An Evaluation of the Applicability of Upland Buffers for the Wetlands of the Wekiva Basin

Final Report

AN EVALUATION OF THE APPLICABILITY OF
UPLAND BUFFERS FOR THE WETLANDS OF THE WEKIVA BASIN

Mark T. Brown and Joseph M. Schaefer, Principal Investigators

with

K. H. Brandt, S. J. Doherty, C. D. Dove, J. P. Dudley,
D. A. Eifler, L. D. Harris, R. F. Noss, and R. W. Wolfe

October, 1987

Center for Wetlands
University of Florida
Gainesville, Florida 32611
(904) 392-2424

Acknowledgements

The authors wish to express their appreciation to the staff of the Center

For Wetlands, for their dedication and service way beyond the call of duty.

Their attitude made completion of this report so much easier. Specifically,

Jenny Carter, Staff Assistant, coordinated personnel and somehow managed to get

the report finalized under difficult odds. Linda J. Crowder processed all the

words over and over again as we edited and polished. Carol Cox proofread each

draft. Stephen Roguski, our expert draftsperson, drafted the figures. Karla

Brandt not only helped with research and writing, but did a fantastic job of

final editing. Steven Tennenbaum devoted two days to derivations of the Theis

Equation. We thank you for your energy.

Ms. Sidney Brinson, spent a day in the field with the authors explaining

the Districts methodology for wetlands determination. Staff from the Florida

Department of Natural Resources, especially Ms. Deborah Shelly, provided

support and transportation.

Glenn Lowe, Chief Environmental Specialist a the St. Johns River Water

Management District was project manager for the District and was extremely

patient and effective in his support.

Preface

This document is the product of a contract between the St. Johns River

Water Management District and the Center for Wetlands, University of Florida to

evaluate the applicability of upland buffers to the wetlands of the Wekiva

River Basin. The purpose of this report is to develop critical insight

concerning the need for, potential applicability of, and criteria for

delineating upland buffers in the Wekiva Basin. In this report, we have

reviewed the literature related to buffers, wetlands, wildlife habitat,

transition zones, water quality and quantity, and other models and criteria for

determining buffer zones. We have reviewed this information as a means of

evaluating the need for an upland buffer to protect the water resources of the

Wekiva River System.

Whether an upland/wetland buffer zone is desirable and/or necessary for

the Wekiva River System is not a question in the minds of the authors. Our

review of the literature; our understanding of the unique character of the

Wekiva River System; our knowledge of the limitations of the criteria for

issuing of permits for construction, operation, and maintenance of stormwater

systems; the limits of jurisdiction under current wetlands regulation; and the

expressed policies of the Governing Board of the St. Johns River Water

Management District suggest that an upland buffer is desirable and necessary if

the wetlands, waters, and wildlife of the Wekiva River Basin are to be

preserved.

Two issues have surfaced during our study which need clarification as a

prelude to the report. Both are related to the language in Chapter 373,

Florida Statutes. In 373.413 and 373.416 a distinction is drawn between

permits for construction and permits for operation and maintenance of water

management systems. Both kinds of permits require that such systems cannot be

harmful to the water resources of the District. Operation and maintenance

permits are subject to an additional criterion: such activities cannot be

inconsistent with the overall objectives of the District. Using one standard

for construction and two for operation and maintenance is confusing. Operation

and maintenance cannot be separated from construction, since once constructed,

a system normally will be operated. Or to put it another way, if constructed,

it seems backwards to then evaluate a permit for operation and find a system

inconsistent with District objectives. Such a finding should be made prior to

construction so that the project may be altered or redesigned to conform to

District objectives. Thus, in our view, the review process for construction

permit applications must also include operation and maintenance concerns. The

District needs to review whether the constructed system will be harmful to the

water resources and whether it will be consistent with the District's overall

objectives during the construction permit review process.

The second issue is related to the definition of "water resources." We

believe that one cannot separate the waters of the District from the overall

aquatic system when determining harm to the resource. The water resource must

include not only the abiotic substance (H20), but also the other abiotic

substances and the biotic organisms that are carried and sustained by it. The

water resource cannot be artificially dissected into parts which are regulated

separately (although it is attempted quite often, much to the detriment of the

resource); for as the water goes, so the organisms, and vice-versa.

TABLE OF CONTENTS

Acknowledgements . . . . . . . . . . . . . ii

Preface .. . . . . . . . . . . . . . .iii

I. INTRODUCTION . . . . . . . . . .

A. The Concept of Buffer Zones . . . . .

B. Buffers for Wetlands, Wildlife, and Water Quality. .

C. Intent and Scope of the Report . . . . . .

II. REVIEW OF THE LITERATURE .. . . . . . .

A. Significance of Buffer Zones . . . . . . .

1. Ecological Significance of Transition Zones . .
2. Physical and Ecological Significance of Wetland For
3. Water Quality Benefits of Buffer Zones . . .
4. Water Quantity Benefits of Buffer Zones . . .
5. Edge Effects on Wildlife . . . . . . .
6. Regional Habitat Needs of Wildlife: Corridors .
7. Between-Habitat Needs of Wildlife . . . . .
8. Within-Habitat Needs of Wildlife . . . . .

B. Review of Buffer Zone Regulations . . . . .

C. Summary: Resource Buffer Zones . . . .. . .

III. LANDSCAPE ECOLOGY OF THE WEKIVA RIVER BASIN.. . . .

A. Landscape Perspective . . . . . . . .

1. Physical Description . . . . . . . .
2. Biological Description . . . . . . .

B. Wetland Communities .

. . . . .1

. . .1

.2

.5

.7

.7

7
est Edges.. 11
.... 14
. 23
S. . . 26
S . . 33
S. . 38
S . . 44

S . . 50

. . 59

S . . 67

. . 67

. . . 67
S . . 74
. . . 74

. . . . . . . . . .82

1. Aquatic/Marsh Communities. . . . . . . . . .82
2. Mixed Hardwood Swamp Community . . . . . . .. .. 84
3. Hydric Hammock Community .... . . . . . . . .86

Transitional Communities: Mesic Hammock and Scrubby Flatwoods. 87

Upland Communities . . . . . . . . . . . 87

1. Pine Flatwoods Communities. .. . . . . . ... 88
2. Wet Prairie Communities . . . . . . . . .. 90

3. Pine Sandhill Communities . . . . . . . . 90
4. Sand Pine Scrub Communities .. . . . . . . 92

IV. STATUTORY & DISTRICT CRITERIA RELATED TO BUFFER ZONES . . .. 95

A. Statutory Criteria, Sections 373.413 & 373.416, FS . . . .. 96

B. Rule Criteria, 40C-4, 40C-41 and 40C-42, FAC... . . . 98

V. DETERMINATION OF BUFFER ZONE REQUIREMENTS.. . . . .. . .11

A. Overview of the Methodology for Determination of Buffer Zones. .111

B. Determining Buffer Zone Requirements . . . . . ... .118

C. Illustration of Buffer Zone Determinations . . . . .. ... 128

LITERATURE CITED .. . . . . . . . . . . . . .145

APPENDIX A. WILDLIFE ASSOCIATED WITH WEKIVA RIVER BASIN . . .. ...169

I. INTRODUCTION

I.A. The Concept of Buffer Zones

The landscape is a mosaic of uplands and wetlands, developed lands and

natural lands, and forests and fields. Somehow, without complete knowledge

about how these pieces fit together, society must make decisions concerning how

best to assemble the puzzle. Some decisions are easier than others, especially

when there are no conflicting elements. Unlike a puzzle, however, the ease of

adding a new land use to the landscape is inversely related to the number of

pieces already on the board. As more and more land uses are added to the

landscape, the decisions concerning placement require more and more thought.

The more pieces there are on the board, the more conflicts there will be

between pieces. Most often, it is at the borders between pieces where

conflicts arise. Decisions are easier to make when adjoining uses are not

significantly different, i.e., when the existing land uses abutting the

boundaries of a decision piece are similar to the proposed use of the piece in

question. However, when adjoining uses are significantly different,

consideration must be given to the interrelationships that are being created

and to how one piece might affect surrounding pieces.

When uses are significantly different, or where the potential for conflict

is serious, it is common practice to create a buffer between them. Thus we

have buffers around airports, nuclear power plants, and bombing ranges. De-

militarized zones (DMZ's) between warring nations are often used to maintain

peace. Zoning in urban areas uses successively lower-density districts to make

1

2

transitions between high-density land uses and low-density residential

districts. Buffer zones can be relatively narrow, like the grassed shoulders

of highways which separate high-energy roads from forested landscapes, or quite

wide, like the DMZ between North and South Korea. Generally, as the density of

activity or the potential for conflict increases, the width of the buffer

necessary to contain the negative effects increases proportionally. In

addition, as the difference in activity level between bordering land uses

increases, the width of the border must increase, since it is not the absolute

magnitude of activity but rather the relative magnitudes of activity between

neighboring uses that require buffering.

I.B. Buffers for Wetlands, Wildlife, and Water Quality

The differences between developed lands and wild lands are significant;

the more intensely developed, the more the differences. Frequently, on

developed lands the native vegetation is removed and replaced with exotics,

drainage is improved, soil is compacted or covered with impervious materials,

wildlife habitat is replaced with human habitat, and activity levels increase

10 to 100 times those found in the wild. The gradient in the intensity of

noise, wastes, temperature, light, structure, and activity from undeveloped to

developed lands is remarkable. It is this gradient that affects the edge

between developed and wild lands, creating a new environment unlike the

original and not at all similar to the developed land.

The edge between wild lands and developed lands is characterized by

overflows of materials and energy that "flow" from high density to low. Water

runs off developed land, carrying sediments and nutrients. Noise from

developed land intrudes into the edge, disrupts natural activity, and

interferes with the less intense communications of wildlife that are cues for

3

territorial protection and breeding. Increased temperatures caused by removing

tree canopy and paving the ground surface have marked effects on the forest

microclimate.

Edges are common in the landscape, with or without development. In the

wild landscape, however, sharp edges are hard to find. The lake edge is not a

clean line; as the water level rises and falls throughout the year, it creates

a littoral zone that is neither open water nor dry land, but somewhere in

between. The sea edge is buffered by estuaries and coastal wetlands, which

are neither open ocean nor dry land. The forest edge is a transition from the

mature canopy with moist soils and minimal understory through a zone of more

open canopy, with drier soils and dense understory, to the prairie with its

xeric soils and lack of woody vegetation.

As the landscape is developed, the wild lands that surround cities

retreat. If the world was a homogeneous flat plain and development happened as

a simple outward progression from the central city, the edge would constantly

be moving and little need would be generated for buffer zones between developed

lands and undeveloped lands. However, the world is not homogeneous, but

heterogeneous, having many differences, many patches, and many obstructions to

development. The development process becomes patchy, areas are skipped,

wetlands left, and parks created. Areas are set aside as wildlife habitat, or

protected for water quality purposes, or left undeveloped as remnants to remind

us of what the landscape was once like. The patches that are created in this

manner undergo constant change as they "evolve" or succeed toward ecological

communities that are islands in the developed landscape.

In Florida, under continuing growth pressure, the wild landscape has been

mostly developed, leaving behind large and small islands of managed national,

state, and local parks, along with many wetlands. Wetlands were first left

4

behind because their development required a relatively greater commitment of

funds and time than they were worth. Where the demand was high, in coastal

areas for instance, wetlands were converted into dry lands and developed.

Development skirted the swamps and marshes of inland areas not as a result of

comprehensive planning, but because of piecemeal decisions made on a project-

by-project basis. As land became more valuable and wetlands were threatened,

state and local governments responded to increased pressure by enacting laws to

protect the special values of wetlands recognized by society. Of primary

concern was their function for maintaining clean, productive surface water and

their value as very productive wildlife habitat.

In the minds of many, wetlands are now "protected." The Warren S.

Henderson Wetlands Protection Act of 1983 established state permitting

authority over activities in wetlands by virtue of the fact that they were

defined as waters of the state. A methodology for determining state waters,

developed as a rule (Ch. 17-4 FAC) by the Florida Department of Environmental

Regulation, was required by the Henderson Act. The rule uses vegetation as a

means of determining landward extent of state waters. Plant species normally

found in wetlands are sorted into three broad categories: submerged species,

transitional species, and "invisible" species. All plants not found on these

plant lists are considered upland species. The rule then uses simple formulas

that relate dominance of these species to determine the upland edge of "state

waters." It is important to note that at no time does the rule or the original

legislation (except in the name of the Act) delineate "wetlands". It only

refers to a selective list of plant species. The intent of the rule and

legislation is to protect state waters, which include areas that are dominated

by certain plant species that indicate regular and periodic inundation.

Wetlands can, and do, extend landward of this jurisdictional line.

5

The area immediately adjacent to and upland of the jurisdictional line is

often a transition zone between wetlands and uplands. It is a zone that is

wetland at times and upland at times, exhibiting characteristics of each and

vegetated by species that are found in each. It is important to both the

wetland and the upland as a seed reservoir, as habitat for aquatic and wetland-

dependent wildlife species, as a refuge to wildlife species during high-water

events, and as a buffer to the extreme environmental conditions of sharp

vegetated edges.

To protect the values and functions of areas that are waterward of the

jurisdictional line, attention must be given to the area immediately upland of

the line. Wholesale alteration of this edge has immediate and potentially

large-scale permanent impacts on the waterward area. Wetland-dependent

wildlife species that are frequent users of this area are excluded, silt laden

surface waters are no longer filtered, the microclimate is greatly affected and

groundwaters may be diverted or drained. In the natural state the transition

zone is a buffer zone both for the adjacent wetland and state waters and for

the wildlife species that inhabit them. It should be recognized as such, and

afforded some degree of development control.

I.C. Intent and Scope of the Report

This report is intended to provide a detailed review of the existing

scientific understanding of upland buffer zones, their importance to the

adjacent wetlands and waters, and the effects of alterations of these

transitional areas on downstream water quality and quantity, and wetland

wildlife habitat values. As we researched the topic, we found a dearth of

information directly addressing buffer zones. However, there has been much

recent interest in the literature concerning the impacts of edges on wildlife

6

habitat values, and some recent investigations of the extent and species

composition of transitional zones between wetlands and uplands.

The habitat values of transitional areas are related to between and within

habitat needs of wildlife, the water quality values of buffer zones are related

to protection of surface and groundwater, and their value related to water

quantity issues is in their ability to mitigate adverse drainage impacts.

The report is organized in five parts with an appendix that gives lists of

aquatic and wetland-dependent wildlife species that are characteristic of the

Wekiva Basin. After the Introduction, the second section reviews the

literature and summarizes and relates it to the concept of a resource buffer

zone. The third section gives a brief physical and ecological description of

the Wekiva Basin. The fourth section reviews the Water Management District's

statutory criteria (Chapter 373 FS) and permit criteria in 40C-4, 40C-41, and

40C-42 FAC and relates them to protection of the water resources of the Basin.

The fifth section develops a methodology for determining buffer zone

requirements.

II. REVIEW OF THE LITERATURE

II.A. Significance of Buffer Zones

II.A.1. Ecological Significance of Transition Zones

The zone that lies between what are unconditionally identified as uplands

and wetlands is a zone of transition (sometimes called an ecotone). In this

zone, environmental conditions resemble neither the true wetland nor upland,

but fluctuate and may resemble each during different times of the year or from

one year to the next. Plant species that characterize transition zones may be

an assemblage of both wetland species and upland species and may contain

species that are not found in either of the two adjacent zones. It has long

been held that highest species diversity occurs in areas of habitat overlap

(MacArther and Pianka 1966, Allen 1962, Ranney 1977).

The Florida Department of Environmental Regulation (FDER), in its rules

concerning the delineation and definition of landward extent of waters of the

state, recognizes three main categories of plants: submerged, transitional, and

"invisible" species (Ch. 17-4.022 FAC). In an area dominated by transitional

species, a preponderance of evidence that indicates it is regularly inundated

is necessary before the area is considered a state water. In other words,

without regular inundation, an area dominated by transitional species is not

considered state waters, and by inference, therefore not a wetland.

Under 17-4.022 FAC, Florida does not recognize purely transitional areas

as state waters, and as a result does not recognize them as important or

necessary for protection under the Warren S. Henderson Wetlands Protection Act

7

8

(Chapter 430 Florida Statutes). However, their ecological significance as

wildlife habitat and seedbanks, as well as their role in maintaining the

integrity of the adjacent wetland community, is unrelated to their

classification under 17-4.022 FAC. The FDER has chosen a cut-off based on

dominance of "submerged" species to delineate state waters. The line so

determined in most cases classifies all vegetation to the landward side as

uplands. Most transition zones are to the landward side of the FDER

jurisdictional line.

The role of transition zones as wildlife habitat for wetland-dependent

species will be reviewed in subsequent sections, as well as their role as

seedbanks. Generally, their ecological value stems from (1) direct use by

wildlife and (2) seed sources for plant species that are more upland in growth

requirements, yet are found within wetlands and are an important component of

the community. Transition zones occupy the key upland fringe and as a result

are the main source for seed material that contributes to the spatial

heterogeneity of wetlands.

Wetland communities are not homogeneous. They are heterogeneous

communities whose spatial variation in species composition is probably

controlled by topography and ultimately by periods of inundation. Higher areas

sometimes referred to as hummocks range in size from a few meters to several

hectares. Hummocks have fewer periods of inundation and water depths are

shallower, resulting in conditions that favor more mesicc" species. Lower

areas are favored by those species that are to7':ant of flooding. This

heterogeneity of the wetland landscape is oce of the key factcrs that gives it

its diversity and wildlife value. Wildlife values are greatly enhanced as a

result of the increased diversity of vegetation and undulating topography.

9

Studies of the Florida landscape indicate that the plant species diversity

in transition zones is higher than the diversity of either the adjacent wetland

or upland. Clewell (1982) found five community types on the Alafia River in

central Florida along a gradient from wet to dry of which the "Moist Mesic

Forest" seems to occupy that area that is transitional. The number of plant

species found in this community exceeded all others by 25%. In Gross's (1987)

study of 12 small-stream floodplain ecosystems in north and central Florida, 12

vegetation types were distinguished, using cluster analysis of which the three

most diverse appeared to occur most often in the transitional areas. Studying

methods for determining transition zones, Hart (1984) documented species

composition of the wetland to upland continuum of eight wetland community types

of north central Florida. She found that the transition zones were more

diverse than either the wetland or upland and were different physiognrmically

from adjacent wetlands, yet were "....aligned more closely with wetlands than

with uplands and were composed of species that tended to alternate between

wetland and transition zones." The width of transition zones varied from a

minimum of 10 m to a maximum of 30 m depending on the method used to delineate

the zone and the type of wetland.

Wildlife species richness shows direct spatial relationships to the

increased diversity of transition zones. Studies of forest edges (see Edge

Effects on Wildlife) have documented increases in species richness and density

of individuals within ecotones between habitats. Vickers et al. (1985) found

that species richness and abundance of herpetofauna were greater along the

edges of six wetlands in north central Florida than in either the wetland or

upland habitat. Harris and Vickers (1984) found that virtually all mammals,

because of their cursorial mode of locomotion and frequently herbivorous food

habits, reside in "peripheral" areas, i.e., the transition zones. When water

10

levels were increased, the movement of wildlife to the peripheral areas was

also increased, suggesting the important role that transition zones play in

providing refuge from wet-season increases in water levels in wetlands. The

ecotone seems to play an even more important role when surrounded by clearcuts.

Harris and McKlveen (1982) found a greater abundance and diversity of breeding

birds in the ecotone between cypress wetlands and clearcut areas than in the

ecotone between cypress and pinelands.

Classifying the bottomland hardwood ecosystem into five classes, Wharton

et al. (1981) represented the transition to upland communities as the fifth

class. They listed the importance of environmental factors to fauna in each

zone. Of the five classes, the transitional zone ranked highest in importance

overall and highest in 18 of 27 factors, among which are:

1. Retardation of "side flooding,"
2. Detritus source,
3. Diversity of oaks,
4. Availability of large variety of flora and fauna.
5. Diversity of forest strata,
6. Refuge from high water, and
7. Forage and cover for upland species.

The importance of the highest zones (transitional zones) of bottomland

hardwoods to detrital food chains of downstream systems has been documented

(Livingston et al. 1976, White et al. 1979). These studies showed that the

infrequent pulses of organic matter and detritus form the higher zones in the

floodplain after extreme flood and rainfall events correspond to peaks in fish

production.

Topography plays an important role in controlling the width of transition

zones. The transition zone between wetland and upland may be only a few meters

in width or may extend for several hundred meters. In areas of very low

relief, where the landscape has little or no slope, the transition zone may be

extensive and ill-defined; yet, in landscapes where slopes are more prominent,

11

the transition from upland to wetland may be far more abrupt. In studies of

the central Florida landscape (Brown et al. 1984, 1985, and 1986), transition

zones as wide as 30 m were delineated using species composition as the

controlling variable. Gross (1987) found community types using cluster

analysis that were dominated by transition zone vegetation on central Florida

streams as wide as 80 m, but 40-m zones were more common.

II.A.2. Physical and Ecological Significance of Wetland Forest Edges

When forest is cleared and some forested remnants are left, as when the

landscape is developed for agriculture or urban uses or clearcut during

silviculture operations, the remaining forested patches contain induced edge

where there once was a forest continuum. These forested tracts develop

characteristic edge habitats that differ from habitat of interior forests.

Ranney (1977), in a review of the literature on forest island edges, suggested

that edge habitat differed from interior forests in at least six ways:

1. Tree species composition,
2. Primary production,
3. Structure,
4. Development,
5. Animal activity, and
6. Propagule dispersal capabilities.

The importance of edge is in its effects on the forested habitat and

eventually on the suitability of the habitat for indigenous fauna. The

conditions created when the forest continuum is cleared have detrimental long

term consequences on the remaining forest system. Using current jurisdictional

lines (Ch. 17-4.022 FAC) to determine the landward extent of state waters and

given current development practices, the forest edge created between wetland

and developed upland will coincide with the jurisdictionally determined line of

regularly inundated wetland. In most instances this does not include

transitional zones to the landward side of the jurisdictional line. The

12

creation of an edge at this wetland/"upland" boundary opens the wetland to the

many changes associated with edge conditions.

When a forested edge is created through clearing and/or partial removal of

the forest canopy, steep gradients of solar radiation, temperature, wind speed,

and moisture are incurred between the relative extremes of the open land and

the forest interior (Wales 1967, 1972). Studies of characteristics of newly

created edges (Gysel 1951, Swift and Knoeerr 1973, Trimble and Tryson 1966,

Wales 1972, 1976) suggest that solar radiation is probably the most important

physical parameter influencing conditions at the forest edge. The forest edge

environment is characterized by increased solar radiation which in turn

increases temperatures and decreases moisture content of soils and relative

humidity when combines with effects of increased wind flows. Physical and

structural manifestations of these changes include increase of shade-

intolerant, xeric tree and shrub species and an increase of species associated

with early stages of succession (Gysel 1951, Trimble and Tryson 1966, Wales

1972).

Of particular importance is the effect of increased wind speed on the

newly created forest edge; especially when the edge is created at the

wetland/upland interface. A major consequence of clearcutting is the loss of

buffering of destructive wind speeds by the adjacent tree canopy. Tree wind-

throws are common along newly created edges since edge trees not exposed to

wind velocities of the open landscape are not structurally prepared for them.

Wetland trees do not have deep-rooted growth habits, but instead rely on a root

mat very near the soil surface to provide structural support in the relatively

low bearing-capacity soils of most swamps. Windthrows are relatively more

common in the deep organic soils of wetland ecosystems than in forested upland

communities because of soil instability and difficulty of trees to anchor

sufficiently.

The spatial effects of forest edge conditions are variable and are highly

dependent on the way the edge is maintained. Ranney (1977) summarizes the

spatial influences of newly created edges as between 5 and 20 m deep from the

forest edge toward the interior. In a study of microclimate on Ft. George

Island near Jacksonville, Florida, Hart (1985) determined that most of the

microclimatic change between cleared areas and a forest interior in north and

south directions occurred within 10 m of the edge, although on western facing

edges, 25 m was often required for equivalent microclimatic change. Hart notes

that the eventual closure of newly created edges by shrubs and vines will

ameliorate negative microclimatic conditions and suggests the planting of fast-

growing shrub species in some instances to speed up edge closure. Harris

(1984) discusses the "three-tree-height" rule of thumb for the distance over

which climatic effects of a surrounding clearcut will penetrate into an old-

growth stand. Related to average tree heights of Florida wetland ecosystems,

the three-tree-heights rule would suggest the penetration of influences as deep

as 70 m.

Conceptually, to maintain the integrity of the wetland community and to

insure faunal habitat for wetland-dependent species, buffers from 5 to 70 m in

depth are necessary. Clearing at the wetland edge opens the wetland to

significant changes. However, with a buffer that encompasses the transitional

zone on the landward side of the wetland, negative impacts can be avoided.

Changes in abiotic parameters (due especially to increased solar radiation

and wind) towards more xeric conditions along edges lead to corresponding

changes in plant and animal communities (see Table 9 in Lovejoy et al. 1986)

and in system properties such as nutrient cycling (Ranney 1977). Studies by

14

Levenson (1981) and Ranney et al. (1981) in Wisconsin have documented major

edge effects on the vegetation of forest islands. They found that climatic

structural edge influences extend at least 10 to 15 m into a forest on the

east, north, and south sides and 30 m on the west side. Once established,

characteristic edge associations composed of xeric-adapted and shade-intolerant

species act as sources of propagules that invade the forest interior; forest

islands below a certain size (about 5 ha in Wisconsin) may be entirely edge

habitat in terms of their flora and vegetation structure (Ranney et al. 1981).

Such changes may be essentially permanent. In Ohio, Whitney and Runkle (1981)

found that small but persistent environmental differences associated with

forest edges have greater long-term effects on tree species composition and

structure than did severe but relatively brief disturbances associated with

logging.

II.A.3. Water Quality Benefits of Buffer Zones

The water quality benefits of buffer zones are related to the ability of

the zone to abate destructive water velocities and quantities of pollutants

carried by surface runoff from uplands that may have a negative impact on

downstream water quality, flora, and fauna. Soils of transitional areas are

generally characterized by a low accumulation of organic matter (Gross 1987,

Clewell 1982) but have accumulations of organic debris on the order of 1 to 5

tons per ha, depending on vegetative cover (Hewlett 1982). This organic debris

can absorb about 98% of rainfall energy, miiiinizing erosion potential. Known

as litter detention storage, it disperses the energy of water that would

otherwise break bonds of soil materials and produces a slurry of mud and eroded

soil. The kinetic energy of sediment transport increases rapidly as the second

power of water velocity (Hewlett 1982) suggesting that as a slope increases or

loses obstructions (such as vegetation) to uninterrupted flow, erosion

potential increases dramatically.

Occupying topographic lows in the landscape, wetlands are particularly

susceptible to erosional deposition from higher grounds and to erosional

scouring as a result of increased water velocities from mismanaged upland

surface waters. Maintaining undisturbed vegetation in transition zones can

help to minimize these destructive forces. The Florida Division of Forestry's

Silviculture Best Management Practices (BMP) Manual 1979) recommends a

Discretionary Zone (DZ) for land occurring within 300 ft of a watercourse.

This 300-foot-wide strip of land is considered the zone most influential to

surface water quality. Recommendations are given for varying site

sensitivities regarding the intensity level of activities such as construction

of roads, site preparation, and harvesting practices. Within the DZ, two

Streamside Management Zones (SMZ) are further delineated: a fixed primary and a

variable secondary zone, each of which has been assigned special management

criteria. The primary zone is 10 m from the edge of the watercourse,l while

the secondary zone is from 10 to about 50 m from the water's edge. The width

of the secondary zone is determined by the soil and topographic characteristics

of the site. Within each of the SMZ's, recommendations for site preparation

and harvesting are given relative to slope and soil erodibility.

Since the main objective of Silviculture BMP's is to minimize negative

impacts on water quality, and since silviculture, by definition, is the

1Sometimes the border of the SMZ's are referred to as "...the edge of the
watercourse or lake..." (Florida Division of Forestry 1979); other times the
SMZ is "adjacent to perennially open waters" (Riekerk and Winter 1982); still
others as a "...strip of forest land surrounding all perennial streams and
lakes 10 acres or larger" or as "a vegetated strip adjacent to a
watercourse..." or as occurring "...along all perennial and intermittent
streams..." (Florida Division of Forestry 1979). A strict definition of the
watercourse's edge as related to 17-4 FAC may be important to help determine
from what point the SMZ should be applied.

16

management of forests for their wood fiber, the BMP's are designed to maximize

the forest area that can be harvested. If the goal was to minimize negative

impacts on state water resources,2 the primary and secondary zones would begin

at the wetland edge and extend upland. In this manner the zones would occupy

the area that is commonly termed the transition zone between upland and wetland

and would insure minimum impacts to state waters.

It is generally agreed that the more sensitive a site is to sediment

production, the wider the undisturbed vegetative buffer should be. Since

sediment production can result in the deposition of materials in locations

where they may have negative impacts, and since sediments provide a major

vehicle by which pollutants are transported, every effort should be made to

encourage the use of vegetative buffers to minimize sediment deposition in

wetlands and water courses. The width and efficiency are variable. Karr and

Schlosser (1977), relying on the work of Trimble and Sartz (1957), suggest a

minimum width of buffers for conditions encountered in "municipal conditions"

as 15 to 20 m for lowest (0 to 3%) slopes and as high as 80 m for slopes of

60%.

There has been much research on the effects of logging in close proximity

to streams, especially in the northwest United States. Erman et al. (1977), in

studies of logged landscapes with and without streamside buffers, found that

buffers were very important in minimizing negative impacts as a result of

erosion from logged lands, and that streams with narrow (less than 30 m)

buffers showed effects comparable to streams with no buffers. Generally, their

2Waters of the state are those lands that are regularly and periodically
inundated as defined by 17-4.022 FAC which includes areas dominated by
"submerged" wetland vegetation. Thus to protect state waters and achieve Best
Management Practices as they relate to these waters, the discretionary zone,
primary zone, and secondary zone should be measured as distance from the edge
of the wetland instead of the edge of the watercourse.

17

research on 62 northern California streams showed a strong correlation between

buffer width and stream community diversity index. Streams with highest

diversity indices had buffers greater than 50 m in width.

Streamside biffers in the high-relief landscapes that characterize most of

the continental United States are generally measured from the channel bank.

Where landscape relief is high, channel slopes are high, and associated

floodplain ecosystems are narrow; and as a consequence, streamside buffers

measured in this way probably adequately protect the resource. However, where

landscape relief is low, as in the Southern Coastal Plain, terrain is flatter,

stream channel slopes are lower, and associated riparian ecosystems are wider.

The measurement of streamside buffers in low relief landscapes needs to reflect

these important differences and begin at the upland edge of the riparian

ecosystem.

Effects of Construction Activities

Darnell et at. (1976) summarized the effects of construction activities on

riparian ecosystems and grouped them into three time related categories:

1. Direct and immediate results which take place during the
construction process;

2. Effects which occur during the period of stabilization
following completion of the construction; and

3. Long-term effects of more or less permanent changes
brought about by the construction itself or by
subsequent human use and environmental management occasioned
by the constructed facilities.

The water quality effects of construction activities within upland

boundaries of riparian ecosystems can be grouped into the same time-related

categories, although for the sake of simplicity, the first two may be combined.

The overall impacts of any construction activity vary depending on

environmental features (e.g., physical characteristics such as slope, soils,

18

and vegetative cover), timing, type of construction activity, and the care

taken during the active construction phases. Long-term impacts associated with

subsequent uses vary depending on environmental features, intensity of uses,

and management practices.

During and immediately following construction the major impacts on water

quality can be grouped into three broad classes:

1. Impacts associated with erosion of loose soils and their
subsequent deposition in downslope wetlands;

2, Suspended sediment increases in surface waters,
resulting in increased turbidity; and

3. Introduction of unusual levels of chemical compounds
that may have negative effects on resident fish and
wildlife populations.

The most obvious results of the deposition of eroded sediments in wetlands

are the impacts associated with filling. When wetland soils are covered with

additional depths of fill material, the most serious impact is loss of oxygen

penetration to the root zone and resulting stress on vegetation. If the accu-

mulation of sediment is deep enough, or if the sediment is fine enough to plug

up soil pores so as to effectively starve roots of oxygen, death of vegetation

results. Secondary impacts are experienced by wildlife as a result of the loss

of habitat.

The authors have on numerous occasions witnessed the results of inad-

vertent filling along wetland edges throughout Florida. In most cases these

results were caused by erosion of unstable soils during clearing and con-

struction activities. The deposition of eroded material resembles a wedge of

sediments where the bigger particles drop out of the sediment stream first and

successively finer materials drop out farther into the wetland system. Depths

of sediment are greatest along the periphery and taper off to shallower and

shallower accumulations as the sediment stream penetrates into the wetland. In

19

most instances the effects are immediate and dramatic. Where sediment depths

are greatest, both understory and canopy plants are killed, and as the sediment

wedge tapers off, tree kills are less noticeable, but the many varieties of

understory vegetation with their shallower root systems are killed. Finally,

at the farthest extreme of the sediment wedge, only the herbaceous vegetation

is affected.

The net result of the deposition of eroded sediments within the limits of

wetland communities is to shift the conditions of the transitional edge farther

into the wetland. Dead and dying vegetation is replaced over time by early

successional upland vegetation. The impacts are less a loss of water storage

capacity (since the volume of sediments may be small compared to the overall

volume of water stored in the wetland) as they are a loss of valuable wetland

function and wildlife habitat.

If the disturbed area is close enough to surface waters, or if the

configuration of the wetland is such that stormwater runoff from the upland

edge can collect and form an intermittent stream, sediments eroded during

construction phases may directly affect aquatic environments. In their review

of impacts of construction activities in wetlands, Darnell et al. (1976)

grouped the biological impacts of suspended and sedimented solids under the

topics of turbidity, suspended solids, and sedimentation. Within each group

they review the biological effects that have been documented in the literature.

Their review is summarized in Table II.A.I.

The final water quality effect during the construction phase is related to

the release of chemicals, the levels of which may be harmful to downstream fish

aand wildlife or negatively affect ecosystem function. When areas are cleared,

runoff increases, carrying with it increased volumes of soil and sediment.

Large quantities of dissolved cations and other minerals are lost from the

Table II.A.1 Biological Effects of Sediments and Suspended Solids on Aquatic
Environments (after Darnell et al. 1976)

Biological Effect

Turbidity

Suspended Solids

Sedimentation

* Reduction of photosynthesis eliminates
phytoplankton, attached algae, and rooted
vegetation, thus eliminating base of aquatic food
chain
* Decreased visibility interferes with normal
behavior patterns of higher aquatic organisms

* Temperature effects water absorbs more radiant energy,
inhibiting vertical mixing in calm waters or uniformly
heating moving waters modifying oxygen content
* Oxygen reduction through inhibition of photosynthesis,
heating, or increased COD and BOD
* Reduction of primary production inhibits light
penetration, absorption of critical nutrients, removal of
algae and zooplankton from suspension through adherence to
particles
* Effects on respiration oxygen reduction can selectively
eliminate aquatic animals that require oxygen for
respiration, suffocation through clogging of gill filaments
* Other effects potential starvation of filter feeding
animals, decreased visibility, interference with migration
patterns of fish

* Primary production smothering and scouring, physical
barrier to gas exchange
* Bottom animals scouring and blanketing with sediments
* Fish populations reduction of food supply, destruction
of habitat, elimination of spawning areas, smothering of
eggs and larvae

Topic

cleared areas and deposited in downstream areas. Borman et al. (1968) found

cation losses from 3 to 20 times greater from clearcut watersheds than from

undisturbed control watersheds. Riekerk et al. (1980) showed significant

increases in sediment and higher nutrient cation loading of runoff from a

clearcut, relatively low-gradient watershed after one year, but no significant

differences after the second year. In watersheds composed of wet savanna

forests of Florida's Lower Coastal Plain, Hollis et al. (1978) recorded

significant loss rates of dissolved minerals including ammonium-nitrogen as

well as phosphorus and suspended sediments.

Of less concern, unless a large-scale accident occurs, is the increase in

petroleum products that may pass from construction sites into downstream

wetlands and watercourses. Spills in maintenance yards and the general

operation of equipment in and near wetlands increases the likelihood of

increased contamination. In all, however, contamination from petroleum

products during the construction phase should be of minor concern if proper

care is taken.

Effects of Operation and Maintenance

Long-term water quality impacts associated with construction activities in

areas adjacent to wetlands and watercourses is a function of subsequent human

uses and management. Stormwaters running off developed lands carry both

organic and inorganic materials, both suspended and in solution. Many of the

organic are degraded through biological pathways or through chemical

oxidation. As degradation occurs, oxygen is consumed causing decreases in

available oxygen. Thus the breakdown of these compounds "demands" oxygen.

Their presence and potential for negative water quality impacts is measured by

tests for Biological Oxygen Demand (BOD) and Chemical Oxygen Demand (COD).

The final land use of a developed area has a marked effect on the quality

of stormwater runoff. The Florida Department of Environmental Regulation

(FDER) (1978) summarized the work of Sartor and Boyd (1972) that compared the

total solids loading from residential, commercial, and industrial land uses.

Industrial areas were shown to have the highest loadings and commercial areas

to have the lowest. Another stormwater study of three sub-basins (a forested,

a suburban, and an urban basin) of Lake Jackson in northern Florida as reported

in Wanielista (1975) and summarized by the FDER (1978) compared mean

concentrations for various constituents in stream water under stormflow and

baseflow conditions. The findings showed that the export from the urban

watershed was greater for all constituents than from the forested or suburban

watersheds, and that mean concentrations of suspended solid loadings from the

urban basins and suburban basins were approximately 10 times and 5 times,

respectively, those of the forested watershed.

In most time-related studies of the quality of urban runoff, a common

finding is the difference in concentrations of constituents between the initial

"first flush" and later runoff. Typically, the initial rainfall flushes the

stormwater system, carrying with it large quantities of suspended solids and

materials in solution. Later discharges have significant, but much lower,

concentrations of pollutants.

Stormwater management rules generally recognize the potential for

downstream water quality impacts from urban and suburban areas, and while there

is some variation in regulation throughout the state, generally they address

the problem through retention of a prescribed amount of the initial rainfall

from a storm event. This treatment method assumes that through retention of

the first flush, a majority of the constituents that would enter surface waters

are retained. However, most stormwater systems, while retaining this first

flush, accumulate the remaining stormwater in the same retention basin,

diluting initial concentrations with the added waters. Under these conditions,

if the rainfall event is large enough, the diluted first flush is discharged

along with remaining stormwaters to receiving water bodies. The net result is

little if any retention of the pollutant carried by the storm runoff.

Recently, more stringent regulations have been put into effect in the Wekiva

River Basin that require off-line retention to avoid this problem.

The problems associated with degradation of water quality as a result of

the operation and maintenance of a stormwater management system once a

development is in place may be little affected by a buffer requirement. The

potential impacts on water quality from the combined effects of increased

runoff and increased concentrations of pollutants that result from urbanization

are not so much functions of how close the developed areas are to wetlands and

watercourses as they are of land use, size of the development, and stormwater

system design. However, potential impacts from systems designed to allow

portions of a development to discharge without treatment may be mitigated

through a buffer requirement. In many instances, where slopes prohibit

stonrwaters to be collected into a system, portions of the lowest areas in the

development (typically those closest to and abutting the wetland/watercourse

edge) may discharge directly. A buffer requirement in these situations would

allow for some treatment of stormwaters if they were allowed to sheetflow

through the buffer area.

II.A.4. Water Quantity Benefits of Buffer Zones

The construction and subsequent operation of urban land uses within areas

adjacent to wetlands and watercourses affect the quantity and timing of waters

that flow over the lands and often surficial groundwaters beneath the developed

24

lands. The effects are the result of both construction-related activities and

operation and management-related activities. During construction, the volume

of runoff increases, since removed vegetation and surface accumulations of

organic matter can no longer act to retard runoff, and transpiration is greatly

reduced,

Studies of water yield from watersheds having varying degrees of clearing

show strong relationships between the amount of disturbance and water yield.

One study in north central Florida by Riekerk et al. (1980) found water yield

increases of 165% from a clearcut, highly disturbed watershed and over 65% from

a minimally disturbed watershed. In a later study Riekerk (1985) found 2.5-

fold and 4.2-fold increases in runoff from minimally and maximally disturbed

watersheds, respectively. In most studies, the effects of disturbance on

runoff volumes decreases markedly from 2 to 3 years after treatment as

vegetation again plays a role in the site's hydrologic balance. The net

effects of clearing are reflected in higher soil moisture levels which may

contribute to higher stream baseflow and greater direct runoff during storm

events.

Once vegetation has been removed, it is quite common for groundwater table

elevations to increase as a result of decreased transpiration. In areas of

existing high water tables, soils can become saturated and extremely difficult

to work without extensive "drainage improvements." Stormwaters are often

routed through temporary swales and ditches into downstream areas to minimize

flooding while drainage and stormwater systems are being constructed. Such

practices can add significantly to construction-related water quality impacts

and alter quantity and timing of surface water discharge.

Typically, areas immediately adjacent to floodplain and seepage wetlands

are dominated by seasonally high water tables. This is a predominate aspect of

the lands adjacent to the floodplain wetlands throughout the Wekiva River

Basin. Lowering of water tables to accommodate construction-related activities

and as a permanent consequence of development can reduce groundwater elevations

and intercept groundwater flows to adjacent wetlands. Of all the impacts of

development, when the mitigating effects on surface water of properly designed

retention/detention systems are taken into account, the loss and interception

of groundwaters has the most profound consequences. Drainage in areas with

high water tables is a widely accepted practice that allows development of

lands that would otherwise be difficult to develop. Positive drainage

"dewaters" the upper portions of the soil, which lowers the surrounding

groundwater elevations. The effects spread radially from the drainage system

with the actual amount of decline in water table elevation dependent on the

elevation of the drainage system and soil properties and topography of the

surrounding lands. Wang (1978) and Wang and Overman (1981) found the effects

of a 10-foot lowering of the groundwater table to extend up to 1 mile from

drainage ditches in south Florida. While this is probably not typical

throughout Florida, it does indicate the magnitude of the problem.

The impact of lowered water tables on adjacent wetlands is a reduction in

hydrologic function. Where adjacent wetlands are depressional wetlands that

intercept the groundwater table, lower water table elevations reduce depths of

surface waters and shorten hydroperiods. Under extreme conditions the wetland

may remain dry throughout the year. In areas dominated by seepage wetlands,

lowered water table elevations result in a reduction or even a cessation of

groundwater seepage and subsequent drying of the wetland. The long-term, wide-

scale effects of general drainage practices throughout the state have been

discussed by Brown (1986) and are pervasive. While immediate ecological

changes resulting from drainage are subtle--to the untrained eye, no change is

26

observable--the long-term consequences are loss of hydrological functions,

gradual replacement of wetland vegetation with upland types, and consequent

loss of habitat values for aquatic and wetland-dependent wildlife species.

The maintenance of setbacks or buffers in areas immediately adjacent to

wetlands and watercourses can have significant impact in minimizing these

effects. A properly sized buffer between construction activities and

downstream wetlands can slow down and filter runoff and mitigate the drawdown

effects of drainage ditches on adjacent wetlands. It cannot be stated strongly

enough that the lowering of water tables to accommodate development is probably

the single most important impact affecting adjacent wetlands, and that a

properly sized buffer will go a long way toward minimizing these impacts.

II.A.5. Edge Effects on Wildlife

The question of how wide a buffer zone must be to maintain biological

integrity cannot be answered without considering the problem of edge effects.

Edge is defined as the place where plant communities meet or where

successional stages or vegetative conditions within plant communities come

together (Thomas et al. 1979). Edge effects have been investigated intensively

in wildlife management ever since 1933, when Aldo Leopold's classic text was

published. Evidence of a positive edge effect (i.e., an increase in species

richness or density of individuals near an edge in contrast to adjoining

habitat types) has been reported for birds (Lay 1938, Beecher 1942, Galli et

al. 1976, Laudenslaver and Balda 1976, McElveen 1977, Gates and Gysel 1978),

mammals (Bider 1968, Forsyth and Smith 1973), and several other groups of

organisms.

Most of the early studies of edge effects were concerned with documenting

an increase in wildlife abundance near edges. They implied that managing land

to increase edge will also increase wildlife. Edge habitat has high cover

density and food availability for many animals, and it is an index of carrying

capacity for species that require two or more habitat types for survival

(Johnson et al. 1979, Thomas et al. 1979). Hence, Leopold (1933) described

game (wildlife) as "a phenomenon of the edge" because wildlife "occurs where

the types of food and cover which it needs come together." Subsequently, edge

became a fundamental principle of wildlife management (e.g., Allen 1962,

Dasmann 1964), and some wildlife management texts went so far as to urge

managers to "develop as much edge as possible" (Yoakum and Dasmann 1971).

Ghiselin (1977) suggested that an edge index based on the amount of edge or

ecotone present is useful as an indicator of relative animal productivity

between similar habitats.

.Unfortunately, these wildlife management recommendations were naive about

the consequences of edge management on nongame animals, on plant communities,

and on regional diversity patterns (reviewed in Noss 1983). In a management

context, there are two types of edge: inherent edge and induced edge (Thomas

et al. 1979). Inherent edge is a natural phenomenon that represents the

juxtaposition of community types and a continuum of successional stages in the

landscape mosaic. This natural heterogeneity is in large part a product of the

natural disturbance regime of fire, flood, treefalls, windthrows, landslides,

and other events (White 1979, Sousa 1984, Pickett and White 1985). Moreover,

this natural heterogeneity may be the primary determinant of wildlife

composition and diversity in a region (reviewed in Noss 1987).

Induced edge, on the other hand, is a product of manipulation of habitat

by humans. It is closely tied to the process of habitat fragmentation that

produces sharp contrasts between vegetation types. Simple geometry shows that

small habitat islands have larger edge-to-interior ratios than do similar

28

shaped large habitat islands. As a natural landscape is fragmented into

disjunct pieces of habitat in a sea of developed land, the proportion of edge

habitat in the landscape increases. Thus, species that are restricted to

habitat interiors gradually are replaced by species characteristic of edge

habitats. The latter species are generally opportunistic or "weedy" and do not

need reserves for survival (Diamond 1976, Whitcomb et al. 1976, 1981, Noss

1983). Conservation biologists therefore recognize that interior species

dependent on large tracts of undisturbed habitat should receive priority

attention in conservation plans.

The recognition of negative edge effects associated with induced edge was

a major breakthrough in wildlife ecology. Until the mid-1970s, the attention

of most wildlife scientists was focused on relatively small tracts of land

(i.e., on local diversity) and on the small subset of animal species that could

be hunted. Under these circumstances, induced edge habitat can be seen as

favorable. Species such as white-tailed deer, bobwhite quail, cottontail

rabbit, and ring-necked pheasant often increase under such circumstances, as

does the species list of individual management units. But if one focuses on

regional and global diversity, and on entire ecosystems (including nongame

animals and endangered species of plants), the negative consequences of induced

edge are all too clear. The end result is often a landscape that has lost its

native character and is dominated by weeds that are either alien to the region

or were previously restricted to recently-disturbed patches (Faaborg 1980,

Samson 1980, Noss 1981, 1983, Samson and Knopf 1982, Harris 1984, Noss and

Harris 1986).

The negative effects of induced edge are easier to understand when one

considers that edge is more than a one-dimensional boundary between habitat

types. Effects of an open, disturbed habitat penetrate some distance into an

29

adjoining wooded natural area; hence, edge has a width of influence. Much of

this influence is ultimately climatic. Lovejoy et al. (1986), summarizing the

status of their research on forest fragmentation and induced edge effects in

Amazonia, recognize an abiotic class of edge-related changes that includes

temperature, relative humidity, penetration of light, and exposure to wind.

The type of habitat on the outside of a forest edge will determine the

nature of edge effects. A general principle is that the greater the contrast

between habitat types, the greater the edge effect (Harris 1984). When a

forest is fragmented, the matrix which surrounds it is usually some type of

early successional habitat. (Housing subdivisions would generally fall in this

category.) This secondary successional habitat is a constant source of weedy

plant and animal propagules that invade the forest fragment and alter species

composition and relative abundances (Janzen 1983). Opportunistic animals that

achieve artificially high densities in the early successional habitat often

invade pristine areas, exerting abnormally high levels of trampling, browsing,

and seed predation; this can destroy a natural area as surely as can a chainsaw

(Janzen 1983, 1986). Bratton and White (1980) reported that manipulation of

habitat to support a huntable deer herd can result in heavy browsing in

adjacent natural areas and further endanger a number of rare plant species.

The effects of habitat fragmentation and induced edge on animal

communities (especially birds) has been particularly well studied over the last

two decades. Application of island biogeography theory (MacArthur and Wilson

1967) to habitat islands have led to recommendations for the design of nature

reserves. Nature reserves and forest fragments have been portrayed as islands

because they are patches of natural habitat in a matrix or sea of culturally

modified land (Terborgh 1974, Diamond 1975, Sullivan and Shaffer 1975, Wilson

and Willis 1975, Diamond and May 1976, Forman et al. 1976, Galli et al. 1976,

30

and many others). Most early work on this problem was essentially a

confirmation of the familiar species-area relationship (Arrhenius 1921, Gleason

1922, Preston 1960, 1962): larger pieces of habitat support more species.

There are many potential explanations for the species-area relationship. Three

general explanations are (1) larger areas support more kinds of habitats (and

thus more habitat-specific species), (2) larger areas offer bigger "targets"

for organisms dispersing across the landscape, (3) larger areas maintain larger

populations that are less vulnerable to extinction due to random or

deterministic population fluctuations, and (4) larger areas support animals of

large territory and home range size that cannot be supported in small areas.

Any one of these explanations is powerful enough to support the general

recommendation that nature reserves should be as large as possible (e.g., Soule

and Wilcox 1980, Frankel and Soule 1981, Schonewald-Cox et al. 1983, Harris

1984, Soule 1986).

The process of habitat fragmentation is accompanied by insularization of

fragments, i.e., isolated pieces of habitat surrounded by dissimilar habitat

begin to resemble islands in many of their ecological dynamics. Eventually,

fewer native species will be found in a habitat islands than in sample areas of

equal size within extensive blocks of habitat (Miller and Harris 1977).

Alternately, species richness may not change much, or may even increase, with

habitat insularization, but species composition will shift towards edge species

at the expense of area-dependent interior species. These edge species are

generally common in the developed landscape and do not need reserves of any

kind for survival (reviewed in Noss 1983).

In New Jersey, Forman et al. (1976) found that forest islands contained

more bird species than did sample plots of equal size within extensive forests,

but the additional species were primarily birds of the forest edge. Forest

interior birds were limited to the larger forest islands. Since then, the

changes in avifauna that occur with forest fragmentation have been th.oougihly

documented (for the most extensive treatment, see Whitcomb et al. 1981). In

Florida hardwood hammocks, Harris and Wallace (1984) documented area-dependence

in a number of bird species. In the Wekiva River bottomland hardwood foresc,

breeding bird species that have been shown elsewhere to be area-dependent and

vulnerable to negative edge effects include the red-eyed vireo and the Acadian

flycatcher, both near the limits of their ranges in this area (H. Kale, Florida

Audubon Society, pers. comm.).

More generally, the species most vulnerable to extinction in fragmented

landscapes are large animals with large home ranges (i.e., top carnivores),

ecological specialists, and species with variable populations that depend on

patchy or unpredictable resources (Terborgh and Winter 1980, Karr 1982, Wright

and Hubbell 1983). The characteristically small populations of these species

are vulnerable to problems related to environmental stochasticity, demographic

stochasticity, social dysfunction, and genetic deterioration brought on by

inbreeding or genetic drift (Frankel and Soule 1981, Shaffer 1981, Schonewald-

Cox et al. 1983, Soule and Simberloff 1986).

Biotic aspects of edge effects have played a major role in the ecological

deterioration that accompanies fragmentation. Some effects of edge on

vegetation were discussed above. Because animal communities respond to

vegetation structure (e.g., MacArthur and MacArthur 1961, Roth 1976), any edge-

related changes in vegetation will cause corresponding changes in animal

communities for a certain distance into a habitat interior. But there are more

insidious processes at work. The opportunistic animals that are attracted to

edge (or to the early successional habitat outside) often prey on, out compete

32

or parasitize interior species, sometimes with disastrous consequences, to the

latter's populations.

Most of the research on faunal deterioration near edges has been done on

birds. Whitcomb et al. (1976) provided evidence that, in areas where forest

has been reduced to isolated fragments, avian brood parasites (brown-headed

cowbirds), nest predators (small mammals, grackles, jays, and crows), and non-

native nest-hole competitors (e.g., starlings) are usually abundant in the

urban-agricultural matrix and often invade small forest tracts and narrow

riparian strips. Gates and Gysel (1978) found that a field-forest edge

attracts a variety of open-nesting birds, but such an edge functions as an

"ecological trap." Birds nesting near the edge had smaller clutches and were

subject to higher rates of predation and cowbird parasitism than those nesting

in either adjoining habitat interior. Brown-headed cowbirds, which were not

found east of the Mississippi River before widespread forest fragmentation, are

a particularly noxious species. Half or more of the songbird nests within 200

m of the edge may be parasitized, reducing reproductive success for some

species below the level of population sustainability (Brittingham and Temple

1983). Brown-headed cowbirds have been increasing in number and are moving

south in Florida, and they will undoubtedly become more problematic as forests

continue to be fragmented; meanwhile, the ecologically similar shiny cowbird,

which has caused severe problems in parts of the Caribbean (Post and Wiley

1976, Post and Wiley 1977), has been moving north and was reported in Florida

in 1986.

Experimental studies of nest predation have documented significantly

higher predation rates in small forests compared to large forests, in forests

surrounded by suburbs compared to forests surrounded by agricultural land, and

at decreasing distances to forest edge (Wilcove 1985, Wilcove et al. 1986).

This abnormally high predation rate is related to the artificially high

densities of many opportunistic animals near forest edges and in disturbed

habitats (including suburbs; Wilcove et al. 1986). A primary reason why these

opportunistic animals (which in Florida include raccoons, opossums, gray

squirrels, armadillos, house cats, and blue jays) achieve such high densities

is that the large predators (e.g., panthers, wolves, and bears) that once

regulated their populations have been extirpated or greatly reduced (Matthiae

and Stearns 1981, Whitcomb et al. 1981, Wilcove et al. 1986). The deleterious

effects of increased nest predation may extend 300 to 600 m inside a forest

border (Wilcove et al. 1986). More generally, Janzen (1986) suggests that

managers can expect serious edge effects (including problems related to heavy

use by humans and domestic animals) anywhere within 5 km of a reserve boundary.

The studies discussed above suggest that any forest tract has a "core

area" that is relatively immune to deleterious edge effects and is always far

smaller than the total area of the forest (Temple 1986). Relatively round

forest tracts with small edge-to-interior ratios would thus be more secure,

whereas thin, elongated forests (such as those along riparian strips) may have

very little or no core area and would be highly vulnerable to negative edge

effects.

II.A.6. Regional Habitat Needs of Wildlife: Corridors

A great deal of recent literature in the fields of island biogeography and

conservation biology has discussed the effects of inbreeding and genetic drift

on wildlife due to genetic isolation and small population sizes (e.g., Miller

1979, Soule 1980, Senner 1980, Wilcox 1980, and Franklin 1980). Inbreeding has

the effect of decreasing population heterozygosity (genetic variation) by

increasing the probability that progeny will receive duplicate alleles from a

34

common ancestor. This loss of genetic variation can have both immediate and

future implications for a species' survival. Inbreeding can lower species

vigor and fecundity within a few generations (Soule 1980). The very reduced

population of Florida panthers may be suffering from the effects of inbreeding.

All five males examined have had greater than 93% abnormal sperm (Roelke 1986).

Over the long term, inbreeding can also limit the ability of a population to

evolve to meet changing environmental conditions (Soule 1980, Harris et al.

1984).

Other literature has questioned the effectiveness of fragmented parks and

preserves in maintaining viable populations of animals which require large home

ranges or activity areas (Pickett and Thompson 1978, Lovejoy and Oren 1981,

Harris and Noss 1985, Harris 1984, Noss and Harris 1986, Noss [in press]). In

Florida, bears may range over 15,000 ac and bobcats over 5,000 ac. An otter

may require several miles of linear river and riparian habitat (Harris 1985).

To maintain viable populations of these and other far-ranging animals, large

blocks of land are needed.

One proposed management alternative for providing for these wildlife needs

is the use of wildlife corridors (Diamond 1975, Butcher et al. 1981, Forman and

Godron 1981, Harris 1983, Noss 1983, Harris 1984, Harris and Noss 1985, Noss

and Harris 1986, Wilcove and May 1986, Noss [in press]). Wildlife corridors

can be defined as strips or parcels of land which allow safe passage of

wildlife between larger blocks of habitat. This contiguity effectively

increases the size of protected lands and their ability to maintain viable

wildlife populations. Genetic variation is maintained because genetic material

is carried freely back and forth across the corridor and among large habitat

blocks by dispersing wildlife. Dispersing animals can recolonize areas which

have suffered from local extinctions (Fahrig and 1Merriam 1985). Large

carnivores such as panthers, bears, bobcats, and otters which require large

home range sizes would be free to continue their natural movement between

habitat blocks which would otherwise be rendered inaccessible to them by

physical barriers. They gain the increased resource base needed to support

top-level carnivores. In addition to serving as travel routes, corridors also

serve as habitat for some species.

Evidence of wildlife use of travel corridors comes from a variety of

sources at several scales. In the extreme, land bridges between continents

have historically acted as wildlife corridors (Simpson 1940, Simpson 1965).

The Bearing Sea land bridge allowed the migration of many animals, including

man, into North America from Siberia. The Isthmus of Panama similarly allowed

the migration of wildlife between North and South America. Other

paleontological research has pointed to the development of a broad, flat

corridor which formed along the Gulf coast of Florida during the glacial

lowering of the sea level. Webb and Wilkins (1984) found that during these

periods Florida's wildlife was highly influenced by fauna which had migrated

into Florida from South America and southwestern North America. Riverine

gallery forests in Brazil apparently act as mesic corridors which have allowed

Amazonian forest species to invade and become components of the xeric cerrado

fauna (Redford and da Fonseca, 1986).

On a smaller scale, habitat corridors have also been shown to be important

for wildlife. Riverine forest corridors are known to be important habitats for

many species. The maintenance of streamside strips of vegetation is an

important management tool for maintaining gray squirrel populations within pine

plantations (McElfresh et al. 1980). Turkeys have been successfully managed in

a mosaic of poor habitat (short-rotation pine plantations) by maintaining

hardwood and mature pine trees in travel corridors. These corridors allowed

36

the turkeys to move widely among foraging and roosting areas (Gehrken 1975).

Several species of birds have been found to regularly use fencerows and

hedgerows as safe travel routes (Bull et al. 1976). MacClintock et al. (1977)

found that a small forest fragment connected by a corridor to an extensive

forest block was characterized by birds typical of the forest interior, while

similar but isolated forest fragments were not. Johnson and Adkisson (1985)

noted that blue jays used vegetated fencerows as travel routes apparently

because they afforded some escape cover from hawks. Wegner and Merriam (1977)

found that small mammals and many birds travel more frequently along fence

hedgerow corridors than across open fields. Wildlife populations in isolated

blocks of forest have been shown to have lower growth rates than populations in

forest blocks tied together by corridors (Fahrig and Merriam 1985).

The relevance of wildlife corridor management to the Wekiva River Basin

lies in the presence of a large number of parks and preserves in the area.

Many wildlife species in Rock Springs Run State Reserve, Wekiva Springs State

Park, Lower Wekiva River State Reserve, and Kelly Park will be adversely

affected if an adequate corridor is not preserved. While several types of

corridors have been defined (Forman and Godron 1981), two are perhaps most

important for a discussion of the Wekiva River Basin. "Line" corridors are

narrow and are entirely edge habitat, while "strip" corridors.are wide enough

to maintain interior conditions. This distinction between edge and interior

habitat is important (see previous discussion on edge effects) as some interior

species cannot live or even migrate through extensive edge habitats (Forman and

Godron 1981). Edge effects are a function of corridor width. Beyond the

deleterious effects of edge, width also has other effects on corridor function.

A wide corridor may provide actual habitat for an animal while a narrower one

may simply provide a travel route. A study of cypress domes showed that

certain species of birds were excluded from smaller cypress domes (McElveen

1978). Stauffer (1978) found that bird species richness increased

significantly with the width of wooded riparian habitat, where some species

were restricted to wider strips. Tassone (1981) reported similar results from

a study of hardwood leave strips. Interior forest species such as Acadian

flycatchers were only infrequently found in corridors less than 50 m. Hairy

and pileated woodpeckers required minimum strip widths of 50 to 60 m, while the

parula warbler required at least 80 m. He suggested that leave strips must be

a minimum of 100 m on larger streams to take advantage of their intrinsic

wildlife value. Corridors which provide habitat should be much more effective

in connecting larger habitat blocks than those which only provide paths for

travel. Forman (1983) has stated that width is the most important variable

affecting corridor function.

An upland buffer along the Wekiva River floodplain would maintain the

width of the corridor, allowing it to function more efficiently as a wildlife

corridor. This would be particularly true where the river floodplain is

narrow. Narrow sections may act as barriers through which some animals are

reluctant to pass. An upland buffer would allow continued movement and provide

a refuge for some terrestrial wetland species during periods of high water. It

would also encourage use of the corridor by upland species which do not make

regular use of wetland habitats (Forman 1983).

In addition to providing dispersal pathways for animal wildlife,

floodplain corridors are also important in plant dispersal. Flood waters

collect and disperse large numbers of seeds from both wetland and upland plants

(Wolfe 1987, Schneider and Sharitz [in press]). This dispersal of upland seeds

may be very important in maintaining the diversity of upland plant species on

topographic highs or islands within the floodplain (Schneider and Sharitz [in

38

press]). Wharton et al. (1982) noted that these upland areas within the

floodplain are very important for wildlife habitat. Upland seeds carried onto

the floodplain originate in the wetland-upland ecotone and from smaller

tributaries feeding the river. Protection of this ecotone, which may extend 10

to 35 m in Florida (Hart 1981), and the adjacent upland would ensure continued

input of seeds to the floodplain.

In summary, width is probably the most important variable affecting

corridor function (Forman 1983). Several forest interior species are known to

be excluded when corridor width falls below a critical level, which is a

function of edge effects and home range requirements. Increasing the width of

the Wekiva River corridor by maintaining an upland buffer can only increase its

effectiveness in providing both a travel route for dispersing wildlife and

high-quality habitat.

II.A.7. Between-habitat Needs of Wildlife

Between-habitat needs refer to wildlife utilization of more than one

habitat type to fulfill their requirements. Several authors have substantiated

the close association and interaction of wildlife in wetland and adjacent

upland communities. Fredrickson (1978) reported that various species more

commonly associated with either wetlands or uplands depend on seasonal or daily

shifts into different habitat types to escape flooding, to forage, to disperse,

or to hibernate. Examples that he cited are turkey, river otter, swamp rabbit,

deer, bobcat, and gray fox. Other species such as raccoon, gray squirrel, tree

frogs, and many woodland bird species occur with similar frequency in both

wetlands and uplands.

Bottomland hardwoods are integrally coupled to the surrounding uplands

(Wharton et al. 1982). Terrestrial lowland fauna may be coupled to the

39

uplands; for example, deer base their home range in floodplains and graze in

uplands. Conversely, upland forms such as the black racer, slimy salamander,

and pine vole may use the floodplain during drydowns. Although many species

breed in both habitat types, their densities may differ considerably between

adjacent areas. However, the lower-density populations may serve as important

recruitment sources. The greenbelts of bottomland hardwoods also provide

routes for migration and restocking.

Many semiaquatic Florida turtles, such as the mud turtle and snapping

turtle, loaf and feed in marshes but need sandy upland sites to lay eggs

(Weller 1978). The river cooter is an example of a turtle which is largely

confined to water but must trek to adjacent uplands to deposit eggs (Patrick et

al. 1981). Documented cases of Florida aquatic turtles laying eggs several

hundred yards from a river are not uncommon (P. Moler, Florida Game and Fresh

Water Fish Commission, pers. comm.). Weller (1978) also indicated a need for

more information relating to the wetland-upland interface. He wrote, "Upland

areas often serve as buffers, nesting areas, or food resources for wetlands

wildlife but their relative importance is undocumented."

The eastern indigo snake is classified as a wetland species but frequently

occurs in dry, sandy areas (Kockman 1978). Speake et al. (1978) found that

indigo snakes concentrated on the higher ridges of sandhill habitat during

winter and moved down into streambottom thickets in summer. Shelter provided

by gopher tortoise burrows is critical to the survival of this snake while it

is in upland areas.

Many wildlife species need both uplands and wetlands to satisfy their food

requirements. Peak mast production occurs at different times of the year in

uplands and lowlands (Harris et al. 1979). Winter and spring is the fruiting

season for most bottomland species, while upland plants bear fruit in the

40

summer and fall. Correspondingly, both upland and wetland nesting birds often

concentrate in wetland areas during the non-nesting season (Wharton et al.

1981). Landers et al. (1979) found that black bears also respond to seasonal

differences in mast production. In North Carolina, they shift their food

preferences from predominantly bottomland species in the winter and spring to

predominantly upland fruits and nuts in summer and fall. Florida bears

primarily inhabit swamps in the center of the state but are long distance

travellers utilizing both wetlands and uplands (Williams 1978). They eat

acorns, palmetto berries and the terminal bud ("swamp cabbage") of the cabbage

palm.

Wild turkeys may be found in a variety of wet and dry habitats and

normally depend on acorns as a staple food in Florida. But they also have been

known to eat crawfish occasionally. (Wild turkeys were recently reintroduced

into the Rock Springs Run State Reserve on the Wekiva River.) During the 'egg-

laying season, female wood ducks eat a large percentage of invertebrates

obtained from the wetland-upland transitional areas (Fredrickson 1979).

Pileated woodpeckers nest and roost primarily in wet hardwoods and cypress

habitats but forage in uplands (Hoyt 1957, Jackson 1978). Conner et al. (1975)

did not find any pileated woodpecker nest trees farther than 150 m from water

in southwestern Virginia.

Jennings (1951) observed that gray squirrels in the Gulf Hammock region of

Levy County, Florida, were dispersed through all habitats while food was

plentiful in the fall. When red maple and elm began to bud and produce seed in

mid-January, the squirrels began to concentrate in the hvdric hammocks and

swamps to utilize this food source. As upland foods became available in the

spring and the lowland areas flooded, the squirrels moved to higher elevations.

Kantola (1986) found higher fox squirrel densities in ecotone or

transitional areas than in upland areas on the Ordway Reserve in Putnam County,

Florida. But she also reported that home-range size and use within ecotones

and uplands may vary with seasonal food abundance, reproductive activity, and

climate.

Between-habitat needs may vary among sites because of differences in

habitat quality. More than 33% of the 30 small vertebrates species caught by

pit-fall traps in the floodplain of the Chattahoochee River in Georgia were

classified as upland species (Wharton et al. 1981). In contrast, only 14% of

21 small vertebrates sampled by the same method along the Alcovy River in

Georgia received the same classification. This dissimilarity was attributed to

vegetation structural differences in the floodplain.

Numerous researchers have been interested in the response of small mammals

to flooding. Most studies concluded that floodplains were marginal habitats

for these species. However, Batzli (1977) found that Illinois floodplain

populations of the white-footed mouse were remarkably similar in density, adult

survival, and age structure to those in the adjacent upland areas. The

exchange of individuals between these two communities consisted mainly of a few

floodplain mice occasionally moving into the uplands. He suggested that mature

trees with abundant holes and cavities may be necessary refuges for small

mammal survival during flooding.

In a blackwater creek bottom in South Carolina's inner Coastal Plain,

Gentry et al. (1968) found that the abundance of the cotton mouse, short-tailed

shrew, and southeastern shrew were 2, 3, and 10 times greater, respectively, in

the bottomland hardwood than in the adjacent uplands. Golden mouse specimens

were collected only from the hardwoods.

Because wetlands are often the last lands to be developed, some species

normally considered upland wildlife are sometimes forced to adapt to wetlands

that can supply their habitat needs (Schitoskey and Linder 1978). When

uplands required by animals are destroyed, animals may concentrate in the

nearby wetlands. Ozoga and Venre (1968) reported that deer mice, which are

usually abundant in uplands, were also found in wetlands. White-tailed deer,

an edge species, is known to adapt well to the swamps and lowland areas (Verme

1961, Verme 1965, Sparrowe and Springer 1970). Weller and Spatcher (1965)

found that upland bird species such as the meadowlark and mourning dove nested

in unflooded portions of wetlands.

High densities of prey species also attract adaptable upland predators

such as the skunk, raccoon, and red fox: Bailey (1971) found that striped

skunk densities were greater in narrow wetlands than in adjacent uplands where

cultivation and other development adversely affected upland feeding sites.

This situation is suspected to cause an abnormally high predation rate on

waterfowl eggs by skunks.

Bobcats in the Welaka Reserve showed a preference for bottomland hardwoods

(Progulske 1982). More than 20% of the 269 recorded locations of two radio-

collared bobcats from July 1980 to December 1981 were in this type of overstory

habitat. Although the other locations were spread among seven different upland

habitat types, their need for wetlands is obvious.

Melquist and Hornocker (1983) found that although Idaho otters generally

followed stream beds, they often took shortcuts across peninsulas formed by

stream meanders. Overland travel of up to about 3 km was recorded. Extensive

cross-country movements considerably reduced the distance an animal would

normally have had to travel to reach the same destination by water. However,

these movements also subjected the animals to highway hazards. Three of nine

known mortalities were road-kills. In Great Britain, Chanin and Jefferies

(1978) reported that in some areas dead otters were found repeatedly at the

same location on roads over a number of years.

In a report that synthesized extant literature for southeastern bottomland

hardwood swamp habitats, Wharton et al. (1982) stated that bottomland animals

do not occur in the same distinct zonal pattern as plants ranging from aquatic

to upland ecosystems. Wetland wildlife inhabitants move freely into

irregularly flooded or dry zones over the year. They also noted that some

overlap among zones occurs, especially in the transitional areas characterized

by periodic annual flooding and a duration of flooding during a portion of the

growing season. Their examples of overlapping species that might occur along

the Wekiva River are mole salamander, slimy salamander, narrowmouth toad,

spadefoot toad, cricket frogs, chorus frogs, box turtle, five-lined skink,

southeastern five-lined skink, brown snake, garter snake, ribbon snakes, rat

snakes, kingsnake, southern black racer, coachwhip snake, barred owl, downy and

red-bellied woodpeckers, cardinal, turkey, common yellowthroat, wood thrush,

eastern wood peewee, white-breasted nuthatch, Swainson's warbler, carolina

wren, yellow-throated vireo, cotton mouse, golden mouse, short-tailed shrew,

least shrew, southeastern shrew, woodrat, marsh rabbit, pine vole, and eastern

mole.

The use of various bottomland hardwood ecological zones by wildlife

differs by species, season and flooding regime (Larson 1981). Some are site

specific during the breeding period while at other times they may use a broad

range of ecological zones. Larson also referred to many of the species

examples used by Wharton et al. (1981).

Many studies have documented utilization of adjacent uplands by wetland

wildlife species. Removal or alteration of this important habitat type could

44

destroy critical requirements for many species and thus render the riverine

system in that area no longer habitable for them.

II.A.8. Within-habitat Needs of Wildlife

Within-habitat needs refer to requirements of wildlife species within the

habitat areas they occupy. Habitat alterations and land use changes can affect

adjacent resident wildlife populations by fragmenting habitat to nonfunctional

sizes and shapes and by introducing disturbance factors above the tolerance

levels of some species. Because of the paucity of Florida-specific data, we

authors have included information from research conducted in other states.

Although not actual descriptions of local species/habitat relationships, there

is no reason to believe that similar situations do not exist in Florida.

Temple (1986) found that 16 of 43 bird species encountered on 49 Wisconsin

study areas occurred less frequently, if at all, in smaller-sized forest

fragments. Fragment-sensitive birds did not occur regularly in fragments that

were large in total area but lacked a secure core area more than 100 m from

edges at which forest adjoined non-forested habitat. Large stretches of linear

habitats without sufficient core areas were not utilized by sensitive species

which included three (hairy woodpecker, pileated woodpecker, and acadian

flycatcher) that have been recorded in a partial listing of Wekiva wildlife

residents (Florida Dept. of Environmental Regulation, 1983). Of the 13 other

sensitive species, it is highly likely that six (tufted titmouse, blue-gray

gnatcatcher, wood thrush, yellow-throated vireo, American redstart, and hooded

warbler) nest along the Wekiva and the remaining seven (least flycatcher,

veery, chestnut-sided warbler, cerulean warbler, ovenbird, mourning warbler,

and scarlet tanager) overwinter along the Wekiva. Stauffer and Best (1980)

45

predicted that the scarlet tanager would be absent from riparian strips in Iowa

less than 200 m wide.

Galli et al. (1976) and Stauffer (1978) found the number of breeding bird

species increased significantly with the width of wooded habitats in New Jersey

and Iowa. Of the 46 species recorded in New Jersey, 18 were size-dependent

species, limited to a particular range of forest sizes from the largest down to

a specific minimum size. The minimum areas varied from 0.8 to 10.3 hectares

and are considered characteristic of each species. As the habitat size

increased, new species appeared when their minimum habitat size requirements

were fulfilled. An example of a size-dependent species that requires a large

wooded riparian habitat is the red-shouldered hawk. This species, which is a

common year-round resident in Florida, was not found in any of the New Jersey

sites less than 10.3 hectares (about 25 ac). Of the 32 species that occurred

in Iowa wooded riparian habitats, 16 restricted breeding to relatively wide

plots.

Bird species diversity is also strongly influenced by vegetation

composition and structural heterogeneity or diversity within a habitat type

(MacArthur et al. 1962, MacArthur 1964, Weller 1978). This type of

heterogeneity is a function of foliage height and cover diversity. The

vertical and horizontal stratification of plants within a forest habitat is

positively correlated with the number of bird species that reside there.

Therefore, land-use conversion of the most diverse portion of a wooded riparian

habitat would have significantly greater negative impacts on the overall avian

community than conversion of the least diverse portion. Transition or ecotone

areas containing both wetland and upland plant species are typically the most

diverse (Wharton et al. 1982).

Fragment-sensitive species or species that require large tracts of

undisturbed, closed-canopy forest are also referred to as interior species.

Tassone (1981) found that many forest interior bird species in Virginia were

most common in riparian buffer widths above a mean value of 62 m, and certain

species occurred very infrequently below different minima. Examples are the

northern parula (80 m), yellow-throated vireo (70 m), Louisiana waterthrush (60

m), hairy and pileated woodpeckers (50 to 60 m), and acadian flycatcher (50 m).

All of these species also breed in Florida's riparian areas. Tassone (1981)

concluded that several other species, observed too infrequently to estimate

width requirements, may also be sensitive to buffer width. He suggested:

Until more extensive research is done, strips should be left 60 or
more meters in width if it is desired to provide breeding habitat for
these species. They should also be situated as corridors between
other hardwood and pine-hardwood stands whenever possible. Strips
located on larger streams should be 100 or more meters in width to
make full use of the intrinsic wildlife values associated with the
riparian zone.

Small and Johnson (1985) recommended that riparian buffer strips be 75 m

wide to protect woodland bird communities in Maine. Johnson's (1986) study of

the tolerance levels of some woodland breeding species supported this

suggestion.

Fragmentation also affects squirrel populations. In an eastern Texas

study of clearcut effects conducted two to four years after tree harvest, gray

and fox squirrels were abundant in mature woody riparian zones wider than 55 m

but were rare in riparian zones narrower than 40 m (Hedrick 1973). The highest

squirrel density index was calculated for the widest stringer study plot (185

m),

The home-range size requirement for an individual gray squirrel is about

two acres (Burt and Grossenheider 1964). Larger Wekiva wildlife residents

require larger areas. For example, home-range sizes for individual gray foxes

and bobcats in Welaka, Florida, is about 625 and 4,580 ha, respectively

(Progulske 1982). Any actions that reduce suitable contiguous habitat below

these home-range sizes will prevent individual animals from occupying those

areas. Of course, viable self-sustaining wildlife populations require

considerably larger home ranges than do of individuals.

Aside from the fragmentation factor, various human activities may alter

behavior patterns or cause undue stress to some wildlife individuals in

adjacent habitats. This is a relatively new aspect of wildlife biology, and

very little research data are available regarding .the effects of specific human

activities on different wildlife species.

Noise associated with construction, operation, and maintenance of

developments can cause harmful impacts on wildlife. Animals that rely on their

hearing for courtship and mating behavior, prey location, predator detection,

homing, etc., will be more threatened by increased noise than will species that

utilize other sensory modalities. However, due to the complex

interrelationships that exist among all the organisms in an ecosystem, direct

interference with one species will indirectly affect many others.

Unfortunately, few data are available that demonstrate the effects of noise on

wildlife. Much of what is found in the literature lacks specific information

concerning noise intensity, spectrum, and duration of exposure.

A few laboratory studies have documented decreased nesting and aversion

reactions of small mammals in response to various sound frequencies (Sprock et

al. 1967, Greaves and Rowe 1969). Other research has focused on the use of

noise for damage or nuisance control. Diehl (1969), Crummett (1970), Hill

(1970), and Messersmith (1970) reported on the success of using sound to repel

rodents, bats, rabbits, deer, and birds.

The Environmental Protection Agency (1971), Ames (1978), and others have

reported on the physiological responses of laboratory and farm animals to

different noise stimuli. These data suggest that chronic sound may cause

stress on wild animals which in turn may affect reproduction.

Most people can easily relate to auditory interference, annoyance, hearing

damage, sleep interference, and stress-related effects of noise. And in

response to public demands, many noise-control ordinances have been enacted to

reduce these adverse impacts. However, it is doubtful that these laws

sufficiently protect wildlife species.

van der Zande et al. (1980) found that breeding densities of three open

grassland bird species were significantly reduced within 500 m of quiet rural

roads and 1,600 m of busy highways in The Netherlands. They suggested that

disturbance is also caused by farms, other buildings, and plantations.

Although Robertson and Flood (1980) found no significant differences in

several vegetation variables between developed and undeveloped sites along

Ontario shoreline areas, bird species diversity was negatively correlated with

disturbance. A greater number of species was found in the natural areas than

along shorelines where cottages were present and boat use was high. Several

species which also nest in Florida (yellow-billed cuckoo, yellow-throated

vireo, American redstart, and pileated woodpecker) were more abundant in, or

restricted to, the less disturbed transects.

The Florida Game and Fresh Water Fish Commission is in the process of

developing habitat protection guidelines for threatened and endangered species

to provide land-use decisionmakers with recommendations and management

practices that would integrate protection for these habitats in the context of

large-scale developments. Many of these recommendations will address within-

habitat needs such as those identified in The Recovery Plan for the Bald Eagle

49

in Florida (Murphy et al. 1984). This U.S. Fish and Wildlife Service document

states that no human activity should occur within 750 ft (250 m) of active

eagle nests and that the building of housing developments should be restricted

within 750 to 1,500 ft (500 m) of active nests. Potential bald eagle nesting

sites are currently abundant along the Wekiva.

Predation and harassment of wildlife by cats and dogs are other

detrimental effects of development adjacent to wildlife habitat areas.

Although relationships between domestic animals and wildlife are not fully

understood, there is enough information in the literature to justify a

legitimate concern. Several authors have documented the occurrence of wildlife

prey in the diets of free-ranging cats and dogs and the effects of their

predatory behavior on individual wildlife animals and populations (Errington

1936, McMurry and Sperry 1941, Bradt 1949, Hubbs 1951, Parmalee 1953, Eberhard

1954, Korschgen 1957, Smith 1966, Corbett et al. 1971, Gilbert 1971, Jackson

1971, Gavitt 1973, George 1974, Gill 1975, van Aarde 1980).

Cats can be especially devastating on local wildlife populations. Hunting

is a feline instinct, and predation rates are not related to hunger (Davis

1957, Holling 1966, Holling and Buckingham 1976). Bradt (1949) reported that a

single cat, who regularly consumed domestic food, killed over 1,600 mammals and

60 birds in Michigan during an 18-month period.

van Aarde (1980) found that free-ranging cats associated with households

utilized open fields, edges, and timber areas. Home range sizes for females

and males were 30 to 40 ha (about 100 ac) and four to eight square km (about

four sections), respectively. In an Illinois study, the average home range for

females and males was 112 and 228 ha, respectively (Warner 1985). These free-

ranging rural cats spent about 40% of their time on farmsteads. Approximately

50

73% of the radiolocations of these cats occurred in edge areas such as

roadsides, field interfaces, farmstead perimeters, and waterways.

Beck (1973) reported that the average home range for free-ranging dogs in

Baltimore was about four acres. Schaefer (1978) reported that dogs in rural

Iowa traveled up to two miles to kill pastured sheep.

Alteration of natural vegetation or other components of wildlife habitat

may make modified areas unsuitable for some species. As these unsuitable areas

encroach upon natural areas, core habitat fragments become smaller and the

number of supported species is reduced. Various forms of disturbance can also

adversely affect adjacent wildlife populations. An upland buffer would reduce

or eliminate negative disturbance impacts associated with development.

II.B. Review of Buffer Zone Regulations

In this section methods and models for determining requirements and

physical dimensions of buffer zones are presented. Buffer zones are normally

considered to protect quality and wetland-dependent wildlife and to reduce the

potential for negative impacts on a wetland resulting from adjacent upland

development.

The determination of buffer requirements (i.e., widths in relation to land

use, soils, topography, etc.) for the most part are not quantitative, but rely

on what is felt to be qualitative evidence for the necessity of buffering

development impacts. Never-the-less, there are numerous rules and regulations

that have been promulgated throughout the country that require buffers around

wetland areas.

The Pinelands Area

One of the most noteworthy buffer zones is that developed by the New

Jersey Pinelands Commission. The Pinelands Area (also known as the Pine

Barrens) is an area of about 445,000 ha of an interrelated complex of uplands,

wetlands, and aquatic communities in southeast New Jersey. It is a largely

undeveloped region within the Northeast Urban Corridor and was designated as

the country's first National Preserve in 1978. In the next year the State of

New Jersey passed the New Jersey Pinelands Protection Act that in essence

called for the preservation, protection, and enhancement of the significant

values of Pinelands land, water, and cultural resources.

In response to both federal and state legislative mandates, the Pinelands

Commission developed a Comprehensive Management Plan to preserve and protect

the unique and essential character of the Pinelands ecosystem. In the plan,

wetlands protection is fostered through a regional land allocation program, a

land acquisition program, and a wetland management program. The management

program prohibits most development within wetlands and also requires an upland

buffer around wetlands. The required "transition or buffer area" established

between proposed development and adjacent wetlands measures 300 ft wide, and

except for those regulated uses which have been described, no development may

occur within the area.

Rationale for the width of the Pinelands buffer is derived from areal

nutrient dilution models (Harlukowicz and Ahlert, 1978; Trela and Douglas, 1979

and Browne 1980) that were used to predict the travel distance necessary for

groundwaters laden with nutrients (from septic tank leachate) to be diluted to

background levels. Depending on the values used for input variables to the

model, it predicted distances to attain background concentrations of 2 mg/1

NO3-nitrogen of between 325 ft and 600 ft.

Since the implementation of the Pinelands Comprehensive Management Plan,

327 development applications that involved wetland issues have been reviewed by

the Pinelands Commission. More than one third of these were considered

inconsistent with the plan because they proposed development within wetlands or

within 300 ft of wetlands (Zampella and Roman, 1983).

Figures compiled by Zampella and Roman (1983) show that more than over

half of the development applications that involved wetland issues were

consistent with the wetland management provisions of the plan. However, among

the conditions imposed on these projects was the need to establish and maintain

a buffer adjacent to wetlands. Approximately one third of these applications

were required to maintain the maximum buffer of 300 ft, while the remainder

were required to maintain variously sized buffers which averaged about 135 ft

in width.

The Pinelands scheme of determining wetland buffers uses a Wetlands Buffer

Delineation Model that evaluates relative wetland quality, and relative impacts

of development. Prior to evaluating wetland quality a determination of the

presence of threatened or endangered species is made, and if the wetland is

known to support resident and/or breeding populations and if the wetland area

is critical to their survival, the wetland is ranked having the maximum

relative wetland value index and is, assigned a buffer of 300 ft. The

following qualities, values, and functions are evaluated as part of the

wetlands evaluation scheme and are evaluated after determination of presence of

threatened and/or endangered species is made:

Vegetation quality,
Surface water quality,
Potential for water quality maintenance,
Wildlife habitat, and
Socio-cultural value.

Each of the five general factors listed above has a score from I to 3,

where 3 is high value. A wetland is scored for each factor, and scores are

summed and divided by 5 to obtain a Relative Wetland Value Index.

Next, Potential for Impacts is evaluated by ranking the following factors:

Potential for site-specific wetland impacts,
Potential for cumulative impacts on a regional basis, and
Significance of watershed-wide impacts.

Once ranked, the scores for each of the potential impact factors are summed and

divided by 3 to obtain an average score for the Relative Potential for Impacts

Index.

To assign buffer areas, the final step is to average the Relative Wetland

Value Index and the Relative Potential for Impacts Index to obtain a Buffer

Delineation Index. Thiqsfinal index is used to determine the buffer distance

in three different Land Capability Areas using a table of assigned values.

The strengths of the methodology are: (1) Wetland attributes and the

potential for on- and off-site impacts are evaluated for each individual

wetland and each development proposal on a case-by-case basis. Depending on

the type of wetland, existing terrain, existing land use designations, and type

of development, the buffer varies from a minimum of 50 ft to a maximum of 300

ft. (2) The method is quantitative and repeatable thus insuring some degree of

consistency. (3) It does not require detailed scientific information about a

wetland.

State of New Jersey

The House of the State of New Jersey has passed a bill whose short title

is "Freshwater Wetlands Proctction Act." As of this writing the State Senate

has not acted upon the bill the ace would establish

...a transition area adjacent to all wetlands having the following
purposes:

(1) Ecological transition zone from uplands to wetlands which is an
integral portion of the wetlands ecosystem providing temporary refuge
for wetlands fauna during high water episodes, critical habitat for
animals dependent upon but not resident in wetlands, and slight
variations of wetland boundaries over time due to hydrologic or
climatic effects:

(2) Sediment and stormwater control zone to reduce the impacts of
development upon wetlands and wetland species.

The bill would also establish the width of the transitional area as no greater

than 150 ft and no less than 75 ft for freshwater wetlands of exceptional

resource value, no greater than 50 ft and no less than 25 ft for freshwater

wetlands of intermediate resource value, and no transitional area for wetlands

of ordinary resource value. The act prohibits the following activities in

transition areas:

Removal, excavation or disturbance of the soil;
Dumping or filling with any material;
Erection of structures;
Placement of pavements; and
Destruction of plant life which would alter the existing pattern
of vegetation.

New York State Wetlands Protection Act

The recent adoption of the Freshwater Wetlands Act (6 NYCCR) protects

wetlands 12.4 ac or greater in size. The Act recognizes ten specific functions

and benefits which freshwater wetlands provide for the public and for the

environment. To achieve the goals of the Act to protect, preserve, and

conserve New York's freshwater wetlands, the statute makes provisions for

mapping, classification, regulation and local government protection. The act

provides for the regulation, of a 100-foot buffer adjacent to wetland

boundaries. Provisions exist in the law to extend the 100-foot adjacent area

if necessary to protect the wetland.

California Coastal Act of 1976

The California Coastal Act (Public Resources Code, Division 20)

established the California Coastal Commission which in turn prolmugated the

Statewide Interpretive Guidelines (SIG) to assist in the application of various

Coastal Act policies to permit decisions. The SIG has set standards for siting

development adjacent to environmentally sensitive habitat areas (ESHA).

Wetlands are considered environmentally sensitive habitat areas, as are

estuaries, streams, riparian habitats, lakes, and portions of open coastal

waters. "Adjacent to" is defined as situated near or next to, adjoining,

abutting, or juxtaposed to an ESHA. This usually means that any proposed

development in an undeveloped area within of up to 500 ft of an ESHA will be

considered to be adjacent and subject to critical review.

Criteria for establishing buffer areas around ESHA's are designed to

provide essential open space between the development and the ESHA. Development

allowed in a buffer area is limited to access paths, fences necessary to

protect the habitat area, and similar uses which have either beneficial effects

or at least no significant adverse effects. The buffer is not itself part of

the ESHA, but it is a "screen" that protects the habitat area from adverse

environmental impacts caused by development. Widths of buffer areas are

variable depending upon site analysis. It is usually a minimum of 100 ft;

however, it may reduce for small projects (defined as single family home or one

commercial office building). Additionally, provisions are suggested that the

100 ft may be reduced if the applicant can demonstrate that it is not necessary

to protect the resource. Standards for determining the appropriate width of

the buffer area:

I. Biological significance of adjacent lands,
2. Sensitivity of species to disturbance,
3. Susceptibility of parcel to erosion,
4. Use of natural topographic features to locate development,

5. Use of existing cultural features to locate development,
6. Lot configuration and location of existing development, and
7. Type and scale of development proposed.

Humboldt County, California

Humboldt County's Southcoast Plan defines riparian corridors and limits

development within them. It defines riparian corridors to include the uplands

on both sides of streams as follows:

Riparian corridors on all perennial and intermittent streams shall
be, at a minimum, the larger of the following: (i) 100 ft, measured
as the horizontal distance from the stream transition line on both
sides; (ii) 50 ft plus four times the average percent of slope,
measured as a slope distance from the stream transition line on both
sides; (iii) where necessary, the width of riparian corridors may be
expanded to include significant areas of riparian vegetation adjacent
to the corridor, slides, and areas with visible evidence of slope
instability, not to exceed 200 ft. (Humboldt County Planning
Department 1981, as cited by Ray et al. 1984)

Uses in the riparian corridor have been limited to minor facilities,

minimal timber harvest, maintenance of existing facilities, residential wells,

road and bridge replacement, and tree removal for disease control. There must

be, at a minimum, replanting of vegetation for any approved disturbance, and

any trees that have visible evidence of current use as nesting sites by hawks,

owls, eagles, osprey, herons, or egrets must be retained.

Massachusetts Wetlands and Floodplain Protection Act

The Massachusetts Wetlands and Floodplain Protection Act recognizes three

types of wetlands: 1) land areas bordering water bodies; 2) land under water

bodies; and 3) land subject to tidal action, coastal storm flowage, or

flooding. These wetlands must border on a listed water body with 50% or more

wetland vegetation or within the 100-year floodplain. Permits are required for

virtually any development activity in, near, or affecting jurisdictional areas.

Work within a 100-foot buffer zone around wetlands may require a permit since

57

the Act defines "bordering" as a distance of 100 ft. The regulation provides

that the extent of protection is to be 100 ft landward from wetland resources

or 100 ft landward from the water elevation of the 100-year flood, whichever is

greater.

Rhode Island Coastal Resources Management Program

The General Laws of the State of Rhode Island, Title 46, Chapter 23,

establishes the Coastal Resources Management Council (CRMC). The CRMC has

developed the Coastal Resources Management Program. The Program establishes

setbacks ranging from 50 to 180 ft from the inland boundary of coastal features

such as beaches, wetlands, cliffs, banks, rocky shores, and existing manmade

shorelines and apply to the following activities:

Filling, removal, grading;
Residential buildings;
Sewage disposal systems;
Industrial, commercial, recreational structures; and
Transportation facilities.

The setback minimum in areas not designated as Critical Erosion Areas is 50 ft,

and within Critical Erosion Areas it is 30 times the calculated average annual

erosion rate, which has been determined to vary between 75 and 180 ft.

In addition to setback requirements, the program establishes Buffer Zones

defined as land areas on or contiguous to a shoreline feature that is retained

in its natural and undisturbed condition by the applicant. Buffer zones must

be tailored to on-site conditions and the specific alterations and activities

that are taking place. Further, the determination of the boundaries of a

buffer zone need to balance the property owners' rights to enjoy their property

with the Council's responsibility to preserve and, where possible, to restore

ecological systems. There are four benefits of buffer zones as follows:

1. Erosion control,
2. Preventi.-n of water body pollution,

3. Preservation and enhancement of scenic qualities, and
4. Protection of flora and fauna.

The buffer zones are established according to values and sensitivities of the

site as assessed by the Council's staff engineer and biologist, and the area

may be wider than the setback distance. The buffer must be maintained by the

applicant as an undisturbed area and in its natural condition.

Marion County, Florida

The Board of County Commissioners of Marion County, Florida adopted an

emergency ordinance (Marion County Ordinance No. 73-4) in June of 1973

"...prohibiting dredging, filling, earth moving, and
landclearing, and underbrushing except mowing, pruning, and
care of existing lawns and planted trees and shrubs for a
distance of 500 ft from the water's edge upon either side
of Rainbow River or Blue Run in Marion County, Florida,
between Rainbow Springs and the northern city limits of the
city of Dunnellon...."

Cross Creek Buffer Requirement

In an amendment to its Comprehensive Plan, the Alachua County Commission

adopted a 300-foot setback line from lakeshores for most development in a

designated area in the southeast part of the county known as Cross Creek. The

only construction allowed within the setback zone are structures such as docks

built for access to the lake. The setback line is measured upland from the

jurisdictional line set by the Department of Environmental Regulation or the

St. Johns River Wacer Management District, or, where the lakeshore has been

cleared, from the 100-year floodplain line. The setback width of 300 ft was

determined qualitatively. The plan amendment is currently being rewritten. A

new version expected to be completed by the second week of October 1987, may

require a variable setback line based on such factors as soils, slope, habitat,

and groundwater, with a minimum setback of 75 ft. (John Hendricks, Alachua

County Department of Environmental Services, pers. comm.)

Hillsborough County Wetland Buffer Requirement

A 30-foot buffer around "conservation areas" and a 50-foot buffer around

"preservation areas" are required by Hillsborough County. The former category

includes most freshwater wetlands and Class III waters; the latter includes

saltwater wetlands, Class I waters, and critical habitats. These categories

originated in the Conservation Element of the county's Comprehensive Plan

adopted in the mid-1970s. That document set a policy to require "appropriate

buffers" around those areas, but no width or formula to calculate width was set

forth. However, in the definition of "natural shoreline"--another category of

conservation area--a 30-foot width was mentioned as one way to delimit natural

shorelines, and that number was applied by county zoning staff in rezoning

applications involving freshwater wetlands. The larger buffer was required

around preservation areas because they are considered to be more sensitive than

conservation areas, but no formula was used to establish that width.

Both buffers were adopted by the Hillsborough County Commission in a 1985

ordinance (revised in 1987) and as part of its zoning code. A variety of uses

are permitted within the buffer zone, such as detention ponds, stilted

structures, and boardwalks, but impervious surfaces are prohibited. Removal of

natural vegetation is discouraged but not prohibited. (Charner Benz,

Hillsborough County Planning and Growth Management Group, pers. comm.)

II.C. Summary: Resource Buffer Zones

Effects of activities in the upland areas immediately adjacent to

floodplain wetlands may be separated into three time-related categories:

1. Direct and immediate impacts which occur during the
construction phase,

2. Impacts which occur immediately following the construction
phases during the period of stabilization, and

3. Long-term impacts that are permanent as a result of
the alterations themselves or subsequent use and/or
management of the altered areas.

These development impacts affect three broad sets of parameters normally

considered of particular importance from a regulatory perspective. They

include:

1. Water quality,
2. Water quantity, and
3. Aquatic and wetland dependent wildlife.

Thus each of these attributes may be affected during construction, immediately

following construction, or in the long term through human uses and management.

Water Quality. The need for a buffer to insure that water quality is not

degraded as the result of upland development in areas immediately adjacent to

floodplain wetlands of the Wekiva River Basin is related to two potential

sources of degraded waters. The first is sediments carried by surface waters

from developed lands into down-gradient wetlands and waters. The second is

dissolved pollutants and suspended particulate matter that degrade water

quality through what might be termed chemical pathways. These include

nutrients, pesticides, and particulate organic matter which increase nutrient

loading, may be harmful to aquatic plant and animal life, or increase BOD.

Both sources of degraded water may come from the result of runoff during

construction, during the period immediately following construction as the

landscape stabilizes, or as a result of urban stormwater runoff long after the

period of initial urbanization.

Loss of valuable wetland functions as a result of sedimentation may be the

most serious impact during construction activities. Sedimentation causes loss

61

of soil properties and vegetation that are characteristic of wetlands and loss

of wildlife habitat. In addition, sediments may carry significant amounts of

pollutants and nutrients. Siltation and sedimentation of the wetland edge

result from increased water velocities acting on soils susceptible to erosion,

removal of ground cover in adjacent areas, or increased volumes of stormwaters

entering the wetland that carry increased amounts of sediments.

Where the waterward lots of a development are not incorporated into the

stormwater management system, the potential for long-term decreased water

quality after the construction phase is related to surface runoff from

driveways and lawns and groundwaters carrying increased levels of nutrients and

pesticides/herbicides from lawn care practices. While the rule criteria of the

District requires specific stormwater management practices to mitigate water

quality impacts from stormwaters, the rules allow 10% of the developed area to

be "unconnected," that is, to receive no treatment.

Water Quantity. The water flowing into the waters and wetlands of the

Wekiva River Basin is derived from three general sources. The first is

groundwater (or surficial groundwater) that becomes surface seepage where

ground surface elevations intersect the zone of saturation. This usually

occurs on inclines near the base of sandy slopes and is characterized by

surface soils that are saturated, shallow ponding of seepage waters, and small

seeps that coalesce into undefined flowing surface streams. The second source

is surface runoff as sheet flow from adjacent higher ground during and

immediately following rainfall events, and the third source from artesia flow

of deep groundwaters.

Development in the zone immediately upland from the floodplain wetland

edge may have dramatic impacts on the quantity of water from surface and

surficial groundwater sources. Seepage is usually derived from rains that have

62

recharged groundwaters in higher sandy soils and are moving downslope over a

broad front thus, seepage areas tend to extend over the toe of the slope in a

perpendicular line across the slope where the ground surface intersects the

groundwater elevation. The site of seepage is usually dominated by floodplain

vegetation. Alteration of the ground surface elevation through excavation in

areas immediately upslope from the site of seepage may intersect the

groundwater plain, and if positive drainage is maintained, can lower

groundwater levels and essentially cut off further downslope seepage. If no

positive outfall is maintained, the effects of excavations immediately upslope

from seepage sites may have only minor impacts, since waters may be intercepted

and discharged on the upslope side of the excavation, but may become

groundwaters through recharge on the downslope side.

In most areas where seepage is present, groundwater elevations immediately

upslope are quite near the soil surface and are not conducive to many land

uses. General engineering practices in areas with high water tables is to

lower water table elevations by drainage or raise ground elevations by filling

to make such areas more suitable for development. While filling may have

little impact on subsurface water flows, drainage has the overall effect of

short-circuiting groundwater flows, cutting off potential groundwater seepage

downslope, and increasing surface discharge elsewhere.

The effects of development on surface water flows in areas immediately

upslope of the wetland edge are more easily seen and more direct. Impervious

surfaces and sod increase runoff coefficients significantly over coefficients

characteristic of natural lands. If these waters are not intercepted by a

stormwater management system, they enter the wetlands and eventually the waters

of the river. The increased volumes and velocities may cause significant

degradation of downslope communities. However, under most development

conditions, these waters are required to be routed through a stormwater

management system, thus lessening the burden of the downslope communities. Yet

these very requirements may in effect decrease sheetflow from rainfall events,

since intercepted surface waters are routed elsewhere. The obvious

implications of altered surface runoff and water quality have already been

discussed.

The temporal differences of impacts on ground and surface water quantity

before, during, and after construction may be significantly different.

Construction practices may dictate that surface waters and groundwaters be

routed differently than will be the case after construction, or that

groundwaters be temporarily drained to facilitate construction activities.

Once construction is completed and the stormwater system is under operating,

the impacts on surface and groundwater quantity may vary little over time.

Generally, stormwater discharge requirements within the Wekiva River Hydrologic

Basin are designed to minimize negative impacts to surface waters through

detention/retention, off-line treatment, and additional criteria. However,

these requirements, and the fact that typically all lands within a project are

engineered as part of the stormwater system, have the effect of isolating

developed land. Contributions of overland flow runoff to downslope communities

are minimized under such circumstances, since normal engineering practices are

to discharge collected stormwaters at one or several point sources.

Aquatic and Wetland-Dependent Wildlife. Construction, operation, and

maintenance of developed areas immediately adjacent to Wekiva Ri-.r wetlands

will result in both direct and indirect impacts on aquatic and wetland

dependent wildlife. Noise during construction will interfere with

vocalizations necessary for courtship, mating, prey location, and predator

detection. Although this hindrance will be only temporary, there is no way to

64

predict long-term consequences. After individuals disperse to less noisy

territories, they may or may. not return. Chronic and acute noises of occupied

residential communities will cause less severe impacts. However, auditory

interference and stress-related effects will still occur in close proximity to

these developed areas. A vegetated buffer will help to attenuate harmful

sounds.

Domestic animal harassment and predation on wildlife will also increase

when pets are brought into adjacent developments. Development covenants and

local laws usually require owners'to leash their pets. However, the

probability of a few free-ranging dogs and many free-ranging cats is quite

high. Because cats tend to spend most of their prey-searching time in edge

areas, an upland buffer will help to alleviate cat predation on wetland

wildlife species.

Indirect impacts on wildlife will be related to habitat alteration

resulting from construction. The removal of natural uplands vegetation

adjacent to the wetlands will cause the originally mesic community along the

outer edge of the wetlands to become more xeric. Drier community species will

invade and eventually dominate this area. The ultimate result will be a

reduction in the wetlands and a corresponding decline in those species that are

wetland-dependent. In some areas, the width of the wetlands may become too

small, and the species composition will change because those with larger

spatial requirements will no longer be present. Increased predation and

parasitism associated with sharp induced edges will also encroach farther into

the forest.

Wetland-dependent species that require uplands to fulfill at least some of

their life functions will extirpated from the Wekiva River Basin if there is no

65

access to these uplands. An upland buffer adjacent to the wetlands will be

needed to protect this necessary habitat element.

III. LANDSCAPE ECOLOGY OF THE WEKIVA RIVER BASIN

III.A. Landscape Perspective

III.A.I. Physical Description

Topography. The landscape of the Wekiva River Basin is a complex mosaic of

wetland and upland habitats distributed over approximately 130 sq mi in north

central Orange County, southeastern Lake County, and western Seminole County.

The basin contains the physiogeographic province known as the Wekiva River

Plain. Its topography is flat to gently rolling. Elevation declines towards

the northeast, where the Wekiva Plain merges into the valley of the St. Johns

River.

On the western side of the upper Wekiva River, the land slopes gently

upward to the sandhills, from whence it rises abruptly to elevations of 120 to

195 ft above sea level. Sandhills and sloping terraces separate large tracts

of swampy lowlands and determine the circuitous route by which the waters of

Rock Springs Run State Reserve flow southeastward. These waters turn north

below the confluence of Wekiva Springs Run and the Little Wekiva River. North

of this confluence, the floodplain narrows and turns to the northeast above

State Road 46. The floodplain broadens out again where the Wekiva River joins

Blackwater Creek and the St. Johns River in the Lower Wekiva River State

Reserve (DNR, 1985, 1987).

Along the west bank of the Wekiva River upstream (south) of its confluence

with the Little Wekiva, elevation rises gradually towards the northwest from 15

ft above sea level along the river to approximately 35 ft above sea level.

67

68

Beyond the 35-foot contour, the land surface rises abruptly into upland

sandhill tracts in the north central section of Rock Springs Run State Reserve.

Downstream (north) of the Wekiva/Little Wekiva confluence, the floodplain

narrows abruptly, and the flow gradient increases rapidly to 1.6 ft/mi. The

flow gradient in this stretch of the central Wekiva River is one of the

steepest in east-central Florida, and the elevation along this section of the

Wekiva rises much more steeply above the narrowed floodplain than it does in

the flat swamps of the upper and lower sections. The width of the

Wekiva/Little Wekiva floodplain narrows to about 1,000 ft along the central

portion of the river, decreasing rapidly from its maximum dimensions of about 5

mi east/west by 3 mi north/south in the Wekiva Swamp. The floodplain broadens

out once again in the Lower Wekiva River State Reserve (near the confluence

with Blackwater Creek) to a width of 1 mi at the Wekiva's confluence with the

St. Johns River (DNR, 1985, 1987).

Blackwater Creek is a major tributary of the Wekiva which drains the

watershed lying just north of the Wekiva Basin. Its watershed covers

approximately 125 sq mi, nearly all of which is in southern Lake County.

Blackwater Creek's headwaters are in Lake Dorr, and its floodplain varies in

width from 800 ft at State Road 44A to over a mile along its upper and lower

reaches. Another major tributary, Seminole Creek, joins Blackwater Creek at

the eastern edge of Seminole Swamp. Seminole Swamp extends north and west from

the confluence of Blackwater and Seminole creeks, covering an area of about 3

mi east/west and 1.5 to 2 mi north/south. This swamp lies between State Road

46A and State Road 44 in southern Lake County. Blackwater Creek flows into the

lower Wekiva near the confluence of the Wekiva and the St. Johns. The waters

of the Wekiva River Basin merge with those of the St. Johns and flow northward

to enter the Atlantic Ocean at Jacksonville, Florida (DNR, 1985, 1987).

The wetlands of the Wekiva Swamp's northern end connect with the lower

reaches of Seminole Swamp through a narrow wetland corridor (not shown on all

maps) which bisects the lands lying above the west bank of the Wekiva River and

the northern borders of Wekiva Swamp. This wetland strip is an extremely

important feature of the basin. It provides a wildlife corridor through which

animals can move between these otherwise disjunct wetland communities (see Noss

and Harris, 1986).

The basin's diverse topography engenders a wide range of habitats, ranging

from submerged aquatic plant associations to upland sandhill communities.

Local soil characteristics and elevational effects combined with past and

present disturbances influence the development of the various plant communities

found within the Wekiva Basin.

Geology and Hydrology. The Wekiva Plain was formed by Pleistocene terrace

sands deposited during times of elevated sea level which occurred during

interglacial periods of the Pleistocene. The most recent extended saltwater

incursion probably occurred during the Sanamonian interglacial about 100,000

to 200,000 years ago, when sea levels were about 20 ft above those of the

present time (Bloom, 1983). Dunes were built up during the dry and windy

periods of earlier and later glacial periods (Watts and Hansen, in press).

These sands overlie the older marine limestones, clays, and marls of the

Hawthorne Formation, which rest upon the deeper, porous Paleocene limestones

and dolomites ccnt3ining the Floridan Aquifer (Heath and Conover, 1981). The

non-sedimentary bedrock of the Florida peninsula upon which these marine

deposits have accumulated lies at least 4,100 ft below the present ground

surface and ties in with the crustal rock of the continental plate.

Soils in the Wekiva Basin can be grouped in five associations. The

Freshwater Swamp Association occurs in the Wekiva River floodplain and the

70

bottomlands around the upper part of Rock Springs Run State Reserve. These

soils are usually saturated with water and are flooded much of the time. They

are adjoined by the Manatee-Delray Association, which grades uphill into the

P.omp-no-Delray Association. The two latter groups include soils which are

sandy, nearly level, poorly drained, and subject to periodic flooding. The

Myakka-Placid Association soils of the north-central and northeastern portions

of the Rock Springs Run State Reserve are sandy, nearly level, poorly drained,

and not normally subject to flooding. The St. Lucie-Pomello Association soils

of the uplands are sandy, moderately to excessively well-drained, and not

subject to flooding. They are restricted to the upland sandhill areas of the

Wekiva Basin (DNR, 1985; see also USDA, 1960, 1966).

Organic matter content of soils in the Wekiva Basin is high only in

wetland soils. Silt deposition and decayed plant materials have formed the

rich organic soils of river alluvium, swamp muds, and peat bogs in the basin.

The general absence of well-developed topsoil horizons in sandy upland

soils of the region encourage the rapid infiltration of rainwater. However,

the lack of organic matter, clays, and silts in the surface soils restricts

nutrient availability and limits the water-holding capacity of upland soils.

Although highly permeable for the most part, the upper horizons of upland sandy

surface soils quickly reach saturation during heavy rains, and sheetflow

movement of rainwater can result. Problems associated with surface runoff and

erosion may develop in conjunction with the heavy rains accompanying major

storm fronts, especially in areas where the protective vegetation cover has

been disturbed or removed. Ecosystem disturbances of historical importance to

the disruption of the vegetative cover of Wekiva Basin soils include logging,

livestock grazing, high-intensity recreational use, construction and

development, and wildfire. Wildfire is by far the least destructive of these

disturbances to the natural vegetation and wildlife. Indeed, fire is an

essential ecological component of several endemic plant communities native to

the region.

Lakes and sinkholes are scattered throughout the basin as the result of

the formation and subsequent collapse of solution cavities within the upper

layer of the underlying Hawthorne Formation limestone bedrock. Sudden surface

collapses due to sinkhole formation have in recent years swallowed buildings,

automobiles, and sections of roads in the suburbs of nearby Orlando. The

prevalence of sinkhole formations in the regions bordering the southern edge of

the Wekiva Basin dictate caution in siting industrial and residential

development.

The same forces which create sinkholes also created the many artesian

springs for which the Wekiva Basin is renowned. The water in these springs

comes from the Florida aquifer in deep Paleocene limestones and dolomites via

faults in the overlying Hawthorne Formation. Crevices and solution channels

have penetrated the otherwise impervious clay and marl layers of the Hawthorne

Formation and dissolved the underlying limestone. These provide the means for

developing surface outlets for artesian springwaters.

The major springwater sources of the upper Wekiva River are Wekiva Springs

(48 million gallons per day [MGDI), Rock Springs (42 MGD), Sanlando Springs (14

MGD), and Sheppard Springs (11 MGD). More than two-thirds of the Little Wekiva

River's flow is provided by a combined discharge of 14 MCD from Seminnle, Palm,

and Starbuck Springs. Springflows from the Floridan Aquifer comprise most of

the Wekiva River's water volume, which averages 186 MGD at State Road 46 (DUR,

1986, 1987). Reductions in flow of several major springs during the past two

decades have been attributed to water withdrawals from numerous deep

72

groundwater wells in the region which supply water for residential, industrial,

and agricultural uses (FWR, 1985; see also Heath and Conover, 1981).

Two man-made springwells at Wekiva Falls with a combined flow of 40 MGD

enter the Wekiva just south of State Road 46. Withdrawals from these deepwater

wells may be partly responsible for the recent decrease in springflow volume

from various natural springs in the Wekiva Basin. However, cutting off

upstream springflows from Wekiva Falls may not have particularly adverse

effects on lower Wekiva River wetlands. The proximity of this source to the

confluence of Blackwater Creek and the St. Johns River and the steep flow

gradient in this section of the Wekiva River tend to reduce the importance of

this water source to the ecology of the Lower Wekiva Aquatic Preserve region.

The natural channelling of the river into a relatively restricted wetland zone

caused by the increase in flow gradient tends to mitigate the negative, impacts

on wetlands of the lower Wekiva River from reduced flows caused by the Wekiva

Falls wells.

Further drawdowns in the regional water table due to over-exploitation of

water resources from the Floridan Aquifer will have major impacts on the

wetland habitats of the Wekiva Basin. Since about 70% of the Wekiva River's

water comes from springs, any reduction in springflow volumes will adversely

affect the size, distribution, character, and quality of state-domain wetlands,

waterways, and aquatic habitats in the Wekiva Basin (DNR, 1987).

Climate. The Wekiva Basin lies in the southernmost latitudes of the humid

subtropical climate zone of the southeastern United States. About two-thirds

of the precipitation in the basin falls between June and September. On

average, 10 to 15% of the annual rainfall occurs from December to February.

Annual rainfall averages 52 inches, with most rain deposited during brief,

heavy showers and thunderstorms (Heath and Conover, 1981). Tropical storms,

73

including hurricanes, contribute on average about 15% of the precipitation from

June to November. Rainfall events greater than 30 inches within a single 24-

hour period have been recorded during hurricanes at some Florida locations, and

such events are possible in the basin if a severe hurricane moves through

central Florida (Yoho and Pirkle, 1985). Tornadoes are not uncommon in central

Florida, and torrential rains and high winds typically accompany the storm

fronts which spawn these dangerous and sometimes highly destructive storms.

The average temperature for the central Florida region is 72 degrees F.;

ambient temperatures of 85 degrees F. or higher can occur during all months of

the year. Arctic air masses sometimes penetrate into the region during winter

months, bringing usually brief but occasionally intense episodes of sub-

freezing temperatures (Yoho and Pirkle, 1985; Heath and Conover, 1981). Record

freezes occurring during the past few years have devastated the citrus industry

in Lake County and most of north central Florida. Tens of thousands of acres

of orange groves were severely damaged or killed outright during sometimes

week-long periods of sub-freezing temperatures. Whether these recent spates of

unusually cold winter weather in central Florida represent a long-term trend

toward harsher winter climates is unknown.

The effects of these hard freezes do not seem to be as severe on native

wildlife and plants as they are on agricultural crops, with one major

exception. Populations of manatees (Trichechus manatus) in northern and

central peninsular Florida suffer high mortality rates from cold stress and

related causes during extended periods of freezing weather (T.J. O'Shea, USFWS,

pers. comm.; see O'Shea et al., 1985). Although the artesian springheads of

the St. Johns River valley create warm water refuges (72 degrees F.) for

manatees, the animals must leave areas of warmer water to feed. As food plants

become increasingly scarce near warm-water refuges, manatees are forced to

74

travel farther and farther in search of food, increasing their exposure to

colder water temperatures and the risk of collisions with boats (see Hartman,

1979). Radio-tagged manatees from Blue Spring on the St. Johns River have been

sighted in the Wekiva River, and residents have reported seeing manatees in the

Wekiva (Beeler and O'Shea, 1978).

Solar insolation is generally high in central Florida. Clear skies

predominate throughout much of the year, but morning ground fogs and rain

clouds often reduce direct sunlight penetration for brief periods. The

relatively high insolation and temperature regimes characteristic of Florida

combine to produce high evapotranspiration rates. In peninsular Florida,

surface water and groundwater losses to evaporation and transpiration may

approach or exceed the amount of water gained from precipitation (Heath and

Conover, 1981). The potential for such a water deficit demonstrates the

importance of springflows from the Floridan Aquifer to the maintenance of

surface-water levels in the Wekiva Basin. The preservation of existing wetland

ecosystems in the central and upper Wekiva River Basin is absolutely dependent

on continued supply of springflow waters at present levels (DNR, 1987).

III.A.2. Biological Description

Ecological Communities. The Wekiva Basin is a unique natural ecosystem.

It lies in the zone where tropical and temperate floras overlap. Tropical

plant species at the northern limits of their range grow side by side with

temperate-zone plants (DER, 1983). The distribution of plant species in

various plant communities varies according to specific site characteristics,

and species typical of one community may frequently occur in another (see Hart,

1984). Some species have fairly rigid microhabitat requirements, while others

occur in wide ranges of soil, shade, and moisture conditions. The soil

characteristics and disturbance history of a given area may result in the

anomalous presence or absence of species or species groups. A knowledge of

prior disturbance history (fires, logging, farming, ranching, etc.) is

important in determining the extent to which present vegetation conditions

reflect native vegetation patterns or the presence of human-altered plant

communities (Noss, 1985). For example, selective logging has greatly reduced

the abundance of cypress trees in the mixed hardwood swamp community of Rock

Springs Run State Reserve (DNR, 1985).

Rigid separations among plant communities do not typically occur within

most natural biotas. The distinct edges (abrupt ecotones) associated with

fire-maintained communities and human-altered landscapes are exceptions. For

the most part, categories for community types are general constructs devised

for analysis and identification of gross habitat characteristics. Despite the

overlap between community types, these recognized associations reflect

important biological differences among various communities and special

interactions or interdependencies among the constituent species of each

community.

The flows of numerous artesian springs from the Floridan Aquifer, together

with groundwater drainage from the surrounding watershed, have created the vast

network of stream channels and associated floodplains, lakes, and sinkholes

which support extensive areas of hydric and mesic habitats in the Wekiva River

Basin. The basin's landscape is dominated by wetland and lowland plant

communities but is interdigitated by relatively limited and patchily

distributed areas of transitional and upland habitats.

Deep-water and shallow-water herbaceous marsh communities inhabit the

stream channels and flooded regions, grading into the hardwood swamp vegetation

which also occurs on permanently flooded sites. Bottomland hardwood

76

communities (swamps and bayheads) grade into hydric hammocks where groundwater

seeps riverward from upland communities and into mesic hammocks and flatwoods

in areas of higher elevation. Drier upland communities in the region include

pine sandhill, xeric hammock, and sand pine/oak scrub. Sinks and ponded areas

support cypress domes and patches of wet prairie and mesic hammock (DNR, 1985,

1987; Brown and Starnes, 1983).

Wildlife. Many of the Wekiva Basin's wildlife species are locally

ubiquitous mammals which are common throughout the region or not restricted to

one specific plant community or habitat type. These include black bear (Ursus

americanus), white-tailed deer (Odocoileus virginiana), feral hog (Sus scrofa),

bobcat (Felis rufus), river otter (Lutra canadensis), striped skunk (Mephitis

mephitis), opossum (Didelphis marsupialis), raccoon (Procyon lotor), two

shrews, several bats, cotton mouse (Peromyscus gossypinus), wood rat (Neotoma

floridiana), and two cottontail rabbits. Most of the ubiquitous species of the

Wekiva Basin frequent swamps, and other species--including wood duck (Aix

sponsa), pileated woodpecker (Dryocopus pileatus), and many songbirds--require

forested wetlands for nesting and/or foraging habitats.

Many of these species require large expanses of diversified habitat

mosaics. For example, black bears, sandhill cranes (Grus canadensis), and wild

turkey (Meliagris gallipavo) need between-habitat diversity (presence of a

patchwork of different habitats) within their home ranges to support the

extensive daily or seasonal shifts in activity among various plant communities.

Species such as bats, pileated woodpeckers, and water turtles, which typically

forage and live within one habitat, may utilize quite different habitats for

dispersal, reproduction, or hibernation. Other species, such as river otters,

have large home ranges and travel widely among various habitat patches, even

though most of their activities are conducted in one habitat type (in this

case, aquatic wetlands).

In describing the faunal communities of the Wekiva Basin, the focus'will

be upon typical species associations within given habitat types. As noted in

reference to plant communities, these associations are, for the most part,

flexible. Overlaps and omissions in the presence of species within community

types may occur. However, the upland communities of the Wekiva Basin are

exceptional in having several characteristic and interdependent commensall)

species which are entirely restricted to specific upland habitats. The wetland

faunal communities include many species of restricted distribution (such as

fishes or the endemic snail and crayfish species), as well as a large

complement of more freely ranging forms which utilize several wetland habitat

types.

Aquatic wildlife communities are comprised of species which are largely or

entirely dependent upon aquatic ecosystems (rivers, lakes, springs, ponds:

places where standing or flowing water is present to some degree at any given

time) for at least part of their life cycle. These communities include animals

such as toads and tree frogs which spend their adult lives in terrestrial

habitats but which require aquatic habitats for breeding and the development of

young (larvae or tadpoles). Aquatic communities in the Wekiva Basin region

support a great diversity of wildlife, ranging from microscopic protozoans and

invertebrates to the massive American alligator. Literally hundreds of species

of snails, mussels, crayfish, insects, fishes, birds, amphibians, reptiles, and

mammals share aquatic habitats with a broad spectrum of aquatic plant life.

The aquatic plant community provides critical resources such as food, shelter,

and oxygen for animals in aquatic environments and is the base of the aquatic

food chain.

The aquatic plant communities in the rivers, springs, and floodplains of

the Wekiva Basin support populations of numerous fishes, including various

types of bass, sunfish, mullet, catfish, pickerel, shad, crappie, perch,

shiner, topminnow, killifish, molly, and bowfin (see appendix for scientific

names of species not detailed in this discussion).

The wetland herpetofauna of the region is exceptionally diverse and

contains many aquatic forms. Amphibians which are largely or entirely

dependent upon aquatic habitats include sirens, amphiumas, newts, bullfrogs,

pigfrogs, and leopard frogs. The largely aquatic reptiles of wetland habitats

in the region include American alligators, softshell turtles, cooters, mud

turtles, alligator snapping turtles, stinkpots, water moccasins, and several

non-poisonous water snakes. .In contrast to amphibians--which typically lay

their eggs in water--reptiles need terrestrial or semi-terrestrial sites in

which to incubate their eggs. Some reptile species (especially turtles) must

travel some distance from the water into adjacent upland habitats to find

patches of sunlit, sandy soil necessary for successful nesting.

Some of the invertebrates of the Wekiva Basin (including three endemic

snails of the genus Cincinnatia and the threatened Wekiva Springs Aphaostracon

snail) are restricted to springhead habitats. The threatened Orlando Cave

crayfish is an endemic form restricted to habitats in the Little Wekiva River

system. Other important invertebrates of the basin are apple snails and

burrowing crayfish. The apple snail is the primary food of the limpkin, a

large snipe-like bird found in wetland habitats that support populations of

this golf-ball-size snail. Burrowing crayfish spend part of the year in

vertical burrows which are dug in muddy soils in non-flooded wetland and

transitional zones. Excavated soil from the burrow is formed into a nearly

cylindrical cone rising as much as 6 inches above ground level. Crayfish

emerge from these burrows periodically to forage and breed. Crayfish are an

important food source for many larger vertebrates, including various fish,

bird, and mammal species.

Many species of birds found within the Wekiva Basin are dependent upon

aquatic and wetland habitats for foraging and/or nesting habitats. Osprey,

anhinga, wood stork, egrets, herons, ibises, bitterns, ducks, snipe,

gallinules, coot, limpkin, and grebes are characteristic of the open wetland

and aquatic habitats of the region. Bald eagles (USFWS: Endangered; FGFWFC:

Threatened) are expected to occur as transients in the aquatic communities of

the area even if no resident individuals are known at the present time. The

resident wood stork (Mycteria americana) has been classified as an endangered

species by both state and federal agencies (FGFWFC, 1987). Sandhill cranes

(threatened) utilize marsh, wetland, and upland sites in the Wekiva Basin.

Other resident species such as the little blue heron (Florida caerulea),

tricolored heron (Egretta tricolor), snowy egret (Egretta thula), and limpkin

(Aramus guarauna) have been designated as Species of Special Concern (SSC) by

the Florida Game and Fresh Water Fish Commission (FGFWFC, 1987). The wetlands

bird community of the Wekiva region is exceptionally diverse. Encroachment

upon wetland habitats will directly and indirectly affect these attractive and

often highly visible components of the Wekiva Basin wildlife community.

Typical mammals of the aquatic marsh communities in this region include

river otter, round-tailed muskrat (Neofiber alleni), raccoon, rice rat

(Oryzomys palustris), and marsh rabbit (Silvilagus palustris). Otters and

raccoons range widely among other habitats, while round-tailed muskrat, marsh

rabbit, and rice rat are usually associated with marsh, swamp, or wet prairie

communities.

The bottomland hardwood associations (hardwood swamp and bayhead

communities) typically occur as closed-canopy woodlands which are distributed

either as extensive wetland forests in flooded basins or as gallery forests

along stream and lake margins. These communities represent a transitional

gradient between strictly aquatic and terrestrial environments and retain

characteristics of both. Consequently, the wildlife associated with swamp and

bayhead communities includes species of both aquatic and terrestrial

affinities. The Wekiva Basin swamps and bottomland hardwoods support not only

numerous aquatic forms, but also many typically terrestrial animals, including

black bear, white-tailed deer, turkey, opossum, gray squirrel (Sciurus

carolinensis), cotton mouse, wood rat, gray fox (Urocyon cinereoargentus), and

striped skunk. The Wekiva Basin and the nearby southern portion of the Ocala

National Forest constitute perhaps the best remaining areas of suitable bear

habitat in central Florida (DER, 1983). The Florida black bear (Ursus

americanus floridanus) is unique among North America's extant black bear

populations in that its large size is commensurate with that of the ancestral

black bears of Ice Age times (Kurten and Anderson, 1980). Management

strategies keyed to protection of black bear populations and habitat

requirements should benefit virtually all other wildlife in the Wekiva Basin.

The largely terrestrial habitats of the uplands include pine flatwoods,

pine sandhills, and sand pine/oak scrubs. Flatwoods are used by bear, deer,

spotted skunk (Spilogale putorius), cotton rat (Sigmodon hispidus), cotton

mouse, diamondback rattlesnake (Crotalis adamanteus), and various other

ubiquitous species. Flatwoods provide good cover and shelter for animals that

forage in adjacent habitats, move between habitats, and seek refuge from high

floodwaters during storms. The truly upland habitats of the region are the

pine sandhill and sand pine/oak scrubs of the dry dune-ridges. Typical

81

wildlife species of these upland habitats include the gopher tortoise (Gopherus

polyphemus, SSC) and its many associated commensal (interdependent) species,

including the gopher frog (Rana sphenocephala, SSC), the Florida mouse

(Peromyscus floridanus, SSC), and the eastern diamondback rattlesnake. Pocket

gophers (Geomys pinetis; also known as salamander or sandy-mounder) occur in

more open habitats of deep, well-drained, sandy soils on high ridges in the

Wekiva Basin. Longleaf pine sandhills are the preferred habitat of the

Sherman's fox squirrel (Sciurus niger shermani, SSC). This species has been

decimated over the past century by logging of mature pines in the sandhill

communities. Reduced populations survive in remnant turkey oak forests, and

the species may occur around cypress domes in flatwoods and in a few other

upland ecotone situations (DER, 1983). The logging of mature longleaf pines

and conversion of sandhill communities into citrus and slash pine plantations

have severely affected the structure, composition, and distribution of sandhill

ecosystems. The preponderance of listed species among the typical wildlife of

the once widespread sandhill community is an indication of the extent to which

this habitat has been disrupted by development.

The sand pine/oak scrub community contains several highly specialized

endemic species such as scrub lizard (Sceloporus woodi), sand skink (Neoseps

reynoldsi, threatened), and Florida scrub jay (Aphelecoma coerulescens

coerulescens, threatened). Many sandhill wildlife species also utilize scrub

habitats and range freely between adjacent scrub and sandhill habitats. The

acorns of various oak species common in these habitats are an important

seasonal food for black bear, deer, turkey, and feral hog populations of the

Wekiva Basin.

82

III.B. Wetland Communities

Wetlands habitats of the Wekiva Basin can be subdivided loosely into three

major community types. These are the aquatic herbaceous marsh community of

stream channels and springheads, the mixed hardwood swamp community of the

river floodplains, and the hydric hammock community of groundwater and surface

runoff seepage wetlands. Marshes and swamps are characterized by saturated or

flooded soils which are subject to regular and extended periods of inundation

by standing or flowing water. Hydric hammock soils are saturated but only

occasionally inundated.

Marsh vegetation consists of submerged, emergent, and floating herbaceous

plants. Swamp vegetation consists of water-tolerant trees and shrubs. Gradual

shifts in plant species composition occur according to the regularity and depth

of immersion for ground-level vegetation in floodplain habitats. The presence

of surface water will greatly affect the types of wildlife species which are

present at any given time.

III.B.I. Aquatic/Marsh Communities

The freshwater marshes of Florida are critically endangered. Freshwater

marshes are vulnerable to changes in water levels, water quality, hydroperiod,

and fire regimes. Drainage of marshes for development and agriculture has

eliminated much of Florida's freshwater marsh habitat. Heavy recreational use,

particularly when associated with off-road vehicle and boat traffic, can

drastically alter the structure and composition of vegetation within marshes

and inhibit the vital ecosystem functions of these communities.

The freshwater marsh is a critical resource for many species of birds,

fishes, reptiles, mammals, and amphibians in the Wekiva Basin. Undisturbed

areas of freshwater marsh provide excellent cover and travel routes for many

83

wildlife species. The vegetation of freshwater marshes filters surface runoff

that feeds rivers and lakes. Soil particles are trapped within the vegetation

matrix, and plant roots stabilize the soil surface both above and below the

water level. Dense rafts and mats of marsh plants protect the stream banks

from strong current flows and help prevent erosion during periods of high water

and storm floods. Marsh plants fix and store nutrients and soil particles

filtering in from the surrounding landscape, thus helping to prevent

eutrophication.

The deepwater marsh communities (Water Hyacinch/Coontail/ Watergrass:

Eichhornia/Ceratophyllum/Echinochloa) are dominated by rooted or free-floating

aquatic herbaceous plants, and they typically occur on permanently flooded

sites with water depths of 3 to 6 ft (Brown and Starnes, 1983). This community

is found in the main channels and deeper sections of spring runs in the Wekiva

Basin, with maximum development in sites not shaded by streamside tree canopy.

These high-gradient, fast-moving streams tend to flush out the organic debris

and compounds which stain the waters of blackwater swamp systems. High flow

rates and steep gradients also create clearwater stream conditions within large

tracts of hardwood swamps.

The large springflows of the upper Wekiva River system generate clearwater

alkaline streams with high flow rates and clean, sandy bottoms. These

calcareous springflows are rich in phosphorous and other nutrients and support

luxuriant aquatic plant growth. Other representative plants of this habitat

include water lily (Nymphaea odorata), hydrilla (Hydrilla verticullata),

water-lettuce (Pistia stratioides), and duckweed (Lemna spp.). The common

water hyacinth (Eichhornia crassipes) is a weedy exotic which has invaded

aquatic habitats throughout Florida. This species has become the dominant

aquatic plant in many localities within the region, severely encroaching upon

84

important native species such as eelgrass and coontail. Elimination of water

hyacinth is a desirable but presently unachievable management goal.

Shallow-water marsh communities (Cutgrass/Maidencane/Cattail/ Arrowhead:

Leersia/Panicum/Typha/Sagittaria) typically occur in sites which may be flooded

only seasonally but which have water depths of 6 inches or more during the

growing season (Brown and Starnes, 1983). Other representative species include

sawgrass (Cladium jamaicense), sedges (Carex spp.), and rushes (Juncus spp.,

Elocharis spp., and Rynchospora spp.) These shallow-water herbaceous plant

associations typically occur in the shallower sections of stream channels and

along stream margins within the Wekiva Basin.

Wet prairie communities are similar to shallow marsh communities in terms

of both their constituent species and flooding regimes. Water depths range

from 0.5 to 2 ft in the growing season. Sometimes called grassy ponds,

sloughs, or prairie lakes (DNR, 1985), the wet prairie is a grass/herb plant

community associated with seasonally and permanently flooded areas of upland

habitats (see below for a more detailed discussion of this community).

III.B.2. Mixed Hardwood Swamp Community

The shallower waters in floodplains and along the margins of lakes and

streams in the Wekiva Basin support large tracts of swamp and bottomland

hardwood vegetation. This community is comprised of cypress species and a

variety of deciduous hardwood trees that grow on saturated and flooded soils.

Nutrient availability is generally low, but wildlife habitat value is high.

Swamps assimilate organic and inorganic wastes and pollutants, store water, and

impede the movement of floodwaters. By slowing down the rate of water flow,

swamps help to minimize erosion, promote infiltration, and facilitate the

settling out of debris, silt, and nutrients carried by floodwaters. These

functions improve water quality, stabilize water flows in streams and rivers,

and help to maintain high groundwater levels in the region. Swamps provide

habitat for many wildlife species and are used as travel corridors by wildlife

in developed areas.

The mixed hardwood swamp (cypress/tupelo/water ash: Taxodium/

Nyssa/Fraxinus) and bayhead (red bay/loblolly bay/pond cypress:

Persea/Gordonia/Taxodium) community occurs on poorly drained sites which are

regularly or permanently flooded (Brown and Starnes, 1983). Tannins and humic

acids from plant debris color and acidify the waters in swamps and in their

drainage streams. These blackwater systems tend to occur in the low-gradient

streams and basins of cypress and hardwood swamps where organic debris

accumulates on stream bottoms and in backwaters. Blackwater Creek is a typical

example of this distinctive stream type.

Pure stands of cypress stands and old cypress trees are no longer common

in the swamps of the Wekiva Basin. Old-growth cypress has been selectively

logged from swamps throughout Florida during the past century (see DNR, 1985;

Monk, 1968). Fluctuations in water level are required for cypress

regeneration: cypress seeds must be soaked before they will germinate, but

they will not germinate if they are underwater. Other important species of the

mixed hardwood swamp are red maple (Acer rubrum), American elm (Ulmus

americana), water hickory (Carva aquatica), buttonbush (Cephalanthus

occidentalis), and bluestem palmetto (Sabal minor) (Brown and Starnes, 1983).

The American elm was formerly common and widespread throughout eastern and

central North America, but in the past century the introduction of Dutch Elm

disease, an exotic pathogen, has nearly exterminated this species throughout

much of its range (see Davis, 1981). However, this species still survives as a

common element in the floodplain forests of the Wekiva Basin.

III.B.3. Hydric Hammock Community

In some areas between the mixed bottomland hardwoods and the upland

communities along the Wekiva River, seepage of groundwater from uplands is the

dominant source of water. The topography of these areas is flat, and the soils

are poorly drained and almost constantly saturated with seepage water.

Although subject to occasional flooding, the hydroperiod in hydric hammocks is

shorter than in the adjacent swamps.

Hydric hammock communities support more plant species than any other

wetland type in the Wekiva Basin. Among the species are popash (Fraxinus

caroliniana), live oak, laurel oak (Quercus laurifolia), red maple, southern

magnolia (Magnolia virginiana), red cedar (Juniperus siliciola), cabbage palm,

saw palmetto (Serenoa repens), tulip poplar (Liriodendron tulipfera), pond pine

(Pinus serotina), slash pine (Pinus elliotii), sweet gum (Liquidambar

styraciflua) wax myrtle (Myrica cerifera), and several vines and ferns (Brown

and Starnes, 1983). The hydric hammock is excellent wildlife habitat and is

used by ubiquitous mammals and birds as well as by numerous reptiles and

amphibians.

This community is very sensitive and vulnerable to changes in groundwater

hydrology in upland areas. Drawdowns caused by nearby wells and diversion of

groundwater via retention ponds and ditches in uplands can interrupt the supply

of seepage water to hydric hammocks, causing the soil to dry out and,

eventually, species composition to shift toward more mesic and fewer wetland

species. Lowering the water table also makes hydric hammock more vulnerable to

fire, a disturbance to which these ecosystems are not adapted.

III.C. Transitional Communities: Mesic Hammock and Scrubby Flatwoods

In the Wekiva Basin, the transitional communities between the wetlands and

the uplands are largely ecotonal in character, occurring as edges or borders

around the larger areas of swamp. Distinct separations between the swamp,

bayhead, and lowland hammock communities occupying floodplain habitats of the

Wekiva Basin are typically absent. Transitional species of the wetland/upland

ecotone include hydric and mesic hammock forms such as cabbage palm, saw

palmetto, tulip poplar, pond pine, slash pine, and sweet gum (Brown and

Starnes, 1983).

Scrubby flatwoods associations form ecotones between the mesic flatwoods

and the xeric sandhill and scrub communities. Scrubby flatwoods occupy sites

which are sufficiently well-drained so that there is no standing water present

even under extremely wet conditions. Pines (slash, sand, and longleaf) are

typically present but scattered within a matrix of shrubby oaks and palmettos.

Scrub oak (Q. inopina), Chapman's oak (Q. chapmanii), and sand live oak

intermix with saw palmetto and scrub palmetto (Sabal etonia) to form a

generally thick shrub layer 1 to 2 m high. Herbaceous cover tends to be

sparse. Lichens (Cladonia spp.) and spike moss (Selaginella arenicola) provide

considerable ground cover except on recently burned sites. Fire is important

in determining the structure and composition of scrubby flatwoods vegetation,

which is intermediate in character between flatwoods and sandhill/scrub

communities.

III.D. Upland Communities

The drier, sandy soils of the uplands support scrubby flatwoods on their

lower edges and longleaf pine sandhill and sand pine scrub on better-drained

and slightly higher sites. Longleaf pine-turkey oak sandhills and sand

88

pine-oak scrub are distinct in structure and composition from flatwoods

communities. These communities occur as savanna or scrub formations on the

well-drained upland soils of the Wekiva Basin and are subject to frequent

burning. Differences in past burning regimes and. local soil characteristics

strongly affect the particular vegetation characteristics of these communities

(see Kalisz and Stone, 1984). These habitats are home to a number of important

endemic species and are particularly sensitive to changes in natural

disturbance regimes (see Monk, 1968).

Upland habitats have been severely disturbed by past and present

management practices, and major ecological components cannot become

reestablished following certain types of human-caused disturbances (Means and

Grow, 1985). In many respects, these specialized uplands communities are at

the greatest risk from development at present (see Means and Grow, 1985). They

occupy substrates and soils which are often preferred over those of wetland

sites for agriculture, silviculture, industry, and residential development. As

a consequence, few large tracts of native upland vegetation and wildlife

survive at the present time. Those upland communities which remain have been

subjected to major changes in vegetation regimes and wildlife species diversity

as the direct result of human disturbance. Such human-caused changes in the

ecology of upland terrestrial communities can greatly alter the movement of

water, soil, and nutrients into adjacent wetlands and regional watersheds (see

Odum, 1971, 1983).

III.D.1. Pine Flatwoods Communities

Flatwoods are fire-adapted communities which occur on moderately to poorly

drained soils of terraced lands above the floodplain in areas subject to

periodic burning. Wildfires tend to kill off most typical hardwood hammock

tree species while leaving the fire-tolerant forms, especially pines and

palmetto, as local dominants. Exclusion of fire will result in succession to

bottomland hardwoods or mesic hammock on most flatwoods sites (Monk, 1968).

The dominant species in the basin's mesic pine flatwoods community are

pines, scattered oaks, and saw palmetto (Pinus spp./quercus spp./Serenoa

repens). A dense shrub layer carpets the understory level. Pines of flatwoods

habitats include longleaf pine, slash pine, and pond pine. Live oak and dwarf

live oak (_. minima) are often present. Saw palmetto is usually the dominant

shrub species, with dwarf live oak, fetterbush (Lyonia lucida), staggerbush

(Lyonia fruticosa), dwarf huckleberry (Gaylussacia dumosa), and wax myrtle as

secondary components. Saw palmetto, threeawn grass (Aristida patula),

greenbriar (S~ilax spp.), and gallberry (Ilex glabra) are typical understory

species.

Flatwoods often contain patchily distributed areas of cypress dome,

bayhead, mesic hammock, wet prairie, sandhill, and scrub vegetation due to

variations in topography, soils, and soil moisture regimes. Standing water may

be present in flatwoods during periods of high precipitation and elevated

groundwater levels.

Longleaf pine was formerly dominant on drier sites, while slash pine and

pond pine tend to be dominant on more mesic soils. Repeated selective

harvesting of longleaf pines, changes in fire regimes, and conversion of

flatwoods and sandhills to agriculture during the past century have greatly

affected the distribution and abundance of native longleaf pine communities

(Monk, 1965, 1968; Means and Grow, 1985). Slash pine is now the dominant tree

on many former longleaf sites. Pond pine becomes dominant in acidic, poorly

drained flatwoods where soil pH is less than 4.5, while slash pine occurs

primarily on more neutral soils (Monk, 1968).

III.D.2. Wet Prairie Communities

Wet prairies typically occur in sinks and depressions within areas of low

topographic relief which receive water through runoff from adjacent higher

ground. Patches.of wet prairie vegetation are commonly found scattered

throughout the flatwoods and sandhills of the region. Shrubs and small trees,

such as primrose willow (Ludwigia spp.) and elderberry (Sambucus simpsonii),

are sometimes present in wet prairies. Characteristic herbaceous plants

include panicum grasses (Panicum tenerum, P. dichotomum, etc.), sloughgrass

(Scleria spp.), swamp lily (Crinum americanum), and sundew (Drosera spp.). The

wet prairie is fire-adapted, and frequent burning will maintain wet prairies on

sites which would otherwise succeed to hydric or mesic woodland. These grassy

patches increase habitat diversity in flatwoods landscapes and provide

accessory foraging habitats for important wetland wildlife species. Because

wet prairies are dependent on surface water runoff to supply water (Brown and

Starnes, 1983), drawdowns of water tables in flatwoods by ditching can

eliminate wet prairies and reduce habitat diversity in upland regions.

III.D.3. Pine Sandhill Communities

The pine sandhill community is a savanna-type formation of scattered trees

and open woodlands having a well-developed ground cover of grasses, forbs, and

scattered shrubs. Characteristic species are longleaf pine, turkey oak, and

wirtgrass (Pinus palustris/Quercus laevis/Aristida stricta). Fire is an

important ecological component of sandhill environments and maintains the

characteristic open pine woodland and sparse understory vegetation. Burns

occur naturally in sandhill communities at about 3- to 5-year intervals,

usually started by lightning strikes (Means and Grow, 1985). These frequent

91

fires preserve the dominance of longleaf pine and inhibit succession to xeric

hammock vegetation (Kalisz and Stone, 1984). Sand pine may also invade

sandhill communities when fire is suppressed (Kalisz and Stone, 1984).

The upland dune formations of the Wekiva Basin are dominated by pine

sandhill communities containing patchily distributed areas of sand pine/oak

scrub (Pinus clausa/Quercus spp.) and xeric hammock. Pines and turkey oaks are

typically the principal overstory components. The relative dominance of turkey

oaks and pine species other than longleaf and the absence of old-growth

longleaf in the region's sandhill communities are due to selective logging in

these communities.

Sandhills have a more or less continuous herbaceous ground cover dominated

by wiregrass. Shrubs and hardwoods other than oaks are typically present at

low densities (Laessle, 1958). Other common plants of sandhill habitats are

bluestem grasses (Andropogon stolonifera, A. tenarius), persimmon (Diospyros

virginiana), prickly pear (Opuntia ammophila), blueberries (Vaccinium spp.),

gopher apple (Chrysobalnus oblongifolius), and spike moss. Shrub rosemary

(Ceratiola ericoides) is often common in sandhill communities but does not

usually occur in dense stands. Individual rosemary bushes are usually ringed

by narrow perimeters of bare soil. This phenomenon is caused by the release of

an allelopathic chemical from the rosemary which inhibits the growth of other

plants within its immediate vicinity.

The high primary productivity of sandhill habitats provides a rich food

base for a wide variety of wildlife. Turkey oak acorns are important to

mast-feeders such as deer, turkey, and feral hog. Sandhill communities support

a characteristic fauna which includes Sherman's fox squirrel and gopher

tortoise with its host of commensal wildlife species.

Sand pine scrub and oak scrub occur on various upland sites within the

Wekiva Basin (DNR 1985, 1987). Sand pine and oak scrub are characterized by a

mixed scrub oak understory dominated by scrub live oak, myrtle oak (Q.

myrtifolia), and Chapman's oak (Q. chapmanii). Sand pine is variably present

in the scrub habitats of the Wekiva Basin. Differences in burning and

disturbance regimes probably account for the absence of sand pine in some scrub

communities of the Rock Springs Run State Preserve (DNR, 1985). Shrubs such as

staggerbush, rosemary, and wild olive (Osmanthus americana) may also be

present. Scrub hickory (Carya floridana), a Florida endemic, is present in the

scrub habitats of the Wekiva Basin. Common herbaceous plants of scrub habitats

in the region include blueberries, gopher apple (Chrysobalnus oblongifolius),

prickly pear (Opuntia compressa), spike moss, and wiregrass. Lichens (Cladonia

spp., Usnea spp., and others) grow on patches of bare soil or as epiphytes on

the bark of woody plants.

Patches of xeric hammock (live oak/sand live oak/red oak/American

holly/deer moss: Quercus virginiana/Q. geminata/Q. falcata/Ilex opaca/Cladonia

spp.) occur within the upland sandhill habitats. Pine sandhill and sand pine

scrub communities will be succeeded by xeric hammock vegetation if the absence

of fire is prolonged (Monk, 1968). Exclusion of fire allows regeneration of

hardwoods and prevents the continued recruitment of pines. The leaf litter of

xeric hammock plants is fire-resistant, so that patches of hammocks within

large sandhill tracts typically remain unburned even though ground fires have

spread throughout the surrounding pine-dominated landscape. The prevalence of

natural and human-caused wildfires in the Wekiva Basin over the past centuries

has limited the significance of this community type in the region.

Sand pine scrub and oak scrub occur on various upland sites within the

Wekiva Basin (DNR 1985, 1987). Sand pine and oak scrub are characterized by a

mixed scrub oak understory dominated by scrub live oak, myrtle oak (q.

myrtifolia), and Chapman's oak (Q. chapmanii). Sand pine is variably present

in the scrub habitats of the Wekiva Basin. Differences in burning and

disturbance regimes probably account for the absence of sand pine in some scrub

communities of the Rock Springs Run State Preserve (DNR, 1985). Shrubs such as

staggerbush, rosemary, and wild olive (Osmanthus americana) may also be

present. Scrub hickory (Carya floridana), a Florida endemic, is present in the

scrub habitats of the Wekiva Basin. Common herbaceous plants of scrub habitats

in the region include blueberries, gopher apple (Chrysobalnus oblongifolius),

prickly pear (Opuntia compressa), spike moss, and wiregrass. Lichens (Cladonia

spp., Usnea spp., and others) grow on patches of bare soil or as epiphytes on

the bark of woody plants.

Patches of xeric hammock (live oak/sand live oak/red oak/American

holly/deer moss: Quercus virginiana/g.. geminata/Q. falcata/Ilex opaca/Cladonia

spp.) occur within the upland sandhill habitats. Pine sandhill and sand pine

scrub communities will be succeeded by xeric hammock vegetation if the absence

of fire is prolonged (Monk, 1968). Exclusion of fire allows regeneration of

hardwoods and prevents the continued recruitment of pines (see Williamson and

Black, 1981). The leaf litter of xeric hammock plants is fire-resistant, so

that patches of hammocks within large sandhill tracts typically remain unburned

even though ground fires have spread throughout the surrounding pine-dominated

landscape. The prevalence of natural and human-caused wildfires in the Wekiva

Basin over the past centuries has limited the significance of this community

type in the region.

IV. STATUTORY & DISTRICT CRITERIA RELATED TO BUFFER ZONES

Questions have been raised concerning the authority of the District to

enact requirements which restrict or preclude the construction and operation of

systems within the Wekiva River Basin. The rulemaking related to the Wekiva

Basin has been the subject of serious concern, and much debate has resulted.

An earlier draft of the Wekiva Basin Rule had requirements that buffered the

river from impacts, but was withdrawn following considerable controversy, and

staff was directed to revisit the issue. Following several drafts of proposed

language for buffer requirements, a Petition for Administrative Determination

of Invalidity of Proposed Rule was filed that alleged the proposed rule

constitutes an invalid exercise of delegated legislative authority or is

otherwise invalid.

The statutory authority granted the Governing Board of the District

related to the enactment of requirements that restrict or preclude the

construction, operation, or maintenance of systems is clearly stated, but

somewhat general. Broad rulemaking authority has been granted by the

Legislature (Chapter 373, FS). The Board may adopt reasonable rules that are

consistent with the law and reasonably necessary to efftecuate its powers,

duties, and functions. Rules adopted by the Board must be reasonably related

to the purposes of Chapter 373. The Board cannot act in an arbitrary and

capricious manner and should have before it competent and substantial evidence

to support a proposed rule.

Two questions related to the statutory authority given in Chapter 373 (FS)

have occurred to us. First, in Sections 373.413 and 373.416 the Legislature

enacted two statutory provisions related to water management systems; the

former addressing construction of systems and the later addressing operation

and maintenance of systems. As a result, an arbitrary distinction is made

between construction and operation and maintenance. Second, the statute does

not provide a definition for the term "water resources."

IV.A. Statutory Criteria, Sections 373.413 & 373.416, FS

Sections 373.413 and 373.416 draw a distinction between permits

required for construction or alteration of dams, impoundments, reservoirs, or

works (373.413, FS) and permits required for maintenance and operation of same

(373.416, FS). In the former, the District may require permits to assure that

construction or alteration of any system will not be harmful to the water

resources of the District. In the latter, the District may require permits to

assure that the operation or maintenance of a system will not be harmful to the

water resources of the District and will not be inconsistent with the overall

objectives of the District. The distinction between the harm that may result

from construction activities and harm that may arise from operation and

maintenance is a confusing one in light of the review a permit application may

receive. Can the District evaluate an application for construction based only

on harm to water resources, or can the application also be evaluated relative

to the overall objectives of the District?

Conceptually it may be relatively easy to separate impacts on water

resources that result from development activities into those that occur during

construction and those that may result from the operation and maintenance of

that which has been constructed. From a resource management perspective,

MILLISECOND	CLASS.METHOD	MESSAGE
0	sobekcm_page_globals.constructor
0	sobekcm_page_globals.constructor	Application State validated or built
0	sobekcm_database.verify_item_lookup_object
0	sobekcm_page_globals.constructor	Navigation Object created from URI query string
0	sobekcm_database.verify_item_lookup_object
0	sobekcm_page_globals.display_item	Retrieving item or group information
0	sobekcm_page_globals.get_entire_collection_hierarchy	Retrieving hierarchy information
0	sobekcm_assistant.get_entire_collection_hierarchy
0	cached_data_manager.retrieve_item_aggregation
0	cached_data_manager.retrieve_item_aggregation	Found item aggregation on local cache
0	item_aggregation_builder.get_item_aggregation	Found 'all' item aggregation in cache
0	system.web.ui.page.page_load (ufdc.page_load)
0	sobekcm_page_globals.constructor.on_page_load
0	html_echo_mainwriter.add_style_references	Adding style references to HTML
0	html_echo_mainwriter.add_text_to_page	Reading the text from the file and echoing back to the output stream
45	html_echo_mainwriter.add_text_to_page	Finished reading and writing the file