Ecological Observations on Dialommus Juscus (Labrisomidae), the "Four-Eyed Blenny" of the Gahipagos Islands1

JfJRGEN NIEDER2

AB~TRAC~: Information is presented on the behavior, abundance, and distri butIOn of Dzalommus fuscus Gilbert in its intertidal habitat on the island of Santa Cruz, Galapagos Archipelago, Ecuador.

IT IS SURPRISING THAT Dialommus fuscus Gil- Behavior bert, a labrisomid endemic to the Galapagos Islands with a vertical eye division (whence As soon as incoming waves covered the came the name "four-eyed blenny"), has rock flat only approximately 50% ofthe time, been studied almost exclusively for its unique D. fuscus emerged from rock pools. The dark eye structure (Munk 1970, Stevens and Par- color of the fish provided excellent camou sons 1980), and its natural history is only in- flage on the black stone substrate. Individual cidentally referred to (Clark 1936). This note fish moved to the top of basalt rocks and presents some observations made in July appeared to search for food. Occasionally 199.1 on the behavior, abundance, and distri- they bit the rock surface, presumably at some butIOn of D. fuscus in its intertidal habitat on prey (which in itself was invisible to the the island of Santa Cruz, Galapagos Archi- observer because of the distance). The fish pelago, Ecuador. The observation site was a clearly did not scrape off algae from the flat reef zone west of Tortuga Bay, 3 km rocks. No particular posture of their heads from Puerto Ayora, Santa Cruz, Galapagos, could be distinguished in the moment of between the beach of Tortuga Bay and the picking at their prey, and D. fuscus did not entrance to the adjoining lagoon, at 0° 45' S move the head before biting. and 90° 25' W. The fish also searched for prey in the tide The black basaltic rock flat of approxi- pools, slowly searching the pool wall surface mately 200 m length (along the natural dam and occasionally picking at some prey. The composed ofbasalt boulders) and 25 m width he~d was kept straight forward, the mouth is flooded during high tide. A tidal difference facmg the prey. of approximately 1 m leaves the rock flat Dialommus fuscus also jumped from rock dry during low tides. Shallow depressions ~o ro~k or on top of water surfaces, flicking form rock pools where D. fuscus and a goby Its tad so rapidly that it "surfed" across the (Bathygobius [G. Merlen, pers. comm.]) can water, very much like a flat stone thrown at be found. a low angle across a water surface, touching The uppermost fringe of the section ob- the water surface every 20 to 30 em. served (12 m wide) was dry 50% of the time Most terrestrial excursions took place 2_hLafteLhigh tide (but _r~:Reatedly flooded where the rock flat was still frequently by waves). The lower limit of the section-was -was~ed--over--by-waves~Qn-aveFage-the-ex dry 50% of the time 5 hr after high tide. cursIOns lasted no more than about 5 min, because the fish returned to pools of water frequently (either by way of an incoming wave or by slowly crawling into the water 1 Manuscript accepted 5 August 1998. a.gain). Fish did not stay in the water for pe 2 Botanisches Institut,· Universitat Bonn Meck enheimer Allee 170, 53115 Bonn, Germany (teL 49-228 rIods longer than approximately 5 min either. 467238; fax 49-228-733120; E-mail jnieder@uni On overcast days they may, however, spend bonn.de). 15-20 min out of water, always in the splash 286 Dialommus fuscus of the Galapagos Islands-NIEDER 287

zone and sometimes among weeds such as Worldwide, comparatively few fish species Ulva (G. Merlen, pers. comm.). Peak terres leave the water for terrestrial sojourns. The trial activity of D. fuscus was observed where amphibious behavior of D. fuscus is most waves occasionally covered the substratum. similar to that of Mnierpes macrocephalus The behavior of D. fuscus reflects a limited (Labrisomidae) of the eastern tropical Pacific ability to endure the physiological stresses of Ocean (Graham 1973). Both fish species terrestrial sojourns. spend only short periods of time outside the No interactions between individuals were water. Bennett and Griffiths (1984) discovered noticed. Distance between individuals was that in rock pools on the coast of South never smaller than 0.5 m, but no active avoid Africa, pool size and number of fishes in ance of intraspecific contact was noticed. habiting each pool were positively correlated. At very low tide, excursions out of water The number of juvenile Coryphoblennius were restricted to the vicinity of tide pools, galerita rises with tide pool volume on the and the presence of D. fuscus in the area was coast of Portugal (Nieder 1993). In the Gala much less conspicuous. pagos observation area, this is true for gobies but not for D. fuscus. Conspicuously, their number is independent of pool volume. Distribution in Tide Pools Water quality (temperature, oxygen content) Ten tide pools did not contain any fishes. probably does not affect the semiamphibious The remaining 23 tide pools contained a total D. fuscus to the same degree as it affects of 34 D. fuscus, 37 gobies, and 1 juvenile nonamphibious fishes such as the gobies. Abudefduf sp. The gobies were never seen Dialommus fuscus becomes accustomed to leaving the water. There were usually not human presence remarkably quickly and can more than three D. fuscus in each occupied be approached with little difficulty (G. Mer tide pool (in one pool three fish of over 6 cm len, pers. comm.). It should be rewarding to total length and six juvenile fish under 2 cm study this unique fish in the context of the total length were found taking refuge under a simply structured intertidal fish community stone). of the Galapagos Islands.

Abundance ACKNOWLEDGMENTS Abundance was estimated to be as high as one fish per square meter. Whenever a wave I thank Rodrigo H. Bustamante, Head of washed over an individual, it did not remain Marine Research at the Charles Darwin in the same place, but afterward could be Research Station, Santa Cruz, Galapagos, seen in a pool of water where the wave pre- very much for his kind help and useful hints. sumably had carried it. The water turbulence, Special thanks are given to Godfrey Merlen however, made it impossible to distinguish (Charles Darwin Research Station) for ex between active and passive movement. tremely valuable advice and observations. There was no correlation between tide C. Dieter Zander (Universitat Hamburg), pool volume and number of D. fuscus Karen Martin (Pepperdine University), and (r = 0.01, P < 0.05), but a strong correla- W. Barthlott (Universitat Bonn) gave valu- -tion-between-tide--pool-volume- ana-number-- able eomments~and-helped--to-improve-the of gobies (r = 0.89, P < 0.05). Gobies were first draft of the paper. much more frequent in larger tide pools, whereas D. fuscus were found in both small and large tide pools. There seemed to be no LITERATURE CITED negative interference between gobies and D. fuscus: no significant correlation between BENNETT, B. A., and C. L. GRIFFITHS. 1984. number of gobies and number of D. fuscus Factors affecting the distribution, abun (r = 0.13, P < 0.05) was found. dance and diversity of rock-pool fishes on 288 PACIFIC SCIENCE, Volume 53, July 1999

the Cape Peninsula, South Africa. S. Afr. fish Dialommus fuscus (Pisces, Blennioidei, J. Zool. 19 (2): 97-104. Clinidae). Vidensk. Medd. Dan. Naturhist. CLARK, R. W. 1936. The Templeton Crocker Foren 13: 17-24. expedition of the California Academy of NIEDER, J. 1993. Distribution of juvenile Sciences, 1932. No. 29. New and note b1ennies (Pisces, Blenniidae) in small tide worthy fishes. Proc. Calif. Acad. Sci. 21 pools: Result of low-tide lottery or strate (29): 383-396. gic habitat selection? Bonn. Zool. Beitr. 44 GRAHAM, J. B. 1973. Terrestrial life of the (3-4): 133-140. amphibious fish Mnierpes macrocephalus. STEVENS, J. K., and K. E. PARSONS. 1980. A Mar. Biol. (Berl.) 23: 83-91. fish with a double vision. Nat. Rist. 89 (1): MUNK, O. 1970. The eye of the "four-eyed" 62-67.