Rapid Assessment of Nonindigenous Marine Species on Coral Reefs in the Main Hawaiian Islands1

S. L. Coles,2 F. L. M. Kandel,3,4 P. A. Reath,4 K. Longenecker,2 and L. G. Eldredge2

Abstract: Coral reefs at Kaua‘i, O‘ahu, Moloka‘i, Maui, and Hawai‘i were surveyed using a rapid assessment method for marine nonindigenous and cryptogenic species commonly found in Hawaiian harbors and embayments with restricted circulation. In 41 sites surveyed by rapid assessment 26 nonindigenous and cryptogenic species (three algae, 19 invertebrates, and four fishes) were recorded from a total of 486 total taxa identified, and 17 of the nonindigenous and cryptogenic species occurred at only one or two sites. No more than six nonindigenous and cryptogenic species were recorded at any one site, and 21 of the 41 sites had fewer than three. By comparison, laboratory identification of samples collected from seven of the sites closest to harbors found 6–23 nonindigenous and cryptogenic species per site. Values for nonindigenous and cryptogenic species from rapid assessment were compared with factors potentially influencing spread and proliferation of introduced marine species. These factors included distances from harbors, boat-launching ramps, stream mouths, and shorelines; degree of shoreline urbanization; quantity of artificial surfaces in the water; reef condition and isolation from the open ocean; and native species richness. A best subsets regression model explained over 65% of the variance in nonindigenous and cryptogenic species from two predictor variables and their interaction: isolation from the open ocean and number of native taxa, with most of the variance explained by a highly significant relationship of nonindigenous and cryptogenic species with isolation from open-ocean conditions.

Introductions of nonindigenous (intro- Sommer 1995, Daehler and Strong 1996, duced) marine species are considered to have Greenberg et al. 1996, Ruiz et al. 1997, 2000, escalated in the last 30 yr (e.g., Carlton 1985, Bax et al. 2001), sometimes with serious neg- Carlton et al. 1990, Carlton and Geller 1993, ative consequences when introduced species Cohen et al. 1995, Gosliner 1995, Mills and become invasive (i.e., compete with native species to the point that they alter the structure or function of the invaded ecosystem). 1 This project was conducted with the ﬁnancial sup- Marine species invasions have been ranked port of the Hawai‘i Coral Reef Initiative and National among the most serious potential perturba- Oceanic and Atmospheric Administration grant no. tions of marine ecosystems (Carlton 1994), NA03NOS4260044. This is contribution number HBS 2006-001 from the Hawai‘i Biological Survey and contri- and alteration of habitats and food webs by bution 1224 from the Hawai‘i Institute of Marine Biol- invasive species has been proposed as a poten- ogy. Manuscript accepted 30 December 2005. tial major factor contributing to degradation 2 Department of Natural Sciences, Bishop Museum, of coral reefs (Birkeland 2004). 1525 Bernice Street, Honolulu, Hawai‘i 96817 (e-mail: In Hawai‘i there is a substantial pool of [email protected]). 3 Department of Zoology, University of Hawai‘i at potentially invasive marine organisms that Ma¯noa, Honolulu, Hawai‘i 96822. have reached the Islands over at least the last 4 Hawai‘i Institute of Marine Biology, P.O. Box 1346, century. Eldredge and Carlton (2002) desig- Ka¯ne‘ohe, Hawai‘i 96744. nated 343 Hawaiian marine or brackish-water species as introduced or cryptogenic (i.e., of Paciﬁc Science (2006), vol. 60, no. 4:483–507 uncertain geographic origin sensu Chapman : 2006 by University of Hawai‘i Press and Carlton [1991] and Carlton [1996]). A All rights reserved number of comprehensive surveys for marine

483 484 PACIFIC SCIENCE . October 2006 introduced species have been completed in the likely to occur, (2) oligotrophic open-ocean Hawaiian Islands (Coles et al. 1997, 1998, reef environments may not provide sufficient 1999a,b, 2002a,b, 2004a,b, DeFelice et al. food to support the filter-feeding inverte- 1998, 2002), Johnston Atoll (Coles et al. brates that prevail among nonindigenous 2001), American Samoa (Coles et al. 2003), organisms, (3) generally higher native species Guam (Paulay et al. 2002), and port areas of richness on ocean-exposed coral reefs may act northern Australia (e.g., Hewitt et al. 1998, to limit introduced species that proliferate in Hoedt et al. 2000, 2001a,b, Russell and Hew- the less-diverse communities of harbors or itt 2000, Neil et al. 2001). Most of these estuaries. Few introduced species have been studies have focused on harbors or disturbed noted in higher-diversity tropical regions for areas with limited oceanic circulation. ports in northern Australia (Hewitt et al. Invasive introduced algae have monopo- 1998, Hoedt et al. 2000, 2001a,b, Russell and lized nearshore reefs (Rodgers and Cox 1999, Hewitt 2000, Neil et al. 2001), Guam (Paulay Woo et al. 1999, Smith et al. 2002) through- et al. 2002), and at American Samoa (Coles out the main Hawaiian Islands. An invasive et al. 2003), where native species richness is octocoral was recently found overgrowing substantially higher than in Hawai‘i. Hawaiian deep-water black corals (Grigg In this study we compared the occurrence 2003, 2004), and an abundant introduced of introduced species among 41 coral reef sponge may be similarly impacting shallow- sites in various environmental conditions and water reef corals in southern Ka¯ne‘ohe Bay distances from harbors, piers, and boat ramps (Coles and Bolick 2006). An introduced, ag- throughout the main Hawaiian Islands. We gressive mantis shrimp has displaced native evaluated the results in terms of important mantis shrimp species from coral rubble hab- factors that may influence the occurrence of itats in Hawai‘i (Kinzie 1968, 1984), and reef introduced species at the reef sites. fishes purposely introduced to Hawai‘i in the 1950s are common to abundant on reefs throughout the Hawaiian archipelago (Ran- materials and methods dall and Kanayama 1972, Randall 1987). Field Techniques However, despite the potential importance of invasive introduced species on coral reefs, A total of 41 coral reef sites distributed across little is known about the occurrence or im- five islands was surveyed using a rapid assess- pact of most marine introductions in Hawai‘i ment technique. Sites were selected to include or elsewhere in the tropical Pacific (Coles and a variety of environmental characteristics that Eldredge 2002). Further, there has been no might influence the establishment of intro- evaluation of the factors that may influence duced species, such as proximity to harbors the establishment and proliferation of intro- and boat ramps, distance from shore or duced species on coral reefs. streams, presence of artificial structures in We used a standardized rapid assessment the water or degree of shoreline develop- technique to determine the presence of intro- ment, reef condition, and exposure to the duced marine species on coral reefs for the open ocean. Selection criteria also included five largest Hawaiian Islands and evaluated utilization of sites established by ongoing factors that may influence introduced species reef-monitoring programs that could provide occurrence. Previous findings (Coles et al. historical and future data on reef commu- 1997, 1998, 1999a,b, 2001, 2002a,b, 2003, nities. Twenty-four of the 41 rapid assessment DeFelice et al. 1998, 2002) suggested that sites coincided with Hawai‘i Coral Reef As- three principal factors may influence the dis- sessment and Monitoring Program (CRAMP) tribution and proliferation of most nonindig- sites ( Jokiel 2002, Brown et al. 2004), and enous species in Hawai‘i and other tropical eight of the 10 surveys on Hawai‘i were areas: (1) isolation may limit recruitment made at West Hawai‘i Aquarium Project from likely sources such as harbors and (WHAP) sites (Tissot et al. 2001). boat landings where introductions are mostly Anticipating that introduced species were Nonindigenous Marine Species on Hawaiian Coral Reefs . Coles et al. 485

Figure 1. Kaua‘i coral reef stations. likely to be low in frequency and abun- and F.L.M.K.) swam in tandem for 15 min dance, we utilized a replicable methodology along the perimeter of the triangle, recording to examine large reef areas within no- the first occurrence of all invertebrates, fishes, decompression time limits for up to three and identifiable macroalgae along a swath up scuba dives per day at 10–20 m. Locations of to 2 m on either side. For the next 15 min we the 41 sites surveyed are shown in Figures recorded organisms that occurred within the 1–5, and their station numbers, locations, 312-m2 triangular area. Finally, we spent 15 depth ranges, and coordinates are listed in min searching outside the triangle recording Table 1. Our timed search approach provided all taxa not previously observed. In addition a standardized and sufficiently large search to the observations made by these two ob- area to assure that most macrobiota at the servers, a third diver (P.A.R.) searched sub- site were observed. We recorded the location habitats known to support nonindigenous and of each reef site using Global Positioning cryptogenic species (i.e., overhangs, crevices, System and deployed a 50-m transect line dead coral heads, coral rubble, algal turf and parallel to the shore for 25 m, then turned tufts) at 28 sites on Kaua‘i, Moloka‘i, Maui, the line at a right angle for the remaining 25 and Hawai‘i (Table 1) and recorded cryptic m, with the resulting triangular observation organisms. We recorded all field-identifiable area established by the hypotenuse thus ap- organisms on underwater paper, and we col- proximating 312 m2. Two observers (S.L.C. lected species suspected to be introduced and 486 PACIFIC SCIENCE . October 2006

Figure 2. O‘ahu coral reef stations. transported them to the laboratory for iden- native status and the degree to which they tification and/or verification by taxonomic are widespread in Hawai‘i. This approach re- experts. We designated species as nonindig- sulted in conservative estimates for the num- enous, cryptogenic, or native according to bers of known or possible introductions for J. T. Carlton and L.G.E. (unpubl. data) each site. and the Checklist of the Marine Inverte- brates of the Hawaiian Islands (http://www2 Evaluation of Rapid Assessment Technique .bishopmuseum.org/HBS/invert/list_home .htm). Taxa not identified to species were Our visual rapid assessment technique did not designated native unless they were previously include minute organisms such as amphipods, known unnamed introductions (e.g., Ascidia isopods, tanaids, and small hydroids that are a sp. A [Abbott et al. 1997; J. T. Carlton and major component of the introduced species L.G.E., unpubl. data]). New reports for Ha- community but can only be identified by wai‘i were designated cryptogenic or ‘‘of un- microscopic examination. However, the ob- certain origin that is neither demonstrably servations included introduced macrobiota native or introduced’’ (Carlton 1996:1653). such as algae, fishes, and invertebrates such We consider this an appropriate category for as sponges, larger hydroids, decapods, bryo- new species reports until sufficient informa- zoans, and tunicates. To evaluate the sensitiv- tion is available to verify their introduced or ity of the rapid assessments for determining Nonindigenous Marine Species on Hawaiian Coral Reefs . Coles et al. 487

Figure 3. Moloka‘i coral reef stations. all nonindigenous and cryptogenic species, from rapid assessment surveys at the same one of the field team (P.A.R.) made system- sites. atic collections of benthic organisms at seven stations on islands other than O‘ahu (Table Variables Influencing Introductions 1). These sites were on reefs closest to the main interisland shipping ports of Kaua‘i A number of both natural and anthropogenic (KA1), Moloka‘i (MO7), and Hawai‘i (HA1 factors may influence the introduction, distri- and HA10) or adjacent to small-craft harbors bution, and proliferation of nonindigenous (KA5, MO3, and MA7). Samples were col- marine species. We evaluated nine potential lected opportunistically as a composite sam- contributing factors as part of this study: (1) ple from all available subhabitats within the Hawaiian harbors have a large complement reef in the vicinity but outside the 312-m2 tri- of introduced species (Coles et al. 1997, angle (i.e., bases of coral heads, coral rubble 1999a,b, 2004b) and proximity to harbors macroalgae, and turf algae, and from cryptic may increase the probability of occurrence areas under ledges and in crevices) and pre- of introduced species on reefs. (2) Similarly, served in 70% ethanol. These organisms small boats may inadvertently transfer intro- >0.5 mm were identified in the laboratory to duced species, thus proximity to boat- species or the lowest practical taxon. These launching areas may also be a factor in results were compared with those obtained spreading introductions. (3) Flooding and 488 PACIFIC SCIENCE . October 2006

Figure 4. Maui coral reef stations. sedimentation from streams may result in a sidered were as follows: (5) the amount of stressed environment more conducive to es- artificial substrate in the water or along the tablishment of introduced species on de- shoreline; (6) the degree to which the adja- nuded or disturbed reef surfaces. Studies in cent shoreline was urbanized; (7) the degree estuaries of O‘ahu (Englund et al. 2000) to which the surveyed area was isolated from showed a high incidence of introduced spe- the open ocean; and (8) the general condition cies. Therefore, proximity to stream mouths of the coral reef itself, again on the assump- was evaluated as a factor potentially influenc- tion that a disturbed reef was more suscep- ing nonindigenous and cryptogenic species. tible to introductions and invasions. We (4) Some introduced species, especially alien ranked, on site, the condition of these latter algae and the invertebrates that occur with four factors by the criteria in Table 2. Finally, them, are most abundant along or near based on previous observations that numbers shorelines, and distance from shoreline was of introduced species appeared to be inverse- evaluated as a possible influence on intro- ly correlated with the numbers of native spe- ductions. We used Geographic Information cies (Coles et al. 1997, Coles et al. 1998, System–based maps (Figures 1–5) to deter- 1999a,b, 2002a,b), we included (9) the num- mine the distances of these four features from ber of native taxa, or native species richness, each survey site. at each site as a predictive variable in our Other potential determining factors con- analyses. Nonindigenous Marine Species on Hawaiian Coral Reefs . Coles et al. 489

Figure 5. Hawai‘i coral reef stations.

results bution of the 26 nonindigenous and cryptogenic species among stations surveyed is Numbers of Nonindigenous and Cryptogenic shown in Table 3. Most nonindigenous and Species and Total Taxa cryptogenic species occurred infrequently Using rapid assessment, we identified a total among the 41 stations, with only eight species of 486 taxa, ranging from a maximum of 151 recorded at more than four sites. The most taxa at Station 6 on Hawai‘i to a minimum of frequent nonindigenous and cryptogenic spe- 40 taxa at Station 2 on Kaua‘i and at Station 3 cies were the fishes roi (Cephalopholis argus on Moloka‘i. All taxa identified by rapid as- Wooster, 23 sites) and ta‘ape (Lutjanus kas- sessment at the 41 sites are listed in Appendix mira (Forsska˚l), 10 sites), the conical hoof B of the full report (Coles et al. 2004a). The shell Hipponix australis Lamarck (17 sites), numbers of total taxa recorded by rapid the didemnid ascidian Didemnum cf. candidum assessment at each station are shown in Savigny (14 sites), the spiny alga Acanthophora Figure 6. Of these, 26 were cryptogenic spicifera (Vahl) Børgesen (7 sites), the sponge or introduced species (collectively termed Mycale cf. armata Thiele (5 sites), the Christ- nonindigenous and cryptogenic species), mas tree hydroid Pennaria disticha (Goldfuss) comprising approximately 5% of all taxa (5 sites), and the feather duster worm Sabellas- identified in rapid assessment surveys. Distri- tarte spectabilis (Grube) (5 sites). The remain- 490 PACIFIC SCIENCE . October 2006

TABLE 1 Locations, Depths, and Coordinates for Stations Surveyed

Island Station Code Location Depth (m) Latitude N Longitude W

Kaua‘i 1*† KA1 Marriott Hotel Reef 2.5–6 2157 036.3 00 15921 014.6 00 2 KA2 Beach House Reef 1–1.5 2153 05.9 00 15928 045.5 00 3 KA3 Ho‘ai Bay 7.5–10.5 2152 048.7 00 15928 028.3 00 4 KA4 Ho‘ai Bay 2.5–3.0 2152 053.4 00 15928 034.2 00 5*† KA5 Port Allen Harbor 7.5–10.5 2154 03.5 00 15935 049.9 00 6 KA6 ‘‘Tiger’s’’ 9.0–11.0 2155 045.5 00 15939 011.5 00 7 KA7 No¯milu Pond 7.5–9.0 2153 013.7 00 15931 058.8 00 8 KA8 Kukui‘ula 6.5–7.5 2153 08.4 00 15929 016.8 00 O‘ahu 1† OA1 Ka¯ne‘ohe Bay, Waia¯hole 1.0–5 2128 035.2 00 15749 055.0 00 2† OA2 Ka¯ne‘ohe Bay, He‘eia 0.5–6 2126 048.0 00 15748 037.0 00 3† OA3 Na¯na¯kuli Point 4–5 2122 020.3 00 15808 032.1 00 4† OA4 Kahe Beach Park 3–4 2121 034.1 00 15808 06.1 00 5† OA5 Pu¯pu¯kea 1–6 2138 046.6 00 15803 052.4 00 6† OA6 Hanauma Bay 5–7 2116 06.4 00 15741 043.5 00 Moloka‘i 1 MO1 Pu¯ko‘o Nearshore 1–6 2104 018.8 00 15648 016.1 00 2 MO2 Pa¯la¯‘au 10 2105 030.4 00 15706 038.7 00 3*† MO3 Hale o Lono Reef 9–12 2105 01.4 00 15714 057.7 00 4 MO4 Pu¯ko‘o Offshore 1–3 2103 055.8 00 15647 038.4 00 5 MO5 Kamalo¯ 3–17 2102 040.6 00 15654 01.1 00 6 MO6 Kamiloloa 5–7 2104 018.2 00 15700 09.7 00 7*† MO7 Kaunakakai Reef 7–10 2104 059.6 00 15702 034.9 00 8 MO8 Hotel Moloka‘i 0.25–1 2104 028.7 00 15659 048.7 00 Maui 1 MA1 Kahekili 1.5–5 2056 022.2 00 15641 045.8 00 2 MA1 Olowalu 2.5–3 2048 042.2 00 15636 045.5 00 3 MA1 Papa‘ula Point 9–12 2055 039.3 00 15625 044.9 00 4 MA1 Honolua Bay 2.5–8.5 2101 08.9 00 15638 033.2 00 5 MO8 Puamana 3–3.5 2051 029.9 00 15640 008.4 00 6 MA1 Ma¯la Wharf 3–9 2053 023.2 00 15641 026.8 00 7*† MA1 Ma¯‘alaea Reef 2.5–5.5 2047 031.6 00 15630 045.8 00 8 MA1 Kanahena Bay 3–8 2038 05.3 00 15629 057.6 00 9 MA1 Molokini Crater 7–8.5 2037 014.9 00 15626 030.8 00 Hawai‘i 1*† HA1 Kawaihae Reef 4–9 2001 055.5 00 15550 09.4 00 2 HA2 Puako¯ 6–9 1958 023.2 00 15551 004.4 00 3 HA3 ‘Anaeho‘omalu 8.5–9 1957 02.2 00 15552 008.4 00 4 HA4 Keawaiki 12–14 1953 039.0 00 15554 045.4 00 5 HA5 Kualani Point 9.5–10.5 1933 005.7 00 15557 054.2 00 6 HA6 Red Hill 5.5–13.5 1930 028.8 00 15557 019.5 00 7 HA7 North Keauhou 6–6.5 1934 017.8 00 15558 020.2 00 8 HA8 South Oneo Bay 8–9 1938 004.8 00 15601 033.4 00 9 HA9 Papawai Bay 7–13.5 1939 000.6 00 15601 033.4 00 10*† HA10 Leleiwi Bay 6.5–10 1944 014.3 00 15501 016.5 00

*, Sites of sample collection and laboratory identiﬁcation; †, sites where rapid assessment surveys were conducted by two instead of three observers. ing 18 nonindigenous and cryptogenic species Ma¯la Wharf, Maui (MA6); and Ka¯ne‘ohe occurred at only 1–3 of the 41 sites surveyed. Bay, O‘ahu (OA2). Five nonindigenous and The number of nonindigenous and cryp- cryptogenic species occurred at the other togenic species recorded by rapid assessment Ka¯ne‘ohe Bay site (OA1); the Marriott Hotel at each site surveyed is also shown in Figure Reef, Kaua‘i (MA1); and at the Puako¯ and Red 6. A maximum of six nonindigenous and Hill sites on Hawai‘i (HA2 and HA6). No cryptogenic species occurred at one site each nonindigenous and cryptogenic species were at Port Allen Harbor Reef, Kaua‘i (KA5); detected at the Kamiloloa, Moloka‘i (MO6); TABLE 2 Description of Characteristics for Each Category of Index Values

Index Value Parameter

Artificial substrate 1 No modification of shoreline or artificial structure in water 2 <25% of shoreline hardening, no structure in the water 3 26–50% shoreline hardening or structures in the water 4 51–75% shoreline hardening 5 Shoreline >75% modified or hardened Urbanization 1 Remote: no habitation or shoreline development in view 2 Rural: 1 to 5 houses or buildings, no shoreline development 3 Residential: 6 to 30 houses or buildings, some development at or near shoreline 4 Urbanized: >30 houses or buildings, high shoreline development 5 Industrialized, commercial shoreline usage and development Ocean restriction 1 Open ocean 2 Semiexposed coastline 3 Embayment 4 Semienclosed harbor 5 Highly enclosed harbor with narrow access Reef condition 1 Excellent: coral cover >50% 2 Good: coral cover 26–50%, macroalgae rare 3 Fair: coral cover 11–25%, some algae present 4 Poor: coral cover 6–10%, algae common to abundant 5 Highly degraded: coral cover 0–5%, high sedimentation

Figure 6. Distributions of numbers of total taxa and nonindigenous and cryptogenic species determined by rapid assessment. TABLE 3 Introduced and Cryptogenic Species Recorded by Rapid Assessments

Site

Island Species Status 1 2 3 4 5 6 7 8 9 10

Kaua‘i Mycale cf. armata Thiele, 1903 Introduced x Gelloides cf. ﬁbrosa (Wilson, 1925) Introduced x Dynamena crisoides Lamaroux, 1824 Cryptogenic x Pennaria disticha (Goldfuss, 1820) Introduced x x x Chaetopterus sp. Cryptogenic x Sabellastarte spectabilis (Grube, 1878) Introduced x x Salmacina dysteri (Huxley, 1855) Introduced x Hipponix australis (Lamarck, 1819) Cryptogenic x x x x Schizoporella cf. errata (Waters, 1878) Introduced x Didemnum cf. candidum Savigny, 1816 Introduced x x x x x x x Lutjanus kasmira (Forsska˚l, 1775) Introduced x x x Cephalopholis argus Bloch & Schneider, 1801 Introduced x x x Total nonindigenous and cryptogenic species 5 1 4 4 6 2 3 3 O‘ahu Acanthophora spicifera (Vahl) Børgesen Introduced x Kappaphycus alvarezii (Doty) Doty Introduced x x Mycale cf. armata Thiele, 1903 Introduced x x Sigmadocia caerulea Hechtel, 1965 Introduced x Sabellastarte cf. spectabilis (Grube, 1878) Introduced x x Anomia nobilis Reeve, 1859 Introduced x x Phallusia nigra Savigny, 1816 Introduced x Didemnum cf. candidum Savigny, 1816 Cryptogenic x Lutjanus kasmira (Forsska˚l, 1775) Introduced x Centropyge loricula (Gu¨nther, 1873) Cryptogenic x Cephalopholis argus Bloch & Schneider, 1801 Introduced x Total nonindigenous and cryptogenic species 5 6 1 0 1 2 Moloka‘i Acanthophora spicifera (Vahl) Børgesen Introduced x x x Cassiopea cf. andromeda Light, 1914 Introduced x Hipponix australis (Lamarck, 1819) Cryptogenic x Didemnum cf. candidum Savigny, 1816 Introduced x x x x Cephalopholis argus Bloch & Schneider, 1801 Introduced x x x Total nonindigenous and cryptogenic species 3 2 1 1 0 2 1 2 Maui Hypnea musciformis (Wulfen) J. V. Lamour Introduced x Acanthophora spicifera (Vahl) Børgesen Introduced x x x Hipponix australis (Lamarck, 1819) Cryptogenic x x x x Mycale cf. armata Thiele, 1903 Introduced x x Pennaria disticha (Goldfuss, 1820) Introduced x Sarcothelia, undescribed sp. Cryptogenic x Carijoa riisei (Duchassaing & Michelotti, 1860) Introduced x Crassostrea sp. Introduced x Didemnum cf. candidum Savigny, 1816 Introduced? x x Cephalopholis argus Bloch & Schneider, 1801 Introduced x x x x x x Lutjanus kasmira (Quoy & Gaimard, 1825) Introduced x x Lutjanus fulvus (Forster, 1801) Introduced x x Total nonindigenous and cryptogenic species 4 0 4 3 1 6 4 1 3 Hawai‘i Pennaria disticha (Goldfuss, 1820) Introduced x Plumularia ﬂoridana (Nutting, 1905) Cryptogenic x x Plumularia strictocarpa Pictect, 1893 Cryptogenic x Sabellastarte spectabilis (Grube, 1878) Introduced x Hipponix australis Lamarck, 1819 Cryptogenic x x x x x x x x Lutjanus fulvus (Forster, 1801) Introduced x Lutjanus kasmira (Forsska˚l, 1775) Introduced x x x x Cephalopholis argus Bloch & Schneider, 1801 Introduced x x x x x x x x x x Total nonindigenous and cryptogenic species 3 5 2 2 3 5 3 2 2 1 Nonindigenous Marine Species on Hawaiian Coral Reefs . Coles et al. 493

Evaluation of Rapid Assessment Technique

The numbers of total taxa, total nonindigenous and cryptogenic species, and total native species richness recorded at the seven sites where samples were collected at Kaua‘i, Mo- loka‘i, Maui, and Hawai‘i are shown in Table 4 for rapid assessments and for total taxa that were identified in the laboratory. Six of the 54 species were new reports for the Hawaiian Islands designated cryptogenic, and 20 more were not previously reported outside O‘ahu in published or unpublished literature or in Bishop Museum collections. Numbers of total taxa, native species richness, and nonindigenous and cryptogenic species were many times greater for the laboratory-identified samples than for the rapid assessment observations, as was anticipated because all minute and cryptic species could be identified in the laboratory. However, there was good agreement for each site between both methods for the ratios of nonindigenous and cryptogenic species to native species richness. Values for nonindigenous and cryptogenic species ranged from 2.4 to 9.1% for the collected samples and 1.0 to 11.1% for the rapid assessments (Table 4), with a significant correlation Figure 7. Histograms of numbers of nonindigenous and between the values from each site (Figure 8). cryptogenic species (N/CS) at all sites with fish data in- Similar agreement in results was found in a cluded and excluded. previous study of 25 stations in Ka¯ne‘ohe Bay where both on-site observations and laboratory identifications showed 19% of the Olowalu, Maui (MA2); or Kahe, O‘ahu total identified organisms to be composed of (OA4) sites. Histograms of nonindigenous nonindigenous and cryptogenic species (Coles and cryptogenic species occurrence values et al. 2002a; S.L.C., unpubl. data). (Figure 7) indicate that 21 of the 41 sites As a further determination of the sensitiv- (51%) had fewer than three nonindigenous ity and consistency of the rapid assessment and cryptogenic species. Many of these were method for detecting introductions, we calcu- sites where only introduced fishes (Lutjanus lated, for the seven stations where samples kasmira, Lutjanus fulvus (Forster), and Cepha- were collected, the ratios of numbers of spe- lopholis argus) occurred. These three species cies recorded from rapid assessments (RA) to were intentionally introduced in the 1950s numbers of taxa identified from laboratory (Brock 1960, Randall 1987) and are wide- identifications (LID) for both introduced/ spread throughout the main Hawaiian Is- cryptogenic (N/CSRA :N/CSLID) and native lands (Randall and Kanayama 1972, Maciolek species richness (NSRRA :NSRLID) data (Ta- 1984). When these three species are excluded ble 4). These ratios would be constant if from the results (Figure 7), 22 of the 41 sites nonindigenous and cryptogenic species de- (54%) had fewer than two nonindigenous and termined from rapid assessments were a cryptogenic species. consistent percentage of the total biota as TABLE 4 Introduced and Cryptogenic Species (N/CS) Identified by Rapid Assessments (RA) Versus Laboratory Identifications (LID) at Survey Stations

KA1 KA5 MO3 MO7 MA7 HA1 HA10

Taxa Species Status RA LID RA LID RA LID RA LID RA LID RA LID RA LID

Rhodophyta Acanthophora spicifera Introduced x x Hypnea musciformis Introduced x x Porifera Mycale armata Introduced x x x Mycale cecilia Introduced x x x Gelliodes fibrosa Introduced x x x Hydrozoa Pennaria disticha Introduced x x x x x x x Plumularia floridana** Cryptogenic xx Plumularia strictocarpa** Cryptogenic x Dynamena crisoides Cryptogenic x x Anthozoa Culicia, undescribed sp.** Cryptogenic x Polychaeta Platynereis abnormis Cryptogenic x Chaetopterus sp. Cryptogenic x x Branchiomma nigromaculata Cryptogenic x x Sabellastarte spectabilis Introduced x x Hydroides crucigera Introduced x Pomatoleios kraussii Introduced x Salmacina dysteri Introduced x Neodexiospira foraminosa Cryptogenic x x Neodexiospira nipponica Cryptogenic x x Simplicaria pseudomilitaris Cryptogenic Gastropoda Hipponix australis Cryptogenic x x Crepidula aculeata Introduced x x x x Crucibulum spinosum Introduced x Eualetes tulipa Introduced x x Hinemoa indica Introduced x Bivalvia Crassostrea sp. Introduced x Hiatella arctica Introduced x xxxxx x Stomatopoda Gonodactylaceus falcatus Introduced x x Amphipoda Cerapus sp.* Cryptogenic? x Ericthonius brasiliensis Introduced x x x Photis hawaiiensis Cryptogenic x Paraleucothoe flindersi Introduced x x Elasmopus rapax Introduced x xxxxx Stenothoe valida Cryptogenic x Tanaidacea Leptochelia dubia Cryptogenic x x x x x Parapseudes pedispinis** Introduced x Decapoda Discias musicus* Cryptogenic? x x Periclimenes cf. amymone* Cryptogenic? x Nikoides multispinatus* Cryptogenic? x x Glabropilumnus seminudus Introduced x Pilumnus oahuensis Introduced x x x Panopeus pacificus Introduced x Ectoprocta Caberea cf. boryi** Cryptogenic x Esharoides angela* Cryptogenic x Schizoporella cf. errata Introduced x Amathia distans Introduced x x x Ophiuroidea Ophiactis savignyi Cryptogenic x x x Ascidiacea Didemnum cf. candidum Introduced x x Ascidia sp. A Introduced x Eusynstyela hartmeyeri** Introduced x Herdmania pallida Introduced x x Osteichthyes Lutjanus kasmira Introduced x Cephalopholis argus Introduced x x x x x Total taxa 50 251 62 219 41 204 65 247 62 304 81 235 101 216 Total N/CS 5 21 6 14 1 7 1 13 4 20 3 10 1 5 Total NSR 45 230 56 205 40 197 64 234 58 284 78 225 100 211 N/CS:NSR 11.1 9.1 10.7 6.8 2.5 3.6 1.6 5.6 6.9 7.0 3.8 4.4 1.0 2.4 N/CSRA :N/CSLID .24 .43 .14 .08 .20 .30 .20 NSRRA :NSRLID .20 .27 .20 .27 .20 .35 .47 Total N/CS w/o fish 5 21 4 14 0 7 0 13 4 20 2 10 0 5 Total NSR w/o fish 25 230 29 205 11 197 26 234 47 284 38 225 32 211 N/CSRA :N/CSLIDþRA w/o fish .24 .29 — — .19 .30 — NSRRA :NSRLIDþRA w/o fish .11 .14 .06 .11 .17 .17 .15 N/CSRA site rank w/o fish 1 2 6524 6 N/CSLID site rank w/o fish 1 3 6425 7

*, New report for Hawaiian Islands; **, new report outside O‘ahu; NSR, native species richness. 496 PACIFIC SCIENCE . October 2006

Figure 8. Comparison of percentage nonindigenous and cryptogenic species (N/CS) recorded on rapid assessment (RA) surveys with percentage nonindigenous and cryptogenic species from laboratory identiﬁcations (LID) þ rapid assessments from the same sites.

determined from laboratory identifications. curred and where fishes were a dominant por- However, the scatter of these ratios (Figure tion of the total identified biota. Calculating 9a) indicates that rapid assessments detected and plotting the same ratios for the data with an inconsistent percentage of the biota to be fish excluded reduced the number of sites to nonindigenous and cryptogenic species. Fur- the four where only nonfish nonindigenous ther, most of the points fall below the line and cryptogenic species occurred (Table 4). representing a 1:1 ratio, suggesting that rapid These show a more consistent pattern (Fig- assessments tended to underestimate nonin- ure 9b), with ratios ranging from 0.11 to 0.30 digenous and cryptogenic species relative to and all points plotting above a line repre- native species richness. senting equivalent ratios for nonindigenous Inspection of the data suggested that much and cryptogenic speciesRA :nonindigenous and of this inconsistency was due to inclusion of cryptogenic speciesLID and native species rich- fish species, because the introduced fishes nessRA :native species richnessLID. This sug- usually occurred at open-ocean reef sites gests that, in contrast to the pattern shown where few to no other introduced species oc- in Figure 9a, the rapid assessment method Nonindigenous Marine Species on Hawaiian Coral Reefs . Coles et al. 497

Figure 9. Ratios of native species richness (NSR) recorded on rapid assessment (RA) surveys and from laboratory identiﬁcations (LID) þ rapid assessments from the same sites (NSRRA :NSRLIDþRA ) plotted against nonindigenous and cryptogenic species (N/CSRA :N/CSLIDþRA ) for a, all sample collection sites; b, same data with ﬁsh excluded.

Figure 10. Regressions of nonindigenous and cryptogenic species (N/CS) on native species richness (NSR) for taxa identified by rapid assessment (black squares, solid lines) and by laboratory identification þ rapid assessment (gray circles, dashed lines) for a, all sample collection sites; b, same data with fish excluded. with fish excluded detected slightly higher full data set (Figure 10a) shows a positive re- percentages of nonindigenous and crypto- lationship for the laboratory identification re- genic species relative to native species than sults but a negative relationship for the rapid did laboratory identifications. assessment results, suggesting that rapid as- A second analysis showed a similar incon- sessments underestimated nonindigenous and sistency between rapid assessment and labora- cryptogenic species at the most species-rich tory identification results that is reduced sites. Excluding fish data from both nonindig- when the fish data are excluded. A plot of enous and cryptogenic species and native numbers of nonindigenous and cryptogenic species richness results rectified some of this species against native species richness for the inconsistency (Figure 10b). The resulting 498 PACIFIC SCIENCE . October 2006 regression for rapid assessment results is pos- to promote nonindigenous and cryptogenic itive and more comparable with the labora- species for these indexes). Over half of the tory identification results. Finally, both rapid values were in categories 3 to 5 for the ur- assessments and the laboratory identifications banization index (more than six houses or showed similar rank orders of the stations for buildings visible on the shoreline). number of nonfish nonindigenous and cryp- Values for ranked variables; distances from togenic species reports (Table 4), with Sta- harbor piers, boat-launching areas, shore- tions KA1 and MA7 ranking highest and lines, and streams; and numbers of native Stations MO3 and HA10 ranking lowest for taxa excluding fishes (Table 5) were used as both rapid assessments and laboratory identi- potential predictor variables in best subsets fications. regression analysis (Minitab Release 14.1, Minitab, State College, Pennsylvania) to identify the best-fitting regression model ex- Variables Influencing Introductions plaining numbers of nonindigenous and cryptogenic species. The initial model with Given the above results, fishes were excluded the most explanatory power (adjusted from the following analysis. Also, because the r 2 ¼ 0:52) included the three predictor vari- third field team member did not make rapid ables ocean isolation, ramp distance, and na- assessment observations at all of the sites, tive species richness in order of variance the rapid assessment data from that observer explained. These three predictor variables were not included in the analysis of relation- and all of their possible two-way interactions ships between numbers of nonindigenous and were then used in a best subsets regression cryptogenic species and environmental vari- analysis to identify the most predictive and ables (Table 2) or native species richness to logical model (i.e., if an interaction term was compare data based on consistent effort included in the model, the variables used to throughout the study. These exclusions re- calculate the interaction term were also in- sulted in the total number of taxa used in re- cluded). This best regression model ac- gression analysis being reduced to 208 from counted for 67.2% of the variance explained the 486 reported by all three observers. How- and included as significant factors ocean iso- ever, algae and invertebrate nonindigenous lation, native species richness, and their inter- and cryptogenic species (Table 3) were re- action (Table 6). Their relationship with duced by only three species: cryptogenic numbers of nonindigenous and cryptogenic Hipponix australis that occurred at 20 sites, species is illustrated graphically in Figure 11. and introduced polychaete Salmacina dysteri Most (45%) of the variance in nonindige- (Huxley) and introduced bryozoan Schizopor- nous and cryptogenic species was explained ella cf. errata (Waters), each of which oc- by a highly significant positive relationship curred at only one site. (P <:001) between numbers of nonindige- The sites surveyed encompassed a wide nous and cryptogenic species and isolation range of environmental conditions and prox- from open-ocean conditions (i.e., nonindig- imity to anthropogenic influences on each enous and cryptogenic species increased island (Table 5). The stations ranged from significantly going from open coastlines to 100 m to 4.6 km from the shoreline and semienclosed locations). This effect was from 300 m to 26 km from the nearest stream modified by a highly significant (P <:001) mouth. Distances from harbors ranged from negative isolation–native species richness 300 m to 71 km and those from boat- interaction term where nonindigenous and launching ramps from 200 m to 27 km. Most cryptogenic species decreased as native spe- of the sites (18–25) were in the lowest index cies richness and isolation increased. category (1) for artificial substrata, reef con- Examining numbers of nonindigenous and dition, and ocean restriction (i.e., most sites cryptogenic species and native species rich- were in the category considered least likely ness in the four categories of isolation index Nonindigenous Marine Species on Hawaiian Coral Reefs . Coles et al. 499

TABLE 5 Matrix of Predictor Variables Used in Best Subsets Regression Analysis at 41 Reef Sites

Harbor Ramp Shore Stream Artiﬁcial Reef Urban- N/CS Distance Distance Distance Distance Substrate Health ization Isolation (w/o Island Station (km) (km) (km) (km) Index Index Index Index NSR ﬁsh)

Kaua‘i 1 1.2 1.5 0.3 0.3 4 3 3 3 45 5 2 12.7 1.7 0.1 1.1 2 4 4 2 32 0 3 13.1 2.1 0.2 0.5 1 2 3 1 77 2 4 13.4 1.9 0.2 0.7 1 2 3 1 56 0 5 0.6 0.7 0.4 0.7 4 5 2 4 58 3 6 8.5 8.6 0.6 0.7 2 5 2 1 59 0 7 3.9 4.0 0.1 3.1 1 3 1 1 65 0 8 11.9 0.6 0.1 2.0 3 3 4 1 74 0 O‘ahu 1 4.4 4.5 1.1 1.4 1 3 4 3 46 5 2 0.3 0.5 0.3 0.7 2 4 3 3 56 5 3 6.2 6.1 0.1 0.3 1 4 3 1 81 1 4 3.4 3.3 0.1 3.2 3 3 3 1 87 0 5 7.5 7.6 0.1 0.6 1 2 3 1 93 1 6 5.4 5.6 0.1 5.8 1 2 2 3 105 0 Moloka‘i 1 0.3 0.3 0.1 0.4 3 4 3 2 37 2 2 8.4 8.4 1.4 8.3 1 1 2 1 60 0 3 0.6 22.7 0.5 23.1 3 1 3 1 40 0 4 1.2 26.9 1.0 1.3 3 3 3 1 54 1 5 12.0 15.3 1.5 1.7 2 1 4 1 33 0 6 3.6 5.3 1.6 5.3 3 1 4 1 43 0 7 1.1 1.5 1.4 1.5 3 1 3 1 64 0 8 4.1 4.3 0.6 4.6 3 4 4 1 56 1 Maui 1 32.3 6.0 0.1 1.6 1 2 4 2 71 2 2 12.4 12.5 0.2 1.9 2 3 2 2 52 0 3 6.2 6.3 1.6 2.7 1 1 1 1 75 2 4 27.4 17.7 0.1 0.4 1 1 1 3 96 0 5 21.1 5.3 0.1 0.5 1 3 2 1 46 0 6 25.7 0.2 0.2 0.3 5 2 3 4 73 4 7 0.3 0.4 0.3 10.4 3 4 4 3 58 4 8 20.6 20.8 0.1 26.0 1 1 2 2 87 0 9 17.7 17.9 4.6 23.0 1 1 1 1 83 0 Hawai‘i 1 1.8 1.5 0.4 1.8 4 1 2 1 78 2 2 8.2 7.9 0.2 5.4 2 1 3 1 89 1 3 10.9 10.6 0.3 8.0 1 1 2 1 85 0 4 18.8 18.5 0.5 16.4 1 1 2 1 58 0 5 68.5 16.2 0.2 9.2 1 1 1 1 75 0 6 71.0 18.9 0.1 12.0 1 1 2 1 106 0 7 64.3 14.1 0.2 6.8 1 1 4 1 83 0 8 57.4 7.4 0.2 1.2 2 1 5 2 74 0 9 53.5 3.9 0.1 4.7 1 1 1 1 111 0 10 8.8 8.9 0.1 8.0 1 2 3 1 100 0

NSR, native species richness; N/CS, nonindigenous and cryptogenic species.

(Figure 12) showed no significant relation- discussion ship in open-ocean or semiexposed coastlines, but nonindigenous and cryptogenic species Our results agree with previous findings in decreased significantly with native species the Hawaiian Islands, Johnston Atoll, Amer- richness in the combined embayment and ican Samoa, and Guam indicating that rela- semienclosed harbor environments. tively few introduced marine invertebrates 500 PACIFIC SCIENCE . October 2006

TABLE 6 Relationships Determined between Numbers of Nonindigenous and Cryptogenic Species (N/CS) and Predictor Variables by Best Subsets Linear Regression Analysis

Potential Determining Expected % Variance Factor Type Range N/CS Effect Result Explained

Ocean isolation Estimated 1–5 þþP < 0.001 45.3 Species richness (NSR) Measured 42–138 þP < 0.030 8.0 Isolation–NSR interaction P < 0.001 13.9 Full model þP < 0.001 67.2

Figure 11. Three-dimensional response surface illustrating relationships between numbers of nonindigenous and cryptogenic species (N/CS), ocean exposure of the sites surveyed, and native species richness (NSR, ﬁsh excluded). N/CS ¼ 3.81*Isolation þ 0.049*NSR 0.040*Interaction 4.19. Solid dots indicate values above the response surface; open circles indicate values below. occur on open-ocean coral reefs, especially introduced marine species found on Guam compared with enclosed harbors and dis- occurred outside Apra Harbor. Coles et al. turbed embayments (Table 7). Paulay et al. (2003) found only six nonindigenous and (2002) determined that only 23% of the 85 cryptogenic species at coral reef sites on the Nonindigenous Marine Species on Hawaiian Coral Reefs . Coles et al. 501

lands, with a maximum of six nonindigenous and cryptogenic species at any one site. In- cluding the organisms identified from collections at seven reef sites nearest harbors, the total increased to 54 nonindigenous and cryptogenic species and a maximum of 23 at any one site. In contrast to the usual circum- stances in Hawaiian harbors, where invertebrate nonindigenous and cryptogenic species are often among the most common or dominant species, this was not the case at any reef site. The alga Acanthophora spicifera (Vahl) Boerg was abundant on Moloka‘i at Pu¯ko‘o (MO1) and at the Hotel Moloka‘i (MO8) site; on Maui at Kahekili Beach Park (MA1), Papa‘ula Point (MA3), and Ma¯‘alaea Reef (MA7); and on O‘ahu at the Na¯na¯kuli Point site (OA3). Another alga, Kappaphycus cf. alvarezii (Doty) Doty, was abundant at the two Ka¯ne‘ohe Bay sites (OA1 and OA2), and the ta‘ape, Lutjanus kasmira, was abundant at 10 sites throughout the survey. The only invertebrate nonindigenous and cryptogenic species abundant at any site was the octocoral Carijoa riisei (Duchassaing & Michelotti), which occurred on the undersides of concrete posts and metal sheet pilings of a collapsed dock adjacent to the coral reef at Ma¯la Wharf, Maui (MA6). Remarkably, this species was absent from KA5 on Kaua‘i, within the breakwater of Port Allen Harbor and only 600 m from where the octocoral was very abundant on the harbor’s main dock Figure 12. Comparisons of numbers of nonindigenous pier pilings (Coles et al. 2004b). All other in- and cryptogenic species (N/CS) with native species rich- vertebrate nonindigenous and cryptogenic ness (NSR) at different isolation index sites: a, open species were rare at all sites and usually con- ocean; b, semiexposed coastlines; c, embayments (closed sisted of a single or few individuals or colo- circles) and semienclosed harbors (open circles). nies in cryptic locations. Rapid assessments detected one to six introduced or cryptogenic species at the seven island of Tutuila, American Samoa, com- stations where collections and laboratory pared with 28 nonindigenous and cryptogenic identifications yielded 5–21 species. By sta- species within Pago Pago Harbor. Similarly, tion, rapid assessments detected 8–43% of in contrast to the approximately 100 nonin- the numbers of micro- and macrobiota non- digenous and cryptogenic species previously indigenous and cryptogenic species deter- recorded from O‘ahu harbors and in mined from laboratory identification, with a Ka¯ne‘ohe Bay (Coles et al. 1999a,b, 2002a) mean for the seven stations of 22.7%. Con- our rapid assessments detected a total of 26 versely, one to four nonindigenous and cryp- nonindigenous and cryptogenic species (in- togenic species (two invertebrates and two cluding three algae and four fishes) at 41 fishes) detected by rapid assessment were not reef sites throughout the main Hawaiian Is- reported by laboratory identifications. Rapid 502 PACIFIC SCIENCE . October 2006

TABLE 7 Numbers of Marine Nonindigenous (N), Cryptogenic (C), and Total Taxa Determined from the Hawaiian Islands, Johnston Atoll, Guam, and American Samoa

Total Total % Location N C N/CSa Taxa N/CS Source

Harbors and embayments O‘ahu, Pearl Harbor 69 26 95 419 23.0 Coles et al. 1997, 1999b O‘ahu, commercial and small boat harbors 73 27 100 585 17.0 Coles et al. 1999a Ka¯ne‘ohe Bay 82 34 116 617 14.5 Coles et al. 2002a Hawaiian Neighbor Island harbors 72 32 104 694 14.9 Coles et al. 2004b Guam, Apra Harbor 27 29 46 682 6.7 Paulay et al. 2002 Pago Pago Harbor, American Samoa 17 11 28 977 2.9 Coles et al. 2003 Coral reefs Waikı¯kı¯ 19 33 52 617 6.9 Coles et al. 2002b Kaho‘olawe 3 0 3 298 1.0 Coles et al. 1998 Midway Atoll 4 0 4 444 1.5 DeFelice et al. 1998 Johnston Atoll 5 5 10 668 1.5 Coles et al. 2001 French Frigate Shoals 2 0 2 617 0.3 DeFelice et al. 2002 Guam 41 44 85 2878 1.5 Paulay et al. 2002 Tutuila, American Samoa 2 5 7 828 0.8 Coles et al. 2003 Hawaiian Neighbor Islands, near harbors 42 23 65 814 7.9 Coles et al. 2004b Hawaiian Neighbor Islands, 41 reef sites 18 8 26 486 5.3 This study

a N/CS, nonindigenous and cryptogenic species.

assessments were therefore clearly less sensi- bors or boat ramps had some of the lowest tive for determining a full complement of values for nonindigenous and cryptogenic introduced species that includes minute or- species. This indicates that if these potential ganisms and species in cryptic habitats. How- sources are important in the spread and pro- ever, rapid assessments were more time and liferation of nonindigenous and cryptogenic resource efficient, requiring about 1 hr in the species, their influence was not spatially con- field and about 1 hr for data entry per site. By sistent, nor did their influence extend very comparison, although the time required for far. Many sites within a few hundred meters sample treatment, sorting, species identifica- from harbors and docks showed few or no tion, and verification by taxonomic experts nonindigenous and cryptogenic species, and varied substantially with organism abundance none exceeded a total of six species detected and species richness, about 1 month per sta- by rapid assessment or 23 by laboratory iden- tion would be a conservative estimate, based tification. No other factors considered likely upon this and previous studies on O‘ahu. to propagate or support nonindigenous and Some sites near harbors or boat ramps cryptogenic species (i.e., proximity to shore- were among those with the maximum num- lines or streams mouths, presence of human- bers of nonindigenous and cryptogenic spe- made structures in the water, development or cies, suggesting that harbors and boat ramps alteration of the shoreline, and apparent de- could act as sources or vectors for the spread clines in reef condition that would suggest of introduced species. This would be the environmental disturbance) showed any sig- expected pattern if boat transport of nonin- nificant relationship with numbers of nonindigenous and cryptogenic species were the digenous and cryptogenic species detected by primary factor in determining their recruit- rapid assessment. ment and final distribution. However, this The most clearly defined relationship be- pattern did not hold for the full array of rapid tween nonindigenous and cryptogenic species assessment sites, and some stations near har- determined from rapid assessment and any Nonindigenous Marine Species on Hawaiian Coral Reefs . Coles et al. 503 predictor variable was the degree to which species richness. Rather, a highly significant the sites were isolated from the open ocean. interaction term between ocean isolation and This factor alone accounted for 45% of the native species richness suggests that the asso- variance in nonindigenous and cryptogenic ciation between the number of nonindige- species numbers. This finding agrees with nous species and native species richness may observations in harbors and embayments be related to the degree of isolation from on O‘ahu (Coles et al. 1997, 1998, 1999a,b, open-ocean exposure, with nonindigenous 2002a,b), which found decreasing numbers and cryptogenic species decreasing as native of nonindigenous and cryptogenic species in species richness increases in semiexposed en- the more ocean-exposed areas of the study vironments and embayments. areas. Increasing ocean exposure also corre- In conclusion, this and previous surveys of lated with increasing native species richness. nonindigenous species in Hawai‘i indicate These relationships were also apparent on that, overall, few introduced marine inverte- Tutuila, American Samoa (Coles et al. 2003), brates have colonized Hawai‘i’s coral reefs where nonindigenous and cryptogenic species and even fewer are invasive (i.e., monopoliz- ranged from 5 to 17 per site in the semien- ing habitats or displacing native species on closed inner Pago Pago Harbor to 1 to 4 reefs). Our results also suggest that the com- nonindigenous and cryptogenic species per petition between native and nonindigenous site on reefs distant from the harbor. By com- species may be influenced by the degree to parison, total taxa there were 102–185 per which the reef or habitat is isolated from site at the inner harbor sites and 403–449 open-ocean circulation, in interaction with per site on reefs outside the harbor. the species richness of the native biota. Previous studies in Hawai‘i (Coles et al. 1999a, 2002a, Coles and Eldredge 2002) acknowledgments have concluded that coral reef systems may be resistant to species introductions and dis- We thank the Hawai‘i Coral Reef Initiative ruptions of native populations. Paulay et al. director and staff for their guidance and assis- (2002) concluded that the results of their tance. Victor Bonito provided valuable assis- study of introduced marine species in Guam tance in the field for the Moloka‘i survey. supported the hypothesis articulated by Ver- The Hawai‘i State Department of Land and meij (1991) that diverse communities are Natural Resources, Division of Aquatic Re- more difficult to invade. Increased species sources generously provided logistic support richness of sessile organisms has been empir- and boat transport on the islands of Maui, ically determined to significantly decrease Hawai‘i, and O‘ahu, and we want to especially invasion success in coastal New England hab- thank Paul Murakawa, Tony Montgomery, itats (Stachowicz et al. 1999). Some studies in Skippy Hau, Russell Sparks, and Bill Walsh Australia also suggest that higher diversity of for providing this service. We are also grate- native biota in reef systems offer fewer op- ful to the Hawai‘i Institute of Marine Biology portunities for successful proliferation of new and the University of Hawai‘i Zoology De- arrivals than is the case for lower-diversity partment for providing logistic support and temperate areas (Hutchings et al. 2002). boat usage to access the Ka¯ne‘ohe Bay sites. However, Hewitt (2002) did not find a signif- Comments by three anonymous reviewers icant negative relationship between native were very helpful in improving the paper’s species richness and introduced species for content and clarity. Specimens of taxa that eight port surveys in tropical to temperate were possible new introductions or that re- waters around Australia, although numbers quired species identification verification were of introduced species did increase signifi- sent to or seen by the following experts, cantly with latitude. who are gratefully acknowledged for their Similarly, in this survey we did not find an assistance—sponges: Ralph DeFelice, Los anticipated decrease of nonindigenous and Angeles County Museum of Natural History; cryptogenic species with increasing native hydrozoans: Dale Calder, Royal Ontario Mu- 504 PACIFIC SCIENCE . October 2006 seum, Toronto, Canada; vermetid gastropod Carlton, J. T., J. K. Thompson, L. E. Sche- mollusks: Anuschka Fauci, Department of mel, and F. H. Nichols. 1990. Remarkable Zoology, University of Hawai‘i; bryozoans: invasion of San Francisco Bay (California, Chela Zabin, Department of Zoology, Uni- USA) by the Asian clam Potamocorbula versity of Hawai‘i; ascidians: Scott Godwin, amurensis. I. Introduction and dispersal. Department of Natural Sciences, Bishop Mu- Mar. Ecol. Prog. Ser. 66:81–94. seum, Honolulu, Hawai‘i. Chapman, J. W., and J. T. Carlton. 1991. A test of criteria for introduced species: The Literature Cited global invasion by the isopod Synidotea lae- vidorsalis (Miers, 1881). J. Crustacean Biol. Abbott, D. P., A. T. Newberry, and K. M. 11 (3): 386–400. Morris. 1997. Section 6B: Ascidians. Cohen, A. N., J. T. Carlton, and M. C. Foun- Bishop Mus. Spec. Publ. 64 (6B). tain. 1995. Introduction, dispersal and po- Bax, N., J. T. Carlton, A. Mathews-Amos, tential impacts of the green crab Carcinus R. L. Haedrich, F. G. Howarth, J. E. maenas in San Francisco Bay, California. Purcell, J. E. Rieser, and A. Gray. 2001. Mar. Biol. (Berl.) 122:225–237. Conserving marine diversity through the Coles, S. L., and H. Bolick. 2006. Assessment control of biological invasions. Conserv. of invasiveness of the orange keyhole Biol. 451:145–176. sponge Mycale armata in Ka¯ne‘ohe Bay, Birkeland, C. 2004. Ratcheting down coral O‘ahu, Hawai‘i. Final Report Year 1. Con- reefs. BioScience 54:1021–1027. tribution No. 2006-02 to the Hawai‘i Brock, V. E. 1960. The introduction of Biological Survey, Bishop Museum, Ho- aquatic animals into Hawaiian waters. Int. nolulu. http://hbs.bishopmuseum.org/pdf/ Rev. Ges. Hydrobiol. 45:463–480. HCRI-report2006.pdf. Brown, E., E. Cox, P. Jokiel, K. Rodgers, W. Coles, S. L., R. C. DeFelice, and L. G. Smith, B. Tissot, S. L. Coles, and J. Hult- Eldredge. 1999a. Nonindigenous marine quist. 2004. Development of benthic species introductions in the harbors of the sampling methods for the Coral Reef south and west shores of O‘ahu, Hawai‘i. Assessment and Monitoring Program Bishop Mus. Tech. Rep. 15. http://hbs (CRAMP) in Hawai‘i. Pac. Sci. 58:145– .bishopmuseum.org/pdf/southshore.pdf. 158. ———. 2002a. Nonindigenous species in Carlton, J. T. 1985. Transoceanic and in- Ka¯n‘eohe Bay, O‘ahu, Hawai‘i. Bishop teroceanic dispersal of coastal marine or- Mus. Tech. Rep. 24. http://hbs ganisms: The biology of ballast water. .bishopmuseum.org/pdf/kbay-report.pdf. Oceanogr. Mar. Biol. Annu. Rev. 23:313– ———. 2002b. Nonindigenous marine spe- 371. cies introductions at Waikı¯kı¯ and Hawai‘i ———. 1994. Biological invasions and biodi- Kai, O‘ahu, Hawai‘i. Bishop Museum versity in the sea: The ecological and Tech. Rep. 25. http://hbs.bishopmuseum human impacts of nonindigenous marine .org/pdf/waikiki.pdf. and estuarine organisms. Pages 5–11 in Coles, S. L., R. C. DeFelice, L. G. Eldredge, Nonindigenous estuarine and marine or- and J. T. Carlton. 1997. Biodiversity of ganisms (NEMO), Proceedings of the marine communities in Pearl Harbor, Conference and Workshop. Government O‘ahu, Hawai‘i with observations on Printing Office No. 0208-C-04, Seattle, introduced species. Bishop Mus. Tech. Washington. Rep. 10. http://hbs.bishopmuseum.org/ ———. 1996. Biological invasions and cryp- pdf/PHReport.pdf. togenic species. Ecology 77:1653–1655. ———. 1999b. Historical and recent intro- Carlton, J. T., and J. B. Geller. 1993. Ecolog- ductions to non-indigenous marine species ical roulette: The global transport of into Pearl Harbor, Oahu, Hawaiian Is- nonindigenous marine organisms. Science lands. Mar. Biol. (Berl.) 135:147–158. (Washington, D.C.) 261:78–82. Coles, S. L., R. C. DeFelice, and D. Minton. Nonindigenous Marine Species on Hawaiian Coral Reefs . Coles et al. 505

2001. Marine species survey of Johnston vertebrates from French Frigate Shoals, Atoll June 2000. U.S. Fish and Wildlife Northwestern Hawaiian Islands with a Service, Pacific Islands Area Office. Bishop note on nonindigenous species. Bishop Mus. Tech. Rep. 19. http://hbs.bishopmuseum Mus. Tech. Rep. 23. .org/pdf/johnstonreport.pdf. Eldredge, L. G., and J. T. Carlton. 2002. Ha- Coles, S. L., R. C. DeFelice, J. E. Smith, D. waiian marine bioinvasions: A preliminary Muir, and L. G. Eldredge. 1998. Determi- assessment. Pac. Sci. 56:211–212. nation of baseline conditions for intro- Englund, R. A., K. Arakaki, D. J. Preston, duced marine species in nearshore waters S. L. Coles, and L. G. Eldredge. 2000. of the island of Kaho‘olawe, Hawai‘i. Nonindigenous freshwater and estuarine Bishop Mus. Tech. Rep. 14. http://hbs species introductions and their potential .bishopmuseum.org/pdf/kahoolawe.pdf. to affect sportfishing in the lower stream Coles, S. L., and L. G. Eldredge. 2002. Non- and estuarine regions of the south and indigenous species introductions on coral west shores of O‘ahu, Hawai‘i. Bishop reefs: A need for information. Pac. Sci. Mus. Tech. Rep. 17. 56:191–209. Gosliner, T. M. 1995. Introduction and Coles, S. L., L. G. Eldredge, F. Kandel, P. A. spread of Philine auriformis (Gastropoda: Reath, and K. Longenecker. 2004a. As- Opisthobranchia) from New Zealand to sessment of nonindigenous species on San Francisco Bay and Bodega Harbor. coral reefs in the Hawaiian Islands, with Mar. Biol. (Berl.) 122:249–255. emphasis on introduced invertebrates. Greenberg, N., R. L. Garthwaite, and D. C. Bishop Mus. Tech. Rep. 27. http://hbs Potts. 1996. Allozyme and morphological .bishopmuseum.org/pdf/hcri-2004.pdf. evidence for a newly introduced species of Coles, S. L., P. A. Reath, K. Longenecker, H. Aurelia in San Francisco Bay, California. Bollick, and L. G. Eldredge. 2004b. Assess- Mar. Biol. (Berl.) 125:401–410. ment of nonindigenous marine species in Grigg, R. W. 2003. Invasion of a deep black harbors and on nearby coral reefs on coral bed by an alien species, Carijoa riisei, Kaua‘i, Moloka‘i, Maui and Hawai‘i. off Maui, Hawaii. Coral Reefs 22:121– Bishop Mus. Tech. Rep. 29. http://hbs 122. .bishopmuseum.org/pdf/tr29.pdf. ———. 2004. Harvesting impacts and inva- Coles, S. L., P. A. Reath, P. A. Skelton, V. Bo- sion by an alien species decrease estimates nito, R. C. DeFelice, and L. Basch. 2003. of black coral yield off Maui, Hawai‘i. Pac. Introduced marine species in Pago Pago Sci. 58:1–6. Harbor, Fagatele Bay and the National Hewitt, C. L. 2002. Distribution and diversity Park Coast, American Samoa. Bishop Mus. of Australian tropical marine bioinvasions. Tech. Rep. 26. http://www.bishopmuseum Pac. Sci. 56:213–222. .org/research/pbs/pdf/pagopago.pdf. Hewitt, C. L., M. L. Campbell, K. M. Moore, Daehler, C. C., and D. R. Strong. 1996. N. B. Murfet, and B. Robertson. 1998. Status, prediction and prevention of intro- Introduced species survey. Port of Hay duced cordgrass Spartina spp. invasions Point, Queensland. CSIRO Centre for in Pacific estuaries, USA. Biol. Conserv. Research on Introduced Marine Pests, Re- 78:51–58. port for the Ports Corporation of Queens- DeFelice, R. C., S. L. Coles, D. Muir, and land, Brisbane. L. G. Eldredge. 1998. Investigation of Hoedt, F. E., J. H. Choat, J. Collins, and J. J. the marine communities of Midway Har- Cruz. 2000. Mourilyan Harbour and Ab- bor and adjacent lagoon, Midway Atoll, bot Point surveys: Port marine baseline Northwestern Hawaiian Islands. Bishop surveys and surveys for introduced marine Museum, Hawai‘i Biological Survey Contr. pests. School of Marine Sciences and No. 1998-014. Aquaculture, James Cook University, Re- DeFelice, R., D. Minton, and S. Godwin. port for the Ports Corporation of Queens- 2002. Records of shallow-water marine in- land, Brisbane. 506 PACIFIC SCIENCE . October 2006

———. 2001a. Port of Lucinda surveys: Port search Centre Report for the Ports Corpo- marine baseline surveys and surveys for in- ration of Queensland, Brisbane. troduced marine pests. School of Marine Paulay, G., L. Kirkendale, G. Lambert, and Sciences and Aquaculture, James Cook C. Meyer. 2002. Anthropogenic biotic in- University, Report for the Ports Corpora- terchange in a coral reef ecosystem: A case tion of Queensland, Brisbane. study from Guam. Pac. Sci. 56:403–422. Hoedt, F. E., J. H. Choat, J. J. Cruz, K. M. Randall, J. E. 1987. Introductions of marine Neil, and J. Collins. 2001b. Survey of the fishes to the Hawaiian Islands. Bull. Mar. Port of Weipa: Baseline surveys and sur- Sci. 41:490–502. veys for introduced marine pests. School Randall, J. E., and R. K. Kanayama. 1972. of Marine Sciences and Aquaculture, Hawaiian fish immigrants. Sea Front. James Cook University, Report for the 18:144–153. Ports Corporation of Queensland, Bris- Rodgers, S. K., and E. F. Cox. 1999. Rate of bane. spread of introduced rhodophytes Kappa- Hutchings, P. A., R. W. Hilliard, and S. L. phycus alvarezii, Kappaphycus striatum, and Coles. 2002. Species introductions and Gracilaria salicornia and their current dis- potential for marine pest invasions into tributions in Ka¯ne‘ohe Bay, O‘ahu, Ha- tropical marine communities, with special wai‘i. Pac. Sci. 53:232–241. reference to the Indo-Pacific. Pac. Sci. Ruiz, G. M., J. T. Carlton, E. D. Grosholz, 56:223–233. and A. H. Hines. 1997. Global invasions Jokiel, P. L. 2002. Hawai‘i Coral Reef of marine and estuarine habitats by non- Assessment and Monitoring Program indigenous species: Mechanisms, extent (CRAMP). Hawai‘i Coral Reef Initiative and consequences. Am. Zool. 31:621–632. final report. http://www.hawaii.edu/ssri/ Ruiz, G. M., P. W. Fofonoff, M. J. Wonham, hcri/files/jokiel_final_report_00-01.pdf. and A. H. Hines. 2000. Invasion of coastal Kinzie, R. A., III. 1968. The ecology of the marine communities in North America: replacement of Pseudosquilla ciliata by Gon- Apparent patterns, processes, and biases. odactylus falcatus (Crustacea: Stomatopoda) Annu. Rev. Ecol. Syst. 31:481–531. recently introduced into the Hawaiian Is- Russell, B. C., and C. L. Hewitt. 2000. Base- lands. Pac. Sci. 22:465–474. line survey of the port of Darwin for ———. 1984. Aloha also means goodbye: A introduced marine species. Northern cryptogenic stomatopod in Hawaii. Pac. Territory Department of Transport and Sci. 38:298–311. Works Marine Branch, Darwin, Northern Maciolek, J. A. 1984. Exotic fishes in Hawaii Territory, Australia. and other islands in Oceania. Pages 131– Smith, J. E., C. L. Hunter, and C. M. Smith. 161 in R. Courtenay Jr. and J. R. Stauffer 2002. Distribution and reproductive char- Jr., eds. Distribution, biology and manage- acteristics of nonindigenous and invasive ment of exotic fishes. Johns Hopkins Press, marine algae in the Hawaiian Islands. Pac. Baltimore. Sci. 56:299–315. Mills, C. E., and F. Sommer. 1995. Inverte- Stachowicz, J. J., R. B. Whitlatch, and R. W. brate introductions in marine habitats: Osman. 1999. Species diversity and inva- Two species of hydromedusae (Cnidaria) sion resistance in a marine ecosystem. native to the Black Sea, Maeotias inexspec- Science (Washington, D.C.) 286:1577– tata and Blackfordia virginica, invade San 1579. Francisco Bay. Mar. Biol. (Berl.) 122:279– Tissot, B. N., W. A. Walsh, and L. E. Hal- 288. lacher. 2001. The effectiveness of marine Neil, K. M., F. E. Hoedt, J. Sheaves, J. J. protected areas and the impacts of Cruz, and J. H. Choat. 2001. Survey of aquarium fish collecting in Hawai‘i. Ha- the Port of Karumba: Port marine surveys wai‘i Coral Reef Initiative final report and surveys for introduced marine pests. year 2001. http://www.hawaii.edu/ssri/ James Cook University, CRC Reef Re- hcri/files/tissot_final_report_00-01.pdf. Nonindigenous Marine Species on Hawaiian Coral Reefs . Coles et al. 507

Vermeij, G. J. 1991. When biotas meet: Un- vasive Kappaphycus striatum in Ka¯ne‘ohe derstanding biotic interchange. Science Bay, a tropical reef. Pages 186–191 in J. (Washington, D.C.) 253:1099–1104. Pederson, ed. Marine bioinvasions. MIT Woo, M., C. M. Smith, and W. Smith. 1999. Sea Grant Program, Cambridge, Massa- Ecological interactions and impacts of in- chusetts.