AnEvaluation of refuge habitats and relationship to water quality, quantity and hydroperiod

An Evaluation of Refuge
Habitats and Relationships
to Water Quality, Quantity,
and Hydroperiod

Synthesis Report

November 1990

by
John R. Richardson
Wade L. Bryant
Wiley M. Kitchens
Jennifer E. Mattson
Kevin R. Pope

_

An Evaluation of Refuge Habitats and Relationship
to Water Quality, Quantity and Hydroperiod

A Synthesis Report

Prepared for

Arthur R. Marshall Loxahatchee National Wildlife Refuge
Boynton Beach, Florida

Prepared by

Florida Cooperative Fish and Wildlife
University of Florida
Gainesville, Florida

Authors:
John R. Richardson
Wade L. Bryant
Wiley M. Kitchens
Jennifer E. Mattson
Kevin R. Pope

Research Unit

November 1990

TABLE OF CONTENTS
TABLE OF CONTENTS........................................ i
LIST OF TABLES......................................... ii
LIST OF FIGURES....................................... iii
INTRODUCTION..... .......... ............................. 1
HISTORICAL PERSPECTIVE AND LITERATURE REVIEW ............ 6
Everglades System Overview............................. 6
Pre-drainage Hydrology ................................ 12
Plant community descriptions........................... 14
Changes in Everglades system........................... 18
LOXAHATCHEE NWR STUDY .................................. 33
Historic Vegetation Mapping and Surveys ............. 34
Classified Vegetation Map from SPOT Imagery ........... 38
Intensive Vegetation Survey............................ 47
Nutrient/ion Distribution............................. 51
Water Quality Maps ................................... 70
Grid Survey/Topographic map............................ 91
Hydrology.............................................. 94
Water Surface Profiles................................. 104
Hydrological Simulation Model.......................... 107
ANALYSIS ................................................ 121
Vegetation Characterization
With Classified Satellite Map....................... 121
Vegetation Changes................................... 131
Historic Transect..................................... 133
Vegetation Communities and Distance from Canal........ 137
Multivariate Analysis of Species Data ................. 139
SUMMARY AND DISCUSSION... .............................. 150
Water Quantity........................................ 150
Hydroperiod.................................................. 151
Water Quality .................................. ........ 152
Vegetation Changes............ ...................... 153
Multivariate Relationships............................ 153
Future Research Questions.............................. 155
Outlook. ................................. .............. 157
BIBLIOGRAPHY... ............................ ............. 158
APPENDIX.... ............. ................................ 167

LIST OF TABLES

Table 1. Vegetation types found in the Everglades and other
freshwater marshes listed by Davis (1943)............... 16
Table 2. Relative charcoal content of some peat types of South-
ern Florida. (from Cohen 1984).......................... 17
Table 3. Generalized 1987/1988 Land Use/Land Cover Types and
Acreages by Watershed for major Sub-basins
within the EAA............ .............................. 21
Table 4. Acreage and Values of Agricultural Production in the
Everglades Agricultural Area............................ 22
Table 5. Water control structures associated with Loxahatchee NWR.24
Table 6. Average annual water and nutrient budgets (1979-1988)
for 3 WCAs............................................... 29
Table 7. Comparison of Marsh Surface Area, Total Phosphorus
Loading and Areal Loading Rates, Percent Retention
and Water Residence Times Among the Three Water
Conservation Areas...................................... 30
Table 8. Vegetation communities in supervised classification..... 40
Table 9. Signature euclidian distance............................ 42
Table 10. Signature statistics listing............................ 43
Table 11. Plant species list ( > 5% cover)........................ 50
Table 12. Areal loading rates of three WCAa and Lake Okeechobee... 66
Table 13. Phosphorus, nitrogen and water budget (1979-1988)
for Loxahatchee NWR..................................... 67
Table 14. Wet and dry season averages for water quality
site groups.............................................. 72
Table 15. Means of vegetation classes from map variables.......... 122
Table 16. Summary of historic vegetation photoplot data
with 1987 classified map data.......................... 134
Table 17. Summary of classified vegetation map data for photo-
plot areas and historic transects....................... 135
Table 18. Summary of historic vegetation transect data with
comparisons to 1987 classified vegetation map data..... 136
Table 19. Vegetation distribution by zones from canal............ 138
Table 20. Averages weighted by percent cover for selected species 141

LIST OF FIGURES

Figure 1. Physiographic provinces of Southern Florida............ 7
Figure 2. Surficial geology of Southern Florida exclusive of
organic materials ...................................... 8
Figure 3. Surficial geology of Southern Florida exclusive of
organic materials and sand ............................ 9
Figure 4. Elevation contours (Lake Okeechobee datum) on bedrock
in the Everglades...................................... 11
Figure 5. Everglades map showing direction of surficial drainage. 13
Figure 6. Schematic diagram of water movement through the SFWMD.. 19
Figure 7. Map of SFWMD showing EAA, canals, and pumpstations..... 20
Figure 8. Vegetation cover map for 1952.......................... 35
Figure 9. Vegetation cover map for 1968.......................... 36
Figure 10. Map showing location of vegetation photoplots, vege-
tation transects, and gaging stations .................. 37
Figure 11. Nearest-neighbor analysis of vegetation map classes.... 46
Figure 12. Map of grid survey and vegetation site locations........ 48
Figure 13. Map of water extractable ammonium from peat
sediment at Loxahatchee NWR............................ 52
Figure 14. Map of water extractable chloride from peat sediment
at Loxahatchee NWR..................................... 53
Figure 15. Map of water extractable nitrate from peat sediment at
Loxahatchee NWR........................................ 54
Figure 16. Map of Mehlich extractable phosphorus from peat
sediment at Loxahatchee NWR............................. 55
Figure 17. Map of Mehlich extractable magnesium from peat sedi-
ment at Loxahatchee NWR................................ 56
Figure 18. Map of Mehlich extractable aluminum from peat sediment
at Loxahatchee NWR ..................................... 57
Figure 19. Map of Mehlich extractable potassium from peat
sediment at Loxahatchee NWR............................ 58
Figure 20. Map of Mehlich extractable sodium from peat sediment
at Loxahatchee NWR..................................... 59
Figure 21. Map of Mehlich extractable calcium from peat sediment
at Loxahatchee NWR.................................. . 60
Figure 22. Map of Mehlich extractable zinc from peat sediment at
Loxahatchee NWR........................................ 61
Figure 23. Phosphorus inflow at S5A and S6 pump stations,
moving average 1974-1986............................... 63
Figure 24. Nitrogen inflow at S5A and S6 pump stations, moving
average 1974-1986 ...................................... 64
Figure 25. Atmospheric phosphorus loading (metric tonnes) and water
input (ten thousand acre feet), 1978-1988.............. 68
Figure 26. S5a and S6 pump stations phosphorus loading (metric
tonnes) and water input (ten thousand acre feet),
1979-1988 .............................................. 69
Figure 27. Map of water quality sampling sites ................... 71
Figure 28. Map of wet season specific conductivity in surface
water at Loxahatchee NWR.......... .................... 73
Figure 29. Map of dry season specific conductivity in surface
water at Loxahatchee NWR............................... 74
Figure 30. Map of wet season pH in surface water at Loxahatchee
NWR.................................................... 75

Figure 31. Map of dry season pH in surface water at Loxahatchee
NWR........... ......................................... 76
Figure 32. Map of wet season total phosphorus in surface water at
Loxahatchee NWR......................................... 77
Figure 33. Map of dry season total phosphorus in surface water at
Loxahatchee NWR ......................................... 78
Figure 34. Map of wet season sulfate in surface water at Loxa-
hatchee NWR............................................. 79
Figure 35. Map of dry season sulfate in surface water at Loxa-
hatchee NWR................ .......................... 80
Figure 36. Map of wet season chloride in surface water at Loxa-
hatchee NWR............................. ........... . 81
Figure 37. Map of dry season chloride in surface water at Loxa-
hatchee NWR............................................ 82
Figure 38. Map of wet season total nitrogen in surface water at
Loxahatchee NWR........................................ 83
Figure 39. Map of dry season total nitrogen in surface water at
Loxahatchee NWR ........................................ 84
Figure 40. Map of wet season alkalinity in surface water at
Loxahatchee NWR ........................................ 85
Figure 41. Map of dry season alkalinity in surface water at
Loxahatchee NWR....................................... 86
Figure 42. Topographic map of Loxahatchee NWR...................... 92
Figure 43. Contour Map of Loxahatchee NWR ......................... 93
Figure 44. Monthly rainfall input (1970-1984) ...................... 95
Figure 45. Average monthly rainfall, (1970-1984) .................. 96
Figure 46. Monthly S5A and S6 pump station input.................. 98
Figure 47. Water inflow-outflow (1970-1984) ........................ 99
Figure 48. Evapotranspiration by month ............................ 101
Figure 49. Stage regulation schedule for Loxahatchee NWR ......... 102
Figure 50. Average surface water elevation at three gaging
stations................................ ............. 103
Figure 51. Stage duration curves 1970-1985...................... 105
Figure 52. Surface water profiles on east-west hydrologic
transect............................................... 106
Figure 53. Hydrologic model grid cells........................... 109
Figure 54. Comparison of model and gage data for 1-7 gage........ 111
Figure 55. Comparison of model and gage data for 1-9 gage........ 112
Figure 56. Map of 16 year hydroperiod............................ 114
Figure 57. Map of 16 year growing season hydroperiod .............. 115
Figure 58. Map of 16 mean depth.................................. 116
Figure 59. Map of 16 year growing season mean depth.............. 117
Figure 60. Map of yearly hydroperiod variance ..................... 118
Figure 61. Means of vegetation classes for distance from canal
vs. wet season conductivity........................... 126
Figure 62. Means of vegetation classes for distance from canal
vs. peat extractable chloride ........................ 127
Figure 63. Means of vegetation classes for distance from canal
vs. peat extractable phosphorus...................... 128
Figure 64. Means of vegetation classes for 16 year hydroperiod
vs. distance from canal ............................. 129
Figure 65. Means of vegetation classes for 16 year hydroperiod
vs. 16 year mean depth................................. 130

Figure 66. Means of vegetation classes for 16 year hydroperiod
vs. peat extractable phosphorus....................... 132
Figure 67. Weighted species means of hydroperiod
vs. peat phosphorus.......................... ........ 142
Figure 68 Weighted species means of mean depth
vs. peat phosphorus.... .............................. 144
Figure 69. Weighted species means of hydroperiod vs.
peat nitrogen..... ................................... 145
Figure 70. Weighted species means of peat nitrogen
vs. peat phosphorus.... .............................. 146
Figure 71. Species vs. canonical discriminant axes................ 149

INTRODUCTION

The Arthur R. Marshall Loxahatchee National Wildlife Refuge,

(hereafter referred to as Loxahatchee NWR), includes 141,420

acres (57,234 ha.) of northern Everglades wetland. This land-

scape consists of a complex mosaic of wetland communities that

grade from wetter areas such as sloughs and wet prairies, to

sawgrass (Cladium jamaicense), brush, and finally tree islands

occurring at the dryer end of the scale. These communities occur

on top of a deep bed of peat. This complex spatial interaction

of plant communities provides habitat for wading birds, deer,

alligators and other species of wildlife.

Once part of an uninterrupted expanse of wetlands extending

from Lake Okeechobee to the tip of Florida, Loxahatchee NWR is

now a unique remnant which still functions as a northern refuge

for species of the Everglades ecosystem. The area was enclosed

by a canal and dike in the 1950's and 1960's, and is now sur-

rounded by drained agricultural land, the Everglades Agricultural

Area (EAA), and rapidly growing urban development. As one of

three water conservation areas that receive runoff in the wet

season and provide water during the dry season, the South Florida

Water Management District (SFWMD) uses Loxahatchee NWR as Water

Conservation Area 1 (WCA1). Since 1951, the U.S. Fish and Wild-

life Service has also administered the area as Loxahatchee NWR, a

migratory bird refuge (Van Arman et al. 1984).

Hydrologically, the refuge is isolated from the historic

Kissimmee/Okeechobee/Everglades watershed. Loxahatchee NWR now

functions as a receiving/storage basin for the drainage of a

comparable area of the EAA and is hydrologically driven by the

flows of the EAA and rainfall directly intercepted by the refuge.

In addition to hydrologic alterations, the quality of water is

affected by nutrients, mineral salts and contaminants associated

with the agricultural runoff entering the refuge at SFWMD pump

stations S5A and S6 into the perimeter canals of the refuge.

Important questions have been raised regarding the response

of the vegetative wetland habitats of Loxahatchee NWR to these

alterations in water quality and hydrology over the past 2 to 3

decades. One particular concern is the conversion of large areas

of the Loxahatchee NWR landscape to stands where cattail (Typha

domengensis) dominates the vegetation community replacing the

spatially complex native wetlands. Hydrologic changes in the

northern portion of the refuge may be causing vegetation changes

due to a reduced hydroperiod and in the southern portion of the

refuge longer hydroperiods have drowned vegetation communities.

In 1985, a study was begun by the Florida Coop Fish and

Wildlife Research Unit to investigate the changes taking place

and some of the probable causes. The objectives and the approach

of the study were intended to resolve some of the complex inter-

actions of the vegetative communities and the species that com-

prise them with the long-term and spatial hydrologic and water

quality impacts. The primary objectives of the study were:

1. Document the vegetation/habitat changes that have

occurred in the refuge.

2. Relate the changes in vegetation types to the water

quality, quantity and hydroperiod.

3. Determine the impacts of water management on

vegetation changes and subsequent wetland functions to

faunal species of particular interest to Loxahatchee NWR

4. To identify additional research needs, in particu-

lar nutrient cycling and acid rain effects.

5. To develop a comprehensive computer map-referenced

data base.

An added task for the project was to generate a summary

document of the research in context with the historical perspec-

tive of the Everglades and a review of existing literature con-

cerning the role of nutrients and hydroperiod in the development

of the Everglades.

The study design incorporated several phases to meet the

needs and objectives just described. An overall objective of the

study design was to use a methodology that would integrate all of

the spatial factors that were thought to structure the complex

ecosystem of Loxahatchee NWR. The study design had to match the

appropriate spatial and temporal scales of the hydrologic and

water quality events that were driving the vegetative responses.

An integrated Geographic Information System (GIS) approach was

used to categorize, analyze and relate the complex assemblage of

variables structuring the ecosystem.

The study began with a synthesis of existing water quality,

hydrologic, and vegetation data into appropriate formats for

spatial analysis. Data sets for water quality, hydrology, rain-

fall, and survey benchmarks were acquired from the South Florida

Water Management District (SFWMD). Next, a comprehensive grid

survey of Loxahatchee NWR was conducted to generate new data

concerning topography, hydrology, and general vegetation rela-

tionships to these factors. These data were then augmented with

an independent intensive vegetation analysis, gathering data con-

cerning sediment chemistry, vegetation community structure and

hydrology. Finally, a high resolution vegetation map was de-

veloped using satellite image processing. These databases were

stored in a GIS format.

One concern was the distribution of water over space and

time and how vegetation was being structured on the refuge by

hydroperiod patterns. Hydrologic data for Loxahatchee NWR had

been collected at only a select number of sites (two interior and

several canal gage stations). A hydrologic model was developed

with the ability to generate the needed spatial hydrologic param-

eters from rainfall, topographic and other data.

A vegetation gradient analysis was used to determine species

relationships to water quality and hydrologic parameters. Our

research effort is still ongoing for habitat analysis to deter-

mine the possible effects of vegetation changes on wading birds

and forage fish.

This report is divided into 7 major sections that parallel

the phases described above: (1) Background of the Everglades

system with special reference to Loxahatchee NWR and its unique-

ness; (2) Literature review and synthesis of existing water

quality and hydrologic data pertaining to the refuge; (3) Review

of previous vegetation studies on the refuge coupled with an

analysis documenting vegetation changes in the refuge; (4)

Description of topography, vegetation, and peat data gathered

during the grid survey and peat sampling phase; (5) Description

of the hydrology model and analysis of output from this model;

(6) Analysis and discussion of vegetation relationships to water

quality, soil chemistry, and hydrologic parameters; (7) Discus-

sion of recent questions concerning the refuge in context of this

study and of additional research needs.

HISTORICAL PERSPECTIVE AND LITERATURE REVIEW

Everglades System Overview

Geology and soils

The historic Everglades, a broad, shallow, peat-filled basin

(Figure 1), covered more than 2.41 million acres (975,000 ha,

Parker et al. 1955), and was the most extensive freshwater marsh

in North America (Hofstetter 1983). Commonly referred to as a

"river of grass" (Douglas 1947) six inches deep and 60 miles

wide, the historic Everglades extended from Lake Okeechobee south

to Florida Bay (Figure 1). The basin contains the largest single

body of organic soils in the world, with peat depths ranging from

less than 1 meter at the south end to over 3 m in the north

(Stephens 1956, 1984).

The geologic history of the Everglades area has been de-

scribed by Parker and Cooke (1944), Schroeder and Klein (1954)

and Gleason (1974). The eastern boundary consists of Pamlico

sands underlain by limestone (Anastasia formation and Miami

Oolite (Figures 2 and 3). The western boundary is a combination

of (1) Pamlico sands underlain by limestones of the Fort Thompson

and Caloosahatchee Formations and (2) limestone of the Tamiami

Formation that is not covered by sand (Figures 2 and 3). The

Everglades bedrock floor consists of limestone and limestone-marl

formations derived from a Pliocene sea floor (Parker and Cooke

1944). The bedrock formation names and locations are shown in

Figure 3. Lake Flirt Marl believed to have developed in part

from residue of the Fort Thompson Formation (White 1970, Gleason

et al 1974), overlies parts of the Fort Thompson Formation.

A
NJ

z
U

0

Figure 1.

Physiographic provinces
Parker and Cooke 1944).

V0 10 10
miles
of Southern Florida (from

I r. 1

L.i -1 -Lake Flirt Marl
M ARTIo.. I No
Jp LoNL Pamlico Sands

Okeochobe Miami Oolite

SAnastasia Formation
^ -v I Pipo cht z
01 ' b [ *
SE---ND / / PALM BEACH' . Key Largo Limestone
-- H ME ORY Op '
St / / Talbot Formation

N / Penholoway Formation

S .Ft. Thompson Formation

B" ROAROD Caloosahatchee Formation
-- "" " "* "Q*'!'i Z w
U Tamiami Formation

F -r-~- ^ -

MONROE f

ICO

ook

0 10 7C

Figure 2. Surficial geology of Southern Florida exclusive of
organic materials (from Parker and Cooke 1944).

Sr-

-7 ---- ----

to e
- -"P -- -MA -N
1-\
--- -- Pc
S Okeechobee --D ---

//LL iCPO .^
H---- \Y//o

nEND^- /

Z

m~jm

-J-,

~gL.

Lake Flirt Marl

Pamlico Sands

Miami Oolite

Anaatasia Formation

Key Largo Limestone

Talbot Formation

Penholoway Formation

Ft. Thompson Formation

Caloosahatchee Formation

Tamiami Formation

Buckingham Marl

'1 0 0 20
m77S

Figure 3. Surficial geology of Southern Florida exclusive of
organic materials and sand (from Parker and Cooke 1944).

The topography of the Everglades' bedrock floor (Figure 4)

is thought to be the result of solution of the limestone during

periods of low sea level (Parker and Cooke 1944). The three most

conspicuous features of the bedrock floor are the Loxahatchee

Channel, West Everglades High and the Shark River Slough. Glea-

son et al. (1974) described the Loxahatchee channel as "a narrow

elongate depression which may have functioned as an overflow

valve for Lake Okeechobee". This depression runs underneath

Loxahatchee NWR.

Based on the classifications of Davis (1946) and Gleason et

al. (1974), Wieland (1981) described seven soil types in the

northern Everglades. These types are: Everglades (sawgrass) peat,

Loxahatchee (aquatic slough) peat, Transition (wet prairie) peat,

Gandy (tree island) peat, Okeelanta Peaty-muck (elderberry-willow

forest), Okeechobee muck, and Sapropel. The first four soil types

are found in Loxahatchee NWR. Gleason et al. (1974) give an

excellent description of the nature and distribution of surface

sediments in the Everglades. Cohen and Spachman (1974) have

describe South Florida peats in terms of their plant tissue

composition and depositional environments.

The peat soils of the Everglades are of relatively recent

origin, the oldest formed approximately 5,000 years ago (Parker

et al. 1955, Stephens 1955, McDowell et al. 1969, Gleason et al.

1974, Cohen 1984). The deepest peats in the Everglades are

located over the Loxahatchee Channel (Stephens and Johnson 1951).

Data presented by Gleason et al. (1974) suggest that the deepest

peats may not be the oldest.

Hick

Figure 4. Elevation

A
NI

NI

z

uu
0J
U

contours

in the Everglades

(Lake Okeechobee datum) on bedrock
(from Parker and Cooke 1944).

11

~__ ___

McDowell et al. (1969) used radiocarbon dating to calculate

an average rate of peat deposition of 8.4 cm/century, and showed

that deposition rates have increased exponentially over time. A

rate of 16 cm/century was calculated for the last 1200 years

(McDowell et al. 1969). Data presented by Gleason et al. (1974)

for cores located in Loxahatchee NWR (Water Conservation Area 1)

and Water Conservation Area 2 show rates of approximately 7 and 4

cm/century respectively over the last 3000 to 4000 years. Net

deposition rates may vary considerably due to fires.

Pre-drainage Hydrology

The bedrock topography of the Everglades determined drainage

patterns before peat accumulation and apparently influenced

drainage patterns after peat accumulation (Gleason et al. 1974,

Parker 1974). The influence of bedrock topography on surficial

flow patterns after substantial peat accumulation can be seen by

comparing bedrock topography (Figure 4) and surficial drainage

patterns (Figure 5) as presented by Parker and Cooke (1944).

Gleason et al. (1974) and Gleason et al. (1977) have shown that

the tree islands in the northern Everglades, particularly evident

at Loxahatchee NWR, may be oriented with the flow of water.

Parker (1974) reviewed early accounts of pre-drainage

hydrologic conditions. Parker noted that no comprehensive study

of hydrology was made prior to 1940, but proposed a pre-drainage

scenario based on hydrologic data from 1940-1946, data from soil

profiles, data from over 300 USGS ground water monitoring wells,

and on personal experience. In addition to precipitation, the

Everglades "river of grass" received overflow water from Lake

j31 1 33 4 35 I. 37 8 3 9- 40 41
i* Nos KEEC

a 'wel w % L s
II .-- I
\ DE \/ -
lowAI I

i1 _T. 40

EE CHEE //4
t,-.%s 4'' \ 1.
-T

/ %;c
i t O 44
FEs map in r o n o surf icialda

r GARDEr n C o 1 9

'll \ n1 l oJ \ "

Na,.... --4 2, / -,
400 A

, ,,*' ,, .o. -
I N,," 1 45

'I 7 -,

a ( Pr and-" Coke\ / 4
QA I li 11 1

Ave I 1. 14
L-r \1 \_\\ 2 t % 0 49
'T f r r A 4, I -* 4 /i7
^E^ --

age (From Parker and Cooke, 1944).
i r rr r rI II rd`

~c: o1 000 me ~

41000 40 3

age (From Parker and Cooke, 1944).

13

Okeechobee. According to Parker (1984), the entire southern

shore of Lake Okeechobee flooded the northern Everglades when

lake levels exceeded 18 ft msl. Water flow south was slow due to

the extremely low topographic gradient and the resistance of the

dense vegetation. Slow flow coupled with high evapotranspiration

rates make it doubtful that a significant amount of overflow

water from Lake Okeechobee made it through the Everglades (Parker

1974).

Plant community descriptions

A more detailed description of vegetation community

structure and dynamics on Loxahatchee NWR will be included in a

following segment that documents vegetation changes over the

last thirty years. The following is a general overview of the

vegetation communities of the northern Everglades.

Long (1974) discusses the origin and development of the

vascular flora of south Florida and list four main sources for

the plant associations found in the state: (1) migration from

tropical America (2) migration from the coastal plain of North

America (3) relict species from the Miocene (4) species

introduced by the activities of man. Prior to Davis (1943)

descriptions of the Everglades region were provided by natural-

ists and sportsmen. Tebeau (1974) summarized historical accounts

of early explorations and descriptions. Early accounts coupled

with data from peat cores (Davis 1946, Jones 1948, Gleason et al.

1974, Cohen and Spachman 1974) support the long-standing idea

that like today, sawgrass was the dominant vegetation type in the

past. The existence of tree islands in the northern Everglades

may be a recent phenomenon of the last 500 1000 yrs (Gleason et

al. 1977).

Prior to man's influence, the vast Everglades wetland system

functioned as a nutrient-limited system. Nitrogen and phosphorus

existed in extremely low levels, with the majority of nutrients

being supplied in rainfall (Parker 1974, Waller and Earl 1975,

Mcpherson et al. 1976, Swift 1984, Swift and Nichols 1987, Gund-

erson et al. 1987). Sawgrass, the dominant plant in the Ever-

glades, has low nutrient requirements compared to other macro-

phyte species (Steward and Ornes, 1975).

Davis (1943) identified 13 Everglades vegetation types, only

some of which are present in the northern Everglades (Table 1).

Loveless (1959) listed tree islands, sawgrass marshes, wet

prairies, and aquatic slough communities as the dominant plant

communities in the Everglades and discussed the role of hydrolog-

ic regime and fire in structuring plant communities. It is

widely accepted that hydroperiod and fire were the major determi-

nants of plant community distribution over the scale of the

entire oligotrophic Everglades system. Other factors that influ-

ence Everglades plant community distribution include hurricanes,

animal disturbance, frost, and nutrient regimes (Davis 1943,

Loveless 1959, Craighead 1971, Gentry 1974).

The relationship between hydrologic regime and plant

community structure in the Everglades has been the subject of

research for almost fifty years (Davis 1943a,b,; Loveless 1959,

Egler 1952, Craighead 1971, Mcpherson 1973, Olmstead et al. 1980;

Zaffke 1983, Gunderson 1987). In particular, Loveless

Table 1. Vegetation types found in the Everglades and other
freshwater marshes listed by Davis (1943).

Type Species

Saw-grass Cladium jamaicense

Flag marshes Pontederia Saqqittaria, Thalia

Aquatic-plant marshes Nvmphaea, Utricularia

Cattail Marshes Tvpha

Spikerush-needle grass Eleocharis

Mixed herb-Shrub Mvrica, Cephalanthus, Baccharis

Fern Marshes

Bulrush marshes Scirpus
-----------------------""------------------------""
(1959) discussed both the long term influence of hydrologic

regime on shaping the overall plant community structure and the

short term effect of drought on succession within community

types.

Cohen (1974) provided evidence from peat cores that fire has

been a part of Everglades ecology ever since peat began accumu-

lating. Cohen addressed the relative frequency of fire occur-

rence (Table 2) in the different Everglades peat types. Sawgrass

communities burn frequently (Hofstetter 1974, Wade et al. 1980).

Community composition after a fire is dependent upon the depth

the fire penetrates into the peat layer (Loveless 1959, Yates

1974, Craighead 1971). Wade et al. (1980) reviewed the effects

of fire on all of the major plant communities in the Everglades.

Table 2. Relative charcoal content of some peat types of South
ern Florida. (from Cohen 1984).

Environment Peat type Charcoal content

Marine Swamp Rhizophora Low to Absent
Rhizophora-Avicennia Low to Absent

Brackish Swamp Conocarpus transition Low
Rhizophora-Acrostichum- Low to High
Cladium

Brackish Marsh Acrostichum-Cladium Moderate to High

Bay Hammock Myrica-Persea-Salix Low to Moderate

Fresh Water Marsh Cladium Moderate to High
Cladium-Water lily Low to High
Water lily Low to Absent
-----------------------------------------------------

Changes in Everglades system

As a result of drainage for agriculture, and to a lesser

extent urban development, less than 50% of the original Ever-

glades remain (Schortemeyer 1980, Kushlan 1989). The driving

forces and legislative acts that led to the drainage of the

Everglades have been chronicled by Tebeau (1974) and Degrove

(1984). The extensive levee and canal system includes the Hoover

Dike around the 90+ mile perimeter of Lake Okeechobee, the enclo-

sure of the 3 water conservation areas (Loxahatchee NWR/WCA1,

WCA2A & B, WCA3A & B), and the system of canals and pump sta-

tions. These structures are used to move water that historically

moved as sheet flow across the Everglades or was lost due to

evapotranspiration. The primary function of this levee and canal

system is flood control (Figure 6).

The largest area drained is the Everglades Agricultural Area

(EAA), an area of approximately 526,000 acres (212,870 ha.,

Figure 7). There are a variety of land use/land cover types for

the major sub-basins within the EAA (Table 3). The major crop is

sugar cane which requires substantial irrigation during the dry

season and drainage during the wet season. Other commercially

important crops include vegetables, sod, and rice (Table 4). In

spite of the large storage of nitrogen and phosphorus in the muck

soils, various crops in the EAA are fertilized depending on their

specific needs. Fertilizer application rates vary depending on

crop type, timing, and water management practices. Published

application rates range from 0.0 to 400. kg of P per hectare,

with vegetable crops receiving the highest application rates.

Figure 6. Schematic diagram of water movement through the SFWMD.
(Source: SFWMD)

19

Figure 7. Map of SFWMD showing EAA, canals, and pump stations.
(Source: SFWMD)

20

Table 3. Generalized
Acreages by

1987/1988 Land Use/Land Cover Types and
Watershed for major Sub-basins within the EAA.

Land Use Type

Basin Agriculture,
acres
(X basin)

S-2

S-3

S-4

S-5A*

S-6*

S-7

S-8

L-8

101,242
(95.5)

64,071
(99.1)

36,807
(85.8)

121,657
(97.8)

80,583
(95.0)

72,996
(86.8)

53,981
(47.5)

21,787
(24.5)

Totals 553,124
(77.98)

Urban,
acres
(% basin)

4,053
(3.8)

Rangeland,
acres
(% basin)

43
(.04)

260
(.4)

3,660
(8.5)

1,042
(.8)

941
(1.1)

1,073
(1.3)

566
(.5)

7,194
(8.1)

18,789

283
(.7)

34
(.03)

179
(.2)

18
(.02)

172
(.2)

729

(2.65) (0.10)

Forested,
acres
(X basin)

Wetlands,
acres
(% basin)

134
(.1)

93
(.1)

Water,
acres
(% basin)

505
(.5)

161
(.2)

67 1,465

(.2)

126
(.1)

(3.4)

4
(.003)

--- 2763
(3.3)

9,459
(11.2)

--- 58,517
(51.5)

1,276 58,041
(1.4) (65.4)

1,696 130,249
(0.24) (18.36)

251
(.6)

1,298
(1.0)

336
(.4)

452
(.5)

539
(.5)

325
(.4)

3,867
(0.55)

Barren Land,
acres
(% basin)

67
(.06)

45
(.07)

380
(.9)

206
(.2)

Totals,
acres
%

106,044
(100.0)

64,630
( 99.9)

42,913
(100.1)

124,367
(99.9)

1 84,803
(.001) (100.0)

144
(.2)

843
(0.12)

84,124
(100.0)

113,621
(100.0)

88,795
(100.0)

709,297
(100.7)

* Major source to Loxahatchee NWR

Source: SFWMD Land Use/Land Cover Data Base

21

Table 4. Acreage and Values of Agricultural
Everglades Agricultural Area.

Crop Acreage (X1000)

Sugar cane 433

Vegetables 23

Sod 29

Rice 3

Livestock 32

Other 6

Total 526

Production in the

Farm Sales/yr ($MILLIONS)

$350

110

30

$509

Source: IFAS, Palm Beach County Extension Service

An average application rate in the EAA is 22. kg P/ha (Sanchez

1989). Morris (1975) estimated the total storage in the EAA muck

soils of phosphorus and nitrogen to be about 15 million and 210

million tons respectively. Hortenstein and Forbes (1972) esti-

mated that 88,600 tons of N and 2000 tons of P are in unbound

forms.

Agriculture within the EAA requires drainage of the organic

soils. Early in this century the state of Florida provided

incentives and direction for the draining of the Everglades and

the conversion to agriculture. This drainage has been accom-

plished by an extensive network of canals, levees, pumps, and

water control structures. Many of these were constructed by

farmers, drainage districts and private developers. The U.S.

Army Corps of Engineers (USACOE) became involved in 1948 when the

state of Florida created the Central and South Florida Flood

Control District and helped to oversee further construction and

improvement of the entire south Florida plumbing system. Most of

the major facilities (Figure 7) are currently operated and main-

tained by the South Florida Water Management District in accord-

ance with the USACOE and state of Florida guidelines and permits.

In addition to the major pump stations, several smaller drainage

districts operate pump stations under the regulation of the state

of Florida and the SFWMD. The structures associated with the

Loxahatchee NWR are listed in Table 5.

Drainage and aeration of the EAA soils have caused consider-

able changes in the soil surface elevations. Rates of change

have been carefully documented, and losses average near one inch

Table 5. Water control structures associated with Loxahatchee NWR.

Structure Type Design HW Design TW Optimum Stage Design
Stage Stage (ft NGVD) Discharge
(ft NGVD) (ft NGVD) (CFS)
. . . ..-------------------------------------------------------------------------------------------------------

S-5A Pump Station
6 units-800 cfs
each

Gated Box
Culvert
2-7 ft x
7 ft x 65 ft
Reinforced concrete
box, invert elev =
1.0 ft NGVD

Gated Spillway
2-Gates
19.3 ft high x
22.8 ft wide
net crest Lgth =
44.0 ft
crest elev =
1.0 ft NGVD

Gated Box
Culvert
2-7 ft x 7 ft
x 80 ft
reinforced
concrete box
invert elev =
1.75 to .30 ft
NGVD

13.0

11.5

18.0

13.0

24.1

10.0

17.9

11.5

HW = 10.5
(wet season)
HW = 11.5
(dry season)

Not
used to
control
stage

Not
used to
control
stage

Not
used to
control
stage

Pump Station
3 units-975 cfs

12.5

20.8

each

Gated Spillways
4 Gates each
8.0 ft high x
25.7 ft wide
net crest Igth =
100 ft
crest elev = 10.0
10.0 ft NGVD

17.3

16.4

10.D-12.5
in Hillsboro

Canal

Regulation
Schedule
in WCA 1

24

S-5AS

*S-5AW

4800

700

2000

700

S-6

S-10A
S-1OB
S-10D

2925

4680 each
14000 total

*S-5AE

Table 5. cont

Structure Type Design HW Design TW Optimum Stage Design
Stage Stage (ft NGVD) Discharge
(ft NGVD) (ft NGVD) (CFS)

S-10E Gated Culvert 17.3 16.4 Regulation 438
3-72 in x 40 ft Schedule
CMP, invert elev = in WCA 1
9.0 ft

S-39

Acme
Improve-
ment
District

S.N.
Knight

Gated Spillway
1 Gate
9.2 ft high x
16.0 ft wide
net crest lgth =
15.0 ft, crest elev
= 2.5 ft NGVD

2 Pump Stations
230 CFS (North)
275 CFS (South)

1 Pump Station

11.0

**NA

**NA

9.0

**NA

**NA

Regulation
Schedule
in WCA 1

Not used
to control
stage

Not used
to control
stage***

800

505

325

*Auxillary structures outside WCA-1. They do not directly discharge to or from the area.
**Not Applicable.
*** Removed from service.

25

per year (Stephens 1974, Snyder 1987). Surface elevations around

Belle Glade in 1912 were between 18 and 20 feet above sea level

(Stephens 1974), while 1970 U.S. Geological Survey maps show

elevations of about 12 feet for the same area. Stephens (1974)

estimated that 88% of the total soil volume in the EAA will be

lost by the year 2000 and soil loss threatens the continuation of

agriculture in its present form (Snyder 1987).

Along with losses due to drainage, ecosystem processes in

the remaining Everglades are being altered due to changes in

hydrology, water quality, the proliferation of particular spe-

cies, and the introduction of exotic species (Loveless 1959,

Alexander and Crooke 1974, 1984; Mcpherson et al. 1976, Lutz

1977, Millar 1981, Rosendahl and Rose 1981, Swift 1981, Lowe

1986, Wagner and Rosendahl 1986, Parks 1987, Kushlan 1989).

The present hydrologic alteration of the Everglades system

involves extended hydroperiods in some areas, decreased hydrope-

riod in others, and a change in the natural pattern of water

deliveries (Davis 1943, Leach et al. 1971, 1972; Kushlan 1989).

Ponding of water behind the levees in the southern portions of

the WCA's has caused adverse effects to the plant communities

(Alexander 1971, Hofstetter and Parsons 1974, Dineen 1972, 1974,

McPherson 1973, Worth 1983). Toth (1988) found that sawgrass and

cattail growth was adversely affected by deep and widely fluctu-

ating water levels, with sawgrass being more adversely affected.

The altered hydrology of the Everglades has also affected

the animal communities. Current wading bird numbers are from 5

to 10% of numbers estimated at the turn of the century (Robertson

and Kushlan 1984, Kushlan and White 1977, Frederick and Collopy

1988). Threatened or endangered species such wood storks (Kush-

lan et al.1974, Ogden et al. 1978) and snail kites (Sykes 1983)

have experienced dramatic declines. The impacts of water manage-

ment on the alligator have been identified by Kushlan (1987).

The endangered Florida panther's range is extensive across the

historic Everglades.

Nutrients

Along with the impacts due to changes in the hydrologic

regime of the northern Everglades, changes in the system as a

result of water quality impacts are becoming apparent. The

creation of the EAA affected the nutrient regime of the remaining

Everglades system in four ways. First, the potential for storage

of nutrients in rainfall by accumulating peat in the EAA is

removed. Second, the process of sequestering nutrients in peat

is reversed as nutrients are released due to soil oxidation.

Third, nutrients in the form of fertilizer are being added to the

system. Fourth, surface waters are mixed with highly mineralized

ground water as a result of the depth of the drainage canals.

Some of these nutrients are removed from the system by the export

of agricultural products. The remaining portion of the nutrient

load leaves the EAA either in dissolved or particulate form in

the water or in smoke from burning the cane fields or the bagasse

at the processing plants.

EAA drainage waters are nutrient enriched (with nitrogen and

phosphorus), contain high concentrations of chlorides, dissolved

minerals, iron (from ground water), and measurable levels of

pesticides (McPherson 1973,1978; Waller and Earl 1975, Lutz 1977,

CH2M-Hill 1978, Dickerson et al. 1978, Millar 1981). Prior to

1980, water was pumped from the northern one-third of the EAA

into Lake Okeechobee. Water from the remaining two-thirds was

pumped into the three conservation areas. Dickson et al. (1978)

reported that the areal nitrogen loading rate from the EAA was

the highest and the phosphorus loading rate was the second high-

est of any major basin of Lake Okeechobee.

Based on concerns of degradation of water quality in Lake

Okeechobee, the Interim Action Plan (IAP) was adopted in 1980.

Under the IAP, excess water is pumped into Lake Okeechobee only

when runoff exceeds capacity of the southern pump stations of the

EAA to pump water southward. This plan, while good for Lake

Okeechobee, increased the amount of water and nutrients entering

the WCA's.

The contribution of EAA drainage water on the water and

nutrient budgets of the three water conservation areas is well

documented (Millar 1981, SFWMD data). For the 1979-1988 water

years atmospheric deposition accounted for 66% of the water, 33%

of the nitrogen, 40% of the phosphorus entering the water conser-

vation areas (Table 6). Canal inputs accounted for the remain-

der. The areal phosphorus loading rate for WCA3A is much less

than in Loxahatchee NWR (WCA1) and WCA2A (Table 7). This

Table 6. Average annual water and nutrient budgets (1979-1988)
for WCA1, WCA2A, and WCA3A combined. Total nitrogen and total
phosphorus are in metric tons and volume is acre-feet (SFWMD
data).

INPUT Total Phosphorus Total Nitrogen Volume

S8
59
5140
S150
L281
L3
S5A
56
57
Atmospheric

77.19
4.49
18.03
4.72
8.34
21.08
73.65
23.04
28.34
107.64

(18%)
(1%)
(4%)
(1%)
(2%)
(5%)
(17%)
(5%)
(7%)
(40%)

1663
338
237
197
131
173
3053
1068
1253
4005

(14%)
(3%)
(2%)
(2%)
(1%)
(1%)
(25%)
(9%)
(10%)
(33%)

312,406
135.823
104,373
56,741
71,497
68,020
314,754
156,968
219,463
2,823,401

(7%)
(3%)
(2%)
(1%)
(2%)
(2%)
(7%)
(4%)
(5%)
(66%)

TOTAL 429.52 12,135 4,263,176

OUTPUT

S151
S12A
S12B
S12C
S12D
S333
S38
S39
5144
5145
5146
E.T.

5.31
0.83
0.82
2.35
2.49
4.41
0.86
5.47
0.74
0.65
0.46

(22%)
(3%)
(3%)
(10%)
(10%)
(18%)
(4%)
(22%)
(3%)
(3%)
(2%)

(18%)
(6%)
(6%)
(14%)
(17%)
(14%)
(5%)
(10%)
(4%)
(4%)
(3%)

182,937
79,733
75,344
160,512
167,769
138,733
42,859
77,778
36,493
38,576
30,271
3.331.329

(4%)
(2%)
(2%)
(4%)
(4%)
(3%)
(1%)
(2%)
(1%)
(1%)
(1%)
(76%)

TOTAL 24.39 2,387 4,361,794
TOTAL 24.39 2,387 4,361,794

Table 7. Comparison of Marsh Surface Area, Total Phosphorus
Loading and Areal Loading Rates, Percent Retention and Watei
Residence Times Among the Three Water Conservation Areas,
1979-1988.

Water Size of WCA's
Management Sq. Mi. (km2)
Unit

WCA-1 227(356)

WCA-2A 173(278)

Total Phosphorus___
Average Areal Phoshporus
Loading* Loading Rate** Retention

126 0.22

110 0.25

WCA-3A 786(1,262) 275 0.14

Source: SFWMD, unpublished data, = metric tons/yr, ** = Grams/m2/yr

difference is due to phosphorus being retained in the upper

conservation areas (49% WCA1 and 71% WCA2A).

Swift and Nicholas (1987) documented the relationship be-

tween water quality and the periphyton community of the Ever-

glades, noting the possible adverse effects of a changing pe-

riphyton community. They described three periphyton communities:

hard water high N and P, hard water low N and P, and soft water

low N and P. They also noted that the periphyton community in

areas enriched with nitrogen and phosphorus had pollution toler-

ant species, higher growth rates, and low diversity compared to

the unenriched areas. Other studies have documented the effects

of nutrient regime on the periphyton community of the Everglades

(Swift 1981, 1984; Flora et al, 1987, Scheidt 1988). The native

periphyton communities serve as a major component of the Ever-

glades food chain and are involved in the oxygen dynamics of the

surface waters. Long term changes in the composition of the

periphyton community could ultimately affect higher trophic

levels (Swift and Nicholas 1987).

Besides impacting the periphyton community, it is now appar-

ent that the increased nutrient loading to the water conservation

areas affects macrophyte community structure. In particular,

cattail is expanding into areas previously dominated by native

Everglades vegetation such as sloughs, wet prairies, and sawgrass

in Loxahatchee NWR and WCA2. Increased nutrient loading from the

EAA has been suggested as a major factor in the spread of cat-

tails in areas that were sawgrass (Reeder and Davis 1983, Davis

1989). Steward and Ornes (1975) reported that sawgrass has lower

nutrient requirements compared to other macrophyte species and

later Stewart and Ornes (1983) reported that high levels of

nutrients, particularly phosphorus, may inhibit sawgrass growth.

Davis (1989) showed that low nutrient requirements alone do not

explain the dominance of sawgrass over cattail in the historic

oligotrophic Everglades, but that cattail is better adapted to

take advantage of years of elevated nutrients. Davis (1989) also

states "the responses of cattail to yearly variations in nutrient

supply correspond well to its success in an environment that is

prone to surges of nutrient rich waters which are larger during

some years than in others" and goes on to say that "such an

environment typifies nutrient-enriched areas of the Everglades."

LOXAHATCHEE NWR STUDY

Loxahatchee NWR was established on June 8, 1951. Until this

time the hydrology of the refuge was affected by the major canals

that had been completed in the 1920's (the Hillsborough canal on

the south end, the Palm Beach canal on the north end, and the St.

Lucie and Caloosahatchee canals diverting water from Lake Oke-

echobee). These major canals had marked an end to the period of

historic water flows in the Everglades. During the 1950's the L-

7 canal on the western boundary and the L-40 canal on the eastern

boundary were established along with changes to the Hillsborough

canal to allow water to be gated into WCA2-A. The two large pump

stations (S5A and S6) were also installed during this time peri-

od. The control structures were finished and the ability to

control water was implemented during June/July of 1960. This

marked a radical change in the hydrologic pattern that had been

in place for the 50 years previous to this time.

In the intervening 30 years there have been at least several

major fires in the refuge, all coincidental with drought condi-

tions (1955 10,000 acres (4047 ha), 1962 over 100,000 acres

(40,470 ha) burned; 1981 6500 acres (2,631 ha) burned on the

west side; and 1989 45,000 acres (18,211 ha) burned on the west

side and the north end). Numerous smaller fires have occurred

along with some managed fires.

Historic Vegetation Mapping and Surveys

Since the inception of the refuge, monitoring and classifi-

cation of vegetation types and distributions has been done

through various studies by the U. S. Fish and Wildlife Service

(FWS) personnel. Aerial photography has been flown on several

occasions and at least two rectified photo maps have been gener-

ated from the aerial photography. The first attempt at a vegeta-

tion map was completed in the mid 1950's using approximately 35

phototransects from earlier 1952 aerial photography. This map

was a generalized vegetation map with little fine scale detail

(Figure 8). This was followed by a similar map in 1966 (not

shown) and 1968 (Figure 9).

A series of vegetation transects were started in 1959 and

continued intermittently until 1969. These three transects

(Figure 10) were analyzed quantitatively for species composition.

They provide an excellent baseline for vegetation near the canal.

In the late 1960's a series of 100 acre photoplots were analyzed

by FWS personnel (Thompson 1972) using a series of aerial photo-

graphs. There were a total of seven photoplots demarcated on old

aerial photos of the refuge (Figure 10). Photos from the years

1948, 1952, 1962 and 1968, were analyzed and vegetation was

classified into 5 separate groups (aquatic, wet prairie, saw-

grass, brush, and tree island). Water level stage recorders were

placed in each of the plots and left for a year and a half start-

ing in 1973. Intensive vegetation analysis transects were quan-

tified in each of the plots to determine the relationship between

vegetation and hydroperiod.

Loxhnatchee
National
Wildlife
Refuge

S6

I Sawgr-ss
2 Sawgross/myrtle
3 Myrtle
4 Slough S10D
5 Whitegrass SIZ0C
6 Mixed grosses SI
7 Moidencane

Figure 8. Vegetation cover map for 1952.

LoxfkcLckee
NtE i olO 1
Wi 11 iFe
RReFuie

I Sawgrass Aixed
grasses myrtle
2 Transition
3 WeL prlr-le tree
island slough
4 Ride sMwgross 4
myrtle
5 Sawgross
6 AquLaic
7 Sawgriss picheral weed 5
brush
B Sowgross spikerush
9 Maidencene

Figure 9. Vegetation cover map for 1968.

In 1
dl if
uge

SGI 01-9

A

SC

Figure 10. Map showing location of vegetation photoplots (1-7),
vegetation transects (A-C), and gaging stations.

The National Wetland Inventory constructed a vegetation map

of the refuge in 1985. The base photography for the study was

from 1983-85 National High Altitude Photography, at an approxi-

mate scale of 1:58000. The photography was interpreted using a

classification with 14 different classes, 10 of which occur

within the levee of the refuge. The classification was tran-

ferred and rectified to 1:24,000 scale quad maps by Martel, Inc.,

and provides a detailed record of the extent of cattail coverage

of the refuge for that time period. A final digital product was

never finished after problems in funding the digitizing and

correcting some background/foreground problems.

Classified Vegetation Map from SPOT Imagery

Satellite imagery of the refuge was acquired from SPOT Image

Corp. for April 4, 1987 (scene id 16232978704041559122X and

16232978704041559101P). Both the 10 meter panchromatic and 20

meter multispectral data sets were obtained and used to make a 10

meter multispectral image by merging the two data sets. The

technique for merging the two data sets requires resampling and

registering the 20 meter multispectral data set to the 10 meter

panchromatic data set. The resampled 10 meter multispectral data

is then transformed from a Red/Green/Blue (RGB) colorspace (band

1 assigned to blue, band 2 assigned to green and band 3 assigned

to red) to an Intensity/Hue/Saturation (IHS) colorspace. This is

functionally equivalent to a polar coordinate transformation.

The 10 meter panchromatic band is substituted for the intensity

band in the IHS colorspace and then the data is transformed back

to the RGB colorspace.

A vegetation index band (infrared-red/infrared+red) was

calculated from the registered 10 meter multispectral data before

transforming with the panchromatic band. This added band was

used in the vegetation classification. A set of approximately 25

signatures was obtained by digitizing known areas of vegetation

and using these areas to set the parameters for the signature set

(Jensen 1986, ERDAS 1987). These signatures were analyzed for

similarity and separation capabilities. Several classifications

were made of small test areas of well known vegetation types to

evaluate the set of signatures for their ability to classify and

separate the vegetation communities. Through this iterative

process a final group of 18 signatures were obtained (Table 8, 9,

10). A final classification of the 4-band image was done using a

maximum likelihood classification technique (Niblack 1985, ERDAS

1987). This final classified image (Appendix) was rectified to a

state-plane coordinate system utilizing known survey benchmarks

and digitized points from a 1:24000 USGS 7.5 minute quadrangle

map to a 30 ft by 30 ft pixel size. Generalized ground truthing

was accomplished using aerial surveys and comparing the rectified

image with known areas using a LORAN C standardized to a known

survey benchmark.

In order to make comparisons between the classified vegeta-

tion map and other data such as the historic photo-plots and the

three historic vegetation transects, the 18 classes (Table 8)

were recorded into community types that were similar to the cate-

gories defined in the older analyses. These types were aquatic,

wet prairie, sawgrass, brush, and tree islands. It was necessary

Table 8. Vegetation Communities in Supervised Classification

1. (Sawgrass) High density sawgrass (948 Acres, .7%) very
dense sawgrass with some occurrence of fern tussocks.

2. (Cattail) Sawgrass with invasion of cattail (2124 Acres,
1.5%) sawgrass mostly near the canal and with high incidence of
cattail mixed in the sawgrass.

3. (Sawgrass) Sawgrass (18132 Acres, 13.0%)- primary sawgrass
class occurring on all parts of the refuge including the vast
sawgrass areas on the west side of the refuge.

4. (Brush) Brush/sawgrass (21915 Acres, 15.7%) primarily
sawgrass with large amounts of wax myrtle. Some tree islands
which may have been burned out previously are made up entirely of
this class, particularly in the southern part of the refuge.

5. (Tree Island) Tree island (2387 Acres, 1.7%) lower stature
tree island community made up of a mix of wax myrtle, dahoon
holly and red bay. Occurs along edges of large tree islands with
some smaller tree island made up entirely of this class.

6. (Wet Prairie) Wet prairie (46544 Acres, 33.4%) largest
area of the refuge. This class is the denser wet prairies occur-
ring over all of the refuge but the primary community type of the
central portion of the refuge. This class may contain small tree
islands smaller than 30 ft across and areas with small sawgrass
strands.

7. (Tree Island) Tree island (867 Acres, .6%) core of larger
tree islands, larger stature trees made up primarily of dahoon
holly and red bay with lesser amounts of wax myrtle than class 5.

8. (Brush) Brush (4771 Acres, 3.42%) smaller brush clumps
primarily in wet prairies.

9. (Wet Prairie) Wet prairie (9934 Acres, 7.1%) sparser wet
prairie community often with sparse sawgrass.

10. (Cattail) Cattail (1746 Acres, 1.3%) cattail community
close to the canal.

11. (Aquatic) Open water (282 Acres, .2%) mostly identified
as deep water along Hillsborough canal.

12. (Aquatic) Slough / very sparse wet prairie (272 Acres, .2%)
- a small class, very similar to class 9 but a little deeper
and/or less vegetated.

Table 8 cont.

13. (Tree island) Willow / brush (1160 Acres, .8%) predomi-
nately willow but with some mixed classification with wax myrtle
brush areas.

14. (Tree island) Brush/tree island (16467 Acres, 11.8%) -
class of many of the smaller tree island with mostly wax myrtle,
some sawgrass, occasionally dahoon holly and red bay.

15. (Sawgrass) Sawgrass / brush (2548 Acres, 1.8%) mostly
sawgrass with some invasion of wax myrtle, generally closer to
canal than other sawgrass classes. A small class similar to
class 4.

16. (Sawgrass) Sawgrass (6214 Acres, 4.46%) slightly higher
elevation sawgrass than class 3. Core of sawgrass ridges.

17. (Tree island) Willow (1167 Acres, .8%) willow along canal
edge, some misclassified floating aquatics along hillsborough
canal.

18. (Cattail) Cattail (1856 Acres 1.33) cattail further from
the canal than class 10.

Table 9. Signature euclidian distance table.

Class 1 Class 2 Class 3

0.00
14.36
12.48
4.26
24.05
19.87
28.64
5.57
42.51
15.10
47.68
41.49
13.31
4.32
7.99
14.99
30.98
26.10

14.36
0.00
3.50
16.91
37.90
23.93
42.29
17.19
38.98
13.88
42.66
38.93
27.08
16.47
7.86
2.75
44.62
19.57

12.48
3.50
0.00
14.38
35.88
20.46
40.10
14.47
36.53
15.71
40.53
36.29
25.18
13.91
5.16
2.63
43.05
22.67

Class 4 Class 5 Class 6 Class 7 Class 8 CLass 9

4.26
16.91
14.38
0.00
22.06
16.50
26.23
1.78
40.33
19.34
45.78
39.08
11.93
1.30
9.62
16.94
29.82
30.23

24.05
37.90
35.88
22.06
0.00
31.46
5.36
22.15
56.78
34.73
62.83
54.80
11.19
22.37
31.41
38.40
9.23
45.34

19.87
23.93
20.46
16.50
31.46
0.00
33.51
15.22
25.59
33.22
31.45
23.87
24.77
16.35
18.11
22.02
40.61
42.46

28.64
42.29
40.10
26.23
5.36
33.51
0.00
26.13
58.31
39.83
64.44
56.18
16.05
26.53
35.65
42.60
9.90
50.38

5.57
17.19
14.47
1.78
22.15
15.22
26.13
0.00
39.35
20.45
44.79
38.07
12.17
1.31
9.98
17.04
30.16
31.14

42.51
38.98
36.53
40.33
56.78
25.59
58.31
39.35
0.00
51.99
6.66
2.87
50.18
40.19
37.08
36.25
65.96
58.12

Table 9 continued.

Class 10 Class 11 Class 12 Class 13 Class 14 Class 15 Class 16

15.10
13.88
15.71
19.34
34.73
33.22
39.83
20.45
51.99
0.00
56.10
51.65
24.46
19.18
15.61
16.48
39.00
11.49

47.68
42.66
40.53
45.78
62.83
31.45
64.44
44.79
6.66
56.10
0.00
9.17
55.87
45.55
41.74
39.94
72.00
61.29

41.49
38.93
36.29
39.08
54.80
23.87
56.18
38.07
2.87
51.65
9.17
0.00
48.57
38.97
36.47
36.21
64.00
58.27

13.31
27.08
25.18
11.93
11.19
24.77
16.05
12.17
50.18
24.46
55.87
48.57
0.00
12.05
20.90
27.73
18.28
35.33

4.32
16.47
13.91
1.30
22.37
16.35
26.53
1.31
40.19
19.18
45.55
38.97
12.05
0.00
9.40
16.50
30.13
29.92

7.99
7.86
5.16
9.62
31.41
18.11
35.65
9.98
37.08
15.61
41.74
36.47
20.90
9.40
0.00
7.56
38.73
24.68

14.99
2.75
2.63
16.94
38.40
22.02
42.60
17.04
36.25
16.48
39.94
36.21
27.73
16.50
7.56
0.00
45.51
22.19

Classl7 Class 18

30.98
44.62
43.05
29.82
9.23
40.61
9.90
30.16
65.96
39.00
72.00
64.00
18.28
30.13
38.73
45.51
0.00
48.84

26.10
19.57
22.67
30.23
45.34
42.46
50.38
31.14
58.12
11.49
61.29
58.27
35.33
29.92
24.68
22.19
48.84
0.00

Class 1
Class 2
Class 3
Class 4
Class 5
Class 6
Class 7
Class 8
Class 9
Class 10
Class 11
Class 12
Class 13
Class 14
Class 15
Class 16
Class 17
Class 18

Class 1
Class 2
Class 3
Class 4
Class 5
Class 6
Class 7
Class 8
Class 9
Class 10
Class 11
Class 12
Class 13
Class 14
Class 15
Class 16
Class 17
Class 18

Table 10. Signature Statistics Listing

Signature Name: Class 1
Number of points = 106
Band 1 2
Minimum 36 28
Mean 38.31 30.22
Standard 0.78 0.66
Maximum 40 31
Covariance Matrix
1 0.61 0.33
2 0.33 0.43
3 0.54 0.39
4 -0.07 -0.17

Signature Name: Class 2
Number of points = 190
Band 1 2
Minimum 42 35
Mean 44.89 37.94
Standard 0.95 1.12
Maximum 48 41
Covariance Matrix
1 0.91 0.89
2 0.89 1.26
3 0.79 0.87
4 -0.17 -0.35

Signature
Number of
Band
Minimum
Mean
Standard
Maximum
Covariance
1
2
3
4

Name: Class 3
points =
1
41
43.43 35
1.02 0
46
SMatrix
1.03 0
0.71 0
0.37 0
-0.28 -0

99
2
33
.42
.87
37

.71
.75
.39
.32

Signature Name: Class 4
Number of points = 68
Band 1 2
Minimum 33 24
Mean 36.85 27.81
Standard 1.35 1.57
Maximum 39 31
Covariance Matrix
1 1.83 1.65
2 1.65 2.48
3 0.42 -0.24
4 -2.04 -3.46

Signature Name: Class 5
Number of points = 265
Band 1 2
Minimum 29 19
Mean 31.59 20.46
Standard 0.74 0.69
Maximum 33 22
Covariance Matrix
1 0.55 0.25
2 0.25 0.47
3 0.39 -0.03
4 -0.24 -0.98

Signature Name: Class 6

3
55
59.40
1.08
62

0.54
0.39
1.16
0.23

3
52
55.33
1.26
58

0.79
0.87
1.59
0.20

3
51
53.66
1.13
57

0.37
0.39
1.28
0.31

3
52
56.57
1.74
60

0.42
-0.24
3.04
2.28

3
60
62.77
1.14
66

0.39
-0.03
1.30
1.02

4
37
38.24
0.68
41

-0.07
-0.17
0.23
0.46

4
27
28.92
0.64
30

-0.17
-0.35
0.20
0.41

4
29
29.91
0.70
31

-0.28
-0.32
0.31
0.49

4
35
39.75
2.56
44

-2.04
-3.46
2.28
6.54

4
55
58.89
1.90
63

-0.24
-0.98
1.02
3.60

Number of points =
Band 1
Minimum 33
Mean 35.78
Standard 1.50
Maximum 40
Covariance Matrix
1 2.24
2 0.81
3 0.62
4 -0.77

2
20
22.32
0.91
25

0.81
0.83
0.36
-0.76

Signature Name: Class 7
Number of points = 81
Band 1 2
Minimum 28 17
Mean 29.88 18.40
Standard 0.62 0.60
Maximum 31 20
Covariance Matrix
1 0.38 0.25
2 0.25 0.36
3 0.30 0.08
4 -0.37 -0.65

Signature Name: Class 8
Number of points = 54
Band 1 2
Minimum 36 25
Mean 37.70 27.04
Standard 1.12 0.82
Maximum 40 29
Covariance Matrix
1 1.25 0.57
2 0.57 0.66
3 1.04 0.77
4 -0.14 -0.37

Signature Name: Class 9
Number of points = 80
Band 1 2
Minimum 31 20
Mean 34.93 22.76
Standard 1.32 0.96
Maximum 39 25
Covariance Matrix
1 1.74 0.74
2 0.74 0.93
3 0.40 0.59
4 -0.37 -0.29

Signature Name: Class 10
Number of points = 122
Band 1 2
Minimum 41 37
Mean 44.16 39.94
Standard 1.28 0.97
Maximum 48 42
Covariance Matrix
1 1.64 0.82
2 0.82 0.94
3 1.00 0.49
4 -0.31 -0.46

3
39
41.46
1.33
45

0.62
0.36
1.78
1.01

3
57
60.43
0.94
62

0.30
0.08
0.89
0.49

3
52
55.26
1.32
58

1.04
0.77
1.75
-0.01

3
20
22.17
1.46
25

0.40
0.59
2.12
1.24

3
65
68.22
1.42
71

1.00
0.49
2.01
0.45

4
33
36.20
1.47
41

-0.77
-0.76
1.01
2.16

4
59
62.90
1.51
65

-0.37
-0.65
0.49
2.29

4
39
40.09
0.80
41

-0.14
-0.37
-0.01
0.64

4
18
19.40
1.16
22

-0.37
-0.29
1.24
1.34

4
32
33.62
0.84
36

-0.31
-0.46
0.45
0.71

Table 10 continued.

Signature Name: Class 11
Number of points = 128
Band 1 2
Minimum 36 22
Mean 38.11 23.69
Standard 1.08 0.81
Maximum 40 25
Covariance Matrix
1 1.18 0.49
2 0.49 0.65
3 0.05 0.05
4 -0.24 -0.29

Signature
Number of
Band
Minimum
Mean
Standard
Maximum
Covariance
1
2
3
4

Name: Class 12
points =
1
32
33.89 21.;
0.94 0.
36
e Matrix
0.88 0.
0.14 0.
0.09 0.<
-0.17 0.

18
2
20
22
63
23

14
40
60
11

Signature Name: Class 13
Number of points = 52
Band 1 2
Minimum 35 23
Mean 36.65 24.21
Standard 0.73 0.49
Maximum 38 25
Covariance Matrix
1 0.53 0.19
2 0.19 0.24
3 0.39 0.01
4 -0.23 -0.26

Signature
Number of
Band
Minimum
Mean
Standard
Maximum

1
2
3
4

3
17
18.11
0.49
19

0.05
0.05
0.24
0.13

3
21
22.78
1.27
25

0.09
0.60
1.62
0.89

3
59
62.13
1.18
64

0.39
0.01
1.39
0.54

Name: Class 14
points = 60
1 2 3
36 25 52
38.12 27.67 56.30
1.20 1.03 2.52
41 30 62

1.44 0.87 2.35
0.87 1.06 1.53
2.35 1.53 6.34
0.47 -0.08 2.42

Signature Name: Class 15
Number of points = 49
Band 1 2
Minimum 38 32
Mean 39.80 33.16
Standard 0.88 0.47
Maximum 42 34
Covariance Matrix
1 0.77 0.16
2 0.16 0.22
3 0.75 -0.19
4 0.10 -0.29

3
52
54.90
1.69
58

0.75
-0.19
2.87
1.54

4
15
15.30
0.52
17

-0.24
-0.29
0.13
0.27

4
20
21.50
0.90
23

-0.17
0.11
0.89
0.81

4
47
49.67
0.87
51

-0.23
-0.26
0.54
0.76

4
37
39.82
1.38
42

0.47
-0.08
2.42
1.92

4
31
32.51
1.09
34

0.10
-0.29
1.54
1.19

Signature Name: Class 16

Number of points =
Band 1
Minimum 42
Mean 44.14
Standard 0.93
Maximum 46
Covariance Matrix
1 0.87
2 0.30
3 0.37
4 -0.06

93
2
36
37.15
0.79
39

0.30
0.62
0.25
-0.31

Signature Name: Class 17
Number of points = 70
Band 1 2
Minimum 29 18
Mean 30.73 20.79
Standard 1.47 1.40
Maximum 36 26
Covariance Matrix
1 2.17 1.23
2 1.23 1.97
3 -0.19 0.43
4 -3.74 -4.80

Signature Name: Class 18

Number of points =
Band 1
Minimum 47
Mean 50.72
Standard 1.86
Maximum 56
Covariance Matrix
1 3.47
2 3.47
3 4.17
4 -0.59

99
2
45
47.89
2.08
54

3.47
4.34
4.19
-1.05

3
51
52.90
0.90
55

0.37
0.25
0.82
0.19

3
66
69.81
2.11
75

-0.19
0.43
4.44
2.32

3
66
71.13
2.56
78

4.17
4.19
6.56
-0.29

4
27
28.22
0.64
29

-0.06
-0.31
0.19
0.41

4
51
64.79
3.88
71

-3.74
-4.80
2.32
15.05

4
28
29.44
0.78
31

-0.59
-1.05
-0.29
0.61

to add cattail as a class for the analysis. These general class-

es are in parentheses with the description of the classes (Table

8).

Vegetation nearest neighbor analysis

The spatial relationships of the vegetation communities from

the classified vegetation map (Appendix) were analyzed using a

nearest neighbor type analysis. Software was written to achieve

this end. A 3x3 pixel window was passed over the entire classi-

fied vegetation map. Within this widow a running total was kept

for every occurrence of each class appearing next to the center

pixel of the window. The data was normalized by dividing the

totals of each center class by the frequency of occurrence. This

data is summarized in Figure 11. Each box represents a vegeta-

tion class from the map. The arrows signify the association of

one class with other classes. The numbers on the arrows are the

normalized frequency of occurrence. The vegetation type at the

arrow's head is interpreted as the class surrounding the class

at the tail of the arrow. All arrows signify associations great-

er than .05 while the thick arrows signify associations greater

than .20. The data in the box represents the frequency of occur-

rence on the refuge ( A in legend box), amount of self associa-

tion (B in legend box), vegetation type and class number.

This analysis helps to define some of the vegetation classes

in more detail than the vegetation descriptions. The core tree

island class (7) is only associated with the other tree island

class (5) and the brush/sawgrass class (4). The two willow

classes appear very similar (13 and 17). The various

45

_ __ _~--_11_111___ 1-

iZ n L M 00
~ 1C 4 m a I I o o o

13 C (I3
SL L 0
NJ (n Q) u

> . 03
S 2lp Om< u
- U N

.. .) : <

\ \ 4-. ^ -- I-------,<
---~ ~~ ~~ --s : __ L. ~U

0)0
n ts LO --

/r 0

(-)~ ,*= 3t s^4
g: m

o~ (o
L .1 -. *l
-m L 0 0r

7s .. Q -,-
(D \ r-- __ 0 < --- P -
m --J

cr*' , L 0 Q

"~ (D B )-
-- L-U __ o; ___ L ( -I
0 I- 1 0 0
m, a m-.)
oo 3cno ar n

, iJ o ^H ^^ ,.
M a3 Q L)
U) co >

\n 0 O U
m im

co r-l4 T

L (n
a3 l

/~~~3 L ( r- Il
cn r)

C \D CCo LC
co C3 0 L
:) 3: w, r

0 En6 \ ^^ "

Ln -
10 Uz 1010 (
7 o A L-c o *c r

r; l nj I I '. (i^ r-l .^r O
L i L ci
SQ) *

4P-j
CD fl *H0 0
:; i 4-)

brush/sawgrass/tree island complex types (4,15,8,14 and 1) have

many cross associations indicative of the complexity of the spa-

tial associations of these vegetation map classes. The two

sawgrass classes (3 and 16) lie between the brush/sawgrass (4)

and wet prairie (6) although the class 3 sawgrass has the strong-

est associations. The cattail/sawgrass class (2) is associated

between the two sawgrass classes and one of the cattail classes

(18). The spatial occurrence of cattail (18 and 10) is associat-

ed primarily with the sawgrass communities. This is evident on

the ground where cattail is usually found in areas formerly

vegetated by sawgrass. The cattail takes over from the sawgrass

and fills in the sloughs and wet prairies in the sawgrass commu-

nities. The aquatic slough (12), wet prairie (6 and 9) and open

water (11) are grouped together.

Intensive Vegetation Survey

The observed changes in the vegetation composition on the

Loxahatchee NWR are of major concern. To understand the driving

forces) behind these changes it was necessary to develop the

relationships between species, their spatial distribution and

environmental variables. This was accomplished by an intensive

vegetation/sediment survey.

Methods

A total of 246 vegetation sample plots were surveyed, 184 in

the graminoid marsh and 62 in tree islands (Figure 12). In order

to assure adequate coverage with the spatially heterogeneous site

conditions, sampling of graminoid communities was stratified into

LoxnktLcee
N2r. i ormc- 1
Wi Idi Fe
ReFuig

oGrid survey a 0 o
iLs 00 o00o o o o o c
VegetaLi / 0
sites S1C 0 --0
SI0A OS3

Figure 12. Map of grid survey and vegetation site locations

15 geographic sectors encompassing the refuge. The same tech-

nique was used for tree islands except 4 geographic sectors were

used.

Within each graminoid survey sector, 4 to 7 sites were

located by random time and direction of travel method. The

location of each site was recorded using a standardized LORAN C.

One 5 x 5 m vegetation survey plot was sampled in each subjec-

tively demarcated vegetation type within a 60 m (accuracy limit

of LORAN) radius of each site. This survey was conducted during

March 1987.

Within each tree island survey sector, 6 to 9 tree islands

were randomly chosen. One 5 X 5 meter sample plot was sampled in

each of the tree islands chosen. This survey was conducted during

November 1987.

Species composition in all plots was determined visually in

5% increments of cover. Only those species representing 5% or

greater were recorded (Table 11). Due to the seasonal succession

that occurs in the wet prairies on the refuge, some species were

not present at the March sampling time (i.e. Rhynchospora).

Unvegetated areas in the plots were recorded as the equivalent of

a plant species and all samples totaled 100% cover.

A sample of the top 10 cm of sediment was collected from

each 5 x 5 m plot. Samples were kept frozen after collection

then were dried and ground for analysis. Soil analyses were

performed by the University of Florida Soils Testing Lab. All

analyses except chloride, ammonium and nitrate were performed on

the acid extractable portions (Mehlich 1953). This type of acid

Table 11. Loxahatchee plant species with > 5% cover. Species
with initials in paranthesis were used in data
analysis.

Genus species

Family

Blechnum serrulatum
Cephalanthus occidentalis
Cladium jamaicense (CJ)
Eichornia crassipes
Eleocharis montevidensis (EM)
Eleocharis quadrangulata
Eriocaulon compressum
Fuirena scirpoides
Hydrocotyle umbellata
Hypericum spp.
Ilex cassine (IC)
Ludwigia octovalis
Mikania scandens
Myrica cerifera (MC)
Nymphaea odorata (NO)
Osmunda regalis
Panicum hemitomon (PH)
Persea borbonia (PB)
Phragmites communis
Polygonum hydropiperoides
Pontederia cordata (PC)
Sagittaria latifolia (SL)
Salix caroliniana (SC)
Salvinia spp.
Typha domingensis (TD)
Utricularia spp. (UT)
Xyris spp.
Open Water/Sky
Acrostichum danaeifolium
Annona glabra
Melaleuca quinquenervia
Taxodium distichum
Ficus aurea

Blechnaceae
Rubiaceae
Cyperaceae
Pontederiaceae
Cyperaceae
Cyperaceae
Eriocaulaceae
Cyperaceae
Apiaceae
Hypericaceae
Aquinoliaceae
Onagraceae
Asteraceae
Myricaceae
Nymphaeaceae
Osmundaceae
Poaceae
Lauraceae
Poaceae
Polygonaceae
Pontederiaceae
Alismataceae
Salicaceae
Salviniaceae
Typhaceae
Lentibulariaceae
Xyridaceae

Pteridaceae
Annonaceae
Myrtaceae
Taxodiaceae
Moraceae

=====-==================================

========================================

-I~----~

extraction is commonly used as an indicator of the plant avail-

able nutrients. Chloride, ammonium, and nitrate were water

extractions. Concentrations of Ca, Mg, Zn, Cu, Mn, Al, Fe, P,

were determined by atomic absorption spectrophotometry (Perkin

and Elmer Corporation 1976). K and Na were quantified using

flame emission spectrophotometry. Colorimetric techniques were

used to determine total extractable N and P (Technicon Industrial

Systems 1976) and chloride (Zall et al. 1956, Technicon Industri-

al Systems 1974). Ammonia and nitrate were determined colorimet-

rically (Rhue and Kidder 1983).

Peat nutrient concentrations for each site were used to make

maps of the distribution of peat nutrients over entire refuge

using an inverse distance algorithm in ERDAS. These GIS data

sets were converted into the format used by SURFER and plotted in

3-D plots (Figures 13-22) to show their distribution over the

refuge.

Three water depth measurements were made in each 5 x 5 m

plot. This depth average was then subtracted from the water

elevation to determine the elevation of the peat surface at each

site.

Nutrient/ion Distribution

The influence of canal water on the distribution of many

peat ion and nutrient species is evident (compare the form and

spatial relationships of the peat sediment graphs (Figures 13-22)

and water quality graphs (see next section Figures 28-42). Nutri-

ent and ion rich canal water moves into the refuge causing in-

creased concentrations in the peat (over 75% water). These

PEAT AMMONIUM
WATER EXTRACTABLE

Figure 13.

Map of water extractable ammonium from peat sediment
at Loxahatchee NWR.

'S
3
1

PEAT CHLORIDE
WATER EXTRACTABLE

Map of water extractable
at Loxahatchee NWR.

chloride

from peat

sediment

Figure

14.

ct\ gD

c~i
3
1

PEAT NITRATE
WATER EXTRACTABLE

7s~
00

@0
Go

Figure 15.

Map of water extractable nitrate from peat sediment at
Loxahatchee NWR.

c9

BP\

i'

3
1

PEAT MEHLICH
EXTRACTABLE PHOSPHORUS

QT

Figure 16.

Map of Mehlich extractable phosphorus from peat sedi-
ment at Loxahatchee NWR.

NCI~

PEAT MEHLICH
EXTRACTABLE MAGNESIUM

KqPZ

Figure 17. Map of Mehlich extractable magnesium from peat sedi-
ment at Loxahatchee NWR.

Is
a.
8

cD~
3
1

PEAT MEHLICH
EXTRACTABLE ALUMINUM

8/
8

03
00

Figure 18.

Map of Mehlich extractable aluminum from peat sediment
at Loxahatchee NWR.

ma)

1~

PEAT MEHLICH
EXTRACTABLE POTASSIUM

"4

Figure 19.

Map of Mehlich extractable potassium from peat sedi-
ment at Loxahatchee NWR.

s~j
a>
1

PEAT MEHLICH
EXTRACTABLE SODIUM

ma)

ma)
3'
63

Figure 20.

Map of Mehlich extractable sodium from peat sediment
at Loxahatchee NWR.

-- -
6019

PEAT MEHLICH
EXTRACTABLE CALCIUM

(cLi

Figure 21. Map of Mehlich extractable calcium from peat sediment
at Loxahatchee NWR.

CD~
3;
C~
~Sj

PEAT MEHLICH
EXTRACTABLE ZINC

Figure 22. Map of Mehlich extractable zinc from peat sediment at
Loxahatchee NWR.
Fiur 22 a fMhihetatbezn rmpa eieta
Loaachee NWR.

elevated concentrations certainly affect plant and animal commu-

nity structure. Distribution patterns of certain nutrient/ion

species (aluminum Figure 18, calcium Figure 21) show a greater

north-south gradient when compared to the canal-interior gradi-

ent.

Comparing the distribution of conservative ions (ex. chlo-

ride Figure 14, magnesium Figure 17) and nonconservative nutri-

ents (nitrate Figure 15, phosphorus Figure 16) gives information

on the extent that canal water moves into the marsh and the

ability of the marsh to act as a filter. Conservative ions act

as tracers showing the extent canal water moves into the marsh.

For example, elevated concentrations of chloride and magnesium

(Figures 14, 17) extend further into the marsh than do nitrate

(Figure 15) or phosphorus (Figure 16). If the capacity of the

sediments to remove nutrients decreases, the zone of elevated

concentrations will increase and the distribution patterns of

conservative and nonconservative ions will become more alike.

Water Quality and Nutrients

Prior to the alteration of the Everglades, the majority of

water and nutrient input came from rainfall. The poor quality of

water that drains from the EAA fields is well documented (Parker,

1955, Mattraw 1973, McPherson 1973, Waller and Earle 1975, Lutz

1977, Millar 1981, Swift and Nichols 1987, SFWMD data). Nitrogen

and phosphorus concentrations of water delivered to the refuge

varies considerably. A time series graph of nitrogen and phos-

phorus for the S5A and S6 discharge for the years 1974 through

1985 shows very sharp peaks (Figure 23, and 24, data from SFWMD

Phosphorus Inflow
S5A Moving average
0,5

0,45

0.4

0.35

0.3
2 0.25-
0.25 -

S 0.2 -

0.15 -

0
0. 1

74 76 78 80 82 84 86

Date

S6 Moving average
0.45

0.4 -

0.35 -

0.3
CL
0.25
0.

0.2 -
8

0.15

0.1

0.05

0
74 76 78 80 82 84 86

Date

Figure 23. Phosphorus inflow at S5A and S6 pump stations, moving
average 1974-1986.

Nitrogen Inflow

74 76 78 80 82 84 86

Date

56 Moving avenge

I I I I I I I I I I I I I
74 76 78 80 82 84 86
Date

Nitrogen inflow at S5A and S6 pump
average 1974-1986.

stations, moving

Figure 24.

smoothed with a weighted moving average) with an apparent in-

creasing trend through time. These concentrations are high

compared to rainfall (generally estimated to be .03 to .05 ppm

phosphorus). Flow weighted average concentration for phosphorus

for the S5A and S6 were .190 and .119 ppm respectively.

Areal loading rates of nitrogen and phosphorus for the three

WCA's and Lake Okeechobee are shown in Table 12. Phosphorus

loading on an areal basis for the refuge is similar to Lake

Okeechobee while nitrogen is more than 3 times greater. Lake

Okeechobee is considered to be eutrophic.

From 1979 through 1988, atmospheric deposition accounted

for 25% of the P and 15% of the N entering the refuge compared to

75% of the P and 84% of the N entering via S5 and S6 combined

(Table 13, Millar 1983, SFWMD unpublished data). A comparison of

the contribution of nutrients with the contribution of water

emphasizes the influence of the EAA drainage on the nutrient

budget of the refuge. Atmospheric deposition accounted for 54%

of the water but only 25% of the P and 15% of the N for the years

1979 through 1988. A comparison of the contribution of water

and P by atmospheric deposition and S5+S6 for each of the years

1979 through 1988 makes this point (Figures 25 and 26). Water

inputs from atmospheric deposition were generally greater than

from S5-S6 but P inputs were always much less. Note that the

inputs of both water and phosphorus via atmospheric deposition

were fairly constant. Phosphorus inputs through S5-S6 showed

greater variation than the water inputs, suggesting that water

Table 12. Areal loading rates (g/m2/yr) for the three water
conservation areas and Lake Okeechobee (Data for WCA's from SFWMD
and from Federico et al. for Lake Okeechobee).

Lake
WCA-1 WCA-2A WCA3A Okeechobee

Total Nitrogen 8.56 8.11 3.51 2.64

Total Phosphorus 0.23 0.25 0.14 0.28

Table 13. Average annual phosphorus, nitrogen, and water budget
for years 1979 through 1988.

INPUT TOTAL P TOTAL N VOLUME
(METRIC TONNES) (METRIC TONNES) (ACRE FEET)

S5A 73.7 (57%) 3053 (62%) 314,754 (30%)

S6 23.0 (18%) 1086 (22%) 156,698 (15%)

ATMOSPHERIC 32.4 (25%) 759 (15%) 560,533 (54%)

TOTAL 129.2 4898 1,031,985

OUTPUT TOTAL P TOTAL N VOLUME
(METRIC TONNES) (METRIC TONNES) (ACRE FEET)

S39 5.5 (8%) 230 (11%) 77,778 (7%)

S10'S 59.1 (92%) 1814 (89%) 371,592 (35%)

E.T. 619,214 (58%)

TOTAL 64.6 2044 1,068,584

RAN P LOADING AND WATER INPUT

7 81
N N N7
x x7 x

79 so 8I

82

= PHOSPHORUS

t N

N
N
N
X N 17

84 85 86 87
VOLUME

Figure 25. Atmospheric phosphorus loading (metric tonnes) and water
input (ten thousand acre feet), 1978-1988.

S5A+S6 P LOADING AND WATER INPUT
150 -
140 -

130-
120 /
110 / /
100 /
90/ /

70

| o / \ ,..,
z, 60

30
0-

20
10-
0 /

79 sO B1 B2 03 84 85 86 87 88

PHOSPHORUS 1[ VOLUME

Figure 26. S5a and S6 pump stations phosphorus loading (metric
tonnes) and water input (ten thousand acre feet),
1979-1988.

routing management decisions tied to weather patterns play a role

in the export of nutrients from the EAA.

Water Oualitv Maps

Water quality data was obtained from the SFWMD for 35 sam-

pling stations located in the refuge (Figure 27). These sampling

station were grouped based on location into inflow, outflow,

canal, interior and transition sites. Inflow sites were at the

S5A, S6, Acme-1, Acme-2, and L-7 (now defunct) pump stations,

outflows at the S39 and SlO's (Figure 27). Interior sites are 1-

3, 1-5, 1-9, 1-8, 1-7, 1-11, and 1-13. Transition sites are 1-1,

1-2, 1-4, 1-10, 1-6, 1-12, 1-14, 1-15, 1-16. Wet season (June to

November) and dry season (December to May) averages for each of

the five groups were calculated for specific conductivity,

chloride, pH, total phosphorus, sulfate, total nitrogen, and

alkalinity (Table 14).

The average wet/dry season data for each site was surfaced

into a GIS map database using ERDAS software utilizing an inverse

distance weighting algorithm. The surfaced GIS water quality

maps were then transformed into the format used by the SURFER

software package for plotting in a three-dimensional format to

show spatial distribution over the refuge. These results are

shown in Figures 28-41.

These spatial plots (Figures 28-41) are excellent approx-

imations of the extent to which pumped inflows from the agricul-

tural drainage are distributed over the refuge. There is a large

central core area of water in the interior of the refuge whose

nutrient composition is characteristic of a rain water and

LoxchaLchee
NIaLional
WiIdlife
ReFuge

S6

\/
CAl 3. 1-1I /CAl-1

1-12

Vater Ouality CA -20 1-13 1
CAl-06
Sites \ n

S10CI 1
Si0A s39

Figure 27. Map of water quality sampling sites.

Table 14. Wet and dry season averages for site groups.

DRY SEASON
Inflow Canal Outflow

Transition Interior

1310
7.3
0.14
72.8
210
4.33
284

817
7.4
0.078
30.3
126
2.63
200

518
7.4
0.062
35.8
126
2.77
196

595
6.5
0.030
18.7
96.6
2.89
135

127
5.9
0.019
7.3
25.5
2.86
22

Conductivity
pH
Total P
Sulfate
Chloride
Total N
Alkalinity

Conductivity
pH
Total P
Sulfate
Chloride
Total N
Alkalinity

1328
7.3
0.12
95.5
204
5.31
284

1384
7.3
0.116
71.6
191
4.99
273

1119
7.3
0.097
79.2
158
3.86
245

766
6.7
0.031
29.1
1206
2.88
172

139
6.1
0.033
11.5
25.8
2.61
20

WET SEASON
Inflow Canal Outflow Transition Interior

SPECIFIC CONDUCTIVITY

\\

Figure 28.

Map of wet season specific conductivity in surface
water at Loxahatchee NWR.

"(\ cr>~D

cD-
3`
,3

SPECIFIC CONDUCTIVITY

CP

Figure 29. Map of dry season

water at Loxahatchee NWR.

specific conductivity in surface

5c7

Figure 30. Map of wet season pH in surface water at Loxahatchee
NWR.

DRY SEASON pH

3-'

Figure 31.

Map of dry season pH in surface water at Loxahatchee
NWR.

NET SEASON
TOTAL PHOSPHORUS

\x

e ,

Figure

32.

Map of wet season
Loxahatchee NWR.

total

phosphorus

in surface

water

at

CD~
~3`

DRY SEASON
TOTAL PHOSPHORUS

\ '

Figure

33.

Map of dry season total
Loxahatchee NWR.

phosphorus in

surface water

at

----.M

, mq

NET

SEASON SULFATE

rNq\t
(Db

&&

Figure 34. Map of wet season sulfate in surface water at
Loxahatchee NWR.

DRY

SEASON

SUL

,

FATE

S7D

Figure 35.

Map of dry season sulfate in surface water at
Loxahatchee NWR.

-'IV'

cD
cr>'
'5
8

NET

SEASON CHLORIDE

Figure

36.

Map of wet season chloride in
Loxahatchee NWR.

surface water

at

a~s
3`
C~
8

8.
8

3'
3
1

DRY

&e
&3

KNT~
22

Figure

37.

Map of dry season chloride
Loxahatchee NWR.

in surface

water

at

~1_ ___ _____

CAD
0

WET SEASON
TOTAL NITROGEN

0,9
-^ ^
8

Figure 38.

Map of wet season total nitrogen surface water at
Loxahatchee NWR.

7 (P
cD:
3'

- I --

DRY SEASON
TOTAL NITROGEN

-p

e~ /5

rCb
5D

Figure

39.

Map of dry season total
Loxahatchee NWR.

nitrogen

in surface

water

at

8,

&9

Q(j

Figure

40.

Map of wet season alkalinity
Loxahatchee NWR.

in surface water

at

8~ ~D
C3
8

600

Figure 41.

Map of dry season alkalinity
Loxahatchee NWR.

in surface water

at

CI~

(D
3'
3
1

atmospheric deposition source. This area is surrounded (except

to the south during the dry season) by water characteristic of

the pumped inflows from the perimeter canal. This is a pattern

that is exhibited for the distribution of all the conservative

constituents in the refuge's waters (Figures 36,37 and 40,41).

These plots define the extent to which the refuge is influenced

by the inflow waters. The central core of the refuge, essential-

ly a rainfall driven system is surrounded and isolated via an

abrupt transitional area by a perimeter area primarily associated

with the waters pumped onto the refuge from the adjacent agricul-

tural area.

Conductivity

Field measurements for specific conductivity were not made

on the interior sites or transition sites for the wet season

months but laboratory measurements were made for all of the sites

for all months. The conductivity is directly related to the

total dissolved solids content of the water. Rainfall generally

has a conductivity of less than 100 micro-mhos (mu). The interi-

or sites averaged 126.9 mu for the dry season and 138.8 mu for

the wet season (Table 14). The inflows average 1310.6 mu for the

dry season and 1328.5 mu (Table 14) for the wet season. The out-

flows tend to be lower in the dry season (518 mu) and higher in

the wet season (1119.2 mu). The transition zone also shows this

same type of influence with higher values (766.3 mu) in the wet

season and lower value in the dry season (595.1 mu). The canal

average for the wet season (1383.9 mu) is similar to the inflow

values (Table 14) and the dry season average (816 mu) is lower.

The outflow average for the wet season (1119.2 mu) is slightly

lower than the average for the canal sites while the dry season

outflow average (518 mu) is even lower than the dry season canal

average. This indicates that the outflows are dominated more by

rainfall in the dry season and by pumped inflows during the wet

season (Figures 28 and 29). Note the relative low values in the

center of the refuge and the effect of the canal on the distribu-

tion pattern. The extent of the canals influence into the center

of the refuge may be over extended due to the spacing of the

sampling sites. The fact remains that the canal water does

affect water quality towards the interior of the refuge as evi-

denced by the higher values around the perimeter of the refuge

during the wet season. The averages for canal sites on the lower

east side and structures S10A and S10C during the dry season are

lower than the rest of the canal sites indicating dilution of the

dissolved solids in the canal water at these sites by interior

water during the dry season.

pH

The interior is primarily rain dominated with low pH

averages for wet (6.1) and dry (5.9) seasons. The transition

zone averages wet (6.7) and dry (6.5) are intermediate to the

canal and interior sites. The influence of the canal on pH in

the refuge is illustrated in the spatial distributions (Figures

30 and 31). A similar pattern of lower pH in the center of the

refuge was noted by Swift and Nicholas (1987).

Phosphorus

Total phosphorus values ranged from a mean of 0.019 ppm in

the interior to 0.141 ppm for inflow sites during the dry season.

These values represent the maximum range for phosphorous concen-

trations for all seasons. The fact that the range maximum occurs

during the dry season is indicative of the spatial loading pat-

tern (Figures 32 and 33). In interior sites, less phosphorus

input occurs during the dry season and concentrations are lower

compared to the wet season. On the other hand, higher values of

phosphorus in inflow sites during the dry season show the lack of

dilution that occurs during the wet season and may also be indic-

ative of greater soil oxidation in the EAA during the dry season.

Outflow site concentration was lower than inflow and canal sites

for both wet and dry seasons (Table 14) reflecting probable

phosphorus uptake by the marsh.

The consistent pattern is low values for the marsh interior,

again reflective of an atmospheric source, and highly elevated

values in the peripheral margins of the marsh associated with the

perimeter canal and its pumped inflows.

Sulfate

Sulfate concentrations range from a mean of less than 10 ppm

in the interior sites to greater than 90 ppm in the inflows

(Table 14). Sulfate is an important nutrient in aquatic ecosys-

tems subject to anaerobic conditions. Under these conditions

sulfate can be reduced to sulfide, which has known toxic effect

on certain plant species (Richardson et al. 1981). Mean values

for all sites is greater during the wet season (Table 14).

Chloride

Chloride, another conservative element when compared to

nitrogen or phosphorus, also serves as an indicator of the

extent canal water extends into the marshes of the refuge (Fig-

ures 36 and 37). Mean inflow chloride concentrations for both

wet and dry seasons are more than eight times that of the interi-

or sites and approximately 2 times the transition sites (Table

14). This pattern in the conductivity (Figure 28 and 29) indi-

cates the source and extent of flows associated with canal water

and the marshes of the refuge.

Total Nitrogen

The range of nitrogen values is fairly small, 2.5 ppm in the

interior sites to 3-6 ppm in the canal, inflow and transition

sites (Table 14). Some seasonal variation occurs with higher

nitrogen concentrations during the wet season for inflow, out-

flow, and canal sites (Table 14 and Figures 38 and 39). Interior

and transition sites have little or no variation from season to

season.

Alkalinity

The alkalinity (expressed as ppm calcium carbonate) is low

in the interior sites and high in the inflows (Table 14) and

canal sites (Figures 40 and 41). These values correspond to

surficial drainage in the canal waters and rainfall in the inte-

rior. The pattern here is similar to that of chloride and

specific conductivity with low values in the transition zone

during the dry season and higher values during the wet season.

Grid Survey/Topoaraphic man

The development of an accurate topographical database was

necessary to aid in comprehending the hydrology of the refuge,

particularly as it relates vegetation changes to water levels and

hydroperiod. In order to achieve this end a comprehensive grid

survey of the refuge was completed during early January, 1987.

Elevations were determined by measuring the water depths at

all of the grid locations and then subtracting from a stable

water level. In order to have a stable flat pool condition of

water on the refuge, the U.S. Army Corps of Engineers and the

South Florida Water Management District held water at the 17 foot

level in the refuge during the week that the grid survey was

being conducted.

Topographic data was gathered at a resolution of approxi-

mately 1 minute of latitude and longitude (Figure 12). Data

recorded at each site included water depths in each community

type at every site, percent cover by each vegetation class and

latitude and longitude coordinates from a LORAN receiver which

had been compensated to a known survey benchmark. The vegetative

cover was classified as either aquatic, wet prairie, sawgrass,

brush, cattail, open water or tree island. Water depth measure-

ments were made in all of the available classes except tree

island.

Data from the grid survey was used to generate a topographic

map for the entire refuge within the levee (Figure 42 and 43). A

weighted average depth at each point was calculated by multiply-

ing the depth in each community type by the percent cover

ELEVATION

F 4
ap

Figure 42. Topographic map of Loxahatchee NWR.

Topogrophic

LoxmhoLhee
NaL onol

Refuge

Produced

Floridn
Coop Fi5H
& WildliFe
Re search
Unit
1989

Figure 43. Contour Map of Loxahatchee NWR.

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
27	html_echo_mainwriter.add_text_to_page	Finished reading and writing the file