Zoogeography of Cetaceans Off Puerto Rico and the Virgin Islands

Zoogeography of Cetaceans off Puerto Rico and the Virgin Islands

ANTONIO A. MIGNUCCI-GIANNONI

Department of Marine Affairs, The University of Rhode Island, Washburn Hall, Kingston, Rhode Island 02881 Present address: Red Caribeña de Varamientos • Caribbean Stranding Network, PO Box 361715, San Juan, Puerto Rico 00936-1715 [email protected].

ABSTRACT.—A zoogeographical analysis of cetaceans in the waters of Puerto Rico, US Virgin Islands, and British Virgin Islands was conducted to document the different species found, and to relate their occurrences to patterns of ocean ﬂoor topography. A total of 2,016 sighting records was entered into a specially formatted database system, and analyzed for distributional and temporal patterns. Species included 13 odontocetes and four mysticetes. The hypothesis that the spatial distribution of cetaceans is highly correlated to the area’s bathymetric relief, whether high or low, was generally supported. Through the use of a relative slope index measure, each sighting was characterized by depth classes (shelf, shelf edge, or offshore), and by sea ﬂoor relief.

RESUMEN.—Para documentar la presencia de las distintas especies de cetáceos y relacionar la misma con patrones de topografía del fondo marino, se llevó a cabo un análisis zoogeográﬁco de los cetáceos de Puerto Rico, Islas Vírgenes Estadounidenses e Islas Vírgenes Británicas. Un total de 2,016 avistamientos fueron ano- tados en una base de datos con formato especíﬁco para avistamientos, y los datos fueron analizados en bus- queda de patrones de distribución espacial y temporal. Las especies estudiadas incluyeron 13 odontocetos y cuatro misticetos. La hipótesis de que la distribución espacial de los cetáceos está estrechamente relacionada con el relieve batimétrico del área, ya sea alto o bajo, fue apoyada generalmente. A través de una medida de índice de pendiente relativa, cada avistamiento fue caracterizado por categoría de profundidad (de plataforma, borde de plataforma o mar afuera) y por relieve del fondo marino.

INTRODUCTION 1935), including water temperature, bottom depth, tidal flow, currents, areas of Whales and dolphins are an intricate part upwelling or areas of convergence and of the marine and coastal fauna of the divergence, prey movements or concentra- northeastern Caribbean Sea, with some of tions, and sea floor relief. Of these condi- the islands serving as primary habitat for tions, water temperature seems to be the the mating and calving of endangered spe- most important factor in determining cies. Although the presence of these crea- breeding and calving grounds, at least in tures has been documented (Erdman, 1970; large whales (Gaskin, 1982). In contrast, Erdman et al., 1973; Taruski and Winn, whale and dolphin congregations in high 1976; Mattila and Clapham, 1989), there is latitude feeding grounds appear to be more still a lack of information on the basic biol- related to phenomena such as ocean fronts, ogy, life history and distribution of the spe- eddies and areas of upwelling which con- cies in the northeastern Caribbean. Two centrate prey. These three features can be of basic questions remain: where and when either of a dynamic or topographic nature, are each of the cetacean species found, and with the latter highly induced by surface which factors affect their distribution in the land masses or underwater sea floor relief northeastern Caribbean. (i.e. sea mounts, subsurface ridges and Cetacean distribution may be deter- edges of the continental shelf) (Gaskin, mined by factors which enhance the avail- 1982). ability of food and the suitable conditions There is a direct relationship between required for reproduction (Townsend, cetacean distribution and bottom topogra- 173

174 A. A. MIGNUCCI-GIANNONI phy in some areas (Hui, 1985; Kenney and between 1987 and 1989, 18 fishing villages Winn, 1987; Selzer and Payne, 1988). In were visited throughout Puerto Rico, and many instances, sea floor contour appears locals were interviewed on the occurrence to be the limiting factor affecting species of all species of marine mammals in the aggregations (Gaskin, 1971). Probably as area. State and federal government officials bottom topography increases in complex- and previous researchers (H. E. Winn, D. K. ity, so does the complexity of the total Mattila, J. C. Jiménez and D. S. Erdman) aquatic environment (Hui, 1979). This com- were also interviewed, and their marine plexity is characterized by seasonal areas of mammal data files kindly made available upwelling, bringing high levels of nutrients for analysis. which enhance primary productivity, and By its nature, most of the data presented therefore facilitating the build up of pri- here are non-random and biased in a vari- mary and secondary consumers such as ety of ways, mainly by observational zooplankton, cephalopods and fish (Hui, chance and effort. The location of some of 1979; Gaskin, 1982). Consequently, oppor- the occurrences may relate more to the dis- tunistic apex predators such as whales and tribution of human populations or activi- dolphins appear. The assumption that com- ties, and not necessarily to the distribution plex sea floor relief enhances food avail- of the animals. Therefore, conclusions ability has been well documented in the based on the statistical analysis must be western North Atlantic, especially for the treated with caution. However, since this is area off the northeastern coast of the the only available information, it serves as United States (Kenney, 1984; Kenney and a general overview of expected patterns of Winn, 1986; Hain et al., 1985), where most aggregation, spatial distribution, and sea- large species of cetaceans migrate to feed. sonality of the species. Only Wells et al. (1980), Dorf (1982) and A database management system was Baumgartner (1997), have examined under- established to file sighting records for each water topography as a factor affecting ceta- species in the study area. Opportunistic cean distribution in tropical or subtropical sightings were included after they proved areas of the western North Atlantic. Rela- to be reliable in terms of species identifica- tionships between underwater topography tion, date, and event location. A serial num- and areas of breeding and calving have not ber (NEPSGXXXX) and a quadrant number been analyzed. were assigned to each sighting in the data- The waters of the northeastern Carib- base. bean are feeding grounds for some species, The study area included the insular shelf and reproductive grounds for others at dif- waters of Puerto Rico, the US Virgin ferent times of the year. Some species do Islands, and the British Virgin Islands (Fig. not migrate, utilizing these waters for feed- 1). The area was partitioned into 864 quad- ing and reproduction throughout the year. rants, each measuring five minutes of lati- Given the presence of whales and dolphins tude by five minutes of longitude, and with off Puerto Rico and the Virgin Islands, it is an area of approximately 68 km2. The hypothesized that their spatial distribution bathymetry of the study area was analyzed is correlated with the area’s bottom topog- by calculating a Slope Index (SI) for each raphy. Some cetaceans are expected to be quadrant, following Kenney’s (1984) tech- distributed in relation to areas of high sea nique for measuring absolute bottom slope, floor relief, while in other species the factor and adapted from Evans’ (1975) Depth affecting distribution may be a low-com- Index (DI) and Hui’s (1979; 1985) Contour plexity level in bottom topography. Index (CI) formulas. These formulas characterize sea floor relief as a percentage of maximum depth change for a given section MATERIALS AND METHODS of the study area. Evans’ DI and Hui’s CI Sighting data were obtained from pub- measure this change as a percentage of the lished and unpublished records collected in maximum depth in that section, while the the study area up to 1989. In 1985, and SI developed by Kenney (1984) incorpo-

CETACEAN ZOOGEOGRAPHY 175

N 025 British Virgin Islands Nautical Miles Atlantic Ocean Camuy Toa Baja Loíza Jost van Dyke Aguadilla Arecibo Anegada Dorado San Juan Luquillo Isla de Tortola Culebra Fajardo Isla Desecheo Virgin Gorda Ceiba Rincón Puerto Rico St. Thomas Mayagüez St. John Humacao 18 00'N Isla de ° Cabo Rojo Isla de Vieques Mona Arroyo US Virgin Islands La Parguera Ponce Salinas Guánica Caribbean Sea St. Croix Guayanilla Bahía de Jobos 67°30'W 67°00'W 66°30'W 66°00'W65°30'W 65°00'W 64°30'W

FIG. 1. Study area and localities referred to in the text. rates into a single number the slope of a chi-square (c2) one-sample test of inde- specific quadrant as a percentage of change pendence (as in Evans, 1975; Hui, 1979; of the bottom topography to the whole Selzer and Payne, 1988). The null hypothe- study area. The SI appears as a dimension- sis (Ho) was: the spatial distribution, repre- less formula defined as: sented by the location of sightings of a given species of cetacean in the study area, æöMmÐ is random and proportional to the number SI= 100 ------of quadrants in each SI class. The expected èøMsa frequency of sightings in SI class i (Ei) was where M and m is the maximum and mini- calculated following Evans (1975), and mum depth, respectively, within each quad- defined as: rant, and M is the maximum depth in the sa L entire study area (6,373 m). Quadrants which E = O æö-----i i tèø border the shoreline were assigned a mini- Lt mum depth of 1.0 m. Strandings and mortal- where O is the total number of sighting ity records were not included in the spatial t occurrences for a species in the study area, distribution analysis because it is impossible L is the number of quadrants in SI class i, to determine the origin of the animals. i and L is the total number of water quad- Slope indices (SI) were grouped into cat- t rants in the study area (762). The E is then egories or classes using a cluster analysis i an estimate of the expected number of together with a discriminative procedure. sightings for a species if their spatial distri- The cluster analysis created six distinct bution is random or proportional to the classes of the SI distribution, and the dis- number of water quadrants in the study criminative analysis provided the ranges area. The level for all tests was 0.05 or for each of the six classes. Class 1 character- a lower, thus avoiding a Type II Error ized areas with very small sea floor relief (accepting the hypothesis when it is false). (up to 4.1%), while Class 6 characterized Only species with more than eight occur- areas with the highest percentage of under- rences were included in this analysis. water relief (as much as 65.3%). To test for a significant relationship between species sightings and bottom RESULTS AND DISCUSSION topography, the observed sighting fre- Diversity quency of each species relative to bottom topography was compared to an estimate The total number of sighting records cat- of expected frequency of sightings using a alogued was 2,016. Seventeen species of

176 A. A. MIGNUCCI-GIANNONI

TABLE 1. Cetacean fauna and summary of sighting records catalogued for Puerto Rico and the Virgin Islands up to 1989 (UNE = unknown status but probably non endangered, V = vulnerable, EBR = endangered but probably recovered).

Taxonomic grouping Total no. Percent of all Earliest/latest Species sightings records record Status Suborder Odontoceti Family Delphinidae Delphinus spp. 13 0.64 1971/1986 UNE Globicephala macrorhynchus 69 3.42 1958/1988 UNE Grampus griseus 2 0.10 1957/1967 UNE Orcinus orca 13 0.64 1960s/1988 UNE Pseudorca crassidens 1 0.05 1988/1988 UNE Stenella coeruleoalba 0a UNE Stenella frontalis 31 1.54 1958/1987 UNE Stenella longirostris 41 2.03 1956/1988 UNE Steno bredanensis 9 0.45 1980/1986 UNE Tursiops truncatus 151 7.49 1966/1989 UNE Unidentified Delphinidae 11 0.55 1969/1985 Family Physeteridae Kogia breviceps 0b UNE Physeter macrocephalus 43 2.13 1952/1986 EBR Family Ziphiidae Ziphius cavirostris 8 0.40 1954/1986 UNE Unidentified small odontocete 2 0.10 1968/1984 Unidentified large odontocete 3 0.15 1960s/1989 Suborder Mysticeti Family Balaenopteridae Balaenoptera acutorostrata 3 0.15 1971/1973 UNE Balaenoptera borealis 2 0.10 1967/1982 V Balaenoptera physalus 3 0.15 1975/1982 V Balaenoptera spp. 14 0.69 1964/1981 V Megaptera novaeangliae 1,597 79.22 1957/1989 EBR

aA stranding of a striped dolphin was recorded for the Virgin Islands in 1982. bFive strandings of a pygmy sperm whales were recorded for Puerto Rico and the Virgin Islands between 1976 and 1989. whales and dolphins were identiﬁed (Table species. Three families and thirteen species 1), including two species, the striped dol- of odontocetes were present at one time or phins (Stenella coeruleoalba [Meyen]) and the another throughout the year. The oceanic pygmy sperm whale (Kogia breviceps [de dolphin family Delphinidae is represented Blainville]), which are only known from in the area by eight genera, and ten species strandings. The earliest sighting record (Table 1). Other species of delphinids which dates to 1952 and consists of three sperm may be present, but so far not reported in whales (Physeter macrocephalus Linnaeus) the study area include: pygmy killer sighted on 5 July at the edge of the drop-off whales (Feresa attenuata Gray), Fraser’s dol- south of Isla de Vieques. The latest record, phins (Lagenodelphis hosei Fraser), melon- 16 April 1989, is of a humpback whale head whales (Peponocephala electra [Gray]), (Megaptera novaeangliae [Borowski]) sighted pantropical spotted dolphins (Stenella atten- at the shelf edge south of Bahía de Guánica. uata [Gray]), and clymene dolphins Toothed whales (odontocetes) were (Stenella clymene [Gray]). These have been reported 397 times (19.7% of all occur- observed in other areas of the Caribbean rences). Of the total number of odontocete (Mignucci-Giannoni, 1989), but as of 1989, sightings, 16 (4.0%) were of unidentiﬁed not in Puerto Rico and the Virgin Islands.

CETACEAN ZOOGEOGRAPHY 177

The sperm whale family Physeteridae is they are described here as Delphinus spp. present, with two out of its three known These dolphins are highly gregarious, at species (Table 1). The dwarf sperm whale times in groups of thousands, probably (Kogia simus Owen), although it may be segregated by sex and age (Schmidly, 1981). present as evidenced by records from St. In the study area they were observed singly Vincent (Caldwell et al., 1973), has not been or in groups of up to 50 animals. Their reported for the study area. The beaked mean group size was 9.2 (n = 12), and four whale family Ziphiidae, probably the least types of aggregations were observed: 1-3 known of all the Cetacea, is represented in animals (66.8%), 7-15 animals (6.9%), 25-29 the study area only by the goosebeak whale animals (6.9%), and 30-50 animals (19.4%). (Ziphius cavirostris Cuvier) (Table 1). Tropi- Saddlebacks are notably acrobatic, and they cal species of the genus Mesoplodon, includ- were reported jumping clear out the water, ing the Antillean beaked whale (M. porpoising, and frequently riding bow europaeus Gervais), densebeak whale (M. waves. densirostris [de Blainville]) and True’s In the western North Atlantic, the species beaked whale (M. mirus True), may be is widely distributed from Newfoundland present, but have not been reported as of south to Venezuela, including the Gulf of 1989. Mattila and Clapham (1989) reported Mexico (Leatherwood et al., 1976). In the the sighting of a Mesoplodon sp. east of the Caribbean, sightings of saddleback dol- Virgin Islands, but could not ascertain the phins have been reported from Antigua, species identity. Cuba, Dominican Republic, St. Lucia, St. Baleen whales (mysticetes) were Vincent, and the Grenadines (Mignucci- reported 1,619 times (80.3% of all occur- Giannoni, 1989). rences). The rorqual family Balaenop- Saddlebacks are an offshore species teridae is the only one represented in the found over the continental slope and in area with two genera, and four species proximity to ocean ridges (Winn et al., (Table 1). Since it is difficult to distinguish 1979). Their distribution has been associ- at sea between fin whales (Balaenoptera ated with rich areas of convergence and physalus [Linnaeus]) and sei whales (B. bore- divergence, and with areas of intrusion of alis Lesson), and most occurrences of large warm water into cooler regions (Gaskin, balaenopterids in the study area did not 1968; Au et al., 1979). Evans (1975) and Hui specify which of the two species was (1979) showed that saddleback dolphins in sighted, the two are treated here as a group. southern California are found in areas of Two balaenopterid whales were not high sea floor relief. Selzer and Payne reported in the study area: blue whales (B. (1988) and Dorf (1982) came to the same musculus [Linnaeus]) and Bryde’s whales conclusion for the Gulf of Maine and Gulf (B. edeni Anderson). Of these, only the latter of Mexico, respectively, describing the may be present as it is well known from the areas as of moderate to high bathymetric southeastern Caribbean. The only Carib- slope. Saddlebacks in the study area bean record for a blue whale is from a cer- showed no higher than expected occur- vical vertebrae encountered at San rences for any specific slope class (n =13, p Cristobal in Panamá in 1922 (Harmer, > 0.70). The average SI was 7.2 (n = 13, max 1923). = 40.6, min = 0.3, s = 10.7, SI class = 2). Of all sightings, 69.2% were well within the Species accounts continental shelf, while 30.8% were near the shelf edge. These dolphins have been Delphinus spp.—Saddleback or common observed off Isla de Mona, northwest of dolphins have been reported 13 times. Bajo Gallardo in the Mona Passage, off Isa- Based on the sighting records, no clear indi- bela, north of Manatí, east of Luquillo, at cation exist to assign the species sighted to Bahía de Guánica and Bahía de Jobos, off the shortsnout saddleback dolphin (D. del- Salinas, north of St. Croix, and near Tortola. phis Linnaeus) or to the longsnout saddle- The species was reported throughout the back dolphin (D. capensis [Gray]); therefore, year, except in May and from August to

178 A. A. MIGNUCCI-GIANNONI

October. The only record describing the Virgin Gorda, south of Horseshoe Reef presence of calves was on early June. (northeast of Virgin Gorda), between St. Globicephala macrorhynchus Gray—Sixty- John and St. Croix and north, northwest nine sightings of shortfin pilot whales have and southwest of St. Croix. been reported. This species has a strong Shortfin pilot whales have been observed herding behavior (Nature Conservancy during all months of the year, more com- Council, 1980), including epimeletic behav- monly during winter and spring. Pilot ior (group cohesiveness in stressful or life- whales in the northeastern Caribbean were threatening situations). In the study area, initially characterized as a summer, warm they have been observed in groups of up to month species, seen from May through 100 animals. The mean group size was 11.1 October, and absent from November (n = 66), and four types of aggregations through April (Caldwell and Erdman, 1963; were observed: 1-3 animals (50.0%), 4-7 ani- Erdman, 1970; Erdman et al., 1973). How- mals (19.6%), 8-20 animals (15.2%), and 25- ever, Taruski and Winn (1976) documented 100 animals (15.2%). They were sometimes winter sightings. The cumulative analysis observed logging stationary at the surface of all occurrences indicates that G. macro- or leisurely swimming. rhynchus is present in the study area Shortfin pilot whales are cosmopolitan, throughout the year, but its appearance, found in tropical and warm temperate compared to the summer and fall months, waters of all oceans (Rice, 1977). In the more than doubles during the winter and western North Atlantic, they have been spring months. This seasonal pattern is in reported as far north as Delaware Bay and accord with those described for pilot Virginia in the summer, but are most com- whales between Cape Hatteras and north- monly found from Cape Hatteras and Ber- ern Florida, and in the Lesser Antilles muda, south to Venezuela, including the (Caldwell and Erdman, 1963). Their occur- Gulf of Mexico (Leatherwood et al., 1976). rence in the Gulf of Mexico seems to be In the Caribbean, sightings of shortfin pilot extended as well, but in contrast, with an whales have been reported from Anegada increase in sightings during the spring and Passage, Bequia, Cuba, Dominica, Domini- summer (Moore, 1953; Dorf, 1982). The can Republic, Haiti, Martinique, Petit only record indicating the presence of Nives, St. Lucia, St. Vincent, and Venezuela calves was in late December. (Mignucci-Giannoni, 1989). There is little information on migration Pilot whales were observed 45.3% of the or movements of this species for the study time over the shelf, 26.6% offshore, and area. Taruski and Winn (1976) suggested 28.1% near the shelf edge. Hui (1985) and the occurrence of a slight seasonal inshore- Dorf (1982) showed a significant relation- offshore movement probably related to ship between their distribution and areas of prey habits. This type of movement has high sea floor relief in other areas. Data for been documented for shortfin pilot whales Puerto Rico and the Virgin Islands does not in southern California and Japan, where the show a clear relation to bottom topography species comes fairly close to shore in the (n = 64, p > 0.20). The average SI was 11.6 (n early spring, coinciding with the spawning = 64, max = 52.6, min = 0.3, s = 11.5, SI class of squid (Leatherwood and Reeves, 1983b). = 3). They have been observed scattered In the northeastern Caribbean, the seasonal throughout the study area, particularly appearance of pilot whales could coincide near Isla Desecheo and Isla de Mona, off with the inshore movement of spawning San Juan, south of Salinas, southwest of octopus described by Voss (1960). Although Cabo Rojo, off Mayagüez, off Rincón and several species of squid are present in the Aguadilla, north of Hatillo, offshore from study area, their spawning habits are Loíza, north of Luquillo, off Guánica and unknown. Ponce, south of Juana Díaz, off Yabucoa, Grampus griseus (Cuvier)—Two sight- throughout the Virgin Bank, northwest of ings of Risso’s dolphins have been Anegada, northwest and south of St. John, recorded. One of the sightings reported in south of St. Thomas, between Tortola and Erdman et al. (1973) as an unidentified dol-

CETACEAN ZOOGEOGRAPHY 179 phin, is classified here as G. griseus. Risso’s seems to be pelagic in some areas, and dolphins, were observed once singly and in coastal in others. Killer whales in the study a group of 18-20 individuals. The latter area were observed 53.5% of the time over agrees with herding behavior described for the continental shelf, 36.4% offshore, and other areas (Leatherwood and Reeves only 0.1% near the shelf edge. No relation- 1983b). ship was evident to bottom topography (n Grampus in the western North Atlantic = 10, p > 0.3), although the average SI was has been reported from eastern Newfound- 7.3 (n = 9, max = 20.6, min = 0.3, s = 8.2, SI land south to St. Vincent, including the class = 2). Killer whales have been eastern and northern Gulf of Mexico observed in the Mona Channel, near Isla de (Leatherwood et al., 1976). It is generally Mona, near Rincón, west of Puerto Real in restricted to offshore areas deeper than 200 Cabo Rojo, off Isla de Culebra, off Luquillo, m (Mitchell, 1975). Caribbean records east of St. Thomas, near St. John, in the Sir included only specimens caught from St. Francis Drake Channel, and between St. Vincent (Caldwell et al., 1971). In the study Thomas and St. Croix. area, Risso’s dolphins have been observed Killer whales were sighted most fre- north of St. Croix, and offshore northwest quently during the winter and early spring of Whale Banks. Both locations are areas of months of December, February and March, high underwater relief, with an average SI but also in June. In contrast, killer whales in of 35.1 (n = 2, max = 40.4, min = 29.9, s = the Lesser Antilles were most often cap- 5.3, SI class = 5). Dorf (1982) also found tured within sight of land during the sum- Risso’s dolphin in the Gulf of Mexico dis- mer (Winn et al., 1979). Killer whales tributed in areas of steep bathymetric appear to follow the seasonal movements slope. The two sightings reported were for of their prey (Leatherwood et al., 1976), but the months of April and September. In the no information is available for the study western North Atlantic, a north-south, area. It has been suggested that killer summer-winter migration has been sug- whales move toward the equator during gested (Schmidly, 1981). the winter and away from it during the Orcinus orca (Linnaeus)—Thirteen sight- summer (Minasian et al., 1984). ings of killer whales has been reported. Pseudorca crassidens (Owen)—False killer Killer whales were observed singly or in whales were observed only once in the herds of up to 25 animals, with a mean study area, on March 1988 at Reef Bay in St. group size of 7.7 (n = 6). Two types of John. The animals were in a group of 15-20 aggregations were observed: 1-3 animals individuals. In the western North Atlantic, (67%), and 12-25 animals (33%). This group false killer whales are found in tropical and composition agrees with reports, in which warm temperate waters from Maryland killer whales were frequently observed in south to Venezuela, and in the Gulf of Mex- pairs or trios, but at times up to 40 individ- ico (Leatherwood et al., 1976). They are uals (Minasian et al., 1984). In the study mainly pelagic, but at times may venture area, killer whale attacks on larger whales inshore (Winn et al., 1979). In the Carib- have been observed twice: first on an uni- bean, false killer whales has been reported dentified species of whale south of Isla de from Cuba, St. Vincent, and Tobago (Mig- Culebra, and second, on a pod of sperm nucci-Giannoni, 1989). No information can whales between St. Thomas and St. Croix. be ascertained in relation to bottom topog- Killer whales in the western North raphy from the single sighting. Atlantic, have been reported from Green- Stenella coeruleoalba—The only record of land and Iceland, south to St. Vincent, the striped dolphin is of a skull found in St. including the Gulf of Mexico (Leatherwood Croix in 1982. In the western North Atlantic et al., 1976). In the Caribbean, sightings of the species is restricted to warmer waters killer whales have been reported from Cay- from Nova Scotia south to Venezuela, man Islands, Cuba, Dominican Republic, including the Gulf of Mexico (Leatherwood St. Lucia, Vincent, and Trinidad and Tobago et al., 1976). In the Caribbean, sightings of (Mignucci-Giannoni, 1989). The species striped dolphins have been reported for

180 A. A. MIGNUCCI-GIANNONI

Bonaire, Curaçao, and Venezuela (Mig- Virgin Islands contradicts the patterns nucci-Giannoni, 1989). Dorf (1982) found observed in Florida and the Gulf of Mexico, striped dolphins in the Gulf of Mexico to be where they move inshore in late spring and distributed in areas of low bathymetric summer (Caldwell and Caldwell, 1966). It slope. has been suggested that this movement Stenella frontalis (Cuvier)—Thirty-one may be related to fish movements, in par- sightings of Atlantic spotted dolphins have ticular that of the pelagic blue runner (Car- been reported. They were observed singly anx crysos [Mitchill]), but the relationship or in groups of up to 50 animals. The mean may be more coincidental than associative group size was 14.2 (n = 26), and four types (Caldwell and Caldwell, 1966). The blue of aggregations were observed: 1-4 animals runner, although present in the study area, (30.8%), 5-10 animals (34.6%), 11-29 animals is not particularly abundant (Erdman et al., (15.4%), and 30-50 animals (19.2%). This 1988). While no information is available for agrees with Perrin et al. (1987), who sug- the study area, given their deep water dis- gested that pods are usually of fewer than tribution elsewhere, and their seasonal 50 animals and most typically of 5 to 15 occurrence over the shelf and shelf edge in individuals in coastal waters. the study area, an inshore-offshore seasonal Atlantic spotted dolphins in the Carib- movement is suspected. bean constitute one of the six geographical Stenella longirostris (Gray)—Spinner dol- populations endemic to the Atlantic Ocean phins were recorded 41 times. Individuals (Perrin et al., 1987). In the Caribbean, sight- have been observed singly or in groups of ings of these dolphins have been reported up to 500 animals. The mean group size from the Colombia, Cuba, Dominican was 53.7 (n = 37), and six types of aggrega- Republic, Panamá, St. Vincent, and Venezu- tions were observed: 1-5 animals (32.5%), 6- ela (Mignucci-Giannoni, 1989). Of all sight- 10 animals (18.9%), 11-30 animals (18.9%), ings in the study area, 85.0% were within 31-100 animals (18.9%), 200 animals (5.4%), the shelf. Dorf (1982) found Atlantic spot- and 500 animals (5.4%). They were fre- ted dolphins in the Gulf of Mexico to be quently seen jumping clear out of the distributed in areas of low bathymetric water, spinning in mid-air, and riding bow slope. In the study area they showed higher waves. Spinner dolphins also were seen in than expected frequencies for areas of low association with humpback whales, even sea floor relief (n = 27, p > 0.001), with an riding the whale’s bow wave. average SI of 3.4 (n = 27, max = 28.6, min = Spinner dolphins are found in pelagic 0.1, s = 6.5, SI class = 1). They were and neritic tropical waters (Schmidly, 1981). observed off Mona Island, south of La Par- In the western North Atlantic, they range guera, southwest of Cabo Rojo, west and from Cape Hatteras, throughout the Gulf of northwest of Mayagüez, off Punta Borin- Mexico and the Caribbean, south to Vene- quen in Aguadilla, near Ceiba, south of zuela. In the Caribbean, sightings of spin- Vieques Island, across the Virgin Bank, ner dolphins have been reported from west of Anegada, and most commonly near Bequia, the Grenadines, St. Vincent, and Jost Van Dyke and Tortola. In the study Tobago (Mignucci-Giannoni, 1989). area, these dolphins seem to be restricted to Although in the southeastern United the shelf areas of low sea floor relief, at States they have been characterized as an times moving to the shelf edge, and rarely offshore, deep water species (Winn et al., offshore. 1979), spinner dolphins in the study area Atlantic spotted dolphins were sighted were observed 72.5% of the time over the throughout the year, with the exception of continental shelf, 22.5% near the shelf edge, July, December and April. Their seasonality and only 5.0% offshore. Dorf (1982) found seems to be bimodal, peaking in August, spinner dolphins in the Gulf of Mexico to decreasing gradually until November, be distributed in areas of low bathymetric peaking again in January and February, slope. In the study area, spinner dolphins and then rarely seen in the spring and sum- showed higher than expected occurrences mer. The seasonality in Puerto Rico and the in areas of low sea floor relief (n = 40, p >

CETACEAN ZOOGEOGRAPHY 181

0.05). The average SI was 6.2 (n = 40, max = Virgin Bank, and off St. John and Virgin 40.6, min = 0.3, s = 8.3, SI class = 2). They Gorda. were observed off Isla de Mona, east of Isla The species only has been sighted during de Desecheo, from Rincón to Aguadilla, February and March. Although the data near Banco de Esponja west of Mayagüez, may indicate a seasonal presence during southwest of Cabo Rojo, off Toa Baja, north late winter and early spring, this is incon- of Luquillo, near Fajardo, throughout the clusive. No information is available on Virgin Bank, east of Virgin Gorda, north migration or movements, but given their and east of St. Croix, and most frequently deep water distribution elsewhere, and off Punta Higüero in Rincón, and close to their seasonal occurrence inside the shelf the islands of St. Thomas, St. John and Jost near the islands in the study area, an Van Dyke. inshore-offshore seasonal movement is sus- Spinner dolphins were observed pected. throughout the year, except in May and Tursiops truncatus (Montagu)—Bottlenose August. They are seldom seen in the sum- dolphins were sighted on 151 occasions. mer and fall, but sightings increase steadily They were observed singly or in groups of from November through January, peaking up to 50 animals. The mean group size was in February, and decreasing in March and 7.8 (n = 147), and six types of aggregations April. No information is available for the were observed: single animals (24.9%), 2-4 study area on migration or movements, but animals (31.0%), 5-10 animals (18.6%), 11-19 given their deep water distribution else- animals (14.5%), 20-39 animals (7.6%), and where, and their seasonal occurrence inside 40-50 animals (3.4%). In the Gulf of Mexico, the shelf near the islands of the study area, Leatherwood and Reeves (1983a) found an inshore-offshore seasonal movement is that most groups averaged 5-6 animals. suspected. Group size in this species seems to increase Steno bredanensis (Lesson)—Nine sight- in open water, with depth and during dif- ings of roughtooth dolphins have been ferent seasons (Shane et al., 1986; Würsig, recorded. The species congregated in small 1978; Odell and Reynolds, 1980). In the groups of 3 to 30 animals. The mean group study area, while group sizes in the sum- size was 9.9 (n = 8). Three types of aggrega- mer and fall averaged 5.5, spring and win- tions were observed: 3-4 animals (25%), 5- ter groups averaged between 8.4 animals. 10 animals (50%), and 15-30 animals (25%). Bottlenose dolphins were observed feed- They were observed on four occasions in ing off Ponce by circling and charging association with humpback whales, once toward the shore as if they were going to riding the bow wave of a whale. strand, a commonly reported behavior of Leatherwood et al. (1976) described the the species (Leatherwood, 1975). They were distribution of these dolphins in the North observed traveling in association with Atlantic as deep tropical and warm temper- humpback whales off Rincón and some ate waters. Winn et al. (1979) conﬁrmed this dolphins have been observed interacting for the West Indies, but added that they with humans, some even approaching also occur near the islands. In the Carib- divers and swimmers, and at times allow- bean, sightings of roughtooth dolphins ing themselves to be stroked or petted. have been reported from Cuba, Haiti and In the Caribbean, sightings of bottlenose St. Vincent (Mignucci-Giannoni, 1989). In dolphins have been reported from Cuba, contrast to the generally suggested pelagic Dominican Republic, Grenada and the distribution, all sightings in the study area Grenadines, Trinidad and St. Vincent (Mig- were within the continental shelf. Further- nucci-Giannoni, 1989). Bottlenose dolphins more, roughtooth dolphins showed higher were observed throughout the study area, than expected values for areas of low sea but were commonly recorded for Isleta de ﬂoor relief (n = 9, p > 0.01), with an average San Juan, including inside the San Juan Bay, SI of 2.4 (n = 8, max = 10.6, min = 0.2, s = and between the Virgin Islands. Sightings 3.3, SI Class = 1). They were observed off were recorded for San Juan, Rio Grande, Aguadilla, Rincón, Aguada, Fajardo, on the Fajardo, Isla de Culebra, Vieques, Arroyo,

182 A. A. MIGNUCCI-GIANNONI

Salinas, Ponce, La Parguera, Rincón, Rojo, and St. Croix. Pygmy sperm whales Aguada, Aguadilla, St. Croix, St. Thomas, stranded in January, April, May and St. John, Tortola, Virgin Gorda and Ane- December. A winter and spring seasonality gada. similar to that of the sperm whale is Bottlenose were observed more often expected. over the shelf (85.2%), occasionally near the Physeter macrocephalus—Sperm whales shelf edge (7.5%), and in offshore waters have been sighted on 43 occasions. They (6.9%). They showed higher than expected have been observed singly or in groups of frequencies for areas of low sea floor relief up to 30 animals. The mean group size was (n = 151, p > 0.001), with an average SI of 4.1 (n = 32), and three main types of aggre- 4.6 (n = 151, max = 40.6, min = 0.1, s = 6.6, gations were observed: 1-3 animals (65.5%), SI class = 2). Dorf (1982) found bottlenose 4-10 animals (31.2%), and 30 animals dolphins in the Gulf of Mexico with similar (3.3%). In the Lesser Antilles they have distributional and bathymetry patterns. been found in less coherent groups, scat- Bottlenose from Florida to Cape Hatteras tered over many miles (Watkins et al., exhibit the same distribution pattern, with 1985). Watkins and Moore (1982) found that animals not extending beyond the conti- Caribbean breeding groups usually com- nental shelf (Caldwell and Caldwell, 1973), prised 1 to 3 calves, 8 or more medium and most being found close to shore, some- whales, and 1 or 2 large whales. times in bays, lagoons and even venturing Sperm whales are found in deep waters into rivers (Leatherwood et al., 1976). From throughout the world (Rice, 1989). Their dis- Cape Hatteras to Georges Bank, bottlenose tribution is clumped into “grounds” and dolphins are found on the shelf edge, in dependent on food source and breeding, areas of high bathymetry (Kenney 1984). varying with sex and age (Gosho et al., Although bottlenose dolphins were 1984). They are most often found along the observed during all months of the year, edge of the continental shelf and along ocean they were more often seen between Decem- trenches (Watson, 1981), seldom inside the ber and April. Dorf (1982) found a similar 200 m isobath (Leatherwood et al., 1976). seasonal pattern for bottlenose dolphins in Whales in the southeastern Caribbean have the Gulf of Mexico. The statement by Erd- been observed to be attracted to some areas man (1970) of the great number of Tursiops where sound scattering organisms concen- during the summer as a characterization of trate just above the 80 to 90 m thermocline their seasonality is unsupported. (Watkins et al., 1985). At times, they appear Seasonal inshore (during winter) and off- close to land, in association with areas of shore (during summer) movements for the upwelling, but they also occur in deep Gulf of Mexico are described by Shane and waters, where they search for large species Schmidly (1978), Shane (1980), and Dorf of cephalopods. Females are more restricted (1982), these related to prey concentration to the tropical and subtropical zones, up to (Wells et al., 1980). Given the seasonal pat- the 54°N mark, while males may inhabit tern observed in the study area, a similar waters further north (Gosho et al., 1984). movement pattern is expected for this In the western North Atlantic, they have region. In addition to regional movements, been reported from southeastern Green- short-term movements dependent on land and Iceland, southward to the Lesser floods, ebb tides, and probably diurnal pat- Antilles, including the Gulf of Mexico terns are also expected. (Leatherwood et al., 1976). In the Carib- Kogia breviceps—Pygmy sperm whales bean, sightings of sperm whales have been stranded on five occasions between, 1976 reported from Anegada, Bequia, Canovan, and 1989. These whales are found in tem- Cuba, Grenada, Grenadines, Guadeloupe, perate, subtropical and tropical oceanic Martinique, St. Lucia, and St. Vincent (Mig- waters of the western North Atlantic, from nucci-Giannoni, 1989). Most sightings in Nova Scotia to the Virgin Islands and in the the northeastern and southeastern Carib- Gulf of Mexico. They stranded in Agua- bean appear to be on the leeward side of dilla, Mayagüez, Punta Pitahaya in Cabo the islands.

CETACEAN ZOOGEOGRAPHY 183

In the study area, sperm whales were the vicinity of sea canyons and escarp- observed 64.2% of the time near the shelf ments (Minasian et al., 1984). Caribbean edge, 26.4% offshore, and 9.4% over the records include Barbados, Bonaire, continental shelf. Sightings were recorded Curaçao, and St. Martin (Mignucci-Gian- for Isla de Mona, Mona Passage, off Rincón, noni, 1989). off San Juan and Loíza, south of Ponce, This species has been observed off south of Isla de Vieques, north of St. Croix, Ramey in Aguadilla, near Isla de Caja de along the southern shelf edge of the north- Muertos, south of Bahía de Guánica, in the ern Virgin Islands, between St. Thomas and vicinity of La Parguera, and between St. St. Croix, and off Anegada. Sperm whales Thomas and St. Croix. Goosebeak whales showed higher than expected occurrences did not show a relationship to bottom in areas of high bottom relief (n = 32, p > topography (n = 8, p > 0.5). The average SI 0.05), with an average SI of 14.0 (n = 32, was 6.6 (n = 7, max = 28.1, min = 0.1, s = max = 65.3, min = 0.1, s = 14.0, SI class = 3). 4.8, SI class = 2). Ziphius was recorded dur- Although older data suggested that ing all months of the year, except July and sperm whales were less seasonal in their October, with most occurrences during the appearance than humpback whales (Erd- winter and early spring. A mother and calf man et al., 1973), the present summary of sighting occurred during November. occurrences shows a similar seasonal pat- Balaenoptera acutorostrata Lacépède— tern for both species. Nonetheless, the Minke whales were sighted on three occa- sperm whale season is more extended than sions. A sighting described by Erdman et that of the humpback, and the bimodal sea- al. (1973) and Winn and Perkins (1976) sonality suggested by Mignucci-Giannoni between the northeast corner of Puerto (1988) seems to be more continuous than Rico and San Juan on 15 February 1965, two distinct seasons. Sperm whales occur was found to be, from the coordinates in the northeastern Caribbean from the late given, considerably far from the study area fall, throughout the winter and early and not considered in the analysis of spring, as early as October, increasing in records. Ten additional sightings were November through January, peaking in reported by Winn and Perkins (1976), February, declining in March and are rarely Taruski and Winn (1976) and by Mattila seen, from April through September. The and Clapham (1989) in nearby Anguilla only two records of whales with calves and in the Anegada Passage. Minke whales were in late December and early January. congregated in small groups of 1 to 3 ani- Their temporal distribution is related to mals. Similarly, the mean group size of the migratory movements, which varies minkes sighted in Anguilla and Anegada between bachelor and breeding groups Passage was 2.4, with most animals found (Leatherwood et al., 1976). in pairs or trios and to a lesser extent, sin- Ziphius cavirostris—Goosebeak whales or gly. Although they were reported to appar- Cuvier’s beaked whales, have been sighted ently not feed while in tropical waters eight times. They were observed singly or (Williamson, 1975), Winn and Perkins in groups of 2 or 3 animals. The mean (1976) observed minke whales near group size was 1.5 (n = 8), and most sight- Anguilla in association with a large group ings were of single animals (62.5%). At of shortﬁn pilot whales and sperm whales other locations, Ziphius have been observed feeding. in groups of up to 25 animals (Leather- In the western North Atlantic, minke wood et al., 1976; Leatherwood and Reeves, whales are distributed from Bafﬁn Bay in 1983b; Minasian et al., 1984). Canada, south to Florida, in the Gulf of In the western North Atlantic, goosebeak Mexico (Winn et al., 1979), and southeast whales are found in tropical and temperate through the West Indies, at least to waters, from Cape Cod to the Lesser Anti- Anguilla Bank in the northern part of the lles, and in the Gulf of Mexico (Winn et al., Lesser Antilles. Additional records of 1979). Ziphius appears to be found prima- minke whales in the Caribbean include ani- rily in tropical pelagic waters, especially in mals sighted off the Bahamas, and near 184 A. A. MIGNUCCI-GIANNONI

Turk and Caicos (Winn and Perkins, 1976). south as the northeastern coast of Venezu- In the study area, they have been observed ela (Leatherwood et al., 1976; Mitchell and near Isla de Mona, off Banco Los Placeres Kozicky, 1974). While fin whales in the southwest of Isla Desecheo, and east of northeastern United States are found more Beef Island near Tortola. often over the shelf, sei whales are found Minke whales are supposedly limited to inshore as well as offshore, frequently near shelf waters (Schmidly, 1981) and are the shelf edge. For both species, a separate mostly found inshore, or even entering Caribbean and Gulf of Mexico population bays. Records in the northeastern Carib- has been suggested (Mitchell and Kozicky, bean, were observed equally over the shelf 1974; Mead, 1975; Schmidly, 1981). Large and near the shelf edge, but less frequently rorquals in Puerto Rico have been observed offshore. The average SI for the study area only north of Isla de Mona, and south of was 3.9 (n = 3, max = 6.3, min = 0.4, s = 2.5, Cayo Ratones in Salinas. Most sightings of SI class = 1). Not enough data are available Balaenoptera spp. have been from the Virgin to make inferences on bathymetry. Islands: north of Whale Banks, off Ane- Taking into account, in addition, the gada, Virgin Gorda, Tortola, south of St. sightings of the nearby Anguilla Bank and John and St. Thomas, and west and east of Anegada Passage, minke whales have been St. Croix. In the study area, rorquals were reported equally in January and February, equally distributed over the shelf (30.8%), and only once in May. Other sightings west near the shelf edge (30.8%) and in offshore of the study area indicate that they may waters (38.4%). The average SI was 3.7 (n = appear in Caribbean waters as early as 12, max = 9.4, min = 0.3, s = 3.4, SI class = December (Winn and Perkins, 1976). Calf 1), and no relationship to bottom topogra- sightings were reported in the study area in phy was evident (n = 12, p > 0.2). February and May. The species is known to Rorquals have been sighted at different be migratory (Seargent, 1963). It is almost months during all seasons, but the majority absent from the New England region in the of sightings (66.7%) occurred during the winter, when it is in its southern range. In winter or early spring. This partially agrees the spring they migrate north towards their with what Mitchell and Kozicky (1974) summer feeding grounds, with males found for sei whales in the southern por- reaching farther north, and females staying tions of the western North Atlantic of a near coastal and more southern areas bimodal seasonality of February-March, (Stewart and Leatherwood, 1985). and July-October, and what Gambell Balaenoptera spp., B. borealis, and B. physa- (1985b) described for fin whales, of a breed- lus—Large rorquals, other than the hump- ing season from December-January. Fin back whale, were observed 19 times, whales in the western North Atlantic have including two sightings of sei whales (B. been reported to move south towards Flor- borelis) and three of fin whales (B. physalus). ida, the Gulf of Mexico and the West Indies These congregated in small groups of up to after October, where supposedly they 5 animals. Mean group size was 1.8 (n = 12) spend their winter in offshore waters and the animals were most often sighted (Leatherwood et al., 1976; Mizroch et al., singly (50%) or in pairs (42%). One group of 1984; Gambell, 1985b). During the spring, five animals was reported. The observed they travel north toward their feeding group composition of both fin and sei grounds north of Long Island. Movements whale herds is in agreement to those patterns of sei whales are less apparent that described at other locations (Leatherwood those of fin whales, closely following the et al., 1976; Nature Conservancy Council, movements of their prey. 1980; Gambell, 1985a; Gambell 1985b). Megaptera novaeangliae—A total of 1,597 In the western North Atlantic, fin and sei sightings have been recorded, most whales are found from Greenland south to obtained from unpublished studies by D. K. Florida, in the Gulf of Mexico and the West Mattila and his colleagues between 1978 Indies. Sei whales are more tropical and 1984. Humpback whales usually con- (Schmidly, 1981) and may be found as far gregated in small groups, with a mean size CETACEAN ZOOGEOGRAPHY 185 of 2.0 animals (n = 1,293). They were often areas of concentration; one along the north- found singly (43.2%), in pairs (37.5%), in western coast of Puerto Rico, and the sec- trios (9.2%), or groups of four or more ani- ond widely spread around the northern mals (10.1%). These values are generally in Virgin Islands. Scattered sightings, both on agreement with those reported for the Vir- the Atlantic and Caribbean sides of the gin Islands by Mattila and Clapham (1989), islands, occurred at other locations, but with the exception of their lower values with less frequency. Of interest is the spo- (less than 2.9%) for groups of four or more radic occurrence of humpbacks between St. individuals. Mattila and Clapham (1989) Thomas and St. Croix, off St. Croix itself, reported that singers, always lone individu- and on the southern coast of Puerto Rico. als, composed 24.7% of the sightings in the Humpbacks were also reported near Isla de Virgin Islands and 11.3% of those in Puerto Mona, Isla Desecheo, and along the north Rico. Mother-calf aggregations constituted coast of Puerto Rico, at times close to San 11.9% of sightings in the overall study area, Juan and Arecibo. Off the northwestern but differed considerably between Puerto coast of Puerto Rico, humpbacks aggre- Rico (6.8%) and the Virgin Islands (17.1%) gated more often in two areas, off Punta (Mattila and Clapham, 1989). Balcomb and Higüero in Rincón, and off Punta Nichols (1978) reported similar mother-calf Agujereada (near Punta Borinquen) in ratios to that of the overall study area in Aguadilla. In the Virgin Islands, the largest Banco de la Plata in the Dominican Repub- concentration of humpbacks occurs along lic. Mattila and Clapham (1989) also noticed the northern coast of the islands, from St. that mother and calf groups were com- Thomas to Tortola, and within these, most monly escorted by a third adult-size animal, sightings were reported off the northeast- supposedly a mature male waiting to mate ern coast of St. John. with the female, but indicated that as the Most of the sightings were close to shore, season progressed, less mother and calf or at least over the continental shelf. In very groups were observed in association with a few instances were humpbacks seen on the third individual. shelf edge or offshore, and in those cases, In the western North Atlantic, humpback they were traveling. Mother and calf whales breed mainly along the Antillean groups where observed closer inshore. In chain (Winn and Winn, 1978; Whitehead, addition, humpbacks showed higher than 1982). Although the Caribbean distribution expected frequencies for areas of low sea of humpbacks extends from Venezuela to floor relief (n = 1,558, p > 0.001), with an Grand Turks, it is clearly concentrated in average SI in the study area of 5.4 (n = the north-central and northeastern Carib- 1,558, max = 40.4, min = 0.03, s = 6.7, SI bean, in areas less than 200 m deep (Winn class = 2). The SI was slightly higher in et al., 1975). The total area inhabited by Puerto Rico than in the northern Virgin whales was estimated as 64,752 km2, but Islands (Puerto Rico: mean SI = 7.1, n = 923, the major breeding and calving grounds are max = 33.3, min = 0.3, s = 4.8, SI class = 2; restricted to two relatively small banks Virgin Islands: mean SI = 1.8, n = 620, max north of the Dominican Republic: Banco de = 39.5, min = 0.3, s = 5.7, SI class = 1). The la Plata and Banco de Navidad (Winn et al., higher SI in Puerto Rico is due to the nar- 1975). Other Caribbean localities were rower nature of its continental shelf. Both humpback whale sightings have been areas were highly correlated with areas of reported include Anegada Passage, low bathymetric relief. Anguilla Bank, Antigua, Barbuda, Belize, The distribution of humpback whales in Cuba, Curaçao, Dominica, Dominican the study area appears to be determined by Republic, Grand Turks, the Grenadines, conditions necessary for breeding and calv- Guadeloupe, Isla de Aves, Monseratt, St. ing: warm waters, protection from heavy Eugtaius, St. Kitts, and St. Martin (Mig- sea action, and relatively shallow banks. It nucci-Giannoni, 1989). seems now that areas with low bathymetry Humpback whales were observed may play, in addition, a role in enhancing throughout the study area, with two major the conditions needed for reproduction. 186 A. A. MIGNUCCI-GIANNONI

Puerto Rico and the Virgin Islands provide (1989) classified the whales in Puerto Rico these suitable conditions, with as many as as a transient population, since most indi- 6,183 km2 of habitable banks (Winn et al., viduals, except mother-calf groups, were 1975), although the whales are concen- not usually resighted within the same sea- trated in smaller specific areas. A density of son. This transient nature also seems be 0.044 humpback whales per nautical mile2 true for the Virgin Islands, given that the has been reported for the Virgin Islands longest period between resightings was 13 (Mattila and Clapham, 1989), and is days (Mattila and Clapham, 1989). expected to be similar in Puerto Rico. Humpback whales have a marked sea- Spatial distribution sonality in the northeastern Caribbean, Whales and dolphins were reported being reported in the tropical waters more often near land masses or within the between November and May. In the north- insular shelf (Fig. 2). Each species can be eastern United States, 95% of the hump- grouped zoogeographically into different back whale sightings occur between April habitats, either by depth classes (shelf, shelf and October (Kenney, 1984). The earliest edge, or offshore), or by sea floor relief record of the season in the study area is of (using contour or slope indices). Most ceta- 10 November, and the latest is of 30 May. ceans were found within the continental Most individuals were seen after 15 Janu- shelf, although sightings of some species ary, with 3.5% of the sightings occurring extended into the shelf edge or offshore. before that date. The peak of the season Sperm whales predominated near the shelf occurred by the first two weeks of February edge and Risso’s dolphins were most com- and through the middle of March. The last mon offshore. two weeks in March were characterized by Relative bottom topography, measured a steep decline in sighting frequency, and by Contour Index (CI) and Slope Index (SI), whales were rarely seen after the beginning have been used successfully elsewhere to of April (<3.0%). An analysis by Mattila characterize geographic areas in which sea and Clapham (1989) of the whales sighted floor relief enhances primary productivity, per hour surveyed supports this seasonal the availability of food prey, and conse- pattern. Their analysis also showed a dif- quently, the concentration of cetacean spe- ference in the peak seasonal abundance cies. The bathymetric analysis of the study between the Virgin Islands and Puerto area showed distinct areas of high and low Rico, with the latter occurring slightly later. sea floor relief. Areas of low bathymetry Mother and calf groups have been reported were found in coastal waters and on the as early as the first two weeks of December, wide eastern shelf, while areas of high sea through the last two weeks in May. floor relief were evident west and north- Through the use of photographic indi- west of Aguadilla, and along the shelf edge vidual identification, animals from the four and offshore on the south side of the insu- major North Atlantic feeding grounds have lar shelf. The high productivity of these been identified migrating to Puerto Rico areas is evidenced by their continuous use and the Virgin Islands. Most of the animals for sport and commercial fishing. photo-identified in the study area have The hypothesis that the spatial distribu- been photographed in the Newfoundland/ tion of a given species of cetacean is highly Labrador grounds (44.6%) and in the Gulf correlated to the area’s bathymetric relief is of Maine (10.8%) (Mattila and Clapham, generally supported, although for some 1989). Matches between the study area and species the relationship was not as clear. Greenland and Iceland were uncommon Nonetheless, by using the measurement of (2.3% and 3.1%, respectively). The photo- higher than expected occurrences of each matches between Puerto Rico and Iceland species per SI class, the species in the study document a great circle distance of 5,982 area can be more reliably characterized into km, probably one of the longest recorded distinct bathymetrical SI habitats, whether traveling distances of any marine mammal of a low, medium or high sea floor relief (Martin et al., 1984). Mattila and Clapham (Table 2). CETACEAN ZOOGEOGRAPHY 187

FIG. 2. Spatial distribution of all cetaceans from the study area.

Other factors play important roles in the TABLE 2. Characterization of cetaceans by distribution of whales and dolphins in the bathymetric slope index (SI) habitats in Puerto Rico study area. This is clearly the case of the and the Virgin Islands. humpback whale. Humpback whale distri- Sea floor characterization Species bution seems to be concentrated on the windward side of land masses, in banks of Low sea floor relief Balaenoptera acutorostrata (SI Classes 1 and 2) Balaenoptera spp. enough width, as is the case of Banco Delphinus spp. Borinquen and Virgin Bank. Humpback Megaptera novaeangliae whales observed elsewhere appear to be Orcinus orca simply passing through. The degree of pro- Stenella frontalis tection from rough weather also comes into Stenella longirostris play in the case of birthing and calving Steno bredanensis mothers, with most of them distributed Tursiops truncatus close to shore. Ziphius cavirostris Medium sea floor relief Globicephala macrorhynchus Temporal distribution (SI Classes 3 and 4) Physeter macrocephalus Most species were sighted during the win- High sea floor relief Globicephala macrorhynchus ter and early spring, in a similar seasonal (SI Classes 5 and 6) Physeter macrocephalus pattern to the one well documented for Not analyzed Kogia breviceps humpback whales. This seasonality is clearly Pseudorca crassidens seen in most odontocetes, even in previously Stenella coeruleoalba considered summer species like bottlenose dolphins and pilot whales. The increase in sightings begins in December, peaks in Feb- reported (Voss, 1960). Some of these species ruary, and gradually decreases in March and are known as prey items for odontocetes, April. The rate of sightings from May however, it is impossible to assess from the through November is similar every month, data, whether the inshore movement of averaging 4.0% of all sightings. octopi and the near shore seasonality of The winter and early spring seasonality odontocetes are related. It is reasonable to of odontocetes coincides with the known hypothesize that the same conditions spawning, inshore-movement of octopus in which drive some species of octopi to Puerto Rico. Little is known about the life spawn inshore in the winter and spring histories of cephalopods in the northeast- months, could drive similar tropical species ern Caribbean, although a number of spe- of cephalopods, particularly squid, to do cies of octopus and squid have been likewise. The higher number of odontocete 188 A. A. MIGNUCCI-GIANNONI sightings near shore from December ronmental features in the eastern Tropical Pacific. through April would tend to indicate this, NMFS/SWFC Admin. Rep. No. LJ-79-43, 59 pp. but concrete evidence is needed to support Balcomb, K. C., and G. Nichols. 1978. Western North Atlantic humpback whales. Rep. Int. Whal. this hypothesis. Commn. 28:159-164. Whales and dolphins are an integral part Baumgartner, M. F. 1997. The distribution of Risso’s of the Caribbean fauna. Four endangered dolphin (Grampus griseus) with respect to the phys- cetaceans are found off Puerto Rico and the iography of the northern Gulf of Mexico. Mar. Virgin Islands, with two of these, fin and Mamm. Sci. 13(4):614-638. sei whales, considered vulnerable, and the Caldwell, D. K., and D. S. Erdman. 1963. The pilot whales in the West Indies. J. Mammal. 44(1):113- sperm whales and humpback whales to 115. some extent recovered. Through the efforts ______, and M. C. Caldwell. 1966. Observations on of dedicated researchers, the basis for the distribution, coloration, behavior, and audible understanding the composition of the ceta- sound production of the spotted dolphin, Stenella cean fauna of the northeastern Caribbean plagiodon. Los Angeles County Mus. Contrib. Sci. has expanded. These data must be supple- 104:1-27. mented with additional research on the life ______, and ______. 1973. Marine mammals of the eastern Gulf of Mexico. In J. I. Jones et al. (eds.), A history, stranding and mortality, general Summary of Knowledge of the Eastern Gulf of biology, and zoogeography of the species in Mexico, The State Univ. Syst. Florida, Institute of order to provide important information for Oceanogr., Gainesville. the management and conservation of the ______, ______, W. F. Rathjen, and J. R. Sullivan. whale and dolphin species in the area. 1971. Cetaceans from the Lesser Antillean Island of St. Vincent. Fishery Bull. 69(2):303-312. ______, ______, and G. Arrindeld. 1973. Dwarf Acknowledgments.—I am indebted to all sperm whales, Kogia simus, from the Lesser Anti- researchers and government officials who llean Island of St. Vincent. J. Mammal. 54(2):515- gathered data from Puerto Rico and the 517. Virgin Islands, especially D. S. Erdman, Dorf, B. A. 1982. Oceanographic factors and cetacean J. C. Jiménez and D. K. Mattila, who trusted distributions at two sites in the Gulf of Mexico. me to analyze their findings. Also, I thank Masters thesis, Texas A&M Univ., 122 pp. Erdman, D. S. 1970. Marine mammals from Puerto those who I interviewed, for providing Rico to Antigua. J. Mammal. 51(3):636-639. information about the whales, dolphins ______, J. Harms, and M. Marcial Flores. 1973. Ceta- and fisheries in the Caribbean. I am partic- cean records from the northeastern Caribbean ularly obliged to my graduate committee, region. Cetology 17:1-14. B. E. Martí, L. M. Alexander and the late ______, J. M. Berríos Díaz, I. N. Corujo Flores, and J. H. E. Winn, for their continuous and sin- González Azar. 1988. Los peces marinos de Puerto Rico. In J. L. Vivaldi (ed.), Compendio Enciclopé- cere interest and support during the dico de los Recursos Naturales de Puerto Rico, research and writing of this study as part of Tomo II, Vol. VII, pp. 1-487, Editorial Librotex, San my master’s thesis at the University of Juan. Rhode Island. R. D. Kenney and J. F. Evans, W. E. 1975. Distribution, differentiation of pop- Heltsche gave advice in preparing the ulations, and other aspects of the natural history of study in terms of methodology, and Delphinus delphis Linnaeus in the northeastern Pacific. Doctoral dissertation, Univ. California, 164 M. Christian, J. G. Mead, D. K. Mattila and pp. P. Clapham provided ample advice. The Gambell, R. 1985a. Sei whale Balaenoptera borealis Les- following persons provided logistic help: son, 1828. In S. Ridgway and R. Harrison (eds.), V. M. Burnett, M. S. Rudd, M. J. Palm, Handbook of Marine Mammals, Volume 3: The T. Shean, R. Kolton, A. L. Mignucci-Mattei, Sirenians and Baleen Whales, pp. 155-170, Aca- M. C. Giannoni-Torres, and Elmer and demic Press, Orlando. Edwin Aponte. The manuscript was greatly ______. 1985b. Fin whale Balaenoptera physalus (Lin- naeus, 1758). In S. Ridgway and R. Harrison (eds.), improved by the comments of N. B. Barros Handbook of Marine Mammals, Volume 3: The and R. K. Bonde. Sirenians and Baleen Whales, pp. 171-192, Aca- demic Press, Orlando. LITERATURE CITED Gaskin, D. E. 1968. Distribution of Delphinidae (Ceta- cea) in relation to sea surface temperature off east- Au, D. W. K., W. L. Perryman, and W. F. Perrin. 1979. ern and southern New Zealand. New Zealand J. Dolphin distribution and the relationship to envi- Mar. Sci. Freshwater Res. 2:527-534. CETACEAN ZOOGEOGRAPHY 189

______. 1971. Distributional and movements of Mignucci-Giannoni, A. A. 1989. Zoogeography of sperm whales Physeter catodon L. in the Cook Strait marine mammals in Puerto Rico and the Virgin region of New Zealand. Norw. J. Zool. 19:241-259. Islands. Master’s thesis, Univ. Rhode Island, 448 ______. 1982. The Ecology of Whales and Dolphins. pp. Heinemann Educational Books, Portsmouth, 459 Minasian, S. M., K. C. Balcomb, and L. Foster. 1984. pp. The World’s Whales, the Complete Illustrated Gosho, M. E., D. W. Rice, and J. M. Breiwick. 1984. The Guide. Smithsonian Books, Washington, DC, 224 sperm whale, Physeter macrocephalus. Mar. Fish. pp. Rev. 46:54-64. Mitchell, E. D. 1975. Report of the meeting on smaller Hain, J. H. W., M. A. M. Hayman, R. D. Kenney, and H. cetaceans. Montreal, April 1-11, 1971. J. Fish. Res. E. Winn. 1985. The role of cetaceans in the shelf- Bd. Can. 32(7):889-983. edge region of the northeastern United States. Mar. ______, and W. M. Kozicki. 1974. The sei whale (Bal- Fish. Rev. 47(1):13-17. aenoptera borealis) in the northwest Atlantic Ocean. Harmer, S. F. 1923. Cervical vertebrae of a giant blue Int. Whal. Commn. Doc. SC/SP74/32, 48 pp. whale from Panama. Proc. Zool. Soc. London Mizroch, S. A., D. W. Rice, and J. M. Breiwick. 1984. 1923:1085-1089. The ﬁn whale, Balaenoptera physalus. Mar. Fish. Hui, C. A. 1979. Undersea topography and distribu- Review 46(4):20-24. tion of dolphins of the genus Delphinus in the Moore, J. C. 1953. Distribution of marine mammals to southern California bight. J. Mammal. 60:521-527. Florida waters. Amer. Midland Nat. 49(1):118-120. ______. 1985. Undersea topography and the compar- Nature Conservancy Council. 1980. A World Review ative distribution of two pelagic dolphins. Fishery of the Cetacea. Nat. Conserv. Council Great Britain, Bull. 83:472-475. London, 390 pp. Kenney, R. D. 1984. Distributional biology of the ceta- Odell, D. K., and J. E. Reynolds. 1980. Abundance of cean fauna of the northeast United States continen- the bottlenose dolphin, Tursiops truncatus, on the tal shelf. Doctoral dissertation, Univ. Rhode Island, west coast of Florida. Natl. Tech. Info. Serv., PB80- 144 pp. 197650, 47 pp. ______, and H. E. Winn. 1986. Cetacean high-use hab- Perrin, W. F. et al. 1987. Revision of the spotted dol- itats of the northeast United States continental phins, Stenella spp. Mar. Mamm. Sci. 3:99-170. shelf. Fishery Bull. 84:345-357. Rice, D. W. 1977. A list of the marine mammals of the ______, and ______. 1987. Cetacean biomass densi- World. NOAA Tech. Rep. NMFS SSRF-711:1-15. ties near submarine canyons compared to adjacent ______. 1989. Sperm whale Physeter macrocephalus shelf/slope areas. Cont. Shelf Res. 7(2):107-114. Linnaeus, 1758. In S. Ridgway and R. Harrison Leatherwood, S. 1975. Observations of feeding behav- (eds.), Handbook of Marine Mammals, Volume 4: ior of bottlenose dolphins, Tursiops truncatus, in the River Dolphins and the Larger Toothed Whales, northern Gulf of Mexico and Tursiops c.f. T. gilli, off pp. 177-233, Academic Press, San Diego. southern California, Baja California and Nayarit, Schmidly, D. J. 1981. Marine mammals of the south- Mexico. Mar. Fish. Rev. 37(9):10-16. eastern United States coast and the Gulf of Mexico. ______, and R. R. Reeves. 1983a. Abundance of bot- Natl. Tech. Inf. Serv., PB81-212-508, 103 pp. tlenose dolphins in Corpus Cristi Bay and coastal Seargent, D. E. 1963. Minke whales, Balaenoptera acu- southern Texas. Contrib. Mar. Sci., 26:179-199. torostrata Lacépède in the western North Atlantic. J. ______, and ______. 1983b. The Sierra Club Hand- Fish. Res. Bd. Can. 20(6):1489-1504. book of Whales and Dolphins. The Sierra Club Selzer, L. A., and P. M. Payne. 1988. The distribution of Books, San Francisco, 302 pp. white-sided (Lagenorhynchus acutus) and common ______, D. K. Caldwell, and H. E. Winn. 1976. dolphins (Delphinus delphis) vs. environmental fea- Whales, dolphins and porpoises of the western tures of the continental shelf of the northeastern North Atlantic: A guide to their identiﬁcation. United States. Mar. Mammal Sci. 4(2):141-153. Tech. Rep., Natl. Mar. Fish. Serv. CIRC-396, 176 pp. Shane, S. H. 1980. Occurrence, movements, and distri- Martin, A. R., S. K. Katona, D. K. Mattila, D. K. Hem- bution of bottlenose dolphins, Tursiops truncatus, in bree, and T. D. Waters. 1984. Migration of hump- southern Texas. Fish. Bull. 78:593-601. back whales between the Caribbean and Iceland. J. ______, and D. J. Schmidly. 1978. The population biol- Mammal. 65(2):330-333. ogy of the Atlantic bottlenose dolphin, Tursiops Mattila, D. K., and P. J. Clapham. 1989. Humpback truncatus, in the Aransas Pass area of Texas. Natl. whales and other cetaceans on Virgin Bank and in Tech. Info. Serv. the northern Leeward Islands, 1985 and 1986. Can. ______, R. S. Wells, and B. Würsig. 1986. Ecology, J. Zool. 67:2201-2211. behavior and social organization of bottlenose dol- Mead, J. G. 1975. Distribution of cetaceans along the phins: a review. Mar. Mammal Sci. 2(1):34-63. Atlantic and Gulf coasts of the United States. US Stewart, B. S., and S. Leatherwood. 1985. Minke whale Mar. Mamm. Commn., MMC MM4AC007, Wash- Balaenoptera acutorostrata Lacépède 1804. In S. ington, DC. Ridgway and R. Harrison (eds.), Handbook of Mignucci-Giannoni, A. A. 1988. A sperm whale, Phy- Marine Mammals, Volume 3: The Sirenians and seter catodon, at Cayo Santiago, Puerto Rico. Carib. Baleen Whales, pp. 91-136, Academic Press, San J. Sci. 24(3-4):213-215. Diego. 190 A. A. MIGNUCCI-GIANNONI

Taruski, A. G., and H. E. Winn. 1976. Winter sightings Williamson, G. R. 1975. Minke whales off Brazil. Sci. of odontocetes in the West Indies. Cetology 22:1-12. Rep. Whales Res. Inst. 27:37-59. Townsend, C. H. 1935. The distribution of certain Winn, H. E., and L. K. Winn. 1978. The song of the whales as shown by logbook records of American humpback whale, Megaptera novaeangliae in the whale ships. Zoologica 19:1-50. West Indies. Mar. Biol (Berlin) 47:97-114. Voss, G. L. 1960. Potentialities for an octopus and ______, and P. J. Perkins. 1976. Distribution and squid ﬁshery in the West Indies. Proc. Gulf Caribb. sounds of the minke whale, with a review of mys- Fish. Inst., 12th Ann. Sess. (November 1959):129- ticete sounds. Cetology 19:1-12. 135. ______, R. K. Edel, and A. G. Taruski. 1975. Popula- Watkins, W. A., and K. E. Moore. 1982. An underwater tion estimates of the humpback whale (Megaptera acoustic survey for sperm whales (Physeter catodon) novaeangliae) in the West Indies by visual and and other cetaceans in the southeast Caribbean. acoustic techniques. J. Fish. Res. Bd. Can. 32:499- Cetology 46:1-7. 506. ______, ______, and P. Tyack. 1985. Sperm whales Winn, L. K., H. E. Winn, D. K. Caldwell, M. C. Cald- acoustic behaviors in the southeast Caribbean. well, and J. L. Dunn. 1979. Marine mammals. In A Cetology, 49:1-15. summary and analysis of environmental informa- Watson, L. 1981. Sea Guide to Whales of the World. E. tion on the continental shelf and Blake Plateau P. Dutton, New York, 302 pp. from Cape Hatteras to Cape Canaveral, by Center Wells, R. S., A. B. Irvine, and M. D. Scott. 1980. The for Natural Areas. Vol. I, Book 2, Chap. 12. Natl. social ecology of inshore odontocetes. In L. M. Her- Tech. Info. Serv., PB80-184104, pp. 1-117. man (ed.), Cetacean Behavior: Mechanism and Würsig, B. 1978. Occurrences and group organization Functions, John Wiley & Sons, New York, pp. 263- of Atlantic bottlenose porpoises (Tursiops truncatus) 317. in an Argentine bay. Biol. Bull. 154:348-359. Whitehead, H. P. 1982. Population of humpback whales in the northwest Atlantic. Rep. Int. Whal. Commn. 32:345-353.