Stygologia 2 (112) 1986, E. J. Brill, Leiden MESOZOIC RELICTS IN MARINE CAVES OF BERMUDA RAYMOND B. MANNING ') C. W. HART, JR.') & T. M. ILIFFE2) SUMMARY Evidence is presented supporting the hypothesis that some of the invertebrates inhabiting marine caves in Bermuda and the Caribbean may have invaded those systems before the opening of the North Atlantic in the Jurassic. It is suggested that the cave havitats provide a potential con- tinuum of habitats from the deep-sea to above sea level fresh waters, and that these habitats could have been entered anywhere in the water column. On prCsente des donntes B l'appui de I'hypothbse que certains des inverttbrhs habitant des grot- tes marines aux Bermudes et dans les Caraibes auraient pu envahir ces systbmes avant l'ouver- ture de 1'Atlantique Nord au Jurassique. On suggbre que le domaine des grottes fournit un con- tinuum potentiel d'habitats, depuis les eaux abyssales jusqu'aux eaux douces au-dessus du niveau de la mer, et que ces habitats ont CtC abordables pour la faune B n'importe quel endroit de la colonne d'eau. INTRODUCTION In recent years there has been a marked increase in studies of the occur- rence, biology, and distribution of invertebrates, especially crustaceans, that are found in anchialine pools and marine caves. These include a variety of groundwater habitats in rock, occurring primarily on oceanic islands. One aspect of the findings of these and related, earlier studies has been the puzzling and often inexplicable distribution patterns of some of these crustaceans: the caridean shrimp genus Procaris occurring on Ascension Island in the south Atlantic (Chace & Manning, 1972) and later discovered in Hawaii (Holthuis, 1973); other shrimps from a limestone fissure that recently opened in the southern Sinai as well as from lava pools on Hawaii and elsewhere (Holthuis, 1973); the subsequent discovery of one of the genera of these shrimps, Calliasmata, in the Dominican ReP ublic (Chace, 1975); the occurrence of a species of Munidopsis, a deep-water galatheid genus, from a recently formed I) Department of Invertebrate Zoology, National Museum of Natural History, Smithsonian In- stitution, Washington, DC 20560, U.S.A. 2, Bermuda Biological Station, Ferry Reach 1-15, Bermuda. lava tube in the Canary Islands (Wilkens & Parzefall, 1974); the crustacean class Remipedia, described from a limestone cave in the Bahamas (Yager, 1981) and later found in the same lava tube containing Munidopsis in the Canary Islands (Iliffe, Wilkens, Parzefall & Williams, 1984); the occurrence of a genus of subterranean mysids, Stygiomysis, in southern Italy and in caves in the Bahamas and the Caribbean (Bowman, 1976; Bowman, Iliffe & Yager, 1984); and the distribution of a genus of subterranean amphipods, Pseudoniphargus, found around the Mediterranean Basin and some eastern Atlantic islands (Stock, 1980) and in Bermuda (Sket & Iliffe, 1980). There are many other examples of such distribution patterns in the literature. Unlike the shrimps and the galatheid, many of these animals lack distributive larval stages. In a review of the biology of anchialine shrimps from the Indo-West Pacific, Maciolek (1983: 115) commented that "most if not all of these hypogeal shrimps could have widespread and possibly longstanding populations in submerged as well as emergent rock of the tropical Indo-Pacific," and (1983: 116) "This broadened habitat hypothesis allows that the shrimps could occur in the groundwaters of many isolated and archipelagic islands where they have not yet been found, in shallow reefs and seamounts, and possibly in suitable rock of continental shelves." Longstanding is a key word, groundwater is another. We also had come to a similar conclusion, that some of the crustaceans oc- curring in caves and groundwaters in the Caribbean had been in those habitats before the Tertiary, and that some were closely tied to the deep sea (Iliffe, Hart & Manning, 1983). The occurrence of these animals in Pleistocene limestone rock of fairly recent origin tends to obscure the possibility that these organisms could have inhabited these groundwater habitats for millions of years. Using the term "cave" for these habitats on Bermuda and elsewhere is misleading, for there are not only subterranean caverns but also more widespread flooded interstices in the rock, interconnected with the caverns. Elsewhere (Hart, Manning & Iliffe, 1985) we have suggested that the term crevicular be used for these groundwater habitats in rock. These habitats may occur at the surface in the limestone cap, where the inhabitants are most often encountered because of limitations in sampling methodologies, and the habitats may also occur anywhere on the submerged surfaces of the volcanic pedestals that attach oceanic islands to the sea bed. Subterranean habitats are undoubtedly present in submerged oceanic ridges and other undersea features, including sea mounts. These habitats can be entered anywhere in the water column, providing direct access to habitats with environmental conditions similar to those found in the deep sea. Once invaded, all interconnecting habitats within the rock, including those above sea level in fresh water, are open to the colonizers. Thus habitats ranging from seeps to underground caverns to exposed anchialine pools to fresh water wells are available to 158 BERMUDA MARINE CA VE SYMPOSIUM members of this fauna. There are also potential direct connections between marine crevicular habitats with the oceanic water of the abyss. BACKGROUND The following information is from Sket & Iliffe (1980), who summarized in- formation on Bermuda's marine caves and provided preliminary findings on the cave fauna. About 150 marine caves are known from the Bermuda Islands. They are situated in a limestone cap of Pleistocene age overlying a volcanic platform that is now about 75 meters below present sea level. The island was formed by volcanic activity on the Mid-Atlantic Ridge in the late Cretaceous, about 110 million years ago. At that time the island was closest to the Euro- pean land mass. Although Bermuda has never been part of the North American land mass, it apparently drifted westward from the Mid-Atlantic Ridge with the North American Plate. Because the caves of Bermuda, the Bahamas, and other islands in the Caribbean are generally of Pleistocene age, those studying their fauna generally have sought explanations as to how cave colonization might have occurred in fairly recent times. We suggest that the origins of Bermuda's cave fauna are less closely related to the origin of the caves than to the origin of Bermuda itself in the Cretaceous. We believe that the largely puzzling distribution patterns of crevicular orgnanisms that lack pelagic larvae might best be explained by the thesis that at least some elements of the fauna of marine caves in Bermuda and elsewhere in the Atlantic are Mesozoic relicts that colonized crevicular habitats in rock along the Mid-Atlantic Ridge as the continents separated. These would not be the first examples of Mesozoic relicts associated with a submerged oceanic ridge. Newman (1979) described a scalpellid barnacle from the East Pacific Rise and considered it to be a Mesozoic relict. Bermuda today is near the winter isotherm of 20°C; it has the northernmost extension of coral reefs in the west Atlantic. As Briggs (1974: 61) commented, "nowhere else in the world does a tropical fauna occur at such a high latitude (32"15')." However, during winter, the sea surface temperature may drop to as low as 15°C. During Pleistocene glacial periods, the sea surface temperature was lowered by as much as 6" to 9°C (Briggs, 1966), enough to have caused extinctions of at least some shallow water tropical taxa. Briggs (1966, 1970) has pointed out that various studies have demonstrated that the North Atlantic shallow water fauna has undergone a general "depauperization," and that the Bermudas as well as other North Atlantic islands have a very low level of endemism. This is probably the result of mass extinctions due to cooling of sea surface temperatures during glacial periods. Iliffe, Hart & Manning (1983) noted that the levels of endemism of Ber- muda's cave fauna (30%) are considerably higher than endemism among lit- toral species (2% for the decapods; Markham & McDermott, 1980). They fur- ther suggested (1983: 141) that some of Bermuda's cave species represented STYGOLOGIA 2 (112) 1986 "groups that survived on submergent and emergent sea mounts along the Mid-Atlantic Ridge since the middle Mesozoic," while other species in the caves were related to the fauna of the deep-sea. Even though inshore surface water temperatures around Bermuda may fluc- tuate between 15°C in winter to more than 30°C in summer, within the caves, at depths below three meters or so, the water temperature remains relatively constant, all year, at about 20.5"C (lliffe, Hart & Manning, 1983), suggesting the possibility of geothermal heating. If the deeper waters of the caves re- mained at relatively higher temperatures during the Pleistocene glacial periods, the caves could have served as warm thermal refuges through the periods of reduced sea temperature. Sket & Iliffe (1980) summarized information on the fauna of marine caves in Bermuda and reported that the caves were inhabited by a wide variety of marine invertebrates. In all, they reported the discovery of more than 50 species, representing groups from the ciliates to tunicates. In discussing the zoogeographical affinities of the cave fauna, they pointed out some obvious af- finities of some of the species with east American and Caribbean species. They also noted (1980: 878) that "Less expected were some striking zoogeographical connections of Bermuda to the east." They suggested and rejected four theories that might explain some of the distribution patterns of the animals found in Bermudan caves: plate tectonics, stranding of species on the shorelines of receding fossil seas, connections with the abyssal fauna, and drift- ing in the open ocean.