Bull Mar Sci. 91(4):479–487. 2015 coral reef paper http://dx.doi.org/10.5343/bms.2015.1023 Micropredation by gnathiid isopods on settlement- stage reef fish in the eastern Caribbean Sea 1 Department of Biological John M Artim 1 Sciences and Environmental 2 Sciences Program, Arkansas Joseph C Sellers State University, State Paul C Sikkel 1, 3 * University, Arkansas 72467. 2 Center for Marine and Environmental Studies, University of the Virgin Islands, ABSTRACT.—The transition from a planktonic larval 2 Brewers Bay, St. Thomas, USVI stage to a benthic or demersal juvenile stage is a crucial event 00802. in the life history of coral reef fishes, and recruitment success has a strong influence on reef-fish population size. Post- 3 Water Research Group, Unit for Environmental Sciences settlement predation is thought to limit recruitment success. and Management, North-West Most studies on post-settlement predation have focused on University, Potchefstroom 2520, piscivorous reef fishes. However, recent studies in the tropical South Africa. Pacific Ocean suggest that blood-feeding ectoparasites may * Corresponding author email: also be an important source of predation. Here we provide <[email protected]>, phone: further evidence that a gnathiid ectoparasite is capable of 870-972-3296, fax: 870-972-2638. such “micropredation” on settlement-stage fish in the eastern Caribbean Sea. On 12 occasions in May–August, 2014 and 2015, during plankton light-trap recovery, we observed post- settlement fishes from the families Apogonidae, Bothidae, Gobiidae, Labrisomidae, and Tripterygiidae with attached ectoparasitic juvenile gnathiid isopods. In one laboratory experiment in which 10 settlement-stage French grunt, Haemulon flavolineatum (Desmarest, 1823), were fed upon by an average of one gnathiid per fish, all fish died compared to none in the control group. In a subsequent experiment in which individual settlement-stage French grunt were exposed to 0–3 gnathiids each, all 50 fish that were fed upon by at least one gnathiid died, compared with only one of 78 that were not fed on by any gnathiids. Combined with observations on damselfishes from the Pacific Ocean, these findings suggest that gnathiids are capable of micropredation on larval fish at time of reef settlement and that micropredation should be Date Submitted: 8 April, 2015. Date Accepted: 23 September, 2015. further evaluated as a factor affecting success of settlement Available Online: 19 October, 2015. and thus recruitment success of coral reef fishes. With few exceptions, the early life history of coral reef fishes consists of a pelagic larval stage, followed by a benthic or demersal juvenile stage. The transition between habitats and life history stages is associated with significant changes in morphology and exposure to reef-based predators. Predation on early post-settlement stages can have significant effects on recruitment to and therefore the density and dynamics of reef-fish populations (e.g., Carr and Hixon 1995, Forrester and Steele 2004, reviewed by Hixon 2015). The source of predation has generally been considered to be small piscivorous fishes. However, recent studies in the tropical Pacific Ocean suggest that Bulletin of Marine Science 479 OA © 2015 Rosenstiel School of Marine & Atmospheric Science of the University of Miami Open access content 480 Bulletin of Marine Science. Vol 91, No 4. 2015 ectoparasites may sometimes function as “micropredators” (Raffel et al. 2008) and thus contribute to mortality of recently settled reef fishes (Grutter et al. 2008, Jones and Grutter 2008, Penfold et al. 2008). Gnathiid isopods are small (1–3 mm) protelean parasites that feed on fish blood during each of three larval stages and live in the substratum between feedings. Thus, they are ecologically similar to terrestrial haematophagous arthropods, such as ticks (Ixodoidea) and fleas (Siphonoptera). After the third feeding, gnathiids metamor- phose into non-feeding adults, reproduce, and die (Smit and Davies 2004, Tanaka 2007). Although gnathiids can be found in the benthos from polar regions to the equa- tor (e.g., Klitgaard 1997), they have been reported mostly in warm temperate and tropical waters, and are one of the most common ectoparasites of coral reef fishes (Grutter 1994, Grutter and Poulin 1998), where they infest a wide range of host spe- cies (Arnal et al. 2001, Ferreira et al 2009, Coile and Sikkel 2013). Though small in size, in sufficient abundance gnathiids have been shown to lower the hosts’ blood volume (Jones and Grutter 2005), cause serious tissue damage, and even kill adult host fish (Bunkley-Williams and Williams 1998). Therefore, even in small numbers, they have the potential to damage or kill newly settled reef fishes, especially given that settlement in reef fishes typically occurs at night (e.g., Stobutzki and Bellwood 1998), when the most and largest gnathiids are active (Grutter 1999, Sikkel et al. 2006, 2009). However, the fact that these events occur at night, that gnathiids remain on hosts only during a brief feeding period, and have the potential to kill small hosts, also makes it difficult to document micropredation by gnathiids in the field. Here we provide the first report of gnathiid isopods infesting settlement-stage reef fishes in the tropical Atlantic Ocean, extend the list of affected fish taxa with five additional families, and provide further experimental evidence that as few as one gnathiid can kill settlement-stage fish. Methods Field Observations.—Lighted plankton traps (Fig. 1) similar to those used in studies of gnathiids at sites on the Great Barrier Reef (Jones and Grutter 2007) were set between May and August 2014 and 2015 in Greater Lameshur Bay, St. John, US Virgin Islands (18.31502°, −64.722904°); a bay dominated by a mix of rocky reef, live and dead Orbicella annularis (Ellis and Solander, 1786) coral, sand, and seagrass. These traps are constructed of PVC and use a LED marker light as a light source and translucent white plastic funnels as one-way entry points. The LED marker light is activated and placed within the trap so that the entry funnel is illuminated. Traps were set 2–3 hrs prior to sunset and were retrieved either 3 hrs after sunset that evening, or 3 hrs post-sunrise the next day. The contents of the traps were filtered through 160-μm plankton mesh and placed in seawater in one or more petri dishes; any gnathiids found were removed, counted, and photographed for another study. The traps attracted and collected a wide range of small organisms, including some settlement-stage fishes. For the remainder of the 2014 field season and all of the 2015 season, when we found settlement-stage fish in a sample, the number of attached gnathiids was noted and the fish with attached gnathiids were placed in a separate petri dish for observation or, in one case noted below, the fish and gnathiid were placed in a larger container overnight. In another case, a settlement stage Acanthurus sp. Artim et al.: Gnathiid micropredation on larval reef fish 481 Figure 1. Light trap used to collect gnathiid isopods and settlement-stage reef fishes. was collected from the water column at night and exposed to three gnathiids in the laboratory overnight. Fish standard length (SL) and key major-axis dimensions for attached gnathiids were measured from digital photographs using ImageJ (Abràmoff et al. 2004). Major- axis dimensions for gnathiids and for their blood and plasma bolus were used to estimate gnathiid juvenile stage and volume of blood and plasma removed by the parasite. Blood and plasma volume was approximated by an ellipsoid whose two mi- nor radii are identical, allowing for estimates from head-on 2-D photographic views. Fish were identified at least to family from the photographic images. Laboratory Experiments.—To further assess whether small numbers of gnathi- ids are capable of causing mortality in settlement-stage fish, we conducted two experi- ments using newly settled French grunt, Haemulon flavolineatum (Desmarest, 1823) collected from Brewers Bay, St. Thomas, US Virgin Islands (18.33053°, −64.961472°). Fish, ranging from 7 to 15 mm total length (TL), were collected by escorting them into a plastic container. In the first experiment, 10 fish were randomly assigned to gnathiid-exposed or control treatment. For the former, a total of 48 stage-3 gnathiid zuphea were added to the 1.5-L aerated container of seawater at dusk, and remained with the fish overnight. No gnathiids were added to the control container. The follow- ing day, the number of fish that died and number of gnathiids that fed were recorded. In the second experiment, individual fish were randomly assigned to 0 n( = 32), 1 (n = 35), 2 (n = 33), or 3 (n = 30) gnathiids placed in 0.40-L containers with a single fish. As before, fish were held overnight and the number of fish that died and number of gnathiids that fed were recorded. For both experiments, Chi-squared analyses were performed to test for an effect of exposure to gnathiids on fish mortality rates. 482 Bulletin of Marine Science. Vol 91, No 4. 2015 Table 1. Summary of field observations on gnathiid micropredation on settlement-stage reef fish. Fish fluid volumes and gnathiid blood meal volumes are estimates based on size measurements. Fork length Fish fluid Total blood meal Body fluid Date observed Family in mm volume in μl volume in μl extracted Fatal May 2014 Labrisomidae 10.2 0.96 0.37 38.7% Yes July 2014 Bothidae 14.1 2.42 0.06 2.5% Yes May 2015 Acanthuridae 25.0 16.13 1.09 6.7% No June 2015 Gobiidae 16.9 4.45 0.36 1.5% Yes June 2015 Gobiidae 14.9 5.37 0.14 0.5% No June 2015 Gobiidae 6.8 0.34 0.09 5.1% Yes July 2015 Apogonidae 12.8 6.66 0.21 0.6% No July 2015 Gobiidae 8.3 1.19 0.68 10.8% Yes July 2015 Bothidae 20.2 14.99 0.54 0.7% Frozen July 2015 Triptigeridae 20.1 14.72 0.04 0.0% Frozen July 2015 Labrisomidae 21.9 18.18 0.06 0.1% Frozen July 2015 Gobiidae 11.3 1.10 0.17 2.9% Frozen Results Field Observations.—A summary of field observations is presented in Table 1.