2010 SOUTHEASTERN NATURALIST 9(4):635–648 Disturbance of the Florida Manatee by an Invasive Catfish Melissa Gibbs1,*, Tiffany Futral1, Megan Mallinger1, Desiree Martin1, and Monica Ross2 Abstract - During the winter, Trichechus manatus latirostris (Florida Manatee) depends on long periods of rest in comparatively warm thermal refuges to help conserve energy and maintain stable body temperatures. Pterygoplichthys disjunctivus (Vermiculated Suckermouth Sailfin Catfish) has been observed attached to, and grazing algae from, Florida Manatee in Volusia Blue Spring. We hypothesized that the disturbance caused by grazing armored catfish would significantly alter Florida Manatee behavior. Analyses of 6 hours of underwater video of Florida Manatee behavior, with and without attached armored catfish, revealed that during each observation period, Florida Manatees with attached catfish demonstrated significantly higher activity levels and numbers of active behaviors. Increased Florida Manatee activity caused by the armored catfish may compound the impact of other known threat effects. Introduction Pterygoplichthys disjunctivus (Weber) (Vermiculated Suckermouth Sailfin Catfish) is an armored, loricariid catfish native to the Madeira River drainage of the Amazon River Basin in Brazil and Bolivia. This species appears to have been in Florida since the late 1950s, but armored catfish have only recently become a significant problem, as they overrun springs and waterways, damage fishing nets, and honeycomb riverbanks with nesting burrows (Fuller et al. 1999; Gibbs et al. 2008; M. Gibbs and K.M. Smedley, unpubl. data; Greene and Lee 2009; Hill 2002; Hoover et al. 2004; Nico et al. 2009). Pterygoplichthys disjunctivus first appeared in Volusia Blue Spring in 1999 (Gibbs et al. 2008). It became apparent by the winter of 2000 that P. disjunctivus was utilizing the spring run as a thermal refuge, moving into the 23 °C spring run when St. Johns River temperatures dropped below that of the spring run (M. Gibbs, unpubl. data). In 2002, we first noticed an interaction between P. disjunctivus and the endangered Trichechus manatus latirostris Harlan (Florida Manatee), which also uses Volusia Blue Spring as a thermal refuge. Armored catfish attached to Florida Manatees with their suckermouths, apparently to graze algal epiphytes, and Florida Manatees were observed trying to dislodge the catfish using a variety of body movements. This behavior has since been described elsewhere, but has not yet been quantified (Nico et al. 2009). Winter distribution of Florida Manatee is determined by seasonal temperature isoclines and it does not normally extend north of Florida (O’Shea 1Department of Biology, Stetson University, DeLand, fl32723. 2Sea to Shore Alliance, 4411 Bee Ridge Road, #490, Sarasota, fl34233. *Corresponding author - mgibbs@stetson.edu. 636 Southeastern Naturalist Vol. 9, No. 4 and Kochman 1990). Although Florida Manatees are large herbivores, they have a very low metabolic rate, and high thermal conductance rate compared to similarly sized mammals (Costa and Williams 1999, Irvine 1983, Reep and Bonde 2006, Walsh et al. 2005). Due to these physiological constraints, Florida Manatees are unable to easily maintain their body temperatures in cold water. The lower limit of the Florida Manatees’ thermoneutral zone is between 20 and 23 °C, and they generally seek out thermal refuges when ambient water temperatures drop below 20 °C (O’Shea and Kochman 1990, Reep and Bonde 2006). Irvine (1983) demonstrated that captive Florida Manatees can maintain a stable body temperature for several days in cold water. Longer cold exposure leads to thermal stress and cold stress syndrome (Bossart et al. 2002). Cold stress syndrome (CSS) is a major cause of nonanthropogenic winter mortality in the northern part of their range (Bossart et al. 2002, Deutsch 2000, O’Shea and Ackerman 1995, Walsh et al. 2005). As seasonal temperatures drop, some Florida Manatees move to south Florida for warmer waters; however, approximately 75% of the population utilizes springs or industrial warm water outflows at higher latitudes (central and north-central Florida) for thermal refuge (Halversen and Keith 2008, Laist and Reynolds 2005a, O’Shea and Kochman 1990, USFWS 2001). Most thermal refuges do not contain enough food to support the Florida Manatees that utilize them, thus necessitating periodic foraging bouts into cold rivers or bays (Berger 2007). In apparent anticipation of limited forage in winter, Florida Manatees will increase their daily rate of foraging in the fall (Bengston 1983, Berger 2007, O’Shea and Ludlow 1992, Reep and Bonde 2006). In the warm seasons, Florida Manatees normally spend up to a third of the day feeding leisurely on a wide variety of marine and freshwater plants. During the winter, when Florida Manatees leave the refuge to feed, their feeding bouts are intense and uninterrupted (Berger 2007). Florida Manatees may remain in the refuge for days without feeding when water temperatures outside the refuges drop below 16 °C, likely reducing their metabolic rate and further increasing their susceptibility to cold stress (O’Shea and Ludlow 1992, O’Shea and Kochman 1990, Westerterp 1977). Since Florida Manatees minimize the frequency of feeding bouts during the winter, they must conserve energy in the refuge by resting most of the day (Costa and Williams 1999, Hartman 1979, King and Heinen 2004). Due to the importance of thermal refuges to the survival of Florida Manatees, the stability of refuges is one of the most critical issues for long-term survival of the species (Bossart et al. 2002, Runge et al. 2007, USFWS 2001). Pterygoplichthys disjunctivus may have reduced the suitability of Volusia Blue Spring as a refuge for Florida Manatees. Armored catfish that live in Volusia Blue Spring run throughout the year, but their numbers increase dramatically during the fall as temperatures and human use of the run decrease (M. Gibbs, unpubl. data). When Florida Manatees enter the spring run during the winter, they are met by numerous adult armored catfish, which quickly attach themselves to the surface of the Florida Manatees. The 2010 M. Gibbs, T. Futral, M. Mallinger, D. Martin, and M. Ross 637 armored catfish appear to be grazing algae (probably diatoms) with their comb-like teeth and leave cleared trails in their wake. Close observation of Florida Manatees reveals no wounds, and neither Florida Manatee skin cells nor other Florida Manatee epibionts (copepods, nematodes, or ostracods) have been found in surveys of armored catfish gut contents, thus indicating that the purpose of this grazing is algae acquisition (M. Gibbs and K.M. Smedley, unpubl. data). Although Florida Manatee interactions with other fish species have been described in both marine and freshwater systems, none show the same response as seen with armored catfish. Numerous Florida Manatees have been observed with attached Echeneis sp. (coprophagic sharksuckers); however, the sharksuckers were not reported to elicit any reaction from the Florida Manatees (Mignucci-Giannoni et al. 1999, Williams et al. 2003). Florida Manatees probably do not react to the presence of sharksuckers because they do not move around on the Florida Manatees’ skin, and as they attach with a modified dorsal fin rather than mouths, they are incapable of rasping the Florida Manatees’ skin (the likely irritant). Small fish (Lutjanus griseus (L.) [Gray Snapper], Lagodon rhomboides (L.) [Pinfish], and Lepomis macrochirus Rafinesque [Bluegill]) have been observed pecking at Florida Manatee epiphytes, again without eliciting a visible reaction from the Florida Manatees. However, similar pecking behavior by larger jacks and Archosargus probatocephalus (Walbaum) (Sheepshead) caused Florida Manatees to flinch and swipe at the fish with their flippers (Hartman 1979). It is possible that a catfish-induced increase in activity could cause Florida Manatees to needlessly expend energy required to deal with cold conditions, thus negatively impacting Florida Manatee fitness and reducing the value of Volusia Blue Spring as a wintertime thermal refuge. To begin addressing this question, we filmed, analyzed, and compared the behavior of Volusia Blue Spring Florida Manatees with and without attached catfish. We quantified Florida Manatee-catfish interactions by ranking the intensity of Florida Manatee behavioral responses based on presumed energy expenditure. We hypothesized that P. disjunctivus’s grazing on Florida Manatees would significantly increase normal Florida Manatee activity levels as they tried to rid themselves of the armored catfish. If increased activity is associated with a significant energetic cost, then new management strategies may need to be developed to address an additional stress from this invasive species. Methods The study site, Volusia Blue Spring (28º56'51"N, 81º20'22.5"W), is a first magnitude (discharges > 2.8 m3 sec-1) oligohaline spring. The spring run is 25 m wide on average, is 620 m long, and discharges approximately 4.2 m3 sec-1 from the Floridan aquifer into the St. Johns River (Fig. 1) (Scott et al. 2004). The depth of the run at mid-channel varies seasonally 638 Southeastern Naturalist Vol. 9, No. 4 from about 1–2 m in the upper portion of the run to 2–5 m in the lower portion of the run, and water temperature averages 23 ºC year round. Volusia Blue Spring is the primary natural thermal refuge used by the St. Johns River Florida Manatee population (O’Shea and Kochman 1990). Over the past five years, an average of 260 individually identified Florida Manatees utilized the spring run during each winter season (W. Hartley, Blue Spring State Park, Orange City, flpers. comm.). The number of Florida Manatees utilizing Volusia Blue Spring each winter has increased steadily over the past 30 years, and reproductive rates have remained stable (W. Hartley, pers. comm.). After 15 hours of direct visual observation of armored catfish-Florida Manatee interactions in the spring of 2005 (M. Gibbs and T. Futral, unpubl. data.), we recorded 6 hours of individual Florida Manatee behaviors with and without armored catfish interactions during October 2005 and February 2006 (about three hours per month). All recordings were made with a CCTV CVC-627WP underwater video camera connected to a pole and lowered into the water from a canoe. All footage was collected in the lower third of the spring run, where Florida Manatees were densely concentrated (Fig. 1). The canoe was positioned approximately 7–10 m away from the Florida Manatees, and as we wanted to minimize the effects of our presence, we did not follow Florida Manatees that moved away from us. Florida Manatees that were fixated on the canoe were identified as “under human influence.” Figure 1. Volusia Blue Spring Run. The Florida Manatee refuge at the lower end of the spring run is indicated by the grey rectangle. 2010 M. Gibbs, T. Futral, M. Mallinger, D. Martin, and M. Ross 639 Upon our return to the lab, the video footage was reviewed and Florida Manatee behaviors were assessed. Individual Florida Manatees were identified by physical characteristics (size, scars, algal coat) and assigned an identification number. This number allowed us to track individuals throughout a day’s taping. For each Florida Manatee that was captured on video, all behaviors observed were categorized and timed, from the moment the Florida Manatee was first seen, until it left the field of view (the observation period). During the observation period, we noted the number of armored catfish and other Florida Manatees in the field of view, the Florida Manatee’s behavior when first observed, the time(s) when catfish attached (some Florida Manatees were initially observed with an attached catfish), the number of catfish attached, the Florida Manatee behavior when catfish were attached, the behavior that dislodged the catfish, and the Florida Manatee behavior after the catfish were gone. We identified 11 distinct Florida Manatee behaviors (Fig. 2) and assigned each to one of six categories based on similar activity levels (for video footage of some of these behaviors, see Supplemental Files 1 and 2, available online at https://www.eaglehill.us/SENAonline/suppl-files/s9-4-Gibbs-s1 and http://www. eaglehill.us/SENAonline/suppl-files/s9-4-Gibbs-s2, and, for BioOne subscribers, at http://dx.doi.org/10.1656/S870.s1 and http://dx.doi.org/10.1656/S870.s2). Each activity category was then ranked from least to most active and assigned a numeric score (0 = resting, 1 = nursing or stationary, 2 = flipper hit or flipper walk, 3 = slow travel or breathing, 4 = tail flip or ab crunching, 5 = fast travel or barrel roll). We reasoned that behaviors with higher activity levels would correlate with energy consumption, although that was not quantified in this study. Because Florida Manatees, both with and without attached catfish, were observed carrying out more than one behavior per observation period, we recorded all behaviors and their duration for each observation period. To compare activity scores among observations of varying duration, we calculated a time-averaged activity score by taking the percentage of time the individual spent displaying each behavior, multiplying the percentage by the appropriate activity score, and summing the scores for the entire observation. We used Mann-Whitney U-tests to compare the time-averaged activity scores, as well as the number of observed behaviors of Florida Manatees with and without attached armored catfish. We also compared respiration rates between all three groups with Mann-Whitney U-tests. Finally, we used a t-test to determine whether increased numbers of armored catfish in the field of view of the Florida Manatee (usually within 3–5 m) resulted in an increased likelihood of an armored catfish attachment. Results Seventy-five Florida Manatee observations were taped over the six hours of video; an observation was a period of continuous footage of a single individual. Twenty-seven Florida Manatees, for a total of 31 observations, were recorded with attached armored catfish for an average of 109 seconds ± 9 SE. 640 Southeastern Naturalist Vol. 9, No. 4 Twenty-two Florida Manatees, for a total of 23 observations, were recorded without attached armored catfish for an average of 93 seconds ± 10.4 SE. We recorded 16 Florida Manatees, for a total of 21 observations of humaninfluenced Florida Manatees, for an average of 153 seconds (± 32 SE). Although an observation period often lasted for a minute or less, armored catfish stayed attached for an average of only 15.8 (± 1.98 SE) seconds before being dislodged. Florida Manatees were able to dislodge armored catfish; however, more than half of the Florida Manatees experienced consecutive armored catfish attachments (one was subjected to 10 in a row). Figure 2. Florida Manatee behaviors. A – Resting (level 0): Not moving, head resting on the bottom, eyes closed; B – Stationary (level 1): No movement, but eyes open or body propped up on flippers on the bottom; C – Nursing (level 1): Attachment of calf’s mouth to teat, very little body movement by calf or female; D – Flipper walking (level 2): Slowly moving along the benthos by walking on flippers; E – Flipper hit (level 2): Movement of flipper(s) back and forth along the body; F – Slow travel (level 3): Slow horizontal movement propelled by partial extension/flexion of the tail, flippers may be used to initiate slow travel or to steer; G – Tail flip (level 4): Quick up and down movement of tail, including arching the back; H – Ab crunch (level 4): Contraction of the abdomen until body takes on an upside-down “U” shape; I – Barrel roll (level 5): Lateral rolling of the body to one side, or complete 360° rotation; J – Travel fast (level 5): Rapid horizontal movement propelled by full extension/ flexion of tail, flippers may be used to initiate fast travel, but are then tucked in while swimming. Breathing (level 3) is not illustrated, but consists of vertical movement up and down in the water column to obtain air. 2010 M. Gibbs, T. Futral, M. Mallinger, D. Martin, and M. Ross 641 The time-averaged activity level of Florida Manatees with attached armored catfish was significantly higher than when armored catfish were not attached (Mann-Whitney, df = 55, P < 0.001; Table 1). Florida Manatees under human influence also had higher time-averaged activity levels than Florida Manatees without catfish (Mann-Whitney, df = 38, P < 0.001; Table 1). The time-averaged activity level of Florida Manatees under human influence with attached catfish was not significantly different from Florida Manatees with catfish attached (Mann-Whitney, df = 36, P < 0.05; Table 1), but was greater than activity levels for all Florida Manatees without catfish (including those under human influence) (Mann-Whitney, n = 45, P < 0.05; Table 1). In particular, human-influenced Florida Manatees spent more time stationary, flipper walking, and travelling slowly, and were never observed resting (Table 2). The number of behaviors per minute exhibited by a Florida Manatee during an observation was significantly higher when armored catfish were attached compared to undisturbed Florida Manatees (Mann-Whitney, df = 55, P < 0.001; Table 1). High-level behaviors (activity level 4 and 5) were Table 1. Mean time-averaged activity levels and mean behaviors per minute for Florida Manatee groups with and without attached catfish. Duration of Time-averaged Behaviors/ Number observation activity level minute of (seconds) Florida Manatee group (± 1 SE) (± 1 SE) observations (± 1 SE) No disturbance 0.64 ± 0.13 2.38 ± 0.85 23 93 ± 10.4 Human influence w/ no attached catfish 1.86 ± 0.48 2.04 ± 0.46 16 153 ± 32 Attached catfish 2.16 ± 0.38 4.74 ± 0.85 31 109 ± 9 Human influence w/ attached catfish 2.47 ± 1.0 2.16 ± 0.74 5 153 ± 32 Table 2. Mean duration and frequency of each behavior for Florida Manatees with and without attached catfish. E = number of Florida Manatees exhibiting behavior, O = number of times behavior observed, No = no catfish, CF = catfish attached, HI = human influence, n.a. = either no data, or a single datum % of all behaviors by Mean duration of Behavior Florida Manatees with behavior (sec.) ± 1 SE (activity level) E O No CF HI No CF HI Resting (0) 37 53 54 41 6 59.6 ± 12.6 43.5 ±7.8 n.a. Nursing (1) 3 6 33 0 66 n.a. (1) n.a. 18.5 ±8.9 Stationary (1) 18 27 16 50 33 70.0 ± 4.05 23.1 ± 5.9 59.0 ± 17.7 Flipper hit (2) 1 1 0 100 0 n.a. n.a. (1) n.a. Flipper walk (2) 12 18 8 50 32 n.a. (1) 15.1 ± 5.4 36.0 ± 9.6 Travel slow (3) 36 71 14 58 28 21.3 ± 8.4 15.9 ± 2.1 45.6 ± 5.7 Breathing (3) 22 34 18 55 28 31.7 ± 14.2 16.6 ± 3.9 13.5 ±2.7 Tail flip (4) 8 11 0 100 0 n.a. 3.54 ± 0.69 n.a. Ab crunch (4) 2 3 0 0 100 n.a. n.a. 14.0 ± 10.6 Travel fast (5) 7 9 0 71 29 n.a. 13.0 ± 4.8 22.6 ± 10.6 Barrel roll (5) 20 41 0 95 5 n.a. 9.5 ± 0.95 n.a. 642 Southeastern Naturalist Vol. 9, No. 4 not seen in undisturbed Florida Manatees (Table 2). One high-level behavior, the tail flip (n = 11), was only seen in Florida Manatees with attached armored catfish. Another high-level behavior, the ab crunch (n = 3) was only seen in human-influenced Florida Manatees. The last two high-level behaviors, barrel roll (n = 41) and fast travel (n = 9), were only seen in Florida Manatees either with armored catfish or under human influence. Florida Manatees without attached armored catfish or human influence, by comparison, spent most of their time resting, stationary or traveling slowly. Florida Manatees with attached catfish had a significantly higher respiration rate than did human-influenced Florida Manatees and Florida Manatees with no attached catfish (Mann-Whitney: df = 57, P less than 0.05; df = 41, P less than 0.05). Human-influenced Florida Manatees had a significantly higher respiration rate than did Florida Manatees with no attached catfish (Mann-Whitney: df = 44, P less than 0.001). The average Florida Manatee was surrounded by 4.8 (± 0.25 SE, range = 0–14) armored catfish. The number of armored catfish in the vicinity of a Florida Manatee was significantly higher when armored catfish were observed attached to that Florida Manatee than when catfish were not attached (t-test, P < 0.001). An average of 1.6 (± 0.12 SE) armored catfish attached to each Florida Manatee, but the number of armored catfish simultaneously attached to a Florida Manatee did not significantly affect Florida Manatee activity scores (linear regression, r = 0.009, P = 0.37). In only nine out of 31 armored catfish-Florida Manatee interactions did Florida Manatees fail to change their original behavior after an armored catfish attached. Among those nine Florida Manatees, six individuals reacted to catfish during subsequent observations and three individuals were not observed again. Discussion Our hypothesis, that armored catfish grazing on Florida Manatees would significantly alter their behavior, was supported. Florida Manatees with armored catfish exhibited higher time-averaged activity levels and twice as many behaviors/minute during an observation period than Florida Manatees without attached armored catfish (Table 1). The general differences in behavior and activity levels are illustrated by comparing two observations of equal duration, one with and the other without attached catfish (Fig. 3). This example illustrates what we generally observed, that Florida Manatees with attached catfish had significantly shorter, more frequent breathing bouts than did Florida Manatees without attached catfish, and that the number, frequency, and variety of most behaviors increased when armored catfish were attached (Table 2). We also found that human-influenced Florida Manatees without catfish were more active than Florida Manatees with no disturbance from humans or catfish. However, they were less active than Florida Manatees with attached catfish, whether they were with or without human influence 2010 M. Gibbs, T. Futral, M. Mallinger, D. Martin, and M. Ross 643 (Table 1). Although the time-averaged activity level of human-influenced Florida Manatees with attached catfish was statistically indistinguishable from Florida Manatees with attached catfish, the low sample size may Figure 3. Time budgets for representative Florida Manatees with (A) and without (B) attached catfish. The Florida Manatee with attached catfish changed behaviors 18 times during 329 seconds. Behaviors included 1 flipper walk, 1 flipper hit, 4 stationary, 6 barrel rolls, 1 tail flip, 2 resting, 1 breathing, and 2 travel slow. Each change in behavior to a higher activity level coincided with the attachment of a catfish. The Florida Manatee without attached catfish exhibited 4 changes in behavior during the 443 second observation period. Behaviors included 3 resting and 2 breathing. Each upward swing in activity level coincided with a breathing episode. 644 Southeastern Naturalist Vol. 9, No. 4 have masked the effect of disturbances. One of the most surprising results was that there was no correlation between the number of simultaneously attached armored catfish and Florida Manatee activity levels; even a single attached armored catfish could cause significant behavioral changes. Furthermore, although the average duration of attachment was 15 seconds, armored catfish did not need to be attached long to elicit a response; many Florida Manatees reacted within 3 seconds. High-level behaviors (activity levels 4 and 5), especially barrel rolls and tail flips, were effective at dislodging armored catfish; however, the armored catfish often reattached to the same Florida Manatee. As yet no data are available to allow us to correlate Florida Manatee behavioral responses to energetic costs. Nonetheless, even though catfish attachments were of short duration and attachments may only initiate short-term changes in energy expenditure, the sheer number of catfish interactions that a single Florida Manatee would experience in a day, week, or season could add up to significant stress. Metabolic studies are needed to quantify energetic costs of changes in Florida Manatee behavior. Although not all Florida Manatees carried attached armored catfish, the strong correlation between the number of armored catfish in the viewing area and the likelihood of armored catfish-Florida Manatee attachment suggests that as armored catfish populations in the spring continue to increase a greater proportion of Florida Manatees might be expected to have attached armored catfish. The fact that some Florida Manatees did not react as quickly, or at all, to the catfish leads to the question of whether Florida Manatees can acclimate to the catfish. Although we did not address this possibility, the fact that (1) most Florida Manatees in Volusia Blue Spring are regular winter visitors (W. Hartley, pers. comm.), (2) we have observed catfish- Florida Manatee interactions since 2002, and (3) many Florida Manatees still reacted swiftly to catfish, suggests that acclimatization is not widespread. It is possible that a lack of response means that an animal is not being disturbed by a particular stimulus; however, it could also mean that those animals are in poor physical condition and simply can’t afford to respond (Williams et al. 2006). Although, P. disjunctivus does not cause any visible signs of physical injury, vocalization studies of Volusia Blue Spring Florida Manatees revealed that Florida Manatees vocalized more often, perhaps in distress, when armored catfish were attached (Williams 2005). In general, the response of Florida Manatees to catfish grazing behavior is similar to that of many ectoparasite hosts: initiation of energetically costly avoidance and cleaning behaviors when ectoparasites attach (Dvoretsky and Dvoretsky 2009, Jog and Watve 2005). Clearly, they irritate the Florida Manatees and cause them to divert time and energy away from normal activities. It appears that the armored catfish have made Volusia Blue Spring a less suitable refuge for Florida Manatees. Undisturbed Florida Manatees in thermal refuges spend nearly half of their time resting, only making brief, intense forays out of the refuge to feed during the warmest part of the day (Berger 2007, Hartman 1979, O’Shea and Ludlow 2010 M. Gibbs, T. Futral, M. Mallinger, D. Martin, and M. Ross 645 1992). The behaviors that we observed when armored catfish were attached were atypical for Florida Manatees in a thermal refuge, but were similar to those described when swimmers and boaters disturbed Florida Manatees in Crystal River. King and Heinen (2004), Buckingham (1990), and Buckingham et al. (1999) found that when swimmers and boaters were present, Florida Manatees spent less time resting on the bottom or nursing, and more time milling, slow swimming, cavorting, and playing. The metabolic rate of sleeping marine mammals has been reported to be 40–60% of waking basal metabolic rates, so deviations from the rest regimen of undisturbed Florida Manatees could have a significant energetic effect (Worthy 2001). Any energetic costs of the increased activity that we observed with armored catfish-Florida Manatee interactions are likely to be met through increased foraging in sub-optimal temperatures. Although Volusia Blue Spring is protected as a Florida Manatee refuge, it does not possess plentiful forage. In addition, the Volusia Blue Spring Florida Manatee population has been steadily increasing over the past 10 years (W. Hartley, pers. comm.). When large numbers of Florida Manatee congregate during the wintertime, they can cause localized depletion of forage, forcing Florida Manatees to travel further from the spring to find food (Bengston 1983). A single stressor (e.g., cold temperatures) may have relatively mild effects, but multiple stressors (e.g., human activity, cold temperatures, and, as shown by this study, armored catfish) will have compound effects on the Florida Manatee’s health and susceptibility to CSS (Bossart et al. 2002, Runge et al. 2007). The added stress caused by armored catfish makes the protection of natural Florida Manatee thermal refuges all the more critical (Runge et al. 2007, USFWS 2006). Springflow in Volusia Blue Spring, for example, has decreased due to human demands for water since monitoring began in the 1930s (SJRWMD 2009), and these demands could reduce the amount of available Florida Manatee refuge. The magnitude of water flow out of the spring generally prevents cold river water from intruding up the spring run; however, during seasonal periods of lower flow, river water intrudes more than 200 m up the run (M. Gibbs, pers. observ.). Since Florida Manatees tend to congregate within 100 m of the interface between the cold St. Johns River and warm spring water, a year-round anthropogenic reduction in flow could result in greater coldwater intrusion and less refuge space (Rouhani et al. 2007). Demands for water, declines in vegetation, blocked access, and boat disturbance in and around springs are continuous and increasing (O’Shea and Ludlow 1992). In the long term, declines in the number and quality of thermal refuges may prove to be a bigger threat to Florida Manatees than boat impacts (Laist and Reynolds 2005b, Runge et al. 2007). Because any harmful interaction between an invasive species and an endangered species is a potential threat, the degree of armored catfish interaction with, or disturbance of, Florida Manatees should be considered when improving management strategies for Florida Manatees within thermal refuge sites. Although Park census numbers (W. Hartley, pers. comm.) indicate 646 Southeastern Naturalist Vol. 9, No. 4 that the health of the Volusia Blue Spring Florida Manatee populations (as indicated by population size and birth rates) seems to be unaffected by catfish as yet, the catfish invasion is still relatively recent, so there could yet be a measurable effect. In addition, Volusia Blue Spring is closed to swimmers and boaters during Florida Manatee season, thus making it a low humandisturbance system; other natural thermal refuges are not as well protected (Buckingham 1990, King and Heinen 2004). So, what can be done about the catfish? It is impractical to try catfish eradication. Armored catfish are firmly established throughout Florida, and they are efficient reproducers (Fuller, et al. 1999, Gibbs et al. 2008). We currently know too little about their biology to effectively manage them. Realistically, the best way to improve Florida Manatee habitat and refuges is to focus on threats that can be managed and reduced. Boating and swimmer harassment regulations could receive heavier enforcement. Spring flow should be preserved through minimum-flow regimes, reductions in impervious groundcover, and water conservation. And most importantly, as Florida Manatee populations increase and/or redistribute due to power plant thermal refuges going offline, the need for additional natural thermal refuge protection (to return human-modified springs to their natural state) becomes essential. Acknowledgments We thank Blue Spring State Park for access to the spring run and use of the Park research canoe, and Blue Spring Enterprises for generously loaning us canoeing gear. Ranger Wayne Hartley kindly provided raw data on individual Florida Manatees visiting the spring run each year. Finally, we thank K. Work and T. Farrell for help with the statistics, and F. Gibbs and K. Work for valuable comments on the manuscript. Stetson University students T. Futral (’05), M. Mallinger (’06), and D. Martin (’07) completed portions of this study for their Senior Research Projects. Literature Cited Bengston, J.L. 1983. Estimating food consumption of free-ranging Florida Manatees in Florida. Journal of Wildlife Management 47(4):1186–1192. Berger, R.W. 2007. Seasonal habitat use of the Florida Manatee (Trichechus manatus latirostris) in the Crystal River National Wildlife Refuge with regards to natural and anthropogenic factors. M.Sc. Thesis. Georgia Southern University, Statesboro, GA. 119 pp. Bossart, G.D., R.A. Meisner, S.A. Rommel, S.-J. Ghim, and A.B. Jenson. 2002. Pathological features of the Florida Manatee cold stress syndrome. Aquatic Mammals 29(1):9–17. Buckingham, C.A. 1990. An evaluation of Florida Manatee distribution patterns in response to public use activities in Kings Bay, Crystal River, Florida. Florida Cooperative Fish and Wildlife Research Unit, University of Florida, Gainesville, fl. 49 pp. Buckingham, C.A., L.W. Lefebvre, J.M. Schaefer, and H.I. Kochman. 1999. Florida Manatee response to boating activity in a thermal refuge. Wildlife Society Bulletin 27(2):514–522. 2010 M. Gibbs, T. Futral, M. Mallinger, D. Martin, and M. Ross 647 Costa, D.P., and T.M. Williams. 1999. Marine mammal energetics. Pp. 176–217, In J.E. Reynolds and S.A. Rommel, (Eds.). Biology of Marine Mammals. Smithsonian Institution Press, Washington, DC. 578 pp. Deutsch, C.J. 2000. Winter movements and use of warm-water refugia by radiotagged West Indian Florida Manatees along the Atlantic coast of the United States. Final Report prepared for the Florida Power and Light Company and US Geological Survey. Dvoretsky, A.G., and V.G. Dvoretsky. 2009. Distribution of amphipods Ischyrocerus on the Red King Crab, Paralithodes camtschaticus: Possible interactions with the host in the Barents Sea. Estuarine, Coastal, and Shelf Science 82:390–396. Fuller, P.L., Nico, L.G., and J.D. Williams. 1999. Nonindigenous Fishes Introduced into Inland Waters of the United States. American Fisheries Society, Special Publication 27, Bethesda, MD. Gibbs, M.A., J.H. Shields, D.W. Lock, K.M. Talmadge, and T.M. Farrell. 2008. Reproduction in an invasive exotic catfish Pterygoplichthys disjunctivus in Volusia Blue Spring, Florida, USA. Journal of Fish Biology 73:1562–1572. Greene, G. and D. Lee. 2009. Social and economic impacts of the loricariid catfish in Florida. Ch. 4. Pp. 39–49, In R.E. Mendoza-Alfaro, B. Cudmore, R. Orr, J.P. Fisher, S.C. Balderas, W.R. Courtenay, P.K. Osorio, N. Mandrak, P.A. Torres, M.A. Damián, C.E. Gallardo, A.G., Sanguinés, G. Greene, D. Lee, A. Orbe- Mendoza, C.R. Martínez, and O.S. Arana (Eds.). Trinational Risk Assessment Guidelines for Aquatic Alien Invasive Species. Commission for Environmental Cooperation, Hill, J.E. 2002. Exotic fishes in Florida. Lakeline Spring 2002:39–43. Halvorsen, K.M., and E.O. Keith. 2008. Immunosuppression cascade in the Florida Manatee (Trichechus manatus latirostris). Aquatic Mammals 34(2):412–419. Hartman, D.S. 1979. Ecology and Behavior of the Florida Manatee (Trichechus manatus) in Florida. Special Publication No. 5, American Society of Mammalogists. Hoover, J.J., K.J. Killgore, and A.F. Cofrancesco. 2004. Suckermouth catfishes: Threats to aquatic ecosystems of the United States? Aquatic Nuisance Species Research Program Bulletin 4(1):1–9. Irvine, A.B. 1983. Florida Manatee metabolism and its influence on distribution in Florida. Biological Conservation 25:315–334. Jog, M., and M. Watve. 2005. Role of parasites and commensals in shaping host behaviour. Current Science 89(7):1184–1191. King, J.M., and J.T. Heinen. 2004. An assessment of the behaviors of overwintering Florida Manatees as influenced by interactions with tourists at two sites in central Florida. Biological Conservation 117:227–234. Laist, D.W., and J.E. Reynolds. 2005a. Influence of power plants and other warmwater refuges on Florida Manatees. Marine Mammal Science 21(4):739–764. Laist, D.W., and J.E. Reynolds. 2005b. Florida Manatees, warm water refuges, and an uncertain future. Coastal Management 33:279–295. Mignucci-Giannoni, A.A., C.A. Beck, R.A. Montoya-Ospina, and E.H. Williams. 1999. Parasites and commensals of the West Indian Florida Manatee from Puerto Rico. Journal of the Helminthological Society, Washington 66(1):67–69. Nico, L.G., W.F. Loftus, and J.P. Reid. 2009. Interactions between non-native armored suckermouth catfish (Loricariidae: Pterygoplichthys) and native Florida Florida Manatee (Trichechus manatus latirostris) in artesian springs. Aquatic Invasions 4(3). Available online at http://www.aquaticinvasions.net/2009/index3.html. 648 Southeastern Naturalist Vol. 9, No. 4 O’Shea, T.J., and B.B. Ackerman. 1995. Population biology of the Florida Manatee: An overview. In O’Shea, T.J., B.B. Ackerman and H.F. Percival (Eds.), Population Biology of the Florida Manatee. National Biological Services Information and Technical Report 1, Department of the Interior, Washington, DC. O’Shea, T.J., and H.I. Kochman. 1990. Florida Manatees: Distribution, geographically referenced data sets, and ecological and behavioral aspects of habitat use. Pp. 11–21, In Reynolds, J.E., and K.D. Haddad (Eds.). Proceedings of a Workshop on Geographic Information Systems in Managing Florida Manatee Habitat. Florida Department of Natural Resources. Marine Research Publication 49. O’Shea, T.J., and M.E. Ludlow 1992. Florida Manatees, Pp. 190–200, In Humphrey (Ed.) Rare and Endangered Biota of Florida, Vol. I Mammals. University of Florida Press, Gainesville, fl. 392 pp. Reep, R.L., and R.K. Bonde. 2006. The Florida Manatee: Biology and Conservation. University Press of Florida, Gainesville, fl. Rouhani, S., P. Sucsy, G. Hall, W. Osburn, and M. Wild. 2007. Analysis of Blue Spring discharge data to determine a minimum flow regime. Prepared for the St. Johns River Water Management District, Palatka, fl. Runge, M.C., C.A. Sanders-Reed, C.A. Langtimm, and C.J. Fonnesbeck. 2007. A quantitative threats analysis for the Florida Manatee (Trichechus manatus latirostris). US Geological Survey, Open-File Report 2007–1086. 34 pp. St. Johns River Water Management District (SJRWMD). 2009. Water resource value monitoring for Blue Spring and Blue Spring Run, Volusia Co., Florida. Special Publication SJ2010-SPS. Palatka, fl. Scott, T.M., G.H. Means, R.P. Meegan, R.C. Means, S.B. Upchurch, R.E. Copeland, J. Jones, T. Roberts, and A. Willet. 2004. Springs of Florida. Florida Geographical Survey Bulletin No. 66. Tallahassee, fl. US Fish and Wildlife Service (USFWS). 2001. Florida Manatee recovery plan, (Trichechus manatus latirostris), Third Revision. US Fish and Wildlife Service, Atlanta, GA. USFWS. 2006. West Indian Florida Manatee (Trichechus manatus) 5-Year review: Summary and evaluation. US Fish and Wildlife Service. Jacksonville, fl. Walsh, C.J., C.A. Luer, and D.R. Noyes. 2005. Effects of environmental stressors on lymphocyte proliferation in Florida Manatees, Trichechus manatus latirostris. Veterinary Immunology and Immunopathology 103:247–256. Westerterp, K. 1977. How rats economize energy loss in starvation. Physiological Zoology. 50:331–362. Williams, E.H., A.A. Mignucci-Giannoni, L. Bunkley-Williams, R.K. Bonde, C. Self- Sullivan, A. Preen, and V.G. Cockcroft. 2003. Echeneid-sirenian associations, with information on sharksucker diet. Journal of Fish Biology 63:1176–1183. Williams, L. 2005. Individual distinctiveness, short- and long-term comparisons and context specific rate of Florida Manatee vocalizations. M.SC. Thesis. University of North Carolina, Wilmington, NC. Williams, R., D. Lusseau, and P.S. Hammond. 2006. Estimating relative energetic costs of human disturbance to Killer Whales (Orcinus orca). Biological Conservation 133:301–311. Worthy, G.A.J. 2001. Nutrition and energetics. Pp. 791–827, In F.M.D. Gulland and L.A. Dierauf (Eds.). CRC Handbook of Marine Mammal Medicine: Health, Disease, and Rehabilitation. CRC Press, Boca Raton, fl. 1025 pp.