Home Archived October 29, 2018

South Florida Information Access (SOFIA)

publications > reports > inventory of freshwater fishes of the big cypress national preserve > results and discussion

>Results & Discussion
Topics of Concern
Literature Cited
Appendix 1
Appendix 2
PDF Version

Results and Discussion

map of Big Cypress National Preserve fish Inventory and Monitoring sampling locations
Figure 4: BCNP fish I & M sampling locations, 10/02-10/03, by habitat type. [larger image]
During the Big Cypress inventory work, we sampled 456 sites throughout the preserve (Figure 4). The distribution of those sites among habitats is shown in Table 3. Canals were the most widely sampled habitat, but the effort involved in many of those samples was lower than for natural habitats, because sampling in canals often consisted of a single hoop-net set or electrofishing transect.

Frequency of sampling in each habitat type was based roughly upon a combination of geographic extent within BCNP and hydroperiod. Coastal marshes and tidally influenced canals were sampled only when salinities were less than one part per thousand. Those conditions prevailed in most locations from June to October of each year.

We documented 64 fish species from freshwater habitats within BCNP (Table 4). Appendix 1 contains distribution maps for each of these species

Distribution by Habitat Type

Table 5 shows the occurrence or absence of each species from samples collected in each of the major habitats previously described. Canals held by far the most diverse assemblage; 62 of the 64 species documented in the preserve occurred within this habitat. Although this result is partly an artifact of the greater sampling effort in canals relative to other habitats, it also reflects the presence of euryhaline species penetrating inland from the Gulf of Mexico. Many of these fishes are probably temporary residents and do not appear to enter wetland habitats other than by swimming in channels in coastal marshes. Species in this category included the various needlefish (Strongylura spp.), mojarras (Eugerres and Eucinostomus), and gobies (Bathygobius, Lophogobius, and Microgobius), as well as hardhead catfish (Arius felis), sheepshead (Archosargus probatocephalus), crevalle jack (Caranx hippos), snook (Centropomus undecimalis), sharksucker (Echeneis naucrates), pinfish (Lagodon rhomboides), gray snapper (Lutjanus griseus) and tarpon (Megalops atlanticus). Also, several fishes found in canal samples were freshwater species that prefer deep-water habitats and were also unlikely to be taken in the shallow wetlands. Those fishes included brown bullhead (Ameiurus nebulosus), gizzard shad (Dorosoma cepedianum), and black crappie (Pomoxis nigromaculatus).

Sloughs, ponds, and rivers are the only naturally occurring habitats in the preserve that retain water throughout the seasonal dry period. There are very few natural rivers draining BCNP, although Turner River drains the southwestern portion and allows passage inland to euryhaline wanderers in the same manner as canals. Lakes are also uncommon within BCNP; however, they provide refuge to large euryhaline species if they have a hydrologic connection to the canal system. The best examples of this are the substantial populations of tarpon (Megalops atlanticus) and snook (Centropomus undecimalis) inhabiting Deep Lake, a flooded sinkhole adjacent to the Barron Collier Canal. It appears that individuals of both species permanently reside in the lake, but it is unclear how long they have resided there, or how much interchange there is with populations along the coast.

We collected a variety of euryhaline fish species in the coastal marshes, although these were mainly small species or juveniles. Euryhaline species present in that habitat included clown goby (Microgobius gulosus), gulf killifish (Fundulus grandis), rainwater killifish (Lucania parva), and striped mullet (Mugil cephalus). During the wet season, many of the freshwater species found in the preserve moved into these marshes.

The remaining wetland habitats shared very similar fish faunas. Of the 34 species found in these habitats, 19 species were common to all. It is likely that several other species were present in all habitats as well, but we did not capture them. There seemed to be a general pattern among small fish species to utilize shallow-water areas, such as herbaceous prairies and cypress prairies when those habitats were inundated.

Methods Testing

The heterogeneity of aquatic habitat structure meant that no single sampling method was suitable for providing a complete picture of the fish-community in a given habitat. During the inventory work, we were not permitted by BCNP to use toxicants such as rotenone in enclosed areas to establish baseline-community profiles. This hampered efforts to determine the efficiency of each method in sampling fish populations within each habitat. The use of drop traps to explicitly determine fish population densities during future community-dynamics research will provide much-needed information in this regard. However, the data collected during the inventory work may be used to draw some general conclusions about the effectiveness of various techniques. Also, the efficiency tests performed for methods such as throw traps and drop traps, done in Everglades habitats structured similarly to the preserve habitats, are probably similarly effective in trapping fish and invertebrates in BCNP.

Minnow and Breder traps for small fishes were used most extensively in wetland environments; we took 37 species using them (Table 6). Of the 40 species taken from wetland habitats (Table 5), 35 were obtained in trap sets. The only documented wetland species not taken by these traps were the brook silverside (Labidesthes sicculus), brown hoplo (Hoplosternum littorale), clown goby (Microgobius gulosus), golden shiner (Notemigonus crysoleucas), and striped mullet (Mugil cephalus). These were taken using a combination of dip netting, gill nets, and electrofishing. Table 7 shows the distribution of traps set by habitat and season. Although trap data may not necessarily provide good estimates of relative or absolute abundances, the ability of the traps to capture such a broad cross-section of species, combined with their ease of transportation and use, suggest they are the best available method for providing presence/absence data on fish species in BCNP wetlands, especially if supplemented with opportunistic dip netting.

Electrofishing proved to be the most effective method for surveying fishes in open-water areas. Forty-four species were taken during electrofishing surveys (Table 8). Demersal species such as catfish were difficult to obtain using this technique, thus any future canal sampling should also include the use of baited hoop nets. Gill nets were also useful in canals and other deep-water areas, but the large population of alligators in BCNP made entanglements a serious problem.

Seasonal and Habitat Patterns

For each sample, the total number of species (S), total CPUE, and Shannon-Weiner Diversity were calculated. The values of these parameters, averaged by habitat type and season, are shown in Table 9. Comparisons were then made for each of these parameters between seasons within each habitat type using a two-tailed t-test for between samples of unequal variance. The results of these tests are shown in Table 10.

Average aggregate CPUEs were greatest for the dry season in all habitats except natural deep-water areas. Insufficient numbers of wet-season trap sets were performed in that habitat to allow meaningful comparisons between seasons. Increased catch most likely reflected the concentration of fishes in remaining wetted areas during the seasonal dry-down. This effect was statistically significant in cypress forest (dry season mean CPUE =14.4, wet season mean CPUE =2.56, P = 1.905 x 10-6) and freshwater marsh (dry season mean CPUE =23.47, wet season mean CPUE =7.67, P = 0.033) habitats, both of which were adjacent to shorter-hydroperiod habitats, such as herbaceous or cypress prairies. Cypress-forest samples also showed a significant increase in the number of species captured during the dry season (mean total dry season species =7.68, mean total wet season species =3.91, P = 0.0002). These results reflect the movement of fishes from short-hydroperiod areas to long-hydroperiod refuges during the dry season. This is similar to information reported by Carlson and Duever (1978) for flagfish (Jordanella floridae) and least killifish (Heterandria formosa) populations during the annual dry-down of Corkscrew Swamp. However, swamp-forest samples showed no significant differences between seasons. This may be due to the relatively small number of samples taken from this habitat and the large variability among these samples throughout the year. Additionally, swamp-forest habitats typically occurred within areas of cypress forest and, therefore, were less likely to be directly adjacent to short-hydroperiod wetlands, the likely source of the increased number of fishes seen in cypress and freshwater-marsh habitats.

To examine differences in community composition among habitat types, the per-species CPUE data for all samples were compared using the ANOSIM routine of PRIMER-E nonparametric statistical analysis software. ANOSIM examines a dataset to determine the similarity between samples within predefined strata (in this case, within habitats) and then uses a randomization procedure to compare between strata (Clarke and Warwick, 2001). The resultant statistic is a probability value indicating the likelihood that observed differences between strata could occur by chance. Table 11 shows the ANOSIM results for comparisons between BCNP habitats. We made no distinction between wet and dry seasons in these calculations.

The assemblage of fish sampled with small-fish traps in canals was significantly different from that found in all other habitats except for naturally occurring deep-water areas. This probably reflects the presence of euryhaline species in the canal dataset, because those species also occurred in the natural deep-water areas (specifically in samples from Turner River). Euryhaline species also occurred in coastal marsh samples, which was the next most similar habitat. The canal fish assemblage also was likely influenced by the relatively high abundance of exotic species taken there.

Coastal marsh samples were most similar to those from canals (as mentioned previously). The coastal marsh, however, exhibited significant differences in fish-community structure from all other habitats, and reflects the influence of estuarine species. Naturally occurring deep-water areas (sloughs/ponds/rivers) intergrade between canals and long-hydroperiod freshwater wetlands. We found no significant differences in community structure between this habitat and the canal, cypress forest, or swamp forest habitats. There were significant differences in structure, however, between this and other habitats.

No significant differences in community structure were found between cypress-forest and the other freshwater-wetland habitats except herbaceous prairies. Cypress prairies and swamp forest both routinely intergrade with this habitat, so the lack of differences in community structure is not surprising. Cypress forest and freshwater-marsh habitats have similar hydroperiods, but we cannot provide a simple explanation for their observed similarity in community structure.

Swamp-forest results showed a very distinct separation from short-hydroperiod wetlands. This community was most similar to cypress forest, with which it commonly intergrades. Freshwater-marsh samples were difficult to interpret. The community was significantly different from herbaceous prairies, despite the fact that they commonly border each other. Marsh samples were not significantly different from cypress or swamp forest samples, perhaps owing to the similar hydroperiods of the habitats, but they were also not significantly different from shorter-hydroperiod cypress forest.

Cypress prairie fish-community structure was not significantly different from those of cypress forest or herbaceous prairie habitats. Cypress prairie often forms a continuum between these habitats, so this result is not surprising. As noted earlier, there was also no significant difference between this habitat and freshwater marsh. The herbaceous prairie community was distinct from all others except for cypress prairie, which operates under a similar hydrologic regime and is often adjacent to this habitat. The differences observed between this habitat and freshwater marshes were unexpected, as noted earlier.

Overall, communities with similar hydrology tended to have similar community composition, as did those that were contiguous. In general, geographic separation between habitats, as well as hydroperiod differences, were reflected by differences in community structure.

Community-Dynamics Research Results

Site selection was made during extensive reconnaissance trips. The first scheduled sampling for this monitoring program occurred during April-May 2004, but the sites were completely dry due to a strong and prolonged dry period. Therefore, sampling of the fish monitoring stations occurred for the first time in July 2004. All stations had been completely dry until two weeks prior to sampling because wet season conditions were delayed. Therefore, the results represent very early recolonization of recently inundated habitats. Water levels were sufficient to allow sampling at the L-28 and Raccoon Point sites. The Bear Island sites were dry through the end of July and, therefore, could not be sampled.

Total catch from the monitoring sites is reported in Table 12. The minnow trap arrays are designed to concentrate catch from a large area and, therefore, were more effective given the low densities present at that point in the annual hydrologic cycle. Gambusia holbrooki and Jordanella floridae were the most commonly captured species.

The L-28 sampling stations are located in a large, semi-continuous slough habitat in relatively close proximity to several canals. It is likely that there were small, inundated refuge areas within this habitat in which fishes survived. Although the L-28 canal has levees on both sides that hydrologically isolate it from the surrounding wetlands, the canal bordering Interstate 75 is close to the third replicate plot and may have represented a potential refuge area.

Raccoon Point is an area of large cypress domes that are hydrologically separated during times of low water. There were no discernable refuge areas nearby, and sampling plots were still separated by areas of dry land at the time of sampling. It is therefore not surprising that the fish catch at Raccoon Point was lower than at the L-28 site.

Depending on hydrologic conditions, fish population levels can vary greatly from one July to the next. Therefore, it is possible that future sampling expeditions will find a greater number and diversity of species than were found during this particularly dry year. The drop traps, in particular, should produce more meaningful results at higher population levels.

Non-Indigenous Species

We took nine species of introduced fishes in the preserve, most in the family Cichlidae (the cichlids). These included the oscar (Astronotus ocellatus), black acara (Cichlasoma bimaculatum), Mayan cichlid (Cichlasoma urophthalmus), jewel cichlid (Hemichromis letourneauxi), blue tilapia (Oreochromis aureus), and spotted tilapia (Tilapia mariae). Other families were represented by only one species, including the labyrinth catfishes (Clariidae: Clarias batrachus), livebearers (Poeciliidae: Belonesox belizanus) and the armoured catfish (Callichthyidae: Hoplosternum littorale). Hemichromis and Hoplosternum had not been previously recorded in BCNP and appear to be in the process of colonizing much of southern Florida. Hoplosternum was rarely taken during the inventory sampling, although several juveniles appeared during the July 2004 monitoring. Hemichromis were encountered much more frequently during the 2004 inventory than during previous efforts. In the cypress forest near Pinecrest (in the southeast part of the preserve), Hemichromis were a numerically dominant segment of the wet-season fish catch in 2004. This rapid spike in population has been seen in other species that have recently become established in south Florida wetlands, and is typically followed by a decline to some lower population level (Trexler et al., 2001).

Overall, the most widely distributed exotic species in the preserve were pike killifish, Mayan cichlids, black acara, and spotted tilapia. Oscar, walking catfish, and blue tilapia were captured almost exclusively in or adjacent to canals or other deep-water habitats, although their distribution in the preserve may be more widespread. Oscars, Mayan cichlids, black acaras, and walking catfish were locally abundant in canals during the dry season.

| Disclaimer | Privacy Statement | Accessibility |

U.S. Department of the Interior, U.S. Geological Survey
This page is: http://sofia.usgs.gov/publications/reports/fish_bcnp/results.html
Comments and suggestions? Contact: Heather Henkel - Webmaster
Last updated: 04 September, 2013 @ 02:04 PM(TJE)