Bryozoa) in the Gulf of California1

Home , Bryozoa, Toxopneustes, Toxopneustes roseus

Bryoliths (Bryozoa) in the Gulf of California1

D. W. James,2,4 M. S. Foster,2 and J. O’Sullivan3

Abstract: Populations of Diaperoforma californica (d’Orbigny) bryoliths were dis- covered in rhodolith beds, a sand habitat, and on a cobble bottom in the Gulf of California, Mexico, the ﬁrst known observation of a modern free-living cyclostome bryozoan in the Northern Hemisphere. Densities ranged from a mean of 9.2 to 22.6 individuals/0.06 m2. Bryoliths from the deepest site were irregularly shaped and had the highest variation in shape; those from shallow sites were spheroidal. Water motion and bioturbation move the bryoliths and may determine their morphology. Schizomavella robertsonae (Soule, Soule & Chaney) bryoliths also occurred occasionally in one rhodolith bed sampled. Fossilized bryolith specimens of the cyclostome Diaperoforma californica (d’Orbigny) were found in a Pleistocene deposit near modern habitats.

Bryoliths (free-living bryozoans that colo- channel deposits of the Imperial Formation of nize and surround a pebble, shell fragment, southeastern California (Kidwell and Gyllen- and other objects or are formed by frag- haal 1998) were found. mentation and regeneration) (Cook 1963, Attached colonies of the cyclostome bryo- Winston and Ha˚kansson 1986, Kvitek 1989), zoan Diaperoforma californica (d’Orbigny) have algae (Bosence 1983), and coral (Glynn 1974, been reported from British Columbia to Scoffin et al. 1985) are known to occur world- Costa Rica from below the low-tide level to wide on soft bottoms. Cheilostome and cte- depths over 185 m and are common in the nostome bryoliths are found on sand and Gulf of California (Osburn 1953). This spe- shell-gravel substrates. The Gulf of Califor- cies is morphologically plastic, with shapes nia contains very large populations of rho- ranging from short and palmate to tall with doliths (unattached nongeniculate coralline slender branches. Habitat may affect the mor- algae) (Foster et al. 1997), coralliths (free- phology of D. californica. The branches are living coral) (Reyes-Bonilla et al. 1997), and, usually narrow in proportion to their length as reported here, modern bryoliths in some in deeper waters and shorter, wider, and less locations. In addition, fossil cheilostome erect in exposed habitats (Osburn 1953). bryozoans from upper Pliocene deposits in We describe this species as a bryolith, the Loreto Basin (Cuffey and Johnson 1997, the first description of a modern free-living Dorsey and Kidwell 1999) and Pliocene tidal cyclostome bryozoan in the Northern Hemi- sphere, and present information on its mor- 1 This work was supported in part by National phology, density, cover, and habitat. Geographic Grant No. 5774-96 (to K. Meldahl), Inter- American Institute Grant No. UCAR 597-74025 (to materials and methods M.S.F.), and Moss Landing Marine Laboratories. Manu- script accepted 1 March 2005. 2 Moss Landing Marine Laboratories, 8272 Moss Study Sites Landing Road, Moss Landing, California 95039. Bryoliths were studied at four sites in the 3 Monterey Bay Aquarium, 866 Cannery Row, Mon- terey, California 93940. Gulf of California (Figure 1), and densities 4 Current address: The City of San Diego, Environ- were estimated at two of these, the rhodolith mental Monitoring and Technical Services Laboratory, beds off Punta Chivato (Manto de James) 2392 Kincaid Road, San Diego, California 92101. and off Isla Pata in Bahıá de Los Angeles. Additional observations and specimens were Pacific Science (2006), vol. 60, no. 1:117–124 collected off Isla Pescador in Bahıá de Los : 2006 by University of Hawai‘i Press Angeles and in the Canal de San Lorenzo. All rights reserved Pleistocene Diaperoforma californica (d’Or-

117 Figure 1. Location of study sites in the Gulf of California, Mexico. Bryoliths in the Gulf of California . James et al. 119 bigny) fossil specimens were collected at Isla All bryoliths, algae, and substrate under each Coronado. point were recorded. Bryolith density was Bryoliths were sampled in the rhodolith determined at Manto de James in June 1997 bed at Manto de James in March 1997 and (n ¼ 5) and Isla Pata in June 1998 (n ¼ 10). June 1997. This rhodolith bed occurred on All bryoliths were counted in 0.25 by 0.25 m a shallow (6.7–7.3 m) flat sandy bottom ex- randomly placed quadrats. posed to strong tidal and wind-driven cur- Sphericity of haphazardly collected bryo- rents. Bryoliths were mostly Diaperoforma liths was calculated to determine if mor- californica (Figure 2A,B,C,E), with Schizoma- phology varied among locations. Variation vella robertsonae (Soule, Soule & Chaney) in shape was characterized by measuring the occurring occasionally (Figure 2D). Rhodo- longest dimension (L), the intermediate di- liths consisted of Lithophyllum margaritae mension measured at the widest point 90 to (Hariot) Heydrich and Neogoniolithon trichoto- the midpoint of the axis of L (I), and the mun (Heydrich) Setchell & Mason and cov- shortest dimension measured at 90 to the ered 50–60% of the bottom. Sparse foliose midpoint of I (S). These lengths were mea- red algae were present. sured with a digital caliper (to the nearest 1/ Diaperoforma californica bryoliths were ob- 100 of a mm) and were used to calculate the served on the southwest side of Isla Pescador, coefficient of variation (CV; sample standard Bahıá de Los Angeles, in May 1994. High deviation/mean of L, I, and S multiplied by densities of bryoliths 1.5–3 cm in diameter 100; n ¼ 3), a measure of sphericity (Foster occurred in 2–3 m of water in swell troughs et al. 1997). on the sand. The troughs were several meters The difference between individual mean long and located 20–30 m offshore. Various bryolith lengths at Manto de James in March algae, including rhodoliths, and organic de- and June was tested for significance (indepen- bris filled the troughs and covered the bryo- dent sample t-test) based on the means of all liths. three axes. The means were log(X) trans- A bryolith field consisting of D. californica formed to satisfy the assumption of normal- was sampled at Isla Pata, Bahıá de Los Ange- ity. Normality was determined graphically, les, in June 1998. Bryoliths occurred on a and homogeneity of variances was tested us- cobble bottom in a band 8.5–11 m wide and ing the F test (Zar 1984). 53 m long that paralleled the curving shore- Differences in bryolith sphericity (CV) line at a depth of 2.4 m and 15 m from shore. among sites were tested with analysis of vari- No rhodoliths were observed. ance (ANOVA). The assumption of normally Diaperoforma californica bryoliths from the distributed error terms was tested using a Canal de San Lorenzo were collected from normality plot of residuals. Levene’s test re- a depth of 14 m in a rhodolith bed between vealed minor homoscedasticity of error terms, Isla Espı´ritu Santo and Calerita in June 1997 which was corrected using an arcsine–square (rhodolith bed described in Foster et al. root transformation. [1997]). The rhodolith bed consisted of rhodoliths (primarily Lithophyllum margaritae) results grading into white calcareous sand at 12– 19 m depth. Bryoliths were common in the habitats Bryolith and algal cover was determined at sampled. The percent cover of bryoliths Manto de James in June 1997 using a point (mean % G SE) at Manto de James in June quadrat consisting of a 1-m bar with a 1.5-m was 15 G 4%. Rhodolith cover was 64 G 6%, string attached at both ends with 10 points foliose red algal cover was 12 G 6%, and sand tied on the string (Foster 1982). Five quadrats cover was 8 G 3%. Bryolith density at Manto were randomly placed and sampling points de James in June was 9.2 G 2.33 individuals/ (total ¼ 20) were positioned on each side of 0.06 m2 (mean G SE) and 22.6 G 2.95 the bar. Bars were placed where the rhodo- individuals/0.06 m2 at Isla Pata. Although liths covered 50–60% of the sand bottom. quantitative data were not collected at Isla Figure 2. Representative bryoliths from study sites in the Gulf of California. A, Isla Pescador; B, Isla Pata; C, Manto de James; D, Manto de James; E, Canal de San Lorenzo; F, Isla Coronado (fossil). All are Diaperoforma californica except D, which is Schizomavella robertsonae. Bryoliths in the Gulf of California . James et al. 121

TABLE 1 Shape and Size of Bryoliths

Site Depth (m) Longest Axis Intermediate Axis Shortest Axis CV n

Isla Pata 2 43.69 G 2.29 39.35 G 2.23 33.82 G 2.37 14 G 39 Manto de James March 7 22.89 G 1.32 19.46 G 1.27 15.89 G 1.42 19 G 36 June 7 17.91 G 1.27 14.65 G 0.96 11.19 G 0.81 23 G 214 San Lorenzo 12–19 19.26 G 0.55 15.47 G 0.71 11.45 G 0.69 27 G 320

Note: Lengths, mean G SE mm (untransformed data); CV, coefﬁcient of variation (%; mean G SE; untransformed data). Sphericity increases as CV decreases.

Pescador and the Canal de San Lorenzo, bryoliths from both Isla Pescador and Isla bryoliths were abundant at Isla Pescador but Pata. rare at the Canal de San Lorenzo. Bryoliths differed in size and shape at the discussion various sites (Figure 2). Size (longest dimension) ranged from 30.8 to 56.7 mm at Isla Although there are numerous species of free- Pata, 10.9 to 28.3 mm at Manto de James, living cheilostomate bryozoans, this record of and 14.1 to 22.8 mm at San Lorenzo (Table Diaperoforma californica is the first known rec- 1). Bryoliths from Isla Pata and Isla Pescador ord of a modern free-living cyclostome bryo- (visual observation only) were highly spheroi- zoan occurring in the Northern Hemisphere. dal; those from Manto de James were mostly Taylor and Gordon (2003) found a modern spheroidal with some elongation of branches. cyclostome bryolith from New Zealand. Bryoliths from San Lorenzo were highly Other free-living benthic cheilostomate bryo- branched, irregularly shaped, and had a ten- zoans occur on sand (Cook 1963, Cook and dency toward spheroidal shapes. Bryoliths Chimonides 1978, Winston and Ha˚kansson from Isla Pata were the largest and had the 1986, Kvitek 1989), but D. californica is lowest CV (Table 1). The mean size of bryo- unique in that it occurs on and around hard liths at Manto de James in June was signifi- substrates, such as rhodoliths and cobble, as cantly smaller than in March (all three axes well as on sand. Rhodoliths and free-living used to determine mean; independent sample hermatypic corals, in addition to the afore- t-test: t ¼ 2:75, df ¼ 18, P ¼ 0:013). mentioned bryoliths, occur together on sand A significant difference was found among bottoms in the Gulf of California (Steller sites for the CV (ANOVA: F ¼ 4:89; df ¼ 3, and Foster 1995, Foster et al. 1997, Reyes- 45; P ¼ 0:005). Post hoc tests showed that the Bonilla et al. 1997). CV for bryoliths at Isla Pata was significantly Conditions in the Gulf of California are fa- smaller than those for Manto de James vorable for the propagation of free-living spe- in March (P ¼ 0:045) and San Lorenzo cies. The formation and turning of rhodoliths (P ¼ 0:003). There was no significant differ- and coralliths in the region are caused by wave ence between the CVs for bryoliths at Manto action, currents, and bioturbation (Steller and de James in March and June (P ¼ 0:739). The Foster 1995, Foster et al. 1997, Reyes-Bonilla high CV at San Lorenzo reflects the observa- et al. 1997, Marrack 1999, James 2000). Bryo- tion of irregular bryolith shapes in this cur- liths may also be formed by these processes. rent bed. The presence of bryoliths with nuclei such as The fossil Diaperoforma californica was small pebbles at Isla Pata and Isla Pescador highly spheroidal (Figure 2F). The size, indicates that water motion may have caused shape, and morphology of the fossil D. cali- them to be moved about and develop a spher- fornica are similar to those of the D. californica oidal shape after initial settlement on their 122 PACIFIC SCIENCE . January 2006 substrate. At Manto de James, bryoliths were there. This morphology is more stable than observed being bioturbated by the sea urchin densely branched forms (Bosence 1983) and Toxopneustes roseus and crabs. Currents were may reduce movements by strong currents. often quite strong at Manto de James and Ca- Although most bryoliths in the Gulf of nal de San Lorenzo (@25 cm/sec [Marrack California likely grew from regenerated frag- 1999]) and may have been sufficient to occa- ments, bryoliths may form from larvae that sionally move bryoliths. settle on sand and gravel grains and other Bryoliths and bryozoans are known to re- hard particulate substrates (Rider and Enrico produce through fragmentation by physical 1979, Winston and Ha˚kansson 1986, Win- and biological processes and asexual regener- ston 1988). Occasional bryoliths were ob- ation in addition to larval development (Blake served growing on rhodoliths at Manto de 1976, McKinney 1983, Winston and Ha˚kans- James. Bryoliths were present at Isla Pescador son 1986, Kvitek 1989, Ostrovsky 1997). and Isla Pata that formed by larvae settling on Bryoliths at Manto de James were signifi- rocks or pebbles. cantly smaller in June 1997 than in March Although bryoliths were occasionally 1997 and appear to have fragmented, perhaps buried under rhodoliths, they appeared by a storm event and/or bioturbation. It is healthy. Live encrusting bryozoans have been highly likely that these fragments survived found living 16 cm below the surface in sandy because many healthy bryoliths at Manto de sediments (Winston and Ha˚kansson 1986), James showed evidence of breakage. The and bryozoans are capable of feeding with majority of the bryolith population at San their lophophores in interstitial spaces (Ha˚- Lorenzo probably also propagate through kansson and Winston 1985). Water-born fragmentation. All of the bryoliths collected food may still reach bryoliths buried under at this site had fragments missing from break- rhodoliths and enable them to survive burial age. These bryoliths were similar to a popula- until being moved back to the surface tion of the free-living bryozoan Selenaria in through bioturbation or water motion. Bryo- New Zealand, where all colonies had devel- liths are much lighter than rhodoliths and oped from regeneration (Cook and Chimo- should end up above rhodoliths after both nides 1978). are disturbed. Vegetative reproduction through fragmen- The trends in the coefficient of variation tation may be the most favorable form of suggest that higher variation occurs in deeper reproduction for bryozoans living on soft habitats dominated by currents. This is simi- sediments. Reproduction through fragmenta- lar to the results of Foster et al. (1997) for tion may counteract high larval and juvenile rhodoliths, where the highest coefficients of mortality and allow rapid recovery from epi- variation occurred in the San Lorenzo rhodo- sodic storm disturbances. Asexual reproduc- lith bed, as well as other current beds. The tion is highly developed in free-living corals highly branched shape of bryoliths at San (Highsmith 1982) and free-living bryozoans Lorenzo may reduce water drag and allow in- (Winston 1988) living on sandy substrates. dividuals to lock together and prevent them The branching pattern and shape of the from being swept out of the area by currents. bryolith Diaperoforma californica may increase The spherical shape of bryoliths at Isla its survival. Coral fragments well developed Pescador and Isla Pata may be caused by in three dimensions are most conducive to more frequent turning. Sphericity in bryo- survival after being broken off the main col- liths (Rider and Enrico 1979), rhodoliths ony and lying on sand (Highsmith 1982). (Bosellini and Ginsburg 1971, Bosence 1976), The same should hold true for D. californica, and coralliths (Glynn 1974) is thought to be, because part of a bryolith fragment would be in part, related to turning frequency. Coral- off the substrate regardless of how it was rest- lith survival was found to be positively cor- ing on the bottom, making it less likely to be related with sphericity and may confer a buried or scoured. Bryoliths at Canal de San selective advantage for individuals living in Lorenzo resembled the fruticose rhodoliths disturbed habitats (Lewis 1989). The spheroi- Bryoliths in the Gulf of California . James et al. 123 dal shape of these individuals may enable Bosellini, A., and R. N. Ginsburg. 1971. them to survive rolling about in swell troughs. Form and internal structure of recent algal Bryozoan growth form should assist in fa- nodules (rhodolites) from Bermuda. J. cies analyses of fossil deposits. The discovery Geol. 79:669–682. of a fossil cyclostome bryolith Diaperoforma Bosence, D. W. J. 1976. Ecological studies californica indicates that these bryoliths have on two unattached coralline algae from been present in the Gulf of California since western Ireland. Palaeontology 19:365– at least the Pleistocene. Previous facies analy- 395. ses (except Balson and Taylor [1982]) have ———. 1983. The occurrence and ecology of not considered the existence of free-living cy- recent rhodoliths: A review. Pages 225– clostome bryozoans. Diaperoforma californica 242 in T. M. Peryt, ed. Coated grains. bryoliths are an exception to current classifi- Springer-Verlag, Berlin. cations and models of bryozoan growth forms Cook, P. L. 1963. Observations on live lunu- (Nelson et al. 1988, Bone and James 1993, litiform zoaria of Polyzoa. Cah. Biol. Mar. Hageman et al. 1997). 4:407–413. Bryolith growth form may aid in re- Cook, P. L., and P. J. Chimonides. 1978. Ob- constructing the habitat in which they lived. servations on living colonies of Selenaria Fossil bryoliths with a high coefficient of vari- (Bryozoa, Cheilostomata). I. Cah. Biol. ation may be indicative of deep habitats that Mar. 19:147–158. were dominated by currents. Balson and Cuffey, R. J., and M. E. Johnson. 1997. Taylor (1982) found four species of fossil cy- Bryozoan nodules built around andesite clostome bryoliths from the Coralline Crag clasts from the upper Pliocene of Baja (Pliocene formation) that were usually spher- California: Paleoecological implications oidal in shape. These bryoliths were present and closure of the Panama Isthmus. in facies where bioturbation may have re- Pages 111–117 in M. E. Johnson and J. sulted in their turning but were absent in tur- Ledesma-Va´zquez, eds. Pliocene carbo- bulent sandwave facies. nates and related facies flanking the Gulf of California, Baja California Sur, Mexico. acknowledgments Geol. Soc. Am. Spec. Pap. 318. Dorsey, R. J., and S. M. Kidwell. 1999. Mixed Many thanks to Rafael Riosmena-Rodriquez carbonate-siliciclastic sedimentation on a for his collaboration and logistical support, tectonically active margin: Example from Judy Winston for her bryozoan identifica- the Pliocene of Baja California Sur, Mex- tions, and Rick Rowe for his digital photogra- ico. Geology 27:935–938. phy. Foster, M. S. 1982. The regulation of macro- algal associations in kelp forests. Pages Literature Cited 185–205 in L. Srivastava, ed. Synthetic and degradative processes in marine mac- Balson, P. S., and P. D. Taylor. 1982. rophytes. Walter DeGruyter, Berlin. Palaeobiology and systematics of large Foster, M. S., R. Riosmena-Rodriguez, D. L. cyclostome bryozoans from the Pliocene Steller, and W. J. Woelkerling. 1997. Coralline Crag of Suffolk. Palaeontology Living rhodolith beds in the Gulf of Cali- (Lond.) 25:529–554. fornia and their implications for paleo- Blake, D. B. 1976. Functional morphology environmental interpretation. Pages 127– and taxonomy of branch dimorphism in 139 in M. E. Johnson and J. Ledesma- the Paleozic bryozoan genus Rhabdomeson. Va´zquez, eds. Pliocene carbonates and re- Lethaia 9:169–178. lated facies flanking the Gulf of California, Bone, Y., and N. P. James. 1993. Bryozoans Baja California Sur, Mexico. Geol. Soc. as carbonate sediment producers on the Am. Spec. Pap. 318. cool-water Lacepede Shelf, southern Aus- Glynn, P. W. 1974. Rolling stones among tralia. Sediment. Geol. 86:247–271. the Scleractinia: Mobile coralliths in the 124 PACIFIC SCIENCE . January 2006

Gulf of Panama. Proc. 2nd Int. Coral Reef plication of bryozoan zoarial growth-form Symp. 2:183–198. studies in facies analysis of non-tropical Hageman, S. J., Y. Bone, B. McGowran, carbonate deposits in New Zealand. Sedi- and N. P. James. 1997. Bryozoan colonial ment. Geol. 60:301–322. growth-forms as paleoenvironmental indi- Osburn, R. C. 1953. Bryozoa of the Pacific cators: Evaluation of methodology. Palaios coast of America. Part 3. Cyclostomata, 12:405–419. Ctenostomata, Entoprocta and Addenda. Ha˚kansson, E., and J. E. Winston. 1985. In- Allan Hancock Pacific Expeditions terstitial bryozoans: Unexpected life forms 14:613–841. in a high energy environment. Pages 125– Ostrovsky, A. N. 1997. Rejuvenation in colo- 134 in C. Nielsen and G. P. Larwood, eds. nies of some Antarctic tubuliporids (Bryo- Bryozoa: Ordovician to Recent. Olsen & zoa, Stenolaemata). Ophelia 46:175–185. Olsen, Fredensborg, Denmark. Reyes-Bonilla, H., R. Riosmena-Rodriguez, Highsmith, R. C. 1982. Reproduction by and M. S. Foster. 1997. Hermatypic corals fragmentation in corals. Mar. Ecol. Prog. associated with rhodolith beds in the Gulf Ser. 7:207–226. of California, Me´xico. Pac. Sci. 51:328– James, D. W. 2000. Diet, movement, and 337. covering behavior of the sea urchin Toxop- Rider, J., and R. Enrico. 1979. Structural and neustes roseus in rhodolith beds in the Gulf functional adaptations of mobile anascan of California, Mexico. Mar. Biol. (Berl.) ectoproct colonies (ectoproctaliths). Pages 137:913–923. 297–319 in G. P. Larwood and M. B. Ab- Kidwell, S. M., and E. D. Gyllenhaal. 1998. bott, eds. Advances in bryozoology. Aca- Symbiosis, competition, and physical demic Press, New York. disturbance in the growth histories of Scoffin, T. P., D. R. Stoddart, A. W. Tud- Pliocene cheilostome bryoliths. Lethaia hope, and C. Woodroffe. 1985. Rhodoliths 31:221–239. and coralliths of Muri Lagoon, Rarotonga, Kvitek, R. G. 1989. Hydrodynamic morphol- Cook Islands. Coral Reefs 4:71–80. ogy and behavior of a free-living sediment- Steller, D. L., and M. S. Foster. 1995. Envi- dwelling bryozoan, Alcyonidium disciforme ronmental factors influencing distribution (Smitt). J. Exp. Mar. Biol. Ecol. 125:13– and morphology of rhodoliths in Bahia 32. Concepcion, B.C.S., Mexico. J. Exp. Mar. Lewis, J. B. 1989. Spherical growth in the Ca- Biol. Ecol. 194:201–212. ribbean coral Siderastrea radians (Pallas) Taylor, P. D., and D. P. Gordon. 2003. En- and its survival in disturbed habitats. Coral demic new cyclostome bryozoans from Reefs 7:161–167. Spirits Bay, a New Zealand marine- Marrack, E. C. 1999. The relationship be- biodiversity ‘‘hotspot.’’ N. Z. J. Mar. tween water motion and living rhodolith Freshwater Res. 37:653–669. beds in the southwestern Gulf of Califor- Winston, J. E. 1988. Life histories of free- nia, Me´xico. Palaios 14:159–171. living bryozoans. Natl. Geogr. Res. McKinney, F. K. 1983. Asexual colony multi- 4:528–539. plication by fragmentation: An important Winston, J. E., and E. Ha˚kansson. 1986. The mode of genet longevity in the carbonifer- interstitial bryozoa fauna from Capron ous bryozoan Archimedes. Paleobiology Shoal, Florida. Am. Mus. Novit. 2865:1– 9:35–43. 50. Nelson, C. S., F. M. Hyden, S. L. Keane, Zar, J. H. 1984. Biostatistical analysis. Pren- W. L. Leask, and D. P. Gordon. 1988. Ap- tice Hall, Englewood Cliffs, New Jersey.