BULLETIN OF MARINE SCIENCE, 81(2): 219–234, 2007

REPRODUCTION, ECOLOGY, AND EVOLUTION OF THE INDO-PACIFIC LIMPET FLEXUOSA

David R. Lindberg

ABSTRACT Scutellastra flexuosa (Quoy and Gaimard, 1834) was studied at Temae islet reef on Moorea, Society Islands, French Polynesia between 1998 and 2001 to compare and contrast the respective roles of deep phyletic history with recent adaptations in shaping its current ecological and life history characteristics. Most of the charac- ters examined are consistent with related specialist and are therefore deter- mined by ancestry. These characteristics include habitat restriction, algal gardening, local distribution, home site fidelity, adult/juvenile differentiation, and protandric hermaphroditism. The only character that appears autapomorphic and a possible adaptation to its proximal setting is its small body size. Large body size is often as- sociated with species that maintain and defend territories. However, the variance in size in clades with and without territorial species presents a more complex picture. The putative size reduction of S. flexuosa has not affected many of the specialized traits shared within its lineage and it remains a classic territorial taxon albeit in miniature. The phyletic pattern that emerges here is one of a clade dominated by specialist species that gave rise to generalist species that in turn gave rise to another group of specialists with identical traits albeit in different habitats.

Studies of the evolution of life history characteristics of marine invertebrates have produced a large body of empirical data and theoretical interpretations. Many recent studies of the evolution of life history traits have become grounded in phylogenet- ic analysis (e.g., McHugh and Fong, 2002; Kupriyanova, 2003; Meyer, 2003; Collin, 2004; Kohler et al., 2004; Byrne, 2006; Glenner and Hebsgaard, 2006); a trend that is a welcomed alternative to the former practice of using current taxonomic classi- fications to parse traits. In some cases, the results and interpretations of these more recent studies have corroborated earlier work, while in others they have suggested alternative hypotheses. Patellogastropod limpets (or “true” limpets) have long been studied in a broad ar- ray of biological disciplines. Because of their intertidal occurrence, they were some of the first and remain one of the commonest taxa manipulated in ecological studies (e.g., Jones, 1948; Lodge, 1948; Underwood, 1979; Branch, 1981; Hawkins and Hart- noll, 1983). They have served as physiological “guinea pigs” in studies of desiccation and growth (e.g., Wolcott, 1973; Branch, 1974b, 1975a; Parry, 1978; Creese, 1981; Lowell, 1984; Niu et al., 1998; Gray and Hodgson, 2004). They also have a long and illustrious record as model organisms in life history studies, especially with regard to reproductive cycles and hermaphrodism (e.g., Orton, 1919; Dodd, 1956; Orton et al., 1956; Fritchman, 1961; Rao, 1973; Branch, 1974a; Wright and Lindberg, 1982; Wright, 1988; Creese et al., 1990; Le Quesne and Hawkins, 2006). However, the dis- tributions of these life history characters have seldom been explored except for quick comparisons or corroborations of similarities (and differences) with “other limpets” in another region or genus. Here I report the results of a multi-year study of the patellogastropod limpet Scutel- lastra flexuosa (Quoy and Gaimard, 1834) on Moorea, Society Islands, French Poly- nesia. Scutellastra flexuosa is often a common member of the algal ridge community

Bulletin of Marine Science 219 © 2007 Rosenstiel School of Marine and Atmospheric Science of the University of Miami 220 BULLETIN OF MARINE SCIENCE, VOL. 81, NO. 2, 2007 where it experiences the full intensity of reef-crest wave action. Unlike most other patellogastropod faunas found in high-energy environments (Branch, 1976; Lind- berg, 1988), S. flexuosa is the only patellogastropod species to occur there. Therefore, it presents an opportunity to compare the roles of phyletic history and recent adapta- tions in shaping its ecology and life history uncomplicated by the co-occurrence of closely related taxa and interactions. To compare the ecological and life history characteristics S. flexuosa with other patellid species a standardized list of ecological and life history characters was used. This list was developed by George M. Branch (1975b) during his study of intraspecific competition in southern African spp. Branch was especially interested in the role of migration, differentiation, and territorial behavior and when he scored his study taxa he found that they sorted into two distinct categories with few interme- diates. Branch termed these groups “migratory” and “non-migratory”, but they also denote groupings of characters commonly associated with distinguishing generalist (migratory) and specialist (non-migratory) species. Scutellastra flexuosa was scored for these traits and the characters examined on a phylogenetic tree to determine character distribution and to separate shared historical characters (pleisomorphy) from new unique characters (autapomorphy) that might represent more recent and local adaptation (Irschick et al., 2005).

Materials and Methods

The study site was located on the Temae islet reef along the northeast shore of Moorea, Society Islands, French Polynesia (17°38´S, 149°27´W) (Fig. 1A–C). This location was selected because of the close proximity of the barrier reef to the coast and ease of access to a typically difficult habitat in which to conduct intertidal observations. Mean tidal range is < 0.25 m and low tides occur between 0530 and 0600 daily (NOAA, 2005). Scutellastra flexuosa occurs in patches on the algal ridge that gently slopes towards the ocean side of the reef crest (Fig. 1D). Limpets associated with these patches were observed yearly between 1998 and 2001; except for 1999 when surf conditions prohibited access to the site (Table 1). In addition to general observations, 17 0.25 m2 quadrats were arbitrarily located within patches and the limpets within these quadrats collected, measured, and sexed. Photoquadrats of both before and after removal were also taken and the removed limpets individually located in the quadrat by mapping their position prior to their removal (Table 1). General observations of S. flexuosa were photo-documented and used to score this species in Branch’s (1975b) matrix. Characteristics.—Branch (1975b) categorized nine “behaviors” of 11 patellid taxa from southern Africa. Each behavior was assigned one of four states represent by the symbols  ,  ,  ,  or was scored as unknown ? (Table 2). The solid circles represented “migratory” or gen- eralist characteristics while the open squares represented “non-migratory” or specialist char- acteristics. The size of the symbol represented well-developed and poorly developed traits. Scutellastra flexuosa was scored for these behaviors based on observations of the at Temae. To quantify Branch’s (1975b) symbolic representation each symbol was assigned a value between −1 and 1:  = −1.0,  = −0.5,  = 1.0, and  = 0.5. Data for Scutellastra argenvil- lei (Krauss, 1848), Scutellastra barbara (Linnaeus, 1758), Cymbula miniata (Born, 1758), and Helcion concolor (Krauss, 1848) were updated from Bustamante et al. (1995), Ridgway et al. (1999, 2000), and G. Branch (pers. comm., University of Cape Town, 2007). Median, mean, and standard deviations were then calculated for each species. Phylogeny.—A molecular phylogeny was constructed by reanalyzing 16s rRNA sequence data first presented and analyzed by Koufopanou et al. (1999). AdditionalS. flexuosasequenc - es were provided by C. Meyer, and additional Cellana spp. sequence data were obtained from LINDBERG: REPRODUCTION, ECOLOGY, AND EVOLUTION OF THE INDO-PACIFIC LIMPET 221

Figure 1. Location of study area. (A) Global position of Moorea, Society Islands, French Poly- nesia (17°38´S, 149°27´W). (B) Topographic hi-resolution map of Moorea showing location of Richard B. Gump South Pacific Research Station (diamond) and regional setting of study area along the northeastern coast. Map provided by the Moorea GIS Consortium. (C) Insert from box on (B) showing Temae islet reef area and location of study area. (D) Transect across barrier reef and reef flat at Temae islet reef illustrating geomorphological and biological zonation of the study area and habitat of Scutellastra flexuosa. Redrawn from Galzin and Pointier (1985).

Reeb (1995), Simison (2000), and Begovic (2004). The dataset was subjected to maximum parsimony analysis using PAUP* Ver. 4.0b10 (Swofford, 1998). All characters were equally weighted and unordered. The tree-bisection-reconnection (TBR) branch-swapping algorithm was used with a random addition sequence (500 replicates), and one tree was held at each step during stepwise addition.

Results

Scutellastra flexuosa densities at Temae ranged between 5 and 48 individuals per 0.25 m2 with an average density of 20.4 limpets per 0.25 m2. Sex ratios in the indi- vidual quadrats were highly variable with a mean ratio of approximately 5 males for every 2 females (0.38) (Table 1). There was no correlation between density and sex ratio (r2 = 0.0453). Females were most common in the larger size classes (m = 28.59 mm ± SD = 4.70), while males averaged 23.70 mm in length ± SD = 4.70, and indeterminate individuals averaged 16.36 mm in length (SD = 6.00). The size frequency distribution ofS. flexu- 222 BULLETIN OF MARINE SCIENCE, VOL. 81, NO. 2, 2007

Table 1. Sampling data and summary statistics for Scutellastra flexuosa at Temae islet reef, Moorea, French Polynesia. ∗ indicates females (F) significantly larger than males (M).

Quad Position Density F : M F size (mm) M size (mm) P: F > M TE29IX98.1 17.4750°S, 149.770°W 5 0.25 34.9 29.33 ± 2.6 0.157 TE29IX98.2 17.4750°S, 149.770°W 19 0.36 30.74 ± 3.0 24.74 ± 5.3 0.052 TE7X98.1 17.4750°S, 149.770°W 9 0.50 29.43 ± 2.1 24.65 ± 2.9 0.039∗­ TE7X98.2 17.4750°S, 149.770°W 25 0.47 33.55 ± 3.06 25.47 ± 4.6 0.001∗­ TE7X98.3 17.4750°S, 149.770°W 24 0.33 29.00 ± 3.2 25.07 ± 3.5 0.005∗­ TE1VI00.1 17.4750°S, 149.770°W 26 0.92 25.28 ± 4.33 24.43 ± 3.58 0.021∗­ TE2VI00.1 17.4750°S, 149.770°W 19 0.78 29.81 ± 4.25 25.53 ± 6.03 0.127 TE25V01.1 17.4731°S, 149.773°W 32 0.48 30.50 ± 6.13 22.80 ± 5.02 0.047∗­ TE26V01.1 17.4738°S, 149.772°W 9 0.33 24.70 ± 6.65 24.02 ± 3.83 0.298 TE26V01.2 17.4744°S, 149.771°W 6 0.67 25.50 ± 0.71 21.13 ± 1.29 0.117 TE27V01.1 17.4751°S, 149.770°W 22 0.19 26.10 ± 0.66 20.04 ± 3.18 0.024∗ ­ TE28V01.1 17.4752°S, 149.770°W 40 0.19 28.23 ± 3.19 24.81 ± 1.31 0.009∗ ­ TE29V01.1 17.4757°S, 149.769°W 11 0.10 29.00 28.49 ± 7.43 0.752 TE29V01.2 17.4758°S, 149.769°W 13 2.25 26.49 ± 3.01 24.73 ± 2.29 0.395 TE30V01.1 17.4763°S, 149.768°W 18 0.23 31.60 ± 6.61 21.32 ± 4.02 0.009∗ ­ TE31V01.1 17.4729°S, 149.773°W 20 0.28 28.38 ± 0.29 20.57 ± 5.41 0.003∗ ­ TE1VI01.1 17.4752°S, 149.770°W 48 0.18 27.46 ± 6.20 22.03 ± 3.90 0.002∗ ­ Summary 356 0.38 28.59 ± 4.70 23.70 ± 4.70 0.000 ­

osa by sex (Fig. 2) suggests that this species may be a protandric hermaphrodite like several other patellogastropod limpets which show similar patterns (Patella: Orton et al., 1956; Le Quesne and Hawkins, 2006), (Scutellastra: Branch, 1974a; Creese et al., 1990) (Lottia: Wright and Lindberg, 1982). The mean size of females was always greater than that of males, significantly so in 10 out of the 17 quadrats sampled (Table 1), and were significantly larger than males in the overall sample (t-test: t = 8.885, df = 329, Bonferroni adjusted P-value < 0.001). While there was no correlation between density and sex ratio, there was a correlation between the presence of significantly larger females and quadrat density (Fig. 3). This relationship was best fitted by an exponential curve (r2= 0.52). Characteristics.—Scutellastra flexuosa is zone restricted and occurred only on the algal ridge on the ocean side of the reef crest. The total vertical range of its distri- bution is approximately 20 cm and there was no evidence of up shore migration. The distribution of limpets on the algal ridge was random to dispersed (Fig. 4A). Scutellastra flexuosa maintains an algal garden around the periphery of its home site (Figs. 4B,C). The surfaces of the gardens are highly pockmarked by radular rasp- ing of the coralline algal substrate and the alga within the garden area is restricted to the interstices of the rasped surface (Fig. 4B). Interior to the garden, the limpet forms a permanent home depression which precisely fits the crenulated margins of the shell (Fig. 4C). The inner surface of the home depression is two tiered with an outer crenu- lated shell depression and an inner deeper foot depression (see Lindberg and Dwyer, 1983). The coralline algal surface within the home depression shows both dissolution and radular rasping. It also lacks pink coloration and the area of the foot depression often has a green tinge. These characteristics indicate a specific food source provided by the algal garden, a permanent home depression and rigid homing to the form- fitting home depression. LINDBERG: REPRODUCTION, ECOLOGY, AND EVOLUTION OF THE INDO-PACIFIC LIMPET 223  Non-migratory

Aggressive defense Territorial Sharp adult juvenile differentiation Zonation restricted Scar permanent Homing rigid Low gonad output Specific food Random to dispersed S. cochlear S.

£ £ £ £ £ £ £ £ £ S. longicosta S.

£ £ £ £ £ £ £ £ £ S. tabularis S.

£ £ £ £ £ £ £ £ £ C. miniata C.

£ £ £ £ £ £ £ £ £ C. compressa C.

£ £ £ £ £ £ £ £ £ S. flexuosa S.

? ? ? £ £ £ £ £ £ S. argenvillei S.

£ £ £ £ £ £ £ £ S. barbara S.

£ £ £ £ £ l l   £ C. oculus C. £ £ £ l l 

− 1.00 0.50 0.50 1.00 1.00 0.50 1.00 1.00 1.00 − 0.44 0.17 0.39 1.00 0.83 0.72 1.00 1.00 0.83 H. concolor H. £   

− 1.00 − 0.78 C. granatina C.

− 1.00 − 1.00 = unknown). Redrawn from Branch (1975b). Data updated from Bustamante et al. (1995), Ridgway et al. (1999, al. et Ridgway (1995), al. et Bustamante from updated Data (1975b). Branch from Redrawn unknown). = ? granularis S. l l l l l l l l l l l ll l l l l l l l l l l l l l 0.00 0.00 0.26 0.73 0.66 0.55 0.00 0.25 0.26 0.00 0.00 0.25 − 1.00 − 1.00 = 0.5, = £ = 1.0, = £ − 0.5,  =  Migratory − 1.0, = Table 2. “Migratory” Table and “Non-migratory” classification of African southern patellacean taxa. Summary scores are the median and means of all trait scores by ( l taxon Migrate up shore Aggregate to random No marked adult juvenile differentiation Scores Median 2000), and G. Branch (pers. comm. 2007). Generalized food Homing varies; may be absent High gonad output No territory Non-aggression Scar temporary or absent Mean SD 224 BULLETIN OF MARINE SCIENCE, VOL. 81, NO. 2, 2007

Figure 2. Size frequency distribution of indeterminate (white), male (grey), and female (black) individuals of Scutellastra flexuosa at Temae islet reef, Moorea. n = 356. Box plots above bar graph display sample mean and quartiles for each group; * = outliers.

Figure 3. Scatterplot of probabilities that female size = male size (Kruskal-Wallis One-Way Anal- ysis of Variance) plotted against limpet density by quadrat. Vertical line marks 0.05 significance level. LINDBERG: REPRODUCTION, ECOLOGY, AND EVOLUTION OF THE INDO-PACIFIC LIMPET 225

Figure 4. Scutellastra flexuosa, Temae islet reef, Moorea. (A) Patch on algal ridge dominated by limpets; (B) individual on home depression surround by algal garden. Note radular erosion of gar- den area. (C) Juvenile situated on the shell of an individual with home depression and garden.

Juvenile limpets (≤ 10 mm in length) are typically found on the shells of the larger limpets. Juvenile limpets lack the highly crenulated margins of the adults and have a dark shell with six white ribs arranged in a distinctive star-like pattern () (Fig. 4C). This distinctive shell pattern is in marked contrast with the highly eroded surfaces of the adult shells (Fig. 4B). The behavioral categories of aggression and territorial defense and gonad output for S. flexuosa were not assessed. Six of Branch’s (1975b) behavioral characters were reliably determined for S. flexuosa (Table 2) and the scores for these characters place this species with certainty in the specialist group (Fig. 5). Phylogeny.—The maximum parsimony analysis found 34 trees of 1268 steps on 12 tree islands. The majority rule consensus tree (Fig. 6) differs little from the Maxi- mum-likelihood tree published by Koufopanou et al. (1999: fig. 5). In addition to the taxa treated in Table 2, four additional taxa were scored. Patella vulgate (Linnaeus, 1758) was scored as “migratory” (Lewis and Bowman, 1975; Jones and Baxter, 1986; Chelazzi et al., 1998; G. Branch, University of Cape Town, pers. comm., 2007), while Scutellastra chapmani (Tenison-Woods, 1876) (Lindberg, pers. observ.), Scutellastra laticostata (Blainville, 1825) (Scheibling and Black, 1993), and Scutellastra aphanes (Robson, 1986) were scored as “non-migratory”. 226 BULLETIN OF MARINE SCIENCE, VOL. 81, NO. 2, 2007

Figure 5. Quantitative display of Branch’s (1975b) behavioral characteristics for generalists (“mi- gratory”) (black bars) and specialists (“non-migratory”) (white bars) patellid limpets in the south- ern Africa intertidal ecosystem. Grey bar values for Scutellastra flexuosa from this study. Means of character scores are shown on left and medians of the scores on right for each species. Also see Table 2.

Figure 6. 16s rRNA molecular phylogeny of the constructed by reanalyzing sequence data from Koufopanou et al. (1999). Additional sequences provided by C. Meyer, and from Reeb (1995), Simison (2000), and Begovic (2004). Territorial taxa [“non-migratory” sensu Branch (1975b)] are in bold, data to the left of the taxa represent median sizes (mm) and biogeographical distributions. LINDBERG: REPRODUCTION, ECOLOGY, AND EVOLUTION OF THE INDO-PACIFIC LIMPET 227

Discussion

Ecological and life history traits of S. flexuosa most likely reflect its phylogenetic history rather than being adaptations to its current habitat and distribution. These traits include habitat restriction, algal gardening, local distribution, home site fidel- ity, adult/juvenile differentiation, and protandric hermaphroditism. Large body size is also correlated with these other characters, but in the case of S. flexuosa this trait appears reversed and autapomorphic as a result of selection for small body size re- flecting the topography of the habitat. The scutellastrid clade Scutellastra( , Cymbula, and Helcion), of which S. flexuosa is a member dates at least from the Cretaceous (Kase and Shigeta, 1996) and may extend back to the Triassic (Hedegaard et al., 1997). The basal branches of the living clade are dominated by specialized “non-migratory” species. A total of eleven of the 19 scutellastrid taxa are categorized as “non-migratory” taxa. Of these eleven species, eight are accounted for in the first 10 taxa on the tree, while in the remaining nine taxa, only three species are categorized as “non-migratory”. This pattern suggests that the shared traits of the “non-migratory” species are ancestral and are present in S. flexuosa and the other basal members of this group because they were present in the common ancestor and are not independently derived. This pattern also suggests that “non-migratory” traits have arisen independently in at least two taxa, Cymbula compressa (Linnaeus, 1758) and C. miniata, both of which prominently reside among “migratory” taxa. The sister taxon of the Scutellastra clade is the Patellidae. Most members of the Patellidae show varying degrees of homing behavior (Hawkins, The Laboratory, Cita- del Hill, pers. comm., 2007). In addition to homing, P. vulgata also forms a home depression, shows aggression (Jones and Baxter, 1986), and is protandric (Le Quesne and Hawkins, 2006). However, other characters such as up shore migration, general- ized food and the lack of gardening (Lewis and Bowman, 1975) suggests a “migra- tory” categorization for P. vulgata as well as for the other northeastern Atlantic and Mediterranean taxa that comprise the genus Patella. Thus, the evolution of the “non- migratory” character suite likely marks an apomorphy at the base of the Scutellastra clade and was not shared with the Patellidae through their last common ancestor. The lack of “non-migratory” characteristics in the Nacellidae Cellana( and Nacella species) further supports this scenario. Large body size is often associated with species that maintain and defend territo- ries (Huntingford and Turner, 1987), and this association is also present here, with the territorial “non-migratory” taxa being significantly larger than non-territorial “migratory” taxa (Table 3). Overall, median shell lengths of most patellid and scutel- lastrid taxa are not significantly different within their respective clades and range between 46.5 mm and 55 mm. The only exception is the clade containing the large Scutellastra mexicana (Broderip and Sowerby, 1829) and the small S. flexuosa where median size is 95 mm. However, the variance in size in the patellid and scutellastrid clades presents a more complex picture, and these patterns are not biogeographically driven. In the patellids, which lack well delineated “non-migratory” taxa, the variance in shell length is 244.5. This in contrast to its sister taxa where the median shell length is only 4.5 mm larger (51 mm compared to 46.5 mm), but the range in shell length is 4 times larger. Furthermore, within the scutellastrids, shell length and variance are 228 BULLETIN OF MARINE SCIENCE, VOL. 81, NO. 2, 2007

Table 3. Mean patellid taxon size (mm) and variance by generalist vs specialist category and biogeographical region.

Category Median Variance P “Migratory” 41.0 269.0 0.0120 “Non-migratory” 73.5 1,743.2 0.0120 NE Atlantic, Mediterranean 46.5 244.5 > 0.05 Indo-Pacific 48.0 4,658.3 > 0.05 Southern Africa 54.0 636.0 > 0.05 Australia 37.0 985.3 > 0.05

similar in both clades with and without large numbers of “non-migratory” species. These patterns arise because of convergent trends in body size in both clades, how- ever, there have been at least three reductions in body size in the basal scutellastrid clade—S. flexuosa (33 mm), S. chapmani (23 mm), and S. aphanes (19 mm)—placing these species in a similar size range as the non-territorial Helcion species. While there are insufficient data to speculate on the selective regimes that drove size reduc- tion in the latter two species, the small body size of S. flexuosa may be linked to the surface irregularities of its habitat on the algal ridge that provides only small areas on to which the limpets can affix, and even then the limpets typically enlarge these surfaces for their home depressions and gardens. The small body size of S. flexuosa may be a character that is a recent adaptation and a reversal of a trait associated with its “non-migratory” legacy. In the second Scutellastra clade, only three of the nine taxa are “non-migratory”, and at least two of these—C. compressa and C. miniata—are likely to be indepen- dent convergences. Cymbula compressa occurs on the stipes of a laminarian alga, and therefore almost by default is zone restricted, feeds on a specific food, has rigid homing with a permanent scar, and is aggressive to interlopers on its food source. Although territorial, it is non-gardening, unlike most other territorial species. And while this habitat is substantially different from the hard substrates on which the other “non-migratory” taxa occur, there appears to have been selection for large body size in C. compressa. The evolutionary trajectories of C. miniata and S. argenvil- lei are more difficult to assess. Cymbula miniata appears to have converged with the other “non-migratory” taxa across a wide suite of traits, however the majority of these traits remain only weakly developed in this species. Scutellastra argenvillei has a mixture of both “migratory” and “non-migratory” traits only one of which (gonad output) is strongly a “migratory” characteristic. Moreover, its sister taxon, Cymbula granularis (Linnaeus, 1758), is one of the most robust of the “migratory” species. But regardless of their respective trajectories, both species show a substantial size increase as compared to most of the other members of this clade. Thus, the median values and similar size variance in the two Scutellastra clades are driven by two opposing trends. In the more basal “non-migratory” taxa, several species are substantially smaller than other members of the clade, reducing median size and increasing the variance for the clade. Interestingly, these putative size reduc- tions have not affected the “non-migratory” behaviors of these species, and S. flexu- osa remains a classic “non-migratory” taxon albeit in miniature. In contrast, in the “migratory” portion of the Scutellastra clade several taxa have independently evolved “non-migratory” behaviors and show substantial size increases relative to the other members of the clade, thus increasing median size and the variance of this clade. LINDBERG: REPRODUCTION, ECOLOGY, AND EVOLUTION OF THE INDO-PACIFIC LIMPET 229

Figure 7. Gardening Scutellastra species. (A) Scutellastra mexicana, Tangola-Tangola, Mexico (Photograph from Book of Bays by William Beebe, reprinted by permission of the publisher). (B) Scutellastra cochlear, Dalebrook, South Africa. (C) Scutellastra kermadecensis, Kermadec Islands, New Zealand. Photograph courtesy of R. Creese. (D) Scutellastra longicosta, Dalebrook, South Africa. The presence of gardening byS. flexuosawas unexpected. Most gardening limpets are large species such as S. mexicana (158 mm) (Fig. 7A), Scutellastra kermadecensis (Pilsbry, 1894) (136 mm) (Fig. 7C), Scutellastra tabularis (Krauss, 1848) (127 mm), S. laticostata (83 mm), Scutellastra longicosta (Lamarck, 1819) (70 mm) (Fig. 7D), and Scutellastra cochlear (Reeve, 1854) (59 mm) (Fig. 7B). Gardening limpets come in two forms. First, on coralline algal substrates gardening limpets typically maintain the garden around the periphery of the shell, and in most cases it is not the coralline algae that is maintained, but rather non-coralline filamentous or finely-branched red algae growing on its surface. The second form is species that maintain algal patch- es over which they graze. These species include S. longicosta, S. tabularis, and S. laticostata. In both forms the limpets increase the productivity of their respective gardens through nutrient enhancement. This is done either through the release of nitrogenous excretions (Plaganyi and Branch, 2000) or through nutrient-rich mucus from the foot which is left on the patch (Connor and Quinn, 1984; Connor, 1986). The patterns present in these studies suggest that periphery gardeners will primar- ily use nitrogenous excretions released along the shell edge, while patch gardeners will be more reliant on mucous enrichment. Both patch and periphery gardeners co- occur in the basal portion of the Scutellastra clade and there does not appear to be any phylogenetic pattern related to their distribution. However, the placement of S. tabularis and S. kermadecensis has yet to be determined. Patch and periphery gardeners can also be recognized morphologically. In most periphery gardeners the front of the shell is extended and has an angular rather than rounded profile when viewed from the dorsal aspect (e.g., S. mexicana, S. kerma- decensis). This morphology reflects a head region that is elongated anteriorly from the foot and viscera forming a distinctive “neck”. This morphology is taken to the 230 BULLETIN OF MARINE SCIENCE, VOL. 81, NO. 2, 2007

Figure 8. Scutellastra species. (A) Scutellastra cochlear, Dalebrook South Africa. Length = 65 mm. (B) Scutellastra sp., Pliocene of Venezuela.­

extreme in S. cochlear where the anterior portion of the shell forms a prominent lobe (Figs. 7B, 8A). This modification may allow these limpets to graze their gar- dens without leaving the vicinity of their home depressions, and also provides a clear ecological character in the fossil record (Fig. 8B). In contrast, patch gardeners lack angular anterior and typically have more rounded profiles (e.g., S. tabularis, S. lati- costata) although several have extended ribs at the shell margin (e.g., S. longicosta, S. chapmani). Although conserved phylogenetically at the base of the scutellastrid clade and bio- geographically across the tropical Indo-Pacific from Central America to South Af- rica, the associated suite of “non-migratory” traits present today in S. flexuosa and related taxa, ultimately decomposed in subsequent divergences in both Australia and Southern Africa. The radiation of Helcion and Cymbula taxa in southern Africa was especially striking as size decreased and “migratory” traits came to dominant in the ancestors of these respective living taxa. The trends towards small size and “migra- LINDBERG: REPRODUCTION, ECOLOGY, AND EVOLUTION OF THE INDO-PACIFIC LIMPET 231 tory” traits also occurred in Australia where the recent Scutellastra peronii (Bla- inville, 1825) co-occurs among lottiid and siphonarid limpets without any notable expression of its belligerent ancestry. The larger pattern documented here is one of a taxon dominated by specialist spe- cies (specific food, restricted habitats, complex behaviors for homing and aggression) giving rise to a generalist clade that subsequently gave rise to other specialists with identical traits albeit in different habitats. While exceptions to the usual expectation of specialist taxa arising from generalists were previously suspected (see Futuyma and Moreno, 1988 and references therein), the use of phylogenetics trees and stan- dardized categories to test these assumptions (Irschick et al., 2005) provides a new and exciting methodology which is showing that transitions in either direction occur in both plants (Armbruster and Baldwin, 1998) and animals (Kelly and Farell, 1998; Nosil and Mooers, 2005; Yotoko et al., 2005). Most of these examples come from terrestrial habitats, but it is likely that similar patterns will be common in marine invertebrate taxa as well.

Acknowledgments

This work was made possible by the mentoring and friendship of John and Vicki Pearse, whose exemplary science and understanding of animals without backbones has been unself- ishly shared with generations of students. I also thank S. Moshel-Lynch, P. Bunje, D. Lindberg, L. Crisostomo, M. Kellogg, and E. Begovic for their early morning assistance at Temae; G. Branch and S. Hawkins for their criticism of the manuscript; C. Reeb, W. Simison, C. Meyer, and E. Begovic for sequence data from their dissertation work; and R. Creese for Fig. 7C. N. Davies, F. and H. Murphy and the staff of the University of California Berkeley’s Richard B. Gump South Pacific Research Station, Moorea provided assistance and logistical support for which I am most grateful. This study was funded, in part, by the Committee on Research, University of California Berkeley, and is contribution 1954 from the University of California Museum of Paleontology.

Literature Cited

Armbruster, W. S. and B. G. Baldwin. 1998. Switch from specialized to generalized pollination. Nature 394: 632–632. Begovic, E. 2004. Population structuring mechanisms and diversification patterns in the Patel- logastropoda of the North Pacific. Ph. D. Diss., Department of Integrative Biology, Univer- sity of California at Berkeley. 291 p. Branch, G. M. 1974a. Ecology of Patella Linnaeus from Cape Peninsula, South Africa. 2. Repro- ductive cycles. Trans. R. Soc. S. Afr. 41: 111–160. ______. 1974b. The ecology of Patella Linnaeus from the Cape Peninsula, South Africa. III. Growth rates. Trans. R. Soc. S. Afr. 41: 161–193. ______. 1975a. The ecology of Patella from the Cape Peninsula, South Africa. IV. Desic- cation. Mar. Biol. 32: 179–188. ______. 1975b. Mechanisms reducing intraspecific competition in Patella spp - migra- tion, differentiation and territorial behavior. J. Anim. Ecol. 44: 575–600. ______. 1976. Interspecific competition experienced by South African Patella species. J. Anim. Ecol. 45: 507–529. ______. 1981. The biology of limpets: physical factors, energy flow, and ecological inter- actions. Oceanogr. Mar. Biol. Annu. Rev. 19: 235–380. Bustamante, R. H., G. M. Branch, and S. Eekhout. 1995. Maintenance of an exceptional grazer biomass in South: subsidy by subtidal kelps. Ecology. 76: 2314–2329. 232 BULLETIN OF MARINE SCIENCE, VOL. 81, NO. 2, 2007

Byrne, M. 2006. Life history diversity and evolution in the Asterinidae. Integr. Comp. Biol. 46: 243–254. Chelazzi, G., G. Santini, and P. Della Santina. 1998. Route selection in the foraging of Patella vulgata (: ). J. Mar. Biol. Assoc. U.K. 78: 1223–1232. Collin, R. 2004. Phylogenetic effects, the loss of complex characters, and the evolution of devel- opment in calyptraeid gastropods. Evolution 58: 1488–1502. Connor, V. M. 1986. The use of mucous trails by intertidal limpets to enhance food resources. Biol. Bull. 171: 548–564. ______and J. F. Quinn. 1984. Stimulation of food species growth by limpet mucus. Sci- ence 225: 843–844. Creese, R. G. 1981. Patterns of growth, longevity and recruitment of intertidal limpets in New South Wales. J. Exp. Mar. Biol. Ecol. 51: 145–171. ______, D. R. Schiel, and M. J. Kingsford. 1990. Sex change in a giant endemic limpet, Patella kermadecensis, from the Kermadec Islands. Mar. Biol. 104: 419–426. Dodd, J. M. 1956. Studies on the biology of limpets. III. Hermaphroditism in the three British species of Patella. J. Mar. Biol. Assoc. U.K. 35: 327–340. Fritchman, H. K., III. 1961. A study of the reproductive cycles in the California Acmaeidae (Gastropoda). Part I. Veliger 3: 57–63. Futuyma, D. J. and G. Moreno. 1988. The evolution of ecological specialization. Annu. Rev. Ecol. Syst. 19: 207–233. Galzin, R. and J. P. Pointier. 1985. Moorea Island, Society Archipelago. Pages 75–101 in B. Dele- salle, R. Galzin, and B. Salvat, eds. Vol 1. French Polynesia coral reefs: Proc. Fifth Int. Coral Reef Congress. Antenne Muséum – EPHE, Mooréa French Polynesia. 556 p. Glenner, H. and M. B. Hebsgaard. 2006. Phylogeny and evolution of life history strategies of the parasitic barnacles (Crustacea, Cirripedia, Rhizocephala). Mol. Phylogen. Evol. 41: 528–538. Gray, D. R. and A. N. Hodgson. 2004. The importance of a crevice environment to the limpet Helcion pectunculus (Patellidae). J. Moll. Stud. 70: 67–72. Hawkins, S. J. and R. G. Hartnoll. 1983. Grazing of intertidal algae by marine-invertebrates. Oceanogr. Mar. Biol. 21: 195–282. Hedegaard, C., D. R. Lindberg, and K. Bandel. 1997. A patellogastropod limpet from the Trias- sic St. Cassian Formation of Italy. Lethaia 30: 331–335. Huntingford, F. A. and A. K. Turner. 1987. Conflict. Chapman and Hall, London, 448 p. Irschick D., L. Dyer, and T. W. Sherry. 2005. Phylogenetic methodologies for studying special- ization. Oikos 110: 404–408. Jones, N. S. 1948. Observations and experiments on the biology of Patella vulgata at Port St. Mary, Isle of Man. Proc. Trans. Liverpool Biol. Soc. 56: 60–77. Jones, A. M. and J. M. Baxter. 1986. The use of Patella vulgata L. in rocky shore surveillance. Pages 265–273 in P. G. Moore and R. Seed, eds. The ecology of rocky coasts. Columbia University Press, New York. 467 p. Kase, T. and Y. Shigeta. 1996. New Species of (Mollusca) From the Creta- ceous of Hokkaido, Japan and Sakhalin, Russia. J. Paleontol. 70: 762–771. Kelley, S. T. and B. D. Farrell. 1998. Is specialization a dead end? The phylogeny of host use in Dendroctonus bark beetles (Scolytidae). Evolution 52: 1731–1743. Kohler, F., T. von Rintelen, A. Meyer, and M. Glaubrecht. 2004. Multiple origin of viviparity in Southeast Asian gastropods (Cerithioidea:Pachychilidae) and its evolutionary implications. Evolution 58: 2215–2226. Koufopanou, V., D. G. Reid, S. A. Ridgway, and R. H. Thomas 1999. A molecular phylogeny of the patellid limpets (Gastropoda : Patellidae) and its implications for the origins of their antitropical distribution. Mol. Phylogen. Evol. 11: 138–156. Kupriyanova, E. K. 2003. Live history evolution in serpulimorph polychaetes: A phylogenetic analysis. Hydrobiologia 496: 105–114. LINDBERG: REPRODUCTION, ECOLOGY, AND EVOLUTION OF THE INDO-PACIFIC LIMPET 233

Le Quesne, W. J. F. and S. J. Hawkins. 2006. Direct observations of protandrous sex change in the patellid limpet Patella vulgata. J. Mar. Biol. Assoc. U.K. 86: 161–162. Lewis, J. R. and R. S. Bowman. 1975. Local habitat-induced variations in the population dy- namics of Patella vulgata L. J. Exp. Mar. Biol. Ecol. 17: 165–203. Lindberg, D. R. 1988. The Patellogastropoda. Malacol. Rev., Suppl. 4: 35–63. ______and K. R. Dwyer. 1983. The topography, formation, and role of the home depres- sion of Collisella scabra (Gould). Veliger 25: 229–234. Lodge, S. M. 1948. Algal growth in the absence of Patella on an experimental strip of foreshore, Port St. Mary, Isle of Man. Proc. Trans. Liverpool Biol. Soc. 56: 78–83. Lowell, R. B. 1984. Desiccation of intertidal limpets: effects of shell size, fit to substratum, and shape. J. Exp. Mar. Biol. Ecol. 77: 197–207. McHugh, D. and P. P. Fong. 2002. Do life history traits account for diversity of polychaete an- nelids? Invertebr. Biol. 121: 325–338. Meyer, C. P. 2003. Molecular systematics of cowries (Gastropoda: Cypraeidae) and diversifica- tion patterns in the tropics. Biol. J. Linn. Soc. 79: 401–459. Niu, C-J., S. Nakao, and S. Goshima. 1998. Energetics of the limpet Lottia kogamogai (Gas- tropoda: Acmaeidae) in an intertidal rocky shore in southern Hokkaido, Japan. J. Exp. Mar. Biol. Ecol. 224: 167–181. NOAA. 2005. Tide Tables 2006: Central and Western Pacific Ocean and Indian Ocean. Inter- national Marine/Ragged Mountain Press (McGraw-Hill). 496 p. Nosil, P. and A. O. Mooers. 2005. Testing hypotheses about ecological specialization using phylogenetic trees. Evolution 59: 2256–2263. Orton, J. H. 1919. Sex phenomena in the common limpet (Patella vulgata). Nature 104: 373–374. ______, A. J. Southward, and J. M. Dodd. 1956. Studies on the biology of limpets II. The breeding of Patella vulgata L. in Britain. J. Mar. Biol. Assoc. U.K. 35: 149–176. Parry, G. D. 1978. Effects of growth and temperature acclimation on metabolic rate in the lim- pet, Cellana tramoserica (Gastropoda: Patellidae). J. Anim. Ecol. 47: 351–368. Plaganyi, E. E. and G. M. Branch. 2000. Does the limpet Patella cochlear fertilize its own algal garden? Mar. Ecol. Prog. Ser. 194: 113–122. Rao, M. B. 1973. Sex phenomenon and reproductive cycle in the limpet Cellana radiata (Born) (Gastropoda: Prosobranchia). J. Exp. Mar. Biol. Ecol. 12: 263–273. Reeb, C. A. 1995. Molecular insights into the evolution of a circumtropical fish (Corypha- ena hippurus) and an Indo-Pacific group of molluscs (Cellana). Ph. D. Diss., University of Hawai‘i. 237 p. Ridgway, T. M., B. A. Stewart, and G. M. Branch. 1999. Limited population differentiation in the bearded limpet Patella barbara (Gastropoda: Patellidae) along the coast of South Af- rica. J. Mar. Biol. Ass. U.K. 79: 639–651. ______, G. M. Branch, B. A. Stewart, and A. N. Hodgson. 2000. Taxonomic status of the ‘Patella miniata’ species complex in southern Africa. Hydrobiologia 420: 103–118. Robson, G. 1986. A new species of South African limpet, Patella aphanes, (Mollusca: Gas- tropoda: Patellidae), with a discussion of P. obtecta Krauss, 1848. Durban Mus. Novit. 13: 305–320. Scheibling, R. E. and R. Black. 1993. The ecology of the giant limpet Patella laticostata on in- tertidal limestone platforms at Rottnest Island, Western Australia. Proc. 5th Int. Mar. Biol. Workshop 2: 565–581. Simison, W. B. 2000. Evolution and phylogeography of New World gastropod faunas. Ph. D. Diss., Department of Integrative Biology, University of California at Berkeley. 202 p. Swofford, D. L. 1998. Paup*. Phylogenetic analysis using parsimony (*and other methods). Ver- sion 4. Sinauer Associates, Sunderland. 302 p. Underwood, A. J. 1979. The ecology of intertidal gastropods. Adv. Mar. Biol. 16: 110–210. Wolcott, T. G. 1973. Physiological ecology and intertidal zonation in limpets (Acmaea): a criti- cal look at “limiting factors.” Biol. Bull. 145: 389–422. 234 BULLETIN OF MARINE SCIENCE, VOL. 81, NO. 2, 2007

Wright, W. G. 1988. Sex change in the Mollusca. Trends Ecol. Evol. 3: 137–140. ______and D. R. Lindberg. 1982. Direct observation of sex change in the patellacean limpet Lottia gigantea. J. Mar. Biol. Assoc. U.K. 62: 737–738. Yotoko, K. S. C., P. I. Prado, C. A. M. Russo, and V. N. Solferini. 2005. Testing the trend to- wards specialization in herbivore-host plant associations using a molecular phylogeny of Tomoplagia (Diptera: Tephritidae). Mol. Phylogen. Evol. 35: 701–711.

Address: (D.R.L.) Department of Integrative Biology and Museum of Paleontology, University of California, Berkeley, California 94720. E-mail: .