165-176 Love r3 11/17/07 12:38 PM Page 165 LOVE ET AL.: FISH ASSEMBLAGE IN THE ANACAPA PASSAGE CalCOFI Rep., Vol. 48, 2007 A CHARACTERIZATION OF THE FISH ASSEMBLAGE OF DEEP PHOTIC ZONE ROCK OUTCROPS IN THE ANACAPA PASSAGE, SOUTHERN CALIFORNIA, 1995 TO 2004, WITH EVIDENCE OF A REGIME SHIFT MILTON S. LOVE AND DONNA M. SCHROEDER Marine Science Institute University of California Santa Barbara, California, 93106 [email protected] ABSTRACT INTRODUCTION During 1995, 1999, and 2001–04, using a manned The fish assemblages of several benthic marine habi- research submersible, we surveyed the fish assemblage tats in southern California have been well described. on rocky outcrops (situated at depths of 45–50 m) in the Trawl surveys (e.g., Allen et al. 2002) have characterized Anacapa Passage, southern California. We observed soft sea floor assemblages and considerable attention has 40,132 fish and a minimum of 32 fish species. Rockfishes been given to the hard bottom fish assemblages in the (Sebastes spp.) dominated the assemblage both in diver- shallow photic zone (30 m and less) (North and Hubbs sity and abundance. Squarespot rockfish (Sebastes hop- 1968; Ebeling et al. 1980; Stephens et al. 1984). However, kinsi), a schooling small-sized species, was the most the fish assemblages of rocky outcrops below 30 m re- abundant taxa, while blue rockfish (S. mystinus), black- main very poorly described. With the exception of a eye goby (Rhinogobiops nicholsii), blacksmith (Chromis semi-quantitative survey of some rocky outcrop fishes punctipinnis), halfbanded rockfish (S. semicinctus), vermil- on Tanner and Cortes banks (Lissner and Dorsey 1986), ion rockfish (S. miniatus), rosy rockfish (S. rosaceus), there have been no published accounts, based on un- senorita (Oxyjulis californica), lingcod (Ophiodon elonga- derwater observations, of the fish communities that in- tus), and sharpnose/white seaperches (Phanerodon atripes habit rocky outcrops in waters below about 30 m in the and P. furcatus) were also characteristic species. The species Southern California Bight. assemblage on these outcrops represented a transition Since 1995 we have conducted surveys of the fish between that of the nearshore kelp beds and those more assemblages of oil platforms and natural reefs in 30 to typical of deeper-water sites. The fish assemblage changed 360 m of water in southern California using a manned over time, due primarily to the addition of some species submersible. Usually, reefs were surveyed once or twice and increases in densities of many taxa. This occurred over this period. In contrast, a rocky area in the deep during a period where the oceanographic regime shifted photic zone (45–50 m of water) in the Anacapa Passage from low productivity and warm water to high produc- (between Anacapa and Santa Cruz islands) was surveyed tivity and cool conditions. during six of the nine years. Our repeated visits to the Figure 1. Location of survey site including sidescan sea floor image. 165 165-176 Love r3 11/17/07 12:38 PM Page 166 LOVE ET AL.: FISH ASSEMBLAGE IN THE ANACAPA PASSAGE CalCOFI Rep., Vol. 48, 2007 Figure 2. Typical habitat and fishes of the study site: (A) boulder and ledge habitat, (B) small boulders and sand with brown algae, (C) juvenile lingcod (Ophiodon elongatus), (D) rosy rockfish (Sebastes rosaceus), (E) squarespot rockfish (S. hopkinsi), (F) blue rockfish (S. mystinus), and (G) juvenile vermilion rockfish (S. miniatus). 166 165-176 Love r3 11/17/07 12:38 PM Page 167 LOVE ET AL.: FISH ASSEMBLAGE IN THE ANACAPA PASSAGE CalCOFI Rep., Vol. 48, 2007 Anacapa Passage site have provided us with an oppor- TABLE 1 tunity to both characterize this previously little-known Common and scientific names of species fish community and to observe how that community observed in this study. changed over the study period; it is that which we ex- Common Name Scientific Name plore in this study. Blackeye goby Rhinogobiops nicholsii Black perch Embiotoca jacksoni Blacksmith Chromis punctipinnis METHODS Blue rockfish Sebastes mystinus Study area: The study area is located at approxi- Bocaccio Sebastes paucispinis mately 33 59.9'N, 119 28.6'W (fig. 1). The outcrops lie California scorpionfish Scorpaena guttata ˚ ˚ California sheephead Semicossyphus pulcher at depths of 45–50 m and some extend more or less un- Copper rockfish Sebastes caurinus broken for at least 1,500 m. The features are sedimen- Deepwater blenny Cryptotrema corallinum tary rock that extends 1–2 m above the sandy sea floor Flag rockfish Sebastes rubrivinctus Gopher rockfish Sebastes carnatus forming ledges with shelter holes and overhangs. The Halfbanded rockfish Sebastes semicinctus rock has been differentially eroded with some layers stick- Honeycomb rockfish Sebastes umbrosus ing up through thin sediment, forming long, linear ridges Kelp rockfish Sebastes atrovirens Lingcod Ophiodon elongatus (fig. 1). The layers that are most easily eroded form low Olive rockfish Sebastes serranoides areas filled with coarse sediments (Cochrane et al. 2003). Painted greenling Oxylebius pictus Some of the outermost sections of exposed strata have Pile perch Rhachochilus vacca Pink seaperch Zalembius rosaceus collapsed, adding boulders to the ledge habitat (fig. 2). Pygmy rockfish Sebastes wilsoni Cobble surrounds some of these features. The outcrops Rainbow seaperch Hypsurus caryi support locally high densities of brown algae (including Rosy rockfish Sebastes rosaceus Rubberlip seaperch Rhachochilus toxotes Laminaria farlowii), foliose red algae (Plocamium carti- Senorita Oxyjulis californica lagineum), articulate coralline algae (Bossiella californica), Sharpnose seaperch Phanerodon atripes branching coralline algae (probably Calliarthron tubercu- Squarespot rockfish Sebastes hopkinsi Starry rockfish Sebastes constellatus losum), and encrusting coralline algae. Sponges, red urchins, Treefish Sebastes serriceps gorgonians, and bryozoans also dot the outcrops. Unidentified ronquil Field sampling: We surveyed fish assemblages using Unidentified seaperches1 Phanerodon sp. Vermilion rockfish Sebastes miniatus the Delta research submersible, a 4.6 m, two-person ves- White seaperch Phanerodon furcatus sel, operated by Delta Oceanographics of Oxnard, Wolf-eel Anarrhichthys ocellatus California. Aboard the Delta, we conducted 15-minute- 1Probably both sharpnose and white seaperches. long (10 minute 1995) belt transects about 2 m from the substrata, while the submarine maintained a speed of the submarine outwards were counted. Fish lengths were about 0.5 knots. We conducted surveys in 1995, 1999, estimated using a pair of parallel lasers mounted on ei- and 2001–04. Four transects were conducted in every ther side of the external video camera. The projected year except 1999 and 2001, when three transects were reference points were 20 cm apart and were visible to made. Within the study area, we selected ridges more both the observer and the video camera. Transect lengths or less randomly and transects were run parallel to these were computed by counting the number of 20 cm laser structures. In every year except 2002, surveys were con- segments in 15 second subsamples (one per minute) ducted in October; in 2002 surveys occurred in throughout the transect, calculating speed based on those November. Late fall is the optimal time to conduct sur- counts and averaging it over the whole transect, and veys because of generally good weather and water clar- multiplying that average speed by the transect duration. ity. In addition, many fish species have completed their An environmental monitoring system aboard the sub- seasonal juvenile recruitment by this time. marine continuously recorded date and time, depth, and In each year, submersible surveys were conducted altitude of the vessel above the sea floor. The environ- during daylight hours between 1400 and 1700. The same mental data was overlaid on the original videotape upon observer (D.S.) conducted all of the transects during all completion of each survey. Transect videos were re- years. During each transect, the researcher made obser- viewed aboard the research vessel or in the laboratory vations from a viewing port on the starboard side of the and observations transcribed into a database. For each submersible. An externally mounted hi-8 mm video fish, we recorded species and estimated its total length camera with associated lights filmed the same viewing in 5 cm increments. All individuals were identified to fields as seen by the observer. The observer identified, species. The common and scientific names of all species counted, and estimated the lengths of all fishes and ver- observed are listed in Table 1. bally recorded those data onto the video tape. All fishes Statistical analyses: Interannual relationships in the in a volume 2 m from the seafloor upwards and from Anacapa Passage fish assemblage were investigated using 167 165-176 Love r3 11/17/07 12:38 PM Page 168 LOVE ET AL.: FISH ASSEMBLAGE IN THE ANACAPA PASSAGE CalCOFI Rep., Vol. 48, 2007 hierarchical cluster analysis and non-metric multi- tions, rockfishes still comprised 59.4% of all fishes. Other dimensional scaling (MDS) plots. We fourth-root trans- particularly abundant species included blue rockfish, formed fish densities (the full species set was used) to blackeye goby, blacksmith, and halfbanded rockfish. reduce the impact of extremely abundant species, and Ver milion and rosy rockfishes, senorita, lingcod, and then constructed a triangular similarity matrix among sharpnose/white seaperches comprised the remaining year-pairs using the Bray-Curtis index (Bray and Curtis top ten species or species complexes. 1957). The Bray-Curtis index is useful in ecological Many of the species we observed recruited as young- analyses because joint absences of species between sam- of-the-year (YOY) to the study reefs (we defined re- ple pairs are not used in similarity calculations (joint ab- cruited fish as those less than 10 cm long, except for sences being difficult to interpret biologically). Of further 5 cm for blackeye gobies). Of the most abundant fishes, benefit, the Bray-Curtis index is robust to non-linear species that at least occasionally recruited as YOYs in- species responses (Faith et al.