University of California, Santa Barbara, CA 93106
Tel. (805) 967-3328; Fax (805) 893-8062
A drifting kelp "raft" off the Catalina coast
© 1997 Dr. Bill Bushing
Islands represent discontinuous habitat for terrestrial taxa yet it is less obvious that they are also "biological islands" for certain marine species as well. Intertidal and subtidal marine invertebrates, especially those which lack a meroplanktonic larval stage, may face significant difficulty in crossing biogeographic barriers such as the San Pedro Channel. Taxa which are asexual, brood their young or have "crawl away" juveniles may have significantly lower inherent dispersal ability, yet are often well-represented and exhibit relatively low rates of endemism on islands. Kelp such as Macrocystis and Pelagophycus provide habitat for numerous marine species. Upon detachment from the substrate, kelp plants drift with the wind and currents, dispersing large numbers of individuals from many taxa over relatively long distances. This mechanism may assist in maintaining these species' geographic ranges as well as increasing genetic exchange between isolated populations. Marine species so dispersed have a high potential for reproductive establishment upon arrival due to the numbers transported, age structure, reproductive status and genetic diversity of the propagules. Of ecological interest is the potential simultaneous introduction of commensals, symbionts, parasites and predators. Colder ocean temperatures following the last glacial period in addition to historic reductions of kelp due to sea otter hunting, kelp harvesting and other anthropogenic activity suggest kelp may have had an even more significant role in the past.
The Channel Islands off southern California offer an excellent opportunity for the investigation of dispersal phenomena in benthic marine invertebrates. The intertidal and subtidal habitats of these islands offer examples of "biological islands." Although the substrate and medium are continuous with those of similar mainland habitats, the depth of the intervening channels creates significant biogeographic barriers.
Three mechanisms are often posited to account for the present-day distributions of benthic marine taxa: vicariance events linked to continental drift, pelagic larval dispersal, and rafting or "epiplanktonic dispersal" (Edgar 1987). The Channel Islands formed in regional tectonic events initiating in the Miocene when local folding and faulting began. The islands represent exposed portions of topographic highs along submarine ridges and most likely had no mainland connections. Therefore vicariance events are unlikely mechanisms for explaining the distributions observed. It is assumed the current biota colonized island intertidal and subtidal habitats through cross-channel pelagic dispersal, either by a dispersal stage or rafting.
Dispersal in marine invertebrates is often linked to reproductive mode and life histories (Grant 1990). Most sexually reproducing taxa offer a wide range of dispersal options. Some produce free-swimming meroplanktonic larvae, either planktotrophic or lecithotrophic, that may disperse with ocean currents. Other species brood, either ovoviviparously to a larval stage or viviparously to a benthic subadult. Non-swimming, "crawl away" juveniles are not likely to disperse readily (Highsmith 1985). Taxa which reproduce strictly asexually or through autotomy may possess no free-swimming dispersal stage.
Several authors (Fell 1962, Dell 1972, Highsmith 1985) have suggested the often regionally cosmopolitan distributions of species with limited inherent dispersal capabilities may be explained by rafting on drift material including wood and kelps such as Macrocystis. The scarcity of macroalgae in the tropics makes rafting a less likely dispersal mode there than in the temperate zone (Highsmith 1985).
The Southern California Bight, with its historically extensive kelp forests, offers an opportunity to test the hypothesis that "kelp rafting" may help explain the distribution of some marine invertebrates and the low rates of endemism in the region. Our research asked two fundamental questions. First, given local oceanographic conditions, is drift kelp capable of transporting invertebrates over the distances and drift paths required for mainland-to-island or inter-island dispersal? Second, are marine invertebrates without pelagic larval stages, such as brooders, actually found on drift kelp?
The author and his students recovered 126 drifting objects (of which 109 were species of brown algae) from December 2, 1969, to February 18, 1973, off Santa Catalina Island and analyzed the invertebrates on them. Sampling was conducted during the academic year only, with the peak months being November through March. Drift kelp was sampled by boat, superficially investigated at sea and interesting rafts brought into the lab for dissection and analysis. Species observed on the raft were identified, when possible, and recorded along with their numeric frequency. Additional data recorded included the species of kelp, holdfast diameter, overall raft dimensions, geographic location of raft, wind and drift direction, associated rafts and associated species observed at sea.
Rafts were generally sampled along a transect line from the Toyon Bay pier straight out to sea for a distance of two miles. However, opportunistic events resulted in sampling from many areas along the leeward side of the island, primarily between Long Point and the East End, from nearshore to distances of about five miles. One sample was collected two miles off Point Fermin on the mainland. Rafts without holdfasts, or with little obvious informational content, were not sampled and not recorded. Some information loss resulted during the transfer of rafts from the water to the launch and the launch to the lab. Dissection was on a laboratory wet table with the extracted specimens placed in holding tanks for identification.
Because the project spanned a period of more than four years and involved some student help, species identification skills varied. Initially there was no recognition of the different taxa of congeneric kelp and all rafts were assigned to Macrocystis pyrifera or Pelagophycus porra. Although identification of the invertebrates was largely done in our lab, some specimens were sent to experts at the Museum of Comparative Zoology in Cambridge, Massachusetts, and the Los Angeles County Museum of Natural History for verification.
The findings from our sampling of drift kelp are summarized in Tables 1, 2 and 3. The following section focuses specifically on observations made during this study while a subsequent section discusses these findings relative to the literature.
Macrocystis pyrifera was the predominant kelp species recovered during our sampling followed by Pelagophycus porra (Table 1), reflecting expectations based on local species composition. The distribution of these two species over the entire southern California Bight made it impossible to state definitely whether long distance dispersal occurred. Some specimens exhibited high degrees of deterioration or encrustation, suggesting longer periods of drift.
The presence of two rafts of Nereocystis luetkeana, and others assignable to Macrocystis integrifolia, confirmed drift distances of magnitude greater than those in the Bight are achievable. Although Abbott and Hollenberg (1976) report Nereocystis as drift from the San Diego area, and there were mistaken reports from Santa Rosa Island (P. Silva, UC Berkeley, pers. comm.), this species is known only from north of Pt. Conception. Its presence as drift material off Santa Catalina indicates dispersal on drift kelp occurs over distances sufficient to allow mainland-to-island or inter-island dispersal. Drift M. integrifolia, noted from British Columbia to central California by Abbott and Hollenberg (1976), adds further confirmation.
Our studies identified 27 species of benthic algae and vascular plants, 179 species of marine invertebrates and at least 25 vertebrates (fish, birds, marine mammals) on or in the vicinity of sampled drift kelp. Table 2 lists the frequency of the major taxonomic groups found in our samples. The presence on sampled rafts of several invertebrates lacking pelagic larvae (Table 3) validates the hypothesis that drift kelp may be a dispersal vector for them. Many of their distributions suggest patterns expected due to chance dispersal on drift kelp in the California Current system.
The anemone (Epiactis prolifera), which broods its young to a crawl-away benthic stage (Fautin and Fu-Shiang 1986, Highsmith 1985), was found on rafts with young in its external brood pits. Another coelenterate (Balanophyllia elegans) identified on sampled rafts is the only scleractinian that broods its embryos to a large, strictly benthic, crawling planula (Fadlallah and Pearse 1982).
Morris et al. (1980) list several other brooding species that were observed on the sampled rafts. The amphipod (Caprella californica), a strictly dioecious brooder, occurred on several rafts. Females of Pycnogonum stearnsi brood eggs that hatch as non-swimming subadults, and were noted on the rafts. Several species of Phycolimnoria are strictly brooding taxa. Among the molluscs found on our rafts, Crepidula includes several brooding species and Octopus bimaculatus broods eggs that hatch into larvae of short duration.
Echinoderms observed on our rafts are of interest due to the large numbers encountered. The brooder Leptasterias hexactis [= L. aequalis] (Morris, et al. 1980, Highsmith 1985), was recovered on Nereocystis luetkeana indicating long-distance drift. Henricia leviuscula is another brooding asteroid observed on one raft. The ophiuroid (Amphipholis squamata) broods its eggs to a large juvenile and has no swimming stage (Morris, et al. 1980, Highsmith 1985, Walker and Lesser 1989). This species has an often locally patchy but wide distribution and is associated with floating material (Morris, et al. 1980). The ophiuroid (Ophioplocus esmarki) likewise broods its young to a juvenile stage.
Several tunicates found on our rafts either brood or have short-lived larvae (Morris, et al. 1980). Aplidium californicum [= Amaroucium californicum] retains its eggs internally, brooding them to a short-lived tadpole stage. Boltenia villosa releases eggs which develop into short-lived larvae, remaining in the water column only 2-6 days.
Based on life history information (Morris et al. 1980), some invertebrates on the sampled rafts reproduce asexually. Several sponges, which may reproduce through fragmentation, were observed. The coelenterate (Corynactis californica) reproduces strictly by asexual fission. Linckia columbiae may autotomize readily (McAlary 1987).
Some species found on rafts such as abalone (Haliotis fulgens) possess pelagic larvae which settle out too quickly for adequate planktonic dispersal (Tegner and Butler 1985). The presence of Lepas spp. and juvenile Pachygrapsus crassipes suggests species possessing larvae capable of long-distance dispersal may still benefit from kelp rafting if their larvae metamorphose and settle on a drifting kelp raft, facilitating dispersal. Species that postpone metamorphosis until suitable substrate is encountered may especially benefit. Many nudibranchs (Melibe leonina, Triopha spp., Hermissenda crassicornis) transported as adults were observed laying eggs on the kelp raft that could later hatch, releasing pelagic stages into the open water.
Several kelps (Laminariales, Lessoniaceae) from central and southern California may play a role in dispersal. This research focused on Macrocystis pyrifera, M. integrifolia, Pelagophycus porra and Nereocystis luetkeana as recognized by Abbott and Hollenberg (1976). Morphologies similar to species proposed, including Macrocystis angustifolia (Neushul 1971) and Pelagophycus giganteus (Dawson 1962), were observed.
Each of the major kelps offers a unique combination of geographic range, depth and substrate preferences, and holdfast morphology that affect their suitability as rafting vehicles and the species which might be transported by them. These differences are summarized in Table 4 compiled largely from Abbott and Hollenberg (1976).
Geographic range and depth distribution determine the raft's origin and the associated invertebrate species complex that might be rafted. The five kelps range geographically from Alaska to Baja California in depths from the lower intertidal to 90 meters. Differentiation between species found only north (M. integrifolia and N. luetkeana) or only south (P. porra) of Point Conception reflect its recognition as a biogeographic transition zone (Seapy and Littler 1980). Local subspecific morphological variation such as the "island" Macrocystis pyrifera on Santa Catalina (Harger 1983) or local "endemics" like Pelagophycus giganteus (Dawson 1962) may allow more precise identification of raft origin. Substrate preferences for the different kelp species may determine the susceptibility to detachment and the associated invertebrate complex.
Holdfast morphology helps establish the "carrying capacity" of the raft as well as the degree of protection from predation, or other loss, it offers to rafted invertebrates. The massive, complex holdfasts of Macrocystis pyrifera have a higher transport capacity than the smaller, less complex holdfasts of Pelagophycus and Nereocystis or the rhizomatous, spreading holdfast of Macrocystis integrifolia.
Pelagophycus porra and Nereocystis luetkeana morphologies involve single stipes that are more likely to break under storm duress leaving the holdfast attached to the substrate. Macrocystis with its multiple fronds emanating from the holdfast is more likely to detach intact. The structural simplicity of P. porra and N. leutkeana blades makes them less suitable for long distance transport of invertebrates than the more complex fronds of Macrocystis.
The role of kelp as an agent for the colonization of the Channel Islands cannot be adequately evaluated based on current patterns of distribution. Changes in kelp cover over geologic and historic time suggest the role of kelp may have been more significant in the past.
Modern kelp (Julescrania and possibly Pelagophycus) first appeared in the California fossil record during the Miocene (Clayton 1984, Parker and Dawson 1965), when the southern California Channel Islands began forming. As the islands emerged and their subtidal habitats formed, kelp was available as a potential dispersal mechanism for their colonization by invertebrates. Cooler temperatures in the North Pacific during the Pleistocene would enhance the distribution of kelp in the Bight (Luhning 1990). Concurrent sea level lowering likely increased the areal extent of intertidal and subtidal habitats around the emerging islands, and reduced mainland-to-island and inter-island distances, enhancing the probability of transport by kelp.
Aboriginal occupants of Catalina Island harvested marine resources from kelp beds over several thousand years, impacting local kelp (Meighan 1959). Other historic and recent anthropogenic factors including sea otter hunting, commercial and sport abalone and sea urchin harvesting, sewage pollution, kelp harvesting also affected kelp bed distribution (Harger 1983).
Kelp surveys, aerial photographs and satellite imagery dating back to 1911 (Crandall 1912, Jensen et al. 1987) allow quantitative evaluations of historic changes in kelp canopy. Two such studies indicate Santa Catalina Island kelp beds declined in area from a maximum in 1911 to a low in the mid-1950's (Hodder and Mel 1978, Neushul Mariculture Inc. 1981). These same studies found similar trends in the beds around San Clemente Island, Santa Cruz Island and Santa Rosa Island during the same period.
Important causes of detachment for large kelp sporophytes include actual removal by storm waves (North 1991), weakening of holdfasts and increased potential for detachment in storm-damaged plants (McPeak et al. 1988) and grazing on the holdfast by gribbles (Phycolimnoria algarum). Catalina is impacted by NW or WNW winter storms on the windward side, Santa Ana storms on the leeward side or southern swell during the summer months. The seasonal timing of storms and detachment peaks also affects the taxonomic composition, age structure and biomass of rafted invertebrates in the Bight.
The rate of export of detached plants from beds on open coasts is significant. Harrold and Lisin (1989) estimated export rates of Macrocystis pyrifera from Monterey Peninsula forests at 130,000 tons per year. Given such rates, there is a high potential for detachment and drift in the Bight.
Sightings of drift kelp at sea, and speculation on its potential role as a dispersal agent, date back many years. Documented evidence of actual transport is rarer. Two authors (Fell 1962, Arnaud et al. 1976) offer evidence of dispersal by invertebrates on drift kelp over distances of several thousand kilometers. While no comparable distances were observed in our study, duration of drift in local waters is sufficient to effect transport over distances comparable to those in the Bight.
The complex ocean circulation pattern in the Southern California Bight, with its seasonally reversing gyre, significant inter-annual variation and mesoscale eddies has been described previously (Owen 1980, Seapy and Littler 1980, Lynn and Simpson 1987, Poulain and Niiler 1989, Pares-Siera and O'Brien 1989). Velocities for the California Current are variously reported as 4 to 50 cm/sec but typically less than 25 cm/sec (Jennings and Schwartzlose 1962, Reid and Schwartzlose 1962, Owen 1980, Lynn and Simpson 1987, Poulain and Niiler 1989).
This complex current pattern makes it difficult to measure "dispersal distances." Linear distances between islands and mainland sources have less value in predicting dispersal potential than for motile forms such as large marine fishes. The seasonally reversing circulation pattern in the Bight allows transport from both northern and southern origins.
Macrocystis canopy may protrude 1-2 cm above the water surface (Jensen, et al. 1980), suggesting wind patterns may affect drift direction and velocity at sea, as borne out by observations during this research. Harrold and Lisin (1989) also found drifting kelp off Monterey was influenced by seasonally variable winds that dominated current directions.
Once detached, kelp must persist in the open ocean long enough to effect transport over the required distances. Factors such as water temperature, nitrate levels, and population densities of grazing organisms may play as significant a role as the local current and wind patterns in determining drift kelp dispersal (Edgar 1987). Laboratory experiments on Macrocystis pyrifera found pneumatocysts lost their buoyancy after about 7 days (Yaninek 1980). Edgar (1987) noted holdfasts disintegrated in less than 6 months due to grazing by boring isopods (Phycolimnoria), placing a potential upper limit on drift dispersal for infested rafts. Phycolimnoria were also observed on rafts in our study with evidence of significant damage to the holdfasts.
Rafted invertebrates must remain alive until the raft reaches a new habitat suitable for colonization. Physical factors for organisms on drift kelp differ from those in attached forests, and may alter mortality rates. Increased IR, UV and visible light levels, and temperature changes result from the buoying up of the holdfast and the reduction in canopy shading from neighboring plants. Holdfasts entangle in the floating fronds and become desiccated. Salinity, O2 and CO2 levels, dissolved nutrients and water turbulence may also exhibit departures from those normally experienced at the bottom.
Trophic interactions which may sustain or deplete the rafted organisms include grazing on the kelp or attached algae by herbivores, utilization of organic matter in holdfast sediments, or predation on other rafting invertebrates. Reduced plankton densities in pelagic waters compared to the richer nearshore environment may affect filter feeders. Aquatic birds including gulls, cormorants and great blue herons, as well as fish, were observed feeding on the invertebrates on drifting rafts.
Beaching may involve high mortality. Lovenburg (UC Davis, pers. comm.) found 27 of 46 mussels on 31 tagged algae were lost during drift and beaching. Rocky subtidal species rafted on kelp are often deposited in sandy subtidal areas where they become easy prey for predators or become desiccated and die before locating suitable habitat.
Kelp forests offer a vertically and horizontally structured habitat for many marine taxa. More than 300 algae and almost 800 animals are associated with M. pyrifera in southern California and northern Baja California (Earle 1980, Foster and Schiel 1985). This species complex includes a wide range of candidates for kelp raft dispersal.
Observations of motile invertebrates associated with Macrocystis off Santa Catalina Island by Coyer (1984) noted 114 species with the mean number of species and biomass both increasing from canopy to holdfast. Although species composition did not vary appreciably, biomass showed marked seasonal increases during winter months when most of our sampling was undertaken.
The kelp holdfast provides protection for many species of sponges, worms, crabs and other arthropods, bryozoans, brittle stars, sea cucumbers and other invertebrates (McConaughey 1985). Species inventories on attached holdfasts identified 128 to 175 taxa (Ghelardi 1971, Jones 1971, Foster and Schiel 1985, McPeak et al. 1988). These numbers are comparable to the 179 species identified on drifting kelp in this research.
Kelp rafting enhances the reproductive and ecological establishment of a taxon due to the potential numbers rafted in a single dispersal event. One sampled raft had 1,500-2,000 ophiuroids on it, while another had 400-600 of three different species. Transport of taxa in such high numbers also has consequences for their genetic diversity.
Dispersal on drift kelp may effect a shift in the age structure of propagules towards the adult or subadult stages. Dispersed individuals in reproductive condition on sampled rafts included gravid decapods, brooding Epiactis prolifera, and egg laying and synchronously copulating nudibranchs (Melibe leonina, Triopha spp., Hermissenda). Relatively high numbers of mature organisms offer higher probability of establishment than a flux of non-reproductive pelagic larvae. Dell (1972) states brooding females may be better colonists than larvae. Highsmith (1985) suggests the small adult sizes of most brooding invertebrates is advantageous for dispersal by rafting.
The simultaneous transport and introduction of predator/prey, parasite/host, commensal and symbiont species pairs is more likely on drift kelp. Predator/prey introductions observed in our studies included caprellid amphipods and hydroids, nudibranchs (Corambe pacifica) and bryozoa (Membranipora spp.), gastropods (Tylodina fungina) on sponges (Aplysina fistularis), and a pelagic nudibranch (Fiona pinnata) and barnacle (Lepas sp.). Parasite/host pairs included the isopod (Phycolimnoria algarum) and Macrocystis. Recognized commensals include the pea crab (Pinnixa) transported within annelid (Chaetopterus) tubes, and zooxanthellae within anthozoan coelenterates (Anthopleura spp.) although zooxanthellae are also known from coelenterate planula larvae (Trench 1987).
Kelp distribution is important in understanding the biogeography of the Channel Islands region (Neushul et al. 1967) and dispersal via drift kelp expands this role. Both today and in the past, kelp has probably played a role in the maintenance and expansion of species' geographic ranges, enhanced the genetic diversity of the propagules, helped maintain gene flow with source populations, and promoted relatively low rates of speciation and endemism.
Seapy and Littler (1980) found strong biogeographic affinities in marine invertebrates between Santa Catalina, San Clemente and Anacapa Islands. Regional ocean current patterns may be helpful in explaining this assuming dispersal by pelagic larvae or drift kelp. A similar study on the marine algae in the Bight revealed a similar pattern (Murray et al. 1980). Kelp rafting may play a role in the distribution of the kelp themselves, and attached understory algae with limited spore or gamete dispersal.
The transport of Pugettia producta (= Epialtus productus) and Leptasterias hexactis on Nereocystis luetkeana from the cold-temperate Oregonian Province north of Point Conception to the warm-temperate California Province indicates mixing of marine invertebrates from different biogeographic regions may result. The recovery of Eupenctata quinquesemita on drift Macrocystis off Santa Catalina Island, south of its normal geographic limit at Morro Bay is another example.
There are also paleobiogeographic, paleoecological and paleontological ramifications to kelp rafting. Valentine and Lipps (1963) noted late Pleistocene fossil assemblages from mixed habitats on Anacapa Island which Lovenburg (pers. comm.) suggested may have resulted from kelp transport of material from one habitat type into another.
Figures for this paper are available from this link
This research verified the hypothesis that drift kelp may be partially responsible for the colonization of marine habitats on the southern California Channel Islands from local mainland populations. Kelp rafts from source areas outside the Bight were found to enter our region with living marine invertebrates attached, some not normally found in our waters. Benthic invertebrate taxa lacking pelagic dispersal stages were observed on rafts, verifying the potential role of kelp rafting in their distribution. The transport of large numbers of mature individuals enhances the role of drift kelp in ensuring reproductive and ecological establishment even in taxa which possess dispersal stages. The co-introduction of ecologically linked species (commensals, etc.) may be an important consequence of kelp rafting. The probable higher density of kelp forests in past geologic and historic periods suggests an enhanced role in affecting local biogeography during the critical stages when island marine habitats were evolving.
This research was supported in part by National Science Foundation grant GB-3532 to H. Barraclough Fell, Museum of Comparative Zoology, Harvard University. I am grateful to Dr. Fell and Dr. E. O. Wilson for initiating my interest in kelp rafting and the biogeography of islands. This project would not have been possible without the support of the Catalina Island School, its staff and students, especially John S. Iversen and Barry Aires. Assistance with species identification was provided by Dr. Ruth Turner of the MCZ, Dr. James Morin of UCLA and Dr. Camm Swift of the Los Angeles County Museum of Natural History. Dr. Merv Lovenburg, formerly of the University of California at Davis, also provided access to his data and comments through the early stages of our work.
This research paper is by Conservancy Vice-President Dr. William Bushing. It was published in the Proceedings of the Fourth California Islands Symposium held in Santa Barbara, California. It may be cited as follows: Bushing, W. W., 1994. Biogeographic and ecological implications of kelp rafting as a dispersal vector for marine invertebrates. In Halvorson, W. and G. Maender (eds.), Proceedings of the Fourth California Islands Symposium: Update on the Status of Resources, March 22-25, 1994. Santa Barbara Museum of Natural History (Santa Barbara, CA). pp. 103-110.