Nests and Nest Habitats of the Invasive Catfish Hoplosternum littorale

Transcription

1 2004 SOUTHEASTERN NATURALIST 3(3): Nests and Nest Habitats of the Invasive Catfish Hoplosternum littorale in Lake Tohopekaliga, Florida: A Novel Association with Non-native Hydrilla verticillata LEO G. NICO 1,* AND ANN MARIE MUENCH 2 Abstract - Hoplosternum littorale is a South American catfish (Family Callichthyidae) first discovered in the United States in 1995 in Florida. The presence of H. littorale was documented from early 2002 to late 2003 in Lake Tohopekaliga (Kissimmee River Basin) in central Florida. In this paper, 22 H. littorale nests and nest sites are described. The characteristic bubble nests were present from late May to early September, with number of nests peaking in August when water stage and temperature were both high. Nest habitats (shallow, open marshes) and timing of nest construction (rainy season) were similar to what has been reported for H. littorale in its native range. Most nests (n = 14) were in areas dominated by Hydrilla verticillata and constructed largely from parts of this Asian aquatic plant, representing a unique association between two non-native species. Nevertheless, during August, as water levels increased, nesting shifted from H. verticillata-dominated communities to use of inshore grass zones dominated by Luziola fluitans. Knowledge of H. littorale nesting seasonality and habitat preferences may be useful for any efforts to control or manage this invasive fish. Introduction The plated catfish Hoplosternum littorale (Hancock) (Family Callichthyidae), commonly known as the brown hoplo, is a medium-sized fish (16 30 cm total length) native to lowland areas of South America from Venezuela, Trinidad, and the Guianas, south to northern Argentina (from about 9 N to about 35 S latitude) (Hostache and Mol 1998, Mol 1994, Reis 1997). The species was first discovered in the United States in 1995, in ditches flowing into the Indian River lagoon system near Florida s Atlantic coast (Nico et al. 1996). Subsequently, H. littorale has expanded its range and now occurs in much of central and southern Florida (unpubl. data, L.G. Nico). For a variety of reasons, the presence of this species in the U.S. is cause for concern. Hoplosternum littorale appears to be increasingly widespread, abundant, and exhibits a number of life-history attributes suggesting the species should be classified as highly invasive (Nico et al. 1996, Hostache and Mol 1998). 1 United States Geological Survey, Florida Integrated Science Centers, 7920 NW 71 st Street, Gainesville, FL 32653; 2 Florida Cooperative Fish and Wildlife Research Unit, University of Florida, PO Box , Bldg. 810, Gainesville, FL * Corresponding author - leo_nico@usgs.gov.

2 452 Southeastern Naturalist Vol. 3, No. 3 Because of its wide distribution, abundance, and importance as human food in South America, H. littorale has been studied more than most other neotropical fishes (Hostache and Mol 1998). Nevertheless, little is known about introduced populations in Florida. Within its native range, the air-breathing H. littorale occurs in both freshwater and brackish water habitats, typically occupying marshes, small slow-moving streams, side channels, pools, and various types of disturbed and artificial environments. Of particular interest is the species reproductive biology. In their native range, populations exhibit a marked shift in habitat (Machado- Allison and Zaret 1984, Mol 1994). During the dry season, the catfish survives in deeper, more permanent aquatic habitats (e.g., canals and dry-season pools). As water levels rise with the progression of the rainy season, adult H. littorale move into vegetated littoral areas (e.g., cattail Typha sp. and spikerush Eleocharis sp. marshes, rice fields). Spawning occurs in these newly flooded areas. As part of the reproductive process, adult males temporarily develop enlarged pectoral-fin spines armed with a hook at the distal tip. The hooked spines are used by males to transport plant material during construction of a spawning nest and as a weapon used aggressively in defending the nest against intruders (Andrade and Abe 1997, Machado-Allison and Zaret 1984, Mol 1993; pers. observ., L.G. Nico). Each nest is constructed by a lone male. The typical nest is an elaborate floating structure composed of bubbles overlain by a mound of dead and dying herbaceous plant material (Carter and Beadle 1931; Machado- Allison and Zaret 1984; Mol 1993, 1994). From early 2002 through most of 2003, A.M. Muench studied changes in the fauna and flora of Lake Tohopekaliga in central Florida. The study included setting and checking traps to sample aquatic vertebrates in the Lake s littoral zone. Adult and juvenile H. littorale were frequently taken in traps in 2002, indicating that the species was present in the Lake throughout the year. Also in 2002, A.M. Muench encountered a series of unusual dome-shaped vegetation masses, later recognized as nests of H. littorale. After discovering new nests in May 2003, detailed nest data were recorded. Herein we document the presence and distribution of nests in the Lake, and describe the nests and nest-site habitats during the reproductive season of We also briefly summarize the literature on the subject based on studies of natural H. littorale populations in South America. Study Area Lake Tohopekaliga is a large, shallow natural lake located in Osceola County, FL, at 28 8'27" N to 28 17'43" N latitude (Fig. 1). The

3 2004 L.G. Nico and A.M. Muench 453 Lake is one of a series of large interconnected water bodies, known as the Kissimmee Chain of Lakes, that forms the headwaters of the Kissimmee River. At normal high pool stage (i.e., m National Geodetic Vertical Datum, NGVD), the Lake area covers 8420 ha (pers. comm., D. Tarver, Tallahassee, FL), but varies dramatically with differences in pool stage (see Moyer and Williams 1982). Average maximum depth is typically less than 4.5 m (pers. comm., E. Harris, Orlando, FL) and littoral areas reportedly encompass 50% of the Lake at high pool stage (Moyer and Williams 1982). Based on a bathymetric survey conducted by ReMetrix LLC (Carmel, IN) in 2000, at a surface elevation of m NGVD the surface area for the 0 1 meter contour is 1066 ha (pers. comm., R. Moore, Carmel, IN), representing 13% of the total Lake surface area. Over recent decades Lake Tohopekaliga has become increasingly eutrophic (Moyer and Williams 1982, Moyer et al. 1995). Consequently, littoral habitats of the Lake have enlarged and become overgrown with nuisance vegetation, including hydrilla (Hydrilla verticillata [L. f.] Royle), pickerelweed (Pontederia cordata L.), and cattails (Typha spp. L.). Hydrilla, in particular, has been a problem, covering as much as 45% of the Lake during the 1990s (Feller et al. 1999) and increasing to 7615 ha (63%) when surveyed by the state in October 2003 (pers. comm., D. Barber, Orlando, FL). Associated with excessive plant growth, organic sediment (muck) also is characteristic Figure 1. Map of Lake Tohopekaliga showing route traveled weekly by airboat and approximate locations of 19 of 22 Hoplosternum littorale nests discovered June September Shape of symbol associated with nests indicates primary plant material used in nest construction.

4 454 Southeastern Naturalist Vol. 3, No. 3 of the Lake s littoral zone. Beginning in the early 1970s there have been several attempts to improve Lake habitat, including extreme drawdowns and mechanical removal of vegetation and associated organic sediments. During the present study, the littoral zone was characterized by several major wetland habitats. Vegetation zones were determined by water depth and, from shore to deeper open waters of the Lake center, were characterized by: 1) submersed to low growing emergent grasses dominated by southern watergrass (Luziola fluitans [Michx.] Terrell & H. Robins) or torpedo grass (Panicum repens L.); 2) robust/large emergents consisting mainly of dense rooted stands and floating mats of pickerelweed and cattail; 3) submergents dominated by hydrilla; and 4) islands of the deepwater emergent Egyptian panicgrass (Paspalidium geminatum [Forsk.] Stapf). During the period Jan 2002 through Nov 2003, water levels in the Lake ranged from to m NGVD. Methods Field observations on Hoplosternum littorale nests were made during weekly surveys (deploying and checking traps) as part of a study of reptiles, amphibians, and fishes inhabiting the littoral zone. Trapping activities involved traversing much ( 70%) of the Lake s periphery by airboat (Fig. 1) weekly from January 2002 through November No trapping occurred during periods with no standing water in the pickerelweed zone (May and June 2002), and during an attempted drawdown (November 2002 January 2003). Only nest abundance was recorded prior to June Detailed observations on H. littorale nests and nest sites were made from June 2003 onward while traveling between trap sites. For all nests encountered during that period, vegetation immediately surrounding each nest (i.e., nest site) and materials (i.e., plants) used in nest construction were recorded. Selected nests were closely examined, determining their geographic point location using GPS and documenting the structure and habitat with a digital camera. In addition, several nests were opened by hand to document internal nest materials, structure, and whether eggs were present. Nest height (above water line), nest diameter, and water depth (at the nest) were measured for three nests, and water depth was estimated for several other nests. Dimensions of an egg mass from a single nest were also measured. Air temperature and rainfall data presented here were obtained from the National Oceanic and Atmospheric Administration-National Weather Service Forecast office, Melbourne, FL, for the Orlando airport (available on the intenet at

5 2004 L.G. Nico and A.M. Muench 455 climatology.html). National Geodetic Vertical Datum data for Lake Tohopekaliga were taken from the South Florida Water Management District s corporate environmental database (DBHYDRO) (available on the internet at Extent of littoral zone illustrated in Figure 1 is based on a map appearing in Moyer et al. (1995). Results Nesting cycle A total of 24 Hoplosternum littorale nests was found during 2003 (Fig. 2). The earliest (n = 2) were observed in late May, but most (n = 21) were discovered June through August. The greatest number of nests was discovered in August, when water stages and temperature were both high. No new nests were found following the discovery of a single nest, without eggs, on 4 September. During the nesting season, mean monthly air temperature was above 26.5 C. Nest site habitats All H. littorale nests encountered in Lake Tohopekaliga were restricted to the littoral zone in relatively open, shallow marshes (Fig. 1). None of the nests were grouped together, the nearest pair were 100 m apart. Low intensity searches of dense emergent areas (e.g., pickerelweed communities) yielded no nests, although it is possible that nests went undetected due to low visibility. Similarly, no nests were seen in the more central parts of the Lake, even though hydrilla was common in many of these deepwater areas. Habitat information on the two nests discovered in May 2003 was not recorded. Of the 22 nests for which notes were taken, the majority (n = 14) were constructed in areas dominated by hydrilla. Nevertheless, by late summer we noted a marked shift in habitat selected for nest construction, from hydrilla-dominated zones to more landward communities dominated by watergrass, corresponding to a seasonal increase in water levels (Fig. 2). According to available Lake water level data for those days when individual nests were located, nesting in hydrilla communities (n = 14) occurred when Lake stages ranged from to (mean = 16.26) m NGVD. In contrast, nesting in watergrass communities occurred during stages ranging from to (mean = 16.83) m NGVD. Water depth was measured at 10 nest sites, with values ranging from 30 to 80 cm. Nest description Almost all of the 22 nests examined were composed largely of the dominant vegetation in the immediate area, typically either hydrilla

6 456 Southeastern Naturalist Vol. 3, No. 3 Figure 2. Lower panel: histogram showing number and type of Hoplosternum littorale nests observed each month in Lake Tohopekaliga, FL, during the period Nov 2002 Oct 2003, and monthly mean air temperature for east-central Florida; includes total number of new nests observed, primary nest construction material (bar shading), and number of nests with eggs present/number of nests examined (see numbers above bars Jun Sep). Middle panel: monthly total rainfall for eastcentral Florida. Upper panel: monthly water stages for Lake Tohopekaliga shown as minimum, mean, and maximum National Geodetic Vertical Datum (NGVD). (See methods for sources of air temperature, rainfall, and water level data.)

7 2004 L.G. Nico and A.M. Muench 457 (64% of nests) or watergrass (32%) (Figs. 2, 3). The single exception was a nest composed entirely of bladderwort (Utricularia spp. L.). Six nests were composed of two or more different species of vegetation, although some plants were of relatively minor importance in terms of overall nest biomass. In one nest composed largely of hydrilla, a single American lotus (Nelumbo lutea Willd.) leaf was situated on top of the dome, either placed there by the resident catfish A B Figure 3. Hoplosternum littorale nests: A) constructed of Hydrilla verticillata; B) constructed of watergrass ( Luziola fluitans); C)partially opened nest (A above) revealing foam and eggs. Black and white bars equal approximately 10 cm. Eggs C

8 458 Southeastern Naturalist Vol. 3, No. 3 or inadvertently pushed up from below as the nest was being built. In terms of frequency of occurrence, components of 22 nests included: Hydrilla verticillata (14 nests), Luziola fluitans (7), Utricularia spp. (2), soft rush (Juncus effusus L.) (2), water primrose (Ludwigia spp.) (2), alligatorweed (Alternanthera philoxeroides [Mart.] Griseb.) (1), fragrant water lily (Nymphaea odorata Ait.) (1), Pontederia cordata (1), and Nelumbo lutea (1). Based on 11 dismantled nests, plant material forming the cap of the nest generally was of the same type as that used in the underlying structure. Nests were generally in the form of low domes, typically with circular footprints. Nest sizes appeared to be relatively uniform; the three measured were 30 to 38 cm in diameter and 8.5 to 9 cm high. None of the 22 nests most closely examined appeared to be free-floating. Those constructed of hydrilla were extensively connected to live-rooted hydrilla surrounding them. Similarly, nests of watergrass also were connected to nearby rooted vegetation. Eggs Of 11 nests dismantled, 6 (55%) contained H. littorale eggs (Figs. 2, 3). In each case, eggs were grouped in a large mass (Fig. 3). An egg mass from one nest measured 12 cm in diameter and 2 cm high. In general, each mass contained eggs all in the same approximate stage of development. The only exception was a nest found 28 August, in which eggs in the upper layers were relatively small, possibly infertile. In contrast, the bottom layers of eggs were large and further developed. The larger eggs broke, when prodded, releasing fry that immediately swam away. Discussion Nesting cycle Research in Suriname revealed that Hoplosternum littorale eggs typically hatch within about three days after spawning, and overall life span of the nest structure is less than five days (Mol 1993). Mol (1993) demonstrated that within 1 day of removing the guarding male the nest breaks down, suggesting the male is continually adding bubbles to the nest and if this stops the nest will break apart and sink. Consequently, it is reasonable to assume that all or most nests observed during this study were newly constructed. Based on presence of nests, the 2003 spawning season of H. littorale in Lake Tohopekaliga extended from late May through early September (Fig. 2). Peak nesting and spawning occurred during summer months, a period falling within the wet (warm) season typical for east-central Florida (i.e., late May to October; Lascody 2002). In South America,

9 2004 L.G. Nico and A.M. Muench 459 H. littorale nesting also occurs during the rainy season, a pattern reported from all regions investigated, including the Paraguayan Chaco (Carter and Beadle 1931), Rupununi Savanna of Guyana (Lowe-McConnell 1964), Venezuelan Llanos (Machado-Allison and Zaret 1984), coastal plain of Suriname (Mol 1993, 1996), and Trinidad (Ramnarine 1995). Nest construction has been reported to occur throughout the rainy season, starting with the flooding of marshes and continuing until the habitats dry up (Mol 1993, 1996). In some regions, a single year may be characterized by two distinct rainy seasons and in Suriname, H. littorale nesting takes place during both (Mol 1993). Unlike the short nesting season observed in Florida, Mol (1996) found that at some sites in Suriname nesting occurred in as many as eight months a year. Moreover, outdoor captive breeding experiments in South America have demonstrated that male H. littorale may sequentially construct up to 14 nests during a single breeding season (Pascal et al. 1994, cited in Mol 1996). Nest site habitats Our observations of Florida nests in open shallow marshes are consistent with what has been reported regarding nesting habitat of H. littorale in South America (Carter and Beadle 1931; Hostache and Mol 1998; Mol 1993, 1996). Literature on H. littorale frequently refers to their breeding environments as swamp, but authors generally describe the sites as having very few if any trees or other woody vegetation. In North America, such habitats are more typically referred to as marsh. In the tropical Paraguayan Chaco, Carter and Beadle (1931) noted that nest sites occurred in standing waters deprived of oxygen, situated in open water at edges of swamps among floating weed and other aquatic plants at the surface of water. Based on field research in the coastal plain marshes of Suriname, Mol (1993, 1996) provided the most detailed descriptions of H. littorale nests and nest habitats. He reported that in natural areas the catfish typically spawned in extensive marshes dominated either by Typha sp. or Eleocharis sp. Catfish generally selected more open areas along marsh edges, normally far from trees. Nests were rarely found in dense emergent vegetation in the central region of the marsh and never in holes. Mol (1996) concluded that nests were not constructed at random, but apparently built in carefully preselected sites within the fish s overall habitat. Artificial environments also are used for nesting. In Suriname, Mol (1996) found many nests in flooded rice fields, but only when plants were still small and the habitat relatively open. Once rice stands became dense, no new nests were found. Hostache and Mol (1998) indicated that a few nests in Suriname were built in deep canals, but no details were

10 460 Southeastern Naturalist Vol. 3, No. 3 given. In the Venezuelan Llanos, Machado-Allison and Zaret (1984) discovered H. littorale nests associated with roadside canals and borrow pits, habitats created by humans during road construction. They noted that nests were easily visible due to the sparse vegetation and also because of the great quantity of white foam surrounding the nests. Machado-Allison and Zaret did not state precisely where nests were located in these artificial habitats, although it seems likely the structures were situated within the vegetated flooded margins. The association between H. littorale and Hydrilla verticillata in Lake Tohopekaliga is interesting because it demonstrates use of a nonnative invasive plant by an introduced fish. Hydrilla was imported from southern India (Madeira et al. 1997) in the early 1950s. It entered Florida s inland waters after plants were discarded into canals in both Tampa and Miami in the mid and late 1950s (Schmitz et al. 1991). By the late 1980s the plant had overtaken most of the Kissimmee Chain of Lakes system (Feller et al. 1999). Although first discovered in Florida in 1995, H. littorale was likely introduced to the state years earlier (Nico et al. 1996). The rather unique and common association between both nonnative fish and plant in Lake Tohopekaliga suggests the presence of hydrilla is facilitating, possibly promoting, the spread and successful establishment of the catfish in this and other waters of Florida. As one study suggests, dense stands of hydrilla may contain fewer large predatory fishes (e.g., Micropterus salmoides) (Colle and Shireman 1980), thereby conferring a survival advantage to juvenile H. littorale. Nevertheless, our limited data suggest shallow hydrilla marsh is not necessarily the only preferred spawning habitat. Rather, use of this particular type of wetland (and use of hydrilla as nest material) may simply be that hydrilla is a dominant plant in the Lake that also satisfies H. littorale spawning requirements. In fact, we cannot rule out the possibility that grass-dominated wetlands, though covering much less area than hydrilla, are more desirable as nesting habitat. A state survey conducted 9 October 2003 indicated Luziola covered only about 126 ha (2%) of the entire Lake (pers. comm., D. Barber, Orlando, FL), yet 6 of 9 H. littorale nests discovered in August 2003 were in this grassy habitat. The fact that Lake Tohopekaliga nest sites were located in areas ranging from 30 to 80 cm deep also agrees with information on habitat preferences of native populations. In South America, nesting of H. littorale is closely associated with water levels, with the first nests appearing during the rainy season when water levels reach 25 to 30 cm (Mol 1993, 1996; Singh 1978, cited in Hostache and Mol 1998). Hostache and Mol (1998) speculated that 30 cm minimum depth may be needed for catfish to gain access to spawning areas. Earlier researchers do not discuss maximum depths, but it appears H. littorale typically do

11 2004 L.G. Nico and A.M. Muench 461 not nest in water greater than 1 m deep. For instance, Mol (1996) examined 220 nests in Suriname marshes and reported average depth at nest sites to be 35.5 cm (SE = 0.97). Mol (1996) noted that in Suriname nests were widely dispersed, and grouping did not seem to be important. Even in marshes with the highest nest densities, Mol calculated there were less than nests per 100 m 2 (i.e., 16 per ha). Mol (1993) found that nests were generally widely spaced, never less than 10 m apart in natural marshes and 3 m apart in rice fields. Nests in Lake Tohopekaliga were widely dispersed, the nearest pair 100 m apart. It is likely that Lake Tohopekaliga contained more nests during the 2003 breeding season than were actually observed. Bathymetric data indicate 1066 ha or 13% of the Lake is 1 m. Assuming about half of that area is suitable H. littorale nesting habitat and using Mol s (1996) maximum density estimate of about 16 nests per ha, Lake Tohopekaliga could provide adequate habitat for more than 800 nests during a single spawning season. This crude estimate assumes 500 ha are actually available (based on depth requirements and vegetation composition) as nesting habitat. The low number of nests observed may be due to several factors, including the likely possibility that many nests were overlooked. The Lake s population of breeding adults also may also be kept low because of frequent winter kills. In Florida large numbers of non-native fishes of tropical origin die during winter months, presumably because of winter temperatures below their tolerance (unpubl. data, L.G. Nico). Nevertheless, unpublished trap data suggest the H. littorale population in Lake Tohopekaliga is large. Nest description In Lake Tohopekaliga, nest materials nearly always consisted of parts from the most common plants immediately surrounding the nest. Rarely did nest composition suggest possible selectivity by the builder. In one case, plants nearest the nest included several species (i.e., pickerelweed, water spangles [Salvinia minima Baker], swamp smartweed [Polygonum hydropiperoides Michx.], bladderwort [Utricularia spp.]), but only bladderwort was used in nest construction. Another nest was composed of a combination of watergrass and softrush, although swamp smartweed and alligatorweed were nearby. The literature indicates that filamentous herbaceous plants are the nest materials most commonly used by native populations of H. littorale. One of the earliest descriptions of nest materials is that of Carter and Beadle (1931), who conducted research in the Paraguayan Chaco. Their account notes only that nests were composed of dead sticks and leaves floating weed and other aquatic plants made of the dead and dying leaves and stems of these plants. Machado-

12 462 Southeastern Naturalist Vol. 3, No. 3 Allison and Zaret (1984) reported on nests in the Venezuelan Llanos, documenting the use of pieces of horsetail paspalum (Paspalum fluitans [Ell.] Kunth; reported as P. repens) and branches and stems of Mimosa picra (sic) (i.e., black mimosa [M. pigra L.]). Mol (1993, 1996) provided the most detailed descriptions of nest materials. Examining hundreds of nests in Suriname from different habitats over several years, Mol reported a wide variety of dead or dying plant materials used in nest construction, including mostly grasses (e.g., West Indian marsh grass [Hymenachne amplexicaulis [Rudge] Nees]), narrowleaf cattail (Typha angustifolia L.), fireflag (Thalia geniculata L.), marsh pennywort (Hydrocotyle umbellata L.), water snowflake (Nymphoides indica [L.] Kuntze), water lily (Nymphaea sp.), and floating aquatics such as water lettuce (Pistia sp.), water spangles, mosquitofern (Azolla sp.), and duckweeds (Lemnaceae). Mol (1993) commented that nests were rarely built of leaves from dicotyledonous trees. There is little information in the literature on anchoring of H. littorale nests. In Venezuela, Machado-Allison and Zaret (1984) reported that nests were often anchored to floating stems of garden puff (Neptunia prostrata [Lam.] Baill.). Hostache and Mol (1998) stated that in rice fields, nests were often attached to clumps of grass. Most Lake Tohopekaliga nests were attached to nearby rooted aquatic plants, either hydrilla or watergrass. Anchoring likely prevents the floating nest from drifting and being destroyed during windy conditions. In Lake Tohopekaliga, hydrilla habitats often are subjected to large waves due to wind blowing across the open water. In terms of nest size, the few nests that we measured fit well within the range described for native H. littorale populations. Carter and Beadle (1931) noted nests to be 30 cm diameter, and Machado- Allison and Zaret (1984:143) reported diameters of 20 cm. Based on measurements of more than 100 nests, Mol (1993) reported diameters typically ranged from 25 to 35 cm (mean = 30 cm) and height above water surface 6 cm, with large nests up to 42 cm in diameter and 10 cm high. Eggs In South America, reported numbers of eggs per nest vary. Counts range from 2000 to over 60,000 eggs, higher figures being the result of egg deposition by multiple females (Hostache and Mol 1998, Machado- Allison and Zaret 1984, Mol 1993). Eggs are placed in a chamber situated in the center of the nest, resting on a mass of foam under the dome of plant material. This study documents the presence of two distinct developmental stages in a single nest, contrasting with Mol (1993) who stated that H.

13 2004 L.G. Nico and A.M. Muench 463 littorale eggs in any one nest are always of the same developmental stage. Mol concluded that spawning at a given nest, even if involving multiple females, is performed over a brief period of time (presumably within 24 hours). It is reasonable to conclude that the two stages of eggs discovered in this study represent deposition by two different females. Based on relative position, eggs in the upper half were likely deposited first. If not viable, then the eggs were likely spawned only shortly before the second (lower) set of eggs. Conversely, if the upper eggs were fertile, then the behavior may represent a bet-hedging strategy by the male, possibly in response to central Florida s relatively short breeding season. Conclusions Hoplosternum littorale is one of more than 30 foreign fishes established in Florida (Fuller et al. 1999, Nico and Fuller 1999). This South American catfish shares many ecological and life-history attributes typical of other non-indigenous fishes that are becoming increasingly abundant and widespread in peninsular Florida and other southern states: medium to large body size, parental care of eggs or young, a generalized diet, ability to breathe air, and broad environmental tolerances (Nico and Fuller 1999, Nico and Martin 2000). In fact, H. littorale appears to have spread more rapidly than any other recently introduced Florida fish (unpubl. data, L.G. Nico). Aquarium observations, in combination with gut content analyses, indicate the fish is a voracious predator that consumes a wide variety of benthic invertebrates (pers. obs., L.G. Nico). It has been suggested that H. littorale even preys on eggs of native fishes, but evidence is lacking. Although our Lake Tohopekaliga data indicate possible spawning preferences, H. littorale is also flexible, using several habitats and a variety of plants for nest construction. For instance, a nest with eggs was found in a narrow wetland situated between a cypress dome and hydric flatwoods in Lee County, and was constructed of slash pine (Pinus elliottii Engelm.) needles and live oak (Quercus virginiana P. Mill.) leaves (pers. comm., J. Surdick, Gainesville, FL). In addition, all of the plants identified by Mol (1993) as used in nest construction by native H. littorale also occur in Florida, either as native or introduced species. These include two recent, highly invasive introductions, Nymphoides indica (a native of Southeast Asia) and the grass Hymenachne amplexicaulis, indigenous to South and Central America (pers. comm., C. Jacono, Gainesville, FL). Successful control of native animals typically involves targeting vulnerable stages in their life cycle. Our findings, together with information in the scientific literature, indirectly argue that H. littorale

14 464 Southeastern Naturalist Vol. 3, No. 3 populations may be most vulnerable during their reproductive season. In particular, in places where H. littorale is recognized as a problem, it may be possible to significantly limit population size by locating and somehow eliminating nests and breeding fish. Acknowledgments For field assistance, we thank Scott Berryman, John Davis, and Chris Tonsmeire. A.M. Muench s research in Lake Tohopekaliga was supported by the Florida Fish and Wildlife Conservation Commission. A.M. Muench thanks Wiley Kitchens for administrative support. We also appreciate the assistance of Janell Brush with logistics and Zachariah Welch for plant identification advice. Colette Jacono provided valuable input on aquatic plants. For providing data on lake depth and plant coverage, we thank Ed Harris and Dean Barber of the Florida Department of Environmental Protection, David Tarver of Sepro Corporation, and Ryan Moore of ReMetrix LLC. Amy Benson provided help generating the base map. We thank James Surdick, Jr., for sharing his nest data. William Smith-Vaniz, Walter R. Courtenay, Jr., C. Jacono, Lisa Jelks, and Howard Jelks kindly reviewed various drafts of the manuscript and provided valuable comments. Literature Cited Andrade, D.V., and A.S. Abe Foam nest production in the armoured catfish. Journal of Fish Biology 50: Carter, G.S., and L.C. Beadle The fauna of the swamps of the Paraguayan Chaco in relation to its environment. II. Respiratory adaptations in the fishes. Journal of the Linnean Society of London (Zoology) 37: Colle, D.E., and J.V. Shireman Coefficients of condition for largemouth bass, bluegill, and redear sunfish in hydrilla-infested lakes. Transactions of the American Fisheries Society 109: Feller, E., M. Bodle, and E. Harris The progression of hydrilla management in the Kissimmee Chain of Lakes. Pp , In D.T. Jones and B.W. Gamble (Eds.). Florida s Garden of Good and Evil: Proceedings of the 1998 Joint Symposium of the Florida Exotic Pest Plant Council and the Florida Native Plant Society. Florida Exotic Pest Plant Council, Ft. Myers, FL. 369 pp. Fuller, P.L., L.G. Nico, and J.D. Williams Nonindigenous Fishes Introduced into Inland Waters of the United States. American Fisheries Society, Special Publication 27, Bethesda, MD. 613 pp. Hostache, G., and J.H. Mol Reproductive biology of the neotropical armoured catfish Hoplosternum littorale (Siluriformes: Callichthyidae): A synthesis stressing the role of the floating bobble nest. Aquatic Living Resources 11: Lascody, R The onset of the wet and dry seasons in east central Florida: A subtropical wet-dry climate? National Oceanic and Atmospheric Administration/ National Weather Service Forecast Office, Melbourne, FL. Available at (accessed November 2003).

15 2004 L.G. Nico and A.M. Muench 465 Lowe-McConnell, R.H The fishes of the Rupununi savanna district of British Guiana, South America. Part 1. Ecological groupings of fish species and effects of the seasonal cycle on the fish. Journal of the Linnean Society of London (Zoology) 45(304): Machado-Allison, A., and T.M. Zaret Datos sore la biologia reproductiva de Hoplosternum littorale (Siluriformes: Callichthyidae) de Venezuela. Acta Cientifica Venezolana 35: Madeira, P., T. Van, D. Steward, and R. Schnell Random amplified polymorphic DNA analysis of the phenetic relationships among world-wide accessions of Hydrilla verticillata. Aquatic Botany 59: Mol, J.H Structure and function of floating bubble nests of three armoured catfishes (Callichthyidae) in relation to the aquatic environment. Pp , In P.E. Ouboter (Ed.). The Freshwater Ecosystems of Suriname. Monographiae Biologicae, Volume 70. Kluwer Academic Publishers, Dordrecht, Netherlands. 292 pp. Mol, J.H Effects of salinity on distribution, growth and survival of three neotropical armoured catfishes (Siluriformes: Callichthyidae). Journal of Fish Biology 45: Mol, J.H Reproductive seasonality and nest-site differentiation in three closely related armoured catfishes (Siluriformes: Callichthyidae). Environmental Biology of Fishes 45: Moyer, E.J., M.W. Hulon, J.J. Sweatman, R.S. Butler, and V.P. Williams Fishery responses to habitat restoration in Lake Tohopekaliga, Florida. North American Journal of Fisheries Management 15: Moyer, E.J., and V.P. Williams Effects of lake bottom channelization on invertebrate fish food organisms in Lake Tohopekaliga, Florida. Proceedings of the Annual Conference of the Southeastern Association of Fish and Wildlife Agencies 36: Nico, L.G., and P.L. Fuller Spatial and temporal patterns of nonindigenous fish introductions in the United States. Fisheries 24(1): Nico, L.G., and R.T. Martin The South American suckermouth armored catfish Pterygoplichthys anisitsi (Pisces: Loricariidae) in Texas, with comments on foreign fish introductions in the American Southwest. Southwestern Naturalist 46: Nico, L.G., S.J. Walsh, and R.H. Robins An introduced population of the South American callichthyid catfish Hoplosternum littorale in the Indian River Lagoon system, Florida. Florida Scientist 59: Pascal, M., G. Hostache, C. Tessier, and P. Vallat Cycle de reproduction et fecondité de l Atipa, Hoplosternum littorale (Siluriformes), en Guyane Française. Aquatic Living Resources 7: Ramnarine, I.W Induction of nest building and spawning in Hoplosternum littorale. Journal of Fish Biology 47: Reis, R.E Revision of the neotropical catfish genus Hoplosternum (Ostariophysi: Siluriformes: Callichthyidae), with the description of two new genera and three new species. Ichthyological Explorations of Freshwaters 7:

16 466 Southeastern Naturalist Vol. 3, No. 3 Schmitz, D.C., B.V. Nelson, L.E. Nall, and J.D. Schardt Exotic plants in Florida: A historical perspective and review of the present aquatic plant regulation program. Pp , In T.D. Center, R.F. Doren, R.L. Hofstetter, R.L. Meyers, and L.D. Whiteaker (Eds.). Proceedings of the symposium on exotic plant pests, 2 4 November 1988, Miami, Florida. Technical Report NPS/NREVER/NRTR-91/06. U.S. Department of the Interior, National Park Service, Washington, DC. Singh, T.B The biology of the cascadura Hoplosternum littorale Hancock 1834 with reference to its reproductive biology and population dynamics. Unpublished Ph.D. Dissertation. University of the West Indies, Trinidad. 298 pp.