« Back

Habitat Factors Influencing the Distribution of Small Vertebrates on Unmined and Phosphate-Mined Uplands in Central Florida


The principal causes for the loss of species on Earth today are pollution, which alters habitats so that they are uninhabitable for some species, and outright destruction of habitats — such as by clearcutting ancient forests, strip mining, and urbanization — which alters the habitats so that they are uninhabitable for virtually all species. Reclamation of human-altered habitats is a method used to reduce the long-term effects of habitat disturbance on the native flora and fauna. Before any reclamation efforts are made, however, one must have clear aims in mind. “Restoration” of habitats, which may be considered the ultimate form of reclamation, is accomplished by creating both structural and functional attributes of a damaged ecosystem. True restoration attempts to put back exactly what was thought to have been there prior to any disturbance. We believe that restoration is not a reasonable goal for the reclamation of lands previously mined for phosphate. A second possible purpose for reclamation is to aim for a product which is similar to, but somewhat less than, full restoration. This second level of reclamation is termed “rehabilitation.” We believe that rehabilitation is a reasonable goal for the reclamation of phosphate-mined lands.

“Representativeness” is the presence of the range of ecological variation in a particular habitat. We use the term representativeness to indicate that “rehabilitated phosphate mined lands will include typical or common vertebrate species as well as rare vertebrate species, in their typical relative abundances, to present the full range of the biota.” The goal of such an approach is to produce a system of rehabilitated phosphate-mined lands that, collectively, encompass the broad range of vertebrate species typically found in areas that have not been mined or otherwise significantly modified. We believe that the concepts of rehabilitation and representativeness together provide a reasonable framework for the reclamation of phosphate-mined lands in Florida. Rather than placing extreme importance on rare and/or endangered or threatened species, representativeness emphasizes the value of preserving overall biodiversity.

We studied three kinds of unmined upland habitats and two kinds of previously-mined lands. The three kinds of upland habitats were sandhill, scrub, and scrubby flatwoods; the two kinds of previously-mined lands were those mined prior to the mandatory reclamation laws of 1975 and those mined since 1975. We studied the three unmined habitats and two previously-mined lands to determine the number of species present (species richness) and relative abundances of resident vertebrates. We then compared the lists of species found on previously-mined lands to lists of species from unmined upland habitats. In all, we selected 60 study sites in Hillsborough, Polk, and Manatee Counties, covering a total area of about 900 square miles. Thirty sites were previously-mined lands; of these, 14 were mined prior to 1975 and 16 thereafter. The mined sites had received various amounts of reclamation effort. Thirty study sites were unmined upland habitats; of these, 8 were sandhill, 15 were scrub, and 7 were scrubby flatwoods. All 30 unmined upland habitats were within the vicinity of mined lands. All 60 sites were classified according to their size (small, large), distance to seasonal water (near, far), distance to permanent water (near, far), and distance to upland habitat (near, far), and the habitats (upland habitat, wetland, citrus grove, pasture, farm, active mine, inactive mine, reclaimed land, old field, residential) immediately surrounding the sites were recorded. The 30 mined sites were classified according to the type of soil (overburden, sand tailings), the type of vegetation reclamation (woody species, herbaceous species, topsoil from upland habitats), and the year(s) in which reclamation occurred.

We sampled most species of amphibians, reptiles, mammals, and birds at the 60 study sites. Amphibians, reptiles and mammals (called quadrupeds, hereafter) were captured by trapping at each site. Birds were not captured, rather they were censused twice each year by observation, to determine both resident breeding bird populations and migratory bird populations. “Resident species” were those species who spent a majority of time and reproduced in their respective habitats. Resident species were identified from existing information on species’ distributions and from information we gathered by trapping and observing quadrupeds and birds in unmined habitats within the vicinity of past phosphate mining operations. Resident species are those species that exist primarily in the upland habitats we studied. For example, an alligator may walk through scrub or sandhill once or a few times during a year, but it is not a resident of either habitat; whereas a bluejay that nests in the sandhill habitat is a resident of that habitat. Resident species comprise a subset of the total pool of species known to occur in sandhill, scrub, and/or scrubby flatwoods habitats. To determine how successful the various resident species have been in establishing colonies at previously-mined sites, we collected information on quadrupeds and birds at both unmined and mined sites. Species that were found at unmined sites but were missing or underrepresented at mined sites were termed “focal species.” Focal species are a subset of the resident species. Focal species would be the likely target species for future efforts to entice wildlife to establish residency at reclaimed sites,

During our study we collected data on the vegetation and soil conditions at each study site. Vegetation composition was measured within plots near the traps we constructed to capture quadrupeds. We compiled a list of plant species present at each site. Vegetation density (percent of a prescribed field of view filled with vegetation) was measured at each plot. We made measurements of ground cover and tree canopy. Vertical canopy cover (the amount of tree canopy overhead) was determined with a hand-held densimeter. Horizontal canopy cover was estimated by visually scoring the percent cover of a board placed at pre-selected heights above the ground while being viewed from the ground at a constant distance from the board. We performed a series of tests on soils sampled from selected plots used for the vegetation analyses. We measured soil chemistry, soil texture, sand particle size, soil compaction, and root density. Standard statistical methods, mostly non-parametric, were used to analyze the data. Specific tests used include Spearman’s Rank Correlation Analysis, Mann-Whitney U-test, Kolmogorov-Smirnov One-and Twosample Tests for Goodness of Fit, and G-test of Independence. To assess species associations, we used the Variance Ratio Test and Monothetic Devisive Cluster Analysis.