FLORIDA APPLE SNAIL (POMACEA PALUDOSA SAY) LIFE HISTORY IN THE
CONTEXT OF A HYDROLOGICALLY FLUCTUATING ENVIRONMENT
PHILIP CHARLES DARBY
A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE
UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS
FOR THE DEGREE OF DOCTOR OF PHILOSOPHY
UNIVERSITY OF FLORIDA
Philip C. Darby
The completion of my doctoral research would not have been possible without
contributions from several cooperating agencies and many people. I greatly appreciate
the financial support of the South Florida Water Management District (SFWMD), the St.
Johns River Water Management District (SJRWMD), the Game and Fresh Water Fish
Commission (GFC)--Bureau of Nongame Wildlife, and the US Geological Survey--
Biological Research Division (USGS-BRD).
I am grateful for the support from my committee; Mike Moulton, Katie Sieving,
Clay Montague, Fred Thompson, Wiley Kitchens, and my committee chair, Franklin
Percival. Their unique perspectives, moral support and insight are very much
Steve Miller (SJRWMD), Mary Ann Lee (SJRWMD), Ed Lowe (SJRWMD), Pete
David (SFWMD), Paul Warner (SFWMD), Dale Gawlik (SFWMD), David Cook (GFC),
Don DeAngelis (USGS-BRD), and Ronnie Best (USGS-BRD) handled contractual issues
and provided valuable information and guidance during my work.
I am grateful to the GFC for permitting us to live on the Three Lakes Wildlife
Management Area. Dave Darrow and Joel Pederson (GFC) were very supportive and
I greatly appreciate the hard work of Betty Jordan, Jason Croop, Steve McGehee,
Bill Millinor, Mike Betts, Theresa Sitar, Ron Breeding, Greg Kauffman and Jeff Carter
while working on Lake Kissimmee. Traci Dean and Steve Darby provided valuable
assistance during the laboratory studies in Kenansville. Patrick Dean (GFC) and Guy
Carpenter (GFC) supplied assistance and guidance on numerous occasions in the field.
Laboratory studies were performed at the USGS-BRD facility in north
Gainesville. I appreciate the space and resources provided, and especially thank Nick
Funnicelli, Bill Stranghoener, Anne Keller and Tim Gross for their assistance and support
during those studies. Amanda Stevens, Noah Stevens, Beth Ackman, and Zach Welch
did an extraordinary job of maintaining the lab and collecting data.
I very much appreciate the administrative support provided by Debra Hughes and
Barbara Fesler (Florida Cooperative Fish and Wildlife Research Unit).
Rob Bennetts was a key figure in every aspect of my research. Rob's
contributions ranged from providing prototype transmitters to providing guidance
regarding data analyses and manuscript preparation.
Patricia Valentine-Darby was involved in all aspects of my doctoral program. As
a colleague, Patty provided valuable scientific insights, solved numerous problems, and
skillfully edited every report and manuscript I wrote. Her companionship throughout my
UF tour of duty was invaluable.
TABLE OF CONTENTS
ACKNOW LEDGM ENTS............................................................................................... iii
A B S T R A C T .................................................................................................................... v ii
1 INTRODUCTION ...................................................................................................... 1
Drying Events and W etland Habitats........................................................................ 2
Water Management in Florida and the Concern over Apple Snails .......................... 5
Background Inform ation on Apple Snail Ecology.................................................... 9
Approach to Investigations of Apple Snail Ecology.................................................. 13
2 APPLE SNAIL MOVEMENTS AS RELATED TO DRYING EVENTS AND
REPRODUCTIVE ACTIVITY......................................................................... 15
In tro d u ctio n ............................................................................................................... 15
M methods ..................................................................................................................... 17
R e su lts ................................................................. ...................................................... 3 1
D iscu ssio n ............................................................................................................... 5 1
3 APPLE SNAIL SURVIVAL AS RELATED TO DRYING EVENTS AND
REPRODUCTIVE ACTIVITY........................................................................ 57
In tro d u ctio n .............................................................................................................. 5 7
M methods ... ........................................ ................................................................... 58
Results ............................................................. ..................................... .............. 65
D iscu ssio n ................................................................................................................ 7 9
4 IMPACTS OF THE LAKE KISSIMMEE DRAW DOWN ON
APPLE SNAIL ABUNDANCE...................................................................... 85
In tro d u ctio n .............................................................................................................. 8 5
M methods .................................................................................................................... 88
R e su lts .......................................................................... .. .... .. ...... .... ................... 9 2
D iscu ssio n ................................................................................................................ 9 8
5 APPLE SNAIL TOLERANCE TO DRY DOWN CONDITIONS:
EFFECT OF SNAIL SIZE AND REPRODUCTIVE CONDITION.............. 106
In tro du action ............................................................................................................ 10 6
M e th o d s .................................................................................................................. 10 7
R e su lts.................................................................................................................... 1 1 1
D iscu ssio n .............................................................................................................. 1 16
6 SY N THE SIS....................................................................................................... 122
Conceptual Model of Florida Apple Snail Population Dynamics......................... 122
Impacts of Drying Events on Apple Snail Populations........................................ 124
Apple Snail Reproduction During the Dry Season: A Paradox? .......................... 125
Apple Snail Annual Life Cycle and Sampling Design Considerations................. 133
Implications of Apple Snail Life History for Water Management Practices........ 134
LITERA TU R E C ITED ........................................................................................ 139
BIO GRA PH ICA L SKE TCH .................................................................................... 154
Determining the biological and physical factors that limit population size is a
fundamental challenge in population ecology. A number of terms have been used to
describe these factors (see review by Krebs 1985), but the terms density-dependent and
density-independent continue to dominate the literature (Ricklefs 1990). Density-
dependent factors refer to those environmental components (i.e., disease) which have an
impact on population growth to a degree proportional to population density (i.e., as
population increases/decreases the degree of impact increases/decreases). Density-
independent factors exert influences on population growth to a degree independent of
population size (e.g., extreme temperatures may increase mortality). Generally speaking,
both density-dependent and density-independent factors likely limit population size
(Begon et al. 1986). The relative contribution of density-dependent and density-
independent processes depends upon the spatial and temporal variation in the
environment as well as the life history of the species in question (Pile et al. 1996).
Relative to populations in stable environments, populations experiencing
fluctuating environmental conditions, where at times unfavourable conditions arise, are
influenced to a greater degree by density-independent processes (Haldane 1956,
MacArthur and Wilson 1967, Menge 1974, Krebs 1985, Elliot 1987). The population
effects of density-independent factors (e.g., extreme temperatures, heavy precipitation)
manifest themselves through direct impacts on individuals, causing them to divert energy
away from reproduction (decreasing birth rate) and/or exceeding their limits of tolerance
to environmental conditions (increasing death rate). Therefore, density-independent
factors may constitute a major selective force in shaping species' life histories (Wilbur et
al. 1974, Steams 1976, Suchanek 1981, Greenslade 1983, Orzack and Tuljapurkar 1989).
Life history theory postulates that organisms allocate resources in a way that maximize
their fitness (through enhanced survival and reproduction) over their lifetime (Cole 1954,
Williams 1966, Gadgil and Bossert 1970, Steams 1976, Calow 1978). A species' life
history strategy should, in theory, reflect the nature of the environment occupied,
especially in terms of the stability of available resources (Lack 1954, Cole 1954, Cody
1966, Wilbur et al. 1974, Steams 1976). In general terms, the research described herein
addresses the potential for a density-independent process, a drying event, to influence the
abundance ofPomaceapaludosa, a species of aquatic snail. The life history of this snail
is discussed in the context of the hydrologically fluctuating environments in which it is
Drying Events and Wetland Habitats
Wetlands are ecological systems exhibiting characteristics which span a continuous
gradient between uplands and open water environments (Mitsch and Gosselink 1993).
Specific wetland types evolve from a complex interaction of geologic, climatic, biologic
and hydrologic processes (Gosselink and Turner 1978). Hydrologic patterns are the
primary distinguishing characteristic of different wetland types, and altering the
hydrologic regime will result in changes in wetland structure and function (Mitsch and
Gosselink 1993, Davis et al. 1994). The pattern of inundation and drying events dictates
the composition of the substrate, the plant community, patterns of water availability, and
subsequently the suitability of the habitat for wildlife and fish (Partington 1968, Weller
1978, Loftus and Eklund 1994). One of the most important processes for maintaining
wetland and lake littoral zone communities are drying events (Cooke 1980, van der Walk
1981, Karr and Freemark 1985, Wood and Tanner 1990, Davis and Ogden 1994,
DeAngelis and White 1994).
A dry down is defined as the condition in which the water level falls below ground
level. A dry down is the end result of a drying event. A draw down refers to the water
management practice of removing surface water, thus dropping water levels. A draw
down does not necessarily result in a dry down. Dry downs are considered disturbance
events as described by Pickett and White (1985) and DeAngelis and White (1994),
because they are discrete events which affect the availability of a critical resource (water)
and thereby affect populations, communities, and/or ecosystem function. Disturbance
events such as these are natural formative elements of community structure in many
ecosystems, and are required for the long term maintenance of some systems. Drying
events in wetlands are analogous to fire in some upland ecosystems (although fire can
also be important in some wetland systems as well). Drying events and fires both oxidize
organic material built up in the substrate (although at different rates) (Gosselink and
Turner 1978, Whelan 1995), stimulate seed germination (van der Walk 1981,
Goldammer 1990, Wood and Tanner 1990, Whelan 1995), re-set succession (Odum
1971, Gossselink and Turner 1978, Goldammer 1990) and, depending on the size and
topography of the area impacted, increase the heterogeneity of the habitat components
within a system (Gosselink and Turner 1978, Davis et al. 1994, DeAngelis and White
1994, Whelan 1995). The impact that both fires and drying events have on ecosystems
can be characterized in terms of intensity, spatial extent, frequency and timing
(DeAngelis and White 1994, Whelan 1995). Changing any of these characteristics can
change the structure and function of natural systems. In wetlands, for example,
suppression of otherwise naturally occurring drying events reduces shrubs and emergent
vegetation (Craighead 1971, Alexander and Crook 1984, Gunderson 1994, Richardson et
al. 1995) and increases the proportion of open water areas (Millar 1973, Gunderson 1994,
van der Walk et al. 1994).
Dry downs are essential in the long term preservation of the habitat structure of
many wetlands. However, drying events may also serve as an important density-
independent process in limiting animal populations (Smith 1983, Semlitsch 1987, Elliott
1987, Chow 1989, Lindeman and Rabe 1990, Woolhouse and Chandiwana 1990, Berven
1995). Understanding the relationship between drying events and the impacted animal
populations has become critical to restoring some of the best known wetland
communities in the world, those of central and southern Florida.
Water Management in Florida and the Concern Over Apple Snails
Issues related to altered wetland hydrology are exemplified by two of the nation's
largest wetland restoration projects, both in Florida; the Everglades and the Upper St.
Johns River Basin (Brooks and Lowe 1984, Holling et al. 1994). A major component of
these restoration efforts involves re-establishing, at least to some degree, natural
Distinct seasonal fluctuations in rainfall occur in Florida, providing a natural
hydrologic regime that supports wetland plant communities adapted to cyclical dry down
and inundation (Duever et al. 1994, Gunderson 1994). The dry season generally extends
from December through May or June (Chen and Gerber 1990), and in some years ends
with a dry down. The fish and wildlife inhabiting Florida wetlands have adapted
accordingly and successfully forage and reproduce within the limits imposed by these
hydrologic fluctuations. Species such as the wood stork (Mycteria americana) and the
snail kite (Rostramus sociabilis) are nomadic within the state and move in response to
changing water levels (Kushlan 1986, Snyder et al. 1989, Hoffmnan et al. 1994, Bennetts
and Kitchens 1997). Others, such as the alligator and crayfish, excavate small pools or
burrows to survive the dry season (Kushlan and Kushlan 1979, Kushlan 1990). Wading
birds benefit from the shrinking water supply as the fish they depend on become
increasingly concentrated (Kushlan 1976, Ogden 1994). The Everglades fish community
is dominated by small species with rapid life cycles which are adapted to drying events
(Dineen 1972, Loftus and Eklund 1994). The seasonal flux in water supply and wading
bird predation actually increases fish species diversity in the Everglades system (Karr and
Freemark 1985). Clearly, hydrologic fluctuations are essential to preserving the
community structure of the Everglades, as well as other Florida wetlands.
The natural hydrologic regime of wetlands in Florida has been altered substantially
due to the installation of water control structures since the 1930s, particularly in South
Florida (Light and Dineen 1994). The system of canals, levees and pumping stations
diverts water from wetland areas to agricultural and urban areas during dry times. The
same system controls flooding of developed areas by shunting water into wetland areas.
Anthropogenic alterations to the natural hydrologic cycle to accommodate agricultural
and urban needs now compromise the success of ecological strategies characteristic of
some wetland species. For example, the untimely and unnaturally large influx of water
from agricultural areas into the Everglades system has increased incidences of alligator
nest flooding (Kushlan and Jacobsen 1990) and diluted prey concentrations which affect
wading bird foraging efficiency (Kushlan 1974, 1976, Ogden 1994, Fleming et al. 1994).
Conversely, diverting water to supply agricultural and urban needs has shortened
hydroperiods in the southern Everglades (i.e., Everglades National Park), resulting in
more frequent and longer dry downs which alter macrophyte community structure (Davis
et al. 1994) and depress fish populations (Loftus and Eklund 1994). Altered hydrology
directly affects the fire regime, which is particularly important in maintaining the spatial
heterogeneity of habitat patches within the Everglades ecosystem (DeAngelis and White
1994, Gunderson and Snyder 1994, David 1996).
Research on the hydrologic regime best suited to support wading birds, alligators,
and the endangered snail kite has provided some clues to managers for the appropriate
seasonal water levels to target in support of the wetland wildlife (Kushlan 1974, Ogden
1994, Mazzotti and Brandt 1994, Bennetts and Kitchens 1997). However, little is known
about an important prey species common, to varying degrees, to each of the above
mentioned wetland inhabitants. The Florida apple snail (Pomaceapaludosa, Say) is a
critical food web component in Florida wetlands, one for which further research has been
identified as a high priority in wetland restoration efforts in all South Florida wetlands
including the Everglades (Science SubGroup, South Florida Restoration Task Force,
1996) and the Upper St. Johns River Basin (USFWS 1986, Turner 1994, Miller et al.
1996). The Florida apple snail (also referred to herein as FAS) is the nearly exclusive
food of the endangered snail kite (Snyder and Snyder 1969) and comprises more than
75% of the diet of limpkins (Aramus guarauna) in central and southern Florida (Cottam
1936, Snyder and Snyder 1969). Other avian predators include white ibis (Eudocimus
albus) (Kushlan 1974) and boat-tailed grackles (Cassidix mexicanus) (Snyder and Snyder
1969). In addition, alligators (Alligator missippienisis) (Fogarty and Albury 1968,
Delaney and Abercombie 1986), redear sunfish (Lepomis microlophus) (Chable 1947),
and soft-shelled turtles (Trionyxferox) (Dalrymple 1977) prey on FAS. Despite their
long recognized importance in Florida wetlands, especially with regard to snail kites
(Howell 1932, Snyder and Snyder 1969, Perry 1974, Beissinger 1986, Bennetts and
Kitchens 1997), surprisingly little is known regarding the life history, ecology, and
population dynamics of FAS.
The primary impetus for FAS ecological research is an interest in managing
wetland and lake water levels to support Florida's population of endangered snail kites
(USFWS 1986). The integrity of the current network of wetland and lake habitats suitable
to kites depends on a suitable hydrologic regime. Increased frequency and duration of
dry downs beyond natural levels are generally accepted as negative influences on snail
kite populations (Sykes 1983, Takekawa and Beissinger 1989, Bennetts and Kitchens
1997). The negative impact is assumed to manifest itself through depressed FAS
populations, although scant data exist (Kushlan 1975, Turner 1994). For 30 years
researchers have suggested that continues inundation benefits snails, and therefore snail
kites (Steiglitz and Thompson 1967, Beissinger 1988, 1995). However, Bennetts and
Kitchens (1997) take issue with this logic and argue that dry downs are a natural process
in the evolution and maintenance of Florida's mosaic of wetland plant communities.
They point out that dry downs may be necessary for the long term maintenance of
suitable kite habitat.
The overall goal of the research described herein was to elucidate the relationship
between hydrologic fluctuations and FAS demographics. The specific objectives of this
research were as follows:
1) to determine the behavioral responses of apple snails (e.g., migrate or
aestivate) to drying conditions;
2) to determine the environmental conditions that trigger snail dry down survival
strategy (e.g., water level, water temperature);
3) to estimate survival of apple snails during a drying event and evaluate the
influence of hydrologic parameters (e.g., rate, extent, timing and duration
of dry down) on survival; and
4) to describe the life cycle of apple snails in the context of seasonal fluctuations
in water level.
Background Information on Apple Snail Ecology
Pomaceapaludosa Say (1824) is a freshwater operculate snail, the largest in North
America (Pennak 1989), and belongs to the molluscan family Pilidae (also referred to as
Ampullaridae; see Thompson 1984). The family Pilidae consists of several genera that
occupy tropical and subtropical freshwater aquatic systems (Prashad 1925, Pain 1960,
Aldridge 1983, Thompson 1984). Efforts to clarify pilid phylogeny continue (Berthold
1991), but taxonomic classification schemes and nomenclature are not used consistently
(Michelson 1961, Thompson 1984, Bieler 1993, Cowie 1997). This document includes
discussion of two genera of the Pilidae family; Pila, an Old World (African, Asian) genus
that has a calcified operculum, and Pomacea, a New World (North, South and Central
American) genus which has a corneous operculum.
Some species in the family Pilidae are referred to as "apple snails" because of their
overall large, globose form (Keawjam 1986). The primary distinguishing characteristic
of pilid snails is the presence of both a ctendium (analogous to a gill) and a lung (a
pulmonary sac created by the mantle) for respiration (Prashad 1925, Andrews 1965,
Aldridge 1983). The presence of a lung facilitates survival for pilid snails challenged by
seasonal fluctuations in water levels, including dry downs, which are common in their
range (Prashad 1925, Pain 1950, Visser 1965, Burky et al. 1972, Kushlan 1975, Haniffa
1978a, Keawjam 1986). The dual respiratory system also proves adaptive to diurnal or
seasonal fluctuations in dissolved oxygen (McClary 1964). The presence of a lung
permits FAS and some other pilids to emerge from the water to lay calcified eggs; a
strategy believed to avoid hypoxic conditions for developing eggs, and possibly to avoid
aquatic predators (Aldridge 1983). FAS and various other pilid species deposit their eggs
on emergent vegetation (Winner 1989), whereas some other species deposit their eggs in
exposed sediments (Saxena 1958, Burky et al. 1972, Haniffa 1978b, Albrecht et al.
Data suggest that pilid reproductive cycles, to some degree, are timed such that
young snails reach sufficient size to survive the dry season (Burky 1973, Burky and
Burky 1977, Haniffa 1978b). Oviposition in FAS occurs during March through October,
with peak egg production occurring in late spring and early summer (Odum 1957,
Hanning 1979), typically prior to the onset of the rainy season (Kushlan 1975, Hanning
1979). The average snail size plummets in August due to an influx of small, young of the
year snails and an apparent adult die off (Hanning 1979). Research from the 1970s
suggested that by January the majority of the young of the year have attained sufficient
size and adequate condition to survive the winter dry season (Kushlan 1975, Hanning
Populations of FAS persist within the annual and inter-annual hydrologic
fluctuations which shape the wetland habitats they occupy. With a life span of at least
one year (Hanning 1979, Ferrer et. al. 1990) some apple snails must contend with
seasonal dry downs known to occur within their range. However, very little direct
evidence exists on the impact of dry downs on FAS populations. Kushlan (1975)
documented declines in abundance associated with low water levels in the Everglades.
Laboratory evidence suggests that FAS is intolerant to dry conditions. Turner (1994)
observed mortality rates ranging between 67% to 90% during 7 to 30 days of aerial
exposure and considered FAS tolerance to desiccation as "unimpressive" relative to other
pilid snails (Turner and McCaffree 1994). Little (1968) also examined FAS for
desiccation tolerance. As stated by Little (1968, p. 576), "the mortality rate of P.
depress [= P. paludosa] was considerably higher than that ofPomacea lineata,
probably because the operculum very seldom exactly fitted the mouth of the shell."
Unfortunately, the details of FAS survival (i.e., proportion surviving over time) are not
available from Little's publication. The literature to date indicates that, in the face of a
dry down as short as one to four weeks, FAS populations experience considerable
Anecdotal evidence suggests that FAS burrow into substrata exposed during dry
downs. Snyder and Snyder (1969) reported limpkins excavating FAS from mud flats in
central and southern Florida. Kushlan (1975), in his examination of dry downs and FAS
populations, concluded that snails burrow when the marshes dry. Harmnning (1979) drew
similar conclusions regarding over-wintering apple snails in Lake Okeechobee. Not only
did he believe snails burrowed in response to receding water, but also in response to cold
temperatures (Hanning 1979).
Research on Pilidae snails from South America and India is considerably more
extensive and may provide clues about potential FAS adaptations. Dry down survival
strategies observed involve either burrowing into the substratum or remaining in residual
pools and acclimating to water conditions. P. globosa burrows into the substrate to
depths where elevated humidity and cooler temperatures sustain the snails for several
months (Haniffa 1978a). Meenshaki (1956,1957) demonstrated P. globosa tolerance to
the anoxic conditions associated with burial. Pomacea urceus and Pila ovata (in Africa)
are apparently intolerant to anoxia (Visser 1965, Coles 1968, Burky et al. 1972, Thomas
and Agard 1992). Instead, P. urceus partially burrows with a small portion of the shell
exposed at the surface. Burky et al. (1972) believed that this permits air to enter the
chamber and possibly for the snail to extend its inhalant siphon to breathe. Burky et al.
(1972) documented an evaporative cooling effect to protect the snail from high
temperatures in its depressional burrow at the surface. Because aerial respiration (and
the associated evaporative cooling) requires occasional opening of the operculum, the
snail must contend with moisture loss. P. urceus experiences and tolerates greater
moisture loss than P. globosa (Burky et al. 1972, Meenakshi 1964). Elevated uric acid
excretion in Pila and Pomacea snails has been interpreted as a moisture conserving
strategy during aestivation (Saxena 1955, Visser 1965, Little 1968, Burky et al. 1972,
Reddy et al. 1974). FAS also excretes uric acid along with urea and ammonia, but
elevated uric acid production during dry conditions has not been investigated.
Developing FAS embryos excrete uric acid (Sloan 1964), and detection of the process in
adults may be a remnant adaptation from embryonic development (Sloan 1958).
Approach to Investigations of Apple Snail Ecology
The available information on intensively studied Pilidae snails offers some
opportunity to anticipate possible dry down adaptations for Florida's apple snail.
However, the dry season survival strategies of only two (Pomaea urceus, Pila globosa) of
the more than one hundred species of Pilidae snails have been characterized by studying
snails in their natural habitat (Meenakshi 1964, Burky et al. 1972, Burky 1974, Haniffa
1978a, 1978b). Laboratory studies describe physiological tolerances to desiccation
(Meenakshi 1964, Coles 1968, Little 1968, Turner 1994) but exclude habitat components
such as moisture retaining organic matter and residual water refugia (e.g., gator holes,
topographic depressions) which could enhance a population's dry season survival.
Caution should be exercised in extrapolating information about dry down adaptations
from other snails, even among those in the same genus. For example, anoxic intolerance
was exhibited by Pila ovata yet Pila globosa were tolerant of anoxia. Previously
published works also lack sufficient detail (Little 1968) or controls (Turner 1994) to
permit a firm basis from which to build hypotheses.
Given the paucity of information on FAS ecology, no assumptions were made about
their population dynamics or dry season survival strategy. The investigation into snail
ecology started with exploratory field studies (Chapter 2) to gain a general understanding
of how apple snails respond to drying events and potentially to reveal any unanticipated
behaviors. Laboratory studies were then begun to elucidate the mechanisms behind some
of the field observations of dry season survival (Chapter 3). Assessment of pre- and post-
draw down population abundances (Chapter 4) revealed an important relationship
between dry down timing and population demographics. The results of these earlier
efforts raised additional questions which led to a second round of laboratory studies to
test hypotheses about dry down timing and apple snail survival (Chapter 5). A model of
apple snail life history and population demographics is discussed in the context of a
hydrologically fluctuating environment (Chapter 6).
APPLE SNAIL MOVEMENTS AS RELATED TO DRYING EVENTS AND
In order to survive in aquatic habitats with fluctuating temperatures and water
levels, freshwater snails must either acclimate, migrate, or aestivate (Aldridge 1983).
Freshwater snails have been observed to exhibit each of these strategies for survival
(acclimation, Burky et al. 1972, Skoog 1976; migration, Medcof 1940, Bickel 1966,
Clampitt 1974, Horst and Costa 1975; aestivation; Visser 1965, Little 1968, Haniffa
Declining water levels challenge snails of tropical regions with higher temperatures
and declining dissolved oxygen (D.O.) in residual pools, followed by desiccation and
overheating once water levels drop to ground level (Burky et al. 1972, Haniffa 1978a,
Aldridge 1983). Some species of Pilidae snails can aestivate from several months to
more than a year in dry conditions (Little 1968, Burky et al. 1972, Haniffa 1978a, 1978b).
Migration of pilid snails in response to a drying event has not been reported, although
Haniffa (1978b) reported that a marked snail moved 150 meters from its original location
over a three month period.
Understanding how dry downs impact Florida apple snail (or FAS) populations
required either assessing changes in snail abundance or monitoring individuals in a
population subjected to a dry down. The available methods for assessing snail
abundance were limited to very labor intensive sampling methods (Darby et al. 1997),
and most methods require standing water to be effective. Determining the relationship
between population abundance and water conditions would not be feasible over the short
time periods in which conditions change. Population monitoring does not provide
information on behavioral responses. Monitoring individual snail responses (behavioral
and in sterns of survival) was therefore selected for the initial investigation.
The results presented in this chapter reveal the movement patterns of Pomacea
paludosa as a function of declining water levels in two contexts. In the first context,
snail movements were monitored during a drying event in a graminoid marsh late in the
dry season in 1995. In the second context, snail movements were monitored during a
lake restoration draw down (which resulted in a dry down) which occurred in the winter
of 1996. In each of these two studies, the main hypothesis tested was that snails move to
deeper water to avoid being stranded. Following these two studies, a hypothesis was
developed about the relationship between snail movements and their reproductive
activity. A study was conducted in 1996 to test the following hypothesis: the proportion
of males relative to females captured is positively correlated with egg cluster production
(an index of reproductive activity).
Blue Cypress Water Management Area (BCWMA). The BCWMA is part of the
Upper St. Johns River Basin, located in Indian River County, FL. It is a large graminoid
marsh system which, like the Everglades, is undergoing a substantial restoration effort
(Brooks and Lowe 1984, Miller et al. 1996). The majority of the field investigations of
snail movements were conducted in BCWMA East (Figure 2-1).
The BCWMA East plant community consists of patches of sawgrass surrounded by
mixed emergent macrophytes (Panicum spp., Eleocharis spp., Sagittaria spp.,
Pontederia cordata, and in shallower areas, Xyris sp.). Periphyton forms thick layers
over the sand substrate which characterizes BCWMA East. In 1995, water depths
throughout the study site (excluding canals) were between 0 and 70 cm. A dry down
occurred in May and lasted into June of 1995. In 1996, water depths throughout the
study site ranged from 35 to 90 cm in February-April and from 20 to 70 cm in May-
August (no dry down occurred).
Lake Kissimmee. Lake Kissimmee is an approximately 14,000 ha lake in Osceola
County (Figure 2-1). The extensive littoral zone supports numerous species of
macrophytes including maidencane (Panicum hemitomon), pickerelweed (Pontederia
cordata) and cattail (Typha sp.). Long term stabilization of water levels has resulted in
an accumulation of unconsolidated organic material (UOM) in many localized areas of
the lake. UOM sites supported dense growths of spatterdock (Nuphar luteum). Other
areas of the lake have a predominantly sand substrate with a thin (< 15 cm), patchily
/ niA \Rt. 512
jOrlando R 5
L. KISS. a BCWMAEast
-' < Miami
Figure 2-1. Map of southern Florida showing the location of field study sites as described in
Chapter 2 and referred to throughout the dissertation (L. Kiss. = Lake Kissimmee, BCWMA =
Blue Cypress Water Management Area). Specific sampling sites noted by 0. Cities (*) and
Lake Okeechobee (L. Okee.) included for reference.
distributed, flocculent organic layer. Unimpacted sites have a clean sand substrate. The
snail movement study site initially (November 1995) had water depths of 50 to 200 cm
(depending on the distance from shore). As a result of an extreme draw down conducted
by the Game and Fresh Water Fish Commission, most of the site had no standing water
by February 1996. Water receded steadily from December 1995 through February 1996
at a rate of approximately 13 cm 6 cm per week (based on measurements made at
water depth check stations, see Monitoring Methods for Lake Kissimmee).
In order to draw inferences about FAS responses to declining water levels,
individual snails had to be monitored frequently enough to keep pace with changing
hydrology (e.g., weekly or biweekly). Miniature radios were selected as the tool for
monitoring apple snails in the field. Radio-telemetry was used to examine patterns of
movements correlated with water depth, to locate stranded snails in dry marsh and to find
snails in deep water (tested up to 2 meters deep). Telemetry allowed for repeated
location of snails without labor-intensive sampling.
Before releasing snails into the field, preliminary observations were made on nine
snails, each bearing a transmitter, to assess the impact of the added weight on their
behavior. The snails weighed from 14-28 grams each. Snails bearing the 1.6 g
transmitters (ATS, Inc., Isanti, MN) were observed in an aquarium in each of the
following behaviors: crawling on the bottom, climbing on aquarium sides or vegetation,
feeding, breathing air, burrowing in loose sand and peat (near the substrata surfaces),
mating, laying eggs, and floating freely on the water surface. Snail behavior did not
appear to be compromised by the transmitter. For an adult snail (approximately 35 mm
shell length), the transmitter is about 5 to 15% of the snail body weight. Accommodating
the burden of the transmitter is ameliorated by the snails ability to regulate its buoyancy
(pers. obs), which has been documented in other Pilidae snails (Burky and Burky 1977).
The transmitters were attached to the outside of the snail shell using marine epoxy.
The area for attachment was towel dried and sanded prior to epoxy application.
Transmitters were placed 1-2 cm up from the aperture. Placement at the apex allowed
the snail to remain in an upright position when withdrawn in its shell. Approximately
one-half of the transmitters were equipped with a 12-cm whip antenna positioned to trail
behind the snail as it crawled. The other transmitters had an antenna coiled and
encapsulated in the same protective resin which coats the circuitry and battery. Signals
were received up to 200 m and 100 m from submerged whip antennae transmitters and
encapsulated antennae transmitters, respectively.
The transmitters can readily be located within an approximate 2- to 3-meter-
diameter area, but quantifying snail response to habitat conditions (e.g., D.O. or water
depth) necessitated obtaining a more precise location. Snail survival checks (see Chapter
3) also required picking up the snail. A magnet probe was developed to pinpoint
transmitter locations. When a magnet (10 cm x 3 cm x 2 cm, 25-kg pull) touched or
came within 6 to 13 cm of the transmitter body or antennae, the pulsing signal was turned
off or interrupted, or the pitch of the signal was lowered. The magnet altered the signal
under water, when the transmitter was buried in drained soil, or when buried in sediment
under water. Through use of the magnet probe, transmitter retrieval time was reduced
from hours) (without a probe) to less than 15 minutes.
Tracking snails during the marsh dry down and lake management draw down
required monitoring for several months. However, the maximum battery life of the 1.6
gram transmitters used in this study was only 60 days. Therefore, apple snail movements
for the course of the dry season were documented by releasing transmitter-outfitted snails
in a staggered fashion. Functioning transmitters from dead snails were transferred to
newly captured snails when possible.
Telemetry Study Design
Blue Cypress Water Management Area. BCWMA East was selected as the main
study area since it had the highest ground elevation of the Upper St. Johns Basin
wetlands, and was therefore most likely to dry out. The study was conducted from March
through July in 1995.
Apple snails were collected opportunistically during daytime searches in clear
water or by spotlight at night (exclusively females laying eggs). Snails were also
collected when found mating with snails bearing transmitters.
Snail sex determination was based on shell morphology (Hannrming 1979) and
behavior (laying eggs or mating). Each snail was weighed (using a spring scale) and their
length measured (using a vernier caliper) prior to transmitter placement. All snails
equipped with transmitters had a shell length of at least 33 mm. Each snail was returned
precisely to the spot from which it was taken. These and subsequent snail locations were
marked with a pvc pole or flag bearing the snail's unique identification number.
At the time of collection and subsequent relocations, the following parameters and
habitat conditions were measured: 1) distance and direction from previous location; 2)
water depth at current and previous location; 3) water temperature at current and
previous location; 4) D.O. at current and previous location; 5) substrate and general
plant composition; 6) depth of aperture of shell, if buried; 7) temperature of the
sediment (if exposed).
FAS movements were examined on three temporal scales; 12 hr, 24 hr and 5-8 day
(= weekly) intervals. These data exclude initial snail locations and locations of snails
found dead. Most movements were documented at approximately weekly intervals
(n=98) until the snail or the transmitter battery died. For daily movements, an individual
snail's location was documented two times over a 24 hr interval (e.g., 5 a.m. and 5 p.m.)
(n=- 48). Snail positions were monitored over a 12 hour interval, including a night
location, to see if snail position was affected by the diurnal cycle (n= 20).
Lake Kissimmee draw down. FAS movements were monitored on Lake
Kissimmee over 18 weeks beginning in November 1995. The draw down, part of a lake
restoration project to improve fisheries habitat (GFC 1995), resulted in an approximately
1.7 meter water level drop (from a high water benchmark of 16.7 meters mean sea level).
A study site on northern Brahma island was selected because there was no restoration
activity planned there (e.g., scraping, burning) and there was minimal boat traffic.
The telemetry procedure used to monitor snail movements and survival was similar
to that previously described for BCWMA. D.O. was not measured since gradients tended
to be very small (less than 1 ppm D.O.) and no effect was observed in the BCWMA
study. Water temperature and depth were monitored at seven stations marked by pvc
poles, which included the entire range of depths for all snail locations. As the dry down
proceeded, the stations closest to shore could not be used for water temperatures, but they
were used to monitor the exposed substratum temperature.
Apple snail movements were monitored until the transmitter battery failed or until
the snail was found dead. Throughout the 18 week study period, snails were checked at 7
to 11 day intervals in most cases (8 of 11 site visits); the remaining three intervals were
14, 17, and 17 days.
Trapping Study and Egg Cluster Survey in BCWMA East
The telemetry work was supplemented with a trapping study to test the hypothesis
that snail movement patterns during the dry season are correlated with reproductive
activity (as well as or instead of hydrologic conditions). The trap study was conducted
between January and August, 1996, in BCWMA at the same site as the telemetry study of
Preliminary observations in 1995 revealed two important issues regarding the use of
crayfish traps. First, bait is not needed to lure snails into the traps; snails apparently enter
the trap funnels during horizontal movements and/or vertical ascent to breathe air or lay
eggs. Second, sex ratios of captured snails (n =11 traps) suggested that males were lured
by females that had crawled into the traps; whenever a female was trapped (n =6 traps),
at least one and up to six males were also in the trap. Only lone males were found in
the remaining 5 of 11 traps. Since the snails must move to the traps, and no bait
attractant is provided, increased movements (whether driven by mating, temperature,
hydrology, etc.) should result in increased captures. If male movements are based, at
least partially, on seeking females with which to mate, one would expect the proportion
of males in traps containing females to be related to some index of reproductive activity.
The basic trap unit was a crayfish trap which had three tapered funnel entrances
(Sam Lemmond Enterprises, Palatka, FL). The traps were modified to test whether or not
baiting traps with a live conspecific would attract more snails. An 8-cm diameter
cylindrical enclosure was constructed which fit inside the crayfish trap and permitted the
bait (snail) to reach the surface to breathe air. A cylindrical bait enclosure was installed
in all traps, regardless of bait use. Approximately one-third of the traps (16-19 out of 54)
were baited with adult females (F), a second third were baited with adult males (M), and
the remaining traps were not baited to serve as controls (0). Traps were numbered, and
the same type of bait was used for each trap throughout the study. This was done in case
a chemical left by the bait (e.g., in the slime trail), not just the snail itself, attracted other
snails. Six pvc poles were placed in the study site to mark monitoring stations for the
collection of temperature and depth data. These markers were left in place for the entire
seven months of the study.
Six trapping sessions were conducted from 30 January to 19 August 1996. A
trapping session was initiated by placing the traps on pvc poles, approximately 5 meters
apart, throughout the study site. On the first occasion, the traps were randomly assigned
locations in order to randomize the distribution of bait type (M, F, or 0). The pvc poles
which support the traps were distributed at the same locations for each trapping session
(a map was drawn using temperature/depth monitoring stations and flagged vegetation as
reference points). However, at the initiation of each of the six trapping sessions,
individual traps (and therefore bait type) were randomly distributed among those pvc
poles to avoid any potential interactions between bait type and trap location.
Traps were checked on two occasions per session. Traps were not moved between
checks. During 12 trap occasions (2 occasions for each of six sessions) traps were
checked at 3-day intervals five times and 4-day intervals four times; the remaining
intervals were 5, 8 and 9 days. Snails found in traps were released approximately 2.5
meters from the trap (half the distance between traps). During each occasion, the shell
widths of at least 30 snails were measured.
Egg cluster production was used as the index of reproductive activity over time. A
1 x 5 meter pvc quadrat was deployed 15 times along each of three transects during each
trap session. Each of the three transects were in the same general vicinity at distances of
approximately 20 to 100 meters apart. The same egg cluster transects were sampled
during each session to control for variation in due to sampling location.
General comments. Data sets of continuous measurement data (e.g., meters
moved, shell lengths) were evaluated using Analysis of Variance (ANOVA), if the
assumptions of the ANOVA were met. The assumption of normality was tested using the
Shapiro-Wilks test (Shapiro-Wilks 1965, SAS 1992). Homogeneity of variances was
tested using the Levene's test, which is less sensitive to deviations from normality
relative to the commonly used Bartlett's test (Conover et al. 1981, Sokal and Rohlf 1981,
Schultz 1983). Non-parametric tests were employed if any of the null hypotheses for the
ANOVA assumptions were rejected.
The count data for egg clusters and the number of males, females, and proportion of
males in crayfish traps were analyzed using X2 tests based on a Poisson or bionomial
distribution using a generalized linear model approach (McCullagh and Nedler 1989,
SAS 1992). All analyses conducted in the general models procedure (SAS 1992)
included an evaluation of overdispersion.
For all analyses applicable, statistics resulting from Type In sums of squares or
likelihood ratios, which are adjusted for all other effects in the model (SAS 1992), are
reported. The outcome of all statistical tests were interpreted based on a Type I error rate
Analyses of movements during drying events. Data were grouped by biweekly
intervals to increase the sample size for the analyses. If an individual snail's movements
were measured twice within a class interval, the mean value for the distances traveled
was used for that individual.
The residuals of the BCWMA movement data as a function of time were tested for
normality using the Shapiro-Wilks statistic (Shapiro and Wilks 1965, SAS 1988). The
model effects were time interval (i.e., 12 h, 48 h, 7 d), snail sex, and biweekly period
within which the data were collected. Distances traveled (in meters) required
transformation using the function log10 (meters +1) in order to meet assumptions of
normality for the analyses of variance (W=0.96, p=0.059). Levene's test failed to reject
the null hypothesis that the variances among groups were equal (X2=22.10, 22 df,
p=0.45). Two separate mixed model ANOVAs (individual snails monitored as a random
effect; all other effects fixed) were performed. In the first ANOVA, only 7 day interval
movements were included; the main effects were biweekly interval and sex. In the
second ANOVA, with time interval as a main effect, only the data from April 29 through
June 16 were used because movements over a range of time intervals (12 h, 24 h, 7 d)
was not available outside of this time period. For both data sets, a repeated measures
ANOVA (Crowder and Hand 1990, SAS Inc. 1992) was used to account for repeated
measure of some individuals across intervals. The analysis produces F-statistics that are
not a ratio of sums of squares, but are instead derived from matix products computed to
yield Wald-statistics (SAS 1992, Searle et al. 1992). Sums of squares (SS) and mean
squares (MS) are therefore not reported for those analyses.
The model structure for movements related to water depth for the BCWMA and
Lake Kissimmee data required different approaches. The same data used to test time
effects were used, but they were grouped by water depth (depth in which the snail was
initially found) to test if snails make larger movements to avoid being stranded when the
water level reached some critical depth. Water depths in which snails moved were
divided into depth categories of up to and including 10, 20, 30, 40, 50 cm, and more than
50 cm. Because some depth categories resulted in a disproportionate number of zero
values, the variances among groups (regardless of data transformations) were found to be
unequal (Levene's test; BCWMA data F3,123=15.25, P<0.0001, Kissimmee data
F5,87=l 1.91, p
of the raw data, was conducted (Sokal and Rohlf 1981, SAS 1988).
DEPTH 1 DEPTH 2
on May 31 DEPTH
on-My -31on May 31
GRADIENT = DEPTH 2 DEPTH 1
Figure 2-2. An example gradient data collection. A snail (@) was located on 24 May and its
location marked with a pvc pole. On 31 May, the snail was found in a new location. Depth 1
and depth 2 were measured within a few minutes of each other on 31 May.
Movements along environmental gradients were calculated based on measurements
taken simultaneously at two consecutive snail locations (Figure 2-2). Water depths
varied considerably within the study area (i.e., shallow wet prairies were adjacent to
deeper canals, and areas that eventually went dry were adjacent to inundated areas). A
mosaic of juxtaposed vegetation types (i.e., sawgrass, Eleocharis sloughs, Panicum wet
prairies) also created temperature and dissolved oxygen gradients within the study area.
The primary interest was whether or not snails moved along a gradient (i.e., towards
deeper water), not the actual value of the gradient. Water depth gradient values ranged
from -90 cm to + 76 cm. If mean gradients were calculated, a large change in water
depth for one snail, for example -50 cm, would negate the weight of five snails which
moved along a + 10 cm depth gradient. Therefore the data were structured to test
whether or not snails move along a positive depth gradient, regardless of the magnitude
of the gradient. Gradients were scored as positive (P) or nonpositive (NP) (which
included zero and negative gradients). A frequency table of the proportions of P and NP
was generated. Two associations were tested; 1) the association between gradients and
biweekly time interval and 2) the association between gradients and the initial depth for
each movement (see Depth 1, Figure 2-2). The biweekly intervals and depth categories
(based on initial depths) were the same as those used in the ANOVAs for the movement
data described earlier. Both associations were tested using the Mantel-Haenszel chi-
square (MHIc2) statistic (Mantel and Haenszel 1959, SAS Inc. 1988).
Gradients were also calculated for temperatures and dissolved oxygen levels, as
illustrated for depths in Figure 2-2. Due to the low range of values for temperature and
dissolved oxygen, the data were not grouped into classes of initial temperature or initial
D.O. Temperature and D.O. gradients were analyzed only as a function of time as
described for depth gradients. For temperature, the negative gradients (N) relative to
non-negative (NN) gradients (zero + positive), as well as P vs. NP, were analyzed in
order to see if snails moved to cooler water as water temperature increased.
Analyses of trap and egg cluster data. Counts of snails from crayfish traps and egg
clusters in 5-m2 quadrats were analyzed using a generalized linear models approach
based on a Poisson distribution (SAS 1988, McCullagh and Nelder 1989). The effect of
bait type (M, F or 0) was examined using two separate analyses. In the first, only the
number of males captured was modeled as a function of bait type (males= bait
trap_session). In the second, female catch was modeled (females= bait trap_session).
Egg clusters were analyzed using the same general approach (model eggs=--trapsession).
A preliminary model indicated that a transect effect did not contribute to egg cluster
variation (p=0.822). Based on the criteria for goodness-of-fit of the Poisson distribution,
overdispersion (larger than expected variance) may have been an issue for each of these
analyses. However, correcting for overdisperion by the method of McCullagh and Nedler
(1989) did not affect the conclusion of the tests. The X2 values reported are those based
on the adjusted likelihood ratios (McCullagh and Nedler 1989, SAS 1992).
The relationship between males and females captured was evaluated using logistic
regression. Logistic regressions are based on a binomial distribution, where each trial
has one of two possible outcomes. A trial was considered a trap check in which one or
more females were captured. The outcome for the binomial response was either zero or
some number of captured males per trial. The primary hypothesis tested was that males
were lured into the traps by females that had crawled into the traps. Therefore, traps with
no captured snails were not included in the data analysis (the proportion calculation
would have included division by zero). The sample size (number of trap checks) here
varied from 33 to 54 among the six trapping sessions. The model M/M+F=--trap_session
was tested based on a bionomial distribution using PROC GENMOD (SAS 1992).
Again, overdispersion may have been a slight problem based on the goodness-of-fit
criteria for the expected binomial distribution, but adjusting for it did not affect the
conclusions of the analyses. The relationship between males captured and reproductive
activity, as measured by egg clusters, was analyzed in the same way (model M/M+F =
eggs). The model M/M+F--totalsnail_catch was also tested, in order to see if an
increase in the proportion of males corresponded to increases in total snail catch. The
mean number of egg clusters (based on n=12 quadrats) from the six surveys (February -
August 1996) were used for the regression.
The residuals of the shell size data for snails (model shell size--trap session + sex +
interactions) captured in crayfish traps were normally distributed based on the Shapiro
Wilks test (W=0.98, p=0.07). The Levene's test for homogeniety of variances showed no
significant effect on variance due to trap session (F5,356=1.89, p=0.09), but a sex effect
was highly significant (Fi, 356=14.97, p<0.0001). In order to avoid the problems of
unequal variance between sexes, separate one-way ANOVAs for males and females were
performed (model size--trap_session).
Movements During Drying Events
Blue Cypress Water Management Area. The weekly movements of the 51 snails
monitored in BCWMA varied over time (Table 2-1), with a peak in mid-April (Figure 2-
3). An overall sex effect was not indicated; however an interaction effect (p=0.052)
suggested some differences occurred between sex depending on the time period (Table 2-
1). Sixty-two percent of the 102 weekly movements documented were greater than 10
meters. The greatest weekly distance traveled was 82.5 meters. Distance traveled was
proportional to the time interval between location checks (Figure 2-4) (Table 2-2).
Again, an overall sex effect was not indicated, but a sex*week interaction was
significant. Males and females routinely moved in and out of sawgrass and other thick
vegetation throughout the study. Based on the 20 different day to night observations,
MAR APR MAY JUN JUL
Figure 2-3. Distances traveled by 51 snails in BCWMA in 1995 over an 18 week period. Data
only for snails in water depths > 10 cm. Numbers inside bars are sample sizes. Error bars are
Table 2-1. Analysis of Variance table for weekly snail distances traveled in
BCWMA. Sources of variation were biweekly period (WEEK) and sex (SEX).
The F-value is based on Wald statistics (see Methods: Analyses).
Source df F Prob>F
WEEK 8 2.53 0.016
SEX 1 0.65 0.912
SEX*WEEK 7 2.89 0.052
12 h 24 h 868 h
Figure 2-4. Distances traveled by apple snails over 12 h, 24 h, and 7 d (868 h) intervals in
BCWMA in 1995. Data only for snails in water depths > 10 cm. Numbers inside bars are sample
sizes. Error bars are standard errors.
Table 2-2. Analysis of Variance table for snail distances traveled as a function of
time interval between location checks (12 h, 24 h, or 7 day). Sources of variation
were trap interval (INT), sex (SEX), and biweekly period (WEEK). The F-statistic
is based on the Wald test (see Methods: Analyses).
Source df F Prob>F
INT 2 30.5 <0.001
SEX 1 0.65 0.42
INT*SEX 2 0.17 0.85
WEEK 3 2.29 0.085
INT*WEEK 6 0.36 0.90
SEX*WEEK 3 2.89 0.042
INT*SEX*WEEK 5 1.15 0.34
no notable pattern of females using the sawgrass by night (for oviposition) and then
moving into open slough by day (e.g., to feed) was observed. Some females remained
among sawgrass for the entire life of their transmitter, while others moved day to night,
day to day, and week to week in and out of different vegetation types and densities.
The hypothesis that snails move towards deeper water as water levels decline was
not supported by the data (Table 2-3) (MH y2 = 1.26, p=0.261). Throughout the
monitoring period, water levels declined (water temperatures also increased), but only
part of the study area went dry (Figure 2-5). The analysis of movements along depth
032 rv )
20 -----I I II I 1 I I 7.00
MAR APR MAY JUN JUL
Figure 2-5. BCWMA East water temperatures and water depths in Spring 1995. Temperatures
are means from data taken between 11:30 a.m. and 1:30 p.m. (daylight-savings time). Water
leved data from Gauge S25 1E. MSL = mean sea level (in meters).
gradients as a function of depth indicates that snails do not begin to seek deep water
refuge when some critical depth is reached (Table 2-4) (MH X2 = 2.843, p=0.092).
Although no overall significant difference was found, more than 70% of the snails at
depths of 11-20 cm moved along a depth gradient. A closer look at these depth
Table 2-3. Frequency tables for movements along depth gradients as a function of
time. NP (not positive) refers to the number of movements along a negative (from
deeper to shallower water) and zero (no difference in depth in consecutive
locations) gradient. P refers to the number of movements along a positive depth
gradient (from shallower to deeper water). Data from BCWMA East 1995.
Time Interval NP P
March 11 -March 24 2 1
March 25 April 7 6 5
April 8 April 21 8 8
April 22 May 5 7 12
May 6 May 19 7 9
May 20 June 2 6 5
June 3 June 16 11 5
June 17 June 30 6 3
July I July 14 3 2
ranges and associated movements was taken to address the possibility that these snails
gained some advantage by moving towards deeper water. In the 11-20 cm depth
category, ten snails moved 13 times to deeper water. The fate of these snails suggested
that little or no advantage to their survival was provided. Within 7 days of moving to
deeper water, two often snails moved out of the deep water and became stranded in their
new, shallower location. These locations were within one meter of deep water refuge
(deeper, shaded, cooler water). Two more of the ten snails eventually became stranded
in the dry marsh. Two other snails died within a week of moving along a positive depth
gradient. Even if snails were able to identify a depth gradient along which they moved to
find refuge, the strategy did not enhance survival: 40% became stranded in dry marsh,
and 20% died before the marsh dried down.
Table 2-4. Frequency tables for movements along depth gradients as a function of
previous depth in which a snail was found. NP (not positive) refers to the number of
movements along a negative (from deeper to shallower water) and zero (no
difference in depth in consecutive locations) gradient. P refers to the number of
movements along a positive depth gradient (from shallower to deeper water). Data
from BCWMA East 1995.
Depth Class (cm) NP P
5 32* 0
10 2 1
20 5 13
30 10 11
40 16 14
50 16 12
60 4 0
70 5 0
snails had become stranded
Although snails do not appear to move along a depth gradient or seek deep water
refuge, snail movements were influenced by depth. The distances traveled by snails
which were in less than 10 cm of water were reduced relative to all other depths
(F5,68=6.86, p<0.0001)(Figure 2-6). No tendency for snails to move into areas with higher
dissolved oxygen was observed (MH y,2 = 1.094, p=0.296)(Table 2-5). The dissolved
oxygen between locations in BCWMA East usually differed by less than 2 ppm. A
similar small range in values was observed for temperature; only eleven out of 100
temperature gradients were more than +1 C. Snails showed no tendency to move
towards warmer water (Table 2-6) (MH X2 = 0.629, p=0.428) or towards cooler water
(Table 2-7)(MHX2= 2.192, p=0.139) over the 18 week period.
5 10 20 30 40 50 60 120
Figure 2-6. Distances traveled by snails as a function of depth in BCWMA East in 1995. Error
bars are standard errors. Numbers above or inside bars are sample sizes.
Lake Kissimmee Draw Down. The mean weekly distances traveled by snails in
more than 10 cm water depths ranged from 0.3 to 21 meters (Figure 2-7). Snail
movements in less than 10 cm were greatly reduced (F4, 34=6.04, p
was observed for the BCWMA snails.
Based on an average water level drop of 13 cm per week, the slope of the littoral
zone (an approximately 100 cm drop over 100 meters), and distances traveled, only snails
released in more than 20 cm water would have been able to stay ahead of the receding
water and avoid being stranded. Twenty-two of the 31 snails were released in water
depths of at least 20 cm; the mean ( SD) depth at release was 40.7 cm ( 9.6 cm). Of
the other nine snails, seven were found in water between 10 cm and 20 cm deep. These
seven snails were included with the 22 snails, just described, in the analyses of snail
movements along gradients, but not in the calculations of proportions of stranded snails.
Table 2-5. Frequency tables for movements along dissolved oxygen (D.O.)
gradients as a function of time. NP (not positive) refers to the number of
movements along a negative (from high D.O. to low D.O.) and zero (no difference
in D.O. between consecutive locations) gradient. P refers to the number of
movements along a positive depth gradient (from low D.O. to high D.O. water).
Data from BCWMA East 1995.
Time Interval NP P
March 11 March 24 2 0
March 25 April 7 8 3
April 8-April 21 8 8
April 22- May 5 15 4
May 6 May 19 10 8
May 20 June 2 3 7
June 3 June 16 13 2
June 17 June 30 7 0
July I July 14 1 0
Table 2-6. Frequency tables for movements along temperature gradients (cooler
to warmer water) as a function of time. NP (not positive) refers to the number of
movements along a negative (from high temperature to low temperature) and zero
(no difference in temperature between consecutive locations) gradient. P refers to
the number of movements along a positive temperature gradient (from low
temperature to high temperature water). Data from BCWMA East 1995.
Time Interval NP P
March 11 March 24 2 0
March 25 April 7 11 2
April 8-April 21 9 7
April 22-May 5 16 3
May 6 May 19 10 7
May 20 June 2 8 3
June 3 June 16 12 4
June 17 June 30 8 0
July 1-July 14 4 1
Table 2-7. Frequency tables for movements along temperature gradients as a
function of time (warm to cooler water). N refers to the number of movements
along a negative (from high temperature to low temperature). NN (not negative)
refers to the number of movements along a positive temperature gradient (from
low temperature to high temperature water) and zero (no difference in temperature
between consecutive locations) gradient. Data from BCWMA East 1995.
Time Interval NN N
March 11 March 24 2 0
March 25 April 7 9 4
April 8 April 21 11 5
April 22 May 5 15 4
May 6 May 19 11 6
May 20 June 2 10 1
June 3 June 16 12 4
June 17 June 30 4 4
July I -July 14 2 3
NOV DEC JAN FEB
Figure 2-7. Distances traveled by snails as a function of biweekly time interval on Lake
Kissimmee. Data are from snails in water depths > 10 cm. Error bars are standard errors.
Numbers above or inside bars are sample sizes. Dates on x-axis are the last day of a two week
5 10 20 30 40 50 60
Figure 2-8. Distances traveled by snails as a function of water depth at Lake Kissimmee during
the 1995-1996 draw down. Error bars are standard errors. Numbers above or inside bars are
Table 2-8. Frequency table for movements along depth gradients as a function of
time interval. NP (not positive) refers to the number of movements along a
negative (from deeper to shallower water) and zero (no difference in depth in
consecutive locations) gradient. P refers to the number of movements along a
positive depth gradient (from shallower to deeper water). Data from Lake
Kissimmee drawdown 1995-1996.
Time Interval N P
Nov. 1-Nov. 14 1 2
Nov. 15 Nov.28 8 1
Nov. 29 Dec. 12 3 6
Dec. 13 Dec. 26 ND ND
Dec. 27- Jan. 9* ND ND
Jan. 10-Jan. 23 6 3
Jan. 24-Feb. 6 2 2
Feb. 7 Feb. 20 2 0
*all snails released with transmitters up to this point had died or were
Snails did not exhibit a tendency to move along a depth gradient either as a function of
biweekly time intervals (MH X2 =- 0.081, p= 0.776) (Table 2-8), or as a function of the
depth of the snails' locations (MH x2 = 1.168, p=0.280) (Table 2-9). A closer
examination of the 10 snails which did move along a positive depth gradient (total of 14
"P"movements, see Tables 2-8 and 2-9) revealed little advantage for these snails; eight
later became stranded, one died just before it would have been stranded, and just one
made it to deeper water. Fifteen of the 22 snails (68%) released in greater than 20 cm of
water became stranded. One of the 22 snails (5%) moved into an area which did not
become dry. The remaining 6 of these 22 snails (27%) died just before they would have
become stranded within the flocculent material and vegetation concentrated by the
receding water. Based on this assessment, had all snails survived, 95% of the snails
released in water depths more than 20 cm would have failed to avoid being stranded.
Movements and Population Dynamics Related to the Snail Breeding Season
Lack of evidence for snails seeking out deep water refuge left unanswered the
question of what drove snails to move, on average, 10 to 20 meters per week. Telemetry
data provided the first clues as to why snails might move during the dry season.
For most of the 18 week BCWMA telemetry study, females traveled an average of
10 to 20 meters per week, with no discernible difference over time (Figure 2-9). Male
average weekly distances, however, appeared to vary as a function of time, showing a
peak during 31 March through 5 May; this was followed by a gradual decline. The
Table 2-9. Frequency table for movements along depth gradients as a function of
previous depth in which a snail was found. NP refers to the number of movements
along a negative (from deeper to shallower water) and zero (no difference in depth in
consecutive locations) gradient. P refers to the number of movements along a
positive depth gradient (from shallower to deeper water).
Depth Class (cm) NP P
10 no data no data
20 2 2
30 6 3
40 5 7
50 7 2
60 2 0
changing relationship between male and female distances traveled over time would
explain the sex*week interaction in the repeated measures ANOVA (Table 2- 1).
Different movement patterns (direction included with distances) for males and
females were observed. Pattern assessment was limited to snails traveling in water
depths greater than 10 cm, and to snails with at least 3 consecutive recorded locations.
Using these criteria, patterns were constructed for 11 snails- 6 females and 5 males.
Males appeared to move farther and more linearly relative to females, which moved
within a more defined home range (Figure 2-10).
Figure 2-10. Movement patterns of male (a.) and female (b.) snails in BCWMA East in 1995.
Only snails for which 3 or more movements were documented are included. Patterns are to scale.
a. Male patterns
Additional evidence that snail movement patterns reflect reproductive activity was
provided by the trapping data. The proportions of males captured in traps containing
females varied over the 7 month period (X2=70.45, 5 df, p=0.0001). Egg cluster
production exhibited a similar pattern (X2 =362.2, 5 df, p=0.0001) (Figure 2-11). The
marked initial increase in egg production coincided with an increase in mean ( SE)
water temperature of 17.4 0.9 C in February to 23.3 0.23 C in April. The proportion
of males captured per female was strongly correlated to egg cluster production (X2=69.4,
5 df, p
snail catch (Figure 2-12) as well ( 2=69.4, 5 df, p
were identical for the effect of snail totals and egg clusters.
Although it appeared that males were attracted to females in the traps, baiting the
traps with females did not affect the number of males captured (X2=0.59, 2 df, p=0O.58).
The number of females captured also was not affected by conspecific bait type (control,
male, or female) (X2=1.36, 2 df, p=0.51).
The mean size of snails captured in the traps changed over the 7-month crayfish
trap sampling study for both males (F5, 179=9.39, p<0.0001l) and females (F5,177=9.86,
p<0.0001)(Figure 2-13), generally declining from April through August.
i Clusters 1.5
0 0 0
FEB MAR APR MAY J/J AUG
(n=33) (n=36) (n=27) (n=33) (n=27) (n=28)
Figure 2-11. Egg clusters and male/female ratios per trap eastern BCWMA East in 1996. Error
bars are standard errors. Only traps in which one or more females were captured are included in
the M:F data. Sample sizes for the M:F data are shown under each month. Egg cluster data was
obtained from 5-m2 quadrats (n=12 per session).
FEB MAR APR MAY J/J AUG
Figure 2-12. Total number of snails collected in crayfish traps in BCWMA from February
through August 1996. Totals are adjusted for number of trap days (which varied from 7 to 12
days) to reflect a 7 day catch in 53 traps per session.
u 32 16
29 -- -
FEB MAR APR MAY J/J AUG
Figure 2-13. Sizes of male and female snails captured in crayfish traps in eastern BCWMA East
in 1996. Numbers inside bars are sample sizes.
The first major observation using telemetry was the degree of mobility exhibited by
snails. FAS move sufficient distances over the landscape to routinely encounter a variety
of habitat types distributed over a much larger scale than expected. The spatial scale for
future studies of snail populations and their relationship to their environments should
account for the fact that 10-30 m weekly movements by snails are common.
FAS are routinely on the move, at least during the spring and early summer.
Interpretations of snail habitat preferences and snail distribution data should consider the
fact that individuals in the population frequently (on a daily to weekly basis) move in and
out of many micro-habitats within a wetland system. Snails with transmitters routinely
moved in and out of dense sawgrass patches. This contradicts earlier reports that snails
have difficulty penetrating dense stands of vegetation and that habitat use is clearly
skewed to more open habitats (Owre and Rich 1987, Turner 1996).
Snails exhibited the capacity to move sufficient distances to reach available deep
water refuge. However, no evidence was found to support the idea that snails seek
refugia as water levels drop. Brooks and Lowe (1984) suggested that deep water refugia
would benefit fish and invertebrates, including apple snails, during dry downs. Based on
observations from a minimal draw down in BCWMA in 1995 (in terms of area impacted
and duration), and an extensive draw down on Lake Kissimmee in 1995-1996, the
proportion of a snail population that becomes stranded would be proportional to the
spatial extent of the dry down. This was documented for the Indian apple snail, Pila
globosa (Haniffa 1978a). Compared to riverine environments where snails such as
Campeloma desicum exhibit rheotaxis (Bovbjerg 1952), gradients in large graminoid
marshes, such as the Everglades and Upper St. Johns Marsh, occur very gradually, even
reversing, over very large distances. FAS therefore have few stimuli to respond to in
terms of movements along a hydrologic gradient relative to their movement capacity.
FAS movements appeared to be incidental to dropping water levels; those snails not
stranded were in the right place at the right time.
Based on a lack of evidence that snails move along depth, temperature or D.O.
gradients, and the identification of a sex difference in movement patterns (distances and
direction), FAS reproductive ecology appeared to drive the movement patterns of snails
more so than did hydrology. Hannrming (1979) alluded to observations that males seek out
females for mating, but provided no data to support this conclusion. Male proportions in
crayfish traps were related to egg cluster production, suggesting that movements were
related to reproductive activity. The april-may peak in male proportions and egg clusters
corresponded to the period in which males traveled the greatest distances during the
telemetry study (Figure 2-9). Collectively these data suggest that males followed females
into the traps (and throughout the marsh) in search of a mate. Smaller movements by
females in a more concentrated center of activity may reflect the time and energy
involved in egg production and oviposition. Baiting the traps with female snails did not
increase the catch of males, suggesting that males find females through chemicals in
female mucous trails, rather than via a chemical gradient in the water column.
Chemoreception as a directional guide for movements has been documented for other
gastropods (Chase and Boulanger 1978, Cook 1979, Tomiyama 1992).
The pattern for total snail catch closely follows the proportion of males captured.
This indicated that more snails were being captured when more males were moving into
the traps in search of females. However, observer limitations may have confounded the
strength of this conclusion. Determining the sex of smaller snails was more difficult than
for larger ones. Snails from hatchling size to approximately 20 mm shell length almost
always had female shell characteristics (based on descriptions by Hanning (1979).
Overwintering males may not have acquired their characteristic flair and wider aperture
until they reached adult size and/or reproductive status. Male shell morphology was
observed for 46 of 82 snails between 25 and 30 mm which had over-wintered. It
therefore was apparent that the sex of snails larger than 25 mm could be identified, but
size-dependent observer error rates were unknown. All snails captured in this study were
at least 25 mm. Males tended to be smaller than females, which agreed with Hanning
(1979). Sexually dimorphic sizes have been documented for other pilid snails (Prashad
1925, Burky 1974, Keawjam 1987)
The proportion of males captured may also have been affected by inaccurate
identification of young of the year snails which reached sufficient size to be caught in the
traps. However, given the onset of egg cluster production (see Figure 2-11), snail egg
incubation time and hatchling growth rates (Hanning 1979), 1996 young of the year
snails could not have reached sufficient size to be captured in the traps in February,
March and April.
Size may not be a sufficient criterion to evaluate reliability of determining FAS sex.
Hanning (1979) found that young of the year male gonad maturation occurred as water
temperatures rose in late winter. Sexual dimorphism in shell morphology may not occur
prior to gonad maturation. If this is so, the sampling time, regardless of shell size, may
have affected sex identification. Variation in snail catch and proportions of males
captured probably reflected a combination of increased mating activity (males in search
of females), an increased proportion of sexually mature males in the population, and
some young of the year snails which reached sufficient size for trapping. Increased water
temperatures (Figure 2-4), until exceeding some stressful level (see Chapter 3), may also
stimulate more activity in general.
Egg cluster surveys in the BCWMA in 1996 (this study), as well as earlier surveys
done by Hanning (1979) on Lake Okeechobee and Odum (1957) in Silver Springs,
revealed that peak egg cluster production consistently occurs between March and July,
and the majority of eggs were laid over a 4 to 12 week period (Figure2-14). The
relationship between environmental conditions (hydrology, temperature, photoperiod)
and temporal patterns of oviposition will be explored further in subsequent chapters.
5 N -x- BCWMA (1996)
\ .\ --L. Okeechobee (1976)
co/\ \ -o- L. Okeechobee (1977)
0 i T \ --Silver Springs (1953-1954)
|3 -- // \
2-- ----------- '~
JAN FEB MAR APR MAY JUNE JULY AUG SEPT OCT NOV DEC
Figure 2-14. Temporal variation in apple snail egg cluster production from three areas in Florida.
Lake Okeechobee data are from Hannrming (1979), Silver Springs data are from Odum (1957), and
BCWMA data are from this study. Different survey techniques were used to sample egg clusters
in these three studies, so the data from Hannrming (1979) and Odum (1957) were transformed by
multiplication with a constant in order to place the data on the same scale. ND refers to no data;
these data points were extrapolated to complete the graph and therefore may not accurately reflect
the actual peak production.
FAS movements were not driven by hydrologic conditions, but water depth did
influence snail movements. An approximate depth of 10 cm appeared to be a threshold
level at which snail movements become impeded, probably a result of falling vegetation
and settling suspended materials (organic debris and periphyton). At this point snails
settled in one spot, and, as residual water receded, became stranded and thereby
subjected to dessication. Of the three strategies for dealing with hydrologic fluctuation
(acclimation, migration, aestivation), migration appeared not to be utilized by apple
snails. In the next chapter, the results of studies on snail survival through the course of
drying events are presented.
APPLE SNAIL SURVIVAL AS RELATED TO DRYING EVENTS
Telemetry studies revealed that although Florida apple snails (or FAS) have the
capacity to move sufficient distances to avoid being stranded in ensuing dry downs, they
did not exhibit this avoidance behavior. Instead they became stranded and presumably
aestivated until their temporarily dry habitat became inundated during the rainy season.
The survival data derived from the telemetry studies described in chapter 2 are presented
Snails moved unimpeded over a wide range of depths until the water fell below 10
cm in depth, at which point movements were greatly reduced or halted. To reiterate,
movements in water deeper than 10 cm were not explained by changes in water depth,
but more likely by FAS reproductive ecology. Once water levels fell below ground level,
snail survival depended on dessication tolerance and not migration. Therefore survival of
a study population in dry down conditions could be simulated in shallow experimental
tanks. A laboratory setting permitted control over parameters of interest, and therefore
eliminated confounding factors (e.g. above average rainfall, predation) which
compromised the interpretion of field results. This approach was employed, and the
results are presented in this chapter.
A dry down experiment was designed to emulate typical spring-summer dry season
hydrology experienced by the FAS population in BCWMA in 1995. In addition, high
water temperature impacts on snails were of interest based on observations of high
temperatures in BCWMA. As water levels fell during Spring 1995, water temperatures in
BCWMA rose to as high as 38C, a temperature found to be lethal for some aquatic
snails (Meenakshi 1964, Skoog 1976). FAS have been documented to perish at 40 C
(Freiburg and Hazelwood 1977), but variability within a population was not reported.
The potential for the temperatures measured in the field to kill Pomaceapaludosa was
investigated in a laboratory study. Two replicate experiments were conducted to
determine the temperature that kills 50% of a snail population (an LT50).
Field Studies of Survival
Telemetry data from BCWMA and Lake Kissimmee were used to calculate survival
for each study sample population. Survival was checked each time a snail was located.
If no signs of activity were observed with an intact snail, mild pressure was applied to the
operculum to see if there was resistance. For dead snails this often resulted in easily
breaking the operculum's mucuos seal and exposing dead flesh. If survival was
questionable, live snails were distinguished from dead ones by carefully inserting a knife
(8 mm blade width) between the operculum and the shell in order to see the flesh. This
was done gently, so as not to damage the operculum. If the flesh appeared intact and no
odor was noticed, the snail was considered alive. If the operculum would not open, the
snail was left for the next week. The operculum, often accompanied by snail flesh, was
very easily removed from a dead snail. If an empty shell was found, evidence of
predation was noted. The two primary signs of predation were 1) location in an obvious
snail kite or limpkin shell pile, and 2) finding that a previously stranded snail had been
extracted from the substrate. Survival data for both BCWMA and Lake Kissimmee snails
included those whose movements were monitored with transmitters and any snails
without transmitters found within 7 d of a drying event. These snails were found by hand
searches in exposed substrata.
Laboratory Draw Down Study
A dry down experiment was conducted at the Three Lakes Wildlife Management
Area (Osceola County, FL). Snails for this experiment were collected from BCWMA
East approximately 500 to 1000 meters from the telemetry and crayfish trapping study
site. All snails were collected via traps from 29 April through 10 May 1996 (n =520).
Snails were immediately placed in tanks with aerated water, provided food (bladderwort,
lettuce and spinach), and held until initial loading for the experiment. Snails were placed
into the 24 experimental tanks on 12 May.
Tank construction. Each test unit consisted of a 120 cm L x 61 cm W x 46 cm H,
285 liter capacity polyethylene tank. During some preliminary work with dry downs in
aquaria, a false bottom was found to be necessary to dry the substrate. The false bottom
was inserted 3" above the bottom of the tank and was constructed of 1/2" mesh
polyethylene screen attached to 3" fiberglass beams. A two-inch layer of stone in a
gradation from 1.9 cm (bottom) to 0.6 cm (top) was layered on top of the false bottom to
hold the test substrates (sand or peat). The test substrate layer was 5" thick. Test units
were outdoors and subject to ambient temperatures and limited sun exposure in a shaded
hammock area. Plastic tarps on pvc frames were used to prevent rain from falling into
the tanks. The water supply was from a well (pH 7.5, total hardness 3.7 grams per liter).
Well water was passed through a 5 um particulate filter, through an activated carbon
filter, and into a 950-L holding tank. The holding tank water was aerated and allowed to
reach ambient temperature prior to distribution to the test tanks.
Animal loading and maintenance. Each tank held 255 L of aerated water, providing
a 15 cm water depth above the substrate. A total of 200 to 300 grams of snails (20 to 25
snails) was placed in each tank resulting in an approximate loading of 1 g / L. Snails in
the experimental population ranged in size from 25 to 43 mm (34.2 3.0 mm). Fifty-two
(10%) of the snails had shell lengths less than 30 mm.
The loading (approximate snail g / L of water) and maintenance regime was based
on prior success at the lab, and on the Hanning (1979) laboratory experiments. In
general, however, water was replaced and snail survival was checked more often than in
previous efforts. Dead snails were identified as those floating and/or those failing to
withdraw into the shell when disturbed. Water was replaced in each tank every 3 to 7
days. Each tank received the same approximate wet weight of food as needed. Uneaten
food and waste were removed every 3 days with a net followed by a siphon pump.
Draw down regime. Twenty-four tanks were placed in two rows of six and one row
of twelve. Tank locations were randomized for substrate type and draw down rate. A
depth of 0 cm refers to a water depth at the substrate surface. To dry the substrate, the
water level fell below the substrate surface. Twelve tanks contained peat as a substrate,
the other twelve contained sand. For each substrate (12 tanks per substrate) there were
3 Control Tanks: Water depths were maintained at 15 cm above substrate level
for the duration of study.
3 Slow Draw Down Tanks: Water was dropped from 15 cm to 0 cm over 28 days (this is
the approximate rate of water drop for the 28 day period prior to dry down conditions in
BCWMA in 1995). Dry down conditions were achieved on 10 June 1996. One day later,
all the water was drained and the substrate was allowed to dry.
3 Slow Draw Down Tanks with Vegetation: The same hydrologic regime as described
under slow draw down applied. These three tanks contained vegetation to investigate
their potential to enhance survival by retaining moisture. Sand tanks were planted with
Panicum hemitomon and peat tanks were planted with Sagittaria lancifolia. Plants were
purchased from Horticultural Systems Inc. (Parrish, FL) and were 30 to 60 cm high when
planted. Stems of each plant type were planted with 5 to 8 cm between plants, which
resulted in approximately 30 Sagittaria plants per peat tank and 100 Panicum plants per
3 Fast Drawdown Tanks: Water was dropped from 15 cm to 0 cm over 1 day. One day
later, all the water was drained and the substrate was allowed to dry. These tanks
reached dry down conditions on the same day as the slow draw down tanks.
Checks for survival of control and dry down treated snails were conducted as described
for the telemetry study. All tanks containing water were monitored daily to remove dead
snails. All snails under dry down conditions were monitored every 6-8 days to check for
mortality. The status of snails in dry conditions can not be assessed without disturbing
the snails; therefore, monitoring was less frequent than for snails in water (which
generally does not require disturbing the snails). As opposed to snails in water, dead
snails in dry down conditions cannot foul the tank environment (there was no water to
Substrate temperature, moisture, and humidity at substrate level (in dry down tanks)
were measured three times each week. Humidity was measured using a hand held digital
hygrometer (Oakton Instruments, CA). Substrate saturation was measured on a relative
scale (0 to 100%) using a soil moisture meter (Lincoln Industries, NE) inserted 6 cm
below the substrate surface. The moisture meters were calibrated for each substrate type
using saturated sand and saturated peat as references.
Snail recovery from dry down. Snails exposed to dry down conditions for 3 and 7
weeks were replenished with water, first with 4 cm (for 24 hr) followed by 15 cm water
depth. The initial 4 cm of water did not completely immerse the snail, permitting aerial
respiration without the snail having to immediately crawl to the surface following weeks
of aestivation. No more than 20% of the dry down study population was randomly
selected for the recovery study. Snails were also taken from control tanks to account for
snail handling in the recovery procedure. In order to keep track of individuals in recovery
(and which treatment they had experienced), 2 mm x 3 mm plastic numbered tags were
glued to the snail shells. Tagged snails were placed in 1 of 2 recovery tanks which had a
sand substrate, and kept under conditions as described earlier. Survival of these snails
was monitored for 8 weeks. Survival of those snails placed in recovery tanks were
evaluated separately from the general study population.
Lethal Temperature Experiment
The laboratory temperature studies were conducted at the Florida Carribbean Science
Center, USGS--Biological Research Division facility, Gainesville, FL, in the Fall of 1996.
Snails were collected from BCWMA East via crayfish traps. For the initial lethal
temperature trial, 93 snails were collected from 25 October-28 October. For the second
trial, 104 snails were collected from 1 November-4 November. Snails were immediately
placed in tanks with aerated water, provided food, and held for 9 to 12 days until initial
loading for the experiments.
Tank design and temperature control. The temperature experiments were conducted
in 30 cm x 61 cm x 30 cm, 57-L glass tanks. Tanks were placed on 4 shelves, with 3
tanks per shelf. Temperature regimes were randomly assigned to the 12 tanks. All tanks
were enveloped in 1.9 cm polystyrene insulation on all sides and the bottom. A lid
constructed of the same insulating material covered the top of the tanks. Five hundred
watt heaters with adjustable temperature regulators (CLEPCO, Inc) were used to
maintain water temperatures. All tanks were aerated so that low dissolved oxygen (D.O.)
would not be a factor in survival. Test water was from a well (USGS.-BRD facility,
North Gainesville). Water was checked daily and adjusted to maintain a water level
within 2.5 cm of the top of the tank. With the lid on, this water level prevented snails
from crawling up the sides and escaping the high temperatures.
5 day LT50. Two replicates of 6 temperatures (26C, 39C, 32C, 35C, 38C,
41 C) were used to determine the temperature that killed 50% of the study population
(LT5s0) in 5 days. The upper temperature limit of 41 C was determined from a rangefinder
test in which all snails at 41 C and higher died within 3 days. Snails were not fed during
the study to prevent plant material from getting caught on the heaters. Seven to nine
snails were placed in each of the 12 tanks. All tanks started at 26C. Test initiation was
the point at which temperatures in treatment tanks were raised above 26C. During the
first 24 hours of the experiment, temperatures were raised at a rate of approximately
1.5C per hour until the final target temperature was reached. Temperature and
dissolved oxygen were monitored daily. Temperature and D.O. data were taken from a
point at mid- depth in the center of each tank. Temperatures were maintained within
1C throughout each experiment. The first trial was conducted from 7 November-11
November. The second trial was conducted from 15 November-19 November.
Survival of BCWMA snails, Lake Kissimmee snails, and laboratory snails was
estimated at weekly intervals. The Kaplan-Meier procedure to accommodate staggered
release of transmitters was used to estimate survival with 95% confidence limits (Pollock
et al. 1989). Chi-square tests as described by Pollock et al. (1989) and White and Garrott
(1990) were used to compare rates of survival between males and females in the
BCWMA study and between controls and treatments in the laboratory studies.
An LT50 was derived from each temperature experiment. In the second lethal
temperature trial, only one temperature resulted in partial mortalities, so probit analysis,
which is the primary method recommended by the American Public Health Association
(1985) was not appropriate. The 5 day LT50 with 95% confidence limits was therefore
determined by the moving average method (Finney 1964, American Public Health
Association 1985, Baker and Heidinger 1996).
Blue Cypress Water Management Area Telemetry Study
No snails died during the first six weeks of the study (Figure 3-1). Only male snails
were found dead in weeks 7 and 8; both male and female snails were found dead in each
of the remaining weeks of the study (Figure 3-2). The male and female survival curves
were not statistically different (X2 = 0.057, df= 1, p> 0.90), even when comparing male
and female survival from only April 22 through June 16 (see Figure 2)(X2 = 0.048, df= 1,
p> 0.50). These survival curves included only 8 snails stranded in the dry down area;
therefore, desiccation does not account for the large decline in survival.
Recall that transmitters were released in a staggered pattern. This means that some
snails in week 13, for example, had worn a transmitter for 1 week, others for 3 weeks,
and others for 9 weeks. Snails dying in a particular week had, therefore, carried a
transmitter for a range of different times. Survival was not a function of time bearing a
"> \\\ L \\
\ \ \
0.40 \ \
\ ^^ s
0.20 .... "-
0.00 I I I I I I I I I- I t t-I I-
MAR APR MAY JUN JUL
Figure 3-1. Survival of the snail population monitored in BCWMA East in 1995. Dotted lines
indicate 95% confidence limits.
Figure 3-2. Survival of male and female snails in BCWMA East in 1995
APR MAY JUN JUL
-- Depth Gauge S251 E
S32 C- O
3 30 7.50 >
20 I I I- I -I I 7.00
MAR APR MAY JUN JUL
Figure 3-3. BCWMA East water temperatures and water depths in Spring 1995. Temperatures
are means from data taken between 11:30 a.m. and 1:30 p.m. (daylight-savings time). Water
leved data from Gauge S25 1E. MSL = mean sea level (in feet).
Survival dropped steadily as water depths decreased and as water temperatures
increased (in general) over the 18 week study (Figure 3-3). The steepest continuous drop
in survival occurred during weeks 9 through 14 of the study period (Figure 3-1). During
this same period mean temperatures (between 11:30 a.m. to 1:30 p.m.) consistently
exceeded 30C. In late May, afternoon temperatures of 38 C were recorded. Right after
these high temperatures were recorded, the greatest number of snail deaths relative to any
other week were observed. All dead snails found in water (n=6) were from the open
slough area, which experienced the highest water temperatures. High water temperature
may have contributed to stress and the eventual mortality of these snails. Despite high
temperatures in the marsh, D.O. was quite high (>8 ppm) due to the abundance of
Although water levels dropped steadily during the study, the dry down was not
extensive. Only 8 of the 51 snails which contributed data to the survival curve became
stranded (Table 3-1).
Three of these stranded snails did not have transmitters (found via hand searches),
demonstrating that snails without transmitters became stranded. The mean survival time
in dry down conditions was 3.9 2.2 weeks.
All stranded snails were near the substrate surface with their apertures 1-3 cm from
the surface. Some snails rested on top of the substrate. Even in patches of accumulated
organic material, where for example a mud turtle (Kinosternon sp.) was found
completely buried, snails were near the surface. Snails do not burrow to avoid dry down
conditions, at least in the predominantly sand substratum characteristic of BCWMA East.
Table 3-1. Fate of stranded snails in BCWMA 1995.
Fate of Stranded Snail No. of Snails
Non-predation death 3
Predation death 1
Unknown fate 2
Snail never died' 2
1 Water level rose and snails revived and crawled away.
All stranded snails were found under the protection of some kind of vegetation.
Afternoon substrate temperatures at snail locations and at adjacent patches of substrate
with no vegetation were compared (n =16). Water temperatures in the open slough area
were also taken at the same time (n =9) and compared to stranded snail temperatures.
Shaded substrate temperatures ranged from 29 to 33 C, and were an average ( SD)
3.3C ( 1.4C) cooler than substrates devoid of vegetation (i.e., unshaded). Stranded
snails, whether under vegetation or not, experienced mean temperatures 4.3C ( 1.6C)
cooler than snails which were still in inundated areas of the marsh.
Lake Kissimmee Draw Down Telemetry Study
The first snail deaths did not occur until after snails became stranded sometime
between 4 December and 18 December (Figure 3-4), two weeks into the study. Snail
survival was not dependent on the length of time bearing a transmitter, since at any time
period snails had worn transmitters for 1 to over 8 weeks. Intolerance to dry down
conditions was also not the direct cause of death in the majority of cases (Table 3-2).
The mean ( SD) survival time for 23 stranded snails was 3.9 3.1 weeks, the
same survival time found for snails stranded in BCWMA in 1995 (n =8). Predominant
causes of mortality were different however. Of the 23 stranded snails monitored in the
lake study, 57% were taken by predators. Another difference from the BCWMA study
was that during the Kissimmee draw down, central Florida experienced several freezes;
at least 11 of 23 stranded snails survived at least one of these freezes. A freeze event was
defined as when the mean minimum temperatures of Vero Beach and Orlando, reported
by the National Weather Service (NWS), were 0 C or lower. Based on this criterion,
central Florida experienced a freeze on three occasions: 25 December (0 C), 9 January
(-1.4 C), and 5 February (-2.2 C). [Note that substrate temperatures at snail locations
were not measured; the analysis is based solely on air temperature data from the NWS.]
Freeze events did not appear to increase the mortality rate (Figure 3-5), however
only seven non-predation deaths were confirmed for the sample of 31 snails. The high
incidence of predation deaths, as well as the contributions of dessication, precluded an
evaluation of deaths caused by temperature. Two of the seven non-predation deaths were
snails that were still in 8 and 15 cm of water, and therefore did not experience freezing
temperatures (i.e., the water column did not freeze).
U First stranded ", ' ""
0.4 snails found ~. .. \
NOV NOV NOV DEC DEC JAN JAN FEB FEB
1 15 29 13 27 10 24 7 21
Figure 3-4. Survival of 31 snails during the Lake Kissimmee 1995-1996 draw down. The dotted
lines are the upper and lower 95% confidence intervals. Substrate exposure in the study area
began sometime between December 4 and December 18. Dates indicate the beginning of a two
week time interval within which data were collected.
At the termination of the study, one remaining stranded snail was collected, still
alive after 8-9 weeks in the dry down area, and placed in an aquarium with water. The
snail revived within 24 hours and lived another eight weeks. Two snails were also
revived in water after having survived at least three months in dry conditions. These
snails were inadvertently left in the traps which were stored on dry ground from 10
November 1995 through 15 February 1996. These snails survived three freezes (as
defined above) in the traps without any benefit of substrate moisture or damp organic
Table 3-2. Fates of snails (n= 31) monitored on Lake Kissimmee during the
1995-1996 draw down.
FATE OF SNAIL NO. OF SNAILS
Stranded Snails 23 total
predation death 13
non-predation death 5
unknown cause of death 4
never died* 1
Snails Never Stranded 8 total
predation death 2
non-predation death 2
unknown cause of death 1
never died* 1
*snail was not dead by the end of the monitoring study
**snails found alive, but shells crushed by foraging cattle
Laboratory Studies of Dry Down Survival
FAS survival dropped at approximately the same rate for all treatment types,
including controls (Figures 3-6 and 3-7). For both sand and peat tanks, control survival
remained 10 to 20 % higher than treatment tanks following the dry down. This
difference between controls and treatments within each substrate type was not
statistically significant, but was nearly so for the peat tanks (Table 3-3). Vegetation did
not affect survival in the peat tanks, but in the sand tanks the presence of Panicum
appeared to result in lower survival relative to other dry down tanks. This difference
appeared to be due largely to survival in the sand tanks prior to dry down conditions.
NOV NOV NOV DEC DEC JAN JAN FEB FEB
1 15 29 13 27 10 24 7 21
Figure 3-5. Survival of 31 snails during the Lake Kissimmee 1995-1996 draw down. Non-
predation deaths of stranded snails (n=5) (s) are shown in relation to the three freeze events (*) as
defined in the text. Substrate exposure in our study area began sometime between December 4
and December 18. Dates indicate the beginning of a two week time interval within which data
Draw down rate did not result in significantly different survival for peat or sand tanks,
although the relative difference within sand tanks was greater than for peat tanks (Table
3-3). There was a rapid drop
MAY JUN JUL
Figure 3-6. Survival of snails in sand tanks. Vertical dotted line indicates the day on which water
levels reached substrate level in treatment tanks (fast, slow and vegetation).
0.80 -- slow
oo ^^ -x- vegetation
MAY JUN JUL
Figure 3-7. Survival of snails in peat tanks. Vertical dotted line indicates the day on which water
levels reached substrate level in treatment tanks (fast, slow, and vegetation).
Table 3-3. Comparisons of survival between controls and dry down tanks (slow draw
down tanks with no vegetation), slow versus fast draw down rate, dry down tanks with
vegetation versus without vegetation, and sand versus peat substrates. Fast draw down
tanks were only included in the slow versus fast draw down analyses. For all comparisons
(2 survival curves) df=l.
Comparison %2 PR> X2
Control vs Slow Dry Down
sand tanks 0.98 0.20
peat tanks 3.61 0.06
Slow vs. Fast Draw Down
sand tanks 3.05 0.08
peat tanks 1.72 0.70
Vegetation vs. No Vegetation
sand tanks 5.84 0.02
peat tanks 1.64 0.30
Sand Draw Down vs. Peat Draw Down 0.06 0.99
in soil moisture in sand compared to peat tanks (Figure 3-8), but survival for snails in dry
sand versus dry peat tanks was not significantly different (Table 3-3). Humidity above
the tanks ranged from 50 to 90%, but was only a factor of atmospheric conditions, not
The survival of snails which were removed from control and treatment tanks, marked,
and placed in recovery tanks initially showed the same decline as snails monitored in the
first 12 weeks of the study (Figure 3-9). Aestivating snails exposed to water became
active within 2 to 24 h. Survival of snails in recovery tanks stabilized during the last 4
weeks of the study, during which time no snails died. The ten snails still alive those last
four weeks were smaller than the rest of the draw down study population (T=4.29,
df=-513, p<0.001). The mean size ( S.D.) of these 10 snails was 30.1 5.9 which
indicates that many of these snails were young of the year.
Figure 3-8. Relative substrate moisture in sand and peat dry down tanks following the dry down.
Controls for both sand and peat tanks were saturated throughout the study (controls not shown).
Moisture scale ranges from 0 (completely dry) to 10 (saturated). Relative humidity ranged from
50 to 90% throughout the study. Sample size ranges from 7 to 9 tanks per substrate. Standard
1.00 -*-. snails from peat
-o- snails from sand
Figure 3-9. Survival of snails removed from control and draw down tanks and placed in
The results of the two trials differed substantially, but both studies indicated that
water temperatures above 35C caused extensive snail mortality (Figure 3-10). The
dissolved oxygen in all tanks was maintained at near saturation level for each
temperature (Figure 3-11), and were well above D.O. levels sustained in viable snail
populations in the field (BCWMA West and Water Conservation Area 3A, pers. obs.).
The 5 d LT50 for the first trial was 31.2C with 95% confidence limits of 30.9 and
31.5C. The 5 d LT50 for the second trial was 37.01C with 95% confidence limits of
36.9' and 37.2C.
---trial 1 /
-trial 2 /
-o 0.8 -
26 29 32 35 38 41
Figure 3-10. Median temperatures which killed 50% (LT5o) of snails in two study populations in
two test trials. LT50 were calculated using the moving average method (Finney 1964). LT50 with
95% confidence limits for trial 1 indicated by symbol LTs5o for trial 2 indicated by symbol
-4-. For each temperature n = 15 to 17 snails.
26 29 32 35 38 41
Figure 3-11. Mean dissolved oxygen levels (D.O.) for tank temperatures from two LT50 trials.
For each temperature, n= 8 to 18 (D.O. was measured daily, but only for tanks with live snails
remaining). Error bars are standard errors.
Hydrologic conditions undoubtedly influence survival of FAS, but the results of the
1995-1996 survival studies suggested that vulnerability to desiccation during the dry
season was not a predominant cause of mortality. The survival curve for BCWMA
snails in 1995 reflects survival of 51 individuals, only 8 of which were stranded in dry
down conditions. The observed drop in survival in BCWMA East may, in part, have
been related to a general senescence of the population. Hannrming (1979) observed an
increase in floating dead adult snails during his summer surveys of Lake Okeechobee.
The estimated life span for Pomaceapaludosa is 15 to 20 months (Ferrer et al. 1990,
Hanning 1979, pers. obs. in aquaria), and the BCWMA 1995 telemetry data are
consistent with that estimate. In BCWMA in 1995, 38% of snail deaths occurred over a
one week period in June, bringing to question whether senesence alone could account for
that high a death rate over such a short period. It may be that older FAS are more
sensitive to environmental conditions, but the relative contribution of high temperatures
to snail mortality during the field study could not be determined.
On Lake Kissimmee, 57% of the stranded snails perished from predation, which
confounded drawing conclusions about senescence and/or desiccation tolerance.
Stranded snails were vulnerable to birds which foraged in the exposed lake bottom.
Although foraging opportunities for snail kites were eliminated in the dry down zone,
ibis, grackles, and limpkins, all predators of FAS (Snyder and Snyder 1969, Kushlan
1974), were observed foraging in the telemetry study site following the dry down.
The results of the laboratory dry down study concur with earlier reported apple
snail tolerance to dry doxwn conditions. Following exposure to dry down conditions, in
this case after a 10 June dry down, snail survival dropped by 50% or more after just two
weeks. Turner (1994) found the same trend in his lab study, recording 50% to as little as
2% survival over 7 to 29 days of aerial exposure. Snails provided saturated sand
perished just as rapidly (13% survival after 7 d). Little (1968) also reported a relatively
low tolerance for FAS to dry down conditions, but provided no survival data. Survival in
dry down conditions on the order of a few weeks is considerably less than the months and
even years reported for other Pilidae snails (Meenshaki 1964, Little 1968, Burky et al.
1972, Haniffa 1978a, Swami et al. 1978, Keawjam 1986).
The dry down experiment demonstrated a very important caveat to purported
tolerances of FAS to dry down conditions; tolerance may be contingent upon age, size
and/or reproductive maturity. This study included control snails (always inundated with
aerated water), which was not the case for the studies of Turner (1994) and Little (1968).
Although control snail survival remained 10 to 20% higher than draw down tanks, their
survival still plummeted over the course of 2 months in spring and summer. Dry season
survival declines appeared to follow the peak period of egg cluster production (Fig. 3-
12), indicating a post-reproductive die off. The control survival curves from the
laboratory experiment look remarkably like those from adult snails monitored in
1.00- -o- BCWMA Survival 1995
1.\ -o-Lab Survival 1996
/ V \ -5.0
BCMWA Eggs 1996 5.0
0.00 nl------- 0.0
FEB MAR APR MAY JUN JUL
Figure 3-12. Apple snail egg cluster surveys (BCWMA 1996) and apple snail survival (telemetry
survey 1995 and lab study 1996).
BCWMA in 1995. The lab and field studies confirm earlier unsubstantiated reports that
snails live 1 to 1.5 years (Hanning 1979, Ferrer et al. 1990). Given a 1 to 1.5 year life
span, and knowing that the majority of egg cluster production for any given year occurs
between March and June (Chapter 2), this means that most overwintering adult snails die
between April and August. It is significant that the Turner (1994) desiccation tolerance
experiments were conducted in May and June, and therefore included short lived post-
reproductive adults. The snails in his study likely would have died regardless of
Snails that survived in the lab through August (some in dry down conditions for 7
weeks) were significantly smaller than the snails that had perished earlier in the
experiment. In the last four weeks of the lab study, these smaller snails exhibited 100%
survival, indicating that they can fully recover from dry down conditions and potentially
complete their life cycle through reproduction. The observed rapid recovery of snails
from dry down conditions is consistent with observations of Meenshaki (1964) for Pila
globosa and Burky (1974) for Pomacea urceus.
Interest in the potential impact of high temperatures on adult mortality stemmed
from observations of 33C to 38C water temperatures in BCWMA in May 1995.
However, given that adult survival declined in the moderate temperatures of the 1996 lab
study (23C -30C), high temperatures probably were not the primary cause of snail
deaths in BCWMA in 1995. Post-reproductive snails were likely dying from senesence
during this period, but high temperatures may have exacerbated mortality in these snails.
A median 5 d lethal temperature for FAS of 31 to 38C is somewhat lower than a lethal
temperature of 40C reported for this species by Frieburg and Hazelwood (1977), but
there exposure time was less than one day. Lethal temperatures around 35 to 45C have
been observed in other Pilidae snails (Meenakshi 1964, Burky et al. 1972, Haniffa
1978b). Note however, that the LT50 experiments on FAS were conducted during the fall
season, and therefore should be considered preliminary to understanding the impacts of
temperature on snail populations in the spring-summer dry season. Snails and fish which
are acclimated to higher temperatures have a higher temperature tolerance than
conspecifics acclimated to cooler temperatures (Skoog 1976, Elliot et al. 1981, Hoffman
1983, Baker and Heidinger 1996). The LT50 could be higher for snails acclimated to
typical spring water temperatures. A fluctuating temperature regime was not applied in
the lab studies (due to the expense and difficulty of replicating temperature profiles),
which would have more accurately reflected field conditions. Given that snail tolerance
to dry down conditions appears to be based on age or reproductive status, it seems
plausible that tolerance to high temperatures would depend on the life stage of the snail
Stranded FAS appear to tolerate freezing air temperatures, as exemplified by some
Lake Kissimmee snails and those inadvertently left in traps. Frieburg and Hazelwood
(1977) reported that FAS survived 5 C, but lower temperatures were not tested. Low
temperature survival during dry downs has implications for freeze protection withdrawals
conducted by water management districts (Steve Miller, SJRWVMD, pers. comm.). These
temporary withdrawals (i.e., on the order of a few days), should not result in extensive
apple snail mortalities.
White and Garrot (1990) point out that caution should be exercised in interpreting
movement patterns (Chapter 2) and survival as influenced by environmental and
physiological conditions that vary seasonally, because many of these conditions change
simultaneously. The FAS studies provided good examples of this; as photoperiod
increased, temperature increased, water depth decreased, snail reproductive activity
increased, and the population aged into senescence. The two main telemetry surveys
(Lake Kissimmee and BCWMA East) included changes in all of these parameters, which
makes conclusions about casual relationships difficult, if not impossible, to make. For
example, interpretation of the snail movement data was likely confounded by the fact
that the majority of the study population eventually died, and some likely were moribund
for a week or more immediately prior to death.
Monitoring survival of individual apple snails provided valuable information on
how these animals respond to changing environmental conditions. These initial
investigations also revealed critical aspects of their life history. However, despite two
field studies and a laboratory experiment, the impacts of a drying event on snail
populations as a whole had not been examined. Drying events do not just impact the
movements and survival of snails exceeding 25 mm in shell length, the only size class
studied thus far. Dry downs may affect recruitment and also survival of smaller age/size
classes of snails. Knowing the Lake Kissimmee draw down would occur provided an
opportunity to prepare for a long term assessment of apple snail abundance following an
extreme drying event. The results of the population study are presented in the next
IMPACTS OF THE LAKE KISSIMMEE DRAW DOWN ON APPLE SNAIL
DISTRIBUTION AND ABUNDANCE
Along with Florida's wetland habitats, Florida lakes historically underwent large
intra-annual and inter-annual fluctuations in water levels. As a result, "normal" lake
levels are elusive; the norm is for wide fluctuations, a pattern which earned Florida lakes
the descriptive title, astaticc", or unstable (Brenner et al. 1990). Between the 1940's and
1960's, Lake Kissimmee (Osceola County, FL) had a mean annual water level
fluctuation of 1.4 m, but the mean water level over the course of several years varied as
much as 3.7 m (Grocki 1975). The substrate and vegetative communities, in part, reflect
the pattern of hydrologic fluctuations; their timing, duration, and frequency.
Topographic variability in lake systems, coupled with their hydrologic patterns, yield
zones of shrubby, emergent and submerged plant habitats in the littoral zone of the lake.
Installation and operation of water control structures over the past 30 to 50 years
have dampened the degree of hydrologic fluctuations in many Florida lakes, resulting in
long term stabilization of lake levels (Holcomb and Wegener 1971, Wegener et al. 1974,
Fox et al. 1977, Moyer and Williams 1982, Florida Game and Fresh Water Fish
Commission (GFC) 1995). The high and low water levels historically noted for Lake
Kissimmee are no longer observed (GFC 1995). In addition, increased nutrient inputs to
Florida's watersheds from agricultural and urban development have resulted in lake
eutrophication (Wegener et al. 1974, Moyer and Williams 1982). As nutrients increased
and water levels stabilized, submerged and floating-leaved macrophytes expanded and
organic debris accumulated in some regions of Lake Kissimmee. Fish populations suffer
from drops in dissolved oxygen associated with decaying algae and plant matter, the
alteration of plant community composition, and poor habitat quality of the substrate
(Wegener and Williams 1974).
To remedy these problems in Florida lakes, the Florida Game and Fresh Water Fish
Commission (GFC) initiated a lake restoration program in 1971 which emphasizes the
use of a lake draw down to expose sediments (GFC 1995). The objective of the program
is to reduce undesirable plant species, expand desirable plant communities, and
consolidate flocculent organic sediments (Holcomb and Wegener 1971, Wegener et al.
1974, Fox et al. 1977, Cooke 1980, Tarver 1980, GFC 1995). In addition, "muck"
removal operations following lake level draw downs are meant to resolve the problem of
extensive build up of organic sediments (GFC 1995, Moyer et al. 1995). [GFC Lake
restoration documents use the term "muck" to describe shoreline with an unconsolidated
organic layer, but based on criteria in Cowardin et al. (1979), these areas should be
considered unconsolidated organic bottom (UOB)]. The goal of the draw down and
restoration activity is to improve aquatic habitat in support of fisheries and wildlife,
while improving lake quality for recreational usage (e.g. sport fishing) (Wegener and
Williams 1974, Tarver 1980, GFC 1995).
The GFC, in cooperation with the South Florida Water Management District
(SFWMD), conducted a draw down of Lake Kissimmee from December 1995- June
1996, dropping lake levels by approximately 1.7 m. The Lake Kissimmee draw down
provided an excellent opportunity to investigate the impact of management techniques
on the Florida apple snail (or FAS). Although data on the responses of some invertebrate
species to draw downs in central Florida are available (Wegener et al. 1974, Fox et al.
1977, Butler et al. 1992), no published studies document the responses of P. paludosa.
The impetus for the Lake Kissimmee draw down was habitat improvement for
fisheries and wildlife (GFC 1995). A conservation priority ranking system for Florida
wildlife identifies the snail kite and limpkin, largely due to their specialized diet of apple
snails, as deserving special attention for monitoring and management (Milsap et. al.
1990). Moyer et al. (1991) specifically recommended an evaluation of the impact of lake
draw downs on FAS. Although the majority of the snail kite population occupies wetland
habitats in southern Florida throughout the year, kites have been found to emigrate from
that area when the dry season escalates into a prolonged drought (Beissinger and
Takekawa 1983, Bennetts and Kitchens 1997). The central lakes system has been
identified as a critical refuge for the snail kite during these droughts. Kites have been
consistently nesting (even in non-drought years) on Lake Kissimmee and Lake
Tohopekaliga since the early 1980's (Beissinger and Takekawa 1983, Bennetts and
Kitchens 1997). The objective of the research presented in this chapter was to study the
impacts of lake restoration activity on snail abundance.
The general characteristics of Lake Kissimmee were presented in Chapter 2. Eight
sites were selected in 1995 to study the impacts of the draw down on sample snail
populations (Figure 4-1). Each site was classified as one of three types: 1) clean sand
site, where virtually no organic layer had accumulated and the sand was easily visible; 2)
flocculent organic site, where sand was visible in patches, but a thin (approximately 15
cm or less) flocculent organic layer was present; and 3) UOB sites, where the substrate
was a thick layer of roots and decaying parts of spatterdock (Nuphar luteum). Two sites
had a sand substrate, three had a flocculent organic layer, and three were UOB sites. The
draw down began in December 1995 and the lake was held at low pool stage from early
February through early June 1996 (Figure 4-2).
The inability to sample FAS under natural conditions had been the greatest obstacle
in understanding FAS population dynamics (Darby et al. 1997). A trap system for snails
was developed just prior to the Lake Kissimmee study. The trap design was a modified
version of those used to capture reptiles and amphibians (Dodd 1994,Greenberg et al.
1994, Enge 1997). A trapping method such as this provides an estimate of snail
abundance, not an actual density. The trap approach to determining relative snail
abundance was validated on Lake Kissimmee using throw traps extracted by a suction
dredge (Darby et al. 1997).
SClean Sand Substrate
A Flocculent Organic Layer
Over Sand Substrate
Route 60 Bridge
Figure 4-1. Sampling Sites on Lake Kissimmee, Osceola County, FL, for the pre-draw down
(1995) and post-draw down (1996, 1997) assessment of apple snail abundance.
A A B
2 15.0 -
G 14.5 -
13.0 I I I I
SEP NOV JAN MAR MAY JUL SEP NOV SEP NOV
Figure 4-2. Water levels on Lake Kissimmee from September 1995 through December 1996,
and from September 1997 through November 1997. Box A= 1995 trap array sampling, Box B=
telemetry study, Box C= 1996 trap array sampling, and Box D = 1997 trap array sampling. MSL
= mean sea level (in meters). Hydrology data from the South Florida Water Management District,
West Palm Beach, FL.
The general areas of the lake in which UOB removal, disking, and burning activity
were planned were provided by the Lake Restoration Section, GFC, Kissimmee, FL.
However, precise locations of the restoration activity were not known because of the
uncertainty of the low pool stage (contigent on rainfall) and site accessibility of the UOB
removal equipment. Ultimately, three of the eight study sites were scraped following
exposure of the substrates following the draw down; the remaining five sites experienced
only the dry down. The treatments among site categories were not equally distributed
An individual site was approximately 80-100 meters along the lake shore between
water depths of 0.3 and 1 m (the minimum and maximum depths suitable for trap arrays).
Following a trap check, some trap units were moved approximately 20 meters to a new
location within the site to obtain a more representative sample. Moving the traps was a
way to better represent a site while reducing the chance of getting multiple captures of
The shell width of all snails captured in trap arrays was measured. Based on
observations of Hanning (1979), and personal observations of snail growth and
reproductive activity, juvenile snails were defined as those having a shell width less than
25 mm. Snail sex was not considered due to the unreliability of assessing sex on smaller
snails and possibly larger snails prior to reproductive maturation. Based on the 1995
Kissimmee survey, an estimated 10% or less of the snails observed in the Fall exhibited
male characteristics as described by Hanning (1979).
All trap array sampling in the Fall of 1995, 1996, and 1997 occurred when lake
water levels were 15.1 to 15.7 m mean sea level (MSL). Once water levels reached
approximately 14.6 m MSL most of the sites became dry. All traps were set in at least 45
cm and up to 90 cm of water.
Descriptive statistics for the trap array data revealed that increased variances in
snail captures were associated with larger means of snail catch per trap. All trap array
data were therefore transformed using the logo10 (snail catch + 1) prior to analyses to meet
assumptions of normality and homogeneity of variances for the analyses (Sokal and
Despite the transformation, Levene's test for equal variances (with year and site as
grouped effects) revealed a significant difference due to substrate type (UOB, flocculent
organic, and sand) in 1995 (F2, 15=6.67, p=0.0085). Unequal variances would affect all
subsequent analyses, therefore a non-parametric approach for the analyses was taken. All
raw data were ranked, and these ranked data were modeled in a nested ANOVA
framework. Three separate ANOVAs (one for each substrate type) were required to test
for a year effect (1995, 1996, 1997) due to differential treatment of the sites within each
substrate type in 1996. The 1995 substrate effect was modeled as trap_array_catch=
substr site(substr). All muck sites and one flocculent site were excluded from the 1995
substrate analyses since no snails were captured in those sites. The year effects in a
given substrate type were modeled as trap_array_catch=year site(year). The trap array
catch was adjusted for number of trap checks per site.
Pre-Draw Down Snail Abundance
Substrate type appeared to have an effect on snail abundance (Table 4-1). The
analyses were limited to only 2 or 3 sites per substrate type, and there was a significant
site effect within substrate type (Table 4-2). Therefore the analyses did not reveal a
statistically significant relationship. However, it is apparent that UOB sites, where no
snails were found with trap arrays or with the suction dredge (during the validation), did
not support apple snails.
Table 4-1. Snail abundance as a function of substrate type. SAND refers to sites in
which clean sand was easily visible, SAND w/FO refers to sand covered with
flocculent organic material, and UOB refers to substrates dominated by decaying
Nuphar luteum and associated root systems.
Substrate Type Sites Snails / trap check (SE)
SAND 7,8 44 ( 13.0)
SAND w/FO 1,2,5 11.8 (9.70)
UOB 3,4,6 0 (-0)
Table 4-2. Analysis of Variance (ANOVA) table for snail abundance as a function
of type of substrate (sand, sand with flocculent organic) for 4 sites on Lake
Kissimmee in 1995. Sites with no captured snails were excluded from the analysis.
The SITE effect is nested within SUBSTRATE type. Trap array data were
transformed using logl0 (snails + 1) prior to analyses.
Source df SS MS F Prob>F
SUBSTRATE 1 48.5 48.5 1.81* 0.31*
SITE(SUBSTRATE) 2 53.5 26.8 4.94 0.036
error 12 48.7 5.41
*The F-statistic for the SUBSTRATE effect was calculated using the MS for
SITE nested within SUBSTRATE as the error term because SITE(SUBSTRATE)
had a significant effect in the model (Sokal and Rohlf 1981, SAS, Inc. 1992).
Post-Draw Down Snail Abundance
The inability to predict precisely which sites would receive which treatments
resulted in an unbalanced design with regard to substrate type and whether a site was
"treated" by the dry down only, or by the dry down and also by scraping (Table 4-3).
Separate ANOVAs were done for each available substrate/treatment type with year as the
Table 4-3. Treatment of the eight sampling sites showing the imbalanced design
associated with the three substrate categories [CLEAN SAND, SAND with
flocculent organic (FO), and UOB]. Site 6 is not shown here since it was not
sampled in 1996 or 1997.
Category/Site No. Treatment
7 dry down only
8 dry down only
SAND with FO
1 dry down and scrape
2 dry down and scrape
5 dry down only
3 dry down only
4 dry down and scrape
In 1996 and 1997, site 6 was inaccessible due to a combination of higher lake levels
and consolidation of the substrate, which resulted in water too deep (>1.7 meters) to use
the traps (Table 4-4). No area in the vicinity had comparable substrate and vegetation for
meaningful comparisons with the original 1995 trap location.
FAS abundance in the study sites in 1996 and 1997 was reduced to 20% and 13%,
respectively, of the pre-draw down (1995) abundance (Figure 4-3). Highly significant
declines in the sand substrate sites (sites 7 and 8) accounted for the majority of the first
year 80% decline (Table 4-5) (Figure 4-4). The scraped flocculent organic sites showed