xml version 1.0 encoding UTF-8 standalone no
record xmlns http:www.loc.govMARC21slim xmlns:xsi http:www.w3.org2001XMLSchema-instance xsi:schemaLocation http:www.loc.govstandardsmarcxmlschemaMARC21slim.xsd
leader nam 22 Ka 4500
controlfield tag 007 cr-bnu---uuuuu
008 s2010 flu s 000 0 eng d
datafield ind1 8 ind2 024
subfield code a E14-SFE0004621
Maintaining biodiversity with a mosaic of wetlands :
b factors affecting amphibian species richness among small isolated wetlands in central florida
h [electronic resource] /
by Jacquelyn Guzy.
[Tampa, Fla] :
University of South Florida,
Title from PDF of title page.
Document formatted into pages; contains X pages.
Thesis (MS)--University of South Florida, 2010.
Includes bibliographical references.
Text (Electronic thesis) in PDF format.
Mode of access: World Wide Web.
System requirements: World Wide Web browser and PDF reader.
ABSTRACT: The biodiversity value of a wetland is linked not only to its position in the landscape relative to other wetlands, but also to its habitat characteristics. I monitored amphibian species richness among 12 small, isolated, and undisturbed wetlands (which occur on lands permitted for phosphate mining) in central Florida during the 2005 and 2006 breeding seasons. I used seven habitat and landscape variables to characterize the environments of the wetlands and generalized linear models to determine which of these variables had the greatest influence on the occurrence of seven amphibian species (Anaxyrus terrestris, Gastrophryne carolinensis, Hyla gratiosa, Lithobates capito, L. catesbeianus, L. grylio, and Pseudacris nigrita verrucosa). Significant models for each species incorporated six of the seven habitat and landscape variables: distance to permanent water (2 spp.), distance to nearest wetland (3 spp.), vegetation heterogeneity (2 spp.), hydroperiod (2 spp.), presence/absence of fish (1 sp.), and distance to canopy cover (1 sp.). I suggest that source/sink metapopulation and patchy population dynamics in a given year are affected in part by environmental variables of ephemeral wetlands as they affect individual amphibian species. I suggest that a diversity of environmental conditions among wetlands produces the greatest amphibian biodiversity in this system, and that conservation and restoration efforts should emphasize environmental heterogeneity.
Advisor: Henry R. Mushinsky, Ph.D.
t USF Electronic Theses and Dissertations.
Maintaining biodiversity with a mosaic of wetlands: factors affecting amphibian species richness among small isolated wetlands in central Florida. by Jackie Guzy A thesis submitted in partial fulfillment of the requirements for the degree of Mast er of Science Department of Biology College of Arts and Science University of South Florida Co Major Professor: Henry R. Mushinsky, Ph.D. Co Major Professor : Earl D. McCoy, Ph.D. Committee: Peter D. Stiling, Ph.D. Date of Approval: June 30 2010 Keywords: amphibians, wetlands, source sink, metapopulation, patchy populations, stepping stone, wetland mosaic, ecological modeling, resource management, reclamation Copyright 2010, Jackie Guzy
Acknowledgments I would like to extend my deepest gratitude to Drs. Henry R. Mushinsky and Earl D. McCoy for providing me the wonderful opportunity of obtaining an advanced degree. It has been an incredible experience to have not only one, but two distinguished researchers from very different scientific backgrounds who are so innovative and accessible. I would like to extend this sentiment to my committee, Dr. Peter D. Stiling. It has been an absolute pleasure to work with yet another distinguished researcher with such diverse interests; I appreciate t he support and continuous feedback This research would never have been possible without the efforts of Shannon M. Gonzalez and J. Steve Godley (Biological Research Associates Inc. ) who worked with the natural resources staff of Hillsborough and Hardee Counties, Florida and with Ron Concoby ( Mosaic Fertilizer, LLC ) to obtain the extensive funding required for this project. I extend my deepest appreciation to you for such a wonderful opportunity. I cannot thank Travis Robbins enough for his assistance w ith statistical an alyses and ecological modeling or Kristan Robbins for her extensive assistance with the complicated inner workings of database management; you have been absolutely fantastic resources and great friends. Finally, I extend thanks many time s over to my friends and fellow biologists Annie Doyle, Brian Halstead, and Neal Halstead for their help in data collection and interpretation. Aside from myself, I trust no one else to drive 100 miles to a remote field site, retrieve a cassette tape, and then strap on a pair of headphones and translate several hours of mating frog vocalizations .
i Table of Contents List of Tables ........... ii List of Figures ........... i ii Abstract .. ........ iv Introduction .......... ... Methods .......... 8 Results .......... .. 3 Discussion ......... ... 18 Conclusions ......... ... 2 9 References ........ .. 3 6 Appendices Appendix 1: Wetland and landscape covariate values used in modeling amphibian species occurrence at twelve we tlands across west central Florida, 2005 and 2006 ...................... 4 5 About the Author .................................................................................................End Page
ii List of Tables Table 1. W etland and landscape covariates used in modeling amphibian species occurrence ..... 3 1 Table 2. Number of successful sampling nights, percent of time frogloggers functioned, and total number of minutes of amphibian breeding v oca lizations recorded per wetland .. 3 2 Table 3. Presence or absence of each amphibian species at study wetlands from June 2005 September 2006 ........................................ .. 3 3 Table 4. Candidate models constructed from generalized linear model of top three covariates against presence/absence of each s pecies. Bold models represent significant values as tested from Likelihood Type 3 tests .......... ................................................................ 3 4 Table 5. Dir ection of the relationship [positive (+); no relationship (nr); negative ( )] for each covariate as it relates to individual species and the Likelihood Type 3 tests for top three covariates used in building candidate models ........................... .............................. .. .......... 3 5
iii List of Figures Figure 1. Map of study wetlands located in Hillsborough and Hardee Counties, Florida, USA . 30
iv Maintaining biodiversity with a mosaic of we tlands : factors affecting amphibian species richness among small isolated wetlands in central Florida Jackie Guzy A bstract The biodiversity value of a wetland is linked not only to its position in the landscape relative to other wetlands, but also to its habitat characteristics. I monitored amphibian species richness among 12 small, isolated and undisturbed wetlands (which occur on lands permitted for phosphate mining) in central Florida during the 2005 and 2006 breeding seasons. I used seven habitat an d landscape variables to characterize the environments of the wetlands and generalized linear models to determine which of these variables had the greatest influence on the occurrence of seven amphibian species ( Anaxyrus terrestris, Gastrophryne carolinens is, Hyla gratiosa, Lithobates capito, L. catesbeianus, L. grylio, and Pseudacris nigrita verrucosa ). Significant models for each species incorporated six of the seven habitat and landscape variables: distance to permanent water (2 spp.), distance to neare st wetland (3 spp.), vegetation heterogeneity (2 spp.), hydroperiod (2 spp.), presence/absence of fish (1 sp.), and distance to canopy cover (1 sp.). I suggest that source/sink metapopulation and patchy population dynamics in a given year are affected in part by environmental variables of ephemeral wetlands as they affect individual amphibian species. I suggest that a diversity of environmental conditions among wetlands produces the greatest amphibian biodiversity in this system, and that conservation and restoration efforts should emphasize environmental heterogeneity.
1 Introduction Amphibian declines are well documented (Blaustein and Wake 1990, Phi l lips 1994, Stuart et al. 2004) ; one third of all amphibians are now considered th reatened (Stuart et al. 2004) and 168 species have become extinct within the last two decades ( Dodd 2009 ) These declines are no longer considered natural population fluctuations (Dodd 2009) but rather caused by a wide range of human induced factors This loss of biodiv ersity is of concern as it influences economics, ecosystem function, esthetics, and ethics (Noss and Cooperrider 1994, Groom et al. 2006). Habitat loss and degradation are now considered among the greatest threats to amphibians worldwide (Cushman 2006 Dodd 2009 ). Among some of the most critical habitat to amphibians are small, iso lated wetlands, which are used to support their biphasic life histories. Wetlands of all types are declining worldwide to facilitate draining or filling for human settlements and agriculture, and small, isolated wet lands are the least protected. As a result of a 2001 Supreme Court decision, Solid Waste Agency of Northern Cook County (SWANCC) vs. U.S. Army Corps of Engineers a significant number of wetlands and other waters t hroughout the United States are no longer protected under the Clean Water Act (SWANCC 2001 Comer et al. 2005 ) The SWANCC decision eliminated reliance on the Migratory Bird Rule that included many geographically isolated wetlands within the jurisdiction of the Clean Water Act through their linkages to interstate commerce (Downing et al. 2003). In 2006, Court decisions in Rapanos v. United States and Carabell v. Corps further restricted federal authority over wetlands not directly connected (via surfi cia waters of the United States further undermining remaining federal jurisdiction over isolated wetlands
2 (BenDor 2008) Thus, i n Florida, where my study was conducted, federal regulation of impacts to wetlands is restricted to t hose greater than 1 acre which have significant connection to navigable waters of the state (e.g. rivers and streams). State regulation of alterations to natural, small, isolated wetlands in Florida varies by water management district and is generally res tricted to those greater than one half acre [ F.A.C. 62 340, South West Florida Water Management District ( SWFWMD ) 2009 ] Thus, these critical amphibian habitats continue to be lost as development in Florida escalates. One particular type of disturbance in central Florida is phosphate mining, which disturbs about 2 ,0 23 2 428 hectares of land annually; approximately 25 30% of these lands are isolated wetlands or wetlands hydrologically connected to navigable waters [Florida Institute for Phosphate Researc h ( FIPR ) 2010] Florida provides approximately supply (FIPR 2010). The mining of phosphate for fertilizer is typically conducted using strip mining techniques which involves clearing the site of all vegetation, removal of soil, and mining the underlying phosphate matrix with draglines F ollowing extraction, the site is back filled with sand tailings ( FIPR 2010). Because of the large scale clearing, mining and reclam ation in central Florida, recent emphasis has been on improvement of reclamation techniques for the purpose of maintaining a diverse flora and fauna after mining (Durbin et al. 2008, FIPR 2010). State law requires that land disturbed by phosphate mining m ust be restored to a useful condition, and sometimes to where ecological systems function as they did before the mining (FIPR 2010). In the legislation creating the Florida Institu te for Phosphate Research is the mission that includes the study of reclama tion alternatives and technologies. Goals of FIPR include developing methods to improve wildlife habitat on reclaimed mined lands and facilitate recolonization by wildlife. Thus, it is vital to identify the factors of native wetlands that support amphibi an diversity to assist resource managers.
3 Small, isolated wetlands are critical for amphibians because t hese wetlands dry throughout the year and thus cannot support predatory fish and perhaps support a lesser abundance and diversity of invertebrate pre dators that may consume amphibian larvae (Morin 1983 Wilbur 1987 Semlitsch et al. 1996). Most amphibians are regarded as highly philopatric and because dispersal distance is generally <0.3 km (Gibbs 1993, Semlitsch 1998, Semlitsch and Bodie 1998, Semlit sch and Bodie 2003) they are generally considered to have poor dispersal abilities (Ma rsh et al. 1999 Semlitsch 2000 Smith and Green 2005) despite evidence of long distance movement (1.0 1.6 km) in some frogs and toads (Lemckert 2004) Furthermore, thes e wetlands harbor large numbers of species of other taxa that are less mobile than birds and mammals (Semlitsch and Bodie 1998) and are more affected by their los s; species include wetland plants such as sundew ( Drosera spp ) and pitcher plants ( Sarracenia spp.; Sharitz and Gibbons 1982), microcrustaceans (Mahoney et al. 19 90 ), and aquatic insects (Kondratie ff and Pyott 1897, Sharitz and G ibbons 1982, Gaddy 1994). Perhap s most important is the small, isolated wetlands aggregate role in protecting wetland dependant species through either source sink dynamics (metapopulations) and/or patchy populations A metapopulation is a collection of partially isolated breeding habitat patches, connected by occasionally dispersing individuals where each patch exists wi th a substantial extinction probability; long term persistence occurs only at the regional level of the metapopulation (Smith and Green 2005). Because each wetland in an area may fluctuate in the number of indiv iduals of a species it contains, at times a wetland may act as a sink when the population of a species dies out locally from that wetland, or it may be a source that produces surplus individuals, which can colonize a nearby sink wetland (Semlitsch 2000) An alternative to metapopulation structure a t the local level is the existence of patchy populations that treat local wetlands as habitat patches. Patchy populations often occur where many
4 wetlands are in close proximity to one another and facilitate adaptive habitat switching; movements between we tlands occur at such high rates that local wetland populations do not develop a significant degree of demographic independence (Harrison 1991, McCullough 1996, Smith and Green 2005 Petranka and Holbrook 2006 ). Regardless of how amphibian populations ar e structured, small, isolated wetlands are critical for breeding success of many species. Because these wetlands are more affected by disturbance, and thus more vulnerable than larger wetlands, c onsequences to losing them include alteratio ns to metapopula tion or patchy population dynamics ; two main effects include the reduction of individuals dispersing and the increase in dispersal distances (Gibbs 1993) This loss reduces the total number of sites in which wetland breeding amphibians can reproduce and s uccessfully recruit juveniles (Semlitsch and Bodie 1998) and ultimately can reduce the number of source populations because juvenile recruitment is related to hydroperiods that favor the periodic drying characteristic of small wetlands (Pechmann et at. 198 9 ). D espite support in the literature for the biological importance of small, isolated wetland s ( e.g. Semlitsch and Bodie 1998, Gibbs 2000, Snodgrass et al. 200 0 Paton and Crouch 2002, Comer et al. 2005 ), they rema in unprotected from disturbance Lands cape ecology, conservation biology and restoration ecology aim to promote better management of natural resources including biodiversity and a large literature (e.g. Wiens and Moss 2005 ; Lindenmayer and Fischer 2006) has resulted. M any studies have focuse d on individual patches of habitat or sites within those patches, but patch size effects cannot be divorced from other critical issues such as the role of patch mosaics, a topic poorly understood (Bennett et al. 2006). I suggest that wetland and amphibian conservation would be best guided by landscape conservation that includes a mosaic approach rather than an individual site or patch approach
5 Gibbs (1993, 2000) emphasizes the necessity of evaluating wetland resources as a mosaic rat her than as isolated entities; as human populations shift from rural to urban landscapes, wetland spatial patterns go from many clustered wetlands (2 5 wetlands/km2, 0.2 0.4 km apart) to fewer, more isolated wetlands (<1 wetland, >0.5 km apart). Gibbs found that wetland mosai cs could withstand only modest losses and still provide wetland densities that are minimally sufficient to maintain wetland biota ; wetland mosaics characterized by <1 wetland per km 2 and >0.5 km from other wetlands were not able to sustain metapopulations of wetland dependent animals. P resent understanding of the traits of wetland mosaics important to sustaining metapopulations or patchy populations of wetland organisms and how those traits are altered by mounting wetland destruction and by regulations in tended to restrict it is minimal (Semlitsch and Bodie 1998, Gibbs 2000). Previous studies have explored the importance of the density and distribution of wetlands (Laan and Verboom 1990 Gibbs 1993 Semlitsch and Bodie 1998, Marsh et al. 1999), but only t wo exist that focus on the structure of an entire network or mosaic of wetlands in an area and its role in amphibian persistence. Fortuna et al. (2006) found that the observed spatial structure of ponds in Spain is robust to drought, allowing the movement of amphibians to and between flooded ponds, and hence, increasing the probability of reproduction even in dry seasons. Gmez Rodrguez et al. (2009 ) investigate d spatial and temporal variation in amphibian breeding habitats in Spain during two different hydrologic cycles and found that a large and diverse network of ponds provides different habitat opportunities each year, favoring the long term persistence of the whole amphibian community. There have been a limited number of studies (Bennett et al. 20 06) that present empirical data on the response of one or more faunal groups to agricultural land mosaics ( an area of land containing multiple different landuses ) Studies on faunal responses to land mosaics in forested (McGarigal and McComb 1995, Edenius and Elmberg 1996,
6 Jokimaki and Huhta 1996, Hargis et al. 1999, Gjerde et al.2005), and experimental model landscapes (Collins and Barrett 1997; Collinge a nd Forman 1998 Parker and Mac Nally 2002 With et al. 2002) are very limited and all studies are of i nsects, birds, and mammals. In this study, a mosaic is defined as a group of small, isolated wetlands with different landscape and wetland characteristics. Functioning as stepping stone s wetland mosaics are important buffers against yearly environ men tal variation. E phemeral wetlands act as stepping stones during years with less rainfall, and can link a large number of dry wetlands with short hydroperiods to those with longer hydroperiods that contain water, and thus favor amphibian persistence and/or dispersal. Thus, I suggest p reserva tion of the mosaic with a range of differing ephemeral wetlands intermingled in the landscape is essential to maintain the biodiversity which the network of wetlands supports. My study is designed to elucidate the la ndscape and wetland characteristics of amphibian breeding habitats (small, isolated wetlands) in west central Florida which yield the greatest species richness. Given the high rate of wetland disturbance and/or elimination of small, isolated wetlands in F lorida I investigate which types of wetlands sustain the highest species richness and present them as a target for preservation and/or restoration goals. I hypothesize that while some of the wetlands have lower richness in a given year, the overall high species richness observed at these sites is a result of the presence of a mosaic of small, isolated wetlands with varying landscape and physical characteristics which act as buffers to breeding amphibians against yearly environmental variation. I studie d seven wetland and landscape variables including (1): major structuring factors of wetland communities such as area ( Beja and Alcazar 2003 Burne and Griffin 2005 Werner et al. 2007), fish presence ( e.g. Heyer et al. 1975 Hecnar and M'Closkey 1997 ), ve getation heterogeneity ( Atauri and Lucio 2001, Tews et al. 2004) and
7 hydroperiod (Beja and Alcazar 2003 Snodgrass et al. 200 0 Gonzales 2004, Werner et al. 2007 ) and (2) landscape features important to metapopulations/patchy populations including distanc e to canopy ( deMaynadier and Hunter 1999 Herrmann and Babbitt 2005 ) distance to permanent water ( Dickman 1987 McComb et al. 1993 Semlitsch and Bodie 2003 ) and distance to nearest wetland ( e.g. Vos and Stumpel 1995 Halley et al. 1996 Semlitsch 2000 ).
8 Methods The study areas are located in west central Florida (Hillsborough and Hardee Counties, Figure 1) a n area of high diversity of amphibians that supports 18 of the 31 species of anurans native to the state [ F lorida M useum of N atural H istory ( FMNH) 2010 ] I surveyed 12 native ephemeral wetlands ranging from 0. 1 3 3 acres which occur on lands permitted for phosphate mining. Automated Frog Call Recorders (frogloggers) were installed in each wetland (Barichivich 2003) to monitor frog calling activities during the summer and winter breeding seasons of 2005 and 2006 At the onset of wetland flooding frogl oggers were set to record anuran calls for one minute each evening hour between 1800 0600 hours, each night from 8 June 23 August (2005) a nd 28 June 4 September (2006) for a total of 140 sampling nights (70 sampling nights each year ) This timeframe encompasses the breeding season for all species occurring in the study area [ N orth American Amphibian Monitoring Program (NAAMP) 2010 ] The majority of r ecordings (90%) were interpreted by me and the remaining 10% were interpreted by two other s also trained in Central Florida frog vocalizations and experienced with biology of amphibians Calling male anurans were identified and their choruses were placed into one of four size categories according to according to the North American Amp hibian Monitoring Program (NAAMP 2010) A calling index of zero indicated that no individuals were heard. An index of one indicated that individuals could be co unted but there was time between calls. A calling index of two indicated that calls of individuals could be counted, but there was some overlap, and a calling index of three indicated that there was a full chorus of constant and overlapping calls. To ens ure that frogloggers were
9 detecting the frequency range for calls of all species, wetlands were visited 3 5 times late in the evening during peak breeding season each for five minutes. Data were compared to that collected on froglogger cassette tapes. No additional species were heard during these visits and anuran choruses were comparable to those captured on the frogloggers. Amphibian assemblages within a given wetland are highly dynamic from year to thi s study permitted collection of data during two hydrologic extremes with differing rainfall amounts and timing seasonal extremes. The 2005 study season was characterized by heavy rain and elevated water levels resulting from a particularly severe H urrica ne S eason in fluencing the study area in 2004 [ N ational O ceanic and Atmospheric A dministration (NOAA) 2010 ] while the 2006 season was characterized by infrequent rains and the beginning of a two year drought which had the driest back to back calendar years Florida has experienced, since 1932 ( F DEP 2010 ). Wetlands were characterized using seven habitat variables (Table 1) including area, distance to canopy, distance to permanent water, distance to nearest wetland, fish presence, vegetation heterogeneity, a nd hydroperiod. To obtain landscape variables, I used georeferenced digital 1:100,000 USGS geological Orthophoto Quarter Quadrangle maps ( based on 2004 aerial photographs ) along with National Wetlands Inventory and Florida Rivers shapefiles, each obtained fr om the Southwest Florida Water Management District ( SWFWMD 2010) to build a geographical information system in Arc Map 9.3.1 [ E nvironmental S ystems R esearch I nstitute (ESRI) 2009 ] Hydroperiod was determined from weekly site visits to each wetland and i s considered to be the length of time surface water inundates the wetland. Fish presence was determined using active and passive methods. Each month, u sing D frame dip nets, five 1 m sweeps were conducted in each microhabitat proportional to the fraction of the total area of the wetland that each
10 microhabitat covers (Mushinsky et al. 2004). Passive sampling was performed monthly, using four unbaited minnow traps placed haphazardly throughout each of the wetlands for a period of 24 hours. Vegetation hete rogeneity was determined after extensive surveys by placing each wetland in one of three categories based on plant species that occur within and immediately surrounding the wetland; categories include : herbaceous, herbs and shrub s, and herbs, shrubs, and t rees. For calling amphibians, non detection of the species is not equivalent to absence. Detection probability varies because certain species are not conspicuous and also because of seasonal behavi or patterns, changing environmental conditions, habitat quality, and sampling technique s ; t hus it is necessary to estimate the proportion of sites occupied when species detection probabilities are less than one (MacKenzie et. al 2002, Bailey 2004) I use d the mark recapture like approach of MacKenzie et al. (2002) as implemented in the program PRESENCE (available for download from htpp://www.proteus.co.nz/) to estimate the proportion of sites occupied by each species, accounting for imperfect detection. Because sampling occasions were so numerous (n= 140; 70 each for 2005 and 2006), presence data were sparse relative to absence data. This sparseness often occurs from ove r sampling and even when collapsing /pooling sampling occasions, estimates are unrelia ble Therefore, I ran the simplest occupancy and detection model ( psi(.),p(.) ) for each species to retain the most basic estimates of occupancy given detection ; this model assumes the probability of occupancy and detection for each sampling night is the s ame. Any estimates over 50% were considered in further modeling. For example, if a species was estimated to occur at a site with a 65% chance, and it likely occurred there based on its biology and my extensive site knowledge, I added the species to the s ite. I used generalized linear models (GLZM) to determine which landscape and wetland variables (Table 1) had the greatest influence on individual amphibian species
11 occurrence Models were fit by maximum likelihood and the significance of individual param eters was tested with likelihood ratio tests based on Type III (non order dependant) sums of squares using STATISTICA 7.1 (StatSoft, Inc 2005). Individual amphibian species occurrence was analyzed using a binomial regression (used with presence/absence da ta) and logarithmic link function (McCullagh and Nelder 1989). In multiple regression, co llinearity between predictor variables can con found their independent effects; t herefore, prior to our regression analysis I calculated Pearson correlation coefficien ts for all pairwise combinations of independent variables (Hair et al. 1998, Knapp et al. 2003). Correlation coefficients for three of the seven variables ranged between 0.46 0. 51 and thus w ere included in subsequent mode ling Distance to permanent wate r and fish presence were strongly correlated (r=0.82) as were area of wetland and average hydroperiod (r=0.62). During model building (see below), if either pair of these correlated variables was shown to be important, the stronger of the two was selected Exploratory univariate GL Z were run to assess the im portance of measured habitat variables at each wetland. The resulting models for each species included all possible combinations of the top three covariates with the lowest or significant p values Because my sample size was limited to twelve wetlands adding more than three variables would have overparameterized the models (Doherty and Grubb 2002). Thus, for each species, a resulting seven candidate models were obtained from all possi ble combinations of the top three covariates with low or signif icant p values. I Mazerolle (2006) recommends for herpetological studies; models with lower AIC values are assumed to explain variation in data better (Burnham and Anderson 1998 ). I selected the models with substantial empirical support given the data (model AIC minAIC / 2, following Burnham and Anderson 2002). For each species, only models with
12 AIC values tha t differ .0 were considered in model selection; in I chose the one with the fewest parameters (most parsimonious) as the one best explaining the data (Burnham and Anderson 1998).
13 Results Froglogg ers installed in 12 native, ephemeral wetlands recorded a total of 1 7 ,7 60 minutes of frog calls during peak breeding season in 2005 and 2006 ( Table 2 ). Because of equipment failure, 11% of minutes recorded were unusable, so 15,756 minutes were analyzed. Sampling nights averaged 123 per wetland and ranged from 66 135 and frogloggers functioned between 73 94% of the time (Table 2 ). Fourteen amphibian species were present at the study sites from June 2005 September 2006 (Table 3 ). Amphibian s pecies richness among wetlands ranged from 8 13. Six species occurred at all wetlands : the southern cricket frog ( Acris gryllus dorsalis ), oak toad ( Anaxyrus quercicus ), green tree frog ( Hyla cinerea ), pinewoods tree frog ( Hyla femoralis ), southern leopard frog ( Lithobate s spehnocephela ), and little grass frog ( Pseudacris ocularis ). Another commo n species was the squirrel tree frog ( Hyla squirella ); i nitially occuring at 11 of 12 wetlands. Because occupancy estimates obtained from PRESENCE indicated a 95% chance of occupa ncy which I found biologically probable, it was assumed present at all wetlands. All other occupancy estima tes obtained from PRESENCE ranged from 5 36% and combined with the intense sampling of each wetland, warranted confidence that species were not pre sent Species not occurring at every wetland (Table 3 ), and thus included in modeling were the southern toad ( Anaxyrus terrestris ), eastern narrow mouth toad ( Gastrophryne carolinensis ), barking treefrog ( Hyla gratiosa ), gopher frog ( Lithobates capito ), bullfrog ( Lithobates catesbeianus ), pig frog ( Lithobates grylio ), and southern chorus frog ( Pseudacris nigrita verrucosa )
14 Landscape and wetland variable measurements differed markedly between wetlands With the exception of one outlier in distance to can opy, there was even spread among all variables. Distance to permanent water averaged 0.64 km and ranged from 0.1 1.35 km. Distance to canopy cover averaged 145.6 m and ranged from 3 853m. Wetland a rea averaged 0.48 hectares with a range between 0. 04 1 3 4 acres. D istance to nearest wetland averaged 261 m and ranged 44 765 m (see Appendix 1) Among wetlands, seven (58%) contained fish and five (42%) did not. Wetland hydroperiod during the breeding season averaged 6.42 weeks and ranged from 2 10 weeks ( see Appendix 1). Vegetation heterogeneity within wetlands ranged from 1 3 with five wetlands earning a score of 1 (herbaceous cover), three wetlands earning a 2 (herbaceous and shrub cover), and four wetlands earning a score of 3 (herbs, shrubs, and tree coverage). Wetlands had a high diversity and abundance of herbaceous groundcover species including grasses, sedges, and flowers (e.g. various species of Andropogon Panicum Spartina Juncus Ilex Xyris Rhynchospor a Eleocharis Aesclepias Rhexia Dros era Sagittaria Pontedaria Cladium and Cyperus ). Shrubs included Serenoa repens Hypericum spp Bacchari s halimifolia Cephalanthus occidentalis Ludwigia spp and Myrica cerifera. Trees within wetlands were few and limited to individuals of Nyssa sy lvatica and Quercus laurifolia ; distance to canopy coverage was measured and tree species included those of xeric and mesic hardwood hammocks, predominantly Quercus spp. For six of the seven species the best model selected was significant ly better than the intercept only model (p<0.05 Table 4 ). The remaining species (pig frog) was marginally significant (p=0.058). Although AIC is a robust method for model selection (Burnham and Anderson 1998), I also employed Type 3 likelihood ratio tests to test whic h of the three selected factors (for each species) significantly affected the model (Table 5 ) For five species ( Anaxyrus terrestris, Gastrophryne carolinensis, Hyla
15 gratiosa, Lithobates capito, and L. grylio ), the factors composing the best model signifi cantly affected the model; for Pseudacris nigrita verrucosa only one of the two factors in the best model also significantly affected the model ( Table 5 ). The best model for the southern toad was one that incorporated distance to permanent water, distan ce to nearest wetland, and vegetation heterogeneity ( Table 4 ) A l ikelihood ratio test found all factors in the best model were also significant (p 0.0 3 8 Table 5 ) Response plots were created for these significant factor s and the resulting direction of the relationship suggests that that southern toad occupancy decreases with increasing distance from permanent water (Table 5 and Appendix 1 ) While vegetation heterogeneity and distance to nearest wetland appear to affect o ccupancy in conjunction with distance to permanent water, there is no clear positive or negative relationship. The best model for the eastern narrow mouth toad was one that incorporated only the average hydroperiod between the two sampling years (Table 4 ) This factor was also significant when using a likelihood ratio test (p=0.017, Table 5 ) The next best significant with hydroperiod ; this factor, however, was not significant with a likeliho od ratio test (p=0.72). The direction of the relationship s uggests that eastern narrow mouth toad occupancy increases when hy droperiod is relatively short ( 2 8 weeks; Table 5 and Appendix 1). The distance to nearest wetland was the only factor included in the best model for the barking treefrog (Table 4 ) This factor was also significant when using a likelihood ratio test (p=0.034, Table 5 ). The direction of the relationship suggests barking treefrog occupancy increases when distance to the next wetland is short (within 1 6 0 m ; Table 5 and Appendix 1). The best model for the Florida gopher frog was one that included distance to nearest wetland and fish presence/absence as factors (Table 4 ). All other models had
16 t model were also significant 5 ). The direction of the relationship suggests that Florida gopher frog occupancy increases when wetlands are nearer to other wetlands (within 16 0 m ; Table 5 and Appendix 1 ) While fish presence appears to af fect occupancy in conjunction with distance to nearest wetland, there is no clear positive or negative relationship. Distance to canopy coverag e (with and without the outlier at wetland G Appendix 1 ) was the only factor included in the best model for the bullfrog (Table 4 ). This factor was not significant when using a likelihood ratio test (p=0. 22 Table 5 ). The next best fish presence/absence along with distance to canopy coverage ; this factor, however, was also not significant with a likelihood ratio test (p=0. 18 ). The direction of the relationship suggest s a trend for the bullfrog to be positively associated with wetlands that are closer to canopy and support fish populations because the AIC method found these factors to be significantly better than the intercept only model. The distance to permanent wate r was the only factor included in the best model for the pig frog (Table 4 ), however the model was marginally significantly better than the intercept only model (p=0.058). Conversely, this factor was significant when using a likelihood ratio test (p=0.016 Table 5 ). The direction of the relationship suggests pig frog occupancy increases with increasing distance (>0.4 km ; Appendix 1 ) to permanent water (Table 5). The best model for the southern chorus frog was one that incorporated only the average hydrope riod between the two sampling years and vegetation heterogeneity (Table 4 ). Only vegetation heterogeneity was significant when using a likelihood ratio test (p=0.004, Table 5 ) The n distance to nearest wetland with hydroperiod and vegetation heterogeneity ( Figure 4 ) ; this additional factor, however, was not significant with a likelihood ratio test (p=0.59 ;
17 Table 5 ). The directi on of the relationship for the significant model suggests that southern chorus frog occupancy increases with increasing vegetation heterogeneity
18 Discussion My study suggests that a range of wetland and landscape characteristics resulting in a mosai c of wetlands are important in supporting amphibian diversity through a stepping stone array The models developed provide insight to the limited knowledge base on the structure of a mosaic of wetlands in an area and its role in the factors influencing am phibian occupancy of wetlands. Species not occur ring in all habitat types can be very informative when quantifying habitat value, and in this particular study, of the 14 frog species observed, seven were absent from four or more wetlands. The southern toad appears to be influenced negatively by increasing distance to permanent water. Of the twelve wetlands, the southern toad was found at seven wetlands, four of which were 0.1 0.19 km from a permanent water source; the remaining three wetlands were 0.4 2 1.11 km and toads were not present at wetlands 0.97 1.35 km from permanent water. In part, proximity to permanent water appears important to the toad, perhaps because of its life history. They breed in both temporary and permanent aquatic habitats (Gib bons and Semlitsch 1991) and are unpalatable or toxic to many potential predators, including fish (e.g. Lefcort 1998). Following transformation and prior to emigration, juvenile southern toads forage for several weeks around the edge of the pond from whic h they emerged (Beck and Congdon 1999). Southern toad home range may encompass an area 1.6 km wide (Bogert 1947); they can travel further distances than frogs as they are better able to regulate water loss. Perhaps given their unpalatability and migratio n/dispersal abilities, they can afford to occupy wetlands closer to permanent water sources and risk occasional fish invasion during sheet overflow events because of the advantage conferred by water permanancy
19 Eastern narrow mouth toad occupancy was si gnificantly influenced by shorter hydroperiods. F emale G. carolinensis deposit a small sheet of eggs on the water's surface in highly ephemeral pools of water (Wright 1932 Wright and Wright 1949 Gibbons and Semlitsch 1991). Larval development is rapid and complete metamorphosis has been reported to occur in 6 10 d ays (Anderson 1951) but can also occur 20 70 days after egg deposition (Wright 1932 Martof et al. 1980) and has been also been reported to report ed complete metamorphosis in 30 days (Donnelly 1997) G. carolinensis is the only species in this study to metamorph so quickly and also the only one to be significantly affected by shorter hydroperiods The eight wetlands occupied by breeding G. carolinensis had hydroperiods ranging from 14 56 days five of which held water less than 35 days ( Appendix 1 ) which coincides with its breeding phenology. Barking treefrog occupancy was significantly negatively influence d by increasing distance to next nearest wetland. In part, distance to next nearest wetland appears important to the frog likely because of its life history. Adult H. gratiosa do not migrate seasonally, but remain in the vicinity of breeding wetlands when not engaged in calling or reproduction in water (Neill 1952, 1958). Murphy (1994 ) reported movements of 100 m between breeding ponds by several males in Florida; of the twelve wetlands in my study, H. gratiosa only occurred where distance to next wetland was within 16 0 m. Murphy et al. (1993) suggest multiple ponds in the landscape s hould be protected to allow dispersal because H. gratiosa migrate among breeding sites. Gopher frogs are considered Endangered, Threatened, or of Special Concern in all of the states within their range (Mount 1975 Martof et al. 1980 Moler 1992, Levell 1997). In thus study, L. capito occupancy was significantly negatively influenced by increasing distance to nearest wetland. Several migrations may occur throughout the breeding season resulting in the use of multiple wetlands for breeding, with males
20 a rriving at reproductive sites prior to females and remaining there longer (Bailey 1991). While L. capito have consistently been reported to move long distances from breeding wetlands to upland retreats which is important for conservation efforts (one indi vidual moved 2 km, Franz et at. 1988, see also Roznik et al. 2009), during the breeding season they may move among ponds in close proximity, which has been observed in other pond breeding amphibians (Semlitsch 2008). Roznik et al. (2009) found support for this hypothesis in their study where radio tagged adult frogs oriented toward breeding ponds within 300 m and an adult frog captured at one wetland was recaptured at a nearby pond the next year. L. capito in our study were only present at wetlands wi thin 16 0 m of another wetland and our modeling results suggests that during the breeding season, L. capito occupancy increases in part when wetlands are nearer to other wetlands (within 16 0 m ). Perhaps this is due to a confluence of unexplained reasons ; L capito might be affected at the within beneficial to have other wetlands nearby. Distances to canopy cover and fish presence were incorporated in the best models for the bullfrog; however, th ese factors were not significant when using a like lihood ratio test. T rend in the data exists however, for the L. catesbeianus to be positively associated with wetlands that are closer to canopy ( generally within 20 m ; Appendix 1 ) and also support fish po pulations. This finding could be explained by the increased transpiration rates of wetlands with high hardwood density nearby; if wetlands can withstand high transpiration rates and still support fish populations, the hydrology is likely also suitable for L. catesbeianus tadpoles. The time to metamorphosis for these frogs is among the longest (to confer greater fitness through larger sizes) and varies from a few months in the south in temporary wetlands to 3 yr in Michigan and Nova Scotia (Collins 1979 B ury and Whelan 1984) wher e they must over winter. Unlike many other frogs, bullfrogs can coexist with predatory fishes (Hecnar 1997) as tadpoles are
21 relatively immune to fish predation because of unpalatability (e.g. Walters 1975 Werner and McPeek 1994) and are one of only a few species likely to persist after fish invasion (Seale 1980). The distance to permanent water was the only factor included in the best model for the pig frog and while only marginally significantly better than the intercept only m odel, this factor was significant in a likelihood ratio test. A trend suggests pig frog occupancy i ncrease s with increasing distance (>0.4 km ; Appendix 1 ) to permanent water While L. grylio opportunistically use ephemeral wetlands, they are largely aqua tic, typically remaining within permanent water habitats throughout the year (Wright 1932 Wright and Wright 1949 Lamb 1984) and tadpoles require comparatively longer to metamorphose [ up to 365 days in Florida; ( Donnelly 1997 ) and 365 730 days further nor th ( Wright 1932 Wright and Wright 1949 Dundee and Rossman 1989) ] Wood et al. (1998) found that pig frogs tend to remain in one location when food and water conditions are suitable, but that substantial movement is possible when water conditions change. Thus, the importance of increasing distance to permanent water is perhaps just an artifact of expected natural fluctuations of amphibian populations and a prolonged drought especially during the latter stages of the study. Southern chorus frog occupan cy of wetlands was significantly positively influenced by an increasing degree of vegetative heterogeneity At our study area, wetlands with the highest vegetative heterogeneity score were those with high diversity and abundance of herbaceous groundcover species including grasses sedges, and flowers ( e.g. Andropogon spp., Spartina spp., Juncus spp., Rhync h ospora s pp., Eleocharis spp., Aesclepias spp., Rhexia spp., Drosera spp., and Cyperus spp. ), and presence of shrubs ( e.g. Serenoa spp., Hypericum spp., Baccharis spp., Cephalanthus spp., Ludwigia spp., and Myrica spp. ) within or directly surrounding the wetland, and presence of trees immediately adjacent to the wetland. Males of P. n. verrucosa are
22 secretive and call with their heads protruding above th e water from locations where the vegetation is most dense, generally at the bases of grass tussocks or under overhanging grass and shrubs on the ed ges of wetlands (Einem and Ober 1956, Duellman and Schwartz 1958, Mount 1975 Gartside 1980). This preferenc e contrasts with ornate chorus frogs ( Pseudacris ornata ) who call from open, exposed situations in the same locations (Schwartz 1957). Ecosystem models aim to characterize the major dynamics of ecosystems, to understand systems and to allow predictions of their behavior (whether generally or in response to particular changes). N o single natural scale at which ecological phenomena should be studied exists ; systems generally show characteristic variability on a range of spatial, temporal, and organization al scales and life history adaptations such as dispersal and dormancy alter the perceptual scales of the species and the observed variability. Developing p redictive models of these systems for habitat and species management is important, thus it is necess ary to interface the disparate scales of interest of researchers studying these problems at different levels (Levin 1992). My study has investigated amphibian occurrence at wetlands using a combination of ecosystem scale (wetland variables) and broad scal e (landscape level) characteristics and provided information from species life histories to explain the resulting significant ecological models. In Florida, isolated wetlands are used as breeding habitat by at least 28 species of amphibian ( Sudol et al. 20 09 ). Of these, 14 species are obligates, meaning they breed exclusively in isolated wetlands The presence of isolated wetlands is essential for these species to breed successfully. The remaining species use isolated wetlands opportunistically and have the ability to breed elsewhere. Increasing pressure placed on wetlands caused by low density, sprawl style urban development, agriculture, and
23 phosphate mining have severely reduced the number of wetlands in the United States and particularly Florida Under recent changes to federal regulations, many isolated wetlands that are hydrologically separated from waterways either by berms or great distances are no longer under feder al protection. As a result of two Supreme Court decisions, a significant numbe r of isolated wetlands throughout the United States lost protection al. 2005) and even more lost protection when federal authority over wetlands not directly connected (via restricted (BenDor 2008). These narrow readings by the court increase pressure on local governments forc ing them to plan for and regulate the effects of wetland conversions and subsequ ent relocations, often through the form of local or countywide stormwater ordinances (BenDor et al. 2008). This situation has increased the importance of well formulated wetland regulations and ordinances at the state and local scale. A major disturban ce to isolated wetlands in central Florida is phosphate mining (FIPR 2010). The mining of phosphate for fertilizer is typically conducted using strip mining techniques including clearing the site of all vegetation, removal of soil, and mining the underlyi ng phosphate matrix with draglines Enormous draglines dig 10 m into the earth to get at the phosphate; strip mining may leave 20 m deep valleys interspersed with piles of cast earth, and the resulting landscape must be reclaimed (FIPR 2010). Following e xtraction of phosphate the site is back filled with sand sepa rated from the phosphate ore Because of this large scale clearing, mining and reclamation in central Florida, improvement of reclamation techniques is critical State law requires that land disturbed by phosphate mining be restored to a useful condition, and sometimes reclamation where the ecological systems function as they did before the
24 mining (FIPR 2010). In the legislation creating the Florida Institute for Phosphate Research (FIPR) is the mission to study reclamation alternatives and technologies. Goals of FIPR include developing methods to improve wildlife habitat on reclaimed mined lands and facilitate recolonization by wildlife. To date, reclamation practices include contouring (la nd is reshaped to resemble pre mining topography and drainage) and revegetation (replacement of plant communities which also support agricultural opportunities). Under current practice there is not a standardized, post release, quantitative assessment of phosphate mine reclamation and restoration projects, but each is considered on a case by case basis according to the conditions contained in the permits (FIPR 2010). E stablishing conservation and restoration goals that provide for high quality wetland and upland heterogeneity as a condition for reclamation release is critical. In a given year, individual species metapopulation or patchy population dynamics are affected in part by environmental variables of ephemeral wetlands. Where and when species occup y areas of the landscape is of great importance to conservation biology, particularly when identifying areas for protection and management. The number of species within an area results from a complex interaction of resource availability, habitat complexit y, biogeography, land use history, and phylogenetic history (Dodd 2009). Because it is logistically challenging to estimate changes in absolute amphibian abundance across large areas over time, an excellent option is to measure the presence or absence of the species at a number of wetlands which is the proportion of area occupied (MacKenzie et al. 2006). Because small, isolated wetlands support a diverse array of amphibian species, produce large numbers of metamorphosing juveniles, and can function as ste pping stones for dispersal and recolonization of extinct populations (Moler and Fra nz 1987, LaClaire and Franz 1991 Semlitsch and Bodie 1998), I have
25 attempted to elucidate the factors affecting individual species occupancy of a range of native, small, is olated wetlands in central Florida wh ere wetland disturbance is high With the worldwide decline in amphibian species richness, conservation targets for amphibians have been a priority for resource managers and are well discussed in the literature. Amphi bian conservation requires an integrated landscape approach to management, rather than solely a species oriented approach (Dodd 2009) because of their complicated biphasic life cycle. When a ttemptin g to conserve amphibian habitat, wetland breeding sites ( core habitat), retreat sites, dispersal corridors, and meta/patchy population structure must be considered Semlitsch and Jensen (2001) advanced the idea of core habitats for wetland breeding amphibians and suggest a core wetland should be surrounded by t hree areas of protection including the aquatic buffer zone, core habitat plus aquatic buffer zone, and a terrestrial buffer zone that is critical for feeding, growth, maturation, and maintenance of the juvenile and adult population, some of which lay eggs and overwinter in this zone. Dodd (2009) suggest ed this concept could also be expanded to include unique habitat including caves, rock faces, steeply sided slopes, and areas that restrict populations including waterfall spray zones and mountain tops. Sem litsch (2000) suggest ed that as the distance between wetlands increases, the potential for migration and recolonization by amphibians decreases as well as the chance for recolonization by source populations from nearby wetlands. Furthermore, many pond bre eding amphibians show high site fidelity and return each breeding season to the same pond (Shields 1982) and do not emigrate long distances. In addition to considerations of distance to neighboring wetlands, it is important for regulatory agencies interes ted in protecting pond breeding amphibians to consider wetland isolation and hydroperiod (Paton and Crouch 2002).
26 Beyond individual wetland conservation goals, however, should be considerations that include a mosaic of wetlands with different wetland an d landscape characteristics, including several of the factors I measured in our study (distance to permanent water, hydroperiod, distance to nearest wetland, and vegetation heterogeneity). In average years, amphibians are equipped to handle specific breed ing environments (e.g. whether it is a short hydroperiod, contains fish, near a permanent water source, or is far from neighboring wetlands). With unavoidable environmental variability, however, amphibians must work harder to find suitable breeding sites. Some amphibians require a variety of vegetative structure around a wetland (structure composition often more important than species composition); some require elevated calling sites, shallow emergent vegetation for cover, or woody debris to deposit eggs (Dodd 2009). Canopy cover is often important as it affects thermal regimes and many species do not breed in enclosed canopy. Spatial and temporal variations in rainfall patterns can have significant effects on amphibian breeding success since dry years r educe the chance of larval amphibians developing to metamorphosis, whereas excessively wet years increase the connectivity among wetlands and allow occupation by predatory fish (Babbitt and Tanner 2000, Barber 2001). Rainfall in Florida during 2004 was ex tensive, with four major named hurricanes (Hurricanes Charlie, Frances, Ivan, and Jeanne; NOAA 2010) passing over Florida, and slightly above average during 2005 when the study began. The following year, 2006, was characterized by infrequent rains and was the beginning of a severe two year drought comprising two of the driest back to back calendar years Florida has experienced, dating back to 1932 (FDEP 2010). Thus, preserving or creating a mosaic of wetlands with varying wetland and landscape characteris tics acts as a buffer to breeding amphibians during environmental fluctuations. Important for preserving and especially when restoring wetlands, it is necessary to implement designs that accommodate adult anti predator behaviors and
27 adaptive habitat shift ing (Petranka and Holbrook 2006) ; this data shows that this can be achieved by arrays of wetlands that vary markedly in hydroperiod, vegetative heterogeneity (structure) and spatial proximity. This mosaic approach is especially relevant to pra ctices that disturb large areas where mitigation and restoration goals could include restoration of wetlands that provide each resident species with many potential breeding sites. Sites that contain one or only a few wetlands with similar characteristics (e.g. hydrop eriod) may severely constrain the ability of adults to seek out high quality habitats that have low densities of predators. Thus, a diverse array of wetlands on site increases spatiotemporal variability in predation risk and increases the likelihood that juveniles will be recruited annually into the adult population, which should enhance the long term persistence of (patchy) populations (Petranka and Holbrook 2006). Wetland breeding amphibians have often been characterized as having strong site fidelity low vagility, and metapopulation structure (Alford and Richards 1999 Smith and Green 2005). Although conservation guidelines have emphasized the need to establish habitats t o support metapopulations (e.g. Semlitsch 2000), emerging research suggests some amphibians are more vagile and less philopatric than previously suspected (Petranka et al. 2004 Smith and Green 2005). An alternative to metapopulation structure at the local level is the existence of patchy populations where movements between wetlands occur at such high rates that local wetland populations do not develop a significant degree of demographic independence (Harrison 1991 McCullough 1996 Smith and Green 2005). Thus, when restoring wetlands ecologists must decide on the appropriate number and spatial arrangement of habitats, which is strongly influenced by the nature of population organization at the local level. According to Petranka and Holbrook (2006) at sites where wetlands are in close proximity (e.g. <500 m apart), restoration succe ss may be enhanced by creating spatial arrays of
28 wetlands that are designed to support patchy populations rather than metapopulations. For example, a metapopulation design would likely entail the installation of relatively few wetlands that are spaced the maximum distance apart to increase demographic independence. In contrast, a patchy population design would likely incorporate more wetlands with many in close proximity to one another to facilitate adaptive habitat switching. At this scale, metapopulati on designs will likely fail to establish local metapopulations (Smith and Green 2005) Instead metapopulation or landscape level conservation, in general, should be focused on dispersal among populations at spatial scales >1 10 km, longer periods of time and on the importance of pond density and distributions, terrestrial connectivity, and isolation effects due to land use (Marsh and Trenham 2001 Semlitsch 2008).
29 Conclusions My study suggests that a diverse range of wetland and landscape characte ristics resulting in a mosaic of wetlands provides different habitat opportunities each year, favoring the long term persistence of amphibian diversity. The models developed in this study provide insight to the limited empirical knowledge base on the stru cture of a mosaic of wetlands in an area and its role in the factors influencing amphibian occupancy of wetlands I concur with Snodgrass et. al. (2000 ) and Paton and Crouch ( 2002 ) that regulatory agencies should strive to maintain a diversity of wetlands with varying hydroperiods and minimal nearest neighbor distances among wetlands and also with Petranka and Holbrook (2006) who advocate restoring wetlands as arrays that vary markedly in hydroperiod and spatial proximity to one another. Further, I sugges t preservation and restoration of mosaics of wetlands with a wider variety of landscape and wetland characteristics including distance to permanent water and vegetation heterogeneity. In this system, the diversity of amphibian species supported by small, isolated, ephemeral wetlands probably relies on the wide envi ronmental gradient the wetlands encompass
30 Figure 1. Map of study wetlands located in Hillsborough and Hardee Counties, Florida
31 Table 1. Wetland and landsca pe covariates used in modeling amphibian species occurrence
32 Table 2. Number of successful sampling nights, percent of time frogloggers functioned, and total number of minutes of amphibian breeding vocalizations recorded per wetland.
33 Table 3. Presence or absence of each amphibian species at study wetlands, from June 2005 September 2006
34 Table 4. Candidate models constructed from a generalized linear model of top three covariates against presence/absence of each s pecies. Bold models represent significant values as tested from Likelihood Type 3 tests
35 Table 5. Direction of the r elationship [positive (+); no relationship (nr) ; negative ( ) ] for each covariate as it relates to individual species and the Likelihood Type 3 tests for top three covariates used in building candidate models. Bold values are significan t
36 References Alford, R. A. and S. J. Richards. 1999. Global amp hibian declines: a problem in applied ecology. Annual Review of Ecology and Systematics 30:133 165. Anderson, P.K. 1951. Albinism in tadpoles of Microhyla carolinensis Herpetologica 7:56. Atauri, J.A. and J.V. de Lucio 2001. The role of landscape structure in species richness distribution of birds, amphibians, reptiles and lepidopterans in Mediterranean landscapes. Landscape Ecology 16:47 159. Bailey, M.A. 1991. The dusky gopher frog in Alabama. Journal of the Alabama Academy of Science 62:28 34. Bailey, L.L., T.R. Simmons, and K.H. Pollock. 2004. Estimating site occupancy and species detection: probability parameters for terrestrial salama nders. Ecological Applications 14(3):692 702. Barichivich, W.J. 2003. Guidelines for building and operating remote field recorders (automated frog call data loggers), Appendix IV. In: Monitoring Amphibians in Great Smokey Mountains National Park. C.K. Dodd, Jr. U.S. Geological Survey Circular 1258. Beck, C.W. and J.D. Congdon. 1999. Effects of ind ividual variation in age and size at metamorphosis on growth and survivorship of southern toad ( Bufo terrestris ) metamorphs. Canadian Journal of Zoology 77:944 951. Beja, P. and R. Alcazar. 2003. Conservation of Mediterranean temporary ponds under agric ultural intensification: an evaluation using amphibians. Biological Conservation 114:317 326. BenDor, T. 2008. The social impacts of wetland mitigation policies in the United States. Journal of Planning Literature 22(4):341 357 Bennett, A.F., J.Q. Radford, and A. Haslem. 2006. Properties of land mosaics: implications for nature conservation in agricultural landscapes. Biological Conservation 133:250 264. Blaustein, A. R. and D. B. Wake. 1990. Declining amphibian populations: a global phenomen on? Trends in Ecology and Evolution 5:203 204. Bogert, C.M. 1947. A field study of homing in Bufo t. terrestris American Museum Novitates, Number 1355, American Museum of Natural History, New York.
37 Burne, M.R. and C.R. Griffin. 2005. Habitat assoc iations of pool breeding amphibians in eastern Massachusetts, USA. W etlands Ecology and Management 13:247 259. Burnham, K. P. and D. R. Anderson. 1998. Model selection and inference: a practical information theoretic approach. Springer Verlag, New Yor k, New York, USA. Bury, R.B. and J.A. Whelan. 1984. Ecology and management of the bullfrog. U.S. Fish and Wildlife Service, Resource Publication Number 155, Washington, D.C. Collinge, S.K. and R.T.T. Forman. 1998. A conceptual model of land conversi on processes: predictions and evidence from a microlandscape experiment with grassland insects. Oikos 82:66 84. Collins, R.J. and G.W. Barrett. 1997. Effects of habitat fragmentation on meadow vole ( Microtus pennsylvanicus ) population dynamics in exper imental landscapes. Landscape Ecology 12:63 76. Collins, J.P. 1979. Intrapopulation variation in the body size at metamorphosis and timing of metamorphosis in the bullfrog, Rana catesbeiana Ecology 60:738 749. Comer, P., K. Goodin, G. Hammerson, S. M enard, M. Pyne, et al. 2005. Biodiversity values of geographically isolated wetlands: an analysis of 20 U.S. states. NatureServe, Arlington, VA, USA. Cushman, S.A. 2006. Implications of habitat loss and fragmentation for the conservation of pond breed ing amphibians: a review and prospectus. Biological Conservation 128:231 240. deMaynadier, P.G. and M.L. Hunter, Jr. 1999. Forest canopy closure and juvenile emigration by pool breeding amphibians in Maine. Journal of Wildlife Management 63:441 450. Dickman, C.R. 1987. Habitat fragmentation and vertebrate species richness in an urban environment. Journal of Applied Ecology 24(2):337 351. Dodd, C.K., Jr. (ed.). 2009. Amphibian Ecology and Conservation: A Handbook of Techniques. Oxford Universi ty Press, Oxford, UK. 556 pp. Doherty, P.F. and T.C.Grubb Jr. 2002. Survivorship of the permanent resident birds in a fragmented forested landscape. Ecology 83(3):844 857 Donnelly, M.A. 1997. A preliminary survey of amphibians in isolated wetlands In S. Mortellaro, Inventory of freshwater biota, South Florida Water Management District's isolated wetlands monitoring program. South Florida Water Management District. Downing, D., C. Winer, and L. D. Wood. 2003. Navigating through Clean Water Ac t jurisdiction: a legal review. Wetlands 23:475 493.
38 Duellman, W.E. and A. Schwartz. 1958. Amphibians and reptiles of southern Florida. Bulletin of the Florida State Museum 3:181 324. Dundee H.A. and D.A. Rossman. 1989. The Amphibians and Reptiles of Louisiana. Louisiana State University Press, Baton Rouge, Louisiana. Durbin, D.J., S. Gonzalez, H. Mushinsky, E. McCoy, et al. 2008. Wildlife Habitat and Wildlife Utilization of Phosphate Mined Lands. FIPR Publication #03 147 230. Available: http://www1.fipr.state.fl.us/Publications (Accessed May 5, 20 10) Edenius, L. and J.Elmber. 1996. Landscape level effects of modern forestry on bird communities in North Swedish boreal forests. Lands cape Ecology 11:325 338. Einem, G.E. and L.D. Ober. 1956. The seasonal behavior of certain Floridian Salientia. Herpetologica 12:205 212. ESRI Inc. 2009. ArcInfo User's Manual, Version 9.3.1. Environmental Systems Research Institute Inc., Redlands, CA, USA. FDEP (Florida Department of Environmental Protection). 2010. "Florida Drought Conditions". Available: http://www.dep.state.fl.us/Drought (Accessed January 20, 2010). FIPR (Florida Institute for Phosphate Research). 2010 http://www1.fipr.state.fl.us/phosphateprimer (Accessed May 10, 2010). FMNH (Florida Museum of Natural History). 2010 http://www.flmnh.ufl.edu/herpetology (Accessed May 10, 2010). Fortuna, M., C. Gmez Rodrguez, and J. Bascompte. 2006. Spatial network structure and amphibian persistence in sto chastic environments. Proceeding s of the British Royal Society 273:1429 1434 Franz, R., C.K. Dodd Jr ., and C. Jones. 1988. Natural history notes: Rana areolata aesopus (Florida gopher frog). Movement. Herpetological Review 19:33. Gaddy, L. L. 1994. The ecology of seven rare wetland species in Carolina bays and bay like depressions in the western Coastal Plain of South Carolina. Report. U.S. Forest Service, Savannah River Forest Station, Aiken, South Carolina. Gartside, D.F. 1980. Analysis of a hybrid zone between chorus frogs of the Pseudacris nigrita complex in the southern United States. Copeia 1980:56 66. Gibbons, J.W. and R.D. Semlitsch. 1991. Guide to the Reptiles and Amphibians of the Savannah River Site. University of Georgia Press, Athens, Georgia. Gibbs, J. P. 1993 Importance of small wetlands for the persistence of local populations of wetland associated animals. Wetlands 13:25 31. Gibbs, J. P. 2000. Wetland loss and biodiversity conservation. Conservation Biology 14:314 3 17.
39 Gjerde, I., M. Saetersdal, and T. Nilsen. 2005. Abundance of two threatened woodpecker species in relation to the proportion of spruce plantations in native pine forests of western Norway. Biodiversity and Conservation 14:377 393. Gmez Rodrguez, C., C. Daz Paniagua, L. Serrano, M. Florencio, and A. Portheault. 2009. Mediterranean temporary ponds as amphibian breeding habitats: the importance of preserving pond networks. Aquatic Ecology 43:1179 1191. Gonzalez, S. M. 2004. Biological indicat ors of wetland health: comparing qualitative and quantitative vegetation measures with anuran measures. Masters thesis. Univ. of South Florida. Tampa, Florida. Groom, M. J., G. K. Meffe, C. R. Carroll, and contributors. 2006. Principles of conservati on biology. 3rd edition. Sinauer Associates, Sunderland, Massachusetts. Hair, J.F., J.E. Anderson, R.L. Tatham, and W.C. Black. 1998. Multivariate Data Analysis. 5th ed. Upper Saddle River, NJ: Prentice Hall. Halley, J. M., R. S. Oldham, and J. W. Arnt zen. 1996. Predicting the persistence of amphibian populations with the help of a spatial model. Journal of Applied Ecology 33:455 470. Hargis, C.D., J.A. Bissonette, J.L. David. 1998. The behavior of landscape metrics commonly used in the study of h abitat fragmentation. Landscape Ecology 13:167 186. Harrison, S. 1991. Local extinction in a metapopulation context: an empirical evaluation. Biological Journal of the Linnaean Society 42:73 88. Hecnar, S.J. 1997. Amphibian pond communities in southw estern Ontario. Pp. 1 15. In Green, D.M. (Ed.), Amphibians in Decline: Canadian Studies of a Global Problem. Hecnar, S.J. and R.T. M'Closkey. 1996. Regional dynamics and the status of amphibians. Ecology 77:2091 2097. y. 1997. The effects of predatory fish on amphibian species richness and distribution. Biological Conservation 79:123 131. Herrmann, H. L., K. J. Babbitt, M. J. Baber, and R. G. Congalton. 2005. Effects of landscape characteristics on amphibian distri bution in a forest dominated landscape. Biological Conservation 123:139 149. Heyer, W.R., R.W. McDiarmid, and D.L. Weigmann. 1975. Tadpoles, predation, and pond habitats in the tropics. Biotropica 7(2):100 111. Jensen, J.B. 1995. Monitoring dusky gopher frog breeding success and management effectiveness, Conecuh National Forest, Alabama. U.S.D.A. Forest Service, Unpublished report, Andalusia, Alabama. Adapted from AmphibiaWeb:
40 "Information on amphibian biology and conservation". Available: http://amphibiaweb.org (Accessed May 1, 2010). Jokimaki, J., and E. Huhta. 1996. Effects of landscape matrix and habitat structure on a bird community in northern Finland: a multi scale approach. Ornis Fennica 73:9 7 113. Knapp, R.A., K.R. Matthews, H.K. Preisler, and R. Jellison. 2003. Developing probabilistic models to predict amphibian site occupancy in a patchy landscape. Ecological Applications 13(4):1069 1082. Kondratieff, B. C. and C. J. Pyott. 1987. Th e anisoptera of the Savannah River Plant South Carolina U.S.A. thirty years later. Odonatologica 16:9 24. Laan R. and R. Verboom. 1990. Effects of pool size and isolation on amphibian communities. Biological Conservation 54:251 262. LaClaire, L.V. an d R. Franz. 1991. Importance of isolated wetlands in upland landscapes. Pp. 9 15. In Kelly, M. (Ed.), Proceedings of the Second Annual Meeting, Florida Lake Management Society, Winterhaven, Florida. Lamb, T. 1984. The influence of sex and breeding co nditions on microhabitat selection and diet in the pig frog, Rana grylio American Midland Naturalist 111:311 318. Lefcort, H. 1998. Chemically mediated fright response in southern toad ( Bufo terrestris ) tadpoles. Copeia 1998:445 450. Lemckert, F. L. 2004. Variations in anuran movements and habitat use: implications for conservation. Applied Herpetology 1:165 181. Levell, J.P. 1997. A Field Guide to Reptiles and the Law. Revised Second edition. Serpent's Tale, Lanesboro, Minnesota. Levin, S.A. 1992. The problem of pattern and scale in ecology. Ecology 73(6):1943 1967. Lindenmayer, D.B., and J. Fischer. 2006. Landscape Change and Habitat Fragmentation. Island Press, Washington, DC. Lindenmayer D., R.J. Hobbs, R. Montegue Drake, J. Alexandr a, A. Bennett, M. Burgman et al. 2008. A checklist for ecological management of landscapes for conservation. Ecology Letters 11:78 91. MacKenzie, D. 2006. Occupancy estimation and modeling: inferring patterns and dynamics of species occurences. Acad emic Press. 344 pages. MacKenzie, D.I. and W.L. Kendall. 2002. How should detection probability be incorporated into estimates of relative abundance? Ecology 83(9):2387 2393.
41 Mahoney, D.L., M.A. Mort, and B.E. Taylor. 1990. Species richness of calan oid copepods, cladocerans and other branchiopods in Carolina bay temporary ponds. American Midland Naturalist 123(2): 244 258. Marsh D.M. and P.C. Trenham 2001. Metapopulation dynamics and amphibian conservation. Conservation Biology 15:40 49. Mart of, B.S., W.M. Palmer, J.R. Bailey and J.R. Harrison III. 1980. Amphibians and Reptiles of the Carolinas and Virginia. University of North Carolina Press, Chapel Hill, North Carolina. Mazerolle, M.J. 2006. Improving data analysis in herpetology: usin Information Criterion (AIC) to assess the strength of b iological hypotheses. Amphibia Reptilia 27:169 180. McComb, W.C., K. McGarigal, and R.G. Anthony. 1993. Small mammal and amphibian abundance in streamside and upslope habitats of matur e Douglas fir stands, western Oregon. Northwest Science 67:7 15. McCullagh, P. and J.A. Nelder. 1989. Generalized linear models, second ed. Chapman and Hall, London. McCullough D. R. 1996. Introduction. In: Metapopulations and Wildlife Conservation (ed Dale R. McCullough) P p. 1 10. Island Press, Washington, D.C. McGarigal, K. and W.C. McComb 1995. Relationships between landscape structure and breeding birds in the Oregon Coast Range. Ecological Monographs 65:235 260. Moler, P.E. and R. Franz. 1987. Wildlife values of small, isolated wetlands in the southeastern coastal plain. Pp. 234 241. In Odum, R.R., K.A. Riddleberger and J.C. Ozier (Eds.), Proceedings of the Third Annual Nongame and Endangered Wildlife Symposium, Georgia Department of N atural Resources, Atlanta, Georgia. Moler, P.E. 1992. Rare and Endangered Biota of Florida. Volume III. Amphibians and Reptiles. University Press of Florida, Gainesville, Florida. Morin, P. 1983. Predation, c ompetition, and the c omposition of l arv al a nuran g uilds. Ecological Monographs 53(2):120 138. Mount, R.H. 1975. The Reptiles and Amphibians of Alabama. Agricultural Experimental Station, Auburn University Press, Auburn, Alabama. Murphy, C.G. 1994. Determinants of chorus tenure in barking treefrogs ( Hyla gratiosa ). Behavioral Ecology and Sociobiology 34:285 294. Murphy, C.G., S.T. Emlen and P.W. Sherman. 1993. Reproductive strategies of the treefrog Hyla gratiosa: implications for management. Florida Game and Freshwater Fish Commissi on, Nongame Wildlife Program, Final report, Gainesville, Florida.
42 Mushinsky, H. R., E. D. McCoy, and S. M. Gonzalez. 2004. Measuring wetland health: comparing vegetation and anurans as indicators. Report to Southwest Florida Water Management District 55pp. NAAMP (North American Amphibian Monitoring Program). 2010. "Protocol Description: Index and Code Descriptions". Available: http://www.pwrc.usgs.gov/naamp (Accessed May 5, 2010). Neill, W.T. 19 52. Burrowing habits of Hyla gratiosa Copeia 1952:196. Neill, W.T. 1958. The varied calls of the barking treefrog, Hyla gratiosa LeConte. Copeia 1958:44 46. NOAA (National Oceanic and Atmospheric Administration). 2010. "Hurricane History". Availa ble: http://www.nhc.noaa.gov/HAW2/english/history.shtml (Accessed: Januray 20, 2010). Noss, R.F. and A. Cooperrider. 1994. Saving nature's legacy: protecting and restoring biodiversit y. Defenders of Wildlife and Island Press, Washington, D.C. Parker, M. and R. MacNally. 2002. Habitat loss and the habitat fragmentation threshold: an experimental evaluation of impacts on richness and total abundances using grassland invertebrates. B iological Conservation 105:217 229. Paton, P.W.C. and W.B. Crouch. 2002. Using the phenology of pond breeding amphibians to develop conservation strategies. Conservation Biology 16(1):194 204. Pechmann, J. H. K., D. E. Scott, J. W. Gibbons, and R. D. Semlitsch. 1989. Influence of wetland hydroperiod on diversity and abundance of metamorphosing juvenile salamanders. W etlands Ecology and Management 1:3 11. Petranka, J. W. and C. T. Holbrook. 2006. Wetland restoration for amphibians: should local si tes be designed to support metapopulations or patchy populations? Restoration Ecology 14:404 411. Petranka, J. W., C. K. Smith, and A. F. Scott. 2004. Identifying the minimal demographic unit for monitoring pond breeding amphibians. Ecological Applica tions 14:1065 1078. Phillips, K. 1990. Where have all the frogs and toads gone? BioScience 40:422 424. Roznik, E.A S.A. Johnson, C.H. Greenberg, and G.W. Tanner. 2009. Terrestrial movements and habitat use of gopher frogs in longleaf pine forests: a comparative study of juveniles and adults. Forest Ecology and Management 259:87 194.
43 Schwartz, A. 1957. Chorus frogs ( Pseudacris nigrita LeConte) in South Carolina. American Museum Novitates, Number 1838, American Museum of Natural History, New York Seale, D.B. 1980. Influence of amphibian larvae on primary production, nutrient flux, and competition in a pond ecosystem. Ecology 61:1531 1550. Semlitsch, R.D. 1998. Biological delineation of terrestrial buffer zones for pond breeding salamanders Conservation Biology 12:1113 1119. Semlitsch, R.D. 2000. Principles for management of aquatic breeding amphibians. Journal of Wildlife Management 64:615 631. Semlitsch, R.D. 2008. Differentiating migration and dispersal processes for pond breedin g amphibians. Journal of Wildlife Management 72(1):260 267. Semlitsch, R.D. and J.R. Bodie. 1998. Are small, isolated wetlands expendable? Conservation Biology 12:1129 1133. Semlitsc h, R. D. and J. R. Bodie. 2003. Biological criteria for buffer zon es around wetlands and riparian habitats. Conservation Biology 17:1219 1228. Semlitsch, R. D. and J. B. Jensen. 2001. Core habitat, not buffer zone. National Wetlands Newsletter 23:5 6. Semlitsch, R.D., D.E. Scott, J.H.K. Pechmann and J.W. Gibbons. 1 996. Structure and dynamics of an amphibian community: evidence from a 16 year study of a natural pond. Pp. 217 248. In Cody, M.L. and J.A. Smallwood (Eds.), Long term Studies of Vertebrate Communities. Academic Press, San Diego, California. Sharitz, R. R. and J. W. Gibbons. 1982. The ecology of southeastern shrub bogs (pocosins) and Carolina bays: a community profile. FWS/OBS 82/04. U.S. Fish and Wildlife Service,Washington, D. C. Smith, M. A. and D. M. Green. 2005. Dispersal and the metapopulatio n paradigm in amphibian ec ology and conservation: are all amphibian populations metapopulations? Ecography 28:110 128. Snodgrass J.W., M.J. Komoroski, A.L. Bryan Jr., J. Burger. 200 0 Relationships among isolated wetland size, hydroperiod, and amphibi an species richness: implications for wetlands regulations. Conservation Biolology 14:414 419. StatSoft, Inc. 2005. STATISTICA data analysis software system. Version 7.1. Available: www.statsoft.com Stuart, S .N., J.S. Chanson, N.A. Cox, B.E. Young, A.S.L. Rodrigues, D.L. Fischman, and R.W. Waller. 2004. Status and trends of amphibian declines and extinctions worldwide. Science 306:1783 1786.
44 Sudol, T.A., E.V. Wilcox, and W. Giuliano. 2009. Isolated W et lands and Breeding Amphibians. Wildlife Ecology and Conservation Department, Florida Cooperative Extension Service, Institute of Food and Agricultural Sciences, University of Florida. Publication #WEC268. Available: http://edis.ifas.ufl.edu. SWANCC (So lid Waste Agency of Northern Cook County) v. U.S. Army Corps of Engineers. 531 U.S. 314. SWFWMD (Southwest Florida Water Management District). 2010. "Data and Maps". Available: http://www.swfwmd.state.f l.us/data (Accessed January 20, 2010). Tews, J., U. Brose, V. Grimm, K. Tielbrger, M. C. Wichmann, M. Schwager and F. Jeltsch. 2004. Animal species diversity driven by habitat heterogeneity/diversity: the importance of keystone structures. Journal of Biogeography 31:79 92. Vos, C.C. and A .H.P. Stumpel. 1995. Comparison of habitat isolation parameters in relation to fragmented distribution patterns in the tree frog ( Hyla arborea ). Landscape Ecology 11:203 214. Walters, B. 1975. Studies of interspecific predation within an amphibian com munity. Journal of Herpetology 9:267 279. Werner, E.E. and M.A. McPeek. 1994. Direct and indirect effects of predators on two anuran species along an environmental gradient. Ecology 75:1368 1382. Werner E.E., D.K. Skelly and R.A. Relyea et al. 2007 Amphibian species richness across environmental gradients. Oikos 116:1697 1712 Wiens, J. and M. Moss., editors. 2005. Issues and Perspectives in Landscape Ecology. Cambridge: Cambridge University Press. Wilbur, H. M. 1987. Regulation of structu re in complex systems: experimental temporary pond communities. Ecology 68:1437 1452. With, K.A., D.M. Pavuk, J.L. Worchuck, R.K. Oates, and J.L. Fisher. 2002. Threshold effects of landscape structure on biological control in agroecosystems. Ecologic al Applications 12:52 65. Wood, K.V., J.D. Nichols, H.F. Percival and J.E. Hines 1998. Size sex variation in survival rates and abundance of pig frogs, Rana grylio in northern Florida wetlands. Journal of Herpetology 32:527 535. Wright, A.H. and A. A. Wright. 1949. Handbook of Frogs and Toads of the United States and Canada. Third edition. Comstock Publishing Associates, Ithaca, N Y Wright, A.H. 1932. Life Histories of the Frogs of the Okefinokee Swamp, Georgia. North American Salientia (Anura) Number 2. Macmillan Press, New York.
45 Appendix 1. Wetland and landscape covariate values used in modeling amphibian species occurrence at twelve wetlands across west central Florida, 2005 and 2006
About the Author Jackie Guzy is f rom Chicago, IL and earned a B.S. Degree in Biology (marine science concentration) from the University of Tampa where she participated in honors research travel in Honduras (coral reefs ) and Costa Rica (tropical ecology and conservation). At UT she publis hed her research on effects of groundwater withdrawal on amphibians at Morris Bridge Wellfield. She obtained a M.S. Degree in Biology (ecology and evolution concentration) from the University of South Florida where she worked as a research assistant study ing amphibian populations among wetlands affected by varying degrees of urbanization She has twice presented research at the international Joint Meeting of Ichthyologists and Herpetologists conference and been recognized for Outstanding Teaching by a Gr aduate Student