xml version 1.0 encoding UTF-8 standalone no
record xmlns http:www.loc.govMARC21slim xmlns:xsi http:www.w3.org2001XMLSchema-instance xsi:schemaLocation http:www.loc.govstandardsmarcxmlschemaMARC21slim.xsd
leader nam Ka
controlfield tag 001 001796857
007 cr mnu|||uuuuu
008 070723s2006 flu sbm 000 0 eng d
datafield ind1 8 ind2 024
subfield code a E14-SFE0001614
Barton, Sheri L.
The influence of habitat features on selection and use of a winter refuge by manatees (Trichechus manatus latirostris) in Charlotte Harbor, FL
h [electronic resource] /
by Sheri L. Barton.
[Tampa, Fla] :
b University of South Florida,
ABSTRACT: Investigating alternate winter refuges for Florida manatees is increasingly important as sustained warm-water discharges from industrial and some natural sites becomes more uncertain. This study examined habitat features of possible importance to manatees by comparing a winter refuge in Charlotte Harbor, FL (the Matlacha Isles canal system) to two nearby, seemingly similar sites that are not frequented by manatees during winter. Water temperature, salinity, boat traffic, canal depth, and tidal flushing were assessed at these sites. Additionally, this study examined when and how manatees use the Matlacha Isles refuge by documenting movements, habitat use, and behaviors of manatees during the winters of 1999/2000 through 2001/2002. Water temperatures had a profound influence on manatee selection of Matlacha Isles over the two comparison canal systems. Matlacha Isles did not experience the sudden drops in water temperature following cold fronts, extreme low temperatures, o r long periods of temperatures below manatees' reported thermal tolerance of 18-20Â¨C that were recorded in Matlacha Pass (ambient) and the two comparison canal systems. Heat retention within Matlacha Isles may be associated with greater water depth and lower tidal flushing. Salinity and boat traffic did not seem to influence site selection by manatees. During moderately cold weather, manatees occupying Matlacha Isles forage at night in nearby Matlacha Pass and return early in the morning to Matlacha Isles, where they primarily rest all day. Neither tidal state nor boat traffic levels affected manatee travel patterns into or out of Matlacha Isles. Manatees may passively thermoregulate in the warmer waters of Matlacha Isles during the day (when they are inactive) and sustain their body temperatures at night through the heat generated during traveling to feeding sites and during ingestion (chewing) and digestion. During extreme or prolonged cold weather, Matlacha Isles provides inad equate warmth for manatees; during such times, most of them travel to a power plant on the Orange River, approximately 50 kilometers away. Findings from this study may inform resource managers as they consider attributes manatees find desirable or necessary in winter. Such information will help managers create new or enhance existing winter refuges to protect manatees.
Thesis (M.A.)--University of South Florida, 2006.
Includes bibliographical references.
Text (Electronic thesis) in PDF format.
System requirements: World Wide Web browser and PDF reader.
Mode of access: World Wide Web.
Title from PDF of title page.
Document formatted into pages; contains 67 pages.
Adviser: Henry R. Mushinsky, Ph.D.
t USF Electronic Theses and Dissertations.
The Influence of Habitat Features on Select ion and Use of a Winter Refuge by Manatees ( Trichechus manatus latirostris ) in Charlotte Harbor, Florida by Sheri L. Barton A thesis submitted in partial fulfillment of the requirements for the degree of Master of Science Department of Biology College of Arts and Sciences University of South Florida Co-Major Professor: Henry R. Mushinsky, Ph.D. Co-Major Professor: John E. Reynolds, III, Ph.D. Earl D. McCoy, Ph.D. James A. Powell, Ph.D. Date of Approval: May 11, 2006 Keywords: Habitat selection, Th ermoregulation, Temperature, Foraging, Activity Patterns Human Disturbance Copyright 2006, Sheri L. Barton
Acknowledgements I would like to thank my committee, Dr s. Henry Mushinsky, John Reynolds, James Buddy Powell, and Earl McCoy for all of your guidance, support, patience, and time. This project would not have been possible wi thout the help of current and former Mote Marine Laboratory staff Rachel Nostrom, Ke rri Scolardi, and Tere sa Kessenich, as well as numerous interns and volunteers, who e ndured many long and often cold hours of fieldwork. I want to thank Jay Sprinkel for assistance with data an alyses and helping me find the light at the end of the tunnel. I also want to thank Maija Gadient, as well as Judy Dorsman (The Sun and the Moon Inn) for allowing us to use your docks and homes during fieldwork. I am grateful to Jessi ca Koelsch for introducing me to manatee research, Dr. Ernest Estevez for advice on habitat sampling, and my employer, Mote Marine Lab, for financial suppor t and flexibility. I am also grateful to Florida Fish and Wildlife Conservation Commission staff for a ssistance with capturing the manatees for this study. Finally, I want to thank my family and friends for all of your encouragement and support. This project was funded through a series of contracts with the Florida Fish and Wildlife Conservation Commission Fish and Wi ldlife Research Inst itute and through the Mote Scientific Foundation. Fieldwork wa s conducted under U.S. Fish and Wildlife Service permit number MA773494-1 issued to the Florida Fish and Wildlife Conservation Commission Fish and Wildlife Research Institute The capture and tagging of manatees for this study were certified by the Mote Marine Lab Institutional Animal Care and Use Committee under protocol number 99-11-SB1.
i Table of Contents List of Tables iii List of Figures iv Abstract vi Introduction 1 Objectives 5 Methods 6 Study Area 6 Data Collection 7 Habitat Characterization 7 Temporal Use of Matlacha Isles 10 Capture and Tagging 11 Behavioral Observations of Tagged Manatees 12 Data Analysis 14 Habitat Characterization 14 Temporal Use of Matlacha Isles 14 Behavioral Observatio ns of Tagged Manatees 16 Results 18 Habitat Characterization 18 Water Temperature 18 Salinity 20 Boat Traffic 21 Canal Depths 22 Tidal Flushing 23 Ground-water Seeps and Submerged Aquatic Vegetation 23 Temporal Use of Matlacha Isles 23 Behavioral Observations of Tagged Manatees 25
iiDiscussion 28 Why Do Manatees Use Matlacha Isle s Rather Than Other Seemingly Similar Sites During the Winter? 28 When and How is Matlacha Isles Used By Manatees? 34 Conservation and Management Implications 37 Literature Cited 39
iii List of Tables Table 1 Descriptive statistics for daily mean water temperature in the three canal systems and Matlacha Pass for days when water temperature in Matlacha Pass was <18oC (N=99 days). 47 Table 2 Descriptive statistics for c ounts of boats per day in the three canal systems and Matlacha Pass. 47 Table 3 Descriptive statistics of de pth (m) measurements in the three canal systems. 47 Table 4 Descriptive statistics for seep detection transects in the Matlacha Isles canal system. 48 Table 5 Results of linear regressi ons correlating counts of manatees traveling into the Matlacha Isle s canal system with average water temperature during the 24, 72-, and 120-hour periods prior to sampling. 48 Table 6 Sampling effort for behavioral observations of tagged manatees. 48 Table 7 Summary of behaviors recorded and descriptive statistics for behavioral observations. 49 Table 8 Results of T-tests comparing percentages of each behavior when water temperatures were warmer versus colder the day of sampling than the previous 24or 72-hours. 50 Table 9 Results of Mann-Whitney rank sum tests comparing percentages of each behavior wh en water temperatures were colder versus warmer the day of the sampling than the previous 24or 72-hours. 50
iv List of Figures Figure 1 Map of the study area, whic h includes Matlacha Pass from the northernmost tip of Pine Island to the power lines just south of channel marker 30. 51 Figure 2 Maps of the three canal systems used in the study. 52 Figure 3 Daily minimum bottom water temp eratures at each sampling site. 53 Figure 4 Frequencies of daily mean bottom water temperatures at each sampling site during the three winters. 54 Figure 5 Histogram showing the differe nces in daily mean bottom water temperature between each sampling site and Matlacha Pass (ambient). 55 Figure 6 Surface and bottom salinities in Matlacha Isles, northern Pine Island, and West Island. 56 Figure 7 Surface and bottom salinities in Matlacha Isles, northern Pine Island, and West Island when all sites were sampled on the same day. 56 Figure 8 Surface and bottom salinities at each sampling station within the Matlacha Isles canal system. 57 Figure 9 Mean number of boats obse rved per hour at each location during 24-hour surveys (N=2 days). 58 Figure 10 Mean number of boats obse rved per hour at each location during daytime surveys (N=4 days). 58 Figure 11 Estimated volume of water (m3) remaining within each canal system during a tidal cycle. 59 Figure 12 Total number of manatees obs erved traveling into and out of the Matlacha Isles canal system each hour, plotted on a circular scale. 60
vFigure 13 Frequencies of tid e readings at different heights recorded at the entrance of the Matlacha Isle s canal system during 24-hour sampling periods. 61 Figure 14 Number of manatees obser ved traveling into Matlacha Isles versus average water temperat ure in Matlacha Pass 24-, 72-, and 120-hours prior to each sampling period. 62 Figure 15 Total frequencies of each activity observed over all sampling periods. 63 Figure 16 Total frequencies of each ac tivity observed during the 11 sampling periods that focal manatees us ed Matlacha Isles during the day and Matlacha Pass at night. 64 Figure 17 Total frequencies of each ac tivity observed dur ing the 8 sampling periods that focal manatees remained in Matlacha Pass 65 Figure 18 Percentages of time manat ees were engaged in the observed behaviors versus the difference between the temperature on the day of the sampling and the prior 24-hour period. 66 Figure 19 Percentages of time manat ees were engaged in the observed behaviors versus the difference between the temperature on the day of the sampling and the prior 72-hour period. 67
vi The Influence of Habitat Features on Selecti on and Use of a Winter Refuge by Manatees ( Trichechus manatus latirostris ) in Charlotte Harbor, Florida Sheri L. Barton ABSTRACT Investigating alternate winter refuges for Florida manatees is increasingly important as sustained warm-water discharges from i ndustrial and some natural sites becomes more uncertain. This study examined habitat featur es of possible importance to manatees by comparing a winter refuge in Charlotte Harb or, FL (the Matlacha Is les canal system) to two nearby, seemingly similar sites that are not frequented by ma natees during winter. Water temperature, salinity, boat traffic, canal depth, and tidal flushi ng were assessed at these sites. Additi onally, this study examined when a nd how manatees use the Matlacha Isles refuge by documenting movements, habita t use, and behaviors of manatees during the winters of 1999/2000 through 2001/2002. Water temperatures had a profound influence on manatee selecti on of Matlacha Isles over the two comparison canal systems. Matlacha Isles did not experi ence the sudden drops in wate r temperature following cold fronts, extreme low temperatures, or long periods of temperatures below manatees reported thermal tolerance of 18-20oC that were recorded in Matlacha Pass (ambient) and the two comparison canal systems. Heat retention within Matlacha Isles may be associated with greater water depth and lowe r tidal flushing. Salinit y and boat traffic did not seem to influence site selection by ma natees. During moderately cold weather, manatees occupying Matlacha Isles forage at night in nearby Matl acha Pass and return early in the morning to Matlacha Isles, where they primarily rest all day. Neither tidal state nor boat traffic levels aff ected manatee travel patterns in to or out of Matlacha Isles. Manatees may passively thermoregulate in th e warmer waters of Matlacha Isles during the day (when they are inactive) and sustai n their body temperatures at night through the
viiheat generated during traveling to feedi ng sites and during ingestion (chewing) and digestion. During extreme or prolonged cold weather, Matlacha Isles provides inadequate warmth for manatees; during such ti mes, most of them travel to a power plant on the Orange River, approximately 50 kilome ters away. Findings from this study may inform resource managers as they consider attributes manatees find desirable or necessary in winter. Such information will help managers create new or enhance existing winter refuges to protect manatees.
1 INTRODUCTION The Florida manatee ( Trichechus manatus latirostris ), a subspecies of the West Indian manatee, occupies the coastal, estuarin e, and riverine habitats of the southeastern United States (Domning and Hayek 1986). Outs ide of winter, Florida manatees may be found as far west as Louisiana and as far nor th as Virginia, with occasional wanderers outside this range (Reynolds and Odell 1991 ; Reynolds and Powell 2002). Their winter range, however, is much more restricted becau se of physiological traits that limit their thermal tolerance (Irvine 1983, OShea 1988). Manatees have a metabolic ra te 17-22 % of that expected for an animal their size, a high rate of thermal conductance, and a limite d capacity for thermogenesis (Irvine 1983). These traits allow manatees to subsist on a low-energy food source of aquatic and semiaquatic vegetation, while also causing them to be vulnerable to cold. Water temperatures of 15-20oC prompt physiological and behavioral changes in manatees, such as an increase in metabolic rate and migra tions to warmer water (Irvine 1983). When exposed to water temperatures of 18-20oC for several days, captive manatees feed erratically, and feeding may cease altogether at temperatures below 15-18oC (Campbell and Irvine 1981). Examinations of carcasses suggest that extended periods of exposure to cold can cause manatees to die slowly as they stop feeding, use fat reserves, and eventually starve (Buergelt et al. 1984; OShea et al. 1985; Bossart et al. 2002). Shortterm exposure to extreme cold can result in death from hypothermia (OShea et al. 1985), although Hartman (1979) observed manatees entering water as cold as 13.5oC. Manatees appear to be unable to survive for long peri ods of time in water temperatures lower than 16oC; however, lethal temperat ures and exposure periods ha ve not been well documented (Ackerman et al. 1995). Young manatees app ear to have a higher vulnerability than adults to cold temperatures, perhaps because of excessive amounts of heat loss during cold weather, caused by such physical charact eristics as their surf ace area:volume ratios
2and lower amounts of insulation, and inadequa te experience with seeking warm water (OShea et al. 1985). To survive periods of cold weather, mana tees need to seek refuge in areas with warm water. Generally, when water temperatures fall below 20oC, many manatees migrate to warm-water refuges (Hartman 1979; OShea 1988; Lefebvr e et al. 1989; Laist and Reynolds 2005a and b). Moor e (1951) suggested th at the historical winter range of Florida manatees existed within the southern regions of Florida and at natural warmwater springs, with the northernmost boundari es being the Sebastian River on the east coast and Charlotte Harbor on the west coast. Recent development of coastal habitats, including the introduction of industrial warm -water effluents and obstruction of access to some springs, has helped to alter and expa nd the winter range of Florida manatees. Whereas most of the winter refuges remain w ithin central and southe rn Florida (U.S. Fish and Wildlife Service 2001), at least 60% of all Florida ma natees now rely on 10 major power plant effluents and 15% rely on 4 major warm-water springs to survive periods of cold weather (Laist and Reynolds 2005a and b). Some currently used springs (at Crystal and Homosassa Rivers and Blue Spring) lie well north of th e boundary of the historical winter range suggested by Moore (1951), a nd Laist and Reynolds (2005a and b) suggest that springs in the northern and central parts of Florida could become important warmwater refuges in the future, if access to them is restored for manatees. One of the major threats facing manatees is the potential loss of warm-water refuges (U.S. Fish and Wildlife Service 2001). Efflue nts from power plants can be unreliable as they may cease temporarily because of routine maintenance and equipment failure. When these temporary cessations have occurred in the past, some manatees remained at or near the site, as if they were waiting fo r the effluent flow to resume (Reynolds and Wilcox 1986; Packard et al. 1989; Reynolds 200 0). A more serious situation would be the complete shutdown of a plant, particular ly older power plants that are due to be retired or are no longer economically productive. Some of the natural springs providing wa rm-water refuges for manatees may also become unreliable (U.S. Fish and Wildlife Se rvice 2001). The depletion of the aquifer, caused by increased human demands for water and occasional periods of drought, has
3lowered some important spring flow rates and resulted in less warm-water output (Vergara 1994; Sucsy et al. 1998). If flows c ontinue to decrease, these springs may be insufficient at warming the surrounding water to the level at which manatees can survive during intensely cold periods. Manatees relying on sites impacted by these alterations may succumb to cold stress if they are una ble to find a suitable alternative site. Whereas most of the winter refuges fo r manatees, consisting of natural warm springs or industrial discharges, have been well documented, many lesser known alternate sites also ex ist and require further investigati on (U.S. Fish and Wildlife Service 2001). Many of the alternate s ites are not influenced by indus trial discharges and consist of dredged boat basins and canal systems, apparently capable of retaining water temperatures above those in adjacent wate rways (Laist and Reynolds 2005b). Possible explanations for the heat retention at thes e sites include ground-wa ter seeps or springs and/or the structure and confi guration of certain boat basins or canal systems that may be able to maintain warmer water temperatures through limited tidal flushing and solar heating of the water held within them (U.S. Fish and Wildlife Service 2001). The current Florida Manatee Recovery Pl an (U.S. Fish and Wildlife Service 2001) has identified the pro tection, enhancement, and invest igation of non-industrial warmwater refuges as Objective 3.2.3; and the role th at winter refuges play in the survival of the species is considered extremely impor tant for managers and researchers to understand. The Florida Manatee Recovery Plan advocates using data collected on attributes of non-industrial warm-water refuges that are attractive to manatees to develop a series of additional sites for manatees as a safeguard in the event that warm water at a power plant or spring ceases to exist (U.S. Fish and Wildlife Service 2001). Several habitat features have been identifie d as being important to Florida manatees. Access to warm water is perhaps most infl uential in determining their distribution, particularly during the wint er (Hartman 1979; Irvine 1983; OShea 1988; Lefebvre et al. 1989). Other key habitat features are abundant seagrasses or other food source, access to freshwater, and absence of hu man disturbance (Hartman 19 79; Reynolds 1999; U.S. Fish and Wildlife Service 2001).
4 Manatees are herbivorous, feeding prim arily on submerged, floating, and emergent plants, and incidentally on items such as t unicates and epiphytic organisms growing on the plants (Hartman 1979; Packard 1981). Beca use of the relatively low quality of this food source, manatees need to spend a cons iderable amount of time foraging to fulfill their energy requirements. Studies by Be ngtson (1983) and Ethe ridge et al. (1985) estimate that manatees may spend approxi mately 5 hours per da y during the winter consuming between 4 and 9% of their body wei ght. As hindgut dige sters, the bulk of their digestion takes place in an enlarged large intestine (hindgut) where microbes break down the cellulose in the plant material. Th is microbial activity is thought to generate enough heat to help regulate th e body temperature of well-fed manatees during the winter (Rommel et al. 2003). Florida manatees are euryhaline; howeve r manatees frequently seek sources of freshwater (Hartman 1979; Belitsky and Belitsky 1980; Powell et al. 1981; Powell and Rathbun 1984; U.S. Fish and Wildlife Service 2 001). Scientists have even used this to their advantage by baiting manatees with fres hwater hoses to specific sites for capture and tagging as part of telemetry studies (Reid et al. 1995; Deutsch et al. 2003). Manatees may not necessarily need to drink freshwater, however, as the structure of the manatee kidney suggests manatees are able to excret e excess salt by producing concentrated urine (Hill and Reynolds 1989). Manatees also posse ss renal and endocrine mechanisms that would allow them to maintain sodium balan ce and avoid dehydration (Ortiz et al. 1998). These mechanisms, however, may be unable to prevent dehydration during extended periods without freshwater (Ortiz et al. 1998) It has also been shown that manatees exposed to saltwater for extended periods ha ve reduced body mass from the oxidation of fat, suggesting that manatees are able to produce the water they need by oxidizing fat stores (Ortiz et al. 1999). Manatees may fa vor habitats where osmotic stress is minimal and freshwater is accessible, because it may be more metabolically advantageous (OShea and Kochman 1990). Regardless of th e reason, it is apparent that manatees are attracted to sources of freshwater. Human waterborne activities, such as boa ting and swimming can have adverse effects on manatees, both indirectly and directly. Manatee habitat is affected by the scarring of
5seagrass beds, increased turbidity, an d noise pollution caused by over 980,000 boats registered in the state of Florida (Sargent et al. 1995; Florida Fish and Wildlife Conservation Commission 2005a) that share these areas with manatees. Collisions with watercraft constitute the single largest identifie d cause of mortality of manatees in Florida (Florida Fish and Wildlife Conservation Commission 2005b); and those manatees that survive these collisions bear scars and mutila tions from their injuries. The disturbance and harassment by boats, as well as other human waterborne activities, such as swimming and scuba diving, affect manatee distributi on, habitat use, energetics, and behavior (Buckingham et al. 1999; Re ynolds and Powell 2002; King and Heinen 2004; Nowacek et al. 2004). Objectives There were two main objectives of this study. The first ob jective was to collect data on physical and chemical features of a non-indus trial winter refuge and compare them to those of nearby sites that are not frequented by manatees during the winter. This first objective provides information on habitat ch aracteristics that manatees may find attractive in a winter refuge and may also help to determine why manatees use some seemingly similar sites over others. The second objective was to investigate whether environmental factors, such as time of day, tidal height, and water temperature influence site use and behavioral patterns of manatees This second objective further suggests why manatees use certain sites duri ng the winter, while also giving insight into when and how these sites are used.
6 METHODS Study Area The Matlacha Isles canal system is a winter refuge for manatees in southwestern Florida. Located in western Cape Coral, Lee County, approximately 26.64o north latitude and 82.05o west longitude (Figure 1), the Matlacha Isles canal system is an approximately 5-km2 series of brackish canal s through a residential ne ighborhood. Although dredged in the 1970s as part of the northern Cape Coral canal system, Matlacha Isles remains separated from the remainder of this larger sy stem by a small dam and boatlift, and it is the only section that has open water access. Aerial surveys have documented manatees using Matlacha Isles year-round and especially during the winter (Florida Department of Environmental Protection 1998; Mote Ma rine Laboratory unpublished data). Photographic identification studies have furt her shown that manate es consistently use Matlacha Isles throughout the winter and spri ng, a high proportion of individuals return annually, and many of these individuals travel the approximately 50 kilometers between this site and the Florida Po wer & Light Company power pl ant in Ft. Myers throughout the winter (Koelsch and Barton 1999). The study area centered on the Matlacha Is les canal system and encompassed Matlacha Pass, from its northernmost boundary to the power lines south of Little Pine Island (Figure 1). Matlacha Pass is a 65-km2 body of water in the lower Charlotte Harbor estuary, between Pine Island and the city of Ca pe Coral. It contai ns abundant sea grass beds, and consists of numerous small ba ys, mangrove islands, backwater areas and shallows. The Matlacha Isles canal system is accessed by a dredged channel, which runs along the northeastern side of the Matlacha Bridge and causeway and continues east, through a small shallow bay (Site 3), locate d between Matlacha Pass and Matlacha Isles (Figure 2a). The canal system begins in the so utheastern corner of this bay as a single 1.5 m deep and 8-10 m wide entrance/exit canal. The entrance/exit canal then branches into
7multiple finger canals, with a lake (Site 2) at the easternmost end. Manatees can usually be found in three of the dead-end canal s (Sites 1, 2, and 6) from late fall through early spring. Two other dead-end canals (Sites 4 and 5) appear to be very similar in structure to Site 6; however they have very limited use by manatees. Sites 7a-c, 8, and 9 lead to other canals and are used prim arily by manatees as travel corridors. Two nearby canal systems were included in the study area and used for comparison with the entire Matlacha Isles canal system because they are generally similar in configuration but have little to no manatee us e. The canals located at the northeastern end of the northern Pine Island canal system (Figure 2b) are physica lly the most similar to those of Matlacha Isles. A small shallow b ay is located to the north of the entrance to the canals. The canals are branched, forming multiple finger canals, with similar shorelines to those in Matlacha Isles. The canals in West Island are located directly across Matlacha Pass from Matlacha Isles (Fi gure 2a). The canals on West Island that were used for comparison are less similar to those in Matlacha Isle s. The West Island canals open directly into Ma tlacha Pass, are much shorter and have less branching than those in Matlacha Isles, and also have shorelin es that are almost enti rely sea-walled. All three canal systems are just a short distance from dense sea grass beds: the Matlacha Isles and northern Pine Island canal systems are both less than 1.5 km from the nearest dense sea grass bed and the West Island can al system is approximately 0.75 km away. Data Collection Habitat Characterization Data on habitat features were collected during three field seasons (winter early spring 1999/2000 2001/2002). These features included bottom water temperatures, surface and bottom salinity, amount of boat tr affic, tidal flushing, and bathymetry. Additionally, presence/absen ce of ground-water seeps was investigated within the Matlacha Isles canal system to determine if there was a source of warm-water input and/or a source of freshwater. Availability of aquatic vegetation was also investigated within the Matlacha Isles canal system to determine whether submerged aquatic
8vegetation (SAV) was available for manatees to consume within the canal system (which would eliminate manatees need to leave the site to forage). Temperature loggers measuring bottom wate r temperatures were placed throughout the study area. Water temperatur e loggers were used to comp are: 1) Matlac ha Isles to ambient (Matlacha Pass), 2) Matlacha Isles to the West Island and northern Pine Island canal systems, and 3) specific sites within Matlacha Isles (sites that have frequent vs. infrequent/no use by manatees). A total of 11 loggers was used within Matlacha Pass, Matlacha Isles, and comparison canal systems in northern Pine Island and West Island. Water temperature loggers within canal systems were attached to the ends of residential docks. Water temperature loggers within Matlacha Pass were attached to channel markers in three different sections of the pa ss: northern (marker 71), middle (marker 56), and southern (marker 30). All temperatures were collected at 40-minute intervals using Optic StowAway Temp data loggers (Onset Com puter Corporation; accuracy = + 0.2oC). The water temperature loggers were housed in casings made of PVC with multiple holes, which allowed ample water flow to reach the loggers, while protecting them from biofouling. Each logger and casing was cl eaned and data from each logger were downloaded approximately once per month. To measure salinity in the three canal sy stems, sampling stations within Matlacha Isles, northern Pine Islan d, and West Island were esta blished. Surface and bottom salinity and temperature were collected between October and April 1999/2000 2001/2002 at each sampling station using a YSI 30 handheld SCT meter with a 7.6 m cord (accuracy = + 0.8 ppt for salinity and + 0.1oC for temperature). Ancillary data collected at each station included time, ai r temperature, surface and bottom water temperature, and depth (m). Boat traffic surveys were conducted to m easure the amount of human disturbance in four sections of the study ar ea: 1) the entrance of Matlacha Isles, 2) West Island canal system, 3) northern Pine Island canal system and 4) Matlacha Pass. The surveys were
9conducted for approximately 10-hour periods to assess daytime boat use and for 24-hour periods to assess both day and nighttime boa t use during winter and spring 2002. Each time a motorboat was observed, the time of da y was recorded along with the boats origin (i.e., first observed location), destination (i.e., last observe d location), and relative speed (i.e., plane, plow, slow, idle). Differences in water temperature and sa linity between bodies of water can be partially explained by the amount of tidal ex change or flushing the bodies of water experience. The amount of tidal exchange or flushing in each of th e three canal systems was examined to determine whether the Ma tlacha Isles canal system has lower tidal exchange or flushing than the canals in West Island and northern Pine Island. The estimated total volume that does not change during a given tidal cycle was calculated for each of these canal systems using a simple tidal prism model (Dyer 1973): volume not exchanged = total canal system volume tidal prism Canal system volumes were estimated by multiplying the mean canal depth by the estimated surface area of each canal. Mean de pths were obtained by measuring depths at three points (2 edge and 1 center) along 35 cross-sections of each individual canal, correcting for tide, and calculating the average for each canal system. Depths were measured using a standard lead line. Cana l surface areas were es timated from digital maps with ArcView GIS software (version 3.3). The tidal prism is the domain volume between high and low tide, and was estimated by multiplying the tidal range (from predicted tide charts) by th e surface area of each of th e canal systems (Dyer 1973; Monsen et al. 2002). To detect evidence of ground-water seeps, transects measuring surface and bottom salinity and water temperature were conducte d within the Matlacha Isles canal system March 2000. Transects were performe d in a zigzag pattern using a YSI 30 (YSI, Incorporated) handheld SCT (salinity, conduc tivity, temperature) meter with a 7.6 m
10cord. Because manatees are attracted to wa rm-water (during the winter) and freshwater, it is reasonable to assume that if warmand/ or freshwater seeps exist within the canal system, manatees would be concentrated near t hose areas. Therefore, transects were only conducted within the three canals where manat ees are regularly obser ved (sites 1, 2, and 6). Transect surveys of submerged aquatic ve getation were conduc ted in the Matlacha Isles canal system and the small bay out side in August 2000. These surveys were conducted during the summer to allow any seag rasses or other vegetation that may have been grazed and/or uprooted by foraging mana tees to grow. Snorkelers inspected the canal bottoms and seawalls. In areas known to have re sident alligators, however, transects were conducted usi ng a long handled rake to feel along the bottom and limit the amount of time snorkelers were in the water. Temporal Use of Matlacha Isles Photographic identification st udies in Matlacha Isles have provided information on site fidelity and movement and travel patterns of manatees th at use this site (Koelsch and Barton 1999). In addition to a seasonal patt ern of manatee presence within the Matlacha Isles canal system, preliminary observations su ggested that they also have a distinctive pattern of daily use of Matlacha Isles, as i ndividuals were document ed traveling into and out of the canal system during certain times of the day. To examine the conditions under which manatees use Matlacha Isles, observati ons of manatees were documented at the entrance of the Matlacha Isles canal system Observations were conducted from a balcony and/or dock at the entrance to the can al system. These observations were carried out in 24-hour periods, generally beginning around 0900 hours. Data were collected each time a manatee passed the entran ce and/or at half-hour interv als. These data included: time of day, tidal height, tidal direction, sighting conditions (how well manatees could be spotted), weather, wind speed and direction, num ber of manatees sighted, and their travel direction (into or out of the can al system). Tidal height was measured as the distance to
11the surface of the water from a fixed point on the survey dock. Most surveys were conducted for 2-3 consecutive days. During most of the sampling periods, mana tees could be easily spotted as they traveled into or out of the canal system. Fortunately, a sa ndy shoal was just outside the mouth of the canal system and the entrance canal was narrow, shallow, and had a sandy bottom. These features enabled observers to see to the bottom so that the manatees could be counted even when they were not breaki ng the surface of the water. At higher tides and lower light levels, manatees were not as easily observed or counted; under these conditions, manatees had to be closer to or break the surface of the water in order to be detected; or more subtle cues were used, such as manatee footprints (smooth swirls on the surface of the water caused by the swimming action of the manatees tail) or the sound of breathing. At night, a dual shop-light on a 3-ft stand wa s used to provide enough light to see the cues listed above. Capture and Tagging Tagging of manatees was used in this study to provide a means of facilitating visual contact of individuals during be havioral sampling. The taggi ng technique used in this study involved capturing and temporarily restraining indi vidual manatees while the tagging assemblies were fitted. Sp ecific manatees that were known to return annually to the study area during the winter were targeted for capture and tagging. The number of manatees often present in the canals duri ng the winter, however, made it difficult to capture only targeted individuals. One area within Matlacha Isles (Figure 2) has a few sloping sandy shorelines, allowing captured manatees to be pulled onto the banks. The method used to capture manatees at this site involved a 122-m l ong, 9-m deep net with 10-cm mesh that was stretched across the mouth of Site 1, which is known to have high manatee use during the winter. The net was positioned so that one end was kept on th e net boat at one side of the canal, while the other was st ationary at the bank where the captured manatees were hauled. Positioning the net in this manner formed an open pocket across the canal into which the manatees swam as they tried to en ter this canal. When one or more manatees
12approached the net, the net boat circled to the opposite bank, enclosi ng the manatee(s) in the net. The net was then used to secure the manatees and pull th em onto the bank. If more manatees were netted at one time th an available personnel could safely handle, some of the individuals were released immediately or soon after capture. Individuals recognized as targ et animals were fitted with a standard manatee peduncle belt, tether of appropriate st rength, and either a single ve ry high frequency (VHF) radiotransmitter or platform transmitter terminal (PTT) with VHF and ultrasonic transmitters, as described by Reid et al. (1995) and Deut sch et al. (2003). Each tag was uniquely colored in order to distinguish individuals and aid obtaining vi sual locations. Because of the method of capture that was used, target manatees were captured infrequently; therefore it was not uncommon for a previously unknown manat ee to receive a tag. Once data collection and/or taggi ng procedures were completed, the manatees were moved back into the water and released. A tota l of 14 manatees was tagged during the study (2000: 3 females, 2 males; 2001: 4 females, 1 male; 2002: 2 females, 1 male). Tags were removed or lost from the manatees by the summer of the y ear each manatee was tagged. Behavioral Observations of Tagged Manatees Focal animal observations were conduc ted on tagged indivi duals during 24-hour sampling periods to determine how manatees th at use Matlacha Isles distribute their time among different activities throughout the day a nd examine the affect of temperature on behavioral patterns. Prior to the onset of each sampling period, the focal manatee was chosen randomly among the ta gged individuals present with in the study area. The focal manatee was chosen randomly because of the small sample size and uneven ratios of males and females, age classes, and reproduc tive status of the tagged individuals. Instantaneous sampling methods (Altmann 1974) were used to collect data on the focal individuals activity at fou r-minute intervals (following Koelsch 1997). A watch with a countdown timer was used to mark the four-min ute sample points. If the focal manatees activity could not be determined within one minute of the sample point, no behavioral data were collected for that point. Additiona lly, if the focal manatees activity appeared
13to be altered by observers presence, no beha vioral data were co llected during those intervals and responses were noted in co mments. Focal manatee observations were conducted from 6-7 meter outboard motorboats with observation towers. Boats were anchored or docked nearby, drifting, or mane uvered with electric trolling motors to minimize disturbance of the manatees. The data collected at each sample point included: time of day, activity, confidence level, location, habitat type, and sighting c onditions. Confidence level was a measure of the observers certainty of the manatees activity at the sample point. The range was from 1 to 4, with 1 being the highest level of confidence and 4 being the lowest. The activities (from Hartman 1979; Urian and We lls 1996; and Koelsch 1997) that were recorded include: Rest encompassed both surface re sting/basking and bottom resting Surface rest floating motionless at or near the surface of the water without changing location Bottom rest submerged for extended periods of time, coming to the surface in the same general area only to breathe, then submerging again Travel directed movement within or among sites Mill non-directed move ment within a site Feed evidence of feeding was used to determine this, i.e. grass in mouth, chewing; focal manatee within a s ea grass bed, surrounded by cropped sea grass blades and/or sediment cloud Possible Feed -used when the focal manatee was in a sea grass bed, either remaining in the same spot or movement was very slight, and conditions made it difficult to confirm feeding. This activity category was designated frequently at night. Clues that were used to determine this include: feeding confirmed at least once while the focal manatee was at the same location, but conditions made it difficu lt to confirm feeding throughout the time the manatee was in that spo t; cropped sea grass blades floating downwind of focal manatee; and/or focal manatee in subgroup in which other manatees were confirmed to be feeding, but conditions made it difficult to confirm focal manatees behavior. Chewing sounds recorded by a hydrophone were also used to he lp assign this behavioral category during the first year of the study.
14Socialize tactile or active (grabbing, rolling, splashing, et c.) interaction with at least one other manatee Play tactile or active (grabbing, m outhing, pushing, etc.) interaction with an inanimate object Habitat type was based on definitions used by Koelsch (1997) and included the following: seagrass bed (GB), dredged basin (DB, >50% altered shoreline), dredged channel (DC, <50% altered shoreline), shoa l/sand bar (SB, unvegetated, <1.5 m deep), open bay (>1.5 m deep), and grass bed/shoal-sa ndbar [GB/SB; used in areas <1.5 m deep when bottom composition (vegetated vs. unvege tated) could not be determined or in areas where seagrass cover was very sparse]. Data Analysis Habitat Characterization Descriptive statistics are reported for bottom water temperature, surface and bottom salinity, boat traffic, depth, and seep detection transects. On ly bottom water temperatures in December, January, and February of each wi nter were used to calculate descriptive statistics comparing the three canal syst ems and Matlacha Pass, because the coldest periods primarily occurred during these months. Surface and bottom salinity were infrequently sampled in all three canal syst ems on the same day, therefore descriptive statistics were calculated on ce for all data points and a s econd time using only those data that were collected on days that all three canal systems were sampled. The boat traffic data collected from 0700 to 1800 hours during the 24-hour surveys were combined with those from the daytime only surveys in the analyses of daytime boat traffic. Tidal flushing estimates are presented as the tota l volume of each canal system that is not exchanged with a given tidal cycle. Findings of aquatic vegetation transects within the Matlacha Isles canal system are reported as spec ies present at each spec ific location. Temporal Use of Matlacha Isles Data on time of day, tide height, and water temperature were each compared to counts of manatees traveling in to and out of the Matlacha Isle s canal system to determine
15if these environmental factors influence si te use. A relationship between manatees traveling into and out of the Matlacha Isles canal system and time of day was tested using circular statistics. Time of day is a special type of interval scale, a circular scale, meaning it has equal intervals (hours), an arbi trarily assigned zero point (midnight), and no true designation of high or low values. Hours of day were converted to angular directions (a, in degrees) usi ng the following formula (Zar 1999): (360o)(X) a = 24 where X is the unit of time (hour). Raylei ghs test of circular uniformity (Zar 1999) was applied separately to the count s of manatees traveling in each hour and to those traveling out each hour to test whether either is randomly distributed around the clock. The Watson-Williams test (Zar 1999) was employed to determine if the circular distribution (mean time of day) of counts of manatees en tering was different from those leaving the Matlacha Isles canal system. To determine if manatees choose certain tidal heights when traveling into or out of the Matlacha Isles canal system, data collect ed on manatees traveling in and out of Matlacha Isles at each tide height measuremen t were analyzed. Counts of manatees were continuous, whereas tide heights were record ed only at half-hour intervals when no manatees were observed; theref ore data were condensed into half-hour intervals for each sampling period (counts of manatees were summ ed to give half-hour totals). Data on tidal heights were then grouped into two cat egories, heights recorded when manatees were observed and those when no manatees were observed. Both categories of tidal heights were then each condensed into 7 groups of 12.7-cm (i.e., 5-inch) increments. A chi-square contingency table was used to test whether observations of manatees traveling into/out of Matlacha Isles were in dependent of tidal height. A correlation analysis was used to determ ine if counts of manatees are correlated with water temperature. Bottom water temper ature data from the three sites in Matlacha Pass were averaged and used to calculate av erage water temperatures for the 24-, 72-, and 120-hour periods preceding each sampling date. The one-, three-, and five-day period lengths were chosen arbitrar ily. Total counts of manatees traveling into and out of
16Matlacha Isles during each sampling date could not be summed because of lack of independence (i.e., a manatee coun ted traveling in may also be counted traveling out later in the day or visa versa), therefore only dail y counts of manatees traveling into Matlacha Isles were used in the analyses. Also, some of the sampling periods were conducted consecutively. To prevent violating the assu mption of independent samples, only the first 24-hour sampling period of multi-day surv eys were used in the analysis. Behavioral Observations of Tagged Manatees Behavioral data were used to determine manatees diel distribution of time spent among activities as well as whether water temp erature influenced the percentage of time manatees spent in each activity per day. A ll analyses of behavioral data included only those activities recorded with a confidence level of 1 or 2 (the 2 highest levels of confidence). Only five of the seven activity ca tegories recorded in the field were used in the analyses. Socialize and play were excluded from the analyses because of inadequate sample sizes. The remaining activities were co ndensed into four cate gories: 1) Rest, 2) Feed, 3) Mill, and 4) Travel. Feed and po ssible feed were combined into one activity category, feed, for the analyses. To determine daily activity budgets of manatees using Matlacha Isles, total frequencies of each activity per hour we re calculated for each sampling period, combined, and then graphed. Percentages of each of the four beha vior categories were calculated for each sampling period and then av eraged for each behavior. The number of hours spent in each behavior was calculated by multiplying the number of intervals for each behavior by four (since behaviors were r ecorded at four-minute intervals) and then dividing by 60. To examine the influence of change in water temperature on activity, temperature differences were calculated by subtracting th e average water temperature in Matlacha Pass during the 24or 72-hour period precedi ng the behavioral sampling from the daily average water temperature in Matlacha Pa ss on the day of the behavioral sampling. Percentages of behaviors when water temper ature was colder the day of sampling than the previous 24or 72-hours were compar ed to those when water temperature was
17warmer. Percentages of each activity category were calculated using the total number of intervals with confidence levels of 1 or 2 for each of the sampling periods. T-tests for independent samples were used on behaviors that were normally distributed with equal variances. Mann-Whitney rank sum tests for independent samples were used on those that were not normally distributed and/or lacked equal variance.
18 RESULTS Habitat Characterization Water Temperature Ambient (i.e., Matlacha Pass) water temperat ures within the study area dropped to a daily minimum of 18oC or less 108 days, or 40% of the da ys, over the three winters of the study (Figure 3). The comparison canal syst ems only experienced slightly fewer days than ambient when daily minimum water temperatures were below 18oC, with 100 days in West Island and 92 days in northern Pine Island. The shallow bay outside the mouth of Matlacha Isles (MI site 3) had simila r daily minimum temperatures to the two comparison canal systems. The Matlacha Isles canal system, however, experienced less extreme cold temperatures, with 34 of the to tal days over the three winters having daily minimum water temperatures less than 18oC. The entire Matlacha Isles canal system did not experience daily minimum wa ter temperatures less than 15oC. The examination of daily minimum temper atures showed the low temperature on a given day, but did not give an indication of th e duration of that temp erature (i.e., if there was just one reading that co ld or if it was most of th e day); therefore daily mean temperatures were also calculated for each s ite. Cold fronts that maintained ambient daily mean water temperatures below 20oC occurred 9 times during the three winters and lasted between 4 and 32 days (X = 18.3 days + 11.65 SD). Cold fronts that maintained ambient daily mean water temperatures below 18oC occurred 14 times during the three winters and lasted between 2 and 28 days (X = 6.9 days + 6.91 SD). Matlacha Pass (used in this study as ambien t) had a total of 99 days over the three winters when daily mean water temperatures were less than 18oC, and 96 days when at least one of its sites had daily mean temperatures less than 18oC (Figure 4). West Island and northern Pine Island had 79 and 73 da ys, respectively, of daily mean water
19temperatures less than 18oC, whereas Matlacha Isles only experienced 17 days over the three winters when the entire canal system averaged less than 18oC. More than 62% of the daily mean water temperatures within Ma tlacha Isles were greater than or equal to 20oC and approximately 90% were greater than 18oC. Individual canals within the Matlacha Isles canal system had similar numbers of days when daily mean water temperatures were 18oC or below, with site 6 having onl y slightly fewer than the rest. On days when the daily mean water te mperature in Matlacha Pass (ambient) was less than 18oC, water temperature in the entire Matlacha Isles canal system averaged greater than 2oC warmer than those in northern Pine Island and West Island (Table 1). Water temperatures in Matlacha Isles also averaged more than 3oC warmer than Matlacha Pass during these colder days. The differences between daily mean temper atures in all three canal systems and ambient were calculated in order to assess each canal systems insulating ability. Temperatures in northern Pine Island and West Island were generally similar to ambient, with 83% and 87%, respectiv ely, of their daily mean temperatures within 1oC of ambient (Figure 5). Northern Pine Island never ha d daily mean temperatures more than 1.6oC warmer than ambient and West Island neve r had daily mean temperatures more than 2.4oC warmer than ambient. The small bay ou tside the mouth of Matlacha Isles (MI site 3) demonstrated a similar pattern to norther n Pine Island and West Island, although it had a few more days of the warmer temperatures. Within the Matlacha Isles canal system, da ily mean water temperatures remained warmer than ambient following cold fronts, when ambient water temperatures dropped suddenly (Figure 5). Specific canals within Matlacha Isles varied slightly in their differences from ambient temperatures. Site 1 maintained daily mean water temperatures at least 2oC warmer than ambient 40% of the tota l days over the thr ee winters and ranged up to 6oC warmer than ambient. Site 2 experi enced 43.6% of the total days with daily mean water temperatures at least 2oC warmer than ambient, however the majority of these days were 2-3oC warmer, with the largest difference being 4.8oC. Sites 4, 5, and 6 experienced 48.7%, 59.5%, and 61.8%, respectivel y, of the total days with daily mean water temperatures at least 2oC warmer than ambient and ranged up to 5.6 oC (sites 4 and
205) and 5.8 oC (site 6) warmer. Temperature differences of 3oC or more were typically when ambient daily mean water te mperatures were less than 18oC and as low as 11.7oC. There were 33 days over the three winters when daily mean temperatures in site 1 were up to 0.8oC colder than ambient, however these days were during periods when ambient temperatures were increa sing and already above 20oC. Overall, the Matlacha Isles canal system wa s able to maintain warmer, more stable water temperatures than the two comparison canal systems. Unlike ambient, northern Pine Island, and West Island, Matlacha Isles di d not experience the sudden drops in water temperature following cold fronts, extrem e low temperatures, or long periods of temperatures less than 18oC. Daily mean water temperatures within Matlacha Isles ranged up to 6oC warmer than ambient and were warmer than 18oC over 90% of the days of the three winters of the st udy. Water temperatures in northern Pine Island and West Island, however, were generally similar to ambient, dropping suddenly with passing cold fronts and most temperatures within 1oC of ambient. Salinity Salinity was sampled a total of 21 days in the Matlacha Isles canal system, 15 days in the northern Pine Island canal system, a nd 16 days in the West Island canal system. The Matlacha Isles canal system exhibited slightly lower salinity than the canal systems in northern Pine Island and We st Island. Northern Pine Is land had the highest salinity readings, ranging from 15 to 31.8 ppt, for bot h surface and bottom (Figure 6). Surface salinity in Matlacha Isles aver aged 8.7 and 6.0 ppt lower th an northern Pine Island and West Island, respectively, and bottom sa linity averaged 6.7 an d 4.3 ppt lower than northern Pine Island and West Island, respectiv ely. Seventy-five percent of the surface salinities recorded in Matlacha Isles were below 19 ppt and 75% of its bottom salinities were below 21.5 ppt. In northern Pine Isla nd, however, 75% of its surface and bottom salinities were greater than 19 ppt. Surface and bottom salinities in West Island were greater than 20 and 21 ppt, respective ly, for 75% of the samples. There were 8 days during the combined th ree winters that all three canal systems were sampled on the same day. The data from these days were used to further compare
21salinity among the canal systems (Figure 7). Overall, these censored data followed the same patterns as the complete dataset, with descriptive statistics generally remaining within 1 ppt. A few notable exceptions occurr ed within Matlacha Isles and West Island. The minimum bottom salinity in Matlacha Isles was nearly 6 ppt higher than that of the complete dataset. The minimum surface and bottom salinities in West Island were 8.9 and 8.5 ppt, respectively, higher than those of the complete dataset. Salinity readings recorded at the six salinity sampling stat ions within the Matlacha Isles canal system were compared to each othe r to assess salinity within Matlacha Isles (Figure 8). Salinity at each of the sampli ng stations was generally similar, with all stations having average salinities within 1 to 3 ppt of each other. Station 3 had the highest mean bottom salinity and station 2 had the highest mean surface salinity. All stations had higher bottom salinities than surface salinities, ranging from a mean difference of 0.6 ppt at station 6 (MI site 2) to 4.5 ppt at station 3 (M I site 1). Station 3 was almost always stratified, whereas station 6 was almost never stratified. Surface and bottom salinities in station 4 were nearly e qual for most of the samples, with only 3 exceptions. Stations 1, 2, and 5, varied in their levels of stratification. Overall, Matlacha Isles had slightly lowe r salinity readings than the two comparison canal systems; however, salinities in all three canal systems remained within brackish to estuarine levels. Additionally, Matlacha Isles showed some stratifica tion in certain areas, whereas northern Pine Island and West Island had little to no stratification. Boat Traffic Of the three canal systems, northern Pine Island had the highest mean number of boats traveling through it per day (Table 2). West Island had the fewest boats each day and counts in Matlacha Isles were intermediate. Few to no boats were observed in any of the locations between 0000 and 0500 hours (Figur e 10). Boating activity in all locations generally began just prior to dawn, peaked in Matlacha Pass around 1400 hours, and then rapidly declined around sunset. Northern Pine Island was the only canal system that had boats in use between 2100 and 0600 hours. During the day, northern Pine Island had peaks in counts of boats in the morning and afternoon, whereas Matlacha Isles, West
22Island, and Matlacha Pass appeared to slowly increas e in the number of boats throughout the day and decrease at the end of the day (Fi gures 9 and 10). All of the boats that leave Matlacha Isles and West Island travel into Ma tlacha Pass. Boats that leave northern Pine Island travel into Bokeelia, Charlotte Harbor, Pine Island Sound, or Matlacha Pass. Manatees encountered boats as they were entering and leaving Matlacha Isles. These encounters were observed repeatedly during the 24-hour surveys from the entrance to Matlacha Isles, as well as during focal observations. The entrance/exit canal is approximately 1.5 m deep at high tide and 810 m wide for 40-50 m. A boat ramp is located about 30 m inside the entrance/exit canal. Manatees entering the canal system when boats were exiting were observed turning around after st arting to enter and traveling to the far side of either an oyster bar or shoal beyond th e channel, where they remained until the boat(s) passed during 12 different occurrences affecting 34 individual manatees. One of the tagged manatees was observed waiting outside Matlacha Isles for over two hours one morning while boats con tinued to pass by. Manatees entering the canal system when boats were also enteri ng increased their swimming speed or turned away from the entrance canal at the last mi nute and waited for the boat(s) to pass during 2 different occasions affecting 5 indivi dual manatees. The responses of manatees exiting the canal when boats were either en tering or leaving, howeve r, were not observed because the opposite end of the canal was beyond view. Canal Depths The Matlacha Isles canal system averages ove r 0.3 meters deeper than northern Pine Island and West Island (Table 3). The shallo west areas in all three canal systems tended to be at the edges/shorelines and terminal ends of canals, whereas the deepest areas tended to be along the center of canals. The West Island canal system had the most uniform depths, with the greatest difference being 0.5 meters. The entire West Island canal system and 91% of the northern Pine Is land canal system were less than 2.0 m in depth. The Matlacha Isles canal system had the grea test variation in de pths of the three canal systems; however the average depth in ea ch of its canals was still greater than those
23of northern Pine Island and We st Island. Although the mean depth in Matlacha Isles was only 0.3 m deeper than the other two systems, 46% of Matlacha Isles was 2 m or greater in depth. Site 1 of Matlacha Isles was the d eepest canal in the system and depths along the center of this canal ranged from 2.3 to 5.1 m. The mean depths of the other four canals were within 0.3 m of each other. Site 2 was the most uniform of the high use and little/no use canals in Matlacha Isles. Sites 4 and 5 (little/no use canals) had the least uniformity in depth. Tidal Flushing The total volume that is not exchanged dur ing a tidal cycle was calculated for each of the three canal systems. Matlacha Isles ha d the lowest estimated tidal flushing, as the volume of water that remains within the can al system is 4 times greater than that remaining in northern Pine Island and nearly 14 times greater than that remaining in West Island with a given tida l cycle (Figure 11). Ground-water Seeps and Submerged Aquatic Vegetation No notable differences in salinity or temper ature readings were detected within any of the three canals in Matlach a Isles that are regularly us ed by manatees. All four variables that were measured (i.e., surface sa linity, bottom salinity, surface temperature, and bottom temperature) remained fairly uni form within each site (Table 3). Any differences greater than 2 ppt or oC generally occurred at opposite ends of the canals. No bottom salinity readings less th an 20 ppt were recorded in any of the canals during the transects. No submerged aquatic vegetation was found w ithin the Matlacha Isles canal system. The alga Cladophora and very sparse patches of sea grass, Halodule wrightii, were found in the small bay just outside the mouth of Matlacha Isles (MI site 3). Temporal Use of Matlacha Isles A total of twenty-two 24-hour surveys of manatees traveling into and out of the Matlacha Isles canal system was conducte d during January through March of 2000 and
24January through February of 2001 and 2002. Ma natees followed a distinctive daily pattern of traveling into and out of Matlacha Isles that wa s significantly different from being uniformly dispersed around the clock. Co unts of manatees traveling into Matlacha Isles were concentrated in the morning hours of the day (Rayleighs z = 651.95, N = 1321, p < 0.001), with the majority of them entering the canal system between 0600 and 1000 hours (Figure 12). Counts of manatees traveling out of Matlacha Isles were concentrated in the late afternoon/eveni ng hours of the day (Rayleighs z = 738.12, N = 1336, p < 0.001), with the majority leaving between 1600 and 1900 hours. Mean times were also calculated for both c ounts of manatees traveling in and those traveling out. The mean time of day that manatees were entering Matlacha Isles (0725 hours) differed significantly from the mean tim e of day that manatees were leaving Matlacha Isles (1750 hours; Watson-Williams F = 6437.73, df = 1, df2 = 2655, p < 0.00001). Manatees did not choose certain tidal hei ghts when traveling into or out of the Matlacha Isles canal system. Tidal heights, ranging from 25 cm (the highest tide) to 114 cm (the lowest tide) below a reference point on the survey dock, were recorded a total of 1,056 half-hour intervals during the 22 sampli ng periods. The proportions of half-hour intervals with and without manatees observed did not vary significantl y with tidal height (x 2 = 6.14, df = 6, p = 0.41). Additionally, frequencies of available tid al heights were graphed against tidal heights when manatees were entering (Fi gure 13A) and leaving (Figure 13B) Matlacha Isles. A range of heights, which included high and low tides, was observed in both periods of high manatee movement. The fre quencies of the tidal heights that were available versus those when manatees were observed are nearly identical for both periods and show that manatees used all tidal hei ghts when traveling both into and out of Matlacha Isles. Scatter plots of counts of manatees trav eling into Matlacha Is les against average water temperatures during the 24-, 72-, a nd 120-hours prior to sampling revealed a difference in the relationship between counts and water temperatures when temperatures
25were below and above 18oC (Figure 14). When water temperatures were below 18oC, counts were extremely low, w ith little variation. When wa ter temperatures were above 18oC, counts were higher, with much more variation. The data points above 18oC also appear to have a negative correlation with 24-, 72-, and 120hour average water temperatures. Results of simple linear regr essions (Table 5) run on the data points above 18oC showed that there was a significant re lationship between ma natee counts and 72and 120-hour water temperature (p = 0.016 and 0.007, respectively). Behavioral Observations of Tagged Manatees Behavioral data were co llected over a total of 7,268 intervals during 27 focal observations of 10 individual manatees (Table 6). Seven of these individuals were sampled more than once. A to tal of 6,327 intervals had behavior al data recorded with the two highest levels of confidence (1 and 2), wh ich were used in the analyses. Eighty-one percent of the sampling periods were 24-hour s in duration, and only 2 of the sampling periods were less than 20 hours in durati on. Three additional sampling periods were initiated, but were not used because of their short durations (each wa s less than 8 hours in length). When behaviors were examined over all sa mpling periods, manatees spent most of their time resting and feeding, both in te rms of percentage of total number of observations recorded and number of hours (T able 7). Resting was the primary activity within Matlacha Isles (7 6.3% of recorded behaviors within Matlacha Isles). Observations of rest were highest between 0600 and 1800 hour s (Figure 15a), whereas those of feeding were highest from 1900 to 1200 hours, with only a slight decrease between 1300 and 1800 hours (Figure 15b). Observations of milling peaked during the afternoon hours (Figure 15c) and frequencies of observed traveling were high est between 1600 and 1800 hours (Figure 15d). During 11 of the sampling periods, the focal manatees demonstrated the distinctive diel pattern of using Matlach a Isles during the da y, traveling into Matlacha Pass during the late afternoon/early eveni ng, and then returning to Matl acha Isles in the morning. The average time that the focal manatees entered Matlacha Isles was 0814 hours and the
26average time they left was 1709 hours, which corresponded with the peak travel periods demonstrated in Figure 12. Histograms of each behavior per hour during these 11 sampling periods further demonstrated the diel pattern (Figure 16). Manatees primarily rested during the day (Figur e 16a) and feeding occurred pr imarily during the night and early morning hours (Figure 16b). Of the habitats recorded between 2000 and 0600, 65100% were grass beds. Little to no feed ing was recorded between noon and 1800 hours. Daily average water temperatures in Matl acha Pass during these sampling periods were between 18.3 and 24.6oC (mean = 20.7oC). During four of the sampling periods (1-2, 4, and 29), the focal manatees never left Matlacha Isles. They primarily rested (mean = 84% of observed behaviors), milled or traveled within the canal system during th ese sampling periods. Daily average water temperatures in Matlacha Pass during these four sampling periods were between 15.5 and 17.7oC. During eight of the sampling periods (911, 14, 16, 23-25), the focal manatees were in Matlacha Pass the entire time. Histograms of each behavior per hour revealed that a diel behavioral pattern also existed during these 8 sampling periods (Figure 17). Focal manatees primarily fed (mean = 49% of obs erved behaviors) throughout these sampling periods, with an increase in feeding fr equency between 0100 and 1200 hours (Figure 17b). The majority of resting occurred during the day and dropped off after 2100 hours (Figure 17a). Daily average water temper atures in Matlacha Pass during these eight sampling periods were >20oC (five of these were >23oC) and had remained >20oC for an average of 9 days prior. Percentages of the sampling periods that each manatee engaged in each of four behavior categories were also used to examine whether ch ange in water temperature influenced activity patterns. Average daily water temperatures in Matlacha Pass during the behavioral sampling peri ods ranged from 15.5 to 24.9oC. The focal manatees spent a significantly higher percentage of time resting when water temperatures were colder during the sampling period than the previ ous 24(Mann-Whitney rank sum test, p = 0.0328; Table 9) and 72-hours (t-test, p = 0.0017; Table 8). Except for a single outlying
27data point (Figure 18a), manatees rested between 21 and 95% of the sampling period when water temperatures were colder duri ng the survey than th e previous 24-hours. Additionally, manatees rested 25.7 95% of the sampling period when water temperatures were colder during the survey than the previous 72-hours (Figure 19a). Conversely, manatees spent a significantly highe r percentage of time feeding when water temperatures were warmer during the sampli ng period than the previous 24or 72-hours (t-tests, p = 0.0300 and 0.0148, respectively; Tabl e 8). When water temperatures were warmer during the sampling period than the pr evious 24or 72-hours, manatees spent 13.6 73.5% and 23.9 73.5%, respectively, of th eir time feeding [except for 1 outlying data point in the 24-hour comparison (Figure 18b) and 2 outlying data points in the 72hour comparison (Figure 19b)]. Manatees also spent more time milling when water temperatures were warmer during the sampli ng period than the 24and 72-hours prior (ttests, p = 0.0099 and 0.0001, respectively; Table 8). The percentages of time manatees spent traveling did not vary significantly between relatively warmer and colder days (Tables 8 and 9).
28 DISCUSSION The Florida Manatee Recovery Plan (U.S. Fish and Wildlife Service 2001) advocates the collection of data on the attributes of non-indus trial warm-water refuges in order to better understand why these sites are at tractive to manatees, as well as what role they may play in the survival of the species if warm water at nearby refuges is no longer available. This study addressed the need fo r data on these winter refuges by providing (1) information on several habitat features that appear to be important to Florida manatees, which suggest why manatees use so me, but not other, seemingly similar sites during the winter, (2) insight into when a nd how certain winter refuges are used by manatees, and (3) criteria for managers to use when selecting and potentially altering/enhancing additional alternative winter refuges, in the event that the creation of such sites becomes necessary. Why Do Manatees Use Matlacha Isles Rath er Than Other Seemingly Similar Sites During The Winter? Organisms exhibit preferential use of habita ts that contain features necessary for survival, such as access to food and shelter, whether the shelter is from predators, environmental conditions, or both (aquatic invertebrates: Gallien 1985; Moran 1985; Williams and Morritt 1985; amphibians and reptiles: Downes and Shine 1998; Schlesinger and Shine 1994; Lecis and Norris 2003; birds: Holmes and Robinson 1981; Martin 1995; Yanes et al. 1996; Sachot et al 2003; Gavashelishvili 2004; and mammals: Craighead et al. 1973; Collins et al. 1978; Cowlishaw 1997; Milner and Harris 1999; Alvarez-Cardenas et al. 2001; Lyons et al. 2003; Marin et al 2003; Tweheyo et al. 2004). Habitat features that are frequently noted as being important to manatees include access to warm water in winter, availability of s eagrasses or other vegetation on which to forage, freshwater to drink, and ab sence of human disturbance (Hartman 1979; Irvine 1983;
29OShea 1988; Lefebvre et al. 1989; Reynolds 1999; U.S. Fish and Wildlife Service 2001). When these habitat features were examin ed at a winter refuge (the Matlacha Isles canal system) and compared to those of tw o similar appearing, but rarely used sites nearby, results indicated that water temperature had the most influence on site selection. No submerged aquatic vegetation was found with in Matlacha Isles and all three sites are a short distance from dense sea grass beds (e ach is less than 1.5 km away); therefore access to foraging sites did not play a role in site selection in this study. Although slight differences in salinity did occur between Matl acha Isles, northern Pine Island, and West Island, these differences were not great and therefore did no t show any obvious influence on site selection. Human disturbance (measured as levels of boat tr affic) also differed only slightly between the three canal system s, with Matlacha Isles having fewer boats than the northern Pine Island canal system but more than the West Island canal system. Thus, it appears that human disturbance di d not influence manatees selection of Matlacha Isles over the two comparison canal systems. Environmental temperature is a substant ial determinant of the distribution of organisms because of its effect on bi ological processes and the imprecise thermoregulatory capabilities of most orga nisms. Although endotherms are able to regulate their body temperatur e through metabolic heat production, insulation, and vascular adjustments, each species functi ons best within certain environmental temperature ranges (Campbell and Reece 2002). Manatees possess phys iological traits (low basal metabolism, low-energy food sour ce, high rate of thermal conductance, and limited capacity for thermogenesis) that limit their thermal tolerance and cause them to be vulnerable to cold (Irvine 1983). Exte nded exposure to temp eratures below 15-20oC causes manatee cold stress syndrome, a co mplex multifactorial disease process that involves a compromise to metabolic, nutri tional, and immunologic homeostasis, and culminates in secondary opportunistic and id iopathic diseases which can be lethal (Buergelt et al. 1984; OShea et al 1985; Bossart et al. 2002). Ambient water temperatures with in the study area ranged from 10.0oC to 26.8oC during the three winters of the study (Decem ber February 1999-2002). Ambient daily average water temperatures dropped below 18oC a total of 99 days, with some of these
30periods lasting up to 28 days. It is therefore critical that manatees in this region find warm water during cold periods in order to survive (Campbell a nd Irvine 1981; Irvine 1983; Bossart et al. 2002). The Matlacha Isles canal system was able to maintain warmer water temperatures than nearby monitored locations during co ld periods, with differences of 2-6oC warmer than ambient (i.e., Matlacha Pass), and the lowest temperature within the site only dropping to 14.7oC. Approximately 90% of the days during the 3 winters of the study had daily average water temp eratures greater than 18oC within Matlacha Isles. Extended periods of extreme cold did, however, drop water temperatures within Matlacha Isles below 18oC during 4 periods, which lasted betwee n 2 and 7 days. Thus, Matlacha Isles has warmer, more stable water temperatures than ambient and maintains temperatures within or above manatees thermal threshold of 18-20oC (Irvine 1983) throughout most of the winter. During extended periods of extr eme cold, however, the canal system appears unable to maintain sufficient warmth and dr ops below manatees thermal tolerance. Matlacha Isles is therefore an inadequate refuge during the coldest periods. Water temperatures in the comparison canal systems in northern Pine Island and West Island, however, were very similar to each other, did not maintain warmer water temperatures following cold fronts, and generally remained within 1oC of ambient. The canal systems in northern Pine Island and We st Island, therefore, do not provide adequate thermal shelter from even moderately cold ambient temperatures. Results suggest that the difference in wa ter temperatures between Matlacha Isles and the comparison canal systems may be a result of the deeper canals and lower tidal flushing in Matlacha Isles. Ground-water seep s, however, were not found and therefore do not contribute to the warmer temperatures within Matlacha Isles. Nearly half of the Matlacha Isles canal system is deeper than the deepest areas of either comparison canal system. Deeper canals have smaller surface ar ea:volume ratios, which limit the effect of decreases in air temperature on water temper ature. Additionally, the decreased tidal flushing within Matlacha Isles allows a much gr eater volume of water to remain within
31the canal system and not be mixed with colder ambient water. As stated above, however, these features are not able to retain suffici ent warmth during extended periods of extreme cold. The model used in this study to estimate tidal flushing only estimated the volume of water that could potentially be exchanged with the receiv ing water body (i.e., Matlacha Pass) during a given tidal cycle in an id ealized circumstance (Monsen et al. 2002; Sanford et al. 1992). It did not, however, give any information about age (the amount of time water remains within each canal system; Zimmerman 1988), residence time (the amount of time it takes to effectively flush each canal system), or the return flow factor (the fraction of water leaving during ebb th at returns during flood; Monsen et al. 2002; Sanford et al. 1992). These factors are ofte n difficult to measure (Monsen et al. 2002; Sanford et al. 1992) and require the use of tracer dyes. A hydrologic study of this kind, however, was beyond the scope of the current study. Water temperatures alone seem unable to explain fully the differential distribution of manatees within the Matlacha Isles canal sy stem itself, as individual canals within Matlacha Isles experienced very similar water temperatures to each other. Manatees use Sites 1 and 2 consistently and in high numbers. Site 6 is also used regularly by manatees, but not by as many. Water temperatures in Site 1 were much more stable than those of the other canals within Matlacha Isles, likely a result of the greater water depths in this site. When ambient water temperatures dr opped suddenly and dramatically, temperatures in Site 1 remained warm and took much l onger to decrease (1-4 days longer) than the other canals in Matlacha Isles, which averag e 2.8-3.1 m shallower th an the deepest areas of Site1. Conversely, when ambient water temp eratures, as well as those in the rest of Matlacha Isles, increased quickly following co ld fronts, temperature increases in Site 1 generally lagged by 1-2 days. Water temperatur es in the other two canals used regularly by manatees (Sites 2 and 6), however, increase d and decreased at si milar rates to the canals that are not used by ma natees (Sites 4 and 5). Although Florida manatees can tolerate a wide range of salinities (Ortiz et al. 1998), they are attracted to sour ces of freshwater (Hartman 1979; Belitsky and Belitsky 1980; Powell et al. 1981; Powell and Rathbun 1984; U. S. Fish and Wildlife Service 2001). No
32fresh ground-water seeps or any other freshw ater sources were detected within the Matlacha Isles canal system that might serve as any obvious attractant of manatees to this site. Differences in salinity and levels of stratification among sampling stations within the Matlacha Isles canal system may be explained by their proximities to potential freshwater/nearly freshwater sources, water depths, and exposure to wind energy. For example, two of the sampling stations are located near breaks in the spreader canal system (part of the north Cape Coral canal sy stem to the north of Matlacha Isles). These two breaks provide potential sources of freshwater/nea rly freshwater; however no manatees were observed congregated at thes e sites, and average salinities remained within 1-3 ppt of the rest of Matlacha Isles. Overall, Matlacha Isles had the lowest salinity readings among the three canal systems; however the differences between Matlacha Isles and the comparison canal systems were not great, all r eadings were within brackish to estuarine levels, with mean differences less than 9 ppt, and ranging in difference between 1 and 18 ppt for surface salinity and and 11 ppt for bottom salinity. It seems unlikely, therefore, that salinity had a strong influence, if any, on manatee se lection of Matlacha Isles over the other two canal systems during the winter. Human disturbance is another important factor potentially affecting manatees habitat selection. This study sampled hu man disturbance by recording numbers of motorboats observed at each location. Boat tr affic in Matlacha Isles was lower than in Matlacha Pass and northern Pine Island; however boat traffic in West Island was less than that in Matlacha Isles. Mean counts of boa ts per day did not differ greatly between the three canal systems, so it seems unlikely that levels of boat traffic had any influence on manatee selection of Matlacha Isles over the other two canal systems during the winter. This study, however, only examined levels of human disturbance at the entrance of the canal systems, not within the canal sy stems themselves. Buckingham et al. (1999) and King and Heinen (2004) found that manat ees near the springs of the Crystal and Homosassa Rivers increased their use of sa nctuaries that are off-limits to boats and swimmers as the number of boaters and swimme rs in the vicinity increased. Future habitat studies of non-industrial warm-water refuges should address human disturbance
33within refuges to assess whether manatees e ndure encounters in order to gain access to a site, but may be forced to use suboptimal regions of the site in orde r to avoid disturbance once they are inside. This may be an im portant conservation concern because studies have indicated that use of suboptimal habita t can affect reproductive rates, predation intensity, growth rates, and even survival of species (S ullivan 1979; Adolf and Porter 1993; Muller et al. 1997; Loeb 1999; Krijgsveld et al. 2003). The extent to which species respond beha viorally to a disturbance, however, may depend on their perception of the severity of th e threat (Frid and Dill 2002; Cassini et al. 2004) and the costs of the res ponse (West et al. 2002; Sti llman and Goss-Custard 2002; Beale and Monaghan 2004). The use of beha vioral responsiveness as an index of disturbance effects may be inappropriate, because animals might make state-dependent decisions about responding to disturbances (Gill et al. 2001; Beal e and Monaghan 2004). Two studies that examined cost s of behavioral responses in shorebirds showed that the level of response depends on the state of the individual as well as the state of the current habitat, regardless of the level of distur bance (Stillman and Goss-Custard 2002; Beale and Monaghan 2004). Birds in good condition an d/or in a richer habitat responded more to human disturbance than birds in poorer c ondition and/or habitat, presumably because the better condition/richer habitat individuals had more response options open to them and less to lose fitness-wise by responding. In contrast, individuals in poorer condition and/or habitat may be more constrained by their current requirements, which may force them to endure certain levels of disturbance in order to survive (Gill et al. 2001; Beale and Monaghan 2004). If this is true for Flor ida manatees (for whic h constraints include the need for warm-water and adequate forage during the winter, and limited sites that provide one or both of these) individuals may tolerate some human disturbance at a winter refuge if that site provides the warm water and access to foraging sites that are necessary for their survival.
34When and How is Matlacha Isles Used By Manatees? Manatees exhibited a distinc tive daily pattern in their use of Matlacha Isles during the winter. Manatees entered the canal sy stem during the morning (primarily between 0600 and 1000 hours), where they rested througho ut the day, and left during the late afternoon/evening (primarily between 1600 and 1900 hours) to travel to Matlacha Pass to feed. Bengtson (1981) observed a similar di el pattern at Blue Spring Run, off the St. Johns River, FL. Daily patterns of site use in manatees have been attributed to tidal cycles at some locations either because indivi duals cannot cross a shallow barri er to enter/exit a site at lower tidal levels or specific feeding areas are only accessible during higher tidal levels (Hartman 1979; Zoodsma 1991). Tides, however, did not explain daily site use patterns at Matlacha Isles. Although some of the lowest ti dal heights did occur during the morning and some of the highest occurred dur ing the afternoon/evening hours, both high and low tides took place during the two peak periods when manatees traveled into and out of Matlacha Isles, and anal yses showed that there was no tidal range that was used disproportionately to its availability. Ma natees were even observed fighting strong outgoing tidal currents at the mouth of the can al system and pulling themselves across the shoal at the mouth of the canal system during extreme low tides in order to enter in the morning and leave in the afternoon/evening. Another possible explanation fo r the diel pattern of site use is avoidance of the high levels of boat traffic in Matlacha Pass duri ng the day. The two peak periods when manatees enter and leave the Matlacha Isle s canal system do not correspond to peak periods of boat use, but boat counts are still much higher during daylight than nondaylight hours. Manatees entering and leav ing Matlacha Isles encountered boats during the periods they traveled to and from Matl acha Isles, and many were observed waiting for boats to pass or quickly swimming out of th e way. It appears, th erefore, that although human disturbance (boat traffic) may have some effect on manatee behavior and habitat use, it is not sufficient to deter them from us ing Matlacha Isles. As stated above, the
35physiological constraints of needing warm water and access to forage may outweigh the perceived costs of enduring certain levels of disturbance (Gill et al. 2001; Beale and Monaghan 2004). If, however, avoiding boat traffic played a strong role in determining the temporal use of this winter refuge, it seems plausibl e that manatees would wait until dusk or later to leave Matlacha Isles and return prior to or at dusk. Traveling during these time periods would allow them to travel and feed at ni ght with few to no encounters with boats as opposed to frequently crossing paths with them. Bengtson (1981) suggested that manatees exhi bit this diel pattern in their use of certain warm-water refuges to take advantag e of daily fluctuations in ambient water temperatures (i.e., travel to feeding areas wh en ambient water temperatures are warmest). Ambient water temperatures during this study peaked late in the day; however water temperatures quickly dropped after sunset and were already increasing before most manatees returned to Matlacha Isles. It s eems more likely that they would travel to feeding sites before water temperatures had p eaked and return to Matlacha Isles prior to the coldest period of the day if they were indeed taking advantage of the warmest water temperatures in order to feed. Measuremen ts of manatee body temperatures at different water temperatures suggest that manatees can temporarily maintain core temperature during movements from warm-water refuges in to colder adjacent waters (Irvine 1983). Manatees have been observed leaving winter refuges to feed in areas where ambient temperatures are < 16oC (Hartman 1979; Powell and Waldron 1981; Shane 1981). Irvine (1983) suggested that these fora ging trips into cold water mi ght be possible if they can later digest in warmer water. Animals can obtain heat from their envi ronment and help regulate their body temperature behaviorally by adjusting habitat selection, body posture, and timing of activity. Some species may rest during the day in warmer sites, allowing them to maintain a higher body temperature at minimal co st, and forage at ni ght, generating heat through muscular activity and digestion. Pregnant female solitary bats ( Myotis evotis ) forage at night, therefore roost site selecti on is based on daytime temperature, even if the site cools rapidly at night (Chr uszcz and Barclay 2002). When th ey return to the roosts at
36dawn they are warmed passively as the roosts warm during the day, thus reducing energetic costs. Trumpeter sw ans and many other species of wa terfowl also feed at night (Jorde and Owen 1988). Squires and Anders on (1997) suggested that because swans wintering in the greater Yellows tone area do not appear to ha ve factors that might force nighttime feeding (such as human disturbance, predator avoidance, or hunting), they may forage at night to conserve energy. If wa terfowl are able to generate heat during nocturnal foraging bouts through muscular activ ity and digestion and use solar radiation to help thermoregulate while they rest dur ing the day, the total thermoregulatory costs would be reduced (Jorde et al. 1984; Squi res and Anderson 1997). If manatees also thermoregulate in this manner, then it may explain the diel pattern observed at Matlacha Isles. Manatees may have been taking advantag e of the warmth within Matlacha Isles during the day in order to passively ther moregulate during periods of inactivity. Manatees within Matlacha Isles typically bottom rested during the morning (when bottom temperatures were warmer than those at the surface) and surface rested or even basked with their backs out of the water during mid-day to late afternoon. By traveling to Matlacha Pass at night to feed (even though water temper atures may have been cooler) manatees may sustain body temperatures so mewhat through the heat generated during muscular activity and digesti on. Conversely, the thermoregulat ory cost of resting within Matlacha Isles at night (when temperatures are cooler) would be much higher as the manatees would have to expend much more energy in order to thermoregulate in the cooler water while inactive. This activity pa ttern may continue to a lesser extent as water temperatures continue to warm above 20oC (as shown by the decrease in the number of manatees entering and exiting Matlacha Isles and the 8 sampling periods when the focal manatees remained outside of Matlacha Isles). As temperatures increa se further, the diel activity pattern may be abandoned, as Bengts on (1981) found that Blue Spring manatees no longer exhibited a strong diel cycle outside of the winter season. Water temperature also influenced the number of manatees that traveled into and out of Matlacha Isles among days. A clear distinc tion existed in the re lationship between the number of manatees traveling into Matlach a Isles and ambient water temperature above
37and below 18oC. Few to no manatees were obser ved when water temperatures were below 18oC and there was an inverse relationshi p between manatee counts and water temperature above 18oC. The mean ambient temperature of the three-day period prior to sampling appears to be the best predictor of the number of manatees that enter the canal system. As ambient temperatures continued to increase above their thermal threshold, the number of manatees observed decreased, like ly due to a higher proportion remaining in the grass beds feeding when they were not as thermally dependent on the refuge. This change in site use with warmer water temper atures was also confir med by the behavioral data, as individuals that were observed during 8 sa mpling periods when water temperatures were >20oC remained within Matlacha Pass (primarily feeding) and overall, manatees spent a significantly higher percen tage of time feeding when ambient water temperatures were warmer. Conservation and Management Implications Water temperature appears to be the primar y attractant of manatees to the Matlacha Isles canal system during the winter. A lthough the Matlacha Isles canal system has no direct thermal input, it mainta ined water temperatures within or above manatees thermal threshold of 18-20oC (Irvine 1983) approximately 90% of the three winters of the study, making it an adequate winter refuge for most of the winter. Because of its inability to maintain water temperatures high enough dur ing prolonged periods of extremely cold weather to provide sufficient warmth for manatees, however, Matlacha Isles is not a reliable winter refuge. During these times, most of the manatees leave Matlacha Isles and travel the approximately 50 km to the Flor ida Power & Light plant on the Orange River (Koelsch and Barton 1999). Between these ex tremely cold days however, Matlacha Isles provides a warm-water refuge that is close to abundant sea grass beds, as opposed to the limited forage at and near the power plant effluent. Findings from this study may allow res ource managers and conservationists to develop a more informed approach to creating sanctuaries or refuges. Sites that provide access to warm water and adequate forage shoul d be the most important habitat features to provide. Additionally, little to no disturbance (especially hu man) would be ideal, even
38though manatees at Matlacha Isles appear to tolerate certain le vels in order to have access to the warmer water and proximity to sea gr ass beds. If the creation of new winter refuges becomes necessary, similar sites to Matlacha Isles may need to be modified and/or enhanced with environmentally se nsitive, non-industry-de pendent methods of maintaining temperature or heating water during extended periods of extreme cold. These sites should also be deep with limited tidal flushing to retain the warm water.
39 LITERATURE CITED Ackerman, B.B., S.D. Wright, R.K. Bonde, D.K. Odell, and D.J. Banowetz. 1995. Trends and patterns in mortality of ma natees in Florida, 1974-1992. Pages 223-258 in T.J. OShea, B.B. Ackerman, and H.F. Pe rcival (eds.). Population Biology of the Florida Manatee. U.S. Fish and Wildlif e Service, National Biological Service, Information and Technology Report 1. Adolf, S.C. and W. P. Porter. 1993. Temper ature, activity, and lizar d life histories. American Naturalist 142:273-295. Altmann, J. 1974. Observational study of behaviour: sampling methods. Behaviour 49(3-4):227-267. Alvarez-Cardenas, S., I. Guerrero-Cardenas, S. Diaz, P. Galina-Tessaro, and G. Gallina. 2001. The variables of physical habitat selection by the desert bighorn sheep ( Ovis canadensis weensi ) in the Sierra del Mechudo, Baja California Sur, Mexico. Journal of Arid Environments 49:357-374. Beale, C.M. and P. Monaghan. 2004. Behavior al responses to huma n disturbance: a matter of choice? Animal Behaviour 68:1065-1069. Belitsky, D.W. and C.L. Belitsky. 1980. Distribution and abundance of manatees Trichechus manatus in the Dominican Republic. Biological Conservation 17:313319. Bengtson, J.L. 1981. Ecology of manatees ( Trichechus manatus ) in the St. Johns River, Florida. Ph.D. Dissertation, Univer sity of Minnesota, Minneapolis, MN. Bengtson, J.L. 1983. Estimating food consump tion of free-ranging manatees in Florida. Journal of Wildlife Ma nagement 47(4):1186-1192. Bossart, G.D., R.A. Meisner, S.A. Romm el, S. Ghim, and A. Bennet Jenson. 2002. Pathological features of the Florida ma natee cold stress syndrome. Aquatic Mammals 29:9-17. Buckingham, C.A., L.W. Lefebvre, J.M. Sch aefer, and H.I. Kochman. 1999. Manatee response to boating activity in a therma l refuge. Wildlife Society Bulletin 27(2):514-522.
40 Buergelt, C.D., R.K. Bonde, C.A. Beck, and T.J. OShea. 1984. Pathologic findings in manatees in Florida. Journal of the American Veterinary Medical Association 185:1331-1334. Campbell, H.W. and A.B. Irvine. 1981. Ma natee mortality during the unusually cold winter of 1976-1977. Pages 86-91 in R.L. Brownell, Jr., and K. Ralls (eds). The West Indian manatee in Florida. Pro ceedings of a workshop held in Orlando, FL, 27-29 March 1978. Florida Department of Na tural Resources, Tallahassee. 154 pp. Campbell, N.A. and J.B. Reece. 2002. Biology, 6th Edition. Benjamin Cummings, San Francisco, CA. Cassini, M.H., D. Szteren, and E. Fernande z-Juricic. 2004. Fence effects on the behavioural responses of South American fur seals to tourist appr oaches. Journal of Ethology 22:127-133. Chruszcz, B.J. and R.M.R. Barclay. 2002. Thermoregulatory ecology of a solitary bat, Myotis evotis roosting in rock crevices Functional Ecology 16:18-26. Collins, W.B., P.J. Urness, and D.D. Austin. 1978. Elk diets and activities on different lodgepole pine habitat segments. Journal of Wildlife Management 42:799-810. Cowlishaw, G. 1997. Trade-offs between fora ging and predation risk determine habitat use in a desert baboon populati on. Animal Behaviour 53:667-686. Craighead, J.J., F.C. Craighead, R.L. Ruff, and B.W. OGara. 1973. Home ranges and activity patterns of nonmigrato ry elk of the Madison Drainage herd as determined by biotelemetry. Wildlife Monographs 33:1-50. Deutsch, C.J., J.P. Reid, R.K. Bonde, D.E. Easton, H.I. Kochman, and T.J. OShea. 2003. Seasonal movements, migratory behavi or, and site fidelity of West Indian manatees along the Atlantic coast of th e United States. Wildlife Monographs 151:177. Domning, D.P. and L.C. Hayek. 1986. Inters pecific and intraspecific morphological variation in manatees (Sir enia: Trichechus). Marine Mammal Science 2(2):87-144. Downes, S. and R. Shine. 1998. Heat, safe ty or solitude? Using habitat selection experiments to identify a lizards prio rities. Animal Behaviour 55:1387-1396. Dyer, K.R. 1973. Esturaries: A physical introduction. 2nd Ed. John Wiley.
41Etheridge, K., G.B. Rathbun, J.A. Powell, a nd H.I. Kochman. 1985. Consumption of aquatic plants by the West Indian manatee. Journal of Aquatic Plant Management 23:21-25. Florida Department of Environmental Protecti on, Florida Marine Rese arch Institute. 1998. Atlas of Marine Resources, Version 1.2. Florida Fish and Wildlife Conservation Commission. 2005a. FWC Releases 2004 Boating Statistics. Retrieved from the World Wide Web August 2005: http://myfwc.com/whatsnew/05/st atewide/boatingstats.html Florida Fish and Wildlife Conservation Commission. 2005b. Manatee Mortality Statistics. Retrieved from th e World Wide Web August 2005: http://research.myfwc.com/feat ures/category_sub.asp?id=2241 Frid, A. and L. Dill. 2002. Human-caused disturbance stimuli as a form of predation risk. Conservation Ecology 6:11-26. Gallien, W.M. 1985. The effects of aggregations on water loss in Colisella digitalis Veliger 28:14-17. Gavashelishvili, A. 2004. Habitat selection by East Caucasion tur ( Capra cylindricornis ). Biological Conservation 120:391-398. Gill, J.A., K. Norris, and W.J. Southerland. 2001. Why behavioural responses may not reflect the population conse quences of human disturbanc e. Biological Conservation 97:265-268. Hartman, D.S. 1979. Ecology and behavior of the manatee ( Trichechus manatus ) in Florida .American Society of Mammalogi sts, Special Publication No. 5. 153 pp. Hill, D.A. and J.E. Reynolds, III. 1989. Gr oss and microscopic anat omy of the kidney of the West Indian manatee, Trichechus manatus (Mammalia: Sirenia). Acta Anatomica 135:53-56. Holmes, R.T. and S.K. Robinson. 1981. Tree sp ecies preferences of insectivorous birds in a northern hardwoods forest. Oecologia 48:31-35. Irvine, A.B. 1983. Manatee metabolism and its influence on distribution in Florida. Biological Conservation 25(4):315-334. Jorde, D.G., G.L. Krapu, R.D. Crawford, and M.A. Hay. 1984. Effects of weather on habitat selection and behavior of malla rds wintering in Nebraska. Condor 86:258265.
42Jorde, D.G. and R.B. Owen. 1988. The need for nocturnal activity and energy budgets of waterfowl. Pages 169-180 in M.W. Weller (ed.). Waterfowl in winter. University of Minnesota Press, Minneapolis, MN. King, J.M. and J.T.Heinen. 2004. An assessm ent of the behaviors of overwintering manatees as influenced by interactions with tourists at two sites in central Florida. Biological Conservation 117:227-234. Koelsch, J.K. 1997. The seasonal occurr ence and ecology of Florida manatees ( Trichechus manatus latirostris ) in coastal waters near Sarasota, Florida. M.S. Thesis. University of South Florida, Tampa, FL. 121 pp. Koelsch, J.K. and S.L. Barton. 1999. Photoidentification and beha vioral studies of Florida manatees in Sarasota Bay and southw est Florida. Report to Florida Fish and Wildlife Conservation Commission, St. Peters burg, FL. Mote Marine Laboratory Technical Report No. 659. 35 pp. Krijgsveld, K.L., G.H. Visser, and S. Daan. 2003. Foraging behavior and physiological changes in precocial quail chicks in res ponse to low temperatures. Physiology and Behavior 79:311-319. Laist, D.W. and J.E. Reynolds, III. 2005a. Florida manatees, warm-water refuges, and an uncertain future. Coastal Management 33:279-295. Laist, D.W. and J.E. Reynolds, III. 2005b. Influence of power plants and other warmwater refuges on Florida manatees. Marine Mammal Science 21(4):739-764. Lecis, R. and K. Norris. 2003. Habitat co rrelates of distributi on and local population decline of the endemic Sardinian newt Euproctus platycephalus Biological Conservation 115:303-317. Lefebvre, L.W., T.J. OShea, G.B. Rathbun, a nd R.C. Best. 1989. Di stribution, status, and biogeography of the West I ndian manatee. Pages 567-610 in C.A. Woods (ed.). Biogeography of the West Indies. Sandhi ll Crane Press, Gainesville, FL. Loeb, S.C. 1999. Responses of small mammals to coarse woody debris in a southeastern pine forest. Journal of Mammalogy 80:460-471. Lyons, A.L., W.L. Gaines, and C. Sevheen. 2003. Black bear resource selection in the northeast cascades, Washington. Biological Conservation 113:55-62. Marin, A.I., L. Hernandez, and J.W. Laundre. 2003. Predation risk and food quantity in the selection of habitat by black-tailed jackrabbit ( Lepus californicus ): an optimal foraging approach. Journal of Arid Environments 55:101-110.
43Martin, T.E. 1995. Avian life history evolution in relation to nest site s, nest predation, and food. Ecological Monograph 65:101-127. Milner, J.M. and S. Harris. 1999. Habitat use and ranging behaviour of tree hyrax, Dendrohyrax arboreus in the Virunga volcanoes, Rw anda. African Journal of Ecology 37:281-294. Monsen, N.E., J.E. Cloern, L.V. Lucas. 2002. A comment on the use of flushing time, residence time, and age as transport time scales. Limnology and Oceanography 47(5):1545-1553. Moore, J.C. 1951. The range of the Florida ma natee. Quarterly Journal of the Florida Academy of Sciences 14:1-50. Moran, M.J. 1985. The timing and significance of sheltering and fo raging behaviour of the predatory intertidal gastropod Morula marginalba Blainville (Muricidae). Journal of Experimental Mari ne Biology and Ecology 93:103-114. Muller, K.L., J.A. Stamps, V.V. Krishnan, N.H. Willits. 1997. The effects of conspecifics attraction and habitat quality on habitat selection in territorial birds ( Troglodytes aedon ). American Naturalist 150:650-661. Nowacek, S.M, R.S. Wells, E.C.G. Owen, T.R. Speakman, R.O. Flamm, and D.P. Nowacek. 2004. Florida manatee, Trichechus manatus latirostris respond to approaching vessels. Biologi cal Conservation 119:517-523. Ortiz, R.M., G.A.J. Worthy, and F.M. Byers. 1999. Estimation of wate r turnover rates of captive West Indian manatees ( Trichechus manatus ) held in fresh and salt water. Journal of Experime ntal Biology 202:33-38. Ortiz, R.M., G.A.J. Worthy, and D.S. Macken zie. 1998. Osmoregulation in wild and captive West Indian manatees ( Trichechus manatus ). Physiological Zoology 71:449-457. OShea, T.J. 1988. The past, present, and future of manatees in the southeastern United States: Realities, misunderstand ings, and enigmas. Pages 184-204 in Odum, R.R., K.A. Riddleberger, and J.C. Ozier (eds.). Proceedings of the Third Southeastern Nongame and Endangered Wildlife Symposiu m. Georgia Department of Natural Resources. Social Circle, Georgia. OShea, T.J., C.A. Beck, R.K. Bonde, H.I. Kochman, and D.K. Odell. 1985. An analysis of manatee mortality patterns in Fl orida 1976-1981. Jour nal of Wildlife Management 19(1):1-11.
44OShea, T.J. and H.I. Kochman. 1990. Flor ida manatees: distribution, geographically referenced data sets, and ecological and beha vioral aspects of habitat use. Pages 1122 in J.E. Reynolds, III and K.D. Haddad, eds. Report of the workshop on Geographic Information Systems as an ai d to managing habitat for West Indian manatees in Florida and Georgia. Flor ida Marine Research Publications 49. Packard, J.M. 1981. Abundance distributi on, and feeding habits of manatees ( Trichechus manatus ) wintering between St. Lucie and Palm Beach Inlets, Florida. Report to U.S. Fish and Wildlife Service Si renia Project, Gainesville, Florida. 139 pp. Packard, J.M., R.K. Frolich, J.E. Reynolds III, and J.R. Wilcox. 1989. Manatee response to interruption of a thermal effl uent. Journal of Wildlife Management 53(3):692-700. Powell, J.A., D.W. Belitsky, and G.B. Rat hbun. 1981. Status of the West Indian manatee ( Trichechus manatus ) in Puerto Rico. Jour nal of Mammalogy 62(3):642646. Powell, J.A. and G.B. Rathbun. 1984. Distri bution and abundance of manatees along the northern coast of the Gulf of Mexic o. Northeast Gulf Science 7(1):1-28. Powell, J.A. and J.C. Waldron. 1981. The manatee population in Blue Spring, Volusia County, Florida. Pages 41-51 in R.L. Brownell, Jr. and K. Ralls (eds). The West Indian manatee in Florida. Proceedings of a workshop held in Orlando, Florida, 2729 March 1978. Florida Department of Na tural Resources, Tallahassee, Florida. Reid, J.P., R.K. Bonde, and T.J. OShea. 1995. Reproduction and mortality of radiotagged and recognizable manatees on the A tlantic coast of Florida. Pp. 171-191 in T.J. OShea, B.B. Ackerman, and H.F. Perc ival (eds.). Population Biology of the Florida Manatee. U.S. Fish and Wildlif e Service, National Biological Service, Information and Technology Report 1. Reynolds, J.E., III. 1999. Efforts to conserve the manatees. Pages 267-295 in J.R. Twiss, Jr. and R.R. Reeves (eds.). Conservation and Management of Marine Mammals. Smithsonian Instit ution Press, Washington, D.C. Reynolds, J.E., III. 2000. Possible locations for long-term, water-water refugia for manatees in Florida: Alternatives to pow er plants. Report to Florida Power and Light Company, Juno Beach, FL. 33 pp. + app. Reynolds, J.E., III and D.K. Odell. 1991. Manatees and Dugongs. Facts on File, Inc., New York, NY.
45Reynolds, J.E., III and J.A. Powell. 2002. Manatees. Pages 709-720 in W.F. Perrin, B. Wursig, and J.G.M. Thewissen (eds.). Encyclopedia of Marine Mammals. Academic Press, San Diego, CA. Reynolds, J.E., III and J.R. Wilcox. 1986. Distribution and abundance of the West Indian manatee Trichechus manatus around selected Florida power plants following winter cold fronts: 1984-1985. Bi ological Conservation 38:103-113. Rommel, S.A., J.E. Reynolds, and H.A. Lynch. 2003. Adaptations of the herbivorous marine mammals. Pp. 287-306 in L. t Mannetje, L. Ramirez-Aviles, C. SandovalCastro, and J.C. Ku-Vera (eds.). Matc hing Herbivore Nutrition to Ecosystems Biodiversity. Proceedings of the VI Inte rnational Symposium on the Nutrition of Herbivores held in Meri da, Mexico, 19-24 October 2003. Sachot, S., N. Perrin, and C. Neet. 2003. Wi nter habitat selecti on by two sympatric forest grouse in western Switzerland: im plications for conservation. Biological Conservation 112:373-382. Sanford, L.P., W.C. Boicourt, and S.R. Rives. 1992. Model for estimating tidal flushing of small embayments. Journal of Waterw ay, Port, Coastal, and Ocean Engineering 118(6):635-654. Sargent, F.J., T.J. Leary, D.W. Crewz, and C.R. Kruer. 1995. Scarring of Floridas seagrasses: assessment and management op tions. FMRI Tech. Rep. TR-1. Florida Marine Research Institute, St. Peters burg, Florida. 37 pp. plus appendices. Schlesinger, C.A. and R. Shine. 1994. Selectio n of diurnal retreat sites by the nocturnal gekkonid lizard Oedura lesuerii. Herpetologica 50:156-163. Shane, S.H. 1981. Abundance, distribution, and use of power plant effluents by manatees ( Trichechus manatus ) in Brevard County, Florida. Fi nal Report to Florida Power & Light Company, Miami, Florida. Contract Number 61552-86540. National Technical Information Service PB81-147019. Squires, J.R. and S.H. Anderson. 1997. Cha nges in trumpeter swan activities from winter to spring in the great er Yellowstone area. The American Midland Naturalist 138:208-214. Stillman, R.A. and J.D. Goss-Custard. 2002. Seasonal changes in the response of oystercatchers ( Haematopus ostralegus ) to human disturbance. Journal of Avian Biology 33:358-365.
46Sucsy, P., R. Hapalo, and B. Freeman. 1998. Minimum flow determination for Blue Spring, Volusia County: the relationsh ip between ground water discharge and winter refuge for manatees. Report to Department of Water Resources, St. Johns River Water Management District, Palatka, FL. 77 pp. Sullivan, T.P. 1979. Demography of populations of deer mice in coastal forest and clear-cut (logged) habitats. Canadian Journal of Zoology 57:1636-1648. Tweheyo, M., K.A. Lye, and R.B. Weladji. 2004. Chimpanzee diet and habitat selection in the Budongo Forest Reserve, Uganda. Forest Ecology and Management 188:267278. Urian, K.W. and R. Wells. 1996. Bottlenose Dolphin Photo-Identification Workshop: March 21-22, 1996, Charleston, South Carolin a. Final Report to the National Marine Fisheries Service, Charleston La boratory, Charleston, SC. Contract No. 40EUNF500587. NOAA Technical Me m. NMFS-SEFSC-393. 92 pp. U.S. Fish and Wildlife Service. 200 1. Florida Manatee Recovery Plan ( Trichechus manatus latirostris ), Third Revision. Atlanta, Georgia. Vergara, B. 1994. Water supply needs and s ources assessment. Technical Publication SJ94-7, St. Johns River Water Mana gement District, Palatka, FL. West, A.D., J.D. Goss-Custard, R.A. Stil lman, R.W.G. Durrell, S.E.A.1.V.d., and S. McGrorty. 2002. Predicting the impacts of disturbance on shorebird mortality using a behaviour-based model. Biological Conservation 106:319-328. Williams, G.A. and S. Morritt. 1985. Habita t partitioning and thermal tolerance in a tropical limpet, Cellana grata Marine Ecology Progr ess Series 124:89-103. Yanes, M. J. Herranz, and F. Suarez. 1996. Ne st microhabitat selection in larks from a European semi-arid shrub-ste ppe: the role of sunlight and predation. Journal of Arid Environments 32:469-478. Zar, J.H. 1999. Biostatistical Analysis. 4th Edition. Prentice Hall, Upper Saddle River, New Jersey. Zimmerman, J.T.F. 1988. Estuarin e residence times. Pages 75-84 in B. Kjerfve (ed.). Hydrodynamics of Estuaries. V.1. CRC Press. Zoodsma, B.J. 1991. Distribution and behavi oral ecology of manatees in southern Georgia. M.S. Thesis. University of Florida, Gainesville, FL. 202 pp.
47 Table 1. Descriptive statistics for daily m ean water temperature in the three canal systems and Matlacha Pass for days wh en water temperature in Matlacha Pass was <18oC (N=99 days). Table 2. Descriptive statistics for co unts of boats per day in the three canal systems and Matlacha Pass. Only daytime hours (0700-1800 hours) were used in the calculations. Table 3. Descriptive statis tics of depth (m) measurem ents in the three canal systems. Data for specific canals with in Matlacha Isles are also included. LocationMeanSDMinMaxRange Matlacha Isles126.96.36.199.14.6 northern Pine Island188.8.131.52.62.1 West Island184.108.40.206.80.5 MI site 220.127.116.11.14.0 MI site 22.00.21.52.51.0 MI site 18.104.22.168.91.6 MI site 52.00.61.12.71.6 MI site 62.30.32.02.80.9 LocationMeanSDMinMaxRange Matlacha Isles 18.9oC 1.2216.021.75.7 West Island 16.6oC 1.5912.619.26.6 northern Pine Island 16.8oC 1.5313.019.86.8 Matlacha Pass 15.8oC 1.6511.718.06.3 LocationMeanSDMinMaxRange Matlacha Isles125.588.4346206160 West Island78.846.203212694 northern Pine Island178.074.7886257171 Matlacha Pass415.8237.93195649454
48Table 4. Descriptive statistics for seep detection transects in the Matlacha Isles canal system. Table 5. Results of linear regressi ons correlating counts of manatees traveling into the Matlacha Isles can al system with average water temperature during the 24-, 72-, and 120-hour periods prior to sampling. Table 6. Sampling effort for behavior al observations of tagged manatees. PeriodN R2p 24-hours120.3040.063 72-hours120.4530.016 120-hours120.5340.007 Average water temperature MI site 1 MI site 2 MI site 6 Salinity (ppt)SurfaceBottomSurfaceBottomSurfaceBottom average22.626.722.722.820.522.0 SD1.01.30.10.40.80.7 min22.214.171.1240.019.720.7 max126.96.36.1993.624.023.0 range4.16.00.33.64.32.3 N606059592929 Temperature (oC) SurfaceBottomSurfaceBottomSurfaceBottom average24.124.925.5188.8.131.52 SD0.70.40.40.40.40.4 min22.623.624.724.725.625.4 max25.525.726.425.827.526.8 range184.108.40.206.11.91.4 N606059592929 200020012002Total Number of individuals34310 Number of sampling periods117927 Mean length of sampling period (minutes)1337.81426.11418.21387.5 Number of intervals: ----All intervals2115228028737268 Confidence 1 or 2 only1941184525416327
49 Table 7. Summary of behaviors recorded and descriptive statistics for behavioral observations. Sampling periods 18, 19, and 21 we re not included in any of the analyses because each was less than 8 hours in duration. Sampling PeriodTotalFeedMillRestTravelFeedMillRestTravelFeedMillRestTravel 1300023261160.07.787.05.30.01.517.41.1 22400922830.03.895.01.30.00.615.20.2 31031413571913.612.655.3220.127.116.11.81.3 4112102361188.920.554.518.104.22.168.11.2 5137523741738.027.029.95.22.214.171.124.5 61485029452433.819.630.4126.96.36.199.01.6 71603528554221.917.534.4188.8.131.52.72.8 823368251241629.210.7184.108.40.206.78.31.1 91133412521530.110.646.013.32.30.83.51.0 1021364261081530.012.250.77.04.31.77.21.0 111827139581439.021.4220.127.116.11.63.90.9 122839267933132.523.732.911.06.14.56.22.1 132547641845329.918.104.22.168.22.214.171.124 142199149463341.622.421.015.16.13.33.12.2 1531115627606850.28.719.321.910.41.84.04.5 161991263603763.318.10.018.68.42.40.02.5 17286053152810.018.553.128.30.03.510.15.4 202937071886423.924.230.021.84.126.96.36.199 2229213737754346.912.725.7188.8.131.52.02.9 2334223010001267.329.20.03.515.36.70.00.8 24162119370673.522.80.03.77.92.50.00.4 2526315052164557.019.86.117.110.03.51.13.0 2635862831591.77.888.02.50.41.921.00.6 2725476221243229.98.748.8184.108.40.206.32.1 28324108331394433.310.242.9220.127.116.11.32.9 2934304928590.014.383.12.60.03.319.00.6 3020350211102224.610.318.104.22.168.47.31.5 Average30.416.041.012.74.22.214.171.124 SD20.96.7126.96.36.199.35.71.4 Min0.03.80.01.30.00.60.00.2 Max73.529.295.028.315.36.721.05.4 Range73.525.595.027.115.36.121.05.2 PercentNumber of Hours Number of intervals
50Table 8. Results of T-tests comparing percentages of each behavior when water temperatures were warmer versus colder the day of sa mpling than the previous 24or 72-hours. Table 9. Results of Mann-Whitney rank sum tests comparing percentages of each behavior when water temperatures were colder versus warmer the day of the sampling than the previous 24or 72-hours. PeriodBehaviorcritical tdfp value Feed-2.30250.0300 Travel-1.32250.1988 Mill-2.79250.0099 Feed-2.62250.0148 Rest3.51250.0017 Travel-0.83250.4168 Mill-4.61250.0001 72-hours prior 24-hours prior PeriodBehaviorGroupMedian25%75%p value colder80.633.0085.70 warmer30.825.0041.10 24-hours priorRest0.0328
51 Figure 1. Map of the study area, which incl udes Matlacha Pass from the northernmost tip of Pine Island to the power lines just sout h of channel marker 30. Dotted lines indicate the boundaries of the study area.
52 Figure 2a. The Matlacha Isles and We st Island canal systems, including canal identifications within Matlacha Isle s and locations of salinity stations (numbered circles) within both canal systems. Figure 2b. The northern Pine Island can al system, including locations of salinity stations (numbered circles). Figure 2. Maps of the three canal systems used in the study.
53 Figure 3. Daily minimum bottom water temp eratures at each sampling site. Data are for December through February for all three years of the study. Sites are arranged graphically in rank de scending order of the mean of their daily minimum temperature. Canal system names have been abbreviated as MI for Matlacha Isles, nPI for northern Pine Isla nd, and WI for West Island. The three sampling sites in Matlacha Pass are labele d as M30 (channel marker 30), M56 (marker 56), and M71 (marker 71). M I 6 M I 1 M I 2 M I 4 M I 5 nP I WI M I 3 M 56 M 30 M 71 Daily Minimum Temperature 10 12 14 16 18 20 22 24 26 28 Outliers 25th 75th Median Outliers 90th 10th Percentiles:
54 Figure 4. Frequencies of daily mean bottom water temperatures at each sampling site during the three winters. 0 20 40 M 56 0 20 40 M 71 0 20 40 MI 1 0 20 40 MI 2 0 20 40 MI 3 0 20 40 MI 4 0 20 40 MI 5 0 20 40 MI 6 0 20 40 nPI 121416182022242628 0 20 40 WINumber of DaysDaily Mean Temperatures (C) 0 20 40 M 30
55 Figure 5. Histogram showing the differences in daily mean bottom water temperature between each sampling s ite and Matlacha Pass (ambient). 0 25 50 MI 1 0 25 50 MI 2 0 25 50 MI 3 0 25 50 MI 4 0 25 50 MI 5 0 25 50 MI 6 0 25 50 nPI -2-101234567 0 25 50 WINumber of DaysDifference in daily mean temper ature (C) from Matlacha Pass
56 Figure 6. Surface and bottom salinities in Matlacha Isles, northern Pine Island, and West Island. Figure 7. Surface and bottom salinities in Matlacha Isles, northern Pine Island, and West Island when all s ites were sampled on the same day. Outliers 25th 75th Median Outliers 90th 10th Percentiles: M I S u rface M I Bo tto m nPI Su rf ace n P I Bo tto m WI Surface W I Bottom Salinity 5 10 15 20 25 30 35 MI Surfa c e MI Bottom n PI S u rface nPI Bottom WI Su rf a c e WI Bo tto m Salinity 5 10 15 20 25 30 35 Outliers 25th 75th Median Outliers 90th 10th Percentiles:
57 Figure 8. Surface and bottom salinities at each sampling station within the Matlacha Isles canal system. Black circles represent surface salinity readings and wh ite circles represent bottom salinity readings. Data are graphed in rank order of salinity. The gray line represents the average of all Matlacha Isles surface and bottom salin ities. Mean ppt + SD and sample size for each sampling station are reported on each graph. The surface and bottom data plots are slightly staggered so that both can be seen when they are similar or equal. Number of Observations 05101520 0 10 20 30 Surface = 17.58 S.D. 4.17 (n=21)Station 6Bottom = 18.19 S.D. 3.86 (n=21)Number of Observations 05101520 Salinity 0 10 20 30 Surface = 16.20 S.D. 3.86 (n=21)Station 5Bottom = 19.52 S.D. 4.06 (n=21) 05101520 0 10 20 30 Surface = 16.21 S.D. 3.96 (n=21)Station 4Bottom = 17.91 S.D. 3.66 (n=21) 05101520 Salinity 0 10 20 30 Surface = 15.73 S.D. 3.71 (n=20)Station 3Bottom = 20.24 S.D. 3.70 (n=20) 05101520 0 10 20 30 Surface = 14.99 S.D. 4.00 (n=21)Station 2Bottom = 17.91 S.D. 3.98 (n=21) 05101520 Salinity 0 10 20 30 Surface = 16.55 S.D. 5.72 (n=20)Station 1Bottom = 18.99 S.D. 5.29 (n=20)
58 Figure 9. Mean number of boats obs erved per hour at each location during 24-hour surveys (N=2 days). Figure 10. Mean number of boats ob served per hour at each location during daytime surveys (N=4 days). Hour 0700 0 800 09 0 0 1000 11 0 0 1 20 0 1300 14 0 0 1 50 0 1600 17 0 0 1 80 0 Mean Number of Boats 0 10 20 30 40 50 60 70 MP nPI MI WI Hour 0 000 0 30 0 06 0 0 09 0 0 1 200 150 0 180 0 2 1 0 0 Mean Number of Boats 0 10 20 30 40 50 60 70 MP nPI MI WI
59 Figure 11. Estimated volume of water (m3) remaining within each canal system during a tidal cycle. M a tlach a I sle s N. Pine Islan d West Is la nd Water Volume (m3) 0 50000 100000 150000 200000 250000 300000
60 Figure 12. Total number of manatees observe d traveling into and out of the Matlacha Isles canal system each hour, plotted on a circ ular scale. Counts were pooled from all sampling periods. The ranges of times of sunr ise and sunset during the three winters are also noted on the graph.
61 Figure 13. Frequencies of tide readings at different heights recorded at the entrance of the Matlacha Isles canal system during 24-hour sampling periods. The half-hour intervals that had the highest counts of manatees during the morning (A) and afternoon/evening (B) are repr esented on the graphs. Tide Height (cm Below Reference Point) -110-100-90-80-70-60-50-40-30Number of Observations 0 5 10 15 20 Tide Height (cm Below Reference Point) -110-100-90-80-70-60-50-40-30Number of Observations 0 5 10 15 20 All Readings Readings with Manatees Observed A B
62 Figure 14. Number of manat ees observed traveling into Matlacha Isles versus average water temperature in Matlacha Pass 24-, 72-, and 120-hours prior to each sampling period. Regression lines are also plotted. Data points when the average water temperature was <18oC are not shaded to differentiate them from data points that were used in the regression analyses. 1214161820222426Number of Manatees 0 50 100 150 200 24-Hour Temperatures 1214161820222426Number of Manatees 0 50 100 150 200 72-Hour Temperatures Water Temperature (C) in Matlacha Pass 1214161820222426Number of Manatees 0 50 100 150 200 120-Hour Temperatures A B C
63 Figure 15. Total frequencies of each ac tivity observed over all sampling periods. Hour of Day 061218 Number of Observations 0 50 100 150 200 061218 Number of Observations 0 50 100 150 200 Hour of Day 061218 0 50 100 150 200 061218 0 50 100 150 200 Rest Mill Travel FeedAB CD
64 Figure 16. Total frequencies of each activ ity observed during th e 11 sampling periods that focal manatees used Matlacha Isles during the day and Matlacha Pass at night. Hour of Day 061218 Number of Observations 0 25 50 75 100 125 061218 Number of Observations 0 25 50 75 100 125 Hour of Day 061218 0 25 50 75 100 125 061218 0 25 50 75 100 125 Rest Mill Travel FeedAB C D
65 Figure 17. Total frequencies of each activity obse rved during the 8 sampling periods that focal manatees remained in Matlacha Pass. Hour of Day 061218 Number of Observations 0 20 40 60 80 061218 Number of Observations 0 20 40 60 80 Hour of Day 061218 0 20 40 60 80 061218 0 20 40 60 80 Rest Mill Travel FeedAB C D
66 Figure 18. Percentages of time manatees were engaged in the observed behaviors versus the difference between the temperature on the day of the sampling a nd the prior 24-hour period. -2-101 Percent of Time Resting 0 25 50 75 100 Relatively Cool Relatively Warm Day of Obs. Temperature minus Prior 24 Hours -2-101 Percent of Time Feeding 0 25 50 75 Relatively Cool -2-101 Percent of Time Milling 0 5 10 15 20 25 30 35 Relatively Cool Day of Obs. Temperature minus Prior 24 Hours -2-101 Percent of Time Traveling 0 5 10 15 20 25 30 35 Relatively Cool A C D B Relatively Warm Relatively Warm Relatively Warm
67 Figure 19. Percentages of time manatees were engaged in the observed behaviors versus the difference between the temperature on the day of the sampling a nd the prior 72-hour period. -4-3-2-1012 Percent of Time Resting 0 25 50 75 100 Relatively Cool Relatively Warm Day of Obs. Temperature minus Prior 72 Hours -4-3-2-1012 Percent of Time Feeding 0 25 50 75 Relatively Cool -4-3-2-1012 Percent of Time Milling 0 5 10 15 20 25 30 35 Relatively Cool Day of Obs. Temperature minus Prior 72 Hours -4-3-2-1012 Percent of Time Traveling 0 5 10 15 20 25 30 35 Relatively Cool A C D B Relatively Warm Relatively Warm Relatively Warm