. -Ii --pl-L - I. 1=i.'IIff" 4.;11 1 1, ' 2*- c.-~i r .~ 7 1.1 1 -11 1 -¢r* 2 ' *1 ------ .-1- ----- - -r-13 - . , --- , 14 of the FLORIDA MUSEUM OF NATURAL HISTORY WILDLIFE IN SOUTHERN EVERGLADES WETLANDS INVADED BY MELALEUCA (Metaleuca quinquenervia) Nancy K. O'Hare and George H. Dalrymple Volume 41 No. 1, pM 1-68 1997 94 1 ' 'r T.,r, 4 , Uri ", '' ,, ' 7 - '., ': , 04', ' UNIVERSITY OF FLORIDA GAINESVILLE Numbers of the BULLETIN OF THE FLORIDA MUSEUM OF NATURAL HISTORY are published at irregular intervals. Volumes contain about 300 pages and are not necessarily completed in any one calendar year. JOHN F. EISENBERG, ED/TOR RICHARD FRANZ, CO-EDITOR RHODA J. BRYANT, AONAGNG EDITOR Communications concerning purchase or exchange of the publications and all manuscripts should be addressed to: Managing Editor, Bulletin; Florida Museum of Natural History; University of Florida; P. O. Box 117800, Gainesville FL 32611-7800; U.S.A This journal is printed on recycled paper. ISSN: 0071-6154 CODEN: BF 5BA3 Publication date: December 31, 1997 Price: $ 6.50 WILDLIFE IN SOUTHERN EVERGLADES WETLANDS INVADED BY MELALEUCA (Me/aleuca quinquenervia) Nancy K. O'Hare and George H. Dalqmplet ABSTRACT In the Everglades region of southeastern Florida invasion of graminoid/herh=„.ius wetlands by the invasive, non-native tree melateuca (Melaleuca quinquenervia) results in a closed=canopy forested wetland, with a sparse understory. Intermediate stages in this transformation include a savannah with scattered mature melaleuca trees, and mature dense melaleuca heads surrounded by areas with moderate to low levels of melaleuca. Intermediate levels of melaleuca invasion have not received any attention and were the rationale for our study. Wildlife was surveyed monthly for two years to detennine species richness and abundance in wetlands with different melaleuca coverages. Wildlife included all vertebrate classes, as well as selected macro-invertebrates such as crayfish (Procambants allem) and grass shrimp (Pateomonetus paludoms). Species richness was highest in areas with moderate metaleuca coverage. Higher species richness is typical of sites with greater vegetative structural diversity, i.e., as in the savannah stage of invasion, as well as areas in an early stage of disturbance. The higher species richness was primarily the result of an increased number of migratory, upland birds. Many of these transient and winter-resident birds occurred at much lower abundances than in native forested habitats such as cypress swamps (Taxodium distichum), tropical hardwood hammocks, and pine (P,nus elhomi var. densa) rocklands. In contrast to the birds number of species and the abundance of herpetofauna varied little across the melaleuca gradient There was no shift in species composition from wetland to upland species as the melaleuca coverage increased. The number of fish species was similar across the melaleuca gradient Unlike the herptiles, fishes were less abundant in the closed-canopy melaleuca forests indicating poorer habitat quality. Complex patterns of hydrology and gapping in the forest canopy due to wind storms and fires permitted light penetration and the persistence of productive pockets of aquatic life even within dense stands ofmelaleuca. The mosaic of areas with low to moderate infestations of melaleuca surrounding mature dense melateuca stands allowed higher numbers of individuals and species to persist in or seasonally use mature dense melaleuca stands. This interspersion of habitats resulted in stands of metaleuca with ecotonal edges that provided marginal habitat for species characteristic of natural conununities. Higher degree of interspersion (more edge) may also mean that the natural areas experience higher exposure to melaleuca seed source, which may result in a faster rate of spread of melaleuca. The results demonstrated that animal populations persisted in areas with disturbed vegetation, as long as critical abiotic factors (in this case hydrology) remained in operation. Areas with moderate levels of melaleuca retained species composition and productivity typical of the natural wetland community. The dominant characteristic of the faunal shifts along the gradient of increasing melaleuca coverage was increased numbers of upland, arboreal, and/or forest species, not the loss of wetland species. Regional 'Cunent address: Everglades Research Group, Inc- 35250 SW 212 Avenue, Florida City, Flofida 33034-4016 C)'Hare, N. K, and G. H. Dalrymple. 1997. Wildlife in southern Everglades wetlands invaded by melaleuca (Melaleuca quinquenervia). Bull. Florida Mus. Nat. Hist 41(1):1-68. 2 BULLETIN FLORIDA MUSEUM NATURAL HISTORY VOL 41(1) permitting and natural resource agencies should recognize that lands with moderate levels of melaleuca may retain significant habitat quality. Restoration of such lands will demonstrate higher levels of success if the method used for melateuca removal allows for retention of the in situ wildlife community. RESUMEN La invasiin de los humeclala graminoides/herbaceos en la regi6n de los Everglades del sureste de la Florida por el arbol no nativo melaleuca (Melaleuca quinquenervia) results en un humedal forestado de dosel cerrado y de solobosque ralo. Los estados intermedios de esta transformaci6n incluyen una savana con melaleucas maduras y dispersas y bosquetes maduros y densos de melaleuca rodeados de areas con moderados a bajos nRmeros de melaleucal Estos estados intermedios han sido poco estudiados, y por edo, fueron el foco de nuestro egtudio. Con el objeto de determinar el numero de especies y su abundancia en humedales con diferentes coberturas de melaleuca, se realizaron reconocimientos mensuales de vida silvestre durante dos ahos La vida silvestre estudiada incluy6 todas las clases de vertebrados, asi como algunos invertebrados tales como dos especies de camar6n (Procambarus alleni y Paleomonetus paludosus). El mayor numero de especies se encontrd en *reas con una cubierta moderada de melaleuca. Un mayor nOmero de especies es tipico de dreas con una mayor diversidad estructural vegetal como por ejemplo, en el estado de invasi6n tipo savana, asf como lambidn en keas con un estadio de perturbaci6n mAs temprana. El mayor numero de especies fue i' ' 'el resultado de un mayor numero de ava migratorias de tierras mas altas. Muchas de estas epecies de aves en Wnsito o residentes invernales se encontraron en abundancias mucho menores que en bosques natives, como pantanos de ciprds (Taxodium distichum), bosquetes de madera dura y bosques de pino (Pinus elliottii var. densa). En contraste a las aves, el numero de especies y la abundancia de anfibios y reptiles vari6 poco a travas del gradiente de melateuca. No hui,o cambio en la composici6n de especies a medida que la cobertura de melaleuca aument6. El numero de especies de peces tambi6n fu6 similar a medida que la cobertura de melaleuca aument6. A diferencia de los anfibios y reptiles, los peces fueron menos abundantes en bosqtics de melaleuca de dosel cerrado, indicando una calidad de habitat inds pobre. La presencia de claros en el bosque producidos por tonnentas de viento y fuegos, asi corno la compleja hidrologia, permitieron la penetracidn de luz y la persistencia de bolsones de productividad de vida acukica, incluso dentrobosques demos de melaleuca. El mosaiw de Areas con infestacioncs de melaleuca moderada a baja rodeando bosquetcs maduros y dmsos de melaleuca permitieron la persistencia o uso estacional en 6stos Oltimos de un numero mayor de individuos y esl,ecics. El entrelazamiento de habitats resultd en bosqtleta de melaleuca con bordes ecotonales, los cuales proveyeron habitats marginales para especies caracteristicas de comunidades naturales. Un mayor nivel de entrelazamiento (mas bordes) tambian significa que las dreas naturales tienen una mayor exposici6n a las fuentes de semillas de metaleuca, lo cual puede resultar en una tasa de avance mayor para la melaleuca. Los resuitados demostraron que las poblaciones animates persistieron en *eas con vegetaci6n alterada, siempre y cuando factores abi6ticos criticos (en este caso hidrologia) continden operando. Las areas con una cobertura moderada de melaleuca mantuvieron la . ' ~' de especies yla productividad lipica de la comuni€laci natural del humedal. La caracteristica dominantede los cambios fhunisticos a lo large del gradiente de melaleuca fue el incremento del numero de especies de tierras altas, art,6reas. o de especies del bosque; no la p6rdida de especies de humedal. Las agencias que administran recumos naturales deben reconocer que beas con niveles moderados de melaleuca pueden retener niveles significtivos de calidad de habitaL La restauraci6n de estas dreas puede resultar mas exitosa si el m6todo usado para remover melaleuca permite la retencitn de la comunidad silvestre presente en dicha drea. TABLE OF CONTENTS Introduction 3 Acknowledgmrn'. 4 Description of Study Area and Cover Type= 4 Sampling Methods 6 Hydrological Assessment 8 Statistical Methr~ 8 Results and Discussion 11 DALRYMPLE & DALRYMPLE: MELALEUCA IN EVERGLADES WETLANDS 3 Literature Cited Fir ,"9 R*$2Tabl* Appendix INTRODUCTION In the Everglades region of southern Florida invasion of an open canopy, graminoid/herbaceous wetland by the non-native pest tree melaleuca (Melaleuca quinquenervia) results in a closed-canopy forested wetland, with a sparse understory. Intermediate stages in this transformation may include a savannah with scattered mature melaleuca trees and mature dense melaleuca heads surrounded by areas with moderate to low levels of melateuca. Previous surveys of wildlife in metaleuca-infested areas have focused on either a few species of mammals (Mazzotti et al. 1981; Sowder and Woodall 1985) or surveyed only dense melaleuca stands (Schortemeyer et al. 1981; Repenning 1986). Each of these studies was of short duration (few months). Therefore, relatively little is known regarding the use of melaleuca-invaded wetlands by native wildlife. Disturbance of natural communities typically results in an increase in species richness as "weed" species, non-native, migratory and/or species uncommon to the natural community increase in numbers (Odum 1983). Furthermore, areas with higher vegetative structural diversity, such as the intermediate stages of melaleuca invasion of graminoid wetlands, are likely to have ~higher species diversity compared to areas with lower vegetative structural diversity (c.f. Cody 1985a). Therefore, the number of species (species richness) and the number of individuals (species abundance) are not, by themselves, a good measure ofthe environmental value of a habitat (Van Horne 1983). Which species are using a habitat and the manner in which they use the habitat (foraging, breeding) are more important to final evaluation of habitat quality (Stauffer and Best 1980; Keller et al. 1993). A fair analysis of habitat quality of disturbed areas should evaluate the types of species (e.g., wetland versus upland animals, native versus non-native), as well as their abundances. Our goal in this study was to determine species richness and relative abundance along the single gradient of melaleuca coverage, without presuming to explain between-taxa differences, or variation within a single cover type. Wildlife was broadly defined to include selected macro-invertebrates and all vertebrates. Some of these groups are not traditionally included in wildlife assessments. However, they were included in this study since the abundance of these animals indicates the ability of a habitat to support higher trophic level animals, such as wading birds, alligators, snakes, and mammals. 4 BULLETIN FLORIDA MUSEUM NATURAL HISTORY VOL 41(1) ACKNOWLEDGMENTS We thank Kenneth L Krsyko for his dedication to the field work for two wet years. Joseph A Wasile,%,ki. Carlos Pages, and Doug Barker also assisted in the field during various times. We also thank Sue M. Alspach, Frank S. Bernardino, and Jean Evoy of the Dade County Depalnent of Environmental Resources M t and the members of the South Florida Limestone Mining Coalition. Funding was provided by South Florida Lin,estone Mining Coalition. U. S. Army Corps of Engineers, South Florida Water Management District and Metro-Dade Water and Sewer Depaitment. Project managemen< technical support. and publication costs were provided by Dade County Department of Enviromnental Resources Management DESCRIPTION OF STUDY AREA AND COVER TYPES The study was performed in northwest Dade County in a 19,400 ha region known as the Lake Belt Study Area (LBSA). The area is bounded by the Dade- Broward County line on the north, the Homestcad Extension of Florida's Turnpike on the east Tamiami Trail (US 41) on the south, and Krome Avenue on the west (Fig. 1). The area is the single, largest tract of land in the South Florida Water Management District's proposed East Coast Buffer/Water Preserve Areas between the urban areas and the remaining Everglades. The western one-third of the stu* (between the Dade-Broward Levee and Krome Avenue) is commonly referred to as the Pennsuco wetlands or Pennsuco Everglades. The classic vegetation survey by Davis (1943) characterized most of the area as "saw-grass marshes (medium dense to sparse)," with the southeastern corner characterized as "saw-grass marshes (with wax myrtle thickets)." Reconstruction of pre-drainage conditions by Everglades National Park, the Army Corps of Engineers, and the South Florida Water Management District include most of the LBSA as part of the long hydro-period marsh of northeastern Shark River Slough (also see Fennema et al. 1994). Recent hydrological records demonstrate that the Pennsuco wetlands (west of the Dade-Broward Levee) are still flooded for more than six months a year under "normal rainfall" (e.g., 1986; Davis et al. 1994). Soils in the region are classified as muck or peat soils, with depths up to 1 m (EAS Engineering, Inc. 1995). A map of existing cover types in the LBSA wa generated from 1992 1:300 aerial photographs (Fig. 2; EAS Engineering, Inc. 1995). The region included approximately 3000 ha of sawgrass marshes with little to no invasion by melaleuca, 3300 ha of low to moderate coverage by melaleuca (10% to 75% melaleuca) and 7000 ha with greater than 75% coverage by metaleuca. The remaining 6100 ha were composed of lakes, littoral zones, agricultural lands, canals, levees, correctional facilities, electrical power facilities, and power line right-of-way (EAS Engineering, Inc. 1995). There was a geographical gradient in the density of melaleuca within the study area. Areas with the highest coverage by melateuca tended to be located in the eastern two-thirds of the region, while areas with lower melaleuca coverage were located in the western one-third (Pennsuco Everglades). Many of the areas DALRYMPLE & DALRYMPLE: MELALEUCA IN EVERGLADES WETLANDS 5 with highest melaleuca coverage were adjacent to developed lands or near structures that alter local hydrology. In the eastern one-third of the study area land uses included a municipal well field and rock-mining. Both of these uses affected adjacent lands by altering hydro-period albeit the type of effects differed. The municipal well field had little effect on the ground surface topography. However, associated canals and the effect of ground water pumping altered local hydrology. The manner in which hydrology was altered was not predictable based upon seasonal weather patterns, but rather was determined by water supply needs. Therefore, the region may have standing water during the traditional "dry" season of southern Florida. In contrast, rock-mining substantially altered surface topography, creating permanent aquatic habitats up to 20 m deep. While the lakes draw water from surrounding areas, shortening their hydro-period, annual hydrological patterns fluctuated with normal seasonality of wet-dry periods. Five cover types were designated for sampling based upon percent coverage by melateuca. The following abbreviations were used in the text tables, and figures. 1) DMM: 75-100% mature dense melaleuca coverage; DBH of trees>8 cm; stem density of 5000/ha (Hofstetter, unpubl. as cited by Hofstetter 1991) 2) SDM: 75-100% sapling dense melaleuca coverage; DBH of trees<8 cm; stem density of 250,000/ha (Alexander and Hofstetter 1975) 3) P75: 50-75% melaleuca coverage 4) P50: 10-50% melaleuca coverage 5) MAR (Marsh): 0-10% melaleuca coverage The detailed vegetation map referenced above was not available when site selection for the Wildlife Studies began. Potential study sites were identified from the vegetation map in Larsen (1992) and 1992 aerial photographs. Actual site selection was determined by ground-truthing. Cover types with intermediate levels of metaleuca coverage (10%-50% and 50%-75%) were the most difficult to delineate on the ground and also occurred in smaller, less discrete parcels relative to the other three cover types. The spatial distribution of melaleuca in these areas usually consisted of a heterogenous mix of melaleuca heads, and savannahs. Since the minimum extent for cover type designation in the vegetation mapping was 0.40 ha (one acre), sites selected for wildlife sampling, were a minimum of 0.40 ha of homogenous melaleuca coverage, embedded in a matrix that we judged to be of the same cover type based upon ground-truthing. For each of the five cover types, ten sites were selected (50 sites total). Each site selected for sampling had to be readily accessible on foot from an existing grade (e.g., up to 1 km from a levee, or right of way). Areas with melateuca seemed to be related to developed areas or areas with altered hydrology. Approximately 75% of the area available for sampling (excluding cover types not 6 BULLETIN FLORIDA MUSEUM NATURAL HISTORY VOL 41(1) sampled. such as lakes or agriculture areas) was within 1 km of some type of human disturbance (e.g., a primary or secondary road, existing grade, building, canal or lake). Approximately 20% of the available area that was greater than 1 km from a grade was MAR. Thus only 5% of available area was more than 1 km from a disturbance, and it was distributed unequally among four cover types. Prim=g and secondary roads were located only on the boundaries of the study area. Vehicular travel within the study area was confined to narrow gravel grades. Access to these grades was restricted by locked gates at all entry points. The major north-south grade was the Florida Power & Light (PPL) powerline right-of-way. Portions of the FPL right-of-way were flooded during the wet season. SAMPL]NG METHODS Drift Fence Amys Drift fence sampling required intensive site preparation and permanent installation of the trapping arrays (see below). Therefore, three sites for each cover type were repeatedly sampled each month from January 1994 through December 1995. Preliminary surveys of the entire region indicated that a hydrological gradient might exist from north-south. Hydrological data to either support or refute these field observations were unavailable. Since sample sizes were low (three sites per cover type), these sites were located in the northern one-third of the study area to minimize variation in factors other than melaleuca coverage (e. g., hydrology) as a precautionary measure. Drift fence amys were checked four days per month, generally, every other day over an eight day period beginning the second week of each month. All 15 arrays were checked on the same days. In studies of the amphibians and reptiles of the Everglades National Park, drift fences designed to trap amphibians and reptiles also regularly trapped high numbers of aquatic macro-invertebrates (e.g., crayfish, Procambarus aUeni; grass shrimp, Paleomonetus paludosus; and fishes (Dalrymple 1988; G.H. Dalrymple and F.S. Bernar(lino, unpubl. data; Dal~ymple 1994). Therefore, drift fence trapping in this study was used as a sampling method for all aquatic, semi-aquatic and terrestrial vertebrate animals (including fishes), as well as selected aquatic macro-invertebrates. Drift fences were constructed of shade or ground cloth. Each array had four 15-m-long by 1-m-high arms arranged as a cross [+1, with a total of four funnel traps per amy. Traps and funnels were constructed of 1/8" gauge (approximately 3 mm) galvanized hardware cloth, with two funnels at one end of each trap. One trap was placed at the end of each arm of the array, so that one funnel rested on each side of the fence (as done by Dalrymple 1988). Pitfall traps were not feasible since most sites were flooded six to nine months each year. Amys were maintained so that the fencing remained upright and no gaps developed between the fencing material and the ground. Funnel traps were DALRYMPLE & DALRYMPLE: MELALEUCA IN EVERGLADES WETLANDS 7 repaired or replaced as needed. When the traps were not being checked, they were removed from the end of the fence and the funnels were blocked to prevent animals from entering the traps. Standing water did not preclude trapping. However, when traps were completely underwater, the time period between trap-check days was modified to minimize mortality of amphibians and reptiles, i.e., traps were checked four consecutive days rather than every other day for eight days. The number of check days remained the same (4 days per month). Trap rates were calculated using the number of days the arrays were open (array days), not the number of times the traps were checked. Bird Strip Transects Bird transects did not require site preparation and, therefore, allowed sampling to occur in a random subset of 3 of the 10 sites in each cover type each month. This procedure permitted a wider range of sites to be sampled. Transects were a fixed length of 100 m. The width of each transect was determined by the farthest distance to a bird observed during the transect. If the bird was flying overhead or could not be positively identified, it was not recorded. Sampling of the 15 sites occurred over a 2 -3 day period during the third week of each month. All data were collected between sunrise and 11 a.m. The order in which cover type sites were sampled was randomly chosen each month. Sites were sampled regardless of standing water conditions. Strip transects for birds in this study were designed to focus on the birds that have limited daily cruising radii and, therefore, were most likely to reflect habitat preferences based on vegetative cover rather than hydrology. Perching birds (blackbirds, shrikes, warblers, cardinals), other land birds (doves, woodpeckers), some smaller wading birds (snipe, rails), and some birds of prey usually are studied to evaluate between habitat differences in vegetative cover (Stauffer and Best 1980). Such surveys also allow assessment of habitat use by migratory and/or transient birds versus resident breeders (Keller et at. 1993). Mammal Surveys Mammals were surveyed using Sherman live traps and scent and bait stations on a quarterly basis, with one replicate per cover type. Oats were used to bait 30 Sherman live traps, 15 Sigmodon-, and 15 Peromyscus-sized traps, laid out in a grid, and checked for three consecutive nights. In addition, one scent (mammal urine) and one bait station (oily tuna pet food) were also checked the same three consecutive nights. Sampling generally occurred the first week of the second month in each quarter (February, May, August, and October). In some quarters, sampling was either delayed until later in the quarter or simply not feasible due to high standing water levels. 8 BULLETIN FLORIDA MUSEUM NATURAL HISTORY VOL 41(1) Incidental Observations To generate a complete species list for the LBSA, incidental observations of species within the five defined cover types as well as species noted along roadways were recorded. This information is presented in Appendix I. HYDROLOGICAL ASSESSMENT As a general rule, hydro-pattern (timing, depth, and duration) is a strong determinant of wetland species diversity and abundance (Mitsch and Gosselink 1986; Campbell and Christman 1982; Dalrymple 1988). Therefore, evaluation of biological resources in wetlands must consider hydrological conditions. As a preliminary assessment of gross hydrological patterns, two data sets were gathered through the Water Resources Section of DERM. The first data set was the 11-year period (1985 to 1995) of ground water levels measured at USGS wells in the LBSA. Two gages were randomly selected for more detailed analyses. One gage was located west of the Dade-Broward Levee (G-975) and the other east of the Dade-Broward Levee (G-972; see Fig. 2). The second *t= set was the 1994 and 1995 average monthly ground water levels of the seven USGS gages located in the LBSA (G-594, G-968, G-972, G-975, G-976, G-1488, and G-3253). STATIST[CAL METHODS Statistical analyses followed standard procedures outlined in Zar (1996), Sokal and Rohlf (1995), Gauche (1982), and Krebs (1989). All nnalyses were performed using STATISTICA 5.1 (StatSoft, 1995). Drift Fences and Bird Transects For the drift fencing data on macro-invertebrates, fishes, and amphibians and reptiles, the two year cumulative numbers (e. g., Sokal and Rohlf 1995; Zar 1996) from each of the 15 sites (3 replicates in the 5 cover types) were analyzed by ANOVA. Some of the raw data sets did not follow a normal distribution and neither log nor square root transformations (Krebs 1989) resulted in a normal distribution. Therefore, in all cases, the data were analyzed by non-parametric Knmk:al-Wallis ANOVA. For the bird transects, the two year cumulative numbers for each cover type were analyzed. Cumulative data generated by this sampling protocol could not be analyzed by ANOVA, because site-specific cumulative measures were not available. While ANOVA of each of the 24 monthly samples for each method was possible, most had such low sample sizes as to be or little or no value, and did not address the larger issue of general patterns of habitat use. DALRYMPLE & DALRYMPLE: MELALEUCA IN EVERGLADES WETLANDS 9 The above limitations are not critical to statistical analyses in ecological sciences. "In general, the framework of hypothesis testing has been largely overused by scientists..., especially in the context of environmental decision making" (Steidl et al., 1997: 278). Simple statistical tests for average differences between cover types in numbers of individuals, or numbers of species reveal a limited amount about the ecological nature of cover type differences (Krebs 1989). They are useful for recognizing gross differences in species richness or diversity, but say little about the species composition of the cover types. Therefore, multivariate techniques that simultaneously consider each species' contributions to cover type differences (and vice versa) were used (Gauche 1982). Data sets collected at standard sites, such as drift fence data for fishes, amphibians and reptiles, or macro-invertebrates, were analyzed using the multivariate techniques of cluster analysis, factor analysis and/or multidimensional scaling. These analyses have the same three replicates for each cover type sampled each month, permit the monthly data to be accumulated for tests of total numbers, averages, or medians (e. g., see Sokal and Rohlf 1995, Box 9.8), and allow us to see more of the variation among sites within the same cover type. The plots of these analyses in the figures have three replicates for the five cover types, entered separately and plotted separately. These data sets had enough replicates to permit factor analyses as well as cluster analyses and multidimensional scaling. For example, the herptile drift fence data has a matrix of 34 rows (species) by 15 columns (locations), i.e., 34 x 15 matrix. All multivariate matrices were derived from the raw data sets to include the effects of differences in absolute sample sizes. Cluster analyses were done using the unweighted pair-group average (UPGMA) amalgamation method ofjoining groups (Krebs 1989). The joining was done on a distance matrix generated as the subtraction of each Pearson's product moment correlation coefficient from unity (1.0, i.e., 1-r), to generate the distances. If for example two cover type sites or species had a correlation coefficient of 0.91, then their distance is 1.0-0.91, or 0.09 (i.e., they cluster close together). The factor analysis method used was the unrotated matrix of principal components based on the same matrices of correlation coefficients. These methods are standard procedures, and incorporate the least manipulation of the original data (unlike, e.g., varimax rotations, etc.). Additionally, multidimensional scaling was used to corroborate the results of the factor analyses. Data sets that were collected using randomly located sites do not have the same geographic locations in each sampling period. In these cases the data for each cover type were lumped together to represent the overall pattern for the cover type. For example the bird transect data had a matrix of 46 rows (species) by 5 columns (cover types), i.e., a 46 x 5 matrix. With only five columns, these matrices were analyzable by cluster analysis but not by factor analysis (the latter method requires more than five rows and/or columns). 10 BULLETIN FLORIDA MUSEUM NATURAL HISTORY VOL 41(1) Tests for Divenity, Evenness, and Patterns of Dispersion Species diversity was calculated using the Shannon H diversity index (Zar 1996). Species' patterns of dispersion among cover types were characterized as 1,niform, random, or clumped (also called contagious or aggregated) using the Index of Dispersion. The Index of Dispersion (I) was calculated as the variance divided by the mean, of a sample of locations, where a species was recorded (Krebs 1989; I = variance/mean). The test statistic for this index was chi square (X 2), where df (degrees of freedom) = number of locations minus 1. Interpretations were based upon a two way test, in which the null hypothesis that the distribution was random was accepted if: X2 a,75 < Observed X2> X2 aozs Significant differences less than 0.025 were interpreted as clumped, and greater than 0.975 were uniform. Habitat Quality and Species Composition Habitat requirements for all life history stages of each species were determined based on the literature and personal experience. Each species was then assigned to one of two categories based upon these life history traits. For the purpose of the analysis, species whose respiration, feeding mechanisms, diet reproduction, or larval development require 1 to 12 months of standing water each year were termed "wetland dependent." Species whose respiration, feeding mechanism, diet, reproduction, or larval development are independent of standing water were termed "non-wetland." Animals described as "wetland dependent" use upland habitats, but a population could not persist with6ut suitable wetland habitat. Conversely, animals described as "non-wetland" use wetland habitats, but their life history traits allow them to survive and successfully breed outside of wetlands. Within this group, some species may be highly tolerant of wetland conditions, while others are intolerant. Species assigned to the same category may have different preferences with regard to timing, depth, and duration of flooding. Some species designations were difficult due to insufficient information. Others, mainly birds, required consideration of the relationship between hydrology and vegetation. For example, most woodpeckers use forested wetlands, such as cypress swamps. However, use of cypress swamps is due to the presence of trees, not hydrological conditions, since woodpeckers also successfully live and reproduce in upland forested areas such as pinelands and hardwood forests. Therefore, all woodpeckers were categorized as non-wetland. In contrast breeding common yellowthroats (Geothlypis trichas) are strongly associated with dense, graminoid vegetation. In this region, this habitat type is dependent upon hydrological conditions of standing water approximately six to nine months per year. Therefore, this species was categorized as wetland dependent. The current DALRYMPLE & DALRYMPLE: MEI-ALEUCA IN EVERGLADES WETLANDS 11 assigned wetland association of each species of amphibian, reptile, and bird is listed in Appendix I. Fishes were excluded because they are all, obviously wetland dependent and were trapped in very high numbers. Therefore, they would artificially bias the results toward the wetland dependent categorization in the evaluation of cover types. Hyd rological Assessment Pearson product moment correlation coefficients (r) were calculated for each taxa group with water levels measured at USGS gages 6972 and G975. In graphical analyses, the height of the water column and the number of individuals or species were plotted for each month. RESULTS AND DISCUSSION Macro-invertebrates from Drift fencing During the 24 months of the study, macro-invertebrates were captured at each of the 15 sites over 160 array days. A cumulative number of 9490 individuals of 10 species of selected macro-invertebrates were trapped. At any one site, the number of species of macro-invertebrates trapped ranged from 6 to 10, and the number of individuals ranged from 199 to 2112 (Table 1). Overall, the most abundant specieswere Procambarus alleni and Paleomonetus paludosus gable 2). There were no significant differences in the number of individuals (Kruskal-Wallis H «If = 4, n=15) = 7.6, p=0.11), number of species (H (df = 4, n = 15) = 2.33, p=0.68), or diversity indices (H (df =4, n= 15) = 2.73, p-6.03) of macro-invertebrates between cover types (Fig. 3). In tests of dispersion using the Index of Dispersion, all macro- invertebrates showed random distributions among cover types (Table 2). This indicated that cover type, defined by melaleuca cover, was not as important in the dispersion of the species as were other variables, including standing water. Cluster analyses revealed two main groupings of macro-invertebrates by cover types: Paleomonetus paludosus, Pomacea paludosa, Romalea microptera, Odonate lame, and Stagnicola sp were predominantly found in MAR and some of the intermediate cover type sites (P50, P75 ; Fig. 4). Procambarus alleni, dytiscid beetles (Dytiscidae), gyrinid beetles (Gyrinidae), Biomphalaria havanensis, and Lethocerus americanus were predominant in DMM, SDM and other intermediate sites. 12 BULLETIN FLORIDA MUSEUM NATURAL. HISTORY VOL 41(1) Fishes from Drift fencing During the 24 months of the study, fishes were captured at each of the 15 sites over 160 array days. A cumulative number of 27 species and 8428 individuals of fishes were trapped. At any one site, the number of species of fishes trapped ranged from 10 (DMM site) to 18 (MAR site), and the number of individuals ranged from 156 (SDM site) to 1111 950 site; Table 1). Overall, the most abundant species were Gambusia holbrooki (3803 fishes), Hemichromis letourneauxi (1059 fishes) and Fundulus confluentus (1038 fishes; Table 2). Rarefaction curves for fishes indicated thaL after 24 months sampling approached maximum species richness in some cover types (Fig. 5). The rarefaction curves for MAR and DMM indicated that new species could be expected with additional sampling. MAR had the highest species richness, with the greatest number of species trapped even though a higher number of individuals were trapped in other cover types. During the last quarter of trapping, two new species of fishes were trapped in three of the five cover types. The non-native ckhads Astronotus ocellatus and Tilapia mariae were trapped in DMM. Lepomis punctatus and Clarias batrachus were trapped in SDM, and L. punctatus and T. mariae were trapped in P50. No new species were trapped in P75 or MAR. Kruskal-Wallis ANOVA was used to compare the average number of species and the average number of individuals trapped between cover types (Table 1). There were no differences between cover types in the average number of species trapped (H ((if = 4, n = 15) = 4.03, p=0.40). However, there were higher average numbers of individuals captured in MAR, P50, and P75, than in SDK and DMM (H (df = 4, n = 15) = 10.5, p=0.03; Fig. 6). This pattern of abundance of fishes helped to explain why the intermediate cover types were commonly used by foraging wading birds, and many fish-eating amphibians and reptiles (see below). The Shannon Index was not significantly different between cover types (H (df = 4, n = 15) = 2.27, p=0.69). There were no significant differences in the number of individuals, or species of non-native fishes found among the cover types (p's>0.05; Fig. 7). Of the 27 species of fishes, 16 showed clumped distributions. However, only seven species showed this clumping within a single cover type (Table 2). Lucania goodei, Lepomis punctatus and A. ocellatus dumped in MAR Lepisosteus platyrhinchus and T. mariae dumped in P'15. Belonesox belizanus and Etheostoma fusiforme clumped in SDM. Each of the other taxa that showed clumped distributions, 9 of 16 (or 56%), were clumped in locations in more than one cover type. Since only 7 of 27 species (26%) showed clumped distribution within a single cover type, variables other than melaleuca density were eqltally important in determining species abundance. These variables would include variations in historical patterns of distribution, hydropattern, and access to deep water refugia. DALRYMPLE & DALRYMPLE: MELALEUCA IN EVERGLADES WETLANDS 13 Cluster analyses of the data for fishes showed the three MAR replicates tightly grouped together, but joined by a range of replicates from intermediate cover types, and even DMM. Four of the six dense melaleuca sites (DMM and SDM) clustered together with one P75 site (Fig. 8). This result demonstrated the wide overlap in fish community structure along the melaleuca gradient. In other words, most species of fish were found wherever there was standing water. Seven species of non-native fishes were trapped or observed in the LBSA. These species were Hemichromis letourneauxi (1059 individuals), Cichlasoma bimaculatum (656 individuals), B. belizanus (106 individuals), Cichlasoma managuense (62 individuals), A. ocellatus (21 individuals), Clarias batrachus (12 individuals), and T. mariae (5 individuals). The 19 A. oceUatus trapped in MAR cover type were all juveniles, trapped on the same day in the same trap. Juveniles of ave species were trapped (H. letourneauxi, C. bimaculatum, A. ocellatus, C. managuense, T. mariae). Belonesox belizanus, C. managuense, and H. letourneauxi are predaceous on small forage size fishes. These small to moderate size predators may have an impact on the natural recruitment of many forage fish species in the area. However, it is likely that they are preyed upon by higher level consumers (snakes, wading birds). As was the case for the macro-invertebrates, the distribution of many fishes was not strongly related to the gradient of melaleuca coverage. However, their abundances were lower in dense melaleuca coverages. This translated into a lower forage base for many higher-level consumers (e. g., many amphibians and reptiles, wading birds, some mammals). Amphibians and Reptiles from drift fencing During the 24 months of the study, amphibians and reptiles were captured at each of the 15 sites over 160 array days. A cumulative number of 1265 individuals of 34 species of amphibians and reptiles were captured. At any one site, the number of species of herptile trapped ranged from 10 (DMM site) to 22 (two P75 sites). The cumulative number of individuals ranged from 33 (MAR site; trap rate of 0.21 amphibians and reptiles per amy day) to 175 (SDM site; trap rate of 1.09 amphibians and reptiles per array day; Table 1). Overall, the most abundant amphibians were Rana sphenocephala (218 individuals), Eleutherodactylus planirostris (161 individuals), and Blifo quercicus (94 individuals; Table 2). The most abundant reptiles were Nerodia foridana (89 individuals),Anolis sagrei (83 individuals) and Nerodia,/imciata (46 individuals). Rarefaction indicated that the number of species trapped was near or at maximum levels (Fig. 9). During the last quarter of trapping, no new species of amphibians and reptiles were trapped in any of the cover types. Rarefaction curves were similar for all cover types. Furthermore, rarefaction curves for melateuca invaded wetlands exceeded the short-hydroperiod prairies in Everglades National Park (Dalrymple 1988). 14 BULLETIN FLORIDA MUSEUM NATURAL HISTORY VOL 41(1) Kruskal-Wallis ANOVA was used to compare the average number of species and the average number of individuals trapped between cover types. There were no significant differences between cover types in the average number of species g{(df= 4, n= 15)=7.68, p=O.11), average number of individuals (H (df = 4, n = 15) = 4.87, p=0.30), or Shannon diversity (H (df = 4, n = 15) = 3.77, p=0.44; Fig. 10). Of the 34 species of amphibians and reptiles, 18 (54%) showed clumped distributions. However, only six species showed this clumping within a single cover type (Table 2). Kinosternon bauri and Bufb terretris clumped in MAR Thamnophis sirtalis and Eumeces inexpectatus dumped in P75. Anolis sagrei and Hy/a cinerea clumped in DMM. The other 12 taxa with clumped distributions were clumped in locations in more than one cover type. Since only 6 of 34 species (18%) showed clumped distributions within a single cover type, this indicated that variables other than melaleuca density were also important in determining species abundance. These variables may include variations in historical patterns of distribution, hydro-pattern, and access to either deep water refugia or high ground refugia (c.f. Campbell and Christm,n 1982) When the numbers of individuals of each species were placed in a correlation matrix by cover types for cluster analyses, the sites that shared similar species composition were easily identified. In the cluster analysis by cover types, all three MAR sites separated out with one of the P50. The other two P50 grouped with the P'75. The SDM and DMM separated as a third distinct group (Fig. 11). When the same matrix was analyzed by species composition, Rana go,lio, K bauri, N. jloridana, Regina alleni, and Acris gryllus all clustered toge\her as good indicators of MAR. The majority of snakes, lizards, frogs, and toads used the wide range of intermediate cover types (P50 and P75). This included fully aquatic species such as Farancia abacura, Amphiuma means, and N. fasciata. The non- nauve Osteopilus septentrionalis, Eleutherodactylus planirostris, and Anolis sagrei, together RAth the native Gastrophryne carolinensis, Bufo quercicus, and Siren lacertina grouped together in DMM and SDM. Factor analyses of the loadings of the taxon on the first two principal components showed a broad scattering (Fig. 12). Taxa at one extreme (left side of graph) were typical of MAR and P50. Taxa at the other extreme (right side of graph) were typical of DMM and SDM. The taxa with significant clumped distributions were shaded (I index p's<0.025; Table 2). The presence of so many S lacertina in DMM and SDM habitats was unexpected (Table 2). This salamander is fully aquatic, and, is unable to feed out of the water (Bishop 1962; personal observation). It quickly dies from desiccation on dry land and does not disperse over dry areas. It was trapped at 11 of the 15 drift fence sites. Of the 60 S lacertina trapped by drift Rncing, 22 were trapped in one DMM site which was isolated from areas of lower melaleuca density. This species has a rather limited home range and individuals were trapped as soon as standing water levels existed. Four individuals were trapped at this site two days DALRYMPLE & DALRYMPLE: MEI-ALEUCA IN EVERGLADES WETLANDS 15 after heavy rain resulted in flooding of this site. A fifth, large individual was trapped on the third day following nooding These short intervals indicated subterranean refugia near the trapping sites. Another 11 S. lacertina were trapped at one SDM site. Refugia for this species are known to be subterranean moist soils, where they aestivate in a mucus covering (Bishop 1962). The substrate of porous limestone overlain with up to 1 m of muck soil was readily accessible via numerous crayfish burrows and natural crevices. A similar pattern of rapid exploitation of surface water was found for A. means by Machovina (1994). Two species of non-native amphibians and one species of non-native reptile were trapped. All three species were typical of drier, ruderal or edificarian habitats (Duellman and Schwartz 1958; Dalrymple 1988). Osteopilus septentrionalis (10 individuals) was trapped in P75, DMM and SDM. This treefrog requires standing water for its egg/tadpole stage, yet these stages are of short duration (less than two months). Eleutherodacty/us planirostris (167 frogs) was trapped in 8 separate sites representing DMM, SDM, and P75 habitats. However, 90% of these frogs were trapped atjust two sites (109 frogs at a SDM site and 41 frogs at a DMM site). This frog has no aquatic egg/tadpole stage. Anolis sagrei is highly tolerant of disturbed settings (Wilson and Porras 1983). It was most abundant in DMM (52 lizards from 3 sites), although it was trapped in all cover types (83 lizards total across all habitats). Birds from strip transects When the strip transect data were analyzed as twenty-four month cumulative data, 518 individuals of 46 species were observed across all five cover types (Table 3). P75 had the highest number of species (29) and the highest number of individuals (146; Fig. 13). DMM had the lowest number of species (9) and individuals (39). Marsh had the second highest number of individuals (137) yet had a lower number of species (15) than SDM, P75 and P50 (22, 29, and 27 species, respectively). Species in P75 were a peculiar mix of typical wetland/prairie species and upland species. Species observed in DMM were characteristic foresVedge species. Species observed in Marsh were typical of Everglades wetlands (herons, egrets, red-winged blackbird 0getaius phoeniceus), eastern meadowlark (Stumella magna), and common yellowthroat (Geothlypis trichas); Robertson and Kushlan 1984). The Shannon Index was highest in SDM and lowest in Marsh (Fig. 13). Lower diversity indicated that a fewer number of species accounted for most of the individuals. Evenness was also highest in SDM. It was lowest in P75. Lower evenness indicated that some species were dominant, while others were rare (Odum 1983). The rarefaction curves of all cover types still showed an upward trend indicating that the maximum species richness was not sampled after 24 months (Fig. 14). The numbers of new species recorded in each cover type during the 16 BULLETIN FLORIDA MUSEUM NATURAL HISTORY VOL 41(1) eighth quarter were: DMM, 0 species; SDK 1 species; P75, 1 species; P50, 1 species. MAR, 1 species. Of 46 species of birds observed during transects, 15 showed clumped distributions. Unlike the patterns seen in macro-invertebrates, fishes and amphibians and reptiles, most species (11 of 15 or 73%) clumped in a single cover type g#e 3). Geothlypis trichas, Capella gallinago and A. phoeniceus dumped in MAR. Sayornis phoebe and Quiscalus major dumped in P56. Colaptes auratus, Mimus polyglottos, Dendroica coronata, and Dendroica discolor dumped inP15. Setophaga ruticilia and Pipilo erythrophthalmus dumped in SDM. The remaining species clumped in adjacent seral stages. The 15 species (33% of total species) that had clumped distributions accounted for 76% of all individuals observed in transects (395 of 518). Many species did not show a clumped distribution simply because they occurred only a few times (e.g., Trogio*tes aedon, Melospiza geo,xiana). These results indicated that cover type defined by degree of metaleuca density was very important in the distribution of the many bird species. Cluster analysis demonstrated that the species composition of the cover types was dramatically different (Fig. 15).Geoth6pis trichas (57 individuals), and A. phoeniceus (47) were characteristic of MAR. These two species are resident breeding species typical of long-hydroperiod, marsh habitats. They accounted for 76% of all individuals seen in MAR sites during transects. The DMM sites were characterized by the presence of Carolina wren (77,yrothorus ludovicianus) and bluejay (Cyanocitta cristata). The majority of herons, egrets, perching birds, raptors, and woodpeckers used P50, P75, and SDM. These cover types had the most species represented, but no more individuals than MAR. Of the 46 species observed during transect surveys, 29 were resident species and 17 were wintering species (designations based upon Robertson 1955, Robertson and Kushlan 1984, and Louhglin et al. 1990; see Appendix D. The percentage of individuals that were resident species was highest in MAR (93%) and lowest in SDM (49%; Fig. 16). Most migratory species were wmblers, which prefer thickets or forested areas (Morse 1985). The strip transect method used in this study targeted bird species with small daily cruising radii, which selected habitat based primarily upon vegetative cover (e.g., passerines, some raptors), not standing water conditions (e.g., many wading birds). Yet wading birds are frequently given high profile in wetland assessments in southern Florida. Again, sampling methods in this study were intended to provide gross information on all species. Wading birds observed during transects were generally solitary, foraging individuals. DALRYMPLE & DALRYMPLE: MELALEUCA IN EVERGLADES WETLANDS 17 Mammals Cumulative results of small mammal live trapping and use of scent and bait stations are presented in Table 4. Since the data sets were small, only general statements on the distributions of each species within each cover type are presented below. Dasypus novemcinctus sign was common in DMM. Didelphis virginiana and Procyon lotor tracks were noted in all cover types. Each of these species are abundmit and common throughout their geographic ranges. Sylvilagus pa/ustris tracks and scats were observed in all cover types. On two separate occasions, its scat was found on top of a drift fence funnel trap when sites had standing water. Felis r:fus tracks were noted in P50, P75, SDM and DMM. Urocyon cinereoargenteus tracks were observed in P75, SDM and DMM. Lutra canadensis tracks were noted in MAR, and scat occasionally were found along a levee adjacent to MAR habitat. Odocoileus virginianus tracks were seen in each of the five cover types during the diy season. All of the above species were directly observed on one or more occasions. Live-trapping captured Sigmodon hispidus in P50, P75, and SDM, Oryzomys palustris in all cover types, and Peromyscus gossypinus in SDM and DMM (Table 5). The cover type/habitat preferences of these three rodents observed in this study were similar to trapping results in mature dense melaleuca versus "mixed melaleuca-graminoid" (Mazzotti et al. 1981) and tree islands surrounded by sawgrass marsh (Smith and Vriese 1979). Percent similarity in species composition The species composition of the MAR cover type was used as a standard to evaluate species composition of the other four cover types. The number of species that occurred in both MAR and the comparison cover type was divided by the total number of species found in the two cover types combined. Separate comparisons were made for each major vertebrate group, in each cover type. For fishes and amphibians and reptiles, species composition of each of the four cover types overlapped between 50 and 70 percent with MAR. The mammals showed similarities in species overlap with MAR from 40 to 65 percent. The birds showed the greatest difference in species composition, with between 20 and 30 percent overlap in species composition to MAR (Fig. 17). In general, as metaleuca invasion progressed, fishes and amphibians and reptiles retained a high degree of constancy in community composition. These faunal groups appeared to move in and out of local areas as water levels seasonally shifted regardless of melaleuca density. The birds showed the most dramatic shift from typical marsh inhabitants to progressively greater numbers of forest dwelling species. The mammals showed a progressive change from wetland to upland species as forest cover increased. 18 BULLETIN FLORIDA MUSEUM NATURAL HISTORY VOL 41(1) The percent of taxa that occurred in all of the five cover types varied widely between faunal groups (Fig. 18). Eighty percent of the 10 invertcbrate taxa trapped by drift fencing were found in all cover types. Only 2 of the 46 birds observed in strip transects were found in all cover types (Geothlypis trichas and Dendroica Fin:arum). Changes in species composition There were two principal physical gradients in the Lake Belt Study Area environment: tree density and water levels. Tree density was a geographic gradient, with density varying primarily from east to west. Water level was primarily a temporal gradient, varying with seasonal rainfall. While it has been anecdotally noted in the literature that melaleuca invasion causes secondary increase in ground surface elevation, we observed little evidence of this in the study area. Most sites in the study area were noocled regularly according to existing patterns of rainfall, topography, and water management. The dominant characteristic of the faunal shifts along the gradient of increasing melaleuca coverage was increased numbers of upland, arboreal, and, or forest species, not the loss of wetland species. As melaleuca coverage increased, the habitat became suitable to non-wetland species at a faster rate than it became unsuitable to wetland species. The result was a pattern of increasing species diversity and abundance in the intermediate cover types. Increased use of areas by savannah and forest birds, and mammals played a significant role in creating this gradient. The dominant characteristic of the faunal shifts along the gradient of water level was seasonal variation in abundance of wetland species. The majority of fully aquatic species (the aquatic macro-invertebrates, all the fishes, and some amphibians and reptiles, birds, and mammals) did use habitat with increased canopy cover, primarily as an effect of standing water. The existence of this prey base (invertebrates and forage sized fishes, in particular) permitted higher consumers to use these habitats. Canopy closure occurred when melaleuca cover increigerl beyond 75%, reducing sunlight penetration and primary productivity of the periphyton, submerged and emergent vegetation. This had a dramatic effect on the primary consumers and detritovore macro-invertebrates (e.g., Pomacea, Procambarus), resulting in overall lower abundance and productivity in the understory. However, complex patterns of hydrology, and gapping in forest canopy due to wind storms and fires permitted light penetration and the persistence of productive pockets of aquatic life even within dense stands of melaleuca. DALRYMPLE & DALRYMPLE: MELALEUCA IN EVERGLADES WETLANDS 19 Habitat preference and species composition Gross comparisons of the numbers of species or numbers of individuals found in each cover type did not yield significant differences among the cover types. However, multivariate analyses, which considered the contribution of each species to overall community composition, demonstrated difTerences between cover types. Indices of dispersion indicated that many falinal groups were distributed along a gradient other than melateuca density. To assist in evaluating community composition in terms of hydrology, each species of herpetofauna and bird was categorized based upon their requirement for a particular, gross hydrologic pattern (see Methods). Kruskal-Wallis ANOVA was used to compare the 24 month cumulative number of species and individuals of wetland and non-wetland amphibians and reptiles trapped at the 15 drift fence sites. There was no significant difference between cover types in the number species of wetland and non-wetland amphibians and reptiles (Kn,gk,1-Wallis H (df = 4, 15) = 6.489 and 6.210, p=0.1655 and 0.184, respectively; Table 5). There were also no differences in the number of individuals of wetland and non-wetland amphibians and reptiles (Kruskal-Wallis H ({if = 4, 15) = 5.510 and 8.610, ro.239 and 0.072, respectively). The one SDM and the one DMM site with a low percentage of wetland-dependent individuals were the two sites where the non-native Eleuthrodactylus p/anirostris was abundant (Table 2). As noted earlier, this species does not have a tadpole stage, does not require standing water during any life history stage, and therefore, is a non-wetland species. In contrast to the amphibians and reptiles, when the 24 month cumulative strip transect data for birds were considered, the occurrence of wetland-dependent species of birds demonstrated a more dramatic shift. In MAR, wetland associated species accounted for 80% of the species and 97% of the individuals. DMM had the lowest percentage of wetland associated species (11%) and individuals (5%; Table 6). It is important to recognize that species categorized as "wetland dependent" may require wetlands only during specific life history stages. Most anuran amphibians have an egg/tadpole stage that is dependent upon standing water, yet adults of some species preferentially use upland areas, only returning to water to breed. Many aquatic snakes and turtles are unable to feed out of water, yet require dry areas to lay eggs. Additionally, most species will have a preference for the timing, depth and duration of flooding. Both Geothlypis trichas and Sturnella magna generally have higher breeding densities when climatic conditions indicate low standing water levels during the breeding season (Cody 1985b). Most wetland vertebrates are adapted to using water depths of less than 25 cm (Fredrickson and Laubhan 1994). Fredrickson and Laubhan state (1994:645): "No single wetland or wetland type will provide all the resources needed by a single vertebrate during 20 BULLETIN FLORIDA MUSEUM NATURAL HISTORY VOL 41(1) all of its life-history stages or for all vertebrates adapted to wetlands. Thus, wetland complexes are essential for successful management..". Gross Hydrological Assessment Animals experience interannual variation in abundances. These variations may be partly determined by climatic conditions, such as rainfall (Cody 1985b; Morrow et at. 1997). The South Florida Water Management District has described the 1994-1995 period of rainfall as a "25-year" high rainfall event throughout Dade County. While the timing, depth, and duration of standing water conditions are comlated with rainfall, this relationship may be altered in managed wetlands by regional patterns of water management Pumpage for drinking water well fields, and/or water releases from basin to basin may affect the actual standing water levels in an unnatural manner. Therefore, we considered ground water levels as measured at USGS gages located within the stu* area as an indicator of hydrological conditions throughout the study area. To determine if there were significant differences in the average monthly ground water levels in different regions (sub-basins) of the LBSA 1994 and 1995 data for seven gages were compared. There were significant differences among the mean monthly water levels of the seven gages (ANOVA: F = 46.72, df = 152, p<0.0001), with the lowest mean value at 63253; Fig. 19). Pearson Product- moment comlation coefficients of variation in monthly mean values of all seven gages were highly significant ( all r's>0.79, and all p's<0.05), indicating that all gages followed the same pattern of timing and duration of seasonal water level nuctuation However, surface water depth cannot be extrapolated since ground elevation data were not available. Two wells were selected for more detailed analyses based upon their proximity to the majority of sampling sites. Ground water levels during the two years of the study were compared to the previous nine years for two USGS wells located within the study area (USGS 6972 and 6975). This 11 year period included yeam described as "low" (1989-1991), "average" (1986-1988),and "high" rainfall (1994-1995). Average annual water levels at 6972 and 6975 for the 11 year period showed significant differences (ANOVA; 6972: F = 10.586, df = 10, 114, p<0.0001; 6975: F = 9.891, df= 10,115, p<0.0001; Fig. 20). For each well, Tukey's Honest Significant Difference Tests were done to determine which years were significantly different from 1994 and 1995. At 6972, the average monthly water level in 1994 was only significantly higher than 1989-1991. In 1995 at 6972, it was higher than 1989-1991, plus 1985 (Tukey's Honest Significant Difference Tests). At G975, the average monthly water level in 1994 was only significantly higher than 1989-1990, and 1985. In 1995 at G975, water level was only significantly higher than the three drought years, and 1985 (Tukey's Honest Significant Difference Tests). In summary, even though annual rainfall in 1994 and 1995 was "high," average annual ground water levels measured at two wells in DALRYMPLE & DALRYMPLE: MELALEUCA IN EVERGLADES WETLANDS 21 the study area were not significantly higher than in years of "average" rainfall. However, there was less variation in water level during 1994 and 1995 (i.e., it was wet longer). Correlations of Gauge Ground Water Level with Trap Rates Major peaks in capture of macro-invertebrates, fishes, and amphibians and reptiles by drift fences were generally associated with changing water levels (either rising or falling; Fig. 21). When standing water existed over large areas, aquatic and semi-aquatic animals were more dispersed, and capture rates were generally lower. Successional changes in vegetative structure and faunal implications Melaleuca invasion of native graminoid/herbaceous wetlands changes the vegetational structure of the landscape. It is unclear to what extent melaleuca invasion also changes the hydrological characteristics of an area because variation and shifts in water management and human disturbance are so strongly correlated with the distribution of melaleuca. This study was designed to address only the impact of melaleuca coverage on wildlife species richness and abundance. Prior to the current study, the only information available was based upon either dense melaleuca stands only (Schortemeyer et al. 1981) or were short-term studies that considered only a few species (Mazzotti et al. 1981; Sowder and Woodall 1985; Repenning 1986). As melaleuca coverage increases, a graminoid wetland with low structural diversity becomes a savannah (mix of open prairie/marsh and trees) with increased structural diversity. As melaleuca coverage continues to increase, the savannah becomes a closed canopy forest with sparse understory. Since little understory persists in the forest and most of the trees are of similar size, structural diversity of the forest is lower than existed in the savannah stage of melateuca invasion. Some animals (e. g., many birds, c.f. Cody 1985a) select habitat based upon subtle differences in vegetational structure. However, other animals (e. g., amphibians and reptiles) are less sensitive to vegetative structure but select habitats based upon other characteristics (e.g., soil or hydrological characteristics; Campbell and Christman 1982). The results of this study demonstrated a higher species richness and abundance of birds in the cover types that have moderate levels of melaleuca coverage. As discussed above, these were the cover types with the greatest structural diversity. Notably absent from these areas, though, were resident bird species that are selective about the types of trees they use (e. g., pine warbler (Dendroica pinus)). Many of the transient and winter-resident birds occurred at much lower abundances than in cypress swamps of the Big Cypress National 22 BULLETIN FLORIDA MUSEUM NATURAL HISTORY VOL 41(1) Preserve or the uplands of Long Pine Key, Everglacles National Park (personal observations). In contrast to the birds, a similar diversity of herpetofauna was found across all cover types. However, their abundances generally decreased in the closed-canopy melateuca forest (DMM cover type). The lower abundances indicated poorer habitat quality. This was probably the result of the closed-canopy of the forest limiting the amount of sunlight reaching the water surface. With reduced sunlight, the algae forming the structure of the periphyton mat does not develop. Many species of amphibians and reptiles consume crayfish, grass shrimp, and smaller forage fishes which depend upon a well-developed periphyton mat. However, complex patterns of hydrology, and gapping in forest canopy due to wind storms and fires permit light penetration and the persistence of productive pockets of aquatic life even within dense stands of melateuca. Changes in both structural and wildlife diversity are summarized in Figure 22. Landscape effects Habitat interspersion and melateuca patch size were not explicitly considered in sampling designs because detailed maps of the area were unavailable at the start of the project. The only variable considered was melaleuca coverage. Random sampling of three replicates of each cover type per month did not permit testing of any variable other than melaleuca coverage. However, the mosaic of areas with low to moderate infestations of melateuca surrounding mature dense melaleuca stands may allow higher numbers of individuals and species to persist in, or seasonally use, mature dense melateuca stands. A single stand of melaleuca surrounded by prairie has less habitat interspersion than several, smaller stands of melaleuca which have the same total area as the single large, stand. The smaller stands have more "edge" habitat which is likely to provide at least marginal habitat for species characteristic of the prairie. However, higher degree of interspersion (more edge) may also expose surrounding natural areas to higher seedfall, since seedfall is generally limited to a distance less than 1.5 times tree height (Meskimen 1962). Factors affecting the rate of spread of melaleuca have not been examined. The most widely cited paper on melaleuca expansion rate by Laroche and Ferriter (1992) did not explore causal relationships between melaleuca invasion and biotic or abiotic factors. In calculating expansion rate Laroche and Ferriter only considered land sections that had attained 100% melaleuca coverage. This approach was explicitly recognized by the authors as a constraint on the application of their results, yet their results have been widely cited as the single possible melaleuca expansion rate. Exclusion of sections that had some melaleuca coverage yet had resisted heavy infestation may have led to the calculation of the fastest possible expansion rate. Moreover, "invasion" was interpreted as the presence of one or more melaleuca trees in an acre. This has unfortunately been DALRYMPLE & DALRYMPLE: MELALEUCA IN EVERGLAI)ES WETLANDS 23 improperly interpreted as 100% dense melaleuca coverage, which was not the intended use of the authors. Additional studies should examine land sections that are exposed to metaleuca yet have resisted heavy infestation. Factors influencing the rate of melaleuca expansion, such as habitat interspersion, melateuca patch size, soil, plant cover, human disturbance and hydrology, should also be considered. While the interspersion of areas of varying melaleuca coverage may contribute to the abundance of animals (particularly fishes and semi-aquatic amphibians and reptiles) in dense melateuca sites, it is unlikely that the Pennsuco marshes on the western edge of the area were the sole source of fishes and some fully aquatic amphibians and reptiles in the study area. High levees subdivide the LBSA along north-south (Dade-Broward Levee) and east-west axes (levees associated with Wellfield and Pennsuco Canals). These levees were dispersion barriers to fishes, and some fully aquatic amphibians and reptiles. Therefore, some species were confined to isolated sub-basins, which sustain local populations. The abundance of Siren lacertina (a fully aquatic salamander) in a DMM site isolated from areas with lower melaleuca coverages was a good example of this. The rapid rate at which fully aquatic amphibians and reptiles and fishes exploited standing water in many sites indicated that deep water or subterranean refugia were available even within areas of dense melaleuca. Likewise, the highly vagile mammals and birds were readily capable of exploiting small patches of suitable habitat throughout the entire region. The numerous, recent reviews of the relationships between habitat quality, demographics, dispersal, and metapopulations that are being derived from landscape ecology are all relevant to future research on the impact of melaleuca (c.f. Hansson 1995). For example, to what extent do melateuca invaded habitats function as marginal habitat?, and what effect does the ratio of optimal to marginal patch area (ROMPA hypothesis; see Hansson 1995) play in the dynamics of the various populations in the areas of melateuca invasion? Melaleuca continues to aggressively invade wetland habitats in southern Florida as well as upland habitats in southwestern Florida and parts of Broward and Palm Beach County. While the replacement of native vegetation with a monoculture of non-native species is undesirable, it is important to recognize that animal populations will persist in areas with distufbed vegetation. Therefore, these areas still retain some habitat value. Successful restorations must re-establish native animal communities as well as the native plant communities. Since many native animals may persist in areas with melaleucg preference should be given to restoration methods that are sensitive to the existing on-site animal populations. LITERATURE CITED Alexander, T. R., and R. H. Hohtetter. 1973. Some cumnt ecological aspects ofMelateuca quinquenervia (Cav.) Blake in southern Florida. Presented al the Florida Acad. Sci., 415: Ann. Mtg 24 BULLETIN FLORIDA MUSEUM NA:IURAL HISTORY VOL 41(1) Bishop. S. C. 1962. Hancil,ook of Salamanders. The Salamanders of the United Statca. of Canada, and of L»wer California. Hafher Publi•hing Co., New York. Campbell, H. W., and S. P. Chri,iman 1982. The herpetological components of Florida sandhill and sand pine scrub associations. Pp 163-171 in Norman J. Scolt, ed. U.S. Fish and Wildlife Seivice Wildlife Research Report 13, Washington, D.C. Cody, M. L 19851 Habitat Selection in Birds. Academic Press _. 1985b. Habitat selection in grassland and open-country birds. Pp 191-226 in M. L Cody, ed. Habitat Selection in Birds. Acs,temic Press. Dalrymple, G. H. 1988. The herpetofauna of Long Pine Key, Everglades National Park, in relation to vegetation and hydrology. Pp. 72-86 in R. Szam, K.E. Severson, and D.R. Patton, technical ©oordinators. Management of An*bians, Reptiles, and Small Mammals in North America. U. S. D. A Forest Service. Gen. Toch. Rept RM- 166, Fort Collins, CO. Dalrymple, 0. H. 1994. In-faunat Study of Wetland Restoration in the Hole-m-the-donut, Everglades National Pa,k 1990-1992. Final report to South Florida Research Center, Everglades National Park. Davis, J. H.. Jr. 1943. The Natural Features of Southern Florida. Especially the Vegetation. and the Everglades. Flo,i(la Geol. Surv., Tallahassee. Davis, S. M., L H. Gunderson. W. A Pa,k, J. R Richar,son. and J. E. Mattson. 1994. Landscape dimension, composition, and function in a changing Everglades eoosystem. Pp. 419444 in S. M. Davis and J. C. Ogden, eds. Everglades. The Foosyatem and its Restoration. St Lucie Press, Delmy Beach. FL Duellman. W. E.. and A Schwartz 1958. Amphibians and reptiles of southern Florida. Bull. Florida State Museum, Biol. Sci. 3(5):181-324. EAS F-ginecringk Inc. 1995. Year One Report on Vegetation of the Lake Belt Study Area. Report submitted to Dade County. Dept Environ Res. Mgmt. Anril 1995. Fennema„ R. 1, C. J. Neidrauer, R. A Johnson, T. K MacVicar, and W. A. Perkins. 1994. A computer model to simulate natural Everglades hydrology. Pp. 249-289 in S. M. Davis and J. C. Ogden. eds. Everglades. The Ecosystem and its Restoration SL Lucie Press, Detray Beach, FL Fredrickson, L H., and M. K I.aubhan. 1994. Managing wetlands for wildlife. Pp. 623-647 in T. A Bookhout, ed. Research and M t Techniques for Wildlife and Habitats. Fifth Ed. The Wildlife Society. Bethesda, MD. G=*e, H. G., Jr. 1982. Multivariate analysis in community ecology. Cambridge Univ. Press. Hansson, L 1995. Development and applicaiton oflandscape approaches in mammalian ecology. Pp. 20- 39 in W. Z Lidicker, Jr., ed. Landscape Approaches in Mammalian Ecology. Univ. Minnesota Press, Minneapolis and Ikedon. Hoatetter, R. H. 1991. The current status of Melaleuca quinquenervia in southern Florida. Pp 159-176 in T. D. Center, R F. Doren, R H. Hohtetter. R. L Myera, and L D. Whiteaker, eds. Proceedings ofthe Symposium on Exotic Pest Plank Tech. Rept NPS/NREVERWRTR-91/06. Keller, C. M. E., C. S. Robbins, and J. S. Hatfield. 1993. Avian oommunities in riparian forests of different widths in Maryland and Delaware. Wetlands 13:137-144. Knba. C. J. 1989. Foological Methodology. Harper and Row, Publ. New York. Laroche, F. B.. and A P. Ferriter. 1992. The rate of expansion of metaleuca in South Florida. J. AquaL Plant Mont. 30:62-65. Lars< P. W. 1992. South Florida I.imestone Mining Coalition Year 2050 Fresh Water Lake Belt Plan. Larsen and Associates. Miami, FL Inust,16 M. H.. J. C. 05~ W. B. Robertson, Jr., K Russell, and R W. March. 1990. Everglades National Park Bird Check List Florida National Paiks and Monuments Association, Inc., Homes:cad. FL 18 pp Machovina, B. L 1994. Ecology and life history of Amphiuma means in Everglades National Park. M. S. thesis, Florida International Univ., Miami, FL Manolti, F. J., W. Ostrenkg and A T. Smith. 1981. Eftects ofthe exotic plantsMelaleuca quinquenervia and Comanna equiserfolia on small mammal populations in the eastern Florida Everglades. Florida Sci. 44:65-71 Meskimen, 0. 1962. A silvical study ofthc melateuca tree in South Florida. M. S. thesis, Univ. Florida Gainesville. 177 pp Mitsch, W. J., and J. 0. Gosselink. 1986. Wetlands Van Noe*and Reinhold, New York. Morse, D. H. 1985. Habitatselectionin North American pantlid warblers. Pp. 131-157 in M. L Cody, cd. Habitat Selection in Birds. Academic Press IYALRYMPLE,abAI:RYMPEE: MELAIsEUCA IN~EVERGI:ADESVETLANDS 125 6dum. EDP.: 1283. ~REi# ff#14*9= ~#*FIF#,Bie611§***~i#9& IRepennin~ R. W. 1986 . Miti *ation of,Fish aod Wildlife Values in *6ck*ined:Aidas.of Sauth Florida Pa*Ii: Wildlift: »Coop. FishWildlif*'ResrUnit Report, Univ. Fi*ida,Gaincsville Robertson' W. B.,jr. 1955. An*nalysis«#fthe Breeding-bird*opulaiions»f Tropical Florida:in Relation to t~Ve~*tati6LP6:D.,di~*1*}*1)*iy 1%**'Urbi,* - Robertson,W. 9., Jr.,.and 91 A. Kuslilan. 1984. Tli* sd~,thern,Floridi,a*ihihit Pp. 219-257. in P. J. GleaMon, 4 Enviroi),iients of South,Florida *esent and Pasfi K{ianii Oeol75% melaleuca coverage, sapling trees; DMM = >75% melaleuca coverage, mature trees. DMM'11 A B. D ALR YM PLE,a LA I*O C A IN EVER G LAD ESW ETLAN D S Rornacea paludosa SDM 1 L 1.' pPaleo. paludosus , r 10MARI Romaleamicroptera, MAR2} - 1 MAR3· Stagnicola sppi DMM2 · Dragonfly larvae' DMMS P75/ 1 ' · Procambarus alleni' P50/2 : 1 BySkidae FESDM2· Leth. americanusSDM3 P50/3 Biom. havanensis P75/2 · CyrinidacP75/3 .. 0.0 0.2 0.4 06 0.8 0.0 0.2 0.4 0~ 4.8 ~0 11.2 1:4 Linkage Di5tance Linkage IDibtance Figure 4. Cluster analyses by mdividual replicates and by taxa for macroinvertebrates. Based upon 2 data Rom the 15 repticates (3frepticates per cover type) Cover type abbreviations as in Fig 3 Y 30 B U LLE TIN FLO R ID A M U SEU M N ATU R A L H ISTO R Y V O L 41(1) 24 20 516 SP EC I #*.. MAR 12 , -m. . >•* pso 1-*. P75 8 5\ DMM 0 400 800 1200 1600 2000 2400 2800 3200 INDIVIDUALS Figure 5. Rarefaction curves for fishes trapped in each cover type. Curves based upon 24 month cumulative data from drift fencing. Cover type abbreviationsas in Fig. 3. DALRYMPLE & DALRYMPLE: MELALEUCA IN EVERGLADES WETLANDS 31 1200 1000 A._ . 2(Min-Max 800 Median value 600 IN D IV ID U AL S 400 200 0 . . MAR P50 P75 SDM DMM 19 17 B. T -4- 15 SP EC IE S 13 _Ii 11 D.9 MAR P50 P75 SDM DMM 1400 1200 C. P50/2 P50/3 0 1000 P75/3 IN D IV ID U AL S 800 P7511 Plt/2 P50/1 600 D*lf[.-3 M'~Ai MARI 400 MAR3 DMM1 SD*M%9"l200 • •DMM2 0 8 9 10 11 12 13 14 15 16 17 18 19 20 SPECIES Figure 6A Box-whisker plot ofcumulative number of individuals offishes trapped by *ift fencing for each cover type. Knisal-Wallis H (4, 13) = 10.5, p = 0.03. 68. Box-whisker plot of cumulative number of species of fishes trapped by drift fencing for each cover type. Kruskal-Wallis H (4, 15) = 4.03, p = 0.40. 6C. Plot ofnumber of species versus number ofindividuals trapped f6r each of the 15 replicates (3 per cover type). Cover type abbreviations as in Fig. 3. 32 BULLETIN FLORIDA MUSEUM NATURAL HISTORY VOL 41(1) I Min-Max LU m Median value Li.1 0al (/) ~ 5. 0.025). W SI 38 BULLETIN FLORIDA MUSEUM NATURAL HISTORY VOL 41(1) 32 160 - I NO. spp. 28 D NO. INDS. -140 24 r (11 ·120 0 20 100 ul 6 16 80 LU S l¥nO IA IG N I S S 3N N 3A 3 1 608 ~ ~~40 15.,k E ~ -...» --/$. ,- 0 MAR P50 P75 SDM DMM 4.2 0.95 I DIVERSITY 090 E~ EVENNESS 3.8 31 0.85 0.80 D IV ER SI TY 3.4 .A 0.75 E '6 070 3.0 0.65 0.602.6 ~ ~ 0.55 2.2 0.50 MAR P50 P75 SDM DMM Figure 13. Abundance, species richness. and diversity of birds observed during strip transects in the five defined cover types. Cover type abbreviations as in Fig. 3. D ALRYM PLE : bALEW C A VER LAD ES)W ETLAN D S 32 *28 O 24 'sjLM 5 I PZE 0 20 »:<..... ..............0............ 6 16 : /6. 0 : IUl 0 12 8 ~,50 - -0.: P75 4 =~SDM- --vE DMM 0 0 20 40 60 80 100 120 140 160 INDIVIDUALS Figure 14. Rarefaction curves (number of species versus number of individuals sampled) for birds based upon strip transect data for each cover type; 24 month cumulative data collection. Covertype abbreviations as in Fig. 3. 40 B U LLE TIN FLO R ID A M U SEU M N ATU R A L H ISTO R Y VO L- 41(1) Phalacro. auritus · Circus cyaneus Gallinula chloropus Melospiza georgiana Capella galifnago Geothlygls tnchas MARAgelaius phoeniceus Butondes striatus Chordeiles minor Aidea herodias Lanius ludovicianus Hydranassa tricolor Megaceryle alcyon Icteria virens Stumella magna MycteatteWT::Ts : Colaptes auratus Troglodytes aedon Mimus polyglottos Dendroica coronata bDendroica discolor Picoides Dubescens Polioptila caerulea Cardinalis cardinali P50 & P752-2Dendmica palmarum Flonda caerulea Sayonds phoebe Zenaida macmura Columbina Dasserina Tyrannua tyrannus Quiscalus major Casmerodius albusR]-iQuiscalus quiscula Falco sparverius · Setophaga ruticilla Mylarchtis crinihis Vireo gnsews Pipllo erythfo. Dumetella carol. Parula americanaMe/ane,pes cam/ DMM & SDM ]C)*~rc~%~a~% , Thgo. ludovicianus J-j Mmotilta varia 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 Linkage Distance Figure 15. Cluster analyses by cover type and by tan for the strip transect data for birds. Based upon 24 month cumulative strip transect data from the five cover types (data fromthethree replicates per covertype combined). Covert>~e abbreviations as in Fi* 3. 100 *Ir 100 - T.. /11 * Ae 01 41; Mi~. PE R C EN T O F IN IV ID U AL S BO @ 80 8 0 60 85 60 D A LR Y M P LE '*D A LR Y M P LE : M ELALEU Q A IN EYERG LADES W ETLAN D S 41 '8 40 C31 WINTERING ~0. \MIYIERING 20 RESIDENT 20 RESIJ»IT ~ MAR! P5O' [1'75 SESM D~ DIVIM - Figure 16. Plot ofthe percentage ofindividualsand species ofresident and wintering bird species. Based upon 24 month cumulative strip transect data from the five cover types (data from the three replicates per covertype combined). Covertype abbreviations as in Fig 3. 42 B U LLE TIN FLO R ID A M U SEU M N ATU R A L H IS TO R Y VO L- 41(1) 80 80 A. FISHES B. AMPHIBIANS AND REPTILES, 70 ~ 70 1 60 50~ ~ ~ ~ 950 Z % IN C O M M O N W IT H M AR % IN C O M M O N W IT H M AR 0 40 5 30 ~20 10 ~10 0 P5O P75 SDM DMM 80 80 C. BIRDS D. MAMMALS70 ---· ~ 70 60 = 6050 i ED R so 40 30 ~ ~ --z 10~ ~ ~ ~100 P50 P75 SDM DMM Figure 17. Percent ofspecies in common between Marsh and each other covertype for fishes, herptiles, birds, and mammals. Cover type abbreviations as in Fig. 3. ¤j 50, · 100 A. - B 1551 TOTAL NA ACR058 55'· 1 LILI 5 COvER,TYRES % 80 EE & D A E k¥M P LE : M E LA LE U C A IN 43 40 - ONLY TAXA FOUNC IN ~~ * ALL 5 COVERIEES · 5 # SP EC IE S 30 'T,mn~ if '* 0, d 20 :> ': -.~a'-/-- 1 40 8rl El- - f*' O - OO 51'i, Uf Ig10 IS, 1 85 20 INVIS FISHES HERPS BIRDS MAMMALS ' INVTS EISHES MERPS, BIRDS MAMMALS E Figure 18. Total tan f6r each falinal group and percent oftaxa found in all of the five defined cover types. 2.8 2.6 G968 2.4 .-'6594 1. .0 0 2.2 ./ - 31488 . -, 2.0 - - .47 '.-I.' W AT ER L EV EL (m ) 1.8 : 6978 -0 -· : ' 4 0 69721.6 1.4 1.2 1.0 G3253 0.8 0.6 B U LLE TIN FLO R ID A M U SEU M N ATU R A L H ISTO R Y V O L 41(1) 0.4 1994 1995 0.2 JFMAMJJASONDJFMAMJJASOND Figure 19. Plot ofthe average monthly water level for the seven USGS gages located within the Lake Belt Study area for 1994-1995. DALRYMPLE*DALRYMPLE: MELALEUCAINEYEROLADES'WETLANDS 45 4. USGS G-972 I *1:96·Std Err. A N U A L M E A N W AT E R L EV S, m ( N G V D ) A N N U A L M EA N W AT E R 'L EL S, m N G VD - B *100 Std. Err. 1.8 • Meani - ] *Crl *$ 1.6 - -- - 13=lits .-.-1-- 6 ED 5d FIZI, - r-41 [42i.wi1.2 -A -8 -8 -8 ··- 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 2.0 B. USGS G-975 -7- -+- 1.-r' III 1.1 1:2· Z *1.96:Std. Err, ~ :1.00•Std. Err. "8 -8 "B "A • Mean, 1.0 - 1985 1986 1987 1988 1989 1996 1991 1992 1993 1994 1995 Figure 20A, Plot of monthly average water level at 0-972 for the period 1985-1995. 208. Plot 6fthe monthly average watef level at' 0-975 fbr tli peri6d 1985-1995. Each plotfisintended to demonstrate the range ofconditions for the 11 year periBd,,ratkitr,than.to ode*are specifi6 years. **A - Year significantly different from 1993 only. **B = Year significantly different from both 1994 and 1995. 46 BULLETIN FLORIDA MUSEUM NATURAL HISTORY VOL 41(1) 1400 1.9 63 120)0 -,- mR (U ~ 1.8 "- 1000 - P.(L) -0- P75(L) • - DMM (L) 1 3 A a l 6131*M 13A37 831VAA 13A 31 NE.LVAA ~ 400 S « JFMAMJJASONDJFMAMJJASOND 700 1.9 . 600 + MAR (U 500 -·o-· P50 (L) - 0- P 75 (L) 1 1.7 400 -*-- SoM (L) - H ER PT IL ES #F IS H E S 300 - G972 (R) ' , 1.5 200 ' . 100 ··?*"/ '//Al• •vi .• i '* ·9' ..71.4 0 I / ..#'.,#....... . 1.3 JFMAMJJASONDJFMAMJJASOND 90 1.9 80 -+ MAR (L) 1.8 70 -0- P50 (L) ·•- P75{L) i . 1.7 60 .4.- BDM (L) 1.650 -• · DMM (L) i, A 40 - 6972(R) 1.5 20 -««j4&„,z'«~ Z .i)'~ 'itji~t**§:i=i 1.31.2 JFMAMJJASONDJFMAMJJASOND Figure 21. Plot ofthe monthly number of individuals of macroinve,tebrates, fishes, and herptiles trapped in each cover type by drift fencing (Jan, 1994 thru Dec, 1995). Also plotted is the average monthly water level at USGS gage G-972 for the same period The horizontal line represents the LSD elevation for the gage (1.5 m, NGVD). Actual ground surface elevation is likely to be lower. Cover type abbreviations as in Fig. 3 Lowllevels of Melaleuca Moderate levels,of Melaleuca High levels of Melaleuca No to litt16 tree canopy ! vpen tree canopy ' 'Closed tree 6anopy , A ';A . . D ALR YM PLE & DAISRW M PLE M ]E LA LE U C A N E #fo#A bE S ,W E ¥LA N D S 47 l ~\W - , -if » l NK.. ,_=I.~IczE.,rapi,1.,~&,-ra o'Idup , 1~ Diversity of undemtory plants ~ Divef~tgof understdryplants , Loss of understory plants Well d«veloped pirlphyt9i mal ~ Well developed,perlphyton mat Poorly developed perlphytor mat f High abundaxe of crayfish & grass shrimp, High abundance,of crayfish'& grass shrimp', Fewercrayfish & grass shrimp Hlgi abundance of fishes l' High dbundance of fishes ~ Fewer fishes Maint, .wetland,herptiles ; Mal,Jy welland with some upland,herptlles | Mainly wetland with come uplard herpmes Mainly wetland birds 1 - , Mainly upland birds i Mix of wetland & upland birds , Mainly wetland'mammals ~ Ml» of w.#and »b« .0md mammals ; '*" «'w'""nd »a"« u'lin« m.m„~,i, Figure 22. Summary of changes in vegetation structural diversity and wildlife diversity in native graminoid wetland habitats with increasing,coverage'by melaleuca. 48 B U LLE TIN FLO R ID A M U SEU M N ATU R A L H IS TO R Y VO L- 41(1) Table 1. Summary of drift fence trapping results for each site (15 sites total), 24-month cumulative numbers. Cover type abbreviations: MAR=<10% melaleuca coverage, P50=10%to 50% melaleuca coverage; P75=50% to 75% melaleuca coverage; SDM=>75% melaleuca coverage, saplingtrees; DMM=>73% melaleuca coverage, mature trees. MAR P30 Y15 SDM DMM Site: 123 123 123 123 123 Invertebrates Odonate larvae 12 6 2 -- 1 8 -- 1 3 -- 1 - 1 Romalea microptera 41413 ---- 1 - 15 3 5 4 4 4 Lethocerus americamis 6 9 15 1 66 87 66 40 39 9 17 39 4 1 17 Dytiscid beetles 7 25 7 7 84 86 78 68 32 17 12 45 5 5 14 Gyrinid beettes - --- - 4 - 2 -1 1 1 Biomphalaria havanensis 1 5 4 14 316 -- 1 - 1 Stagnicola sp - 6 - - 11 113 1 - 1 5 - 1 Pomaceapaludosa 2 11 4 1 -- 1 3 - 2 1 -- 166 Paleomonetus paludosus 295 748 690 459 3 19 4 5 10 319 9 17 130 19 1 Procambarus client 68 67 50 104 626 1896 763 352 94 136 243 669 124 292 153 Fishes Lepisosteus platy,hinchus 2 - - Ameiums natalis - 1 Ameiums nebuloms - - - 1 - 1 Clanas batrachus - 1 3 1 1 1 2 3 C»nnodon vanagatus 1 - - - - - - - - - - - 1 Fundulus chgsorus 30 29 34 76 29 35 33 16 14 4 - 34 --3 Fundulus con#uentus 32 43 16 99 62 80 85 164 132 37 118 121 26 11 12 Jordanella floridae 19 43 40 99 53 53 6 17 75 1 7 11 25 13 Lucania goode, 18 11 10 7421121121 9 0 Table l Continued 'Belenosox belizamit ----- 12 - 43 3 31 10 7 Ganibusta hotbrooki 296 264 281 264 598 556 344 130 429 35 17 98 81 10 1*1 ALR YM PLEf& D A LR Y M P 49 Hejerandriaformosa 3 12 2 4 2 - 5 10 3 Poecma latpinna 16 46 50} 39 82 62 38 24 113 1 1 2 130 Labidesthes sicculus Enneacanthus glonosus 1 2 2 3 1 - - Lepe»iii gutoms 11 - 125 2 2 115 2 4 Ldp*mis macrothirus 2 1 1 1 - 1 1 1 - Lepomis marginams 1 1 - - - - - 1*o,Ristinidi·610phus 11 3 10 11 6 5 3335 18146 L Lepomis punctatus 3 5 1 3 1 - 1 - Microptents salmoides 1 1 - - Etheostomahsiforme 1 - 1 3 Astronotus ocellatus* - 19 - - 2 - Hem*hromis letournemuci* 6 19 7 15 132 148 150' 234 96 18 18 133 735 41 Clchiasia,bimatutaium* 5 5 13 12 135 80 41 42 36 10 5 105 42 102 23 Cichiasoma.mauatiense* 8 2 2 22 1 - 3 1234311 ' E hiapia mariae* 1 - 3 1 - Amphibians Pseudobranchus striatus - - - 1 - - - - - M firentacerrina 2 1 2 5 2 3 3 2 711 22 - ~Amphiuma means 2 - 11410433 368 1 4 Notophthalmus viridescens - Bufoterrestris 72 - - - - 13 - 4 Bufoquerclcus 2 - 6 9 6 24 3 15 10 12 2 3 Gastrophgne carotinensts 1 1 3 5 6 14 - 3 8 1 Eleutherodactylusplaniromis®- -121 109 1 41' 10 2 Pseudacris nigrita 1 50 B U LLE TIN FLO R ID A M U S E U M N ATU R A L H ISTO R Y V O L 41(1) Table 1. Continued. Limnaoedus ocularis - 1 - - - - - Acris gryllus 4 1 - 22 13 - 1 - Hyla cinerea 1 1 - 2 - 2 5 Hyla squireMa - 1 - 1 --- Osteopilus septentrionalis - - - - - - 2 - - 2 1 - -23 Rana sphenocephala 13 9 3 15 16 15 34 15 15 9 6 19 17 13 19 Rana grylio 5 4 2 12 8 7 14 2 7 1 6 - 7 Reptiles - Kinostemon bauri 738 214 113 - 111 Terrapene carolina - - - - 1 - 1 1 1 Chelydra serpentina -- 11 -- 1 - --- -- Anolis sagrei* 1 -- 6 -- 17 - 6 4 6 10 Anolis carolinensis 31 - 5 --- 23 1 Ophisaurus compressus - 142 11 - - Eumeces inexpectatus - 12 - - 3 - 1 Nerodia fasciata 624 213 384225 1 Nerodia floridana 19 8 7 15 4 11 2 1 6 1 5 7 3 A l 1 - I l l q 1 - 1 - I S 1 1 I N I 1 - 1 1 - Ill' 1 Regina allem 10 4 249 112 -12 1 Thamnophis sirrahs 2 - - - 5 3 863 423 6 1 Thamnophis sauntus - 1 1 1 1 2 3 5 3 342 2 Diadophispunctatus 1 - - - - 1 1 Farancia obacura - - - 3 1 1 2 --- - Coluber constrictor 1 2 1 - 4 6 1 147 2 Elaphe guttata 2 - - 2 - 1 1 Lampropeltisgetula floridana - - 2 - 2 - 1 - Agkisrrodon piscivorous 1 1 1 3 228 3 Mammals Blarina carolinensis 1 Table l Continued. 1 - 1Sigmodon hispidus Oozomys'*alusms 1 i All anirials Number of species 40 40 37 39 39 36 41 42 45 38 35 40 28 45 39 Number of individuals 991 1,438 1,278 1,284 1,962 3,229 1,748 1,232 1,220 832 521 1,463 573 398 818 Trap rate 6.19 8.99 7.99 8.03 12.26 20.18 10.93 7.70 7.63 5.20 3.26 9.14 158 3.74. 5.11 D A LR Y M P LE *D A LR Y M P LE : M E LA LE U C A IN EVER G LAD ES W ETLAN D S Invertebrates only Number of species 798 678 787 9688710 Number of individuals 394 887 774 575 785 2,112 918 472 206 501 287 778 274' 328 199 Trap rate 2.46 5.54 4.84 3.59 4.91 13.20 5.74 2.95 1.29 3.13 1.79' 4.86 0.93 1.34 0.83 Fishesonly Number ofspecies 18 16 16 16 14 12 12 12 17 13 12 13 10 17 12 Number of individuals 514 505, 471 636 1,111 1,031 734 650 935 156 175 384 198 185 543 Trap rate 3.21 3.16 2.94 3.98 6.94 6.44 4.59 4.06 5.84 0.98 1.09 3.65 '0.91 1.16 0.59 Amphibians & Reptiles only Number.of species 15 15 13 17 18 16 22 22 20 16 16 18 10 21 16 Number of individuals 83 46 33 73 66 86 96 110 78 175 58 100 101 85 75 Trap rate 0.52 0.29 0.21 0.46 0.41 0.54 0.60 0.69 0.49 1.09 0.36 0.63 0.63 0.53 0.47 U„ - 52 B U LLE TIN FLO R ID A M U SEU M N ATU R A L H ISTO R Y V O L 41(1) Table 2. Results of drift fencing trapping summarized for each species by cover type (3 replicates in each cover type). 24 month cumulative numbers. Also indicated, for each species, are the Index of Dispersion (I), Chi-square value (%2), and whether distribution was clumped in any cover type. Cover type abbreviations as in Tab. 1. MAR P30 P75 SDM DMM I Index *2 Cover types Invertebrates Odonate larvae 20 9 - 4 2 5.49 76.86 Romalea microptera 19 3 1 23 12 5.65 79.069 Lethocems americanus 30 154 145 65 22 26.80 375.18 Dytiscid beetles 39 177 198 74 24 29.01 406.16 Gyrinid beetles - 4 3 2 2.10 29.33 Biomphalaria havanensis 6 18 10 1 1 5.94 83.17 Stagnicola cp 6 2 5 2 6 2.43 34.00 Pomaceapaludosa 17 2 5 1 13 3.77 52.79 Paleomonetus paludosus 1733 481 19 345 150 375.90 5262. 35 Procambarus alleni 185 2626 1209 1048 569 618.64 8660.99 Fishes Lepisosteus platyrhinchus - - 2 - 2.00 28.00 Clumped P15 Ameiums natahs - - 1 1.00 14.00 Ameiurus nebulosus - 1 - 1 0.93 13.00 Clanas batrachus - 4 2 1 5 1.46 20.50 Cfnnodon variagams 1 - - - 1 0.93 13.00 Fundulus chgsoms 93 140 63 38 3 18.73 262.15 Cluntped MAR P50 Fundulusconfluentus 91 241 381 276 49 34.85 487.93 Clumped P75 SDM Jordanellaftoridae 162 207 98 19 38 28.49 398.90 Clumped MAR P50 ETable 2 Continued. Lucaniagoodel 39 13 13 4 10 5.84 81.82 Clumped MAR Belenosox betizanus - - 12 77 17 23.88 334.38 Clumped SDM Gambusia holbrooki 841 1418 903 150 491 142.44 1994. 13 Clumped MAR PSO Fls Hete¢«driaformosg 17 6 15 - 3 5.20 72.78 Clumped MAR f'75 Poecitialatipinna 112 183 175 2 33 33.22 465 .01 Clumped MAR P50 F75 Labidesthes sicculus 1 - 1.00 14.00 Enneacanthus gtoriosus 5 4 - - 1.62 22.67 Lepomis'guloms 2.8 4 7 4 1.43 20.00 A D A LR ¥k#LE : M E LA U U G A IN EVER G LAD ES 53 Lepomis macrochints 4 2 2 0.17 10.75 Lepomis marginatus · 1 1 - - - 0.93 13.00 Lepomismicrolophus 24 22 41 9 11 8 . 91 124.73 Clumped MAR P50 915 Lepomis puncrams 9 3 1 1 2.52 35.29 Clumped MAR Micropterussalmoides 1 1 - 0.93 13 .00 Etheostomafusgorme 113- 2.00 28.00 Clumped SDM Astronoms ocellams* 19 - - 2 17.12 239.71 Clumped MAR Hemichromisdetoutneauxi* 32 295 480 169 83 73.85 1033 .96 Clumped P50 P75 SDM Cichlasoma bimacularum* 23 227 119 150 167 '45.92 642:92 Clumped PSO SDM DMM Cichlasoma managuense* 12 4 4 7 5 1}. 80 25. 19 Tdapho hionae* - 1 3 1 2.00 28:00 Clumped P75 Amphibians Pseudhbranthits striatus 1 - - li.00 14.00 3 Siren laterrina 5 7 6 20 22 9.43 132.04 Clumped SDM DMM Amphiuma means 3 15 10 17 5 2.47 34.60 Clumped P50 SDM Notophrhalmus,viridescens 3 1-3 1.80 25.14 Bufo terrestris_ 9 - - 4 4 377 52.71 Clumped MAR Bufo quercicus 2 15 35 37 5 7. 42 103 .87 Clumped P15 SDM 54 B U LLE TIN FLO R ID A M U SEU M N ATU R A L H IS TO R Y VO L- 41(1) Table 2 Continued. Gastrophryne carolinensis 2 3 11 17 9 5.72 80.14 Clumped SDM DMM Eleutherodactylus planirostris* - - 4 110 53 75.79 1061 . 11 Clumped SDM DMM Pseudacris nigrita - 1 - - - 1 .00 14.00 Limnaoedus ocularis 1 - - - 1.00 14.00 Acris gryllus 5 4 4 - 1 1.76 24.57 Hyla cinerea 1 1 - 2 7 2.62 36.73 Clumped DMM Hyla squireUa 1 1 1 0.86 12.00 Osteopilus septentrionalis - - 2 3 5 1.64 23 .00 Rana sphenocephola 25 46 64 34 49 3.44 48.15 Clumped P50 P75 Ranago·ho 11 27 23 7 7 3.74 52.40 Clumped P50 P75 Reptiles Kinosternon bauri 20 7 5 - 3 2.99 41.88 Clumped MAR Terrapene carolina - 2 1 2 0.17 10.00 Chelydra serpentina 1 1 1 - 0.86 12.00 Anohs sogrei* 1 6 8 16 52 9.42 131.88 Clumped DAIA! Anolis carolinensis 6 5 5 1 - 2. 82 40 . 35 Clumped AIAR P50 Ophisaums compressus - 7 2 - 1 .80 25 . 14 Eumeces inexpectams - 1 12 3 1 8.56 119.77 Clumped P75 Nerodiafasciara 12 6 15 9 4 1.48 20.67 Nerodiafloridana 34 30 9 13 3 5.21 72.97 Clumped NIAR P50 Regina QUeni 14 15 4 3 1 3.99 55.84 Clumped A!.AR P50 Thamnophis sirralis 2 8 17 9 7 2.24 31.30 Clumped P75 Thamnophis sauritus 2 4 11 9 7 0.92 12 . 91 Diadophis punctatus 1 - - 1 1 0.86 12 .00 Farancia abacura - 3 4 - 1.80 25.14 Coluber constrictor 1 3 11 12 4 2.31 32.39 Clumped P75 SDI©i Table 2 Continued. Elaphegunata - 2 2 1 1 1.36 19.00 Lampropetrisgetula.#oridana - 2 3 - - 1.57 22.00 Agkistrodon piscivorous 1 5 12 3 4 2.89 40.40 Clumped P30 P75 Mammals Blarina carolinensis - - 1 DALRYMPLE & DALRYMPLE: MELALEUCA IN EVERGLADES W ETLANDS 55 Sigmodon hispidus - - 1 - 1 Oryzomys palustris - - - \ - All animals Number of species 53 53 57 52 57 Number ofindividuals 3707 6475 4200 2816 1989 Trap rate 7.72 13.49 8.75 5.87 4.14 Macroinvertebrates only Number ofspecies 9 9 9 10 10 Number of individuals 2055 3472 1596 1566 801 Traprate 4.28 7.23 3.33 3.26 1.67 Fishesonly Number ofspecies 21 18 19 16 20 Number of individuals 1490 2778 2319 915 926 Trap rate 3.10 5.79 4.83 1.91 1.93 Amphibians and Reptiles only Number ofspecies 23 26 28 24 26 Number of individuals 162 225 284 333 261 Trap rate 0.34 0.47 0.59 0.69 0.54 56 B U LLE TIN FLO R ID A M U SEU M N ATU R A L H ISTO R Y V O L 41(1) Table 3. Results ofbird strip transects summarized by cover type, 24«month cumulative numbers. Within each cover type, there were three replicates. Also indicated for each species are the Index of Dispersion, Chi-square value, and whether distribution wasclumped inany covertype. Covertypeabbreviations as in Table 1. MAR P50 P75 SDM DMM I Index 2 Covertypes Phalacrocorax aurims 1 - - 1.00 4.00 Ardea herodias 1 1 2 - 0.88 3.50 Butorides striatus 3 1 1 1 1.00 4.00 Flondacaerulea - 2 1 1.33 5.33 Casmerodius albus 3 6 - 4.00 16 .00 Clumped MAR P50 Hydranassa tricolor 1 1 0.75 3 .00 Myctena americana 1 1.00 4.00 Buteo lineatus - 2 2.00 8.00 Circus cyaneus 1 - 1.00 4.00 Fatco sparverius - 1 1.00 4.00 Gamnulachloropus 1 - 1.00 4.00 Capena gallinago 5 1 1 3.07 12 . 29 Clumped MAR Zenaida macroura 1 1.00 4.00 Columbina passerina - 1 - 1.00 4.00 Chorde,les minor 2 - 1 1.33 5.33 Megaceryle alcyon 22 1 .50 6.00 Colaptes aurams - 4 - - 4.00 16.00 Clumped I'75 Melane,pes carolinus 1 3 1 3 1.13 4.50 Sphyrapicus vanus - - 2 2.00 8.00 Picoides pubescens 1 1 1 - 0. 50 2 .00 Tyrannus verticalis 1 - 1 .00 4.00 Mplarchus ennitus - 2 2.00 8.00 Sayornisphoebe 4 1 3.00 12.00 Clumped P50 ·Table 3 Continued. Cyanocitta cristata - 1 5 5 5 1.94 7.73 Troglo*tes oedon - - 1 - - 1.00 4.00 77#yothorus tudevicianus - -14 13 8.42 33.67 Clumped SDM DMM Mimuspotyglorros - 1 5 3 .92 15 .67 Clumped Y75 Dumetellacarotinensis - - 23 2 .00 8.00 'Polioptila ca~dea 3 7 7 2 2.55 10.21 Lantus ludovicianus 1 3 6 1 2.59 10.36 Mreo griseus - - - 1 1 .00 4.00 Mnionita *aria - - 1 2 1.33 5.33 A LR Y M P LE *D A E R Y M P LE : M E ' a' vU C *IN E V E R G LA D E S W E TLA N D S 57 P,inda amencana - - 1 2 1.33 5.33 Dendroica coranata - 2 . 18 4 11 .92 47.67 Clumped Pls , Dendroica discolor - - 10 4 6.86 27.43 Clumped P75 Dendroicapalmarum 2 11 22 13 9 4. 59 18 . 35 Clumped P75 SDK{ Geotht*istrichas 57 20 12 4 2 26.42 105.68 Clumped MAR Icte#lai*~ns - 1 1 0.75 3.00 Setophaga nitic,80 - - 3 3.00 12.00 Clumpsd SDM Sturnella magna 8 26 28 - 15.23 60.90 Clumped P50 P75 Agelm'us phoeniceus 47 . 11 1 34.64 138.54 Clumped MAR Quiscalus major 4 21 - 16.60 66.40 Clumped P50 Quiscalus quiscula - 2 - 1 - 1.33 5.33 Cardinalis cardinalis - 1 5 4 1 1.14 8 . 55 Piptio erythrophthalmus - 1 1 5 3.01 12 . 29 Clumped SDM Melospiza georgiana 1 - 1.00 4.00 Number ofspecies 15 27 29 22 9 Number 6findividuals 137 127 146' 69 39 Across all five cover t>pes Number ofspecies 46 Number of indivduals 518 38 BULLETIN FLORIDA MUSEUM NATURAL HISTORY VOL 41(1) C Table 4. Scent and bait stations and small mammal live trapping results by cover type, 24-month cumulative numbers. Cover type abbreviations as in Table 1. MAR P50 f>75 SDM DMM Bait stations Didelphis virginiana + + + + Sigmodon hispidus - - - + Procyon lotor + + + + Urocyon cinereoargenteus - - + M ./ Canis familiaris + - + Felis n®s Number of species 3 2 3 5 Scent stations Didelphis virginiana - + Sigmodon hispidus - Procyon lotor + + Canis familiaris Felis rufus - + Odocoileus virginianus + Number ofspecies 2 2 3 3 Sherman live traps Sigmodon hispidus - 7 4 1 Oozomys polustris 3 4 2 2 Peromyscus gossypinus - - 3 Mus musculus - - Rattus rattus - - Number ofspecies 1 3 2 3 5 Number of individuals 3 11 6 6 5 Al 1 three methods combined Number ofspecies 5 5 6 7 10 Across all five cover types Number ofspecies 11 D ALRYM PEE : )@ iLALEU C A IN EVER G LAD ES W E 59 E] Table 5. Habitat association aild species composition ofamphibian and reptiles in each cover type based upon the cumulative totals ofnumber ofspesies and number of individuals trapped 6y drift fencing. Cover type abbreviations as in Table 1. MAR P50 P75 SDM DMM Site 123 123 1123 123 123 Number ofspecies Wetiand:dependent 12 14 12 13 15 14 18 15 17 12 13 13 7 15 13 Non-wetland 311 432 473 4 3 5 363 % Wetland dependent 80 931 92 76 83 8& 82 68 85 75 81 72 70 71 81 3 Numb~'of individuals ® Wetland dependent 76 45 32 59 59 83 89 78 73 58 47 84 34 60 55 Nan.wetlarid 711 14 7 3 7325 117 11 16 67 25 20 % Wetland dependent 92 98 97 81 89 97 93 71 94 33 81 84 34 71 73 60 BULLETIN FLORIDA MUSEUM NATURAL HISTORY VOL- 41(1) Table 6. Habitat association and species composition of birds in each cover type based upon the cumulative totals ofnumber of species and number of individuals observed during strip transects. Cover type abbreviations as in Table 1. MAR P50 P75 SDM DMM Number of species Wetland dependent 12 11 10 4 1 Non-wetland 3 16 19 18 8 % Wetland dependent 80% 41% 34% 18% 11% Number of individuals Wetland dependent 132 92 59 10 2 Non-wetland 5 35 87 59 37 % Wetland dependent 96% 72% 40% 14% 5% APPENDIX I Glossary of the scientific and common names of each vertebrate species found during the 24 months of the surveys in the Lake Belt Study Area, including areas otehr than the five defined cover types (e.g. canals, levees). Within each class, D ALR YM PLE & D ALR YM PLE: M E LA LE U C A IN EVER G LAD ES W ETLAN D S 61 species are listed alphabetically by scientific name. The Status column indicates whether a species is considered non-native in southern Florida. The GFC column indicates whether a species is listed as Endangered (E), Threatened (T) or Species of Special Concern (SSC) by the State of Florida Department of Game and Freshwater Fish Commission. The FWS column indicates whether a species is listed as Endangered (IE), Threatened (T) or Candidate for Listing (Cl or C2) by the United States Fish and Wildlife Service. For amphibians, reptiles, birds, and mammals only, the Habitat Assoc. column lists whether the species requires wetland habitats at some point in its life histoly for either reproduction, respiration, feeding mechanism or diet. For birds only, it is also indicated whether the species occurs in southern Florida all year (Resident), only during certain seasons (Winter or Summer), or passes through during spring and/or fall migration (Transient). In general, species designated "Resident" or "Summer" breed in southern Florida, although exceptions do exist. 62 B U LLE TIN FLO R ID A M U SEU M N ATU R A L H IS TO R Y VO L. 41(1) Habitat Scientific Name Common Name Status GFC FWS Association Season Fishes Ameiums narWis Yellow bullhead catfish Ameiurus nebuloms Brown bullhead catfish Amia calva Bowfin Astronorus ocellatus Oscar Non-native ~ Belonesox belizanus Pike killifish Non-native Cichla ocellaris Peacock bass Non-native Cichlasoma bimaculatum Blackacara Non-native Cichlasoma managuense Nicaraguan cichlid Non-native Clarias batrachus Walking catfish Non-native Cyprinodon variegatus Sheepshead minnow Enneacanthus gloriosus Bluespotted sunfish Etheostoma jusiforme Swamp darter Fundulus ch,ysotus Golden topminnow Fundulus confluentus Marsh killifish Gambusia holbrooki Mosquito fish Hemichromis letourneauri Jewelfish Nor,-native Heterandria formosa Least killifish Jordanella floridae Flagfish Labidesthes sicculus Brook silverside Lepisosteus platyrhinchus Florida gar Lepomis gulosus Warmouth Lepomis macrochirus Bluegill Lepomis marginatus Dollarsunfish Lepomis microlophus Redear sunfish Lepomis punctatus Spotted sunfish Lucama goodei Bluefin killifish Habitat Scientific Name Common Name Status GFC FWS Association Season Microptems salmoides Large mouth bass Mugil cephalus Striped mullet Poecilia latipinna Sailfin molly Titapia mariae Spotted tilapia Non-native Anwhibians Acris gryllus Southern cricket frog Wetland Amphiuma means Two-toed amphiuma Wetland DALRYM PLE & DALRYMPLE: M ELALEUCA IN EVERGLADES W E 63 Bufo quercicus Oak load Wetland Bufo terrestris Southerntoad Wetland Eleutherodactylus plamrostris Greenhouse frog Non-native Non-wetland Gastrophryne carolinensis Eastern narrowmouth toad Wetland Hyla cinerea Green treefrog Wetland Hyla squirella Squirrel treefrog Wetland Limnaoedus oculants Little grass frog Wetland Notophthalmus viridescens Peninsula newt Wetland Osteopilus septentrionalis Cuban treefrog Non-native Waland Pseudacris nigrita Florida chorus frog Wetland Pseudobranchus striatus Dwarfsiren Wetland Rana grylio Pig frog Wetland Rana sphenocephala Southern leopard frog Wetland Siren lacenina Greater stren Wetland Reptiles Agkistrodon piscivorous Cottonmouth Wetland Alligator mississippiensis American alligator SSC T Wetland Anolis carolinensis Green anole Non-wetland Anolis sagrei Brown anole Non-native Non-wetland Apaloneferox Florida softshell turtle Wetland 64 B U LLE TIN FLO R ID A M U SEU M N ATU R A L H ISTO R Y V O L 41(1) Habitat Scientific Name Common Name Status GFC FWS Association Season WetlandChelydra serpentina Florida snapping turtle Coluber constrictor Black racer Non-wetland Deirochelys reticularia Chicken turtle Wetland Diadophis punctat'us Southern ringneck snake Non-wetland Elaphe guttata Red rat snake Non-wetland Elaphe obsoleta Yellow rat snake Non-wetland Eumeces inapectatus Southeastern five-lined skink Non-wetland Farancia abacura Mud snake Wetland Gopherus polyphemus Gopher tortoise SSC C2 Non-wetland Kinostemon baurii Striped mud turtle Wetland WetlandI«ampropeltis getulafloridana Florida kingsnake Nerodia fasciata Florida water snake Wetland Nerodiaflondana Florida green water snake Wetland Nerodia taxispilota Brown water snake Wetland Opheodrys aetivus Rough green snake Non-wetland Ophisaurus compressus Island glass lizard (2 Non-wetland Pseudemysfloridana Peninsula water Wetland Pseudemys nelsoni Florida redbelly turtle Wetland WetlandRegina alleni Striped crayfish snake Terrapene carohna bauri Florida box tuttle Wetland Thamnophis sauritus Peninsula ribbon snake Wetland WetlandThamnophis sirtalis Eastern garter snake Birds Agelaius phoeniceus Red-winged blackbird Wetland Resident Ajaia ajaia Roseate spoonbill SSC Wetland Resident Ana, fulvizula Mottled duck Wetland Resident Wetland Resident Anhinga anhinga Anhinga Habitat Scientific Name Common Name Status GFC FWS Association Season Archilochus colubris Rubythroated hummingbird Non-wetland Winter Ardea herodias Great blue heron Wetland Resident Bubulcus ibls Cattle egret Non-wetland Resident Buteojamaicensis Red-tailed hawk Non-wetland Resident Buteo Iineatus Red-shouldered hawk Non-wetland Resident Buteo regalis Swainson's hawk Non-wetland Winter DALRYMPLE & DALRYM PLE: M ELALEUCA IN EVERGLADES W E 65 Butorides striatus Green heron Wetland Resident Cairina moschata Muscovy Non-native Wetland Resident Capella gallinago Common snipe Wetland Winter Cardinalis cardinalis Northern cardinal Non-wetland Resident Casmerodius albus Great egret Wetland Resident Cathartes aura Turkey vulture Non-wetland Resident Cathants guttatus Hermit thnish Non-wetland Winter Catoptrophorus semipalmatus Willet Wetland Winter Charadrius vociferus Killdeer Non-wetland Resident Chordeiles minor Con,inonn*tha,Nk Non-wetland Resident Circus cyaneus Northern harrier Wetland Winter Cistothorus palustris Marsh wren Wetland Winter Colaptes auratus Northern flicker Non-wettand Resident Colinus virginianus Bobwhitc quail Non-wetland Resident Columbina passerina Ground dove Non-wetland Resident Coragyps atran£3 Black vulture Non-wetland Resident Cyanocitta cristata Bluejay Winter Non-wetland Resident Dendroica coronata Yellow rumped warbler Non-wetland Dendroica nigrescens Black throated green warbier Non-wetland Winter Dendroica caerulescens Black throated blue warbler Non-wetland Winter Dendroica discolor Prairie warbler Wetland Resident 66 B U LLE TIN FLO R ID A M U SEU M N ATU R A L H IS TO R Y V O L 41(1) Habitat Scientific Name Common Name Status GFC FWS Association Season Dendroica palmarum Palm warbler Non-wetland Winter Dendroica striata Blackpoll warbler Non-wetland Transient Dendroica tigrina Cape may warbler Non-wetland Winter Dryocopus pileatus Pileated woodpecker Non-wetland Resident Dumetella carolinensis Gray catbird Non-wetland Winter Egretta thula Snowy egret SSC Wetland Resident Eudocimus albus White ibis SSC Wetland Resident Falco columbarius Merlin Independent Winter Falco sparverius American kestrel Non-wetland Winter Florida caerulea Little blue heron SSC Wetland Resident Fulica americana American coot Wetland Resident Gallinula chloropus Common moorhen Wetland Resident Geothlypis trichas Common yellowthroat Wetland Resident Himantopus mexicanus Black necked stilt Wetland Resident Hirundo nistica Barn swallow Non-wetland Tramient Hydranassa tricolor Tricolor heron SSC Wetland Resident Icteria virens Yellowbreasted chat Non-wetland Resident Lanius ludovicianus Loggerhead shrike C2 Non-wetland Winter Lophodytes cucullatus Hooded merganser Wetland Winter Megaceryle alcyon Betted kingfisher Wetland Winter Melanerpes carolinus Red-bellied woodpecker Non-wetland Resident Melosdittacus undulatus Budgegriar Non-native Non-wetland Resident Melospiza georgiana Swamp sparrow Wetland Winter Mimus polyglottos Northern mockingbird Non-wetland Resident Mniotilta varia Black white warbler Non-wetland Winter Mycteria americana Wood stork E E Wetland Resident Myiarchus crinitus Great crested flycatcher Non-wetland Resident Habitat Scientific Name Common Name Status GfC FWS Association Season Nyctanassa violacea Yellow crowned night heron Wetland Resident Nycticorax nycticorax Black crowned night heron Wetland Resident Pandion haliaetus Osi)rey Wetland Resident Partla americana Northern parula warbler Non-wetland Winter Passerculus sandwichensis Savannah spafrow Wetland Winter Phalacrocorax auritus Double crested cormorant Wetland Resident Picoides pubescens Downy woodpecker Non-wetland Resident Piranga rubra Summer tanager Non-wetland Transient D ALR YM PLE & D ALR YM PLE: M ELALEU C A IN EVER G LAD ES W ETLAN D S 67 Pipilo erythrophthalmus Rufousilded towhee Non-wetland Resident Plegadis fakinellus Glossy ibis Wetland Resident i Podilymbus podiceps Pied billed grebe Wetland Resident Polioptila caerulea Blue-gray gnatcatcher Non-wetland Winter Prothonotaria citrea Prothonotary warbler Nomwetland Transient Quiscalus major Boat tailed grackle Wetland Resident Quiscalus quiscula Common grackle Non-wetland Resident Rallus elegans King rail Wetland Resident Sayornis phoebe Eastern phoebe Non-wetland Winter Seiums aurocapillus Ovenbird Non-wetland Winter Setophaga ruticilla American redstart Non-wetland Winter Sphyrapicus varius Yellow-bellied sapsucker Non-wetland Winter Sterna albifrons Least tem T Wetland Summer Streptopelia decaocto Eurasian collared-dove Non-native Non-wetland Resident Sturnella magna Eastern meadowlark Wetland Resident Sturnus vulgam European starting Non-native Non-wetland Resider,t Thryothorus ludovicianus Carolina wren Non-wetland Resident Tringaflavipes Lesser yellowlegs Wetland Winter Tringa melanoleuca Greater yellowlegs Wetland Winter 68 B U LLE TIN FLO R ID A M U SEU M N ATU R A L H ISTO R Y V O L 4 Habitat Scienrvic Name Common Name Status GfC FWS Association Season Troglodytes aedon House wren Non-wetland Winter Turdus migratonus American robin Non-wetland Winter Tyrannui tyrannus Eastern kingbird Non-wettand Resident Tyrannus verticalis Western kingbird Non-wetland Winter lireo griseus White eye vireo Non-wetland Resident Vireo philadelphicus Philadelphia vireo Non-wetland Transient Zenaida macroura Mourning dove Non-wetland Resident Mammals Blarina carolinensis Southern short-tailed shrew Non-wetland Canis jamiliaris Domestic dog Non-native Non-wetland Dasypus novemcinctus Nine-banded annadillo Non-native Non-wetland Didelphis virginiana Virginia opossum Non-wetland Felis domesticus Domestic cat Non-native Non-wetland Felis ruji/ Bobcat Non-wetland Lutra canadensts River otter Wetland Mus musculus House mouse Non-native Non-wetland Odocoileus virginiamis White-tailed deer Non-wetland Oryzomys palu:tris Marsh rice rat Wetland Peromyscus gossypinus Cotton mouse Non-welland Procyon lotor Raccoon Non-wetland Rattus rattus Black rat Non-native Non-wetland Sigmodon hispidus Hispid cotton rat Non-wetland Sylvilagus palustris Marsh rabbit Wetland Urocyon cinereoargenteus Gray fox Non-wetland Contributions to the BULLETIN OF THE FLORIDA MUSEUM OF NATURAL HISTORY may be in any field ofbiology. Manuscripts dealing with natural history or systematic problems involving the southeastern United States or the New World tropics are solicited especially. Manuscripts should be of medium length-circa 35-200 pages (10,500-60,000 words). Examination for suitability is made by an Editorial Board. The BULLETIN is distributed worldwide through institutional standing orders and exchanges. It is considered the responsibility of the author to distribute his paper to all interested individuals. To aid in this, the author(s) receive(s) 50 copies free, and he/she(they) may purchase additional separates at cost if ordered when page proof is returned. The author is also responsible for any charges incurred for alterations rn.,1• by him on galley or page proofs. The Museum will send an invoice to the author for this amount upon completion ofpublication. PREPARATION OF MANUSCRIPT Contributors should consult recent numbers of the BULLETIN f6r preferred style and format Highly recommended as a guide is the Scientific Style and Format, CBE Style Manual for Authors Editors, and Publishers, 6th Edition, 1994 (published by the Council of Biology Editors). Manuscripts must be submitted in duplicate and satia& the following minimal requirements. They must be double-spaced throughout, including tables. figure captions, and literature citations. Figure legends and tables should be typed on separate sheets. Also, please submit a copy of the complete text tables, figure captions, and literature cited on a floppy disk (software used does not matter, but we use Word for Windows). All illustrations are ref6rred to as figures. They must comply with the following standards: Photographs should be sharp, with good contrast, and printed on glossy paper, or the originals submitted. If the background of photographs (especially those of specimens) is not desired, amberlith should be cut out and used to cover the backgrouncl Drawings should be made with dense black waterproof ink on quality paper or illustration board. All figures should have a cover sheet- All lettering will be medium weight sans-serif type (e. g. Futura Medium. News Gothic) in cutout, dry transfer, or lettering guide letters. Make allowance so that after reduction no lower case letter will be less than 1 mm high (2 mm is preferred) nor any capital letter greater than 5 mm high. The maximum size for figures is 9" x 14" (twice BULLETIN page size); figures should not be less than typepage width (414"). With soft lead pencil, on the back of each illustration, designate the top and identify each by author's name, manuscript title, and figure number. All manuscripts not submitted in BULLETIN format will be returned to the author for retyping and/or fonnajling Manuscripts, all editorial matters, and requests for more detailed preparation instructions should be addressed to: Managing Editor ofthe BULLE'nN Florida Museum of Natural History University ofFlorida P. 0. Box 117800 Gainesville FL 32611-7800, U.S.A Phone: 352-392-6724 FAX: 352-846-0287 email: rjbryant@flinnh.ufl.edu