1). Furthermore, humans can be removed from the system, i.e. in the case of marine reserves, to restore the "naturally" higher urchin predator densities (Bell 1983), the effects of which have not been intensely studied.

The aim of this paper is to assess the effect that a no-take marine reserve has on the sheephead-urchin-kelp system. Three basic questions will be addressed with respect to the effect of a marine reserve. First, what are the differences in sheephead, urchin, and kelp densities? Second, what variable is most strongly correlated with urchin density? Third, what variable is most strongly correlated with kelp density?





Figure 1. Diagram of the trophic interactions in southern California (adapted from Pinnegar et. al. 2000). Note that in southern California hunting has completely eliminated sea otters as urchin predators.


This opportunistic study took place on the eastern border of the Catalina Marine Science Center Marine Life Refuge, a no-take marine reserve on Santa Catalina Island, California, USA (fig. 2). All observations were done on scuba between August 5 and August 11, 2002. The site was divided into twelve continuous 100m sections, six inside the reserve and six outside of the reserve. With two exceptions deviating only slightly from this pattern, each 100m section was divided into two 30m sections: the first running from the 10m mark to the 40m mark and the second from the 60m mark to the 90m mark.

Figure 2. Study site on Santa Catalina Island. (Image adapted from McArdle 1997).

To determine the density of sheephead, three transects were completed for each 30m section, each corresponding to a different depth contour: 14ft., 26ft., and 38ft. On each 30 X 2 X 3m transect, the total number of juvenile, female, and male sheephead were recorded with their associated estimated total length. Size estimations were pre-calibrated to within approximately ±10% of the actual length of the fish. Due to the curious nature of sheephead, the observer maintained a constant swimming speed and only the fish in front of the observer were included.

To determine the urchin density, kelp density, and substrate information, one of the three sheephead transects of each 30m section was chosen at random to be repeated. This method was chosen due to the time consuming nature of this form of data collection. By completing several passes of these transects in succession, five variables were recorded. First, the total number of urchins on each transect, separated by species, was recorded. Next, the total number of M. pyrifera were counted and divided into two categories: adult plants, those taller than one meter, and juvenile plants, those being shorter than one meter. The number of stipes per plant was also recorded. For adults, this measurement took place at one meter off of the substrate. For juveniles, the measurement was taken at the point of maximum stipes. Next, the substrate size and cover were determined by random point contact. At least 15 stratified random points were sampled directly under the transect tape. The cover measurement was separated into five categories: bare rock, coralline algae, turf algae, invertebrate, and M. pyrifera. The approximate substrate size was determined with the aide of the transect tape. Lastly, relief was measure at three stratified random locations below the transect tape. A one-meter long chain was used to follow the contour of the substrate. The strait-line distance that the chain covered was then recorded from the meter tape (McClanahan and Shafir 1990). This method produced values such that lower numbers corresponded to higher relief. The values obtained for all variables in each 100m section were then averaged, thus eliminating some of the variability associated with mobile creatures like the sheephead.


A comparison of the average densities of the observed species inside versus outside of the reserve showed that all species, with the exception of C. coronatus and S. franciscanus, had a higher density inside the reserve (Table 1). With respect to sheephead, this pattern was also observed for numerous size class groupings. The difference in density of sheephead inside versus outside of the reserve reached a maximum at the size grouping of 37cm or greater, well above the 30cm legal fishing size limit (Fig. 3).


Mean out of reserve


Mean in reserve


All Sheephead





Sheephead >30cm





Sheephead >35cm





Sheephead >40cm





C. coronatus





S. franciscanus





S. purpuratus





M. pyrifera





M. pyrifera Stipes





Table 1. Mean number of individuals per 30m segment of coastline with associated standard deviations.


Figure 3. Difference in average sheephead density per 30m section of coastline based on size grouping.

Sheephead density was then compared to C. coronatus density to confirm a predator-prey interaction. Only sheephead larger than 30cm were included in this correlation because it was presumed that only large sheephead have the ability to eat urchins. Surprisingly, the data obtained from inside the reserve yielded a positive relationship with a very high correlation (R2 = 0.977) (Fig. 4A). Thus, higher C. coronatus densities corresponded to higher large sheephead densities and vice versa. This pattern, however, does not occur outside of the reserve. Beyond the reserve boundary, there was no discernible relationship between large sheephead and C. coronatus (R2 = 0.017) (Fig. 4B).

Figure 4. Comparison of sheephead density to C. coronatus density inside and outside of the reserve. Densities are given per 30m section of coastline averaged over 100m sections. Densities within 100m of the reserve boundary are not shown.

Figure 6. Comparison of M. pyrifera density and C. coronatus density inside and outside of the reserve. Densities are given per 30m section of coastline averaged over 100m sections. Densities within 100m of the reserve boundary are not shown.

Continuing down the trophic cascade, C. coronatus density was compared to total kelp density (Fig. 6). Outside of the reserve, a strong negative correlation was found (R2 = 0.923). Inside the reserve, a much weaker negative correlation was found (R2 = 0.480). In an attempt to understand the reasoning behind this difference a three-dimensional plot was made with an extra variable: relief (Fig. 7). It was found that relief could explain more than 85% of the C. coronatus density variation within the reserve.

Figure 7. Comparison of kelp (M. pyrifera) density, urchin (C. coronatus) density, and relief inside the reserve. Densities are given per 30m section of coastline averaged over 100m sections. Densities within 100m of the reserve boundary are not shown. The three-dimensional surface is an approximation based solely on the points shown.


This study demonstrates that the presence of a no-take marine reserve can have both direct and indirect effects on a trophic cascade. The lack of fishing in the reserve presumably had the direct effect of elevating the average density and size of sheephead, as has been shown in other studies (Bell 1983). As expected, this change had an impact on the next trophic level, namely a lower density of C. coronatus in the reserve. However, some potentially contradictory data was obtained when compared to the results of other research (e.g. Sala and Zabala 1996). A strong positive relationship between sheephead density and C. coronatus density was found within the reserve when looking at 100m sections away from the reserve border. This spatial scale is clearly relevant due to the general lack of long-range movement of sheephead (Irene Tetreault, unpublished data) and relatively sessile nature of C. coronatus. One possible explanation is the feeding strategy of sheephead. It is likely that only very large sheephead, for example those over 45cm, are capable of ripping an urchin off the substrate and cracking the test and that many smaller sheephead may inhabit an area simply to take advantage of a free meal. Cowen (1983) demonstrated that an average of three sheephead preyed upon an urchin once it had been removed from the substrate. This was clearly not the case outside of the reserve where no correlation was found between sheephead and C. coronatus. This is likely due to the very low density of extremely large sheephead.

The next interesting relationship was found to exist between C. coronatus and the number of M. pyrifera. A very strong negatve correlation was found outside of the reserve, indicating successful kelp grazing. This suggests that in the absence of large sheephead the predator-avoidance behavior of C. coronatus has been relaxed. Fish have been shown to alter the behavior of C. coronatus to stay close to a centralized hiding place, thus forcing the it to develop the ability to consume a diversity of prey species (Vance and Schmitt 1979). Thus, one would expect a decrease in predation to cause an increase in consumption of the C. coronatus food of choice: brown algae.

Not surprisingly, the relationship between C. coronatus and M. pyrifera within the reserve was not strong. Large sheephead presumably keep urchin predator-avoidance behavior well intact, minimizing the amount of M. pyrifera grazing. This, combined with the inherently low C. coronatus density, has allowed the competitive dominant species, M. pyrifera, to increase in density, with respect to outside of the reserve (Vance 1979). The actual cause of C. coronatus distribution within the reserve is likely a direct effect of the availability of suitable shelter. The relatively high correlation found between C. coronatus and relief, a measure of available shelter, in the reserve confirms similar research on a congener, C. rodgersii, which faces similar predation risks (Andrew 1993). Because this study did not contain any experimental components, the correlations found do not necessarily confirm causation of the observed patterns. These systems have many other yet unstudied causative variables that could potentially have large effects. For example, larval abundance, post-settlement survival and disease can all drastically influence adult populations (Watanabe and Harrold 1991; Sala et. al. 1998). Tegner and Levin (1983) also point out another potential influence on the trophic cascade: a possible predator-prey relationship between sheephead and lobster. Clearly, the dynamics of this system will require much future study to be fully understood.


I wish to thank Irene Tetreault for the incorporation of this project into her work and for her mentoring, my Dad for his help with data analysis, Marc Carr and Pete Raimondi for their help with experimental design, the staff at Wrigley Marine Science Center, and my numerous dive buddies.







Andrew, N. L. 1993. Spatial heterogeneity, sea urchin grazing, and habitat structure on reefs in temperate Australia. Ecology 74: 292-302.

Bell, J. D. 1983. Effects of depth and marine reserve fishing restrictions on the structure of a rocky reef fish assemblage in the north-western Mediterranean Sea. J. Appl. Ecol. 20: 357-369.

Cowen, R. K. 1983. The Effect of sheephead (Semicossyphus pulcher) predation on red sea urchin (Strongylocentrotus franciscanus) populations: an experimental analysis. Oecolgia 58: 249-255.

Estes, J. A. and D. O. Duggins 1995. Sea otters and kelp forests in Alaska: generality and variation in a community ecological paradigm. Ecological Monographs 65 (1): 75-100.

Harrold, C. and D. C. Reed 1985. Food availability, sea urchin grazing and kelp forest community structure. Ecology 66: 1160-1169.

Jones, N. S. and J. M. Kain 1967. Subtitdal algal colonization following removal of Echinus. Helgolander Wiss. Meeresunters 15: 460-466.

Miller, R. J., K. H. Mann, and D. J. Scarratt 1971. Production potential of a seaweed-lobster community in eastern Canada. J. Fish. Res. Bd. Canada 28: 1733-1738.

Mann, K. H. and P. A. Breen 1972. The relation between lobster abundance, sea urchins, and kelp beds. J. Fish. Res. Bd. Canada 29: 603-609.

Mann, K. H. and P. A. Breen 1976. Changing lobster abundance and the destruction of kelp beds by sea urchins. Mar. Biol. 34: 137-142.

McArdle, D. A. California Marine Protected Areas. California: California Sea Grant College System, 1997.

McClanahan, T. R. and S. H. Shafir 1990. Causes and consequences of sea urchin abundance and diversity in Kenyan coral reef lagoons. Oecologia 83: 362-370.

Nelson, B. V. and R.R. Vance 1979. Diel foraging patterns of the sea urchin Centrostephanus coronatus as a predator avoidance strategy. Mar. Biol. 51: 251-258.

Pinnegar, J. K., N. V. C. Polunin, P. Francour, F. Badalamenti, R. Chemello, M. L. Harmelin-Vivien, B. Hereu, M. Milazzo, M. Zabala, G. D’anna and C. Pipitone 2000. Trophic cascades in benthic marine ecosystems: lessons for fisheries and protected-area management. Environmental Conservation 27 (2): 179-200.

Sala, E. and M. Zabala 1996. Fish predation and the structure of the sea urchin Paracentrotus lividus populations in the NW Mediterranean. Mar. Ecol. Prog. Ser. 140: 71-81.

Sala, E., C. F. Boudouresque, and M. Harmelin-Vivien 1998. Fishing, trophic cascades, and the structure of algal assemblages: evaluation of an old but untested paradigm. Oikos 82: 425-439.

Tegner, M. J. and P.K. Dayton 1981. Population structure, recruitment and mortality of two sea urchins (Strongylocentrotus franciscanus and S. purpuratus) in a kelp forest. Mar. Ecol. Prog. Ser. 5: 255-268.

Tegner, M. J. and L. A. Levin 1983. Spiny lobsters and sea urchins: analysis of a predator-prey interaction. J. Exp. Mar. Biol. Ecol. 73: 125-150.

Vance, R. R. 1979. Effects of grazing by the sea urchin, Centrostephanus coronatus, on prey community composition. Ecology 60 (3): 537-546.

Vance, R. R. and R. J. Schmitt 1979. The effect of predator-avoidance behaviour of the sea urchin Centrostephanus coronatus, on the breadth of its diet. Oecologia 44: 21-25.

Watanabe, J. M. and Harrold, C. 1991. Destructive grazing by sea urchins Strongylocentrotus spp. in a central California kelp forest: potential roles of recruitment, depth, and predation. Mar. Ecol. Progr. Ser. 71: 125-141.