Title: Oriigin of Surface Soils Distribution
Full Citation
Permanent Link: http://ufdc.ufl.edu/WL00003017/00001
 Material Information
Title: Oriigin of Surface Soils Distribution
Physical Description: Book
Language: English
Publisher: Golden Gates Estates Study Committee
Spatial Coverage: North America -- United States of America -- Florida
Abstract: Richard Hamann's Collection - Oriigin of Surface Soils Distribution
General Note: Box 12, Folder 4 ( Golden Gate Estates Redevelopment Study - Phase I - 1975 ), Item 11
Funding: Digitized by the Legal Technology Institute in the Levin College of Law at the University of Florida.
 Record Information
Bibliographic ID: WL00003017
Volume ID: VID00001
Source Institution: Levin College of Law, University of Florida
Holding Location: Levin College of Law, University of Florida
Rights Management: All rights reserved by the source institution and holding location.

Full Text


Catastrophic Restructuring Hypothesis (The Cocohatchee Event)

A working hypothesis has been outlined to help explain
the origin and present distribution of surficial sands of
western Collier County. This hypothesis holds that at.some
time in the past, probably before the coming of Europeans to
Florida, a severe hurricane caused a major break in the high
dune north of Naples which is now the outlet for the Cocohatchee
River drainage.
The genesis of the hypothesis grew out of our
attempts to determine the pre-development limits of small
stream watersheds of the west and southwest coastal areas
of the County. Part of this watershed delineation process
involved mapping, in a simplified manner, certain surface
soils (i.e. St. Lucie fine dune sands = highest coastal
elevations;,Keri fine sands = highest sand deposits of the
upland inland from the coastal dune; Immokalee fine sands =
middle elevations of the coastal dune system; Broward fine
sand, shallow phase = low elevations; and cypress swamp)
which were presented using the colors red, gray, pink, orange
and yellow, respectively (Figure 7).
When this color map was being prepared several
features emerged which were not readily discernible on
the "Leighty" soils maps. First, there is an enormous
sand delta (pink) inside the mouth of the Cocohatchee River.
Second, there were numerous caprock exposures just east and
south of this delta (black dots). These two features
suggested, respectively, deposition of the delta fan by
water rushing inland from the Gulf, and erosion of the
original sand burden from the old caprock east and south
of the delta. The common character (i.e. Immokalee fine
sand) in the main mass of the coastal dune and the interior
delta deposit seemed to us to further support the broken-
dune idea.

T 35 J

; __ ~_ _~____11l___r,~ ~~

Figure 7. Distribution of surface soils.

T 36

_ _IY_ ~ _ji_____~_ ~l~--^-RIl-~1~.1--.1-11~11 ~i-Tii

Following this storm event hypothesis farther, our
eyes were drawn to the series of linear deposits, also of
Broward fine sands shown in pink, along the 15 foot contour,
a distance of 7 to 9 miles east of the present mouth of the
Cocohatchee. These deposits trend southeast for a distance
of about 10 miles where they tend to disappear for an interval
and then occur again in the southeast.
If such a severe storm event did occur it is
likely that the first rush of water through the broken
dune barrier would carry dune sand with it until velocity
was lost or until it encountered an obstacle of sufficient
magnitude to break the rush and cause sand deposition.
Behind this, and following the initial break-through,
successive quantities of inrushing water might begin to
erode the upland soil over-burden, mix it with marl mud
of the Gulf shelf and, carrying this burden, move inland.
At this point in our reasoning, the diagnostic
feature of Keri sands (i.e. that they contain marine marls
in bands and lenses) led us to examine those deposits. They
are remarkably restricted in extent and tend to form several
interesting patterns. The first observation, which has
pertinence to the hurricane theory, is that they lie east
of the unbroken stretch of coastal dune as though in a lee
situation. This is the situation where one might first
expect a diminished wave force and settling out of the
water-borne sand load. Later observation showed that
there is much rock near the surface in this area several
feet above the rock outcrops in the delta region. The
linear nature of the main east-west Keri sand beds and
the rock there suggests that there may actually be an
ancient "reef" upon which the Keri sands were deposited.
Examination of the other Keri sand deposits shows
a 6-mile long curving archipelago of these deposits anchored
on the main east-west ridge. This curves northeast and then
north to terminate at the 15-foot contour due east of the

T 37

piouth of the Cocohatchee. This was taken to be a series of
;deposits left by a slower-moving body of water, more-or-less
.confined within the newly scoured basin north of the main
"reef". Under such a condition there could have been a
rotation of basin flow in a counter-clockwise direction
Smith Gulfward return flow as the storm abated. It will
,be noted that the main direction of deposition of the delta
deposit was southeast. Evidence for such a return flow eddy
was found in the presence of an-atoll-like or circular deposit
Fof Keri sands just inside' the Cocohatchee mouth north of the
Delta which, by its configuration, suggested deposition in a
oocal current vortex.
If the reader turns his attention next to Keri
kiand deposits south of the main east-west ridge he will note
four additional linear aggregations of such deposits all
gunning in roughly north-south direction. The final feature,
allied in our mind with the four north-south sets of Keri
and deposits, lies about three miles east of the eastern
erminus of the main ridge. That latter deposit, also on
High rock base, is terminated on the south by a narrow
ail lying parallel to the "mega-ripples" (in pink) of
mokalee sand and with the four short ridges farther west.
uth of the southern end of these fine Keri sand features
e deposits of Keri sand become noticeably-less extensive.
The reasoning used to explain this decline in Keri
nd in the south was as follows. Once the storm wave
cuntered the solid rise at the 15-foot contour it was
erted south and southeast. This put the moving water in
ee situation relative to the intact coastal dune and
ld have forced it over the old reef'at elevations (present-
*F*) of 9 to 11 feet where most of the marly sand was then
ped. Following this overtopping the water would have
to pick up speed in its southerly move, down-gradient,
| to the sea. In doing this, erosion once again began to
ve sand from off the gently sloping cap rock. As it

T 38

I_1~1I_~ .___Ll_.li____l--- __.1.1 ___.__II

receeded it left vast areas with little or no sand overburden
which resulted in the present distribution of Broward fine
sand, shallow phase, shown in orange.
This leads us to the final observations which have
to do with large areas of Immokalee fine sand in pink and
several peculiar circular or crescent-shaped features shown
in red which lie generally along the present Tamiami Trail
east of Naples. The pink area may be the erosion reduced
remnant of the coastal dune line as it existed at the time
of the storm. It still forms a noticeable ridge feature
which was exploited first by road builders and now by real
estate development. At several places deposits of the high
St. Lucie fine sands remain, as shown in red. Their circular
shape was taken to represent erosion re-deposition of sand
carried by the outward-rushing water as it moved seaward.
East of the eastern edge of this figure there are similar
features, all appearing to support the idea of a powerful
down-hill rush of water. These features stop in the vicinity
of the present-day Collier Seminole State Park.
Finally, the present-day low-lying swamp and
island mass from Naples to beyond Addison Bay does not seem
to have been deposited by longshore currents or by swamp
depositional processes. Its jumbled character along with
several fairly substantial river channels (e.g. Stopper
Creek, Henderson Creek, Macalvane Bay, and the Addison Bay
system) are quite different in character from true sinuous
marsh streams farther east. This jumble of high and low
sand deposits were taken to be the result of the deposition
of Cocohatchee water-borne sediments in an outwash area
among old Pleistocene dunes.
Dr. Frank Craighead, Sr., in a consideration of
this hypothesis, recounted the collection of several samples
of wood from under deep sands east of Naples, generally in
the area we suggest may have been inundated by the seaward

T 39

_ _

moving wave. Certainly, such a natural event could have buried
forest remnants as it eroded and re-deposited sand. Although
the wood samples have not yet been "aged" by C14 dating
techniques, this process may place a date on the event which
buried the wood deposits. We deeply appreciate his interest
in this hypothesis.

Long-Term Deposition

Recognizing that the foregoing hypothesis might be
erroneous, we took it before an assemblage of geologists at
the University of Miami for review and criticism. Out of that
came an alternate hypothesis which is summarized as follows.
At some time in the past, when sea level stood higher
than now, the mainland shore of the Gulf of Mexico was at the
present 15 foot contour. Seaward from that for a distance of
7 to 9 miles there was a shallow bay separated from the open
Gulf by a barrier island of old dune sands.
This bay would have had a nearly level bottom except
where the east-west trending "reef", now shown by the Keri
sands (analagous to the Featherbed Bank of Biscayne Bay),
divided the bay into a northern and southern basin complex.
The northern basin would have contained the present watershed
areas of the Imperial and Cocohatchee Rivers. The southern
basin would have contained all areas on the south, seaward
of the 15 foot contour out to the old Marco Island dune system.
Inside each of the river mouths, which were then tidal passes,
there would have been extensive flood tide deltas. Strong
support for this theory of origin of the interior sand delta
was subsequently found in Dean and Walton (1975: pp. 131-133)
who described sand transport near inlets such as Redfish Pass
in Texas which produced a near-mirror image of the Cocohatchee

T 40


The geologists had no explanation readily available
for the general lack of sands in the southern portion of the
bay basin but could explain the marly content of Keri sand
as marl of bay origin having been deposited on that mid-basin
bank in an area of low current velocity. The Featherbed
Bank in Biscayne Bay has a considerable marl content for this
reason. Finally, they postulated that the circular deposits
of Immokalee fine sand lying along the Tamiami Trail are
"flood tide" deltas. A model of such a delta is beautifully
shown in Hayes (1975, p. 17). The only problem we have with
this is that the inlet would have been seaward of the open
side of the crescent and there is little sign of inlet
configuration seaward from these features at present. In
fact, there is a jumble of very substantial sand dunes
seaward of the western-most of the two "deltas".
Finally, the geologists consider the "mega-ripples"
shown in pink along the 15-foot contour (Figure 7) as normal
offshore bar formations. We could subscribe to this except
that they are so localized. If, as is undoubtedly the case,
sea level receeded slowly down the land slope to the present
position, why are there no similar "ripples" in the inter-
vening space? Certainly there was a sand supply to produce
a continuous progression of such formations out to the present


However the present distribution of sands was formed
is not as important as what they illustrate in the context of
water management. The Keri fine sands are the best water
ponding soils due to their high marl content. St. Lucie fine
sands are drought sands with extremely rapid internal drainage
so that canalization can over- rain such areas rapiazy.-
Immokalee fine sands, with hard_anu._~,.fet be the
surface in the undrained state, are the best agricultural soils.

T 41

MM --- __1_ -

It is our contention that the sands are enclosed in
more-or-less discreet basins formed by undulations in the
underlying rock and these sands form, especially in the
higher portions of the Golden Gate Tract, the main storage
reservoirs for fresh water which led to the present living
systems. Further,, there may be poor vertical communication
between these sand beds and the underlying aquifer in many
areas. The rationale for these statements is given in
following sections.

T 42


Drainage Basins

The most detailed study dealing with watershed
characteristics was done by the original drainage engineering
team on the Golden Gate Tract (Anon., 1962; Anon., 1965) and
shows the closest approximation available of the original
watershed condition (Figure 8). Working with the map and
accompanying engineering data we concluded that this study was
exceptionally useful, but our own observations in the field
suggested certain points of disagreement. This necessitated
a re-evaluation of the Leighty, et al. (1954) soils maps and
a careful evaluation of our own data on soil thickness, flood
depth as indicated by water marks on trees, and more careful
attention to zones of rock exposure or near-exposure.
One part of our close "second look" involved mapping
of several soil associations in the western part of the County
to re-evaluate the size of small watersheds. This was
justified on the basis that, should re-design of the Golden
Gate canal system be considered then it would be ecologically
sound to try and duplicate the original runoff volume as
closely as possible.
The re-mapping involved isolating five soil associations
plus bedrock exposures in the watershed areas. The soils chosen
were St. Lucie fine sand, shown in red (Figure 7); Immokalee
fine sand, shown in pink; cypress swamp soils, shown in yellow;
Keri fine sand, shown in gray; and Broward fine sand, shallow
phase, shown in orange. Bedrock exposures are shown as black
The St. Lucie and Immokalee soils were chosen because
they are of coastal dune origin and reach maximum elevation
along the present coast where they tend to form a confining
boundary against the upland soil bodies. Keri fine sands

T 43

_______1_^__1___ _I

Figure 8. General topography and extent of drainage
basins of central and western Collier
County. (After Smalley, Wellford and
Nalvern, 1962)

T -44
r"T 4


( Y/- ) \ \J

10 / /


,/.. /

\ ~~c i\-^ /I^^-A
^X ,' ^

^^^ /

1~1___ II ___~__I __ _

were chosen for several reasons. They are among the highest
sands in the western Golden Gate Tract and they are said to
be good water-ponding soils because of high marl content.
This is borne out by the abundance of cypress associated with
these sandy deposits which may exceed 8 feet in thickness in
some areas. Broward fine sand, shallow phase, soils were
chosen because they indicate the presence of bedrock very
near the surface, usually within 8 to 12 inches. Their
distribution has proved to be useful in locating minor
drainage divides between the small watersheds.
When produced, the map quickly emphasized several
features which seem useful in outlining watersheds. The most
startling result was the pattern produced by the Keri sands.
The reader will quickly note an east-west trending ridge (also
shown as line C-C1, in Figure 9) nearly 5 miles long. Beyond
that there is a short gap before another section of Keri sands
begins. This extends slightly northeast before curving south
(line D-D1) for a distance of nearly 9 miles. Other Keri sand
deposits also are oriented in linear fashion (F-F1, G-G,, H-H1)
running roughly north-south from the main east-west ridge.
Another alignment (line E-E1) is anchored on the main east-west
axis but this latter series runs northeast to north in a
curvilinear fashion. The final interesting aspect of the Keri
sand deposits can be seen in the large, jumbled mass of deposits
adjacent to the pink Immokalee sands of the coastal dune system.
The Naples Bay headwater cypress occupies a narrow trough
between the two main soil masses. Before drainage rain water
accumulated in this area and ran in a radial fashion north,
west and south.
There are, of course, other sands in the white areas
' between the Keri sand deposits, chiefly Arzell fine sands,
I which are often more than 50 inches deep but are highly
Spermeable while the Keri sands are not. Thus, we believe

T 45


Figure 9. General topography and extent of drainage
basins of central and western Collier
County (as modified by the authors from
Smalley, Wellford and Nalvern, 1962).
Dashed lines indicate drainage divides.

T 46 -

__ _~__ ___~ ._ __








- 1. -- ?:

the Keri sand ridges act as drainage basin divides in the
absence of rock. Lower down on the land profile, particularly
on the south slope, the near-surface rock indicated by extent
of the Broward fine sand, shallow phase, permits extension of
the drainage divides. Rattlesnake Hammock cypress is bounded
on east and west (lines I-I1 and J-J1 in Figure 9) by two such
Another aspect of the Keri sands in an undrained
state was their ability to pond water in a perched situation.
Thus, early rains quickly accumulated on these highlands,
instituting the hydroperiod earlier than in the adjacent,
better drained, Arzell sands.
It will be noted that our estimation of extent of
small watersheds is somewhat different from the engineering
version (Figures 8 and 9). On the basis of the extent of
areas within ridge systems and extrapolation of these ridge
systems to the coast where no relief is visible we made the
following calculations.
The Cocohatchee Basin contained between 35 and 40
square miles of watershed surface. The Gordon River basin
covered about 6 to 7 square miles. The Rattlesnake Hammock-
Stopper Creek drainage basin covered 20 to 21 square miles,
and the Henderson Creek basin may have covered 30 to 31
square miles. Earlier mapping put the entire Belle Meade
Basin into one unit but we believe that at least half of the
water entering the Belle Meade Basin actually flowed in a
south-southeasterly direction toward the Royal Palm Hammock
Creek area. Presently, farm interests are exploiting this
system by opening ditches in that direction.
Having the acreage estimates, we then re-examined
the soils maps to calculate as best we were able, how deep,
on the average, the sands were in each basin. This was done
with the following assumptions, to determine the probable
water storage capacity of each basin at saturation, and thus
to estimate how much runoff might be available as "surplus".

T 47

~-~-casm~c-si-- IILI*-ul~-s--~~*mra~---~~sr~-r~--l- ---- ----- r--- ----------- ---~----- ----- ...~,_~


The assumptions used to make these estimates are
as follows:
1. The caprock under the sands is a "perfect seal"
which prevents downward percolation of rainwater.
2. No loss of rain to evapotranspiration during
the period of sand saturation.
3. Complete storage of water (i.e. no seaward
flow via sub-surface routes) until saturation.
4. Twenty percent porosity of all sands within
the basins.
Considering the problem in light of the above, we
estimated the average depth of sand in the Cocohatchee Basin
to be 2 feet, covering an area of 22,400 acres. Thus the
basin contained 44,800 acre feet of sand leaving 20 percent
pore space or porosity. At saturation this system contained
8,960 acre feet of fresh water.
SThe Gordon River Basin had an average depth of sand
of 4 feet over the basin estimated to cover 6,080 acres, or
S24,320 acre feet of absorptive sand. At saturation this basin
contained 4,864 acre feet of fresh water.
The Rock Creek Basin, with an average sand depth of
2 feet over 3,840 acres, contained 1,530 acre feet of fresh
water at saturation.
The Rattlesnake Hammock-Stopper Creek Basin covering
12,800 acres contained sands averaging 2.5 feet in depth for
a total of 32,000 acre feet of sand. At saturation this basin
had 6,400 acre feet of freshwater in storage.
The Henderson Creek Basin is not so easily circum-
scribed but was estimated to cover 30 square miles. The average
sand depth is estimated to be 3 feet. Thus its 19,200 acres of
sand bed held 11,520 acre feet of fresh water.
The combined acreage of the five basins was calculated
conservatively at 64,320 acres. The Cocohatchee Basin contained
35 percent of the total and 26 percent of water storage
capacity. The Gordon River Basin contained 9.5 percent of

T 48


the total watershed acreage but could store 14.6 percent of
the fresh water in the 64,320 acre area. The Rock Creek
Basin contained 6.0 percent of the area and could hold 4.6
percent of the total water. The Rattlesnake Hammock-Stopper Creek
Basin covered 20 percent of the area total and contained 19.2
percent of the fresh water. The Henderson Creek Basin covered
30 percent of the 64,320 acre total of watershed surface area
but could store 34.6 percent of the fresh water. The estimated
storage potential of the entire series covering 64,320 acres
was 33,280 acre feet. Black, Crow and Eidsness (1974) estimated
that the Golden Gate Canal discharged in excess of 290,000
acre feet of water into the Gordon River alone in 1973.
We carried this reasoning one step further to calcu-
late the amount of rain which would be required to saturate
one acre foot of sand having 20 percent porosity. This comes
to 2.4 inches. Thus, 4.8 inches of rain would saturate the
sands of the Cocohatchee and Rock Creek Basins. More than this
would cause surface flooding and immediate seaward flow. The
4 foot deep sands of the Gordon River, on the other hand, would
accept nearly 8 inches of rainfall before producing surface
flooding, while the Stopper Creek-Rattlesnake Hammock and
Henderson Creek basins would have been saturated by 6 inches
and 7 inches, respectively.
Examination of the long-term rainfall records from
Fort Myers (Table 4) showed that only during the months of
June through September would there be sufficient rain to
saturate the Gordon River and Henderson Creek basins. The
other basins would, theoretically, be full and providing
surface runoff for the additional months of May and October.
In June, with an average rainfall of 8.96 inches,
there would be a surplus of 0.96 inches for runoff from the
Gordon River Basin and 1.96 inches of surplus in the Henderson
Creek system. In July there would be 0.68 inches of surplus
in the Gordon River and 1.68 inches in Henderson Creek. In

T 49


August there would be no surplus from the Gordon River Basin
but there would be 0.91 inches from Henderson Creek. In
September there would be 0.58 inches of surplus in the Gordon
River Basin and 1.58 inches of surplus from Henderson Creek.
There would, under these hypothetical conditions, be no more
surplus during the year. This says, therefore, that the hydro-
period is 4 months long in the Gordon River and Henderson
Creek Basins. In the Cocohatchee and Rock Creek basins where
4.8 inches of rain are required for saturation, the hydroperiod
would probably last 6 months in the undrained situation in an
average year.
The conclusion of this line of reasoning is that
there probably never has been major surplus of water in
Collier County in average rainfall years. This ties in well
with the climatic definition of Hela (1952) and the hydroperiod
thus described of from 4 to 6 months matches the personal
recollection of long-time residents who say that the hydro-
period may have varied over much of the .area between 5 and
7 months.
This also is in agreement with our own observations
that surface flooding was of minor importance over much of the
area, ranging between 7 and 24 inches. Finally, to those who
recollect severe flooding and erosion immediately after
early rains let us point again to the water repellent character
of dry sand and the peculiar Lisse effect which can cause
runoff with the water table several feet below the surface. --
Having done these calculations for the small west
and south coastal watersheds, we turned our attention to the
east. Our study of soil depths, vegetation indicators,
published papers on the hydrology of the Naples area (Klein,
1972; Schroeder and Klein, 1961; Hartwell, 1972), and detailed
study of contours, particularly "outrider islands" on the 10
foot level, led us to the conclusion that the Golden Gate
"Highlands", delineated generally by the Keri sand deposits
(Figure 10), was distinctly separable from the large region

T 50

_I _

Figure 10. Presumed location of Fort Thompson for-
mation caprock over Tamiami formation in
the Golden Gate "Highlands".

T 51



--I-- canal with weir

Golden GatekHighlands

o I
I 1
---- ,
.1 f f
I -
t"- !I

I g








II ~ 1


r -~ ... i


to the east. This suspicion was strengthened by Le Grand's
descriptions of karstification, and by the presence of many
well defined cypress strands east of a line drawn between
Bird Rookery Strand and the Tamiami Trail near Collier-
Seminole State Park (Line A-A1, Figure 9). Our own observa-
tions of soil and rock formations in the cypress strands
suggested that these were old rock-cut water collecting and
distribution features which had been filled by sea-borne
Pamlico sediments to a nearly flat surface, but which are
in reality, sand filled streams with up to 6 or more feet
of relief. Furthermore, LeGrand's explanation of the process
of erosion by solution makes it very likely that these old
channel bottoms have cavernous connection with deeper and
mpre extensive undergo .
In the section dealing with geology the reader will
note that the presence of Deep Lake in the east near S. R. 29
is suggestive of very long-term solution to depths of nearly
100 feet. Cavern systems of the magnitude suggested by Deep
Lake must be extremely extensive; certainly extensive enough to
connect with sub-surface areas in the Fahkahatchee Strand.
LeGrand's comment that once begun, solution continues along
established channels, would suggest that solution or karstifi-
cation process is most active under the largest surface
indicators of tributary systems. On this basis, the sector
south of Immokalee and east of line A-Al to at least the Deep
Lake area would, for all practical purposes, be the "leaky
roof" which permits rapid recharge of the aquifer(s) below,
while the area of the Golden Gate "Highlands" seems to be
highly impermeable.
This raised the question, how are the two areas
different? To know all the answers to this question would
require far more time than we can muster, but once again the
soil mantle and solid rock buried under it may give clues.

T 52

~_ _j*/ 1---. __r __

Examination of the contour map of the Golden Gate Tract
o (Figure 5) will show a number of isolated islands of 10 foot
or higher land sweeping westward toward Henderson Creek from
-- a very large southward bulge on the 10 foot contour at the
point where line A-Al of Figure 9 crosses that contour.
Examination of the Leighty et al. (1954) soils maps shows
.these features are covered by shallow Broward fine sands
(cf. Figure 7) where these "islands" are accentuated with
the orange color. These features of the 10 foot contour
are actually expressions of undulations in the solid under-
lying caprock. Looking further at the soils map the reader
will see an extensive chain of similar rock elevations running
north from the main bulge on the 10 foot contour. Thus we
see what we believe is the main barrier between the Golden
SGate "Highlands" in the west and the strand system in the east.
SIf, as we suspect, this ridge is simply a solidified
mud bank of an ancient bay and its caprock is like that
further west where the Keri sands sit, then there probably
are dense clays and marls under the cap. Several feet below
is the Tamiami caprock, which is itself sealed from downward
percolation by dense clay on its top surface. Such a clay
seal, nearly a foot thick and very dense, can be seen on the
Tamiami caprock exposure in the quarry near the junction of
S. R. 951 and the Tamiami Trail.
Passing east from the 10 foot contour bulge one
Quickly drops to the lower elevations (Figure 11) of the
SPicayune Strand where there seems to be no double caprock
system, and the etched rock of the slough is highly eroded
Tamiami limestone. One other bit of evidence suggests that
the low mounds are effective barriers not only to vertical
penetration by water but also to horizontal movement. A
: similar series of rock mounds separates the Picayune Strand
.. from the Fahkahatchee Strand. The sub-division, called
"Golden Gate Gardens" was platted on this mound system.

T 53


_ __ 1__

Figure 11. Reconstruction of approximate ground
contours between the five and ten foot
contours in the vicinity of the
Picayune Strand.

T 54

__ ____~_~ __1_1__1___ ~_______~~~111_~__~ ___

N "Ip

// wm





During the rapid decline in water levels of almost 3 feet
observed during the period October through December 1975
in the Picayune Strand we noted a similar decline all along
Jane's Scenic Drive eastward to the ridge. Upon crossing
the ridge into the Fahkahatchee water levels remained high
until well into February. The orientation of this mound
system is shown by the line B-B1 (Figure 9). We have no
way of knowing how deep this barrier penetrates into the
ground but it does seem to effectively prevent overdraw of
the Fahkahatchee by the nearby Golden gat( aal System.
We are concerned, however, that there may be gaps in this
ridge due either to solution or, more probably, simple
undulations in its height. One such area lies to the
north of the Golden Gate Gardens subdivision. There is
an obvinilQ -- Qr5-d yPrag -- cztvn 'nnnpction between the
Picayune and the Fahkahatchee just north of S. R. 84
(well north of the existing end of the easternmost canal).
Extension of that canal should not be considered unless
it can be established that there is no cavernous connection
or deep, etched channel in the rock ridge at this point
which would lead from the Fahkahatchee directly to the canal.
We have similar concern that there is at least a partial
connection between the Fahkahatchee and the Picayune south
of Jane's Scenic Drive which has permitted accelerated
overdraw by the canal reach between S. E. 118 Avenue and
Lynch Boulevard.
The paradox faced by builders of development canals
in Collier County was clearly stated by Parker (1955, pp. 203-
204) in reference to the Miami area as follows:

"Except when heavy and prolonged rainfall
results in floods, the canal system of the
Miami area is an effective outlet for dis-
charge of ground water; it lowers the water
table rapidly after a short, heavy rainfall.
However, when the Everglades are flooded and
the aquifer is full to overflowing, the

T 55

_ 1_ ____C__^I~I___

capacity of the canals is not great enough
to remove flood waters in time to prevent
damage to crops and structures on low lands.

"Ground water is free to (discharge) through
the bottoms (provided there is not a thick
layer of sediment) and through both sides of
drainage canals cut into the aquifer. Ground
water can discharge directly into .Biscayne
Bay only along one side of the Bay; however,
this movement is not as effective as discharge
into the canals through their highly permeable
cut-rock faces. As an outlet for ground-water
flow, a mile of deepened canal appears to be
more effective than 4 miles of bay shore,"l/

Although we readily admit that there are differences
in the rock of Dade and Collier counties through which the
drainage canals are cut, the rate of drawdown at the end of
the rainy season is similar in each area at 1 to 2 feet
per month.
Accepting the two areas as similar in this respect
we find that the natural freshwater discharge front to the
estuaries of the Collier County watersheds between the
Cocohatchee River and S. R. 29 near Everglades is about 45
miles. There are 183 miles of canals in theGolden Gate
system. Using Parker's calculations these 183 miles of canals
are discharging stored water at the same rate as 732 miles of
natural shoreline, or, theoretically, at 16 times the rate of
the undrained system.

Surface Water Conditions

Although the major part of the Golden Gate Tract
can scarcely be taken as wetlands at present, there are
abundant signs that it was once much wetter. These signs

S-Underlining added for emphasis.

T 56

most commonly take the form of intact fresh water shell
remnants such as the fresh water clam, tentatively identified
as Popenias popei Lea, ramshorn snails, Heliosoma spp., pond
snails, Physa sp., and apple snails, Ampularia depressa, on
the ground surface in what must now be considered as dry
upland. The widespread presence of cypress, pond apples and
pop ash trees are also indicators of formerly wetter conditions
than presently exist in much of the tract and the older trees
(i.e. greater than 50 years of age) show marks on their trunks
and "knees" which formed in response to long-term flooding
of pre-development times (Figure 12).
In some cases, these indicators" can give insight
into the water condition that occurred during the time the
organisms lived. For example, large apple snails and
fresh water clams require at least two years of growth in
natural shallow waters in order to reach adult size. Water-
marks on cypress trees, on the other hand, require much longer
time spans to form, and, where cypress knees are tall and
well-developed (Figure 13) they indicate, in their own way,
the long-term surface water condition which led to their
Even the presence or absence of trees such as pine,
oaks and cabbage palms in an area can tell much about the
pre-development hydroperiod, especially if the trees are of
considerable age.
Organic soils of varying thickness form the last
important indication that there once must have been near-
permanent saturated conditions where such deposits o-ccur,
Our study of all the indicators available points to
three different sets of surface flood conditions which may be
related to the different stages of land development in south-
west Florida. These are: (1) the pre-drainage period, (2)
the period from first drainage activity to just prior to-develop-
ment of the Golden Gate canal system, and (3) the period
beginning with completion of the GAC canal network.

T 57

II_~__~______________lin___ ji~__~ Il____i_ I___~___ I_ ____ _




Dwarf cypress on Keri sand', deep phase,
in the vicinity of Golden Gate Boulevard.
The 7" flood level 'is plainly visible in
line with helmet; the pre-1900 flood'
level at approximately 24" is indicated
by the pointer at center.

Figure 13. Cypress stump in main flow-way south of
Stewart Boulevard. historic flood level
is indicated by poijier at right.

T 58
4 T

Figure 12.

The pre-development condition is the most difficult
to describe because it is the farthest removed in time. We
do not know exactly what date marks the end of the original
condition but speculate that it ended, as did the original
condition in the Lake Okeechobee-Everglades system, in the
decades 1880-1920 when the first major diversion of the Lake
Okeechobee-Caloosahatchee River-Everglades system was begun
by Hamilton Diston and successive canal builders. Accounts
of those early drainage activities in South Florida and their
effects upon water levels and organic soils were given by
Parker, et al. (1955); Parker (1974); Stephens (1974); Tebeau
(1974); and Tabb (1963). Those activities were marked every-
where by a very rapid decline in groundwater by as much as
6 feet, area wide, loss of vast quantities of organic soils
by oxidation, compaction and outright burning, and by wide-
spread salt intrusion along the coast. Similar effects of
first drainage probably have been imposed on Collier County
but, because development was slower in coming to Collier
County, may not have been as severe there until somewhat
later. We have speculated that first deepening and channeli-
zation of the Caloosahatchee River by Hamilton Diston in the
1880's (Hanna and Hanna, 1948, and Parker et al., 1955) marks
the end of the original condition in southwest Florida. Thus,
any indicator used by us to describe this era such as tree
marks, must be found on trees exceeding 100 years of age.
Some of these still persist, but most were logged in the
1940's and 1950's as described in the section on vegetation.
The second period of alteration probably occurred
with acceleration of land drainage in the period 1920 to
1950. Klein, et al. (1970) described these activities as

"Significant development affecting the Big
Cypress began in the early 1920's, when two
major roads were built -- the north-south

T 59

road. from Everglades City to Immokalee",
completed in 1926, and the Tamiami Trail,
completed in 1928. Both were constructed
of borrow material from continuous pits.
adjacent to the roads; the borrow pits
became canals.

"The north-south Barren River Canal, ...,
was the first major drainage canal in the
Big Cypress.

"Drainage was further modified in the late
1950's by construction of the major Turner
River Canal."

The third period of alteration which is still going
on, but which has been marked by a rise in public concern and
official activity to control over-drainage, seems to have
begun with construction of Levee 28, the western boundary of
Everglades Conservation Area 3A, in 1963 and the Everglades
Parkway (Alligator Alley) and its associated canal in 1967
(Klein, 1972). The Golden Gate canal system completion came
in 1969 with opening of the Fahka Union Canal and precipitated
a concern over possible effect on public water supply in
western Collier County (Klein, 1954; Schroeder and Klein,
1961; McCoy, 1962).
Each of these major canals were quickly exploited
as outlets for a vast net of urban and agricultural canals
which greatly accelerated runoff. The detailed chronology
of such a subsidiary network of canals is beyond the scope
of this study.
What were conditions in pre-drainage Collier County?
For part of the story we must, by inference, rely on the
vivid descriptions of drainage effects on the Everglades
where drainage of lands, once subjected to continuous flooding
and attendant accumulation of partly decomposed plant material
called peat and muck, led to rapid and unexpected results.
Parker (1955) described the situation as follows:

T 60


f __


"It is doubtful that the drainage enthusiasts
ever envisioned that, among other results of
their operations, they would induce or cause
(1) shrinkage, compaction, oxidation, burning,
and general subsidence of the organic soils.
This loss is reported by Jones (1948, p. 79)
to be as much as 5 feet over extensive culti-
vated areas. In some places, where the organic
soils were a couple of feet or less in thickness
they have disappeared completely. (2) Develop-
ment of wide, shallow 'subsidence valleys' along
each drainage canal in the muck and peat soils
(Evans and Allison, 1942, pp. 34-36). (3)
Reduce the original capacity of the canals,
thus contributing to floods, slowing down of
runoff, and general canal efficiency. This
resulted from soil compaction and burning,
thus lowering the land surface near the canals
and reducing the vertical height of the banks;
slumping of canal banks; and blocking of canal
flow by weeds, fallen trees and other debris.
(4) Increase frost damage, which formerly had
been held in check by the large body of water
in the Everglades (Clayton, Neller and Allison,
1942, p. 5). (5) Cessation of the process that
had built up the muck and peat soils in the
first place. (6) Changed ecological conditions
seriously affecting wildlife of the drained
areas. This has resulted in species migration
and extinction, or near extinction of others,
one of these is the Everglades Kite, now a
rarity because of the swamp with resultant
destruction of a certain species of fresh-
water snail upon which the kite feeds solely."

It seems unlikely that western Collier County ever
had area-wide conditions conducive to long-term accumulation
.of peat which demands basically a continuous saturation and
anaerobic conditions. John Stephens (1974) provides an
iup-date on the problem of loss of organic soils and stated
that upon drainage, peats tended to shrink with drying alone
t the rate of about 1 inch per year (p. 352). He comments
further (p. 357) on the water holding capability of peats as

T 61


"One of the important elements, seldom
recognized, predominating in the freshwater-
saltwater balance in the lower Everglades
has been the loss of peat soil and the low-
ering of surface elevations in the hinter-

"The original depth of the absorbent peat
mantle before subsidence as shown by a sub-
sidence line survey along the South New River
Canal at Davie, Florida was about 5 feet.
In 1913, the peat surface was approximately
7 ft. m.s.l. In the early 50's, the surveys
were terminated at elevation 2.4 ft. m.s.l.
when the land exposed was bare sand and no
organic soil was left to subside. The impor-
tance of a peat surface mantle is that while
1 inch of rainfall will raise water levels
only 1 inch above a flat land surface, it
will normally raise groundwater levels about
7 inches in peat soil. The exact water
table rise will depend on initial depth to
water table and antecedent conditions.
Weaver and Speir (1960) have shown that on a
receeding water table 30 inches deep in
Everglades peaty muck, the percentage of
soil volume available for water storage
(air space) ranges from 25% in the surface
3-inch layer to 2% in the 3-inch layer
immediately above the water table. The average
pore space for the entire 3-inch profile was

"Thus the peat soils acted as a sponge to
absorb the rainfall, and the 7 to 1 ratio
of water table rise to rainfall exerted a
powerful leverage in maintaining higher
hydrostatic heads of freshwater to combat
sea-water encroachment."

One other statement by Stephens (1974) is helpful
in our search to learn what the pre-development condition may
have been. In talking about the rate of peat formation he
said (p. 356):

"a initial study in 1951 and a more sophis-
ticated one in 1968 was made using radio-carbon

T 62

IP -I- --

age determinations to determine the rate
of peat formation in the upper Everglades
(Stephens, 1956; McDowell, Stephens, and
Steward, 1969). These age determinations
indicate that peat formation began during
late Hypsithermal time about 4,400 years
ago. About 500'to 1,000 years were required
to build up a 3-inch layer of basal mucky
peat, composed of a mixture of marl and
organic matter. Then, about 3,500 to 4,000
years ago, coincident with a rapid eustatic
rise in sea level, plant growths were pre-
served as fibrous peats with little mineral
admixture, and organic soil developed
relatively rapidly. By 1914, peat had
developed to a depth of about 12 ft., which
represents an average peat development of
about 3.3 inches per century. The value of
turning under cover crops (to replace soil
lost to subsidence) is obviously trivial,
since 1 foot of peat requires nearly 400
years to develop, which is lost to oxida-
tion in about 10 years."

In reading accounts of the drainage of the Everglades
(e.g., Parker, 1955; Tabb, 1963) one is constantly struck by
the similarities to the drainage of the Golden Gate. One
example was called to mind when we first saw a waterfall at
the head of the easternmost through canal where it stops at
S. R. 84. A parallel situation on the southeast coast was
described by Parker (1955) as follows:

"W. S. Jennings (1909), in a letter to the
Trustees of the Internal Improvement Fund
of the State of Florida, dated November 19,
1907, tells of his observations during a
visit that he and Governor N. B. Broward
made to the site of the dredging operations:
'The canals reduce the water level from the
surface to approximately 6 feet below the
surface of the ground as shown by water in
the canal, and the land for a mile on either
side of the canal is entirely reclaimed, and
is practically ready for preparation for
cultivation, and the general influence of
the drainage reaches to a much greater

T 63

distance than one mile ... Generally speaking,
the work is a great success .... The reduction
of the water level to 6 feet below the surface
of the ground is all and more than could have
reasonably been expected ... the superintendent
finds it necessary to keep a sufficient quantity
of water in the canal to float the dredge, while
in front of the dredge is water pouring over the
front of the canal and falling 6 feet over a
perpendicular dam to the water level of the canal
and thus going to the ocean.'"

Another statement (Allison, 1956) has great signifi-
cance to the Golden Gate and adjacent lands, even though we
are now dealing mainly with sandy soils in the Golden Gate.

"According to the surface subsidence trends
in our Everglades soils it is only too
obvious that the time must come when the
depth of the peat over the rock will be
too shallow for most cultivated crops. The
most critical time will be, of course, when
the depth to free soil water must be main-
tained for any particular cropping operation
is such that the water tables must take a
position in the rock itself and the direct
contact between it and the remaining muck thus I
largely lost."

It was in this context that we spent so much time
trying to understand the caprock in the function of the Golden
Gate where canal water levels and groundwater has fallen
below the caprock periodically. This results in loss of
the direct contact with the overlying sands so important
to plant communities.
From all the above we conclude that changes due to
drainage were most serious in the Everglades and came much
earlier than similar change in Collier County. For example,
most of the serious decline in water levels had occurred in
the Everglades by 1940 while this probably did not occur in
Collier County until the 1960's. We would surmise that most
of the soil loss in the Big Cypress, whether by oxidation or

T 64



fire, did not occur on an area-wide scale until the 1960's.
Evidence for this is found in the concerns expressed by the
Audubon Society over drainage effects on peat and water
resources of the Corkscrew Swamp Sanctuary in the late 1960's.
If we have interpreted available records accurately
the three hydrologic regimens, representing three downward
adjustments in water table levels of the Big Cypress, including
Golden Gate, must have occurred about as follows:
1. Pre-drainage period -- ended about 1888-1900
2. First drainage adjustment period -- 1900-1950
3. Overdrained condition period -- 1950-present
On this basis, the highest water marks on ancient
cypress trees and their knees, as well as organic soil deposits
or their marks on the same trees if now absent, would probably
be the only indicators of pre-drainage water levels.
Water marks on cypress less than 100 years old,
larger hardwoods, and old cabbage palms, as well as fresh-
water shell material, should be the best indicator of Phase
2 conditions.
Burned out organic soils, displacement of plant
communities in the direction of upland species dominance,
and level of attachment by cocoplum, ferns and other short-
lived small plants on existing large cypress tree trunks
would give the best clues as to Phase 3 flooding conditions.

Pre-Development Water Levels

Although present drainage systems and development
in low ground locations make it unlikely that pre-development
surface water elevations will ever again prevail in western
Collier County it can be instructive to know what those con-
ditions were. Such information tells much about the conditions
required to obtain maximum recharge of shallow aquifers. It
also can explain how plant communities must have been shaped

T 65

by the period of soil saturation called the hydroperiod and it
can indicate what volumes of water were stored on the surface
prior to initiation of sheet flow seaward.
Lacking historical data, we know of only one method
which can give such information, that obtained by studying
"water marks" on the oldest trees available. Fortunately,
most of the larger cypress still remaining are well over 100
years of age, and for the major portion of their lives they
have undergone a natural regimen of flooding and drying which
has left marks we believe are reliable and which constitute an
integrated pattern of wet season growth due to long-term
presence of water around their bases.
We have conducted cross-country transects measuring
the height of such marks above the ground surface at five
different locations and find indications of two different,
long-term high water situations. These must represent the
original condition and that brought about by completion of
most of the major canals in the GAC network.
The five main transects were run in a west to east
direction along Golden Gate Boulevard, Berson Boulevard, 20th
Avenue, S.E., Stewart Boulevard, and Lynch Boulevard.
Stations were one-half mile apart. At each station an
irregular search pattern was made through one or more acres
of forest, and watermarks measured on enough trees to show
the range of local undulations in ground level.
For example, Station 1 off the extreme west end
of Stewart had a small cypress strand (greatest depth
indicated) bordered on the north by grass prairie with
stunted cypress and pine (least depth indicated). The depth
range indicated was 17.5 to 39.5 cm (6.8 to 15.5 in.) at
that station.
Taking the 26 stations along Golden Gate Boulevard
we found average depth of pre-drainage flooding to be 20.2 cm
or 7.9 inches, and varied from 0 at three stations where rock
was exposed at the surface, to 55 cm (21.5 in.) in deepest
cypress slough situation.

T 66


Likewise, the Stewart Boulevard Transect produced
an average depth of predrainage flooding of 40.7 cm or 16.0
in., with a range of 0 to 75 cm (i.e. 0-29.5 in.). Interest-
ingly, it was at Station 7 on this transect that we found
a few very old cypress, apparently too short to be noticed
by loggers, whose knees towered above the recent watermarks
prevailing on surrounding trees and on those knees we found
marks indicating long-term flooding to depth of 127 cm
(50.0 in.). Such knee growth and watermark formation may
attest to a much earlier high water stage which prevailed
at the time of formation of the major cypress resource which
was cut out in the 1940's. The two high water marks may also
indicate wet cycle and dry cycle conditions prior to any
The Lynch Boulevard transect produced an indicated
average predrainage depth of flooding of 34.1 cm or 13.5 in.
with a range between 0 and 75 cm.
We stress the predrainage character of these marks
because, to the best of our knowledge, water does not now
stand on the surface long enough to foster marks on such
slow-growing trees as cypress. It is still possible for water
to rise to such heights after locally heavy rains but it does
not remain long enough to make water marks.
While the evidence suggests that peats were never
as important to the ecosystem of Golden Gate as in the Everglades
we believe they were important in large cypress stands such as
Bird Rookery and Corkscrew Swamp and that drainage around the
periphery of such swamps will result, inevitably, in eventual
loss of organic soils and their water storage potential.
The watermark information also suggests that there
has been considerable change in hydroperiod. The pre-drainage
hydroperiod was of sufficient duration to produce peat or muck
in all the areas now marked by main sloughs and flag ponds.
This suggests that they seldom dried completely, although the
shallow depths of peats indicate that the basin depth was
insufficient to allow peat development of the order occurring
in the Everglades.

T 67

Salt Water Intrusion

Following completion of the major drainage canal
systems of Dade and Broward counties, but before installation
of salt control structures, serious contamination by seawater
forced abandonment of municipal and private well fields, loss
of coastal agricultural production areas, and many other
problems, which were described by Parker, et al. (1955).
The trend of the salt water to move farther and farther
inland on that coast.was reversed only after the expenditure
of large sums of money on variable-height water control
structures inland to reinstate a higher fresh water head,
and on salt control levees, dams and flood gates near the
seaward ends of major canals. More recent additions to
the drainage canal system in south Dade County, notably the
Canal 111 which was dug to provide barge access to rocket
test facilities located near the entrance to Everglades
National Park, precipitated a crisis over whether it should
be left open, plugged, or provided with a lock to prevent
saline intrusion in the eastern reach of the National Park.
The controversy quickly involved many state, federal and local
agencies, as well as public and private groups and illustrated
the passion that such a drainage feature can arouse.
While Dade and Broward county shallow surface
aquifers differ structurally from Collier County's, we
believe the potential for salt intrusion in Collier County
is severe enough in the karst region to warrant considerable
future study. The papers by Klein (1954), Schroeder and Klein
(1961; 1963) and McCoy (1962) provide the published data base
for consideration of this problem. Klein (1972) gives the
position of ground water contours during the severe drought
of 1971 (Figure 14) and Carter, et al. (1973) discuss aspects
of ground water in the system which may be useful in future
design work to prevent salt intrusion. Finally, the engineering

T 68

Figure 14. Ground water contours during drought
of 1971. (After Klein, 1972)


T 69

__ _

study by Black, Crow and Eidsness (1974) provides data on
weirs now regulating the system. Nowhere could we find
information on the effectiveness of the present weir system
in preventing bypass by water, either salty or fresh, of
weirs in the Golden Gate system.
Salt contamination in Collier County may be expected
from four sources at least. These are (1) sea water intrusion
inland during drought via uncontrolled canals, and thence
into shallow aquifers as happened in Dade County; (2) wind
or tide overflow of low coastal elevations where drainage
and drought lower the surface fresh water levels to near
sea level and fresh water is replaced by salt water; (3)
penetration of artesian salt water, from the deep Floridian
Aquifer, into shallow surface aquifers by way of abandoned
wells suggests that there may be a number of such wells,
now lost and plugged only with a log or rock, which may be
flowing freely into the upper aquifer layers; and (4)
pumping from shallow aquifers which contain large quantities
of residual sea water left in the strata by earlier high sea
stands such as the Pamlico and Talbot Seas.
The only new indication we have seen that salt
intrusion is happening may be observed along the Tamiami
Trail between Ochopee and Collier-Seminole State Park
where extensive stands of cattails have been killed by salt
water north of the road. This kill, if not reversed by
restoration of surface fresh water flooding and control of
tidal penetration via Tamiami Trail culverts, will result
in further inland encroachment of the mangrove forest, a
process that began in the 1930's and 1940's (T. R. Alexander,
personal communication). This is exactly analagous to the
salt contamination of coastal fresh water marl farm lands
in Dade County in the 1940's. Marls, including Ochopee
arls,are extremely "tight" soils and, under unditched
conditions, tend to inhibit direct contact between fresh

T 70

surface water and saline sub-surface water components. They
are equally effective in ponding salt water on the surface
once canals provide direct contact for tides between coastal
bays and the inland fresh-water marshes.
We have noted, with special concern, the strong
inland flow of tide water through the Blackwater River culvert
under the Tamiami Trail. During low groundwater stages this
is a serious potential point source of contamination of sand-
filled basin storage in the southeastern Belle Meade Basin as
well as the southern end of the Picayune Strand. We have also
observed northward flow of tidewater over weir 24 on the
Fahka Union Canal at the Tamiami Trail when water level above
the weir was about 6 inches below the concrete top of the weir.
The top of the weir is set at +2.0 feet (m.s.l.).
We have no evidence that "fossil" sea water is a
problem in agricultural areas but poor water quality in
wells drilled in the Golden Gate "Highlands" were described
by McCoy (1962, p. 51) and further indicate that this is a
"problem area."

"The high mineral content of the groundwater
east of the coastal strip is due primarily to
constituents derived from sea water, in
addition to the calcium and bicarbonate derived
from the limestone in the aquifer. The chloride
content of the water ranges from less than 100 I
ppm to more than 2,000 ppm and may come from
three possible sources: (1) Direct movement
inland from the sea and along tidal reaches of
streams; (2) residual sea water left in the
sediments at the time of deposition or during
former invasions of the sea; and (3) upward
movement of salty water from deeper artesian

In summary, the system seems to have first undergone
marked saline invasion in the 1930's and 1940's, and still
appears to be adjusting to the lowered fresh water head caused
by completion of the Golden Gate Canal net. The adjustment

T 71

1_~~~_ _1~__1_1___ ( ___~

.is toward further landward migration of the fresh water -
palt water interface as indicated by dead fresh water vege-
%ation north of the Tamiami Trail. We would expect that
present agricultural and real estate development in the
Belle Meade Basin will cause further inland salt movement
and, given drought of sufficient magnitude (e.g. 5 to 6 year
duration) may force abandonment of farming activities in
that basin.
We would urge that the Tamiami Trail between Ochopee
and Naples be made an effective salt barrier by use of culvert
* weirs to control water levels above the Trail. Citizens of
Sthe Ochopee area were, we believe, acting in progressive
fashion when they constructed a stone weir in a major canal
just west of the Golden Lion Restaurant.
The threat to water supply seems particularly acute
in the southern sector of the County and deserves special
research consideration by appropriate agencies.

T 72

ll1------------------- *^^^** ----- ------------------- --------l ^



A vegetation map drawn at the scale of 1 inch to the
mile has definite limitations regarding absolute accuracy
and resolution. It is not easy to delimit plant community
margins where species gradually merge, interdigitate, or
become interspersed. This is the case over most of the study
area, exceptions being the margins of the dry prairies and
the tree islands within them. Location, always a problem in
undisturbed ecosystems, was made easy by the extensive
network of roads, old tramlines, and canals. At the same
time this enhancement was negated to some extent because
these man-made structures often obliterated historic and
natural margins. They also tended to control and direct
fires which have locally and severely modified historic
communities and their real margins. The margins of com-
munities represented by lines and the color they delimit
(Figure 15), represent as accurate a presentation of existing
communities as could be obtained under the above conditions
and time restraints. Ground checks were made repeatedly and
revisions followed before the final rendition was made.
From an ecological viewpoint, and for the purpose of this
report, the accuracy is considered fully adequate.
The vegetation map indicates a good vegetational
correlation with the physical features of surface soils and
contours. On a broad scale, three major vegetational zones
can be recognized. The northernmost, around the 15' contour,
extends down to approximately N.E. 26th Street. Here most
of the original pineland vegetation has been removed for
farming. This forest stood on Immokalee fine sands, and no
strong drainage patterns were apparent. Between 26th and
S.R. 84 the dominant forest is pineland-dwarf cypress with

T 73

scattered pine and saw palmetto communities. Except for the
narrow slough between Desoto and Everglades, most of this
extensive acreage is without strong drainage features and
the vegetation is associated with minor changes in soil
elevation. South of S.R. 84, generally below the 10' contour,
the major communities occur in broad parallel bands oriented
slightly west of southward. These reflect a historic
centrally located cypress forest with flanking pinelands.
The former had the largest trees in the deeper sloughs, with
the pinelands occupying higher soil elevations.
It should be noted that the vegetation map legends
do nct have any reference to the cabbage palm (Sabal palmetto),
yet it is one of the ubiquitous plants of the study area,
especially south of S.R. 84. Cabbage palm was not included
as an indicator plant because of its presence in most of
the recognized communities. At present its density ranges
from canopy dominance in some forested areas to rare occurrence
in prairies and ponds. Its dominance and significance are
directly related to the fire history at any given site. It
tends to survive fires when woody species are killed.
Apparently it is undergoing a population explosion and forming
dense stands, especially in forested areas. When one examines
the fire map (Figure 16) the amount of fire damaged acreage is
apparent and this helps one to recognize the importance of
cabbage palms throughout the system. Areas judged fire-free
are limited to the south of Lynch Boulevard, and to a small
area north of Golden Gate Boulevard between 2nd and 6th Streets
(SE). The remainder of the Golden Gate area has been stressed
.by recurrent fires, and is therefore involved in the overall
spread and local abundance of cabbage palm.
The vegetative aspects of farmlands and urban areas
are not discussed in detail. Essentially all remnants of the
original vegetation have been removed from these areas. The
abandoned farmlands now have a native vegetation cover

T 74

L i

representing early stages in secondary succession. The
exact stage or floristic make-up depends on the time lapse
since abandonment. Some fields have reached a stage where
pines are invading. These farm lands will be discussed
later in terms of potential use possibilities.




T 75


Figure 15. Vegetation

Descriptive Legend

Wet prairies, flag and willow ponds:

Communities located on lowest soil elevations and
characterized by long hydroperiods. Soil surface
usually organic, sometimes marl or sand. Trees
(e.g. willow) when present, are of limited size
and shrub-like. Dominant vegetation graminoid
(grass and grass-like) and made up of grasses,
S sedges, cattails, and aquatic related herbs such
as flag, pickerel weed, arrowhead, lizard's tail
and yellow canna.

ied swamp forest:

Replacement community for the original location of
the largest cypress. This bottomland hardwood and
cypress forest is common where the tramline spurs
were aligned and fires have not been severe. Soil
S surface ranges from organic in unburned areas to
sandy-organic or.marl-organic in burned sites.
Dominant tree cover of cypress, maple, oak, ash,
and cabbage palm with an under canopy of young
cabbage palm, wild coffee, myrsine and ferns.
Lower areas, with ash dominating, are the sites
of greatest orchid occurrence. Canopy density
(shade) directly related to fire-history.

Cypress forests:

Growth form large: Canopy ranging from thin to almost
closed, with trees ranging from 30 to 80' in height and
diameter breast high (DBH) from 3" to 12". Cabbage
palms scattered; some hardwood invasion occurring.
Understory variable, commonly dominated by wax myrtle
and Blechnum fern. Woody understory yields to grami-
noids, dog fennel, saltbush and bracken fern, depending
on fire history. Usually associated with the mixed
swamp forests and lower soil elevations.

T 76

r i __j_ __

Growth form small: Canopy as above but with trees
ranging from 5 to 60' in height and DBH from 1" to
9". These trees appear to be growing on slightly
higher and thinner soils than the larger growth form.
Several age classes present but few saplings.

Pine-cypress forests:

Canopy very irregular in height and with many openings.
Cypress very much like that of the small growth form
described above, with the largest growing in heads.
Pines occur interspersed among cypress on higher ground
in a variable fashion; from isolated trees to local
areas dominated by pines. Pines to 60' height and
DBH to 12" are few, and locally several age classes
of pines are common. The general impression is that
pines are invading the cypress, especially on the
higher soils. Understory very variable with cabbage
palm, dog fennel, and wax myrtle related to recent
fire history. Saw palmetto occurs in some of the
highest locations.

Pine-dwarf cypress:

Pines scattered randomly among the more common cypress,
usually forming a very open tree canopy. Pines open
grown 30-35' tall, to 7" DBH with good age-class
representation. The growth form of the cypress, with
their huge buttressed trunks, even on trees of limited
height, characterizes the appearance of the forest.
Associated shrubs, wax myrtle and holly, are dwarfed.
Characterized by a uniform grass cover of maidencane.

Pine-cypress-palmetto mosaic:

Cypress tends to be in local heads among the more
common pines. Cabbage palms of all age classes
commonly dominate sizeable areas. Young pines of
several age classes dominate some sites. In the
pineland proper, saw palmetto and graminoids occur
separately as understory plants. The latter is
commonly muhly and broom sedge. Wax myrtle is a
common shrub throughout, becoming .dense and tall
(to 18') especially near cypress margins.

Pineland-grassland understory:

Pine canopy very open with large trees varying
from absent to scattered. Local areas with young

T 77

I-~--l----li--; -- --r-. .r- ------- ----_I_.~- -.--..----.. ~.--..^---

pines closing the canopy. Pine reproduction generally
good, with several age classes present. Cabbage palms
common with larger palms usually in margins with pockets
of cypress and younger palms common in grassland under-
story. Muhly and broom sedge dominant ground cover in
open areas that are very variable in extent. Where
fires have been infrequent, wax myrtle invasion of the
grassy sites is common. Saw palmetto patches present
but not common.

Pineland-saw palmetto understory:

Largest pines over 65' tall, 14" diameter but very few
and scattered. Pine regeneration limited. Saw
k palmetto verycommon, forming dense cover occasionally
broken by small grassy openings. These openings
characterized by broom sedge and muhly grasses.
Semaphore eupatorium (Eupatorium mikanioides), swamp
fern (Blechnum serrulatum), and milk buckthorn (Bumelia
reclinata) occur in the grass cover. Lyonia (Lyonia
ferruginea), blueberry (Vaccinium myrsinites) and sumac
(Rhus copallina) occur with saw palmetto on the highest
parts of this pineland and are indicators of dry and
acid sands.

Dry prairie:

Graminoid cover relatively dense to 18" tall and commonly
dominated by muhly grass (Muhlenbergia filipes) with
scattered colonies of cordgrass (Spartina bakerii), common
reed (Phragmites australis), and sawgrass (Cladium
jamaicensis) with broom sedge invading where disturbance
has occurred. White top sedge (Dichromena colorata)
and beak rush (Rhynchospora spp.JI common in lower
areas. Common herbs are yellow top (Flaveria linearis),
heliotrope (Heliotropium polyphyllum) and marsh fleabane
(Pluchea rosea.

T 78

_ I_ ____________; ___ I ~_ ____

Figure 16. Severity of fire damage as of April

T 79


s ---- I SR



---- r /:
I 'M An Gate

1 0 1 2 Mi

of l -




EWI~RELQ F DMt b tSStewart-S = iR B=


Historic Aspects of Vegetation

To interpret the several plant communities that are
found within the Golden Gates Study area today, it is
necessary to reconstruct as much as possible the way they
looked under virgin conditions. Then attempts to understand
the existing modified communities can be put into useful
perspective. This,.in turn, requires that the post-1900
history of the area be studied, both from the viewpoint of
natural phenomena as well as anthropogenic activity. There
is, then, some capability for prediction about these communities,
leading to the formulation of vegetation management practices.
Published literature on the virgin condition of
the vegetation is sparse. The earliest published accounts
described the Fahkahatchee rather than the Picayune strand
area (Golden Gate). The brief descriptions of the Fahkahatchee
were from the east side, since early access was mostly limited
to the railroad along the Barron River Canal (S.R. 29).
Davis (1943) in his Florida Geological Bulletin No. 25
"The Natural Features of Southern Florida" treated the
vegetation of t-he entire areas very generally under the
heading of cypress forests. Study of his vegetation map
indicates that the boundaries of the major plant communities
shown were about the same as they are today. The first
published detailed list.of plants in this general region
only appeared when hearings were held in 1972 prior to the
Big Cypress purchase. (Hearings before the Subcommittee on
National Parksl and Recreation; 92nd Congress on H.R. 10410,
13017 and 13115. Serial No. 92-33). The Environmental
Protection Agency itudy on ecosystems of the Big Cypress
swamps and estuaries (Carter et al., 1973) has 21 pages of
plant lists arranged according to habitat. It should be
noted that only in the last 5 years has the ecosystem of
this part of Collier County been systematically studied.

T 80

~yl___l_~l ~__ ~_~_~_ _I _;_s_

This means that the pre-impact status of the vegetation is
not well known.
Another source of information is from early settlers.
Unfortunately this source is limited and, at best, accurate
recollection in old age is difficult. Fortunately Mr.
Randolph Swaim of Lee Cypress still resides in the area. He
was in charge of the lumbering operation which harvested most
of the marketable cypress. Cypress was essentially clear-cut
from the Fahkahatchee and Picayune strands and other smaller
strands from the southernmost locations just north of U. S.
41 northward to the southern boundary of Audubon Corkscrew
Sanctuary. Mr. Swaim stated that the cypress was all virgin,
forming a closed canopy. Oak hammocks were small and confined
to the highest ground, and maple was not common. Cabbage palms
were common but royal palms were mostly restricted to the
deepest part of the Fahkahatchee slough. (A few royal palms
still grow in the southwestern part of the cypress forest
north of the Seminole-Collier State Park). According to
Mr. Swaim the cypress of the strands and heads of the Golden
Gate area was good but not of the quality or density of the
Fahkahatchee cypress. Still, by all standards they were sub-
stantial, and cut stumps persisting in a few locations lend
credence to his statement.
One of the authors (Alexander) visited the logging
operation in the Fahkahatchee on several occasions, and was
able to examine and photograph the forest ahead of the cutters.
These observations of the virgin forest are in agreement
with Mr. Swaim's statements. The only addition is the fact
that the ash and pond apple trees of the deep sloughs produced
the habitat for the best orchid population in the United States.
Associated with these were epiphytic bromeliads, ferns and
peperomias, several of which were, and are, of rare occurrence.
On the southeast flank of the Fahkahatchee the paurotis palm
was common.

T 81


The pinelands to the west of the Fahkahatchee were
heavily logged during cypress cutting and after the cypress
was harvested. The virgin pines were large. Cut stumps in
the range of 2 feet in diameter still exist. It is not
possible to judge their original density since most of the
stumps have been removed for distillation or have been burned
The prairies, ponds, and dwarf cypress communities
probably look very much today as they did in the past,
except where they have been altered by fire. Unfortunately,
most of them have been seriously damaged by fire.
In summary, the original Golden Gate communities
which have been altered the most are the cypress stands
and pinelands. To characterize the virgin condition of
Golden Gate vegetation one has to relate it to the historically
better-known Fahkahatchee strand cypress. Reasoning from
the historic evidence, it is felt that the cypress trees of
the Golden Gate site were less spectacular than those of
the Fahkahatchee. The forest lacked the deep and wet sloughs
that supported the ash, pond apple, and royal palm-stands.
However, the surrounding and contained pinelands in the Golden
Gate system were probably of better quality than those which
flanked the west side of the Fahkahatchee. These generaliza-
tions are supported by the belief that the Golden Gate site
is on higher elevations and is thus a drier site. This would
favor pinelands and restrict cypress growth.
The first area-wide human impact was the construction
of U. S. 41 from 1916 to 1928, along the southern reaches of
the forest ecosystem. It is doubtful that this road and its
canal caused any appreciable effect on the vegetation to the
north. In a secondary way it made penetration possible and
some of the old trails are still in use today. To the east
of the Fahkahatchee the Barron River Canal had begun to drain

T 82

the land southward from the Okaloacoochee Slough by the mid
1920's. Here, also, the impact was probably too remote to
immediately effect the vegetation on the Golden Gate site.
The first major and catastrophic impact by man was
the lumbering of cypress, beginning in the Fahkahatchee in
1943 and ending in remote strands of Golden Gate in 1957.
The wreckage of the canopy was nearly total. Exposure of the
soil surface to the drying sun, together with drought, brought
about severe fire hazard. Fire did start in the slash and
organic soil. It could not be contained. The result was
localized total, and elsewhere partial, kill of plants which
had survived the lumbering. Cabbage palms were the only
survivors in the most severely burned locations. Willows
seeded into large areas under the standing snags of cypress.
However, they are now dying out. Maples also seeded into the
Sburned-out areas and even into less burned sites where the
original cover of cypress and oak had survived -- at least
in part. These maples are in the replacement canopy with
oaks and cypress. This new forest that now grows along the
routes of the tramlines is referred to in this report as the
chfxed swamp fore and occupies areas which were once
Sessentialy pure stands of bald cypress.
SA second major impact, and specifically one that
impacted Golden Gate in the 1960's, was the canal network
built to drain the site. This draw-down of the water level
'has triggered an additional invasion of maples into areas
originally dominated by cypress and a pine invasion of some
of the .cypress on higher sites and prairies communities. Of
( greater importance is the increase of fire hazard associated
With the dry-down. The hazard is both increased frequency
and severity (see fire map, Figure 16). Not only has fire
-completely killed all trees in some stands, it has removed
organic soil generally in the system and some pockets of
eep organic soil have been depleted by as much as 18".

T 83

._ ___ ___.___ ~__;___~_~ ___~_____I___~

the land southward from the Okaloacoochee Slough by the mid
1920's. Here, also, the impact was probably too remote to
immediately effect the vegetation on the Golden Gate site.
The first major and catastrophic impact by man was
the lumbering of cypress, beginning in the Fahkahatchee in
1943 and ending in remote strands of Golden Gate in 1957.
The wreckage of the canopy was nearly total. Exposure of the
soil surface to the drying sun, together with drought, brought
about severe fire hazard. Fire did start in the slash and
organic soil. It could not be contained. The result was
localized total, and elsewhere partial, kill of plants which
had survived the lumbering. Cabbage palms were the only
survivors in the most severely burned locations. Willows
seeded into large areas under the standing snags of cypress.
However, they are now dying out. Maples also seeded into the
burned-out areas and even into less burned sites where the
original cover of cypress and oak had survived -- at least
in part. These maples are in the replacement canopy with
oaks and cypress. This new forest that now grows along the
routes of the tramlines is referred to in this report as the
qi xed swamp forest and occupies areas which were once
essentially pure stands of bald cypress.
A second major impact, and specifically one that
impacted Golden Gate in the 1960's, was the canal network
built to drain the site. This draw-down of the water level
has triggered an additional invasion of maples into areas
originally dominated by cypress and a pine invasion of some
of the cypress on higher sites and prairies communities. Of
greater importance is the increase of fire hazard associated
with the dry-down. The hazard is both increased frequency
and severity (see fire map, Figure 16). Not only has fire
completely killed all trees in some stands, it has removed
organic soil generally in the system and some pockets of
deep organic soil have been depleted by as much as 18".

T 83

practical sense, Golden Gate has three distinguishable cypress
immunitiess: bald, pond and dwarf. Pines co-exist in the
,ast two, especially the dwarf. Since most of the lumber-
Kiluable type was the "bald", it is now in the minority and-
, present mostly in the mixed forest that grew in the lumbered
iites. Most of the cypress forests appear to be of the
rpond" type.
k, To quantify these forests, point-quarter transects
ere used in four selected locations. The technique calls
,or measurement of distances between trees, diameter at breast
eight, height of trees, etc. From such measurements summary
Formation such as trees per acre, relative dominance,
relative frequency, and importance value can be calculated.
,he last is most useful when one wants to derive a single
Figure summarizing several of the measured features of the
'orest. An importance value of 300 means that a species has
produced a pure stand. Where a mixture of species is involved,
.Ae total equals 300 for the community.

Transect on Golden Gate Blvd. at S.R. 858

species Trees/Acre Ave. Ht. Rel. Dom. Rel. Freq. Imp. Value

Press 982 14 90 68 242
.ne 187 11 10 32 58

|ansect on Golden Gate Blvd. east of Canal and west of Everglades

Pypress 762 48 100 100 300

Transect at 74th and Everglades

press 1262 48 100 100 300

Transect at 124th and Desoto

ress 906 51 88 67 238
jk 82 42 10 13 30
p Ash 110 31 2 20 32

T 85


These four transects illustrate the three most common
conditions seen in the normal cypress (probably mostly "pond").
The first is in the upper part of the drainage basin, on
sandy-organic soil, and not very wet. Therefore, it is subject
to pine invasion. The second and third are obviously pure
stands of larger trees. The fourth is in the lnwer reaches
of the drainage, on deeper soils, and has a better hydroperiod.
It is being invaded by broad-leaved species and is moving
toward a mixed forest type.
Further work on the cypress was undertaken to get
some idea of the ages and growth rate of the most common sizes.
By use of an increment borer, a small core of wood was extracted
from the trunk. The core was examined for rates of growth at
different periods and the age of the tree. There is agreement
with the Duever et al. (1975) statement (p. 703) "we feel
reasonably certain that we can detect annual rings and will be
able to age a tree to within about 5% of its actual age."
Several cores were taken at transect sites and elsewhere. From
these it appears that the cypress in the deeper soil sites
grow in diameter about two to three times as fast as cypress
on dry sites.
There are few very old trees that were not cut, and
these are estimated to be at least 400 years old. The corer
was not long enough to reach the center of the tree, so an
extrapolation was made. There appears to be population in
the 200-240 year range, and a group clustering around 120-150
years of age. There are numerous small trees (less than 4"
DBH) and these can be very young or surprisingly old. When
under the shade of mother trees, the cypress sapling is
suppressed and grows very slowly as compared to one seeded
into an opening. When one considers the time involved to
grow such forests, the tragedy of a killing fire becomes
meaningful and stresses the need for forest management

T 86

~_ __ _

A transect in the dwarf (hat rack) cypress gave the
following results:

Golden Gate Blvd. at West 9th Street

Species Trees/Acre Ave. Ht. Rel. Dom. Rel. Freq. Imp. Value

Cypress 533 16 96 84 275
Pine 22 18 4 12 20
Holly 6 10 0 4 5

The age of these trees is known to be great in spite
of their limited size. The larger trees appear to be in the
range of 250 years. These cypress are frequently associated
with pines. In fact, invasion by pines is contemporary. This
community has a unique and aesthetic property. In some ways
one is reminded of the Serengeti National Park, Tanzania, Africa.

Mixed Swamp Forest
This forest is the recovery type for the extensive
sites where cypress was clear-cut in the early 1950's. In a
quarter of a century this secondary forest has become the
outstanding vegetation feature in the system. If it were not
for the palms, the forest would resemble those of the temperate
southeastern United States. It is probably approaching a climax
stage in succession. This means that it should be stable for
many years to come, unless destroyed by fire. The future of
Cypress in terms of a very long time (100 yrs) is not clear.
Certainly, the oaks and maples will continue to reproduce as
they are today.
It is also this part of Golden Gate where the rarer
plants should continue to grow and even become more numerous.
This is especially true of the epiphytic orchids, bromeliads,
peperomias and some of the ferns. The greatest number of
"bald" cypress occur here.

T 87

Spot transects were located in several sites of this
forest in order to get an average for the area. It should be
remembered that the forest is still young, quite variable,
and has been variously affected by fire since lumbering. The
data below represents unburned and some of the best of this
type forest.
Trees/ Rel. Rel. Imp.
Species Acre Ave. Ht. Dom. Freq. Value

Cypress 469 48 46 40 134
Oak 90 59 33 15 59
Cabbage palm 73 35 10 12 30
Pop ash 121 35 2 17 33
Maple 82 50 9 11 30
Red bay 13 18 0 3 5
Fig 9 22 0 2 3
Pond apple 4 15 0 1 2

The understory has a shrub layer that except for wax
myrtle is of tropical affinity. Wild coffee and myrsine are the
most common species. Other tropicals could invade and shift the
ultimate succession toward a sub-tropical climax community.
Pine Forests
Pine forests are well distributed throughout the
entire Golden Gate area and become more extensive as one moves ,
from south to north. They also tend to flank the interior a
cypress and mixed forests. They are an indicator of higher .E4
ground, and the relative elevation can be determined by the
species associated with the pines. In the most southern part a
the pinelands tend to be small islands in cypress, mixed swamp.
forests or dry prairies. In the north the pineland is frequently
a part of the dwarf cypress, so closely interrelated as to mak-t
detailed mapping difficult. There are large wet flatland
pinelands on lesser elevations that intermix with large "pond'
cypress and tend to have a graminoid understory and a large -.f
cabbage palm population. Again community delineation is

T 88

The highest pinelands are easily detected by the
presence of the understory saw palmetto. They tend to support
acid-soil-loving plants such as lyonia, blueberry, pennyroyal,
scrub oak, persimmon and pawpaw. The last, Asimina reticulata,
is endemic to Florida. These plants are not numerous and are
found scattered in openings between the dense colonies of
saw palmetto.
These high pinelands offer a special food supply
and drier habitat to wildlife. Since they are limited in
area, they should receive special attention in management.
The most vexing problem is fire. Where the understory fuel
accumulated over a number of fire-free years, fires have
killed even the largest pines, DBH almost 12 inches. On the
other hand, these pinelands are a fire sub-climax and must
have periodic fire to perpetuate them.

Wet Prairies
Compared to all other communities these are limited
in occurrence, vary in size and vegetational cover, and occur
scattered throughout the study area. Smaller ones are not
shown on the vegetation map (Figure 15). Some are transi-
tional in nature and are developing as new communities in
severely burned-out cypress heads. The selected type example
is located in the vicinity of Desoto and S.E. 48th Avenue.
Field evidence indicates that in fairly recent
times this specific site was dominated by a closed and dense
canopy of flag (Thalia geniculata) and very likely had
localized colonies of healthy sawgrass (Cladium jamaicensis)
and willow (Salix caroliniana) scattered about. Their
specific locations are related to varying soil elevations
And water depth. These plants still dominate overall, but
n most sites the communities are recovering from degrees
f fire stress. In some cases total replacement of sawgrass,
lag and willow has occurred. For example, there are numerous

T 89

local areas now dominated -- at least on a seasonal basis --
by two fire-related species, dog fennel (Eupatorium capillifolium)
and squaw weed (Senecio glabellus). Elderberry (Sambucus
simpsonii), primrose willow (Ludwigia peruviana), and cattail
(Typha sp.) also have colonized sites of severe soil burn.

Note: An unusual situation exists along Patterson between
-6th and 78th. This is included as an "artificial" prairie
created by the almost total loss of tree canopy and substantial
loss of organic soil due to reoccurrence of fires since the
lumbering of cypress. A tramline was extended into this area
to harvest the cypress. This argues originally for large
cypress, deep organic soil and a natural drainage. Hence,
with the loss of shade, slash accumulation, and dry-down of
deep organic soil, excessive loss of soil elevation by fire
occurred. Recovery of a tree canopy did not occur. Instead
a shrub dominated prairie-like community has developed.
The present cover is dense, and is dominated by reproducing
populations of elderberry, willow, primrose willow, dog fennel,
and saltbush (Baccharis sp.). Where fire damage was especially
severe, bracken fern (Pteridium aquilinum) forms large stands.

Ecology of Disturbed Surfaces

In terms of length and acreage the roads and berms
represent an enormous new microhabitat problem within the
general ecosystem. It is one of the most challenging aspects
of the venture into habitat recovery. Physically and
biologically their presence will be felt for many decades.
The challenge is to absorb them into an ecologically mean-
ingful relationship with the relatively undamaged environment
they traverse. Since they traverse every one of the plant
communities, they break the original natural continuum into
a very large number of rectangular units or blocks. Each
of these has considerable potential to become different from
its neighbors. As one views it today, fires have obviously
played a lasting part in developing extremes in ecological
conditions on an "across the road" basis. Other ecological
factors are also at work to make the difference. These are

T 90

__ ___

often subtle and difficult to prove, but their presence is
felt by a trained observer.
Some of the ecological factors that have changed
are the flow, depth and periodicity of water, air circulation,
radiation, relative humidity, temperature, light intensity,
litter and fuel accumulation, and taxonomic make-up. Obvious
effects involve the secondary plant succession on the exposed
artificial surfaces, the numerous miles of edge effect along
these roads, and the new diversity already created makes new
habitat within old established systems. Some of this new
habitat is grassland, which when dry serves as a torch to
carry fire from the roads and berms into the blocks. A few
examples from the above mentioned factors will be discussed
in the following paragraphs.
Temperature -- The protective nature of a closed
tree canopy in reducing temperature extremes is well known.
Within the mixed swamp and cypress forests one sees the
tropical species in the understory. Here the temperate
zone trees form a protective umbrella over the tropicals.
Without this protection the tropicals would not be a part
of the communities. The difference between survival and
death may be only a degree of difference over a short period.
The opened canopy of roads represents an ecological tempera-
ture hazard.
Another aspect of temperature is the exposed rock
surface along the right-of-way. Bare ground on cold wind-
less nights radiates its heat slowly all night and tends to
heat the air close to the ground. At the same time, nearby
grassy or litter covered areas will hold the heat in the soil
and not allow the cold air at ground level to be heated.
Tender plants frequently will be cold-killed at ground level
where the ground is covered but will escape damage on nearby
.open ground. The minimum temperatures (oF) shown below for

T 91

seven randomly selected cold nights.in past years illustrates
the point (Alexander, 1958).

Bare ground 28 45 33 41 27 33 51
Grass covered 21 40 30 33 22 31 48
Difference 7 5 3 8 5 2 3

The roads also represent an elevation of enough height
to also affect temperature near the ground. Cold air on wind-
less nights flows to the low spots and accumulates in them.
Therefore, these roads interfere with historic cold flows
since they "dam off" the flow in the same way water flow is
dammed behind an obstruction. Cold air also accumulates at
ground level as explained above, and a few feet above ground
the air may be warmer. The following recorded minimum tempera-
tures randomly selected from ten cold nights illustrate this
point (Alexander, 1958).

Three feet 36 28 32 32 37 39 43 30 34 38
Ground 33 22 28 27 30 34 40 27 32 34
Difference 3 6 4 5 7 5 3 3 2 4

In this case the roadway would be warmer. Thus, it is clear
that the presence of roads does affect the microclimate temperature
in several complicated ways. The difference could on occasion
bring disaster to sensitive plants.
Light -- As one drives the roads and looks into the
closed-canopy forest, the view is blocked by the density of
plants growing at the forest margin. Some of this is due to
the secondary succession (to be discussed later) on the spoil
created by construction. However, if the forest margin is
penetrated, it is obvious that the plants on the forest floor
are different in appearance and occurrence and sometimes in
species than they are deeper in the forest. Actually the
forest becomes open and easier to walk in as one moves away
from the road. Much of this difference is due to the higher
light intensity that penetrates from the road because of the
loss of tree canopy.

T 92

Evaporational Stress -- Plants are constantly
struggling to retain adequate water internally against the
combined forces of the environment that makes the water
evaporate. Plant wilting is the obvious symptom when water
loss exceeds water intake or retention. An ecological
instrument, the Livingston Atmometer is used to mimic the
plant in the environment and to measure the total evaporative
stress. Some of the stress factors are air temperature,
sunshine, relative humidity, wind, etc. Atmometers were
placed in several locations, and the water loss (stress
factor) measured. Examination of the data shows that the
roadsides are stress environments as compared to all the
native communities except for the dwarf cypress habitat.

Relative Loss of Water from Atmometers as a Measurement
of Evaporational Stress Locations Arranged
in order of Decreasing Stress

Dwarf cypress 4.4
Street intersection 4.0
Burned mixed forest 2.8
Burned cypress 2.5
Cypress 2.1
Mixed forest 1.8

These data suggest that the roads left in any pre-
served area should be carefully altered to minimize their
The new habitats mentioned earlier are the plant
communities along the roads, and on the road and berm spoil.
All of these are at some stage of vegetational recovery or,
ecologically stated, at some point in secondary succession.
The broom sedge border, stretching for miles along the pave-
ment, is often invaded by pines which are now three feet tall,
where they are not mowed. These seeded naturally and have
grown without any man-care in the short interval since the
roads were built. On the spoil outside the road ditches,

T 93

many plants have invaded and succeeded very well on their
own. Florida trema, strangler fig, lantana, oak, cypress,
cabbage palm, saw palmetto along with many vines and shrubs
have produced a sizeable thicket along many of the roads
where fire has not reduced them. These observations point
out clearly that with some help selected roads could be
altered and rapidly returned to an acceptable resemblance
of the original forests. In fact, the natural system of
recovery could be supplemented by planting royal and
paurotis palms, mahogany, and other available native plants.
This would lend more diversity to the habitat.
It is further envisioned that selected roads and
berms could serve as recreational sites for campers, hunters,
etc. Some of them could be incorporated into nature, hiking,
and bike trails and still be allowed to develop a secondary
plant cover that would lessen their effect on the existing
natural communities in the blocks. Careful planting could
remove the artificiality of.the miles of corners and endless
Farmlands have been abandoned long enough for some
secondary plant succession to start naturally. For example
some of the spoils along the drainage ditches are supporting
a young pine population. Willows are also becoming locally
common along the ditches. Old field succession is very well
known in the abandoned cotton fields in the cotton belt but
the direction plant succession in the farmlands of Golden Gate
that might remain abandoned is speculative. If not returned
to active crop production they could be easily converted to
plant nurseries to supply street, park and public buildings
with landscape material. Since they are cleared and level
they might also serve as recreational and vocational sites
such as playing fields, campsites, and vocational farming
programs. Left alone they would very likely become pinelands.

T 94

The vegetation of the canals is typical of the
hundreds of miles of canals throughout southern Florida
(see Table 5). Ecologically, the most significant problem
is the recent invasion and domination of this habitat by
an exotic, Hydrilla verticillata. This plant and other
Oiiatic exotics are currently the concern of water manage-
Went personnel and will continue to be a problem until
some means is found to control them. In the meantime,
Jtricularia spp. (bladderwort) appears to be the only
native plant that competes very well with hydrilla in
the water. Where the banks have sloughed off and a
ue -- --
ahallower water zone created, Typha spp. (cattail) are
rapidly invading. Some of the upper reaches of the canals
ire now dominated by cattail. The decided use of a water-
gay will dictate what kind of aquatic vegetational management
ill be necessary.

,ong Range Predictions on Vegetational Change

This section is.written from two points of view.
Ehe first considers anticipated changes in the existing
plant community if the system were successfully rewatered.
the second addresses the future of the community if the
environmental factors remain essentially static. Since
5ome of the communities exist in damaged and relatively
indamaged condition it will be necessary to discuss some
f them separately, starting with these two initial con-
litions. These predictions also disregard competition
Erom exotic plants. A statement on change related to
xotics concludes this section.

Mixed Swamp Forest
lewatered currently undamaged:
This cypress would continue to grow, mature and
reach maximum size and begin to die out over time

T 95

Table 5. Aquatic and semi-aquatic weeds (exclusive of
the aquatic grasses) of Golden Gate Canal System-

Water hyacinth
Water lettuce
Water pennywort
Primrose willow
Common duckweed
Giant duckweed
Fragrant waterlily
Common cattail
Narrow-leaved cattail
Florida elodea
Brazilian elodea
Common bladderwort
Illinois pondweed
Parrot feather
Spike rush

Eichhornia crassipes (Mart.) Solms
Pistia stratoites L.
Hydrocotyle umbellata L.
Jussiaea peruviana
Lemna minor L.
Spirodela polyrhiza (L.) Schleid
Azolla caroliniana (Carolinian)
Salvinia rotundifolia Willd.
Alternanthera philoxeroides (Mart.)
Nuphar advena (Ait.) Ait.
Nyymphaea odorata Ait.
Pontideria lanceolata Nutt.
T ha latifolia L.
yp a angustifolia L.
Hydrilla verticillata (L.F.) Casp.
Egeria densa (Planch.)
Ruppia maritima L.
Cabomba caroliniana Gray
Utricularia vulgaris L.
Ceratophyllhum demersum L.
Potamogeton illinoensis Morong
Myriophyllum brasiliense Camb.
Myriophyllum sp.
'Chara sp.
Eleocharis sp.

- After Weldon, Blackburn and Harrison, 1973

T 96

measured in centuries. Their place would be taken by
the cypress-shade tolerant maples and oaks. Also the
population of these two would increase. In similar
forests of the Turner River drainage north of the Trail
along 840A, West Indian tropical broad-leaved species
are becoming well established in the shade of oaks and
maples. This change from a temperate zone flora toward
a tropical zone flora would likely occur in time. Wetter
sites would support a lower canopy produced by colonies
of ash and pond apple and these, in turn, would allow
the buildup of an epiphytic flora that could eventually
include all the species previously known from the
Fahkahatchee. A requirement would be very limited fire
damage. Herein lies a paradoz; total freedom from fire
would allow dangerous accumulation of organic fuel
which, ignited in a severe drought, could cause a total
kill in vegetation. Some prescribed burning should be

Rewatered currently damaged:
With relief from fire, the vegetation would slowly
recover. The rate would be related to the degree of
damage existing at rewatering. It is possible that where
almost total tree kill has occurred, willow could invade,
form a closed canopy, and restrict the invasion of other
tree species for years. It is also possible that heavily
burned sites would become invaded by elderberry, primrose
willow and saltbush. Here tree invasion would be slow,
but not prevented as in the case of a closed willow cover,
In less-than-total kill areas, surviving plants will
recover. Cabbage palm fire resistance and reproductive
capacity will allow them to dominate extensive areas
and slow down re-invasion by other tree species. Maples
will become more prominent, with oak replacement being
much slower. Oak trees show very little fire survival
Sso seed sources become limiting. Surviving cypress
would continue to grow and in some open canopy areas

T 97

reseeding might occur to make limited pure stands.
Gradually these damaged areas would approach the con-
ditions described above for the currently undamaged
rewatered forest.

Not rewatered:
Decreased hydroperiod coupled with droughts and
the complete accessibility afforded by roads will spell
the doom of this forest. Examples already exist as
seen north of 126 Avenue where locally almost all trees
are dead and essentially no tree recovery is occurring.
Herbaceous plants such as dog fennel, shrubs such as
saltbush, bracken ferns, and grape and moonvines tend
to invade the treeless sites. Grasses also invade.
These plants collectively produce a fuel sufficient to
reach temperatures that kill tree seedlings but does not
prevent immediate recovery of the non-tree invaders by
root sprouting. Cabbage palms could easily produce
large stands through fire survival and new seedling
populations. The prediction has to be variable and
depends on conditions when and where fire occurs, and
at what frequency a given site is burned.

Cypress Forests

Rewatered currently undamaged:
Cypress tree growth would be greatly enhanced
both in an annual increase in trunk diameter and height.
The latter would be especially noticeable in growth of
the younger classes present in the canopy. Cypress
reseeding and seedling establishment would continue
in special situations such as following storm or fire
damage sufficient to prepare a seed bed. The critical
water regime needed for cypress seed to germinate and
for the seedling to survive has been documented by
foresters (Langdon, 1958). This information is so

T 98

germaine as an example of the complexity of water levels
and plant growth that it is quoted from Langdon.

"Seed germination usually averages 40 to 60
percent, but it may be as low as 9 percent
when conditions are poor or as high as 87
percent when conditions are ideal.

"Under swamp conditions the best seed ger-
mination generally takes place on a sphagnum
moss or a wet muck seedbed. The main require-
ment for germination is however an abundant
supply of moisture for a period 1 to 3 months
after seedfall. Water facilitates germina-
tion by allowing the hard seedcoats to swell
and soften. Seeds covered with water for as
long as 30 months may germinate when the water
recedes. Baldcypress seeds usually fail to
germinate successfully on the better-drained
soils because of the lack of surface water.
The exacting requirements for moisture in
early life seem to furnish the key to the
whole question of baldcypress distribution.

"For seedlings of baldcypress to become
established, the seed must sprout after the
water recedes in the swamps, and the seedling
must grow high enough the first year to stay
above the floods, except for short periods.

"Seedlings often reach heights of 8 to 10
inches during the first season and 16 to 20
inches by the second year ....

"Flooding is also an important factor in the
early growth of baldcypress. Growth is
checked when the seedling is completely sub-
merged by flooding, but submerged seedlings
have produced new shoots when the tips were
exposed to the air. Prolonged submergence,
however, may cause seedling mortality."

Natural succession is a continual process and replace-
ment of cypress over a very long period is normal. Broad-
leaved trees such as red bay, maples and ash, already in
the undercanopy, will also respond to rewatering and
increase the height of their canopy, especially where
the older cypress begin to thin by natural attrition.

T 99

The ubiquitous cabbage palm will also increase and become
an increasing component of the older cypress forest.
Thus, the direction of change would be toward the
establishment of a mixed swamp forest as described

watered currently damaged:
Succession in these damaged forests is difficult
to anticipate. The degree of damage at the start of
recovery would play a large part in determining the
recovery pattern. It is possible a successful cypress
seedling population could be established over extensive
areas. Localized re-establishment is currently common
throughout the area. On the other hand, it is more
likely that maple, ash, red bay, and cabbage palm
would invade the damaged cypress and the forest would
quickly approach the mixed forest type rather than
return to a cypress dominated forest. The recovery
rate and the characteristics of such recovery are
well-illustrated by observations of the Fahkahatchee
Strand. Figure 17 depicts conditions in the heart
of the Fahkahatchee Strand some 25 years ago toward
the end of logging operations. The similarity to
the current scene in portions of Golden Gate is striking.
In any event, rewatering and the associated lessening
of fire damage would greatly speed the recovery of the
tree canopy without regard to species.

Not rewatered:
As stated for the mixed forest, the cypress
forests will continue to degrade and disappear under
the dry-down and fire regime of the past decade.
The fire-plants, dog fennel, wax myrtle, saltbush
and bracken fern, will continue to flourish under
the opened tree canopy and supply adequate fuel to
make each fire a disaster to tree seedlings and saplings.

T 100

_~I~ I _i~_

Figure 17. Center of Fahkahatchee Strand after
logging operations in the late 1940's.


r '

T 101





It appears this type of fire-plant recovery pattern
could continue indefinitely. Cabbage palms would
become increasingly the total tree canopy cover.'
Pine invasion, already locally started, would
change existing pure cypress stands to a pine-
cypress type of forest.


Rewatered currently undamaged:
These pinelands are on the higher and drier soils
and, at best, adequate water could be restored to the
root zone. This should immediately improve their growth
rate and reproduction. They have undoubtedly suffered
drought recently to the extent that these processes
have slowed. Oak trees and cabbage palms are invading
some pinelands as a natural step in succession and this
would continue. In pinelands not excessively burned,
there is usually a population of shrubs, herbs, and
grasses that co-exist with saw palmetto, cabbage palms
and pines. An improved water table would favor retention
of these and lead to desirable diversity of habitat.

Rewatered currently damaged:
The damage situation runs the gamut from total
pine kill to minor and temporary effects. Where seed
mother trees still exist and fire is lessened, reseeding
by pine should occur quickly and naturally. But since
some of the pinelands are without live mother trees
and are isolated from each other, there would bh very
slow natural reseeding between them. In other words,
even with an improved water situation; isolated and
killed-out pinelands might not recover their pine
population for years unless forest re-seeding is

T 102


practiced. Pinelands are fire resistant, and with
reasonable water and fire regime they will perpetuate

Not rewatered:
The end result of continued drought and high
frequency fire is already seen in some of the pine-
lands. The stand may suffer a complete fire-kill of
all pines regardless of age and reduce the site to a
continuous cover of saw palmetto with a few cabbage
palms. In some cases, the large trees survive and
all younger stages die. This leads to a sterile and
non-reproducing system. Usually when the old surviving
trees die or become a wind-fall, there is no replacement.
Even the "fire-proof" saw palmetto has been burned out
in some hot fires. In these localities, grasses and
seasonal herbs tend to dominate the habitat for long
periods of time, following initial invasion by dog
fennel, and its slow decline. Fire could eliminate
much of the pineland and yet it is a part of the
management of these forests in the future.

Pineland Dwarf Cypress Forest

Rewatered undamaged:
This community is unique ecologically in that it
teams two plants with opposing water requirements;
cypress doing best in wetlands and pine best on drier
land. The nature of the sandy soil that provides
quick percolation and dry-down seems to be the answer
for the successful association of the two. It is an
old stable community as judged by the growth rings and
the size of the buttressed trunks of the so-called
"hat-rack cypress". An improved water regime should

T 10 3

favor the growth of the cypress over the pines. But
it is likely that the probable degree of water level
improvement would not significantly alter the present
appearance of this forest nor hinder the pine invasion
that is already underway.

Not rewatered:
Of the two species it appears pines have survived
the dry-down and fires of recent years better than the
cypress. This seems to be the case in spite of the
root sprouting recovery cypress displays after fire
(something pines cannot do for survival). It is also
true that young pines have to reach about 3 inches in
diameter, breast high, in order to survive an intense
fire. Pines have been successful in this community
because of the ease whereby their seedlings become
established as compared to cypress seedlings. The
prediction favors increasing quantities of pines in
this habitat.

Wet Prairies, Flag and Willow Ponds

Rewatered currently damaged:
These communities are hydrophytic and by their
very nature require standing water during a long
hydroperiod. Hence, in order for the surviving pre-
damage and pre-fire vegetation to recover or continue
to survive these habitats must be rewatered. Re-
colonization by such plants as flag, cattails, and
sawgrass would occur at the expense of the fire-
followers such as dog fennel, butterweed and evening
primrose. Since soil loss by fire has been con-
siderable, a recovered water table might allow sizeable
areas of open water. In this case, willows, pond

T 104

apples and ash would probably invade the margins and
thus return these sites to their pre-burn conditions.

Not rewatered:
As stated above, these communities are on organic
soils. In some cases all woody vegetation has been
burned away along with the upper soil. It is likely
that continued drought and fire will allow a
re-occurrence of the annual re-establishment of fire-
plants, dog fennel and butterweed. This would mean
the loss of most of the original species that were
found in these low spots.

Dry Prairies

Rewatered undamaged:
The graminoid vegetation is primarily a mixture
of sedges and grasses that grow best when each year's
hydroperiod brings about shallow inundation for
several months. Like the pinelands, the plants are
fire adapted. They respond well to periodic fires
for nutrient recycling and flowering. The community
is very stable and has survived habitat manipulation
without much change. Restoration of the water table
would help perpetuate the present plant cover.

Not rewatered:
Drought and fire favor the grasses at the expense
of the sedges. Too frequent fire tends to thin the
cover and allows invasion of seasonal herbs. The
significant predictable change is already evident in
the margins of some of the prairies. They are being
invaded by pines in some cases and by wax myrtle in
others. This is a response to dry-down and lack of

T 105

r I


fire to prune or kill the woody invaders. Change will
follow the above, depending on the cycles of fire and
length of dry periods.


Fire is the most dangerous environmental factor as
far as the future of Golden Gate vegetation is concerned.
Drought is the factor that intensifies fire damage. One
measurement of drought stress is to determine the moisture
content of the upper soil (root zone) and the covering finely
divided organic cover as a composite sample. A series of
percent soil moisture readings based on oven dry weight were
made three times during the winter dry season. These were
taken throughout the area and in several communities.
Results are tabulated below:

December 3, 1975

Habitat Soil moisture in %
Cypress 31
Cypress 21
Dry prairie 18
Cypress (average of 5 stations
range 9-11) 10
Pineland 6
Pineland 2

January 16, 1976

Habitat Soil moisture in %

Ash pond 80
Cypress willow 26
Mixed swamp forest 23
Oak hammock 20
Pineland with saw palmetto 20
Pineland with graminoids 9

T 106


March 31, 1976

Habitat Soil moisture in %

Cypress (local shower) 37
Mixed swamp forest 14
Cypress (average of 7 stations
range 5-9) 7
Pineland 3
Burned mixed swamp forest 3
Pine-dwarf cypress 3
Prairie 2
Dwarf cypress 2
Cypress pine 1

These figures take meaning when one considers that
personnel who do prescribed burning in South Florida use 30%
moisture as the critical point in deciding whether or not to
burn. In other words, if the moisture content is below 30%
it is likely the burn will get out of hand and soil fires can
also start. On only three sites was the moisture at a
marginally safe level. This simply means that the entire
Golden Gate area was vulnerable over an extended period during
the recent dry season. Fortunately fire did not get into the
deeper organic soils or forest kills would have occurred as
they have many times in the past. Any improvement in water
level in the larger cypress and mixed swamp forests will help
to prevent additional fire-kills.


Exotic plants are those plants that are introduced
from foreign sources. Some of them can invade a damaged
native plant community, replacing it with pure stands of the
invader. In recent years Melaleuca quinquenervia (cajeput),
and Schinus terebinthifolius (Brazilian pepper) have spread
into thousands of acres in southern Florida and matured into
breeding populations. No native species seem to be able to

T 107

_ ___ ~~~~__~__ ________ _

compete with them and the future of such areas is not
predictable. It appears they are both capable of self-
perpetuation in that their seedlings can tolerate the dense
shade of the parent canopy. When a mature plant dies,
the seedlings already there begin to grow. This makes a
closed community with very little or no species diversity.
Fortunately, Golden Gate has no bad invasion of
either of these exotics. They are rare south of the Alley
and become locally common to the north of the Alley. It
is judged that a population explosion of either or both of
them could follow fire, hurricane damage, or start along the
miles of roads. Management should provide constant
monitoring and removal of these plants in any area set
for preservation.

T -108


_ ____

University of Florida Home Page
© 2004 - 2010 University of Florida George A. Smathers Libraries.
All rights reserved.

Acceptable Use, Copyright, and Disclaimer Statement
Last updated October 10, 2010 - - mvs