Vertebrate species composition of selected scrub islands on the Lake Wales Ridge of central Florida /

MISSING IMAGE

Material Information

Title:
Vertebrate species composition of selected scrub islands on the Lake Wales Ridge of central Florida /
Physical Description:
xi, 325 p. : ill., maps ; 28 cm.
Language:
English
Creator:
Mushinsky, Henry R
McCoy, Earl D
Florida -- Nongame Wildlife Program
Publisher:
Florida Game and Fresh Water Fish Commission, Nongame Wildlife Program
Place of Publication:
Tallahassee, FL
Publication Date:

Subjects

Subjects / Keywords:
Vertebrates -- Florida -- Highlands County   ( lcsh )
Vertebrates -- Florida -- Polk County   ( lcsh )
Xeric ecology -- Florida -- Highlands County   ( lcsh )
Xeric ecology -- Florida -- Polk County   ( lcsh )
Genre:
government publication (state, provincial, terriorial, dependent)   ( marcgt )
bibliography   ( marcgt )
non-fiction   ( marcgt )

Notes

Bibliography:
Literature cited: p. 313-325.
Additional Physical Form:
Electronic version available on the World Wide Web as part of the Linking Florida's Natural Heritage Collection.
Statement of Responsibility:
Henry R. Mushinsky, Earl D. McCoy.
General Note:
"Submitted as project report for Nongame Wildlife Program project GFC-87-149, December 1995."
General Note:
"Submitted as Final Report Number NG87-149, 30 June 1991."--p.i

Record Information

Source Institution:
University of Florida
Holding Location:
University of Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
ltqf - AAA0808
notis - AKS1165
alephbibnum - 002082708
oclc - 34518442
System ID:
UF00001471:00001

Full Text
Vertebrate Species Composition of Selected Scrub Islands on the Lake Wales Ridge of Central Florida
Henry R. Mushinsky Earl D. McCoy
Department of Biology University of South Florida Tampa, FL 33620-5150
Submitted as project report for Nongame Wildlife Program project GFC-87-149
December 1995
UNIVERSITY OF FLORIDA


This report is the result of a project supported by the Florida Game and Fresh Water Fish Commission's Nongame Wildlife Program. Although the report fulfilled the project's contractual obligations, it has not been reviewed for clarity, style, or typographical errors, and has not received peer review. Any opinions or recommendations in this report are those of the authors and do not represent policy of the Commission.
Suggested citation:
Mushinsky H.R., and E.D. McCoy. 1995. Vertebrate species composition of selected scrub islands on the Lake Wales Ridge of Central Florida. Fla. Game and Fresh Water Fish Comm. Nongame Wildl. Program Project Rep. 325pp + xiv. Tallahassee, Fla.


Vertebrate Species Composition of Selected Scrub Islands on The Lake Wales Ridge of Central Florida
Henry R. Mushinsky and Earl D. McCoy
Florida Game and Fresh Water Fish Commission Nongame Wildlife Program
Submitted as
Final Report Number NG87-149 30 June 1991




Abstract
Project Number: NG87-149
Project Title: Vertebrate Species Composition of Selected Scrub
Islands on the Lake Wales Ridge of Central Florida. Date of Final Report: 30 June 1991
Project Directors: Henry R. Mushinsky and Earl D. McCoy
We assessed vertebrate species composition of 16 patches of scrub habitat of three different size categories along the Lake Wales Ridge of central Florida. For 24 months, we monitored vertebrates in three large scrubs (170 to 190 ha), four medium scrubs (25 to 50 ha), and nine small scrubs (2 to 10 ha) using a variety of collecting and sampling techniques. Most analyses were performed on the following four data sets. 1. All non-avian taxa trapped at least once, 2. characteristic non-avian taxa trapped at least once, 3. avian taxa observed breeding, and 4. characteristic avian taxa.
Association and cluster analyses, using qualitative data (presence-absence of taxa), indicated that the size of a patch of scrub was the most important habitat attribute. All four data sets showed that smaller scrubs, which were less-rich scrubs, constituted perfect (nested) subsets of the larger, species-rich scrubs. The quantitative data (relative abundance of taxa) for 34 species, indicated that the relative proportion of taxa in scrubs was more similar in large and medium scrubs than in small scrubs. Some species, however, showed considerable variation in relative abundance in all size categories of scrubs and were found in relatively high numbers even in small patches of scrub. Patch size correlated most strongly to taxonomic richness of scrubs. Distance to water and distance to other scrub habitat, however, influenced richness of amphibians, reptiles, and mammals, but not birds.
Structurally, the 16 study sites divided into "relatively-open canopied scrubs" and "relatively-closed canopied scrubs," but neither type of scrub correlated with species richness. Results of rarefaction and ordination analyses, using quantitative data, indicated that the relative abundances of 32 taxa in 16 the scrubs were influenced by vegetation structure.
When we considered the distributions of officially listed taxa in small scrubs, we found that only the relative abundance of Seeloporus woodi was correlated (positively) with scrub size. Florida scrub jays were correlated positively with the relative abundance of scrub oaks. The abundances of Eumeces egregius and Neoseps reynoldsi were correlated negatively with scrub oaks and were more common in patches where Florida scrub jays do not occur.
We determined that the best single predictor of the number of vertebrate taxa inhabiting a scrub was the size of the scrub, but that scrubs of similar size were not simple replicates. "Open-canopied" scrubs tended to support a different relative abundance of taxa than "closed-canopied" scrubs of similar size. We found that vegetational structure of a scrub influenced the abundances of "rare" species.
iii


Acknowledgments
We thank Paige L. Martin, our field assistant, and Michael MacMillan, our consulting ornithologist, for their dedication to this research project. Fred E. Lohrer, Librarian, Archbold Biological Station, provided us with much information regarding the history of our study sites, we appreciate his commitment to conservation biology. The backing of Dr. John W. Fitzpatrick, Executive Director, Archbold Biological Station, allowed us to extend our data collection period to 24 months, we thank him for his support. We thank Mr. Ney C. Landrum, Director, Division of Recreation and Parks; Mr. William R. Helm, Jr., Chief, Bureau of Forest Management; Mr. Jim Stevenson, Chief, Scientific and Technical Services; Mr. Mark E. Hebb, District Forester; Mr. Frank Montalbano, Director, Division of Wildlife; and Dr. Catherine P. Cornelius, President, South Florida Community College, for their assistance to us as we attempted to gain access to State lands for this research. Several private land owners allowed us access to patches of scrub habitat on their property, we sincerely thank them for their cooperation. Finally, we thank Ms. Janet Camp, Bookkeeper, Department of Biology, for her excellent job of managing our account.
iv


Table of Contents
Introduction....................................................1
Background.................................................1
Objectives.................................................4
The Scrub Habitat and Its Vertebrates......................5
Methods........................................................14
Study Sites...............................................14
Surrounding Habitat and Area Reduction....................16
Vegetation Sampling.......................................18
Trapping Methods for Non-avian Taxa.......................19
Trapping Schedule for Arrays..............................36
Other Sampling Methods for Non-avian Taxa.................39
Avian Sampling............................................40
Analytic..................................................42
Results........................................................46
Habitat Structure...............,.........................46
Vertebrate Records.......................................246
Qualitative Patterns in Taxonomic Composition............252
Nestedness Patterns......................................260
Quantitative Patterns in Taxonomic Composition...........264
Explaining the Patterns..................................275
Discussion....................................................298
Ecological Influences on the Diversity of
Scrub Vertebrates........................................298
Ecological Consequences of Fragmentation of
the Scrub Habitat........................................299
v


"Putting It All Together": Ranking Scrubs
in Terms of "Worth"......................................305
Literature Cited..............................................313
vi


List of Tables
1. Characteristic vertebrates of Florida's scrub habitat. Taxa
marked with an asterisk are discussed in text..............9
2. Study sites.................................................15
3. Trapping schedule for vertebrates in 16 scrubs of central
Florida. "X" means that a scrub was trapped during the sampling period indicated.................................38
4. Sampling schedule for breeding birds (April, May, June) and
migratory birds (January, February, March) in 16 scrubs during 1989 and 1990......................................41
5. Records of non-avian taxa trapped at least once. Symbols are
defined in text..........................................247
6. Records of non-avian taxa never trapped. Symbols are defined
in text..................................................248
7. Records of avian taxa observed breeding in at least one scrub.
Symbols are defined in text..............................249
8. Records of avian taxa never observed breeding. Symbols are
defined in text..........................................250
9. Correlations (Spearman's Rank) among eight kinds of scrub
vertebrates (designations of kinds of vertebrates in the table correspond to numbering in text). * = p < 0.05....254
10. Non-avian taxa trapped at least once, arranged by decreasing
degree of relatedness (see text). * = noncharacteristic taxon....................................................262
11. Avian taxa observed breeding and/or characteristic avian
taxa, arranged by decreasing degree of nestedness (see
vii


text). * = noncharacteristic taxon, but observed breeding; + = characteristic taxon, but never observed breeding.... 263
12. Abundances of selected characteristic taxa................265
13. Correlations (Spearman's Rank) among ten physical attributes
of the 16 scrubs (designations of physical attributes in
the table correspond to numbering in text). * = p <
0.05.....................................................276
14. Correlation (Spearman's Rank) between five physical
attributes of the 16 scrubs and four kinds of vertebrates (designations of vertebrates in the table correspond to the numbering in the text). * = p < 0.05.................... 278
15. Expected numbers of species derived by rarefaction analysis.
Expected numbers in each row were derived from the scrubs listed at the beginning of the row. Underlined numbers indicate that the absolute value of (observed - expected) is one or less...........................................282
16. Scrub vertebrates captured, or observed, or otherwise
determined to be present, at our sites, that are also listed by Wood (1991) or suggested for consideration by McCoy and Mushinsky (in review). Taxa from the later source are indicated by a +. SU's are arranged in order of decreasing size. * = noncharacteristic taxon, X = recorded as present.................................................294
17. Rankings of the 16 scrubs on 10 criteria. Decimal points
indicate half ranks (for Ties)..........................311
vi i i


List of Figures
1. Approximate placement of trap arrays in HEN................20
2. Approximate placement of trap arrays in ARC................21
3. Approximate placement of trap arrays in AB3................22
4. Approximate placement of trap arrays in AB1................23
5. Approximate placement of trap arrays in HH1................24
6. Approximate placement of trap arrays in AB2................25
7. Approximate placement of trap arrays in JOS................26
8. Approximate placement of the trap array in CHU.............27
9. Approximate placement of the trap array in HH2.............28
10. Approximate placement of the trap array in JUN............29
11. Approximate placement of the trap array in EGL............30
12. Approximate placement of the trap array in LAK............31
13. Approximate placement of the trap array in BAR............32
14. Approximate placement of the trap array in COL............33
15. Approximate placement of the trap array in DUF............34
16. Approximate placement of the trap array in CUT............35
17. Number of Bufo terrestris captured, and number "expected,"
in each scrub (see text)................................267
18. Number of Coluber constrictor captured, and number
"expected," in each scrub (see text)....................267
19. Number of Eumeces egregius lividus captured, and number
"expected," in each scrub (see text)....................268
20. Number of Tantilla re 1icta captured, and number "expected,"
in each scrub (see text)................................268
ix


21. Number of Gopherus polyphemus captured, and number
"expected," in each scrub (see text)....................269
22. Number of Sceloporus woodi captured, and number "expected,"
in each scrub (see text)................................269
23. Number of Masticophis flaqellum captured, and number
"expected," in each scrub (see text)....................270
24. Number of Gastrophryne carolinensis captured, and number
"expected," in each scrub (see text)....................270
25. Number of Cnemidophorus sexlineatus captured, and number
"expected," in each scrub (see text)....................271
26. Number of Neoseps reynoldsi captured, and number "expected,"
in each scrub (see text)................................271
27. Number of Podomys floridanus captured, and number
"expected," in each scrub (see text)....................272
28. Number of Bufo quercicus captured, and number "expected," in
each scrub (see text)...................................272
29. Number of Peromvscus polionotus captured, and number
"expected," in each scrub (see text)....................273
30. Number of Florida scrub jays located, and number "expected,"
in each scrub (see text)................................273
31. Relative abundances of 32 species of vertebrates captured in
trap arrays in AB3......................................284
32. Relative abundances of 32 species of vertebrates captured in
trap arrays in ARC......................................284
33. Relative abundances of 32 species of vertebrates captured in
trap arrays in HEN......................................285
x


34. Relative abundances of 32 species of vertebrates captured in
trap arrays in AB2......................................285
35. Relative abundances of 32 species of vertebrates captured in
trap arrays in AB1......................................286
36. Relative abundances of 32 species of vertebrates captured in
trap arrays in HH1......................................286
37. Relative abundances of 32 species of vertebrates captured in
trap arrays in JOS......................................287
38. Relative abundances of 32 species of vertebrates captured in
trap arrays in LAK......................................287
39. Relative abundances of 32 species of vertebrates captured in
trap arrays in JUN......................................288
40. Relative abundances of 32 species of vertebrates captured in
trap arrays in EGL......................................288
41. Relative abundances of 32 species of vertebrates captured in
trap arrays in DUF......................................289
42. Relative abundances of 32 species of vertebrates captured in
trap arrays in HH2......................................289
43. Relative abundances of 32 species of vertebrates captured in
trap arrays in COL......................................290
44. Relative abundances of 32 species of vertebrates captured in
trap arrays in CHU......................................290
45. Relative abundances of 32 species of vertebrates captured in
trap arrays in CUT......................................291
46. Relative abundances of 32 species of vertebrates captured in
trap arrays in BAR......................................291
xi




INTRODUCTION
Background
Species-area relationships have been recognized for more than a century, but interest in them has expanded dramatically over the past few decades. Examples of the relationship are common in the ecological literature: Connor and McCoy (1979) list 100 examples, and more than 100 additional ones have been published since their paper appeared (McCoy, unpublished). The traditional explanation for the relationship, increase in number of available habitats with increase in area (e_.g_. , Williams 1964), has been joined recently in importance by at least two others. Ecologists have realized that equilibrium processes can generate a species-area relationship even within a single habitat, largely because of the pioneering work of Preston (1960, 1962) and MacArthur and Wilson (1963, 1967). Nonequilibrium processes also can generate a species-area relationship within a single habitat (Connor and McCoy 1979, Coleman et �_1. 1982), but much less emphasis has been placed upon such processes to date. Connor and McCoy (1979), Coleman et al. (1982), Vincent and Haworth (1983), McGuinness (1984) and Williamson (1988) review the relationship from several perspectives.
Investigations over the past twenty years have lead to the conclusion that the species-area relationship may have important implications for conservation. The relationship has been shown to apply equally-well to "habitat islands" as to true islands.
1
k


Mountain tops, caves, forest patches, and many other types of isolates often support more resident species when they are relatively large than when they are relatively small (Brown 1971, Culver e_t aJU 1973, and Moore and Hooper 1975 provide classic examples). Nature reserves, for the most part, are habitat islands. It follows logically from the species-area relationship that to preserve the species richness characteristic of habitats, a premium automatically should be placed upon size when nature reserves are planned (Diamond 1975, Wilson and Willis 1975). Recent studies have suggested, however, that this conclusion may be too simplistic; and that serious errors might be made if emphasis were placed solely upon sizes of reserves (e_.g_. , Simberloff and Abele 1976, Connor and McCoy 1979, Kushlan 1979, McCoy 1983, Usher 1985, Zimmerman and Bierregaard 1986). The planning of nature reserves must take into account the autecologies of the species to be preserved.
How the structure of a reserve interacts with the autecologies of resident species is a subject of much current research and debate. For example, will a single large reserve necessarily contain more species than several small reserves of equal cumulative area (the "SLOSS" dilemma)? The answer seems to depend upon the particular situation (�_.�., Jarvinen 1982; Simberloff and Abele 1982; Kindlmann 1983; Simberloff and Gotelli 1984; Kobayashi 1985; Lahti and Ranta 1985, 1986; but see Willis 1984, Murphy and Wilcox 1986), likely as a result of the genetic and population attributes of the species involved (Helliwell
2


1976, Boecklen 1986b, Soule and Simberloff 1986). For another example, will the species composition of a relatively small reserve be a random subset of the composition of a relatively large reserve? The question is related to the one above, and the answer also seems to depend upon the particular situation (�_. g_. , Jones e_t al. 1985, Niemela et a_l. 1985, Nilsson 1986). In this case, population attributes of the species involved (e_.g_. , Gottfried 1979, Humphreys and Kitchener 1982, McCoy 1982, Blake and Karr.1984, Haila �_t a_l. 1987) and the regional setting in which reserves are located (e.g_. , Vaisanen e_t al. 1986, Askins et al. 1987) appear to be especially important in determining the species compositions of reserves. Other questions involve the potential roles of the shapes of reserves (see Game 1980, Blouin and Connor 1985) and corridors between reserves (see Simberloff and Cox 1987, Harris and Gallagher 1989) in influencing the species composition of resident organisms.
While it is clear that habitat fragmentation can contribute to the decline of species (e.g_. , Muggleton and Benham 1975, East 1983, Newmark 1987), precisely how it might do so is not understood well. We do not, in general, know enough about the autecologies of species even to predict for whom, and under what circumstances, island-biogeographic, or other models of species' decline might be appropriate (see Boecklen and Gotelli 1984, Jarvinen 1984, Simberloff and Abele 1984, Burgman et al. 1988, Pahl et al. 1988, Small and Hunter 1988). Some researchers have concluded from this lack of information that models of species'
3


decline, island-biogeographic models in particular, have little to offer conservation planning efforts (�_.g_. , Margules e_t al. 1982, Reed 1983). Other researchers, however, have concluded that potentially flawed models are better than no models at all, given the rapid loss of species and habitat worldwide (�_.g_. , East and Williams 1984, Kent 1987). Regardless of the perspective one chooses to take in this matter, it is clear that reserve design is no simple undertaking. Appropriate reserve design to reduce or prevent species' decline on fragments of habitat must include biological and social considerations well beyond those incorporated in simplistic models (Pickett and Thompson 1978, Margules and Usher 1981, Dony and Denholm 1985, Soule and Simberloff 1986, Usher 1986, Fernald 1989, Grumbine 1990). Hanski and Gilpin (1991), Hansson (1991), Harrison (1991), Saunders et al. (1991), and Sjoberg (1991) review fragmentation within the larger frameworks of landscape ecology and metapopulation dynamics.
Objectives
We shall address the two major questions posed above. In particular, we shall determine if a relatively large patch of scrub habitat tends to support more species than several small patches of equal cumulative area, and if the species' compositions of relatively small reserves tend to be random subsets of the composition of relatively large reserves. To make these determinations, we shall examine the distribution of
4


species of vertebrates among selected patches of scrub habitat in interior peninsular Florida, to determine if the presence of each species in a particular patch is related to the size of the patch (cf, Humphreys and Kitchener 1982, Jones e_t a_l. 1985, Dickman 1987). Next, we shall use the data obtained to hypothesize the role played by habitat fragmentation per se in determining the species compositions of a particular patch of scrub. Finally, we shall explore other kinds of data that are useful in understanding the role that habitat fragmentation plays in determining species composition of a patch of scrub (see Haila and Hanski 1984 ) .
The Scrub Habitat and Its Vertebrates
Christman (1988), Fernald (1989), and Myers (1990) provide detailed information on the scrub habitat in Florida. We provide only a few salient features of the habitat, and some comments about some of the vertebrates found there.
The scrub habitat is confined almost entirely to Florida; a few patches are located in coastal Alabama. The scrub habitat occurs, or formerly occurred, along the east coast of Florida from St. John's County in the north to Dade County in the south, along the northern Gulf Coast from Mobile Bay, Alabama, in the west to Franklin County in the east, and along the west coast of Florida from Levy County in the north to Collier County in the south (Davis 1943, 1967). Scrub habitat also occurs, or formerly occurred, throughout much of the interior of the Florida
5


peninsula. Among all of the scrubs in Florida, Christman (1988) takes particular note of some that are restricted to three elevated ridges near the center of the peninsula, the Lake Wales, Winter Haven, and Lake Henry Ridges, that appear to be previous shorelines. Christman (1988) refers to these scrubs as "ancient scrubs." He says that ancient scrubs may be recognized by the presence of one or more plant species that are both ecologically restricted to the scrub habitat and geographically restricted to the Lake Wales, Winter Haven, and Lake Henry Ridges (Christman 1988, Table 6). Most remaining patches of scrub are small, often less than a few hundred hectares in area. The Ocala Scrub, on The Ocala National Forest in the north-central part of Florida, however, is several thousand hectares in area.
The substratum underlying the scrub habitat are deep, well-washed and sorted, deposits of white siliceous sands. These deposits are extremely well drained and overlie more yellowish sands that are higher in clay content. The whiteness of the sand substratum and its extensive drainage promote a very hot, xeric environment at the surface during the summer. The substratum also loses heat rapidly, however, causing considerable diurnal temperature fluctuation.
A xerophytic plant association, characterized by the presence of sand pine (Pinus clausa), rosemary (Ceratiola ericoides) , and/or scrub oaks (Que reus chapmani, Q_. geminata, Q_. inopina, Q_. myrtif olia) , covers the white sand surface. Ground cover typically is sparse and patchy. Lichens (e.g., Cladonia
6


spp.) sometimes may be found in great abundance, growing on the surface. Most of the plant species growing in the scrub habitat are capable of rapid regeneration after the relatively-infrequent (30-50 year intervals) but intense fires that occur there. In the absence of fire, scrub tends to succeed to xerophytic hammock.
About 300 species of non-weedy plants have been collected from Florida scrubs, and a large percentage of these species (10-40%, depending on exactly how the scrub habitat is delimited in the field) may not occur in other habitats (Richardson 1989, see Christman 1988). The so-called "ancient scrubs" support an unusually-high concentration of rare plant species; about twenty species are found nowhere else. Many of these rare plant species, such as the pygmy fringe tree (Chionanthus pygmaeus), scrub mint (Dicerandra frutescens) , and scrub plum (Prunus geniculata), are listed as threatened or endangered by the federal government.
About 70 species of vertebrates have been collected regularly from Florida scrubs (Richardson 1989), and many more species may be found there on occasion (pers. obs.). Several of these species of vertebrates are not known to occur in other habitats, and nine of them (or subspecies) are protected, or under review for protection by the State and/or federal government. Christman (1988, Table 3) provides a list of vertebrate species "characteristic" of the scrub habitat. Individuals of characteristic species regularly include scrub
7


TABLE 1. Characteristic vertebrates of Florida's scrub habitat. Taxa marked with an asterisk are discussed in text.
AMPHIBIANS
Notophthalmus perstriatus Bufo terrestris
Eleutherodactylus p_. planirostris Hyla femoralis
Scaphiopus h. holbrooki Gastrophryne carolinensis
Bufo guercicus Rana areolata aesopus*
REPTILES
Gopherus polyphemus Rhineura floridana Anolis carolinensis Sceloporus woodi* Sceloporus u. undulatus Ophisaurus attenuatus
longicauda Ophisaurus ventralis
Cnemidophorus �_. sexlineatus Eumeces egregius onocrepis Eumeces egregius lividus* Eumeces inexpectatus Neoseps reynoldsi* Diadophis p_. punctatus Heterodon platyrhinos Heterodon simus
Ophyodrys aestivus Coluber constrictor priapus Masticophis flagellum Drymarchon corais couperi Elaphe g_. guttata Pituophis melanoleucus
mugitus Lampropeltis triangulum
elapsoides Cemophora c_. coccinea Stilosoma extenuatum* Virginia v. valeriae Tantilla r. relicta* Tantilla relicta pamlica* Micrurus f_. fulvius Sistrurus miliarius barbouri Crotalus adamanteus
9


BIRDS
Turkey vulture Sharp-shinned hawk Cooper's hawk
Southeastern American kestrel
Northern bobwhite
Mourning dove
Common ground dove
Eastern screech-owl
Common nighthawk
Chuck-will's-widow
Red-bellied woodpecker
Downy woodpecker
Hairy woodpecker
Northern flicker
Great crested flycatcher
Eastern phoebe
Blue jay
Florida scrub jay* Tufted titmouse Carolina wren
House wren
Blue-gray gnatcatcher Ruby-crowned kinglet Northern mockingbird Gray catbird Brown thrasher Loggerhead shrike White-eyed vireo Ye 1low-throated vireo Northern parula Yellow-rumped warbler Yellow-throated warbler Pine warbler Prairie warbler Palm warbler Northern cardinal Rufous-sided towhee Chipping sparrow American goldfinch
MAMMALS
Sorex lonqirostris Blarina carolinensis Cryptotis parva Scalops aquaticus Lasiurus intermedius Nycticeius humeralis Dasypus novemcinctus Sylvilagus palustris Geomys pinetis goffi Reithrodontomys humilis
Peromyscus polionotus Peromyscus gossypinus Podomys floridana* Ochrotomys nuttalli Microtus pinetorum Ursus americanus Spilogale putorius Felis rufus
Odocoileus virginianus
10


Ridge of peninsular Florida. McDiarmid (1978) suggested thatindividuals tend to be clumped within a scrub, reaching densities as high as 25 per acre in clumps. Its legs are much reduced, but not as much as in a true "sand swimmer," like Neoseps reynoldsi (sand skink). The bluetail mole skink eats invertebrates, such as cockroaches, crickets, and spiders.
Neoseps reynoldsi (sand skink) is highly specialized for a fossorial, "sand swimming," existence. Its eyes are reduced, it has no external ear openings, and its snout is wedge shaped with a partially-shielded lower jaw. The sand skink is restricted to well-drained sandy soils in the interior of peninsular Florida, from Marion County in the north to Highlands County in the south. Its diet consists of beetle larvae and termites (McDiarmid 1978).
Stilosoma extenuatum (short-tailed snake) is among the rarest reptiles in North America, although its cryptic, burrowing habit has prevented the gathering of much information about it. It is confined to Florida, and occurs from Columbia and Suwannee Counties in the north to Hillsborough, Orange, and Highlands Counties in the south. It is found primarily in the longleaf pine/turkey oak association and adjacent sand pine scrub. The diet of the short-tailed snake consists entirely of another snake characteristic of the scrub habitat, Tantilla relicta (Florida crowned snake) (McDiarmid 1978, Mushinsky 1984).
Tantilla relicta (Florida crowned snake) consists of several subspecies in Florida. It is confined to xeric and mesic habitats in peninsular Florida. Smith (1982), for example, found
11


it in principally in the longleaf pine association and young stands of sand pine scrub. He also found individuals, but at lower densities, in the longleaf pine/turkey oak association, xeric hammock, and relatively-thick live oak scrub. Likewise, Mushinsky (1985) has shown that abundance seems to be related to the frequency with which the longleaf pine/turkey oak association is burned. Florida crown snakes feed primarily on larvae of tenebrionid beetles (Smith 1982).
The Florida scrub jay (Aphelocoma c. coerulescens) has extremely narrow habitat requirements. The scrub jay resides in oak scrub, where cooperatively-breeding "families" defend territories. It is restricted largely to scattered, often small, patches of oak scrub in peninsular Florida. Small families (two breeders) may occupy patches of no more than 2 to 14 hectares, while larger families, of, say, eight individuals, may defend patches as large as 22 hectares (Woolfenden and Fitzpatrick 1984). Acorns form the bulk of the scrub jay's diet in autumn and winter, but throughout the remainder of the year, they eat mostly a variety of invertebrates and small vertebrates (Kale 1978).
Podomys floridana (Florida mouse) is the only native species of terrestrial mammal entirely restricted to Florida. It is distributed mostly throughout peninsular Florida, from St. Johns, Clay, Alachua, and Taylor Counties in the north to Sarasota, Highlands, and Dade Counties in the south. It has a broad diet, including seeds, nuts, fungi, and insects (Layne 1978). The
12


Florida mouse has one of the narrowest ranges of habitats among Florida's mammal species. Its primary habitat is early successional stages of sand pine scrub, but it also occurs commonly in longleaf pine/turkey oak and slash pine/turkey oak associations. Thus, the range of habitats of the Florida mouse encompasses relatively xeric open tree stands, with scattered-to-thick ground vegetation overlying well-drained sandy soils.
Decades of development have taken a severe toll on Florida's scrubs. Their perceived lack of value to humans has denied scrubs the sort of protection that has been afforded, say, wetlands. The powerful forces for development in the state have focused their activities on dry upland habitats, such as scrub and sandhill, with little restraint. For example, Peroni and Abrahamson (1985a, 1985b) estimated that by 1981, xeric habitats (scrub, scrubby flatwoods, sandhill) on the Lake Wales Ridge of Highlands County had suffered a loss to agriculture and real estate development of 23,578 ha, or about 64% of the original extent. Christman (1988) estimated the loss of scrub habitat on the Lake Wales, Winter Haven, and Lake Henry Ridges (which together stretch from Lake County in the north to Highlands County in the south) to be some 49,100 ha, or about 66% of the original extent.
13


METHODS
Study Sites
The primary focus of our research is to understand the role of the physical features of a scrub, particularly its size, in determining the taxonomic composition of resident vertebrates. It is clear from numerous studies that many variables other than size are important in determining taxonomic composition of isolates; in particular, habitat heterogeneity .(�_. g_. , Boecklen 1986a, Freemark and Merriam 1986), degree of isolation (�_.�., Adler and Wilson 1985, Jones et. al. 1985, Dickman 1987), and the pool of potential inhabitants present regionally (e_.g_. , McCoy 1983, Vaisanen e_t �_1. 1986). To account for the rich variety of physical features that influence taxonomic composition, we chose scrubs that: (1) were located within a single geographical region of the state, so that repeated and long-term sampling of a large number of scrubs was feasible; (2) varied in size, distance to other scrubs, distance to water, vegetational structure, and in other potentially-important ways; and (3) were made available to us by their owners.
We centered our study at Archbold Biological Station, in Highlands County, using it both as a study site and for housing during sampling trips. We located 15 additional study sites (Table 2), in Highlands and Polk Counties, that met the three criteria listed previously. When possible, we chose scrubs on
14


TABLE 2. Study sites.
CHRISTMAN (1988) ANCIENT PUBLIC
SCRUB LOCATION DESIGNATION HA ARRAYS SCRUB? LANI
LARGE SCRUBS
Hendrie Ranch(HEN) 30E, 39S, HIGH44,46 190* 12 Y N
Sec. 9&10
Archbold(ARC) 30E, 38S, HIGH22(PART), 179 10* * Y N
Sec. 30&31 33
HIGHT19(PART),
20(PART),35,44
Lake Arbuckle(AB3) 29E, 32S, POLK17 170 17 Y Y
Sec. 29&32
MEDIUM SCRUBS
Lake Arbuckle(ABl) 29E, 32S, POLK75 40 4 Y Y
Sec. 34&35
Highlands Hammock(HH1) 28E, 34S, 37 4 N Y
Sec . 32&33
Lake Arbuckle(AB2) 29E, 32S, POLK98 30 3 Y Y
Sec . 30
Josephine Creek(JOS) 29E, 36S, HIGH05 25 2 Y N
Sec. 3
SMALL SCRUBS
Church(CHU) 28E, 34S, HIGHT28 10 1 Y N
Sec . 23
Highlands Hammock(HH2) 28E, 35S, HIGH09 10 1 Y Y
Sec . 4
Lake June(JUN) 29E, 37S, HIGH98 10 1 Y N
Sec. 11
Eagle(EGL) 28E, 32S POLK20 6 1 Y N
Sec. 17
Lakemont(LAK) 29E, 34S, HIGH85 6 1 Y N
Sec. 7
Bar-D Ranch(BAR) 29E, 38S, HIGH65(PART) 5 1 N N
Sec. 9&16
College(COL) 28E, 33S, HIGHT16 3 1 Y Y
Sec. 34
Duffer's(DUF ) 28E, 34S , HIGH03 2 1 Y N
Sec. 3
Cute(CUT ) 28E, 34S HIGHT14 1. 5 1 Y N
Sec . 14
* Probably an overestimate, as not all of the area was considered suitable for placement of arrays.
** Long-term research projects restricted our placement of arrays (i.e., should have been 18 arrays, see below).
15


State- and federally-owned lands, because they provided the best protection possible for our trap arrays (see below) and they will be available for follow-up studies of scrub flora and fauna.
To address the relationship between numbers of taxa and size of scrubs, we used sites that fell into one of three size categories: large scrubs, that were about 200 hectares in area; medium scrubs, that were 25-50 hectares in area; and small scrubs, that were 10 hectares or less in area. Whenever possible, we selected scrubs recognized as "ancient" scrubs by Christman (1988). By doing so, we facilitated our location of potential study sites, and gained background knowledge of the existing plant community. All but one of our study sites were "ancient" scrubs.
Surrounding Habitat and Area Reduction
To characterize the habitat surrounding each scrub, we combined our direct knowledge of the habitat, information gained from reviewing recent aerial photographs, and ground truthing. For each scrub, we determined: (1) distance to the nearest scrub, (2) distance to the nearest larger scrub, (3) the presence/absence of potential "corridor" habitats (habitats with sandy substrate) between scrubs, (4) types of habitats between scrubs, (5) distance to the nearest permanent water, and (6) types, and percent coverage, of surrounding habitats. For the last determination, we
16


arbitrarily selected as "surrounding habitats" of a particular scrub any land within an area approximately four times the size of that scrub. Types of habitats recognized were scrub, upland, low flatwoods, wetland, grove, and disturbed.
To evaluate area reduction over time, we used a temporal series of aerial photographs of each scrub. Very little of the kind of development that noticeably reduces the sizes of scrubs had occurred along the central ridge of Florida prior to the mid-1940's. Aerial photographs of Highlands County taken in the 1940's, 1960's, 1970's, and 1980's were made available to us by Archbold Biological Station. We obtained aerial photographs of Polk County, taken at about the same times as those of Highlands County, from two sources. The oldest aerial photographs of Polk County, from the late 1940's, were located in the Florida Department of Transportation Office in Bartow. More recent aerial photographs were located in the Polk County Property Appraiser's Office.
Photographs of the scrubs were traced, and the tracings were digitized with SIGMASCAN software. We were then able to calculate the size of each scrub at several points in time over the past five decades, using the same software.
17


Vegetation Sampling
We used several methods to characterize the vegetational structure of the scrubs. Habitat structure (i..e_. , the physical arrangement of living and dead plant material) is a potentially important aspect of scrub vegetation to measure, because it potentially could affect the resident vertebrates dramatically. Habitat structure reflects the burn history of a scrub: scrubs that have burned recently have a reduced shrub and canopy layer, compared to scrubs that have not burned in several decades. To assess habitat structure, we took four samples spaced at ten-meter intervals from the distal end of each wing of each trap array (see below). For each sample, we characterized visually the vegetation in a quadrat four square meters in size. We estimated ground cover (0-lm level) as the proportion of the ground covered by living plants or litter; the proportion of the ground without cover we designated 'bare ground'. Canopy cover (l-3m and >3m levels) was determined as simple presence/absence.
We also collected data on the vegetational composition of each scrub. We selected several taxa of characteristic scrub plants, and visually estimated their relative abundances in the vicinity of each trap array. For some of the taxa, we made our estimates more finely, by estimating relative abundances of different size/age classes. The taxa and size classes that we selected were chosen because we
18


judged that they contributed most to the differences in the variety of "types" of scrub that can be recognized.
Trapping Methods for Non-avian Taxa
We used a arrays of pitfall and double-ended funnel traps, in conjunction with drift fences that divert moving animals into the traps (Campbell and Christman 1982). Each trap array consisted of four drift fences, of 7.6m lengths of aluminum valley material, arranged in a plus(+)-shaped pattern. Figures 1 to 16 illustrate the approximate placement of the trap array(s) in each scrub.
Twenty-liter plastic buckets were sunk into the ground, flush with the surface, at the eight ends of the fences. Bottoms of the buckets were drilled with small holes to facilitate drainage. When operational, plastic lids were supported with clothes pins about 5cm above the bucket; at other times the buckets were closed with the lids.
Funnel traps were constructed of 76cm-wide window screen, rolled to form a cylinder about 20cm in diameter. Pre-formed screen cones were fastened into the ends of the cylinders, to reduce the entrance to about 4 cm. Two funnel traps were placed along each drift fence, one on each side. The conical entrance at each end of a funnel was partially buried in the sand to encourage animals to crawl into the trap. Funnels were shaded with pieces of masonite to reduce the death rate of captured animals. The mortality rate in
19


FIGURE 1. Approximate placement of trap arrays in HEN
20


1
54
62
66 ^ ^ ^ 68 ^ 65 67
* 80
85
800'
FIGURE 2. Approximate placement of trap arrays in ARC
21




FIGURE 4. Approximate placement of trap arrays in AB1
23


24


400'
FIGURE 6. Approximate placement of trap arrays in AB2
25


FIGURE 7. Approximate placement of trap arrays in JOS


FIGURE 8. Approximate placement of the trap array in CHU
27


FIGURE 9. Approximate placement of the trap array in HH2
28


FIGURE 10. Approximate placement of the trap array in JUN
29


I
400'
FIGURE 11. Approximate placement of the trap array in EGL
30


I
200'
FIGURE 12. Approximate placement of the trap array in LAK
31


32


FIGURE 14. Approximate placement of the trap array in COL
33


FIGURE 15. Approximate placement of the trap array in DUF
34


35


funnels was still high, however. When not in use, funnels were removed and stored so that animals could not enter them accidentally.
Funnel traps of the same design as those described above, were placed in the mouths of gopher tortoise (Gopherus Polyphemus) burrows to supplement other trapping efforts. Funnel trapping of tortoise burrows was employed at sites where we suspected certain taxa (Rana areolata, in particular) to be present, but had not previously captured or observed them.
Small mammal trapping (Sherman Live Traps) was used, like the supplemental funnel trapping, at sites where certain taxa (rodents, in this case) had not previously been trapped or observed. We placed a pair of mammal traps every ten meters along a transect line in each scrub. Traps were baited with a mixture of peanut butter and rolled oats. Transect lines were 200 meters in length and located near the center of the scrub.
Trapping Schedule for Arrays
We divided the year into three "thermal" seasons, of four months each, to accommodate known differences in the activities of scrub vertebrates. The three seasons were Winter (November, December, January, and February), Spring-Fall (March, April, September, and October), and Summer (May, June, July, and August). Each of these four-month seasons
36


was divided into two two-month sampling periods, yielding a total of six trapping periods per calendar year (Table 3).
We used data accumulated during the first year of the study to modify our trapping schedule for the second year of the study (see Table 3). Based upon our first year's results, we had captured virtually all species anticipated at the large and taxonomically-rich scrub at Lake Arbuckle (AB3), and, therefore, we decided that additional trapping likely would not yield new information. We stopped trapping AB3 at the end of 1989, and added an additional small scrub (Duffer's = DUF). Hence, AB3 scrub was trapped only in 1989 and DUF scrub only in 1990.
Originally, we had anticipated trapping scrubs for ten-day intervals during each sampling period. Furthermore, had we anticipated selecting eight scrubs (two large, two medium, and four small) to be trapped during each period. The experience we gained during the first two sampling periods of 1989 suggested to us that a slightly different regime would be more fruitful. During the remainder of 1989, we sampled all four medium and eight small scrubs, and two of the three large scrubs in each sampling period. In 1990, we shortened the duration of trapping in each sampling period, from ten days to seven. Because we did not trap the large scrub AB3 in 1990, and shortened the duration of trapping, we were able to sample all 15 scrubs in each sampling period of 1990. Under the new trapping regime, approximately half of the
37


Table 3. Trapping schedule for vertebrates in 16 scrubs of central Florida. "X" means that a scrub was trapped during the sampling period indicated.
SCRUB 1989 1990
1 2 3 4 5 6 1 2 3 4 5 6
Hendrie Ranch X X X X X X X X X X
Archbold X X X X X X X X X X X
Lake Arbuckle (3) X X X X
Lake Arbuckle (1) X X X X X X X X X X X X
Highlands Hammock (1) X X X X X X X X X X
Lake Arbuckle (2) X X X X X X X X X X X X
Josephine Creek X X X X X X X X X X X X
Church X X X X X X X X X X
Highlands Hammock (2) X X X X X X X X X X
Lake June X X X X X X X X X
Eagle X X X X X X X X X X X
Lakemont X X X X X X X X X X X X
Bar-D Ranch X X X X X X X X X X
College X X X X X X X X X X X X
Duffer's X X X X X X
Cute X X X X X X X X X X
38


scrubs were trapped for 7 days, and the other half for the following 7 days in each sampling period. The gain in employing this new regime was that each scrub was sampled more often over the course of a year than under the old regime; the loss was that each scrub was sampled for three days less per sampling period.
When open, traps were checked for captured animals daily. Each taxon of captured individual was processed according to the following procedures. Amphibians were identified in the field and released at the point of capture. Reptiles were identified and given a unique number by toe clipping (lizards) or caudal scale clipping (snakes), prior to release at the point of capture. Rodents were identified and a uniquely numbered ear tag was applied to each individual prior to release at the point of capture. These procedures facilitated our use of only original captures to calculate the relative abundance of reptiles and mammals at our study sites.
Other Sampling Methods for Non-avian Taxa
We sampled gopher tortoises (Gopherus polyphemus) by walking a total of 264 transects, each 300 meters long and seven meters wide, in the 16 scrubs. We recorded the number of "active," "inactive," and "abandoned" burrows encountered on each transect (see McCoy and Mushinsky 1988). To convert our counts of burrows to an estimate of the resident tortoise population at each site, we used the method of McCoy and
39


Mushinsky (in review). This method uses burrows classified as "active" (those with footprints or plastron abrasions visible in their mouths) as the basis for population estimates.
We marked areas 10m in length by 5m in width with stake-wire flags at each scrub. We then raked these areas clean, and monitored them for animal signs, footprints, in particular. This method was labor intensive, time consuming, and provided little new information, except to document the presence of several taxa of large mammals.
When appropriate (after an afternoon or evening rain), we drove on roads around the scrubs, to check for road crossings by vertebrates. We had considerable success in spotting individuals, particularly of amphibians, but no new species were added to our species lists by employing this method.
We made numerous walks through each scrub to look for vertebrates or signs of their presence (�_.�., scat, footprints, tracks). This method proved to be time consuming, and yielded few taxa not captured by employing other methods.
Avian Sampling
Birds were censused twice yearly. One census, which included migratory birds, was made in January, February, and March, and the other census, which included breeding birds, was made in April, May, and June (Table 4). Transects were taken through each scrub in the morning (from sunrise to about 1100h) to identify birds visually and by their calls. The amount of
40


TABLE 4.
Sampling schedule for breeding birds (April, May, June) and migratory birds (January, February, March) in 16 scrubs during 1989 and 1990.
SCRUB 1989 1990
MIGRATORY BREEDING MIGRATORY BREEDING
Hendrie Ranch X X X
Archbold X X X X
Lake Arbuckle (3) X X X X
Lake Arbuckle (1) X X X X
Highlands Hammock (1) X X
Lake Arbuckle (2) X X X X
Josephine Creek X X X X
Church X X
Highlands Hammock (2) X X
Lake June X X
Eagle X X X
Lakemont X X X X
Bar-D Ranch X X
College X X X X
Duffer's X X
Cute X X
i
41


time spent, and the number of transects taken, in each scrub were scaled to the size of the scrub. We considered territorial defense and nest building behavior to be indicators of breeding activity. Some of our scrubs could not be included in the 1989 censuses, because we did not obtain permission to use them until after the breeding season.
Analytic
Because not all scrubs were trapped during each trapping period in 1989, we adjusted the capture data to reflect a complete trapping effort for all scrubs. To do so, we used the data on relative abundance from scrubs which were trapped during a given period to calculate an expected number of individuals of each species known to occur in a scrub. This procedure is sensitive to seasonal variation of a species' relative abundance. We used a similarly procedure to adjust the capture data for the two scrubs which were trapped for only one year to estimate the relative abundance of individuals of species known to occur in that scrub during the year in which it was not trapped. We did this for DUF in 1989 and for AB3 in 1990.
We employed several common univariate statistical techniques throughout our analyses. These techniques include the Spearman Rank Correlation Test of Independence, the Wilcoxon Rank Sum Test, and the Kolmogorov-Smirnov Test for Goodness of Fit (all nonparametric techniques), and Multiple
42


Correlation Analysis (a parametric technique). As these are common techniques, we will not discuss them in any detail.
For classification of scrubs (SU's = sampling units), we employed three multivariate techniques. The first was monothetic divisive association analysis (Williams and Lambert 1959). This technique uses contingency tables to split SU's into homogeneous groups repeatedly, based, at each step, on the presence or absence of a particular, "divisor," taxon. We used the BASIC program of Ludwig and Reynolds (1988) to accomplish the analysis. A group was considered "homogeneous" when its chi-square value was less than the critical value for p < 0.05.
The second and third techniques that we employed actually were qualitative and quantitative versions of the same analysis, polythetic agglomerative cluster analysis (Gauch 1982, Pielou 1984). This analysis uses the matrix of Q-mode resemblances among SU's to "sort" the SU's into groups or R-mode resemblances to "sort" attributes (measures of vegetational structure, in this case) into groups. It begins with the collection of individual SU's, and at each clustering step, builds groups of similar SU's. We used the SYN-TAX IV program NCLAS (Podani 1990) to accomplish the analysis. The distance functions we chose were relative measures, either chord distance for binary data (qualitative) or chord distance (quantitative), and the clustering method was flexible (beta = -0.25). We discriminated meaningful groups subjectively.
43


We employed two techniques for detecting associations among taxa. The first was Schluter's (1984) Variance Ratio Test, which tests for overall association, among all of the taxa simultaneously. The other technique was inverse association analysis, which uses contingency tables, like association analysis, above, but to search for associations among taxa, rather than among SU's. We used BASIC programs of Ludwig and Reynolds (1988) to accomplish these analyses.
To determine if the taxonomic compositions of relatively-poor scrubs are nested within those of relatively-rich scrubs, we employed the technique of Patterson and Atmar (1986). By 'nested', we mean "that the (taxa) comprising a depauperate fauna should constitute a proper subset of those in richer faunas..." (Patterson and Atmar 1986, p. 65). Furthermore, "...the archipelago of such faunas arranged by (taxonomic) richness should present a nested series" (Patterson and Atmar 1986, p. 65). The technique is a Monte Carlo simulation that generates an expected level of nestedness by distributing taxa among SU's at random. The mean and variance of a large number of simulations can then be used to determine if the observed level of nestedness deviates from random expectation. We used a modified version of the BASIC program provided by Patterson and Atmar (1986), and weighted the selection of taxa by their actual frequencies of occurrence (RAND0M1 option in the program). We used 1000 simulations to generate the mean and variance of nestedness based on random expectation.
44


We employed Rarefaction Analysis (Hurlbert 1971, James and Rathbun 1981) to determine the number of taxa in relatively-poor scrubs can be predicted from the number of taxa in relatively-rich scrubs. The technique generates a number of taxa expected for samples of individuals drawn from the abundances of taxa in a parent distribution. These samples are equal to the number of individuals in the relatively-poor scrubs of interest. We used the BASIC program of Ludwig and Reynolds (1988) for the analysis. Note that the technique is known to possess a very restrictive set of assumptions (Peet 1974, Tipper 1979 ) .
Finally, we used Principal Coordinates Analysis to order scrubs, based on selected attributes of vegetational structure. The technique computes a score for each SU from the weighted sums of the structural attributes. We employed the SYN-TAX IV program PRINCOOR (Podani 1990) to accomplish the analysis.
45


RESULTS
Habitat Structure
We present our data on habitat structure as a series of sheets, arranged alphabetically by scrub. The data for each scrub include (1) the name we gave it, (2) its location, (3) the alphanumeric designation that Christman (1988) gave it, (4) our estimate of its size, (5) the number of trap arrays that we used, (6) the distance to the nearest other scrub, (7) the habitat type(s) between the scrub in question and the nearest other scrub, (8) the distance to the nearest larger scrub, (9) the habitat type(s) between the scrub in question and the nearest larger scrub, (10) whether or not corridor habitat (well-drained sands) surrounds it, (11) distance to nearest permanent water, (12) the identity of nearby (within an area four times the size of the scrub itself) habitat type(s), and (13) a record of the decline in size of the scrub over the past 45-50 years.
We present some other data on habitat structure separately for each trap array. These data include (1) a record of the relative abundance of selected plant species within the vicinity of the array (asterisks indicate species recorded as present by Christman (1988), but not recorded by us as present near any array), (2) a photograph taken at the array, (3) an estimate of percent cover of low (l-3m) and high (>3m) canopy, (4) an estimate of percent ground ( 46


HENDRIE SCRUB
SURROUNDINGS:
(1) NEAREST SCRUB - 470m
(2) HABITAT BETWEEN - Low Flatwoods, Disturbed
(3) NEAREST LARGER SCRUB - 3520m (Approximately Same Size)
(4) HABITAT BETWEEN - Scrub, Grove, Disturbed
(5) CORRIDOR HABITAT - Yes
(6) NEAREST PERMANENT WATER - 70m
(7) NEARBY HABITAT - Scrub (10%), Low Flatwoods (15%),
Wetland (25%), Grove (20%), Disturbed (30%)
-1000/AREA
-40
1940 1950 1960 1970 1980 1990
YEAR
47


HENDRIE
PLANT CATEGORY
Pinus clausa (young) (mature)
P. elliottii. P. palustris
Quercus inopina (<0.5m)
(0.5-lm) (l-2m) (>2m)
2.. qeminata (<0.5m)
(0.5-lm) (l-2m) (>2m)
Q.. chapmanil (<0.5m)
(0.5-lm) (l-2m) (>2m)
Q_. myrtifolia Carya
Ceratiola ericoides (small)
(large)
SCRUB (ARRAY 1)
ABSENT RARE MODERATE ABUNDANT
X X
X
X X X X X X
X X
X X X X
X X
X X
Erynqium, Hypericum, Polygonella PRESENT
Calamintha, Conradina ABSENT
Persea, Ilex opaca PRESENT
Lyonia, I., glabra, Sabal PRESENT
48


20
15
10
WINTER
PERCENT OF QUADRATS
111
1
0 5 10 15 20 25 30 35 40 45 50 65 60 65 70 75 80 66 90 96100
PERCENT COVER
? BARE GROUND-48%
I GROUND COVER-62%
SUMMER
PERCENT OF QUADRATS
0 5 10 16 20 26 30 36 40 45 60 56 80 65 70 75 80 85 90 96100
PERCENT COVER
? BARE GROUND-51%
I GROUND COVER-49%
49
35


PERCENT COVER
LOW CANOPY HIGH CANOPY
CANOPY LEVEL
50


HENDRIE
PLANT CATEGORY
Pinus clausa (young) (mature)
P. elliottii, P. palustris
Quercus inopina (<0.5m)
(0.5-lm) (l-2m) (>2m)
Q.. geminata (<0.5m)
(0.5-lm) (l-2m) (>2m)
Q_. chapmanii (<0.5m)
(0.5-lm) (l-2m) (>2m)
Q.. myrtif olia Carya
Ceratiola ericoides (small)
(large)
SCRUB (ARRAY 2)
ABSENT RARE MODERATE ABUNDANT X
X
X
X X X
X
X X
X
X X X X X X
X
X
X
Eryngium, Hypericum, Polyqonella PRESENT
Calamintha, Conradina ABSENT
Persea, Ilex opaca PRESENT
Lyonia, I. glabra, Sabal PRESENT
51


WINTER
PERCENT OF QUADRATS
0 6 10 16 20 25 30 35 40 4 6 60 56 80 65 70 76 80 85 90 96100
PERCENT COVER
? BARE GROUND-38% � GROUND COVER-62%
SUMMER
PERCENT OF QUADRATS
0 6 10 15 20 25 30 35 40 45 60 65 60 66 70 76 80 86 90 96100
PERCENT COVER ? BARE GROUND-34% � GROUND COVER-66*
52


PERCENT COVER
LOW CANOPY HIGH CANOPY
CANOPY LEVEL
i
53


HENDRIE
PLANT CATEGORY
Pinus clausa (young) (mature)
P. elliottii. P. palustris
Ouercus inopina (<0.5m)
(0.5-lm) (l-2m) (>2m)
Q.. qeminata (<0.5m)
(0.5-lm) (l-2m) (>2m)
Q_. chapmanii (<0.5m)
(0.5-lm) (l-2m) (>2m)
Q.. myrtifolia Carya
Ceratiola erlcoides (small)
(large)
SCRUB (ARRAY 3)
ABSENT RARE MODERATE ABUNDANT X
X
X X X X
X X X X X
X X X
X X X
X
X
Erynqium, Hypericum, Polyqonella PRESENT
Calamintha, Conradina ABSENT
Persea, Ilex opaca PRESENT
Lyonia, I. glabra, Sabal PRESENT
54


WINTER
PERCENT OF QUADRATS
0 5 10 15 20 25 30 36 40 45 50 55 60 65 70 75 60 65 90 96100
PERCENT COVER ? BARE GROUND-38% � GROUND COVER-62*
SUMMER
PERCENT OF QUADRATS
0 6 10 16 20 25 30 36 40 46 50 65 60 66 70 75 60 86 90 96100
PERCENT COVER ? BARE GROUND-45% � GROUND OOVER-56*
55


PERCENT COVER
LOW CANOPY HIGH CANOPY
CANOPY LEVEL


HENDRIE SCRUB (ARRAY 4) PLANT CATEGORY ABSENT RARE MODERATE ABUNDANT
Pinus clausa (young) X (mature) X P. elliottii. P. palustris X Quercus inopina (<0.5m) X
(0.5-lm) X
(l-2m) X
(>2m) X Q.. qeminata (<0.5m) X
(0.5-lm) X
(l-2m) X
(>2m) X Q.. chapmanii (<0.5m) X
(0.5-lm) X
(l-2m) X
(>2m) X Q_. mvrtif olia X Carya X Ceratiola ericoides (small) X
(large) X
Erynqium, Hypericum, Polyqonella PRESENT
Calamintha, Conradina PRESENT
Persea, Ilex opaca ABSENT
Lyonia, I. glabra, Sabal ABSENT
57


WINTER
PERCENT OF QUADRATS
0 6 10 16 20 26 30 36 40 46 60 66 60 85 70 75 80 86 90 95100
PERCENT COVER n BARE GROUND-38% � GROUND COVER-62%
SUMMER
PERCENT OF QUADRATS
n n n I � I i II I
0 5 10 15 20 25 30 35 40 45 50 55 60 65 70 75 80 85 90 96100
PERCENT COVER ? BARE GROUND-44% � GROUND COVER-66*
58


59


Q.. qeminata
PLANT CATEGORY Pinus clausa (young)
(mature) P. elliottii. P. palustris Quercus inopina (<0.5m)
(0.5-lm) (l-2m) (>2m) (<0.5m) (0.5-lm) (l-2m) (>2m) (<0.5m) (0.5-lm) (l-2m) (>2m)
Q.. myrtifolia Carya
Ceratiola ericoides (small)
(large)
HENDRIE SCRUB (ARRAY 5)
ABSENT RARE MODERATE ABUNDANT X
Q.. chapmanii
X X
X X
X X X X
X
X X X
X X X
Erynqium. Hypericum, Polyqonella Calamintha, Conradina Persea. Ilex opaca Lyonia, I. glabra, Sabal
PRESENT PRESENT ABSENT PRESENT
60


WINTER
PERCENT OF QUADRATS
0 5 10 15 20 26 30 36 40 45 50 65 60 65 70 75 80 85 90 96100
PERCENT COVER ? BARE GROUND-17% � GROUND COVER-83%
SUMMER
PERCENT OF QUADRATS
0 6 10 16 20 25 30 35 40 45 50 65 60 85 70 75 80 85 90 95100
PERCENT COVER
? BARE GROUND-20%
I GROUND COVER-80*
61


PERCENT COVER
LOW CANOPY HIGH CANOPY
CANOPY LEVEL
62


HENDRIE
PLANT CATEGORY
Pinus clausa (young) (mature)
P. elliottii, P. palustris
Quercus inopina (<0.5m)
(0.5-lm) (l-2m) (>2m)
Q.. qeminata (<0.5m)
(0.5-lm) (l-2m) (>2m)
Q_. chapmanii (<0.5m)
(0.5-lm) (l-2m) (>2m)
2_. mvrtifolia Carya
Ceratiola ericoides (small)
(large)
SCRUB (ARRAY 6)
ABSENT RARE MODERATE ABUNDANT X
X
X
X X
X
X
X X X
X
X X X
X X
X
X
X
Ervngium, Hypericum, Polyqonella PRESENT
Calamintha, Conradina PRESENT
Persea, Ilex opaca ABSENT
Lyonia, I. glabra, Sabal PRESENT
63


WINTER
percent of quadrats
0 5 10 16 20 26 30 36 40 46 60 56 60 65 70 75 80 86 00 96100
percent cover
? BARE GROUND-61% � GROUND COVER-39*
SUMMER
percent of quadrats
0 5 10 15 20 26 30 35 40 45 60 55 60 65 70 75 80 85 90 95100
percent cover
? BARE GROUND-38*
I GROUND COVER-67*
64


F
percent cover
LOW CANOPY HIGH CANOPY
canopy level
65


HENDRIE
PLANT CATEGORY
Pinus clausa (young) (mature)
P. elliottii, P.. palustris
Quercus inopina (<0.5m)
(0.5-lm) (l-2m) (>2m)
Q_. qeminata (<0.5m)
(0.5-lm) (l-2m) (>2m)
Q.. chapmanii (<0.5m)
(0.5-lm) (l-2m) (>2m)
Q_. myrtifolia Carya
Ceratiola ericoides (small)
(large)
SCRUB (ARRAY 7)
ABSENT RARE MODERATE ABUNDANT X
X
X
X X X
X X X X
X
X X X
X X
X X
X
Ervngium, Hypericum. Polygonella PRESENT
Calamintha, Conradina PRESENT
Persea, Ilex opaca ABSENT
Lyonia, I. glabra, Sabal PRESENT
66


WINTER
percent of quadrats
0 5 10 16 20 25 30 35 40 45 60 66 60 65 70 75 80 86 90 95100
percent cover
? BARE GROUND-25%
I GROUND COVER-75%
SUMMER
percent of quadrats
n r II i
0 5 10 15 20 25 30 35 40 45 50 55 60 65 70 76 80 86 90 96100
percent cover ? BARE GROUND-24% � GROUND OOVER-76%
67


Q_. qeminata
PLANT CATEGORY Pinus clausa (young)
(mature) P. elliottii. P. palustris Quercus inopina (<0.5m)
(0.5-lm) (l-2m) (>2m) (<0.5m) (0.5-lm) (l-2m) (>2m) (<0.5m) (0.5-lm) (l-2m) (>2m)
Q.. myrtifolia Carva
Ceratiola ericoides (small)
(large)
HENDRIE SCRUB (ARRAY 8)
ABSENT RARE MODERATE ABUNDANT X
Q.. chapmanli
X
X
X X X
X
X
X
X X X X
X X X
X X X
Erynqium, Hypericum, Polygonella PRESENT
Calamintha, Conradina ABSENT
Persea, Ilex opaca ABSENT
Lyonia, J., glabra, Sabal PRESENT
69


WINTER
percent of quadrats
0 5 10 15 20 25 30 35 40 45 50 65 60 66 70 75 BO 65 00 06100
percent cover ? bAREGROUND-31% � GROUND COVER-69*
SUMMER
percent of quadrats
0 6 10 15 20 26 30 35 40 45 60 66 60 66 70 76 80 66 90 95100
percent cover
? BARE GROUND-32%
I GROUND COVER-66%
70


71


Q.. geminata
PLANT CATEGORY Pinus clausa (young)
(mature) P. elliottii. P. palustris Quercus inopina (<0.5m)
(0.5-lm) (l-2m) (>2m) (<0.5m) (O.5-lm) (l-2m) (>2m) (<0.5m) (0.5-lm) (l-2m) (>2m)
Q.. myrtifolla Carva
Ceratlola erlcoides (small)
(large)
HENDRIE SCRUB (ARRAY 9)
ABSENT RARE MODERATE ABUNDANT X
Q.. chapmanli
X X X X X
X X
X X X X X
X
X X X
X
Eryngium. Hypericum. Polvqonella Calamintha. Conradina Persea, Ilex opaca Lyonia. I., glabra. Sabal
PRESENT PRESENT ABSENT PRESENT
72


WINTER
PERCENT OF QUADRATS
0 6 10 15 20 26 30 36 40 46 50 66 80 66 70 76 80 85 00 05100
PERCENT COVER ? BARE GROUND-32* � GROUND COVER-68*
SUMMER
PERCENT OF QUADRATS
0 5 10 16 20 25 30 35 40 45 60 65 80 86 70 76 80 86 00 06100
PERCENT COVER
? BARE GROUND-44*
I GROUND COVER-58%


percent cover
LOW CANOPY HIGH CANOPY
canopy level
74


PLANT CATEGORY Pinus clausa (young)
(mature) P. elliottii. P. palustris Quercus inopina (<0.5m)
(0.5-lm)
Q.. gemlnata
Q.. chapmanii
(l-2m) (>2m) (<0.5m) (0.5-lm) (l-2m) (>2m) (<0.5m) (0.5-lm) (l-2m) (>2m)
Q.. myrtifolia Carya
Ceratiola ericoides (small)
(large)
HENDRIE SCRUB (ARRAY 10)
ABSENT RARE MODERATE ABUNDANT X
X X X X X X
X X
X X X X
X X X X
X
X
Eryngium, Hypericum. Polygonella Calamintha. Conradina Persea. Ilex opaca Lyonia. I. glabra. Sabal
PRESENT ABSENT PRESENT PRESENT
75


WINTER
PERCENT OF QUADRATS
0 6 10 16 20 26 30 36 40 46 60 66 60 66 70 76 80 86 90 96100 PERCENT COVER
? BARE GROUND-G8* � GROUND COVER-82%
SUMMER
PERCENT OF QUADRATS
0 6 10 16 20 26 30 36 40 45 60 66 60 66 70 76 80 86 00 06100
PERCENT COVER ? BARE GROUND-37% � GROUND COVER-63%
76


PERCENT COVER
LOW CANOPY HIGH CANOPY
CANOPY LEVEL
77


Q.. geminata
PLANT CATEGORY Pinus clausa (young)
(mature) P. elliottii. P. palustris Quercus inopina (<0.5m)
(0.5-lm) (l-2m) (>2m) (<0.5m) (0.5-lm) (l-2m) (>2m) (<0.5m) (0.5-lm) (l-2m) (>2m)
Q.. myrtlfolla Carva
Ceratlola ericoides (small)
(large)
HENDRIE SCRUB (ARRAY 11)
ABSENT RARE MODERATE ABUNDANT X X
Q.. chapmanil
X
X X X X X X X X X X
X X X X
X X
Ervngium. Hypericum. Polygonella Calamintha. Conradina Persea. Ilex opaca Lyonia. I. glabra. Sabal
PRESENT PRESENT PRESENT PRESENT
78


Q.. geminata
PLANT CATEGORY Pinus clausa (young)
(mature) P. elliottii. P. palustris Quercus inopina (<0.5m)
(0.5-lm) (l-2m) (>2m) (<0.5m) (0.5-lm) (l-2m) (>2m) (<0.5m) (0.5-lm) (l-2m) (>2m)
Q.. myrtifolla Carya
Ceratiola ericoides (small)
(large)
HENDRIE SCRUB (ARRAY 12)
ABSENT RARE MODERATE ABUNDANT
X
Q.. chapmanii
X
X X
X
X X X
X X X
X
X
X X X
X X
Ervngium. Hypericum. Polygonella Calamintha. Conradina Persea. Ilex opaca Lyonia. I. glabra, Sabal
PRESENT PRESENT PRESENT PRESENT
81


20
16
10
WINTER
PERCENT OF QUADRATS
i
LJI
0 6 10 15 20 26 30 36 40 46 60 65 50 66 70 75 80 86 00 06100
PERCENT COVER
? BARE GROUND-41% � GROUND COVER-SO*
SUMMER
PERCENT OF QUADRATS
0 6 10 16 20 26 30 86 40 46 60 66 60 86 70 76 80 86 80 06100
PERCENT COVER ? BARE GROUND-41% � GROUND COVER-69%
82


PERCENT COVER
LOW CANOPY HIGH CANOPY
CANOPY LEVEL
83


ARCHBOLD SCRUB
SURROUNDINGS t
(1) NEAREST SCRUB - 120m
(2) HABITAT BETWEEN - Upland, Wetland
(3) NEAREST LARGER SCRUB - SAME
(4) HABITAT BETWEEN - SAME
(5) CORRIDOR HABITAT - Yes
(6) NEAREST PERMANENT WATER - 650m (Seasonal Ponds Within)
(7) NEARBY HABITAT - Scrub (36%), Low Flatwoods (29%),
Wetland (8%), Grove (5%), Disturbed (22%)
-1000/area
1940 1950 1960 1970 1980 1990
year
84


ARCHBOLD
PLANT CATEGORY
Pinus clausa (young) (mature)
P. elliottii. P. palustris
Quercus inopina (<0.5m)
(0.5-lm) (l-2m) (>2m)
Q.. geminata (<0.5m)
(0.5-lm) (l-2m) (>2m)
�. chapmanii (<0.5m)
(0.5-lm) (l-2m) (>2m)
Q.. mvrtifolia Carva
Ceratiola ericoides (small)
(large)
SCRUB (ARRAY 1)
ABSENT RARE MODERATE ABUNDANT X X
X
X X X
X
X X X
X
X X X
X
X*
X
X
X
Ervngium. Hypericum. Polygonella PRESENT
Calamintha. Conradina PRESENT
Persea. Ilex opaca PRESENT
Lvonia. I. glabra, Sabal PRESENT
85


WINTER
PERCENT OF QUADRATS
III 1 ll ,1,
0 6 10 16 20 26 30 36 40 45 60 66 60 66 70 76 60 85 90 96100 PERCENT COVER
? BARE GROUND-36% � GROUND COVER-64*
SUMMER
PERCENT OF QUADRATS
0 6 10 16 20 26 30 36 40 45 60 66 80 66 70 75 80 86 90 95100
PERCENT COVER ? BARE GROUND-22% � GROUND COVER-78*
86


87


ARCHBOLD
PLANT CATEGORY
Pinus clausa (young) (mature)
P. elliottii. P. palustris
Quercus inopina (<0.5m)
(0.5-lm) (l-2m) (>2m)
Q.. geminata (<0.5m)
(0.5-lm) (l-2m) (>2m)
Q.. chapmanii (<0.5m)
(0.5-lm) (l-2m) (>2m)
Q.. mvrtifolia Carva
Ceratiola ericoides (small)
(large)
SCRUB (ARRAY 2)
ABSENT RARE MODERATE ABUNDANT X X
X
X X X
X
X X X
X
X X X
X
X*
X
X
X
Eryngium. Hypericum. Polygonella PRESENT
Calamintha. Conradina PRESENT
Persea. Ilex opaca PRESENT
Lvonia. I. glabra. Sabal PRESENT
88


WINTER
PERCENT OF QUADRATS
0 6 10 16 20 26 30 36 40 46 60 66 60 66 70 76 80 86 90 96100
PERCENT COVER
? BARE GROUND-34% � GROUND COVER-88%
SUMMER
PERCENT OF QUADRATS
0 6 10 16 20 26 30 36 40 46 60 66 60 66 70 76 80 86 90 96100
PERCENT COVER ? BARE GROUND-34% � GROUND COVER-66�
89


90