• TABLE OF CONTENTS
HIDE
 Title Page
 Introduction
 Figure 1: Place names of sites...
 Methodology
 Figure 2: Atlantic coast radio-tagging...
 Table 1: Number of radio-tagged...
 Table 2: Number of radio-tagged...
 Table 3: Descriptive statistics...
 Table 4: Causes of termination...
 Table 5: Radio-tacking efforts...
 Figure 5: Sampling effort by year...
 Figure 6: Radio-tracking records...
 Figure 7: Latitude-longitude windows...
 Figure 8: PTT locations vs. time...
 Figure 9: Annual latitudinal movements...
 Figure 10: Annual latitudinal movements...
 Figure 11: Examples of two types...
 Figure 12: Exampls of two types...
 Figure 13: Distance between Summer...
 Figure 14: Northward migration...
 Figure 15: Latitudinal movements...
 Figure 16: The mean Julian date...
 Figure 17: Year-to-year consistency...
 Figure 18: Site fidelity to warm-season...
 Figure 19: Winter distribution...
 Summary
 Acknowledgement
 Bibliography
 Report appendices
 Appendix 1a: Radio-tagged manatee...
 Appendix 1b: Radio-tagged manatee...
 Appendix 2: Description of the...
 Appendix 3: Changes in PTT duty...
 Appendix 4a: Printout of the tagging...
 Appendix 4b: Explanation of TERM...
 Appendix 5: Definition of manatee...
 Appendix 6: Preliminary assessment...
 Appendix 7: Tagged manatee use...






Title: Movements and spatial use patterns of radio-tagged West Florida indian manatees (trichechus manatus) along the Atlantic Coast of Florida and Georgia:
CITATION PAGE IMAGE ZOOMABLE PAGE TEXT
Full Citation
STANDARD VIEW MARC VIEW
Permanent Link: http://ufdc.ufl.edu/UF00067366/00001
 Material Information
Title: Movements and spatial use patterns of radio-tagged West Florida indian manatees (trichechus manatus) along the Atlantic Coast of Florida and Georgia:
Physical Description: Book
Creator: Deutsch, Charles J.
Publisher: Florida Cooperative Fish and Wildlife Research Unit, Department of Wildlife Ecology and Conservation, IFAS, University of Florida
Publication Date: 1996
 Notes
Funding: Florida Historical Agriculture and Rural Life
 Record Information
Bibliographic ID: UF00067366
Volume ID: VID00001
Source Institution: Marston Science Library, George A. Smathers Libraries, University of Florida
Holding Location: Florida Agricultural Experiment Station, Florida Cooperative Extension Service, Florida Department of Agriculture and Consumer Services, and the Engineering and Industrial Experiment Station; Institute for Food and Agricultural Services (IFAS), University of Florida
Rights Management: All rights reserved by the source institution and holding location.

Table of Contents
    Title Page
        Page 0
    Introduction
        Page 1
    Figure 1: Place names of sites along the Atlantic coast of Florida and Georgia that were commonly used by manatees or that are mentioned in the text
        Page 2
        Page 3
        Page 4
    Methodology
        Page 5
    Figure 2: Atlantic coast radio-tagging sites
        Page 6
        Page 7
        Page 8
        Page 9
        Page 10
        Page 11
        Page 12
    Table 1: Number of radio-tagged manatees along the Atlantic coast of Florida and Georgia over the period of May 1986 to May 1996
        Page 13
    Table 2: Number of radio-tagged manatees along the Atlantic coast of Florida and Georgia over the period of May 1986 to May 1996
        Page 14
        Page 15
        Page 16
    Table 3: Descriptive statistics on the number of tagging bouts per manatee...
        Page 17
    Table 4: Causes of termination of 424 radio-tagging bouts along the Atlantic coast of Florida and Georgia over the period of May 1986 to May 1996
        Page 18
    Table 5: Radio-tacking efforts by year and tag type
        Page 19
        Page 20
    Figure 5: Sampling effort by year and tag type
        Page 21
    Figure 6: Radio-tracking records of Atlantic coast manatees by tag type
        Page 22
        Page 23
        Page 24
        Page 25
        Page 26
        Page 27
        Page 28
    Figure 7: Latitude-longitude windows used ti flag putative outliers in the PTT telemetry database during SAS data processing
        Page 29
        Page 30
    Figure 8: PTT locations vs. time of day
        Page 31
        Page 32
    Figure 9: Annual latitudinal movements of subadult male TBC - 27
        Page 33
    Figure 10: Annual latitudinal movements of adult female TBC - 24
        Page 34
        Page 35
    Figure 11: Examples of two types of seasonal migratory patterns (TBC - 24)
        Page 36
    Figure 12: Exampls of two types of seasonal migratory patterns (TBC - 37)
        Page 37
    Figure 13: Distance between Summer and Winter ranges (km)
        Page 38
    Figure 14: Northward migration of adult male TBC - 42
        Page 39
        Page 40
    Figure 15: Latitudinal movements of adult female TBC - 09
        Page 41
    Figure 16: The mean Julian date ar which migratory tagged manatees departed Brevard Country in Fall...
        Page 42
        Page 43
    Figure 17: Year-to-year consistency in annual movement patterns
        Page 44
    Figure 18: Site fidelity to warm-season home range
        Page 45
    Figure 19: Winter distribution (Dec-Feb) of satellite-determined location for eight tagged manatees in north Biscayne Bay
        Page 46
        Page 47
        Page 48
    Summary
        Page 49
    Acknowledgement
        Page 50
    Bibliography
        Page 51
        Page 52
        Page 53
        Page 54
    Report appendices
        Page 55
    Appendix 1a: Radio-tagged manatee identity database
        Page 56
        Page 57
        Page 58
    Appendix 1b: Radio-tagged manatee identity database...showing capture and release location and method of capture
        Page 59
        Page 60
        Page 61
    Appendix 2: Description of the type of sensor data collected from satellite-monitored PTTs
        Page 62
        Page 63
        Page 64
        Page 65
        Page 66
    Appendix 3: Changes in PTT duty cycles and turn-on times used for radio-tracking manatees along the east coast of Florida and Georgia
        Page 67
        Page 68
        Page 69
    Appendix 4a: Printout of the tagging history lookup table for manatees radio-tagged along the Atlantic coast
        Page 70
        Page 71
        Page 72
        Page 73
        Page 74
        Page 75
        Page 76
        Page 77
        Page 78
        Page 79
        Page 80
        Page 81
        Page 82
    Appendix 4b: Explanation of TERM Codes and Off Comments in the tagging history lookup table for the causes for tagging bout termination
        Page 83
    Appendix 5: Definition of manatee activity categories and database codes for VHF radio-telemetry data collection and data entry
        Page 84
        Page 85
        Page 86
    Appendix 6: Preliminary assessment of the precision (i.e. repeatability) associated with digitizing manatee point locations
        Page 87
    Appendix 7: Tagged manatee use of the Cocoa Beach/Thousand Islands area
        Page 88
        Page 89
        Page 90
        Page 91
        Page 92
        Page 93
Full Text








Movements and Spatial Use Patterns of Radio-tagged West Indian Manatees
(Trichechus manatus) along the Atlantic Coast of Florida and Georgia:
A Progress Report



Charles J. Deutsch

Florida Cooperative Fish and Wildlife Research Unit
Department of Wildlife Ecology and Conservation, University of Florida
and
Sirenia Project, Florida Caribbean Science Center, National Biological Service
U.S. Department of the Interior

30 September 1996


Progress Report submitted to the Florida Cooperative Fish and Wildlife Research Unit
under Research Work Order No. 125.


Contact Address:


Sirenia Project, National Biological Service
412 NE 16th Ave., Rm. 250
Gainesville, FL 32601 U.S.A.


* Please do not cite without the permission of the author.









INTRODUCTION


The West Indian manatee (Trichechus manatus) is a large, herbivorous aquatic mammal
that lives in shallow estuaries, rivers and coastal areas of the New World tropics and subtropics
(Lefebvre et al. 1989). The Florida subspecies (T. m. latirostris) occurs at the northern end of
the species' range. Their low metabolic rate makes them susceptible to cold stress in winter,
hence limiting their northward distribution and affecting their behavior and movements (Irvine
1983). During the warm-season, manatees can be found throughout the coastal areas and
freshwater river systems of Florida and Georgia, but in winter most manatees aggregate around
warm-water sources, either natural springs or industrial effluents, in central and southern Florida
(Hartman 1979) (Fig. 1).

The Florida manatee is endangered because of its small population size, low reproductive
rate, increasing human-related mortality, and the impacts of the burgeoning human population
on coastal habitats (U.S. Fish and Wildlife Service 1996). While manatee numbers in some
parts of Florida appear to have increased at a healthy rate in recent years, the population along
the eastern seaboard has apparently grown slowly or remained stable (Eberhardt and O'Shea
1995; but see Garrott et al. 1994), probably as a result of excessive mortality from anthropogenic
causes (Ackerman et al. 1995, O'Shea and Langtimm 1995). Evidence for a decline in the
number and proportion of calves in this population since the early 1980's is also a serious
concern, suggesting a decline in female fecundity or a rise in calf mortality (Reynolds and
Wilcox 1994). Life-history data from photo-identification and radio-telemetry studies along the
Atlantic coast indicate that reproductive rates seem comparable to those found in regions with
growing manatee populations, but firm data on calf and subadult survival are lacking (Reid et al.
1995).

The Sirenia Project undertook a radio-telemetry study starting in 1986 to gather
information on the ecology, behavior and life-history of manatees living along the Atlantic coast
of Florida and Georgia (Sirenia Project 1992). The ultimate goal was to provide managers with
solid data upon which to base decisions regarding the best ways to protect manatees and their
habitats over this extensive region. At that time, manatee distribution and movements along the
east coast were known primarily from aerial surveys of particular areas, anecdotal records based
on interviews and opportunistic sightings, and on resightings of scarred, identifiable individuals.
These studies documented a seasonal migratory pattern, with manatees moving into southeast
Florida for the winter and dispersing northward for the warm season (e.g., Hartman 1974,
Kinnaird 1985, Weigle et al. 1987, Reid et al. 1991). Aerial surveys in Brevard County, for
example, found relatively low numbers in winter and rapid increases during the spring and fall,
suggesting that a wave of migrants moved into central Florida during these seasons (Shane 1983,
Provancha and Provancha 1988). Photo-identification records of uniquely scarred" individuals up
to 1986 confirmed the long-distance (> 500 km) seasonal movements of many manatees along
the Atlantic coast and demonstrated individual site fidelity to winter aggregation sites (Reid et


1


1^1 _~ _~_I _




































S""" .. 0 10 Kilometers
Riviera Beach "'.
N Melbourne

Port Everglades '
SLittle River '"
Coral Gables
0 100 Kilometers Sebastian














Figure 1. Place names of sites along the Atlantic coast of Florida and Georgia that were commonly
used by manatees or that are mentioned in the text. Major and minor winter aggregation sites, mostly
thermal effluents, are indicated by large and small triangles, respectively. Brevard County, the area
most heavily utilized by tagged manatees in this study, is shown in detail.









al. 1991). This work also showed that manatees move among thermal refuges, both between and
within winter seasons.

Radio-telemetry provides a different window on manatee biology than that provided by
aerial censuses or photo-identification. Aerial surveys yield brief snapshots of the distribution of
a large number of manatees in a given area, whereas telemetry provides an intensive time-series
of the movements of a relatively small number of individuals. Aerial survey data can present a
biased picture of manatee spatial use patterns when there is a distinct diel cycle to their
movements (Rathbun et al. 1990) or when environmental (e.g., water turbidity) and behavioral
(e.g., diving pattern) factors that are known to affect manatee visibility vary substantially among
surveyed areas (Lefebvre et al. 1995). Photo-identification is valuable in documenting gross
movements of individuals over long periods of time (Reid et al. 1991) but it lacks the fine
temporal resolution of movements provided by radio-tracking methods, and data collection is
necessarily opportunistic, especially outside of the cold season. During the 1970's and early
1980's radio-telemetry was used successfully to track manatee movements in other regions of
Florida-including the upper St. John's River (Bengtson 1981), the Big Bend area on the Gulf
coast (Rathbun et al. 1990), and southwest Florida (Lefebvre and Frohlich 1986)-but these
early studies relied principally on field-monitored VHF transmitters and were limited in their
spatial and temporal extent (see O'Shea and Kochman 1990 for overview). Typically, animals
were tagged at a thermal refuge in the winter and then tracked until the tag detached or the
battery failed, usually sometime in the summer or fall of the same year. The Atlantic coast
telemetry project departed from previous studies in a number of important ways. First, it has
relied heavily on satellite-monitored transmitters, yielding an approximate ten-fold increase in
the frequency of locations compared to a VHF tracking study, and the temporal and spatial
distribution of fixes are not biased by observer accessibility. The study animals were still
tracked in the field, however, to provide information on reproductive status, activity, habitat use
and tag condition. Second, it has encompassed the entire range of the Atlantic coast population,
from the Florida Keys to Georgia, so inferences are not limited to a small geographic area.
Third, the study has spanned a decade, permitting an examination of interannual variation in
movement patterns as a function of winter severity, reproductive status and other factors. Many
of the same individuals have been tracked for a large portion of this period, providing precise
measures of female reproductive parameters (Reid et al. 1995) as well as allowing us to
determine within-animal variation in seasonal movements, site fidelity and spatial use patterns.
The strength of this research program lies in its emphasis on long-term, longitudinal tracking of
individuals.

Much of the work accomplished over the course of this Research Work Order involved
the development, refinement and quality control of spatial databases relating to manatee radio-
telemetry. While many of the problems and limitations in interpreting and analyzing manatee
distribution and movements from telemetry data were anticipated (White and Garrott 1990,
Deutsch 1994), others were only discovered after thorough scrutiny of the databases. The
general approach that I have taken in analyzing these large and complex sets of data follows.


3









(1) Update Existing Databases: This was a continuing process, of course, but a large backlog
of data had accumulated that required digitizing or time-consuming data processing.
(2) Create New Databases: It was necessary to create new databases on the characteristics of
tagged manatees and on water temperature in order to extract particularly useful
information from the principal telemetry locational databases. The tagged manatee
identity database, for example, has allowed me to create subsets of the data on migratory
movements by sex, age class and rehabilitation status. Several databases on hourly water
temperature at different sites within the study area will be valuable in examining
temperature-induced changes in manatee movements.
(3) Improve and Implement Quality Control of Databases: This effort involved clarification of
variable definitions, standardization of data collection protocols, improvements in
computerized error-scanning routines, and development of new error-detection programs.
(4) Characterize the Databases: A thorough description of each database was required,
including variable definitions, database structure, sampling effort, current status of
updates, and the limitations and biases of individual variables and of the database as a
whole. Descriptive statistics are provided to characterize each database.
(5) Analyze Annual Patterns of Movements: This work involves analyses of migratory
behavior, including distance between summer and winter ranges, site fidelity, travel
rates, and timing of long-distance seasonal movements in relation to temperature change.
These movement variables are examined as a function of sex, age class and rehabilitation
status.
(6) Analyze Within-season Movement Patterns: This work includes analyses of home range
patterns, power plant use in winter, and relationships of spatial use patterns to
environmental features (e.g., seagrass beds) using Geographic Information Systems
(GIS), again as a function of individual attributes.

I will describe the progress made in each of these categories and indicate work that still
remains to be done. Substantial progress has been made toward completing the most important
tasks in the first four categories, that is, in creating, preparing and describing high-quality
telemetry databases for analysis. Relatively less effort has been directed toward spatial analyses.
The time period covered is from May 1986 through May 1996 (December 1986 to September
1995 for the satellite-determined locations). Since this work is still in progress, all results are
preliminary and subject to modification.


4









METHODS


Study Area
Our study area encompassed the entire east coast of Florida and southeast Georgia,
including inland waterways such as the lower St. John's River (Fig. 1). Habitats used by tagged
manatees along this 750-km stretch of coastline varied widely and included: rivers, canals,
estuaries and the ocean close to shore; freshwater, brackish and marine water bodies; polluted
rivers in highly modified urban environments; relatively undisturbed mangrove swamps, salt
marshes, and seagrass flats; and human-created features, such as marinas, residential canal
systems, thermal industrial effluents, and freshwater sewage effluents.

Capture, Tagging and Radio-tracking of Manatees
A total of 83 manatees were radio-tagged and tracked along the Atlantic coast between 8
May 1986 and 31 May 1996. Most (n = 48, 57.8%) of these animals were tagged and released in
the Indian River lagoon system in Brevard County along the central Florida coast (Fig. 2). The
remainder were tagged in about equal numbers in south Florida (n = 16, 19.3%), including three
in the Florida Keys, and in north Florida and southern Georgia (n= 19, 22.9%). One additional
manatee (TBS-01, "Bertram"), tagged and released at Blue Spring and tracked briefly in the
upper St. John's River system in 1995, is not included in this study. Figure 2 shows the number
tagged and released at each specific location.

About half of the study animals (n = 42) were initially captured alongshore and
restrained with nets; three of these involved rescues of manatees in distress, followed by
immediate release on site. Another 17 free-ranging manatees were initially belted and tagged by
a researcher snorkelling quietly up to the unrestrained animal while it was resting. Twenty
manatees in this study had been rescued and brought into captivity because they were sick,
severely injured, or found as orphans; after a period of rehabilitation at oceanaria, these animals
were radio-tagged and released, usually near the point of original capture (Bonde et al. 1995).
Four additional animals were born and raised in captivity. Specific information on capture
location, release location, and method of capture is listed for each individual in Appendix lb.

The tag and harness assembly consisted of a floating radio-transmitter attached with a 4-
foot, flexible nylon tether to a padded belt that fit snugly around the manatee's peduncle (for
details, see Rathbun et al. 1987 and Reid and O'Shea 1989). Two types of radio-transmitters
were used in this study: very-high frequency (VHF) transmitters and ultra-high frequency
(UHF) transmitters, known as platform transmitter terminals (PTTs). Twenty-one manatees
carried only the smaller, field-monitored VHF tags, while 62 animals were tracked using
combinations of VHF tags and the larger, satellite-monitored PTTs. Most of the PTT tags
contained a weaker VHF transmitter within the housing, permitting the manatees to be tracked in
the field. In the latter part of the study, sonic transmitters were also incorporated into the
housing to provide a means by which field personnel could locate a sunken tag or a manatee
whose radio-transmitter had ceased functioning. The battery life of these transmitters was
approximately 2 years for VHF tags and 8-9 months for PTTs. Radio-transmitters that had little


5


~ _~11~-11~_-~














Atlantic Coast Radio-tagging Sites


NORTHERN FL / SOUTHERN GA SITES
SBrunsuck (3)
Kings Bay (1)
Femandina Beach (12)

_ range Park (1)
Guana Lake (2)



CENTRAL FLORIDA SITES
Indian River
Banana Ceek (15)
Pat St John (2)
Banana River
L per Banana River (21)
S dVBannaa River (10)


SOUTHERN FLORIDA SITES
R. Pierce (1)


t Luae(1)
Hobe Sound (1)
Jupter Sound (1)
Riviera Beach (1)


100 0 100


Kilometers




,- a.1


Figure 2. Locations of initial tagging and release sites for 83 manatees tracked along the
Atlantic coast of Florida and Georgia. The number of manatees initially tagged at each site is
given in parentheses. An additional 347 retagging events took place throughout the study area.


Port Everglades (3)
Little River (1)
Bscayne Bay (2)


- Key Largo (1)
Tavenier (2)









battery life remaining, that were encrusted with barnacles (and hence sinking), or that had
electronic or physical malfunctions were often removed and replaced with a fresh tag by
researchers snorkelling up to the manatee; this resulted in the maintenance of continuous
tracking bouts of individuals that were longer than the battery life of the tag. Radio-tags
frequently became detached at the designed weak link in the base of the tether, which ensured
that the manatees could free themselves from a potential tag entanglement. Field researchers
have been very successful in relocating and reattaching tags to these manatees, usually without
recapturing the animal, thus allowing many individuals to be tracked intermittently over a
considerable period of time. A total of 54 (65%) of the study animals have been tracked over
multiple tag deployments.

VHF Tracking Methods and Data Collection
All manatees were tracked from shore, by boat, or occasionally by single-engine airplane
using conventional VHF radio-tracking equipment about once or twice per week. The locations
of visual sightings were plotted in the field on copies of USGS quadrangle maps or NOAA
nautical charts (source map scales = 1:24,000 and 1:40,000, respectively). Locations of tagged
manatees that could not be determined visually were estimated using triangulation or were
recorded as general areas (i.e., a polygon which enclosed the probable location). In addition,
sightings of tagged manatees made by the public were included after verification of information.
The following data were recorded on tracking maps in the field: individual identification, date,
locations) and movements, timess, observerss, observation platform (boat, truck or plane),
precision of location (visual point location, triangulation, general area polygon), bearing and
signal strength (for non-visual locations), manatee group size, presence/absence of calf, animal
activity, condition of transmitter housing, tether and belt, and sometimes associated
environmental information (e.g., water temperature, salinity).

PTT Tracking Methods and Data Collection
The satellite-monitored radio-tags transmitted data to two (or three) polar-orbiting Tiros-
N NOAA weather satellites and the data were processed by Service Argos' Data Collection and
Location System (DCLS) (Service Argos, Inc. 1990). Platform locations were calculated from
the Doppler shift in the perceived frequency of the transmitted signal as the satellite approached
and then moved away from the animal. Service Argos provided the following data for each
satellite pass over a transmitting PTT: PTT ID number, date, time (GMT), latitude and
longitude (for the WGS-84 datum) in decimal degrees, location quality, number of transmitter
messages received, ID of satellite that received message, and a variety of requested sensor data
(see below). With Argos' Location Class Zero Service (LCO, formerly called Animal Tracking
Service) we were able to access the following information as well: number of messages received
with signal strength greater than -120 dB, the best signal strength of any message, the two
possible location solutions (located at equal distances on either side of the satellite's ground
track), and specific information on oscillator drift within and between satellite passes that can
account for poorer locational accuracy. See Fancy et al. (1988) and Harris et al. (1990) for
excellent descriptions of the Argos DCLS and its application to tracking wildlife.


7









PTT Location Quality
Satellite-determined locations were reported to the nearest 0.001 degree latitude and
longitude, constraining locational precision to approximately 100 m; consequently, telemetry
locations displayed on a relatively small area that is heavily used by tagged manatees appear as a
grid of points spaced at 100-m intervals. A number of factors associated with the geometric
conditions of the satellite's orbit (e.g., distance of satellite ground track from PTT, error in
orbital determination) and the PTT itself (e.g., signal strength, number of messages, stability of
frequency of transmitted signal) contributed to the error of the estimated locations from their
true locations. Since 1986, Service Argos has provided a location class (LC) as a measure of the
relative accuracy of each location. While Argos has improved its methods of location
determination, the meanings of the location classes and the algorithms used to assign them have
changed over the course of this study. The major changes are summarized here.

During the first several months of this study, through 31 March 1987, Argos defined a
Location Calculation Quality Index (NLOC) as follows:
Time Period: <01Dec86 31Mar87
NLOC Meaning
-1 = PTT not located, last location determined from 1 pass only.
0 = PTT not located.
1 = Location determined from 2 passes of same satellite.
2 = Location determined from 2 passes, one by each satellite.
3 = Location determined from 1 pass only.

Information on the accuracy and precision of locations associated with these location quality
indices is provided in Fancy et al. (1988). As one would expect from the above criteria, the
mean error of NLOC3 locations is substantially greater than that of NLOC1 or NLOC2
locations.

In April 1987, Service Argos introduced a different location classification system based
on improved software for processing locational data from PTTs. Most of the data in our PTT
database conform to the following location classes, sometimes also termed location quality
indices (LQ or NQ):
Time Period: 01Apr87 14Jun94
LC Precision (1 SD) Required Conditions
3 150 m 5 messages, 420 sec between 1st & last messages,
very good oscillator stability & geometric conditions.
2 350 m 5 messages, 420 sec between 1st & last messages,
good oscillator stability & good geometric conditions.
1 1000 m 2 4 messages, 240 sec between 1st & last messages,
reasonable oscillator stability & good geometric conditions.
0 None given 2 messages, pass duration <240 sec, oscillator unstable,
or satellite ground track too close or too far from PTT.


8









The Sirenia Project started receiving LCO service for manatee tracking in November 1987; it
was essentially a "catch-all" class for any locations rejected by the other classes. Precision is
measured as RMS (root mean square) Error, which is equal to one standard deviation of the
locational errors. Service Argos' technical bulletins claim that 68% of locations would lie within
a circle of radius equal to the RMS error around the true location. The error is apparently
applied to the latitudinal and longitudinal components separately, however, resulting in a square
(not a circle) that contains approximately 68% of the observations (Chris Estes, Service Argos,
pers. comm.). Note that the highest accuracy under this system was obtained with a LC3,
whereas a location quality index of 3 (NLOC3) under the previous classification system gave the
poorest accuracy. To clarify the different meanings given by these classification systems, all
location classes prior to 1 April 1987 have been prefaced with an 'X' (e.g., LC = 'X3' instead of
'3').

The second major change in the way that location classes were determined and reported
occurred on 15 June 1994. Now location classes 1-3 mean basically the same as in the previous
system, although the algorithm for classification has been greatly changed. The original LCO
was redefined and has essentially been subdivided into four classes (0, A, B, Z), the latter three
of which became available under Argos' Auxiliary Location Processing. The company claims
that most of the new LCO locations are accurate to within 5000 m. Service Argos has not
rigorously defined its criteria for classification but has provided the following information in its
bulletins.


Time Period: 15Jun94 Present (31May96)
LC Precision (1 SD)
3 150 m 4 me
2 350 m 2 4 me
1 1000 m > 4 me
0 >1000 m 2 4 me
A None given 3 mess
B None given 2 mess
Z None given Unvali


Required Conditions
ssages, > 2 of the 4 plausibility tests are passed.
ssages, > 2 of the 4 plausibility tests are passed.
ssages, > 2 of the 4 plausibility tests are passed.
ssages, 2 of the 4 plausibility tests are passed.
ages, 2 of the 4 plausibility tests are done.
ages, 2 of the 4 plausibility tests are done.
dated Location. <2 of the 4 plausibility tests are
Implausible velocity, transmitter frequency
lity, or satellite passed directly overhead the
m.


Service Argos introduced the plausibility tests to ensure that the more plausible of the two
possible locational solutions is given as the platform location (Service Argos 1995). The four
types of plausibility checks are: (1) Check that the frequency calculation for the assumed correct
location is more plausible than that for the assumed image location (requires > 4 messages); (2)
check that the calculated transmitted frequency of the chosen locational solution is closer to the
previous calculated frequency (which must be < 12 hr old) than that of the assumed image
location ; (3) compare the distances of the chosen and assumed image locations to the previous
validated location; and (4) check the mean transmitter speed since the previous location.


9









PTT Sensor Data
The satellite-monitored radio-tags provided information about the manatee's behavior
and environment, as well as its location. Depending on the hardware incorporated into the PTT
and how it was programmed, between three and seven types of sensor data were provided with
each location. Data were collected on the following variables: (1) PTT temperature; (2) short-
term (1-hr) activity index; (3) long-term (12-hr) activity index; (4) low battery voltage flag; (5)
average PTT dive time over a 4-hr period; (6) PTT dive count over a 4-hr time period; and (7)
saltwater switch fail-safe flag. The temperature sensor provided counts that must be converted
to temperature in degrees Celsius using a PTT-specific temperature calibration curve provided
by Telonics, Inc. The accuracy of the temperature sensor is given as approximately 2 oC (B.
Burger, Telonics, Inc.). The insulating effect of the tag cannister results in a time lag between
temperature change in the water and in that reported by the PTT; furthermore, the housing is
subject to the influence of solar radiation (resulting in an overestimate of temperature when the
tag remains at the surface on sunny days), warming by the electronic circuitry, and possibly to
evaporative cooling on windy days (potentially resulting in an underestimate of temperature).
Technical information on each of these sensor types and the way in which the data were
formatted in the Argos data transmission stream are provided in Appendix 2.

PTT Duty Cycles
The NOAA satellites pass over the Florida region an average of 9 times (range, 8 12)
per 24-hr period, each overpass lasting about 10 minutes (Fancy et al. 1988). To optimize the
tradeoff between maximizing the number of locations per day and conserving battery power
(hence extending PTT longevity), a duty cycle was programmed into each PTT. This turned the
units on at predetermined intervals that corresponded to the four relatively fixed periods that
satellites passed over the region. Most PTTs used since 1987 were programmed with the
following optimum duty cycle: 2 hr On 4 Off- 2 On 3 Off- 2 On 6 Off- 2 On 3 Off. The
transmitters sent messages for a total of 8 hr over each 24-hr period, providing us with locations
during four 2-hr time periods per day. The duty cycle would start when a magnet was removed
from the outside of the cannister; the turn-on time was varied between 0630-0900 hr EST in
order to correspond to changes in the timing of satellite overpasses. For a PTT started at 7 AM,
the periods of transmission were 0700-0900 hr, 1300-1500 hr, 1800-2000 hr, and 0200-0400 hr.
The PTT's internal clock has sometimes drifted by as much as 0.5 hr over its deployment period,
thus shifting the duty cycle slightly. Two other duty cycles were utilized in the early years of
the study; details on these and on recent changes in turn-on times are provided in Appendix 3.

PTT Data Access
PTT location and sensor data were acquired from Argos' Data Processing Center in a
number of ways. The data that was developed into our spatial databases and used in all analyses
was sent to the Sirenia Project on 9-track tapes (December 1986 September 1989) or diskettes
(since October 1989) on a monthly basis. Project staff usually queried the Argos data every day
using personal computers and modems in order to monitor the manatees' movements and tag
condition (as indicated by the battery flag and activity sensors). In recent years, two-day


10









compilations of PTT data were also provided via electronic mail over the Internet. Methods of
data processing are described below.


Radio-telemetry Databases: Description, Current Status, Sample Sizes, and
Improvements in Quality Control

There are four principal databases associated with the radio-telemetry study of manatees
along the Atlantic coast of Florida and Georgia: (1) the Tagged Manatee Identity Database
provides data on the attributes of the study animals; (2) the Tagging History Lookup Table
gives information on all tagging and retagging events for each individual, including dates and
reasons for detachment; (3) the VHF Radio-telemetry Database provides locational and
behavioral data on manatees tracked in the field; and (4) the PTT Radio-telemetry Database
provides the bulk of data on manatee locations, as determined from satellite-monitored radio-
transmitters. Another PTT database that will not be discussed in this report comprises locational
data collected from fixed radio-tags during two separate experiments designed to determine the
accuracy and precision of the locations provided by Service Argos. All of the above data are
maintained as permanent SAS databases at the Sirenia Project, although the tagging history
lookup table is entered and also maintained as a dBase III file. The spatially referenced datasets
have also been incorporated into a GIS, specifically ARC/INFO and ArcView (Environmental
Systems Research Institute (ESRI)), for map creation and spatial analysis of distribution and
movements in relation to environmental features. Each database is described in detail below,
including the following information: database structure and variables; current status of updates;
sample sizes; improvements made to data collection and data entry protocols; improvements
made to database quality control procedures; and other work accomplished. Information in each
database is further characterized by descriptive statistics, tables and figures.


1. Tagged Manatee Identity Database- Attributes of Study Animals
In order to analyze the telemetry data by individual attributes, it was first necessary to
create a permanent SAS database (tagmanid.sd2) that included the following data for each
tagged manatee: identification number, name, sex, captive before tracking bout?, dates of
capture and tagging/release, duration in captivity, capture and tagging/release locations, capture
method, age class at capture and tagging/release, and total length at capture and tagging/release.
It also includes data on number of days tracked and percent of tracking period (i.e., period from
initial tagging to last day with tag) with functioning transmitter for each manatee through 31
May 1996. For free-ranging manatees, date, location, age class and length have the same values
at capture and at tagging/release, and duration in captivity equals zero months. This
information, when merged with the VHF and PTT telemetry databases by ID, allows the
generation of data subsets according to sex, age/size class, and rehabilitation status. A complete
printout of the tagged manatee identity database is given in Appendix lab.


11









Our sample of radio-tagged manatees included 32 males (38.6%) and 51 females
(61.4%); at initial tagging, most animals were adults (n = 66, 79.5%), followed in number by
subadults (n = 12, 14.5%) and dependent calves (n = 5, 6.0%) (Table 1). Four of the five calves
were tracked with their mothers and subsequent to weaning; one female calf (TBC-39) was only
tracked while dependent on its mother, so the effective sample size of independent tagged
animals is 82. The mean ( SD) total length at tagging and release was 285 ( 32) cm and
ranged from 200 to 350 cm (Fig. 3). The smallest independent animal, an orphan (TBC-34) held
in captivity for 8 months, was only 219 cm at release.

Fifty-nine (71.1%) of the study animals were free-ranging at initial capture and tagging.
Of the 24 manatees (28.9%) that were tagged and released after rehabilitation in captivity, most
were male (n = 14, 58.3%) and adult at time of tagging (n = 17, 70.8%) (Table 1). Half(n = 12)
of the captive group of animals were adults at the time they were rescued from the wild and
brought into captivity; two were subadults, one was a dependent calf (with its mother), five were
orphaned calves, and four were born in captivity (Table 2). One of the captive-born manatees
(TBC-39) was tagged and released with its mother at an age of 10 months, while the other three
were released as independents at ages of 2.4 to 5.0 years. The duration that these rehabilitated
manatees had spent in captivity at oceanaria before release ranged from 6 days to 7.1 years, with
a median duration of 8.6 months (Fig. 4). Most rehabilitated manatees (n = 14, 58.3%) had been
in captivity for less than one year and three (12.5%) had been captive for over 5 years (Fig. 4).

The identity database is current and verified through 31 May 1996. It needs to be
expanded to include data on: year of birth (for calves), the approximate dates that individuals
change age classes (from. calf to subadult and from subadult to adult), and additional length
measurements and associated dates after initial capture and release. A reproductive status
database for tagged adult females also needs to be developed so that movements and habitat use
can be easily compared between females with and without calves and analyzed as a function of
calf age. The information on number of days tracked and percent tracking period currently in
the identity database will be removed once a separate database created from the tagging history
lookup table is completed.


2. Tagging History Lookup Table
Information on tagging and retagging events is found in the tagging history lookup table,
which encompasses all manatees tagged with PTT and VHF radio-tags by the Sirenia Project
since 1986, including the Atlantic coast of Florida and Georgia (n=83), the upper St. John's
River (n=l), Puerto Rico (n=8), southwest Florida (n=3), and test PTTs used in the locational
accuracy experiments. Data are entered and maintained as a dBase III file and then converted to
a SAS database (lkuptags.sd2); a printout of the Atlantic coast and test PTT records is provided
in Appendix 4a. The lookup table contains data on the following 13 variables: animal
identification number, name, tag number (VHF tags have 3 digits, PTTs have 4 or 5 digits), bout
number (0 for VHF tags, numbered sequentially for PTTs), date and time tag was deployed,
comment about deployment or redeployment (e.g., net tag, free tag, replace tag), estimated or


12










Table 1. Number of radio-tagged manatees along the Atlantic coast of Florida and Georgia
over the period of May 1986 to May 1996 by age class at tagging, sex and captive status
prior to tagging.

AGE CLASS AT TAGGING


Calf

_M __ _All

2 1 3

1 1 2


3 2 5


Subadult

M _E_ All

3 4 7

4 1 5


7 5 12


Adult

_M _E All

13 36 49

9 8 17

22 44 66


Total

_M EL All

18 41 59

14 10 24


32 51 83


Free-range

Captive


Total











Table 2. Number of radio-tagged manatees along the Atlantic coast of Florida and Georgia
over the period of May 1986 to May 1996 that had been captive prior to tagging
according to age class at start of captivity and sex.

AGE CLASS AT START OF CAPTIVITY
Dependent Calf


Wild-born
Captive-born Orphan


3


5


Wild-born
with Mother


1


Subadult


0


Adult Total


5


14


1 0 0 2 7


4


5


1


2


12


14


Male


Female


Total


10

24


Duration in Captivity




5-




4 -- -- ----- -
4-_



LL" 2




0
<= 1 >1-6 >6-12 >12-24 >24-60 >60
Duration in Captivity (mos.)

Figure 4. Frequency distribution of the duration that rehabilitated and captive-born manatees
spent in captivity prior to radio-tagging and release on the east coast of Florida.









known date and time tag was replaced or detached, comment about detachment, termination
code (e.g., replace transmitter, tag lost and tether broken at weak link, tag struck by boat), belt
size and color of nylon webbing, and tether strength.

The dates and times of PTT locations are "looked up" for a given tag number against the
tagging history database to filter out locations from non-deployed PTTs and to assign the correct
manatee ID and name to locations from deployed PTTs. The accuracy of the information
(especially tag number, deployment and detachment dates and times) in this database is therefore
crucial to the construction of an accurate PTT radio-telemetry database; one error in the tagging
history database can result in numerous serious errors in the PTT database. Three types of
procedures were implemented to assure quality control of this relatively small but important
database. First, the tagging history database was scanned in SAS (program by H. Kochman) for
logical errors in dates and times (e.g., ondate/time is later than offdate/time for a given
deployment; ondate/time is earlier than offdate/time of previous deployment), for unusually long
deployment durations (i.e., >270 days), and for problems that may arise from the conversion of
dBase to SAS files. In addition, frequency tables of animal IDs and names were examined to
ensure that only one name was used for each ID number and vice-versa. Potential errors
detected with these scans were checked and corrected. Second, the dates and times of
deployment and detachment were carefully proofed against the tagging history files for each
manatee. Third, we realized that because the date and time of tag recovery was often used for
the date and time of detachment, especially in the older records, this resulted in the inclusion of
locations from detached, drifting radio-tags in the PTT database. To remedy this problem, the
locational and sensor data from each premature detachment (e.g., from break of tether at the
designed weak link) were scrutinized to estimate the date and time on which we were reasonably
certain that the tag was still attached to the manatee; the 12-hr activity sensor typically reported
zero tips after detachment. A hypothetical example will illustrate the process: turn-on time for
the PTT duty cycle was 0700 hr; tag was recovered at 1600 hr, 10 Jan.; Argos data indicated
zero tips on the long-term activity counter since 1930 hr, 8 Jan.; a moderate long-term tip count
was reported for two prior locations on 8 Jan. (0800 and 1400 hr) but it is important to realize
that this indication of activity refers to the previous 12-hr reporting period (1900 hr, 7 Jan. to
0700 hr, 8 Jan.); and accessory information (i.e., short-term activity count, movement distances)
are then used, when possible, to estimate the time window within this 12-hr period in which
detachment occurred. Estimates are conservative to minimize the likelihood of locations from
detached PTTs being included in the database; records from detached PTTs were then purged
from the PTT database. This process could not be employed for VHF tags and so some of the
deployment durations are likely to be slightly inflated; however, this does not affect the validity
of data in the VHF telemetry database, which is not dependent on the tagging history lookup
table.

The tagging history lookup table is now updated and verified through 31 May 1996. The
following changes were also made to this database: (a) The animal identification field has been
changed from the old 'S-#'s (originally used to differentiate satellite-monitored individuals from
those with only VHF tags, which were denoted with 'T-#'s) to 'T-#'s for all records. (b) PTTs


15


_I~l~s_ ~









used in the locational accuracy tests are denoted by a leading 'X' in the identification field; this
includes tags deployed on captive animals in the acclimation pens along the NASA causeway.
(c) Animals tagged by the FDEP (formerly FDNR) on Florida's West coast (term code = '88')
were removed from the database.

The tagging history database provides a great deal of information on the tagging
program. Each record represents one tagging event or one "take," as defined by the Marine
Mammal Protection and Endangered Species Acts, and each tag deployment is referred to as a
"tagging bout." Through May 1996, there have been 430 successful tagging or retagging events
(n = 188 for VHF, n = 242 for PTT) on the Atlantic coast (excluding the upper St. John's River),
for a median of 3.0 tagging bouts per study animal (Table 3). The maximum number of tagging
bouts for one manatee (TBC-09, "C-cow") was 39. Excluding six deployments that were still
ongoing at the end of the study period, the median duration that individual tags remained on the
subjects was 27.0 days (range, <1 578 days) for VHF tags and 70.5 days (range, <1 250 days)
for PTTs; one adult male carried a PTT past the end of the study period for a total of 282 days.
Many of the very short durations for VHF tags involved their use as interim "safety" tags until a
satellite-monitored tag could be substituted. Most consecutive tagging bouts (61.7%, n =
214/347) were separated by intervals of less than or equal to 24 hr (n = 210 intervals lasted < 1
hr), meaning that the durations of continuous tracking bouts for individual manatees were
considerably longer than indicated by the durations that particular tags remained on an animal.
Continuous tracking bout durations were highly variable, lasting up to 2.7 years for one
individual (TFP-02, "Ross"). Some manatees were tracked intermittently over long periods of
time; adult female TBC-24 ("Betty"), for example, was tracked for 92% of the time over a 6.1-
year period! The median total duration that study subjects were tracked was about 7 mos., but
this ranged from 2 days to 6.8 yrs (Table 3). The median interval between continuous tracking
bouts was 71.8 days but this was highly variable, ranging up to 2861 days (7.8 years) (Table 3).

The most common reason for the termination of the tagging bout was to replace the
transmitter with another one; this was done for a number of reasons, including low battery life,
electronic or physical malfunction, and simply to replace a VHF tag with a PTT (Table 4). The
next most common cause of termination was the detachment of the tether at the designed weak
link, presumably after the tag had become snagged on something. Occasionally, humans or
alligators were responsible for the tag detachment. Twenty of the tagging bouts were terminated
by boat strikes to the transmitter housing, the tether or the belt. Seven other bouts ended with
the death of the tagged manatee, but none of these deaths were attributable to the presence of the
radio-tag or attachment system.

Over the 10-year time span of this research effort, 83 manatees along the Atlantic coast
carried VHF tags for 10,264 days and PTT tags for 20,547 days, for a grand total of 30,812
days, equivalent to 84.4 animal-years (Table 5). Figure 5 illustrates the pronounced shift from
predominantly field-based VHF tracking during the first half of the study to predominantly
satellite-based tracking during the second half (see also Table 5). Note that manatees carrying
PTTs, however, continued to be tracked in the field from the VHF transmitters incorporated in


16











Table 3. Descriptive statistics on the number of tagging bouts per manatee, tagging bout
duration, duration of intervals between tracking bouts, and total number of days tracked
per individual. Data are included for both VHF and PTT transmitters deployed along the
Atlantic coast of Florida and Georgia over the period of May 1986 to May 1996. All
durations are in days.


Mean Median Std Dev Min


Max Sum N


No. Tagging Bouts1


Tagging Bout Duration1'2

Duration of Intervals
between Tracking Bouts3


Total No. Days Tracked4


5.2


71.2



167.7


3.0


51.7


6.9


1


69.5 < 1


71.8 309.8


371.2 215.7 449.3


39 430 83

578 30812 424


1 2861 22298 133


2 2475 30812 83


1 A tagging bout refers to a single transmitter deployment.
2 Tagging bouts (n = 6) that were ongoing at the end of the study period (31 May 1996) were
excluded from these analyses (except for the calculation of Sum, where n = 430).
3 A tracking bout refers to one or more consecutive tagging bouts, separated by no more than 24
hours; 98% (210/214) of tagging bouts within a tracking bout were separated by < 1 hr.
4 The total number of days that a manatee had a functioning radio-tag.


17


_ I~ ~I











Table 4. Causes of termination of 424 radio-tagging bouts along the Atlantic coast of Florida
and Georgia over the period of May 1986 to May 1996, grouped into eight major
categories (see Appendix 4b for specific categories). Frequency and percent of
occurrence and the median and maximum tagging bout duration (in days) are listed for
each termination cause. Data for both VHF and PTT transmitters are included.


Cause of Tagging Bout Termination


N


Tagging Bout Duration
% Median Maximum


Replaced Transmitter

Replaced Belt and Transmitter

Recovered Transmitter:
Tether or Hardware Broke

Recovered Belt and Transmitter:
Belt Broke

Removed Belt and Transmitter


Lost Transmitter

Manatee Rescued


209

10


138


12


8


34


6


Recovered Dead Manatee with Tag


7


49.3

2.4


32.5


2.8


64.1

53.2


41.3


85.1


1.9 32.5


8.0

1.4


54.6

70.6


1.7 77.5


18


578

121


192


192

250

172

130

134























4,.) 0 '


C. M r






0


S0






So Q4

C, 3 4
co S


-o >- E >.




02 0 o




.0 E





0-
CZts to













o 3 0 -
) .0 0
oo c o
'3 e 3 ""





*a ME

dE
* -b S .







^^^i


ON
C0
O\


o\



o\



CT\


-o


6.
t4-

I-
4)


cl


OC


ON
00




ON
0\



0\
oo







00
cl
cO


00

cO
oO
CT\

^0


S00
00 00 6


o00


Cl
t,-i


-0 --
- Cl 00 m
00 -o


0
ON 0 ON
- -t 00
Cl


0

cl4
I-


d
(M^m
m


oo
Clb~
C14 -- re)
00





Co 00 00



Ir,




CM
~*^ 0 ON
C 4- -



o 0
0 T- tn


o oo
cCl


00






-^ 00



'0 00s








4; V3)


E-4
W5













co
s-
Cd
















cd 0
|.oo E2













-00
'o







66F
szzF
rt >
I
o s 5

S S


o0
s g-i


0
00


o ON
' C) CIr


tnc~


O




00 -


Cl


ON
- If-b



I-


00 Itr



00
- 0

\oo
C)


0
^0 00


N
Cl
cl
Cl


00
0
ON



00
(N


m
-,


0\



oo
ON






Ti
00











6
00


r-
C-
00


"tt











N
C0
00

r-
0
00

If-,

Cl



00
6
>k>


\O


-
ct


en






UC
F-




00




co
r-.
Q
U,
4)
4)
I

6


*

M .2 0

Cd
coF~






00>
zzj


00
If-
00"



0


cl







oo

0)


0
a

4p

4)

0
B


0

-o
v4
0
-
,0
C.)
4)


0 0



















e-
00




o o
'Cl




00
C)








0
>- m


or











0
uO ee


h-












'P Eg
>. Q~o.

TC9 ^^
00 -
i2 as
0- ^4

2" 0s N


_Il~s_~_~~_^_~ __~~_~ ~___I


















0


C\



C\
0\
H/"


ON


m
0\
<^


o
0

4


Clo





OC
oN
O\



oN
oN


O\
00 (c
ON


00
00




00

ON

00
ON




Uo





Q
^Q







r.
cc
CU




-0-


4-.0000

cC U


in 'S0



Pz
5
3 S


00 0
"-' 00 '-


00


m
,I-
Cl4


C4 C4

c~l Cl

m
en





0
C14 01)
l ON
C4 00



C14
00 Cl
^- (l
C14

C14
C M~

3 \


0\
cu
ON


00
oo
0



0o


Ul
CU
-O

03





Pi?
CU


00
00

CR
rr)


C+

4-




0)
CUJ

U) 0
d


C1f

5N


ON


R-


t-
0











0
4-
0



4-i







Ho

6?
cz


J3
>4-
cn
CUU)

-o 0
t:^




[IC
SHo
0


00

Cl\


I I I
i I


00
00

00
0


C)

ON
m









00

r'


Cn
I'.
Cl


c\l
ON


CI


6
o
00


0
0
0


0

0


0

0


0

0


0
6
0
0C


0




C;


00



00


rcl
o'
0
ON


00

Ol




ON


00
Cl
o\


0
0


t





Cl4


00





00


4o



Y
<-*
C3
0c

4o









0
G

o










0
o








t5
a











0
o
to
















o

0
L4.
o
CU




m


E
2



























0..
C3
o
U)
U)







S



o
CU
0
Un







i3
4)

30



-4)




a
r0





4a
0
U)

4)
00









cn
4)
4-






a
U)




I-









M
CU
4)







0
4)








0
0
-o
4)

-o



0
U)


X) CU 00
aj








o .s .2
0
















4)C*
Ci
IcCtltf
.e o~




















0 00
0a


















~0 CZ
.c







CU -u -o





rn





00 00


Q- 1
Q









cd M 03








00
CU -o

















CZ
u 00
N iCO CU -
















M Cn C
0
o^ o o )
4)4)






00,



2 % c)
o -
*4- Pi0











-- 4)x
0 Ij




oo t: o .













.=X- U)4)


Q- *0
to .5 020

- C. a
U) 0 C
4) 0 O
a- U.-













S0 ^ 0
* -4
O! I 2 ^




"n5 _] n
* -l-i


0
4-*


Cl


'-' Cl "




















Sampling Effort by Year and Tag Type


0
6
z


1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996
Year of Study


Figure 5. Radio-tracking effort by year and tag type, measured in terms
number of days manatees carried functioning radio-tags along the
May 1986 to May 1996.


of the cumulative
Atlantic coast from


21


1 I~L1..---.-~l_~____11_111~









SI


TB-01 -
TPE-04 -
TPE-03 -
TPE-02 -
TPE-01 -
TNC-12
TNC-11 -
TNC-10 -
TNC-0 -
TNC-08 -
TNC-07
TNC-06 -
TNC-05
TNC-04 -
TNC-03 -
TNC-02
TNC-1-
TMI-03
TM-02
TM-01


TJX-03 -
TJX-02
TJX-01 -
TGA-04 -
TGA-03
TGA-2 -
TGA-01
TFP-06
TFP-o0
TFP-04 -
TFP-03 -
TFP-02
TFP-01
TFK-03
TFK-02
TFK-01
TBC-46
TBC-47
TBC-46
TBC-45
TBC-43
TBC-41
TBC-40
TBC-39
TBC-38
TBC-37
TBC-38
TBC-335
TBC-34
TBC-33
TBC-32
TBC-31 -
TC-30
TBC-29
TBC-28
TBC-27
TBC-28 -
TBC-25 -
TBC-2 -
TBC-23
TBC-22
TBC-21
TBC-20
TBC-19
TBC-18 -
TBC-17
TBC-16
TBC-1S
TBC-14
TBC-13
TBC-12
TBC-11 -
TBC-10 -
TBC-0 -
TBC-08
TBC-07
TBC-06
TBC-05
TBC-04
TBC-03
Ta-c




TBC-02
TBC 01
TS-0


-*1
-----I l i
--_--,_|__


'' __


I


~-I El I --


U


-uI I-


M
-I


S -I
-I


II


U~111 -


.----


-H U

I


I


I

~ IL


----


-I


I I N I -


- im -i .C I I -


JAN86 JAN87 JAN88 JAN89 JAN90 JAN91 JAN92 JAN93


JAN94 JAN95 JAN96


DATE


Figure 6. Radio-tracking records of Atlantic coast manatees by tagtype (PTT = ) VHF = .)
from 1986 through 1995.


ur


0
z

m





Z
I-


--I
.M A1


.4
- -I --


II ~ I ~U*


-I


U5--


I I I


*


I-


I


-- ~e


I

















B-












i--

--I
-U
*


I


--liil


1


I


I
I


M i


iHM


* : I


i ---


- ---
i


9 -


iI I









the PTT canisters. Between 18 and 26 manatees were tracked in each full year of the study
(1987-1995) (Table 5). Many individuals were tracked over multiple years, as illustrated in
Figure 6. This is also reflected in the generally declining number of "new" manatees tracked
each year (Table 5).


3. Field-monitored VHF Radio-telemetry Database
VHF location records of tagged manatees (including belt-only sightings) were entered
into ASCII files from data recorded on the field tracking maps, and the locations were digitized
on 1:24,000-scale USGS quadrangle maps. The data were then imported into SAS and subjected
to a series of error-scans; after correcting the errors revealed by the scans, two permanent SAS
databases were created (see below). These SAS databases contain information on the following
variables: manatee identification number, manatee name, date, local time (EST or EDT),
precision of location (PCODE = visual point location, triangulation, or general area polygon),
number of digitized points (>1 for polygons), type of observation (staff or public), duration
manatee spent at the digitized location, manatee group size, presence of calf, tagged manatee
activity, quadrangle name, and location (UTM coordinates for zone 17 in the NAD-27 datum).
Continuous observations of a tagged animal that changed locations were indicated in the PCODE
field as a first location (e.g., 'PF'), optional center locations) (e.g., 'PC'), and a last location
(e.g., 'PL').

Implementation of Database Quality Control Procedures: Computerized Error-scanning
Three SAS programs were employed to check for the following types of logical errors in
the VHF ASCII and SAS databases: incorrect formatting (e.g., missing space between variable
values), duplicate records, more than one ID number or name for a given manatee, inappropriate
or extreme values for the variables, incorrect or incomplete sequence of PCODES for continuous
observations, and unreasonable locations (as indicated from a crude X-Y plot of UTM
coordinates). After an update dataset had been scanned and corrected, it was merged with the
main SAS database and the error-scans were run again. Other types of errors, such as incorrect
manatee IDs or dates, were more difficult to detect. We are in the process of developing a new
SAS routine to detect these errors by comparing the dates and times of locations in the VHF
telemetry database to the starting and ending dates and times of tagging bouts in the tagging
history lookup table. Visual inspection of latitude versus date plots showing annual movement
patterns (see Results) has also been useful in identifying potential errors that were not detected
with the computerized scans.

Processing of General Area Locations
Each general area location was composed of several records (n = 2 to 33) in the original
database (atlvhfrw.sd2), each corresponding to a vertex of the estimated polygon. To make
these observations compatible with the rest of the point locations in the database, a centroid (ie.,
visual center) was calculated for each general area polygon with a SAS routine; when the
number of points was only two, (used occasionally for non-visual locations in narrow canals),
the mid-point of the line was taken as the estimated point location. The raw general area data


23


_~P~









were then saved to a separate database (atlvhfar.sd2) and the centroid data were merged back
with the visual locations and triangulations to create the principal VHF point database
(atlvhfpt.sd2). The following variables were added to each general area location in the VHF
point database to provide an index of the relative accuracy of the estimated centroid: Area in
hectares of the location polygon; mean distance (m) of the centroid to the polygon vertices; and
maximum distance (m) of the centroid to a polygon vertex.

Problems and Improvements in Data Collection and Data Entry Protocols
I identified a number of problems with the attribute data in the VHF telemetry database
that resulted from inconsistencies in data collection and data entry protocols. Some of these
problems stemmed from a lack of clear, standardized definitions or criteria. Four variables with
such inconsistencies are discussed below. These problems have either been rectified or are not
serious in terms of their effect on the overall integrity value of the VHF database.

Manatee Activity
A list of manatee activity categories and their corresponding database codes were
established at the beginning of the.telemetry project, but as the study progressed more activities
were added and the meanings of certain terms diverged across the multiple observers. So while
recorded activities were consistent within observers, there were sometimes disparities among
observers in what the activity categories meant. By consensus among field staff and with input
from the literature (Hartman 1979, Bengtson 1981, Zoodsma 1991), we standardized and clearly
defined activity categories and codes for radio-tracking data collection and data entry in late
1994. Appendix 5 provides the complete list of activities and their definitions; seven new
categories were added (four being subcategories of two former ones) and three categories were
deleted. This effort eliminated inconsistencies since that time, but we still had to deal with past
problems in the database. Observations with activities that had been interpreted variously in the
past were verified against field tracking maps and the activity codes were modified, as
necessary, to standardize usage. This lengthy process has resolved most of the problems with
the activity variable. In addition, the size of the activity field in the database was changed to
permit up to three activities to be coded for a given location (previously it was limited to only
one activity).

The following modifications to activity codes in the VHF database have been completed:
(a) The code "Multiple" was deleted and replaced with up to three characters denoting the
sequential order of observed activities; (b) the code "Other" was replaced with the appropriate
code in current use (e.g., categories such as "Socialize" were at one time coded as "Other") and it
now indicates a tagging event or attempt by a researcher; (c) the category for a human (non-
researcher) feeding a tagged manatee was changed from "Feed" to "Human Feed"; (d) the most
problematic activity category, "Idle", was deleted from usage and replaced with either "Mill",
"Rest", or "Slow Travel", depending on the description of the activity on the tracking map and
on the identity of the observer; (e) the seldom-used category "Loiter" was deleted and replaced
with "Mill"; (f) a new category "Mating Herd" was added by checking all tracking maps with
activity codes of "Other", "Multiple", "Socialize" or "Cavort" for possible mating activity; and


24


_I_ _









(g) activity categories used by B. Zoodsma in her thesis research (Zoodsma 1991) were
converted into Sirenia Project codes prior to merging her Georgia VHF data with the principal
database.

Use of habitat and home range will be analyzed as a function of gross activity categories
that combine some of the finer divisions noted in Appendix 5. For example, resting will include
bottom rest and surface rest activities, travelling will include slow and fast travel, and socializing
will include cavorting and mating. More specific questions could also be addressed, such as
where manatees go to drink freshwater and how they respond (in terms of movements) to
tagging interactions with researchers.

Precision Code
The PCODE variable gives information on both the relative precision of the location and
on whether it was an isolated sighting or one of a continuous series of locations. There have
been no clear criteria for deciding between the latter two possibilities when two or more
locations are recorded for the same manatee on a given day by the same observer. Generally, if
the manatee was under continuous observation, the locations are listed as PF, PC ... PC, PL.
This is a minor problem because the times of the locations will be used when determining their
independence for statistical purposes (e.g., home range calculation). Nevertheless, I recommend
that a series of locations be indicated as continuous only if there are no gaps greater than 1 hr
duration in observation.

Duration at Location
This variable indicates the number of minutes that the researcher observed the tagged
manatee at the digitized location. There has been some variation among data entry staff over the
years as to how duration was calculated when applied to a continuous series of locations,
however, including the following methods: the entire observation duration given for the first
location and missing durations for subsequent locations; the duration of the interval between
consecutive points given for each location; and the actual duration the animal spent at each
location, which may have been minimal (or not recorded) for a travelling manatee. The latter
method was the intended one. Without referring back to the 531 original maps that tracked
animals over two or more locations, the inconsistency in calculation method makes this variable
of little value in this context. The variable should be accurate for the vast majority (85.7%) of
the records in the database, however, which consist of single locations.

Group Size
Group size in the SAS database is equal to the number of manatees in a group, including
the focal tagged manatee but excluding its calf For example, the group size of a single manatee
or a solitary cow-calf pair is listed as one. The operational criteria for delineating a group,
however, were not clearly defined; some observers counted manatees within a distance of
approximately 10 m of the focal animal as comprising the group, whereas others estimated the
number of manatees "in the general vicinity" (e.g., in the same residential canal). While neither
of these definitions is wrong, the lack of standardization reduces the value of this variable;


25









perhaps it can be used to indicate whether the tagged manatee was solitary, in a small group, or
in a large group (e.g., a power plant aggregation). I recommend the following definition of a
manatee "group" in this application: the number of manatees within 10 m of the focal animal
plus-using the "chain rule"-the number of additional manatees within 10 m of them, and so
on; this would allow most of a large winter aggregation to be counted as part of the tagged
manatee's group, even though it might be spread out over an area 50 m across.

Sources of Error in VHF Locations
There are three principal sources of error that contribute to the accuracy of a visual
sighting of a tagged manatee. (1) The charts and maps used in the field have a small degree of
inherent error, as well as error associated with natural or human-induced changes that have
occurred since the original cartographic work was conducted. Furthermore, the coastline in the
USGS quadrangle maps (used for digitizing) may not be completely concordant with the NOAA
charts (from which the GIS coverage was digitized). There is little that we can do to reduce
these errors, but we assume that they are negligible compared to other sources of error. (2) Error
in plotting the animal's location on the map in the field is unknown but probably varies
considerably with the manatee's distance offshore, the proximity to landmarks (e.g., islands,
causeways) identified on the tracking maps, the observer's familiarity with the area, and other
factors. (3) Digitizing error is the third component of the overall locational error. A preliminary
assessment of this error source yielded the following results: (a) four manatee locations that
were redigitized differed by an average Euclidean distance of 149 m; and (b) ten locations
digitized on hardcopy quadrangle maps (i.e., the current method) and on-screen in ArcView
(over a digitized NOAA shoreline coverage) yielded UTM coordinates that differed by an
average of about 20-25 m (equivalent to 1 mm on a 1:24,000-scale quad map) (see Appendix 6
for details). The precision indicated by the first result was not satisfactory, but the sample size
needs to be greatly increased before any conclusion can be made. The latter result suggests that
the locations may be digitized on-screen in a GIS with approximately equal precision and
accuracy as on a large-scale hard-copy map.

Current Status of the VHF Telemetry Database
A large backlog of tagged manatee locations has been digitized and processed up through
31 May 1996. There are still three major checks that needs to be completed before verification
of the VHF database is complete: (1) confirmation that all VHF sightings (except belt-only
observations) occurred within the tagging bouts specified in the tagging history lookup table; (2)
identification and addition of researcher tagging events that were omitted from the database; and
(3) comparison of the IDs, dates and times of locations in a printout of the entire database
against the tracking maps to search for typos in these fields. All sighting times will then be
converted to Eastern Standard Time. Numerous sightings by the public have not yet been
plotted onto tracking maps and digitized, but given the lower reliability of these data and the
relatively low benefit-to-cost ratio involved, this task has been deemed a low priority.

We gathered a total of 11,810 VHF locations on 8,586 different tag-days for 83 manatees
over the entire study period (Table 5). Tagged manatees were located on 28% of total tag-days


26









(range across years, 9.2 45.9%, Table 5), for an average of one location every 3.6 days (range
across years, 2.2 10.9 days). (Note that this average excludes multiple sightings on the same
day.) Field effort was necessarily greater during the first half of the study (1986-1991), as
indicated by the higher sighting rate (39.5% of tag-days, or one location every 2.5 days), when
primarily VHF tags were used, compared to the latter half of the study (1992-1996) when greater
reliance on satellite-monitoring was made (16.4% of tag-days, or one location every 6.1 days).
For 75.3% of the location fixes, manatees were visually sighted after "homing in" on the VHF
signal (Table 5); the rest of the locations were either determined through triangulation (7.5%) or
estimated as general area polygons (17.2%). The visual sightings included 1250 locations
(10.6%) provided by the public and verified by Project staff.


4. Satellite-monitored PTT Radio-telemetry Database
Location and sensor data for each PTT were obtained from Service Argos on a monthly
basis via nine-track tapes (Dec. 1986 Sept. 1989) or diskettes (since Oct. 1989). SAS programs
were used to extract the locational data from the raw Argos DS (dispose) files (ASCII format),
to scan for formatting errors and duplicate records, to flag putative outliers, and to create a
permanent PTT telemetry database. The tagging history lookup table (see above) was accessed
by the SAS program when building the location database in order to match the PTT record to the
correct manatee name and identification (ID) number, based on PTT ID number, date and time.
Locations from non-deployed or detached PTTs were removed and archived in a separate
dataset. Locations were converted from the WGS-84 datum to the NAD-27 datum and projected
from geodetic coordinates (latitude and longitude) to Universal Transverse Mercator (UTM,
zone 17) coordinates using the Geographic Calculator software (Blue Marble Geographics,
Gardiner, ME) in order to be consistent with the mapping system used for the Florida shoreline
and other existing GIS coverages. The current PTT telemetry database includes data on the
following variables: manatee ID number, manatee name, PTT ID number, date, time (EST),
tagging bout number (from the lookup table), latitude and longitude in decimal degrees (WGS-
84 datum), UTM x (Easting) and y (Northing) coordinates (zone 17, NAD-27 datum), and
location class. Location class zero and other non-guaranteed locations were included in the
database for some years and not for others (Table 5), depending on how Argos provided these
data.

Prior to January 1991 the PTT sensor data were extracted from the Argos DS files,
processed to yield only one observation per location (i.e., one record per satellite pass, since the
raw data give one record for each transmitted message), and stored as a separate ASCII data file.
This sensor database included the following information: PTT ID number, date, time (which
typically differed slightly from time in the location database), number of messages received,
median PTT temperature (stored in a coded form that must be converted to degrees Celsius with
PTT-specific regression equations), short-term activity index (modal no. minutes PTT tipped in
a 1-hr period, max. = 60), and long-term activity index (modal no. times PTT tipped in a 12-hr
period, max. = 1023). These and other sensor data (see Appendix 2) were still collected


27


~~_~~~1_11









subsequent to January 1991, but they were not processed into one record per satellite pass and
they were left in the raw Argos DS files.

Implementation of Database Quality Control Procedures
The PTT database had fewer quality control problems than the VHF database because it
did not require as much of a human element in its creation. There were two principal types of
potential error that differed in their origin and consequences. The first type of error was
introduced by mistakes in the start and end times of tagging bouts listed in the tagging history
lookup table. This resulted in the inclusion of locations from detached (presumably drifting,
stranded or recovered) PTTs in the telemetry database and, occasionally, in the exclusion of
valid locations from tagged manatees. The lookup table was scrutinized for errors, as described
above, and has been verified for accuracy. Therefore, this source of error has been eliminated or
at least minimized. The second type of error was inherent to the Argos system of PTT location
determination. Locations were reported by Argos to the nearest 1/1000th of a degree latitude
and longitude (approximately 100 m) but the locational error followed a probability distribution
with a variance determined by the assigned location class (see section on Location Quality
above). Potential outliers were identified as those lying outside a series of latitude-longitude
windows (Fig. 7). Different series of these windows were used for tagged manatees in the
Florida Keys, the Atlantic coast, and the upper St. John's River regions. Locations that were
wildly wrong (e.g., southern hemisphere) were immediately deleted, while those in the vicinity
of the study area were checked manually against other PTT and VHF locations to determine
whether the locations were plausible for the individuals and dates in question. Putative outliers
were then either verified as valid or deleted and archived in a separate database. Sometimes the
alternate locational solution (on the opposite side of the satellite's ground track) was plausible
and was substituted for the given location. Nearly all of the outliers were of non-guaranteed
location classes (0, A, B), although a few location classes 1 and 2 were beyond acceptable
bounds. Locations within the defined latitude-longitude windows might also be considered
"outliers" for certain types of spatial analyses, such as calculation of travel rates or use of home
range. We opted to keep such locations as they may be useful in delineating annual movement
patterns (e.g., rapid migratory movements often result in poor location classes); the database can
be trimmed further depending on the requirements of the particular spatial analyses being
conducted.

Current Status of the PTT Telemetry Database
The verified satellite telemetry database now spans the period from 1 December 1986
through 30 September 1995. It will be updated through at least May 1996 by the end of the
year. Additional locational data on five manatees along Florida's east coast were collected by
satellite earlier in 1986 (two in May to June, three in September to November), but they were
not included in the database because of compatibility problems associated with changes in Argos
format. All but one of these manatees (TBC-02, "Trixie") were radio-tracked again during the
period included in the database.


28








GA


N


E


S


BLUE w
SPRING


EAST
COAST




0 50 100 Miles



FLORIDA
KEYS

g 7.Latude-ongitude window utative ouiers inthe teleme database
during SAS data processmg. Different setsofwindows
and Florida Keys manatees.









A total of 52,811 "guaranteed" locations for 61 manatees were obtained through Service
Argos on 17,579 tag-days over the nine-year period represented in the PTT database (Table 5,
Fig. 6). At least one "guaranteed" location for each deployed PTT was obtained on 93% of all
tag-days pooled (Table 5). An average of 3.0 "guaranteed" locations per day were acquired for
each manatee carrying a satellite-monitored transmitter. Locations were spread approximately
equally across four 3-hr time periods representing the programmed duty cycles and the timing of
satellite passes over Florida (Fig. 8).


30












PTT Locations vs. Time of Day


7000 -


6000-


5000-




0
0 4000-
0

S3000-
'%-
0
6
Z 2000-


1000-


0 --I i I I I I I I I
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23
Time of Day (hr)


Figure 8. Number of satellite-determined manatee locations (n = 57,688) as a function of time
of day along the Atlantic coast of Florida and Georgia from December 1986 to
September 1995. Each time of day value includes the 60 minutes following the start of
the hour (e.g., 11 refers to the period from 1100 1159 hr).


31









PRELIMINARY RESULTS AND DISCUSSION


Most of the work to date has focused on the development, quality control, and qualitative
and quantitative descriptions of the radio-telemetry and associated databases (Deutsch et al.
1995, 1996). What follows are the results from some preliminary analyses of these data through
September 1995 (see also Reid and O'Shea 1989, Sirenia Project 1993, Reid et al. 1995). The
first section characterizes the annual movement patterns and migratory behavior of radio-tagged
manatees along the Atlantic coast. Then site fidelity to warm-season home ranges and diel
movement patterns are briefly discussed. Finally, a case study of the how the telemetry data
have been applied to address issues of manatee management is presented.

Annual Movement Patterns
The annual patterns of movement were determined for 41 radio-tagged manatees; data
were insufficient to ascertain patterns for the remaining study animals. These long-term
movement patterns are described according to the occurrence (presence/absence) of seasonal
migrations, the geographic regions encompassed by long-distance movements, the distance
between winter and summer ranges, and the timing of seasonal movements in relation to
temperature change.

Occurrence of Seasonal Migrations
At a gross level, manatee movement patterns could be categorized into two basic types,
year-round residents and seasonal migrants. Year-round residents were defined as those animals
remaining within the sanie 50-km stretch of coastline all year-long, as illustrated in Figure 9 for
a subadult male in Brevard County. They usually moved to local power plant discharges during
winter cold periods. Figure 10 shows the north-south movements of a typical seasonal migrant
over the annual cycle. Migratory manatees exhibited the following general pattern of
movement: they departed their warm-season range and travelled south upon the onset of cold
fronts in late fall and early winter; overwintered in southeast Florida in the vicinity of power
plant thermal discharges or in rivers, residential canals and bays; then migrated back north as air
and water temperatures moderated during late winter and early spring. These observations are
concordant with those obtained from photo-identification studies (Reid et al. 1991) and aerial
surveys (e.g., Provancha and Provancha 1988, Weigle et al. 1987) along Florida's east coast.
Travel between summer and winter ranges was typically direct and rapid, ranging from 20 to 50
km per day of travel. The southbound trip from Brevard County to southeast Florida
(approximately 200-300 km) lasted a median of 10 days (range = 5 66 days, n = 34); this
included a median of 6.5 days (range = 5 13 days, n = 22) of travelling plus brief stopovers
along the way. Most migratory movements occurred in the relatively narrow Intracoastal
Waterway (ICW), but occasionally manatees travelled close to shore in the ocean between inlets.
Of the 41 manatees for which sufficient data on annual movements were available, 85% (n =
35) migrated seasonally and 15% (n = 6) remained resident in a given region year-round.


32


1 _1~~1_1~~~








<" -0
0 coE
E> E
,, rn *-




C, \




0 0
U)0
aw I




*---

F--





co

m E .




( up .
oo

c-
0 h



D a,



SI 0 0 0 %#






C Cl 0 CO
CO CO CO CO CO CO C) C) )


(wj) G1iu!pjooo A vain









< "(3
0 cu
I> E
ciu

Q. Q.










I------------------- (D^ -
Z,,




C\l
E cL


0
U-







-4-.
Eco,

S- *-C -


cz
0- ,_


oa 'c



c:
Eu








<0





cz
Lc u 0..

C=) C'\




(LI 0 CP A 0


5 t c't








(UIuY) eGuipJroo3 A 1/Min









Variation in Annual Movement Patterns: Geographic Region and Migratory Distance
Variation in annual movement patterns was more complex, however, than indicated by a
simple migrant-resident dichotomy. There was considerable variation in the destinations of
seasonal migrations, including four main types. The most common migration pattern, shown by
21 (60%) of the 35 migratory manatees with good tracking records, was from central Florida in
summer to south Florida in winter (Fig. 11). Six (17%) manatees exhibited a similar pattern but
shifted to the North, moving to southeast Georgia in the spring and summer months and
overwintering in Brevard County (Fig. 11). Manatees showing these two migratory patterns
were often simultaneously present in Brevard County during the spring and fall (see Fig. 11).
Five (14%) study animals made short-distance migrations (50-100 km) within a given
geographic region, while three others (9%) travelled the entire length of Florida's east coast and
into Georgia, a straight-line distance of nearly 600 km (Fig. 12; see also Fig. 6 in Reid et al.
1995). These long-distance migrants often spent considerable time in Brevard County during
spring or fall.

The latitudinal distance between "summer" (Apr. to Oct.) and winter (Dec. to Feb.)
ranges was highly variable, ranging from 15 to 550 km (Fig. 13). The bimodal frequency
distribution highlights two groups: (1) year-round residents and short-distance migrants (< 100
km), and (2) migrants that travelled about half the length of Florida's Atlantic coast (200-300
km). The median migratory distance for 26 wild-caught manatees was 250 km. This distance
was similar for eight manatees rehabilitated in captivity as adults (median = 190 km), but three
animals rehabilitated as calves (and released as subadults) showed little seasonal movement
(median = 20 km). Body size and age could not account for the observed variation. Migration
distance was not significantly correlated with body length among wild-caught adult females (r =
-0.09, P > 0.50, n = 18; including rehabilitated adult females: r = -0.04, P > 0.50, n = 24).
There was also no significant difference in migration distance between wild-caught adults (n =
20) and subadults (n = 6) (Wilcoxen two-sample test, P > 0.50).

The migratory movements of one of the study animals not included in the above analyses
deserves special mention. TBC-42, known to the public as "Chessie," was a large adult male
manatee that travelled over 2000 km (straight-line distance along coast) from Florida to Rhode
Island, breaking scientific records for the most northerly location and for the longest migration
of a West Indian manatee (Fig. 14) (Reid 1995). Fearing that he would succumb to cold stress
from the approaching winter, the U.S. Fish and Wildlife Service captured Chessie on 1 October
1994 in the upper Chesapeake Bay, Virginia-where he had been sighted since late
summer-and transported him to Florida on a U.S. Coast Guard C-130 aircraft. Chessie was
radio-tagged and released in the Banana River, Brevard County and travelled to Port Everglades
in southeast Florida, where he spent much of the winter. He returned to the Banana River by
late February 1995 and reached the lower St. John's River by early May. All of these areas are
heavily utilized by manatees, indicating that Chessie was probably quite familiar with important
manatee habitats along the east coast of Florida. In mid-June 1995 he launched into his two-
month-long northward journey, as revealed by Argos locations; he typically covered 40 to 50 km
per day, rarely lingering more than one day in a given area (Fig. 14). Chessie briefly entered the


35


~I_-_---~-~










Cu

m


-I --
L g -- -- -- _^










^-^^- i --r
C'- ^- ^-^^ __ .(


0


0
CN


0
0
CN


0 0 0 0 0 0

( C C) Ce Cu C

(wu>) eieulpjooo A vv n


-C')



Cl
--'C
w


-







<








- -


CE


--


0)
0
(D


0


4-0

0




Ei





ocu
LCI



co



0l -
-C)







0M
,0 I>






C O
|i|I



C''U-
E .9 -
Eo

O CD


a,
0)
Lo


I














(1) V5
o c)

--
as



0' ,
O

S 0


cm

SC

- CI



( 0 .-
~----- OO






alOc-




0
Sx m.o
0 I 0- 0 -0 c

\I as -- a
----0 I .-





2 1.


I E
O o ED0

I---m--13------O-V---- cq


(uwl) emu!pJooO A V1/Ln




















































SI I I I I I I I I I


tO


0


iO


sIfnp!AipuI Jo *ON


-o



Y



D U


4-4


a
(



*


0)

0
0
o
u,


0
0
0




0
o



c)


N


r

c0
0









0)
0)
0
(M






CS


E






d





r/3
a
bo
















C~
.g


Td
(D
C1-<


E


*^






a


0
Ns


0


COM


-0






4a



*SI

=
cc
SE




Q U
" o
O10

















1V

C
. E
c o




00
*^0
=) V) .
JD i -
*E 0 *
^S "-





i- ro (








N


E


S


I


16 August


7 August

24 July
4 July


30 June


26 June


20 June


15 June
May 5 Jun 12


Feb21 Apr23

Jan 12 Feb 15


0


- -


500 Kiloreters


Figure 14. Northward migration of adult male TBC-42 ("Chessie") along the Atlantic seaboard from
Florida to Rhode Island in 1995. He returned to Jacksonville, Florida by mid-Noveber.
Only a subset of PTT locations and dates are shown here.


500









lower Chesapeake Bay in early July 1995 and continued north through New York City and into
Long Island Sound by early August. He finally reversed his northward movement on 16 August
at Pt. Judith, Rhode Island, where he encountered water temperatures of 19-20 C, the coldest of
the trip. The tag detached about one week later off New Haven, Connecticut. This continuous
tracking bout lasted 222 days from retagging in Ft. Lauderdale in January. Chessie's journey
from north Florida to Rhode Island took 64 days and ranged along the coasts of 11 states.
Although the duration, distance and destination of this migratory move were certainly extreme
compared to the rest of the Florida manatee population, Chessie's activity and habitat use
appeared typical for a manatee in more temperate waters, such as those found in north Florida
and Georgia. He chose salt marsh and estuarine habitats for resting and feeding, and travelled
primarily in the ICW or in shallow lagoons and sounds inshore of barrier islands. Furthermore,
manatees are sighted regularly and with increasing frequency during the warm months along the
southeast coast of the United States, including Virginia and the Carolinas (Rathbun et al. 1982,
Schwartz 1995), indicating that Chessie is not an isolated wanderer. Public sighting reports
allowed us to document Chessie's return trip south to Jacksonville, Florida by mid-November.
This manatee Olympian has taught us what manatees are capable of achieving: a round-trip of
well over 3500 km between north Florida and New England in five months; and an average rate
of travel of 50 km/day sustained for 3 weeks over about 1000 km of coastline from southern
Georgia to the Chesapeake Bay.

Timing of Seasonal Migrations
The timing of the fall migration and its relationship to temperature change was examined
for radio-tagged manatees in Brevard County waters. Visual inspection of graphs showing
changes in latitudinal movements and water temperature (in Banana Creek) over time strongly
suggests that manatees initiated migratory moves in response to the passage of strong cold fronts
in late fall or early winter (Fig. 15). There appears to be variation among manatees in the
threshold temperature at which this response occurs, however. The mean date at which
manatees initiated their southward migration from their warm-season range was 8 December, but
there was considerable variability in departure date among individuals (early November to late
January) and across years. Some of the individual variation may be related to differences in
"cold-tolerance" and some to differential use of the Brevard County power plant thermal
effluents. Inter-annual variation in the onset of cold weather apparently accounted for some of
the variation in migratory timing. Mean departure date from Brevard County was positively
correlated with the first date at which mean daily water temperature dropped below 17 C (r
0.76, n = 7 years) (Fig. 16). Note that data for some of the same individuals were included in
multiple years, breaking the assumption of statistical independence required for a hypothesis test
of correlation; a repeated measures type of analysis will be necessary. These findings are
generally consistent with current knowledge of manatee response to cold temperatures, but they
suggest that individual thermal tolerances may be greater than that indicated by previous
experimental studies on the lower limits of thermoneutrality (Irvine 1983).


40










(0) ejnijedwe.l JeeM


0 0
Co 1o


0


CL)
-U)
Clo


c


C0
V-


L.
cui


ILU


0











U) 0)
C )


(wuj) elu!pJooo A vfin


0
qt


0
CO
o~


a_





E
o










0
5-

















E




I-


1=
U-



a
.c











a,
"0















o
E
E





n4

0











a,
m
4-









o,


r


0_


O
0


0
CMl
CO


0
CO


0
0
CO
C'I


~II~~ I~_


V

E
au
.0
0
S
4-

o

C


0
r




fl
a,


5










.0
.4
-o
C
c:















.
ai












m
"a,
4-
C















*o












I.
0
.4-
0
C
0



a,




ia
6.


ca
2E
a,


'0




0








c







CL




.
>1
-c



I-
g




.4

6.










360


S340
,,

:3




ca
a320
2 /
/
/



/

300 _

300 320 340 360

First Date Water Temperature < 17 C


Figure 16. The mean julian date at which migratory tagged manatees departed
Brevard County in fall as a function of the first julian date that mean daily
water temperature dropped below 17 C. Means for seven years are plotted
(n=3-7 individuals per year). Temperatures were recorded in Banana Creek,
Brevard County. The dashed line indicates equality between mean departure
date and the date on which the 17 C temperature threshold was crossed.









Site Fidelity to Warm-season Range
Long-term field studies of recognizable individuals have documented strong site fidelity
to winter aggregation sites in a number of regions (O'Shea and Langtimm 1995) but there has
been relatively little known about the occurrence of site fidelity to summer ranges (Reid et al.
1991). Some radio-tagged manatees showed remarkable consistency in the extent, timing and
destination of their seasonal movements from year to year, as exemplified in a five-year tracking
record for TBC-24, Betty (Fig. 17). This adult female occupied three areas along the east coast:
the Banana River, Brevard County from early spring through late autumn; and two sites, Port
Everglades and Coral Gables, in south Florida during the winter period. I considered a manatee
to show regional site fidelity to a warm-season home range if it returned to the same 25-km
section of the coast in two or more years. Of 18 individuals with at least two years of summer
tracking data, 14 (78%) were site faithful from year to year by these criteria. Three of the other
four manatees were faithful to two widely separated core ranges in summer. Many manatees
also made brief trips outside of their principal home ranges during the warm months.

Although this has not yet been thoroughly investigated, site fidelity to a warm-season
home range was also apparent at a much finer spatial scale. Figure 18 illustrates this interannual
consistency in spatial use pattern for TBC-24 in the mid-Banana River from April through
October. Note how her movements were restricted to a relatively small portion of this water
body. This area provided excellent habitat for manatees: abundant seagrass flats on the east
side; residential canals for resting on the west side; and a reliable freshwater source (Cape
Canaveral sewage discharge) which served as a major manatee attractant. The larger area
indicated by locations in 1992 compared to the previous two years may have been related to
Betty's weaning of her calf during that summer (Fig. 18). It is confounded, however, by the fact
that the first two years relied principally on VHF sightings whereas most locational data were
acquired via satellite in 1992.


Diel Movement Patterns
The spatial distribution of radio-tagged manatees (n = 8) overwintering in northern
Biscayne Bay was examined as a function of time of day. A clear diel pattern was revealed:
manatees spent the afternoon in the Little River and nearby canals, moved downriver and into
the bay in the early evening, spent the night and the morning in the bay up to 3 km from the
mouth of the Little River, and returned to the nearshore and inland areas in the late morning
(Fig. 19). Field observations by Project staff substantiated this pattern and indicated that
manatees mostly rested or socialized in the Little River (K. Curtin, pers. comm.). At night the
manatees presumably foraged on the extensive beds of seagrass (mostly Halodule wrightii) or
algae in the middle of the bay. Bengtson (1981) described a similar diel pattern of movement
between Blue Spring run and foraging areas in the St. John's River during the winter months,
which he attributed to diurnal fluctuations in air and water temperatures. Likewise, manatees
overwintering at the warm-water springs of Crystal and Homosassa Rivers generally move
downriver to foraging areas in the late afternoon or at dusk and return to the springs by dawn


43











LO







C)
0


a J










0 c
Co




2


o9
O )

m C LU
00 r-






a




2
CO





0)
CO


0 0
Co
co
0)
- 0)


0
0
CM
cv


(wLU) etu!pjooo A v.Ln


ooE

E) 2
E w




-me
C lQ.
(D

a> C





0 (0-=
S- 0..
C M





Q I 0.
(D
Ca


e e --
Ed a



(0 .( D
*= m (
0.0



Eo
o ~ o


C 0 0 0
%M C"M 1




-0W0

2 | -Eg






aO -o
I Co
0 t.- a m
o 0(

SI m


I .



C = -a, Co




U-


0
0
CY


0
0
CO
C*


0
0
O
CV
cr


0
0
C


0
0
C\j













II
(U

00's





AFIERNOON: 1300-1600


N


EVENING: 1800-2100


2 C


)


2


4


NIGHT: 0200-0500


6


Kilometers


Figure 19. Winter distribution (Dec. Feb.) of satellite-determined locations for eight tagged manatees
in north Biscayne Bay, Florida between 1986 and 1995 as a function of time of day (EST).
Location classes 2 and 3 only(n = 487).


III


MORNING: 0700-1000









(Rathbun et al. 1990). This type of diel cycle also appeared to occur in Brevard County during
the summer months (S. Tyson, pers. comm.), indicating that factors other than temperature, such
as motor vessel traffic (Reynolds 1981) and sea state, may also be responsible for such patterns.


Management Applications
The radio-telemetry study has yielded data on seasonal variation in manatee movements
and high-use areas along the Atlantic coast which have been valuable in the development of
measures by federal, state and county managers charged with recovering the population and
protecting its habitat. Along with aerial survey and other sources of data, for example, the
telemetry information was used to help justify the designation of manatee protected areas in the
upper Banana River and in the north fork of the Sebastian River. Florida manatees have
demonstrated flexibility and opportunism in their ability to adapt to human-modified
environments and have taken advantage of manatee sanctuaries shielded from human
disturbances (O'Shea 1988).

A case study of how spatial data from tagged manatees have been analyzed to assist
managers in rule-making decisions is summarized here, with the text of the report given in
Appendix 7. An area south of the Minuteman Causeway, Cocoa Beach had been designated a
watersports zone (e.g., for water skiing) prior to 1991 and was changed at that time to a slow
speed zone for motor vessels because of consistent use by manatees in the area, as determined
from aerial surveys (L. Ruhana, pers. comm.). Earlier this year the Brevard County Commission
was considering a recommendation to reopen this area for watersports use. The Florida
Department of Environmental Protection's (FDEP) Bureau of Protected Species Management
(BPSM) requested our assistance in evaluating recent manatee use of the area encompassing
Cocoa Beach and the Thousand Islands, with special emphasis on the proposed watersports zone
(see map in Appendix 7). Special protocols were developed to integrate the spatial analysis
functions of ArcView with the data manipulation and statistical analysis capabilities of SAS.

Based on PTT data between 1992 and 1995, I found that 16 (64%) of the 25 tagged
manatees that used the Banana River were located at least once in the Cocoa Beach / Thousand
Islands area. Half of these individuals were located there on at least 20 days (maximum = 167
days). Pronounced seasonal variation in manatee use was evident, with greatest use during the
spring/early summer and fall. These time periods correspond approximately to the end of the
spring migration and the start of the fall migration, respectively, suggesting that Cocoa Beach
waters may be used as a "staging" area for migrant manatees moving into and out of the Banana
River. The area was rarely used during the winter months. The small size of the proposed
watersports zone made it more difficult to formulate firm conclusions, but it showed the same
seasonal use pattern as for the larger Cocoa area. This zone did not appear to be a hotspot of
tagged manatee use, but it is likely that manatees passed through or near it when travelling
between important foraging and resting areas to the north and south. The residential canals and
other sheltered waterways in the Ten Thousand Islands area were used by tagged manatees for


47









resting; they often moved offshore to feed on seagrass beds, a favorite foraging ground being
located along the southern edge of the Cocoa Beach area. The telemetry data could not be used
to directly assess whether manatees were swimming through the small zone of interest, because
the radio-tag is usually pulled underwater when a manatee is travelling, resulting in few high-
quality locations. Potential travel corridors can be modelled using cost-path analyses in a raster-
based GIS (Weigle and Flamm 1995), but the urgency of the request did not permit more
sophisticated analyses.


48









SUMMARY


The principal objective of this work was to analyze ten years of data on the seasonal and
short-term movements and spatial use patterns of Florida manatees along the Atlantic coast.
Locational data were collected using conventional, field-monitored VHF radio-transmitters and
satellite-monitored UHF platform transmitter terminals (PTTs). Major tasks were identified and
prioritized, and considerable progress was made in updating, error-checking and characterizing
several telemetry databases, with emphasis on improving quality control. A new database on the
attributes of individual tagged manatees was created and has proved useful in analyzing
migratory patterns as a function of sex, age class and rehabilitation status. Database quality
control procedures were improved through modifications to SAS error-scanning programs, and
databases were verified through manual proofing, as necessary. Problems and inconsistencies in
the databases were identified and resolved. For example, manatee activity codes were revised
and standardized across observers, and this may prove useful in understanding habitat use. This
phase of the analysis dealing with the development and quality control of spatial databases on
manatee distribution and movements is nearing completion. The VHF telemetry, tagged
manatee identity, and tagging history databases now encompass the period from May 1986
through May 1996. The PTT telemetry database currently covers the period from December
1986 through September 1995.

The Sirenia Project radio-tagged and tracked 83 manatees along the Atlantic coast, from
the Florida Keys to southeast Georgia, between May 1986 and May 1996. Over half (58%) of
these subjects were tagged in the Indian River lagoon system of Brevard County in central
Florida. Between 18 and 26 individuals were tracked during each full year of the study (1987-
1995). Most study animals were female (61%), adult (80%), and free-ranging (71%) at tagging.
Lengths ranged from 200 cm (for a dependent calf) to 350 cm. Most (58%) of the 24
rehabilitated manatees had been held in aquaria for less than one year (range, 6 days to 7.1
years) prior to their release and four were born in captivity. Tagged manatees were tracked for a
median duration of 7 months (range, 2 days to 6.8 years). Despite frequent tag detachments and
limited PTT battery life (about 8 months), the duration of continuous tracking bouts lasted up to
2.7 years. Such long tracking durations were made possible through replacing or reattaching
radio-tags, typically without recapturing the animal. About two-thirds of the study animals were
tracked over multiple tag deployments (median = 3.0 bouts per animal, maximum = 39). They
carried radio-tags for a total of 30,812 days (33.3% VHF tags, 66.7% PTT tags), equivalent to
84.4 animal-years, yielding over 11,000 VHF locations (on 28% of total tag-days) and over
50,000 PTT locations (on 93% of PTT tag-days). On average, 3 good quality (LC1-3) locations
per day were provided by Service Argos for manatees carrying satellite-monitored transmitters.

Analysis of manatee movement and spatial use patterns is in its initial stages.
Preliminary findings on annual movement patterns include the following: (1) About 90% of
manatees tagged in the wild made seasonal migrations, typically between south Florida in winter
and central Florida during the warm season, while the remainder were resident year-round in a


49








given region. Brevard County, especially the Banana River, was the most heavily utilized area
during the non-winter months. (2) Individuals varied considerably in the extent and destinations
of seasonal movements. Distance between summer and winter ranges (median = 250 km) did
not vary significantly with age class or adult body size. One adult male made a spectacular
journey from north Florida to Rhode Island over a two-month period in the summer of 1995,
breaking records for the most northerly documented location and the longest migration for a
West Indian manatee. (3) Migratory travel between summer and winter ranges was usually
direct and rapid (up to 50 km/day), and was prompted by changing water temperatures. There
was substantial variation among individuals and years in the date that fall migratory movements
were initiated; the hypothesis that individuals respond at different temperature thresholds needs
further investigation. (4) Most individuals exhibited strong site fidelity to warm-season ranges
across years. Cursory analysis of within-season movements revealed a clear diel movement
pattern for manatees overwintering in north Biscayne Bay; they utilized the inshore canals and
rivers during the afternoon and moved offshore in the evening to forage on seagrass beds
through the night and early morning hours. The information provided by this long-term radio-
telemetry study has been and will continue to be valuable to federal, state and county managers
who must make regulatory and permit decisions to protect this endangered aquatic mammal and
its habitat from increasing human encroachment in Florida's coastal zone.



ACKNOWLEDGEMENTS

This research project has truly been a collaborative effort among many researchers, both
in terms of data collection in the field and spatial database development in the laboratory. The
Atlantic coast radio-telemetry study was initiated by Dr. Tom O'Shea, Jim Reid and Bob Bonde
and continued under the leadership of Dr. Lynn Lefebvre. A number of Sirenia Project staff
members contributed greatly to the development of the telemetry databases described in this
report, including Bob Bonde, Susan Butler, Dean Easton, Jim Henderson, Howard Kochman and
Jim Reid. Howard, in particular, deserves credit for most of the SAS programming that was
responsible for database creation and error-scanning. Dean and Susan have tackled numerous
challenges presented by the VHF telemetry database with unflagging devotion, thorough
scrutiny and good humor. Without all of their concerted efforts, little progress would have been
made in updating, processing and analyzing these large, long-term datasets. I thank Dean
Easton, Michele Klaips, and Christina Lucas for assistance with figure preparation. I am
grateful to Dr. Lynn Lefebvre for her support and persistence in keeping my Research Work
Order funded throughout this last trying fiscal year. I also thank Dr. H. Franklin Percival and
Barbara Fesler of the Florida Cooperative Fish and Wildlife Research Unit (National Biological
Service and University of Florida) for their continued assistance and administrative support of
this RWO. Funding for this work was provided by the Sirenia Project (NBS), U. S. Fish and
Wildlife Service, and Save the Manatee Club (SMC). I am very grateful to Judith Vallee and the
Board of Directors of SMC for their willingness to provide funding for this work at short notice.


50









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 manatees in Florida, 1974-1991. Pp. 223-258 in O'Shea, T.J.,
B.B. Ackerman, and H.F. Percival eds., Population Biology of the Florida Manatee
(Trichechus manatus latirostris). National Biological Service, Information and
Technology Report 1. 289 pp.

Bengtson, J. L. 1981. Ecology of manatees (Trichechus manatus) in the St. John's River,
Florida. Ph.D. dissertation, Univ. of Minnesota, Minneapolis. 126 pp.

Bonde, R. K., J. P. Reid, S. L. Tyson, and D. E. Easton. 1995. Captive manatee releases and
monitoring along Florida's east coast. Eleventh Biennial Conference on the Biology of
Marine Mammals. December. Orlando, Florida. Abstract.

Deutsch, C. J. 1994. Proposal for summary and analysis of manatee radio-tracking data.
Unpublished report to the Sirenia Project, National Biological Survey. 34 pp.

Deutsch, C. J., L. W. Lefebvre, and H. F. Percival. 1995. Evaluation of sampling and analytical
protocols for manatee capture-recapture and telemetry data (RWO 125). Florida
Cooperative Fish and Wildlife Research Unit Annual Report: January to December
1994. p. 31. Abstract.

Deutsch, C. J., L. W. Lefebvre, and H. F. Percival. 1996. Evaluation of sampling and analytical
protocols for manatee capture-recapture and telemetry data (RWO 125). Florida
Cooperative Fish and Wildlife Research Unit Annual Report: January to December
1995. p. 26. Abstract.

Eberhardt, L.L. and T.J. O'Shea. 1995. Integration of manatee life-history data and population
modeling. Pp. 269-279 in O'Shea, T.J., B.B. Ackerman, and H.F. Percival eds.,
Population Biology of the Florida Manatee (Trichechus manatus latirostris). National
Biological Service, Information and Technology Report 1. 289 pp.

Fancy, S. G., L. F. Pank, D. C. Douglas, C. H. Curby, G. W. Garner, S. C. Amstrup, and W. L
Regelin. 1988. Satellite telemetry: A new tool for wildlife research and management.
U.S. Fish & Wildlife Service, Resource Publication 172. 54 pp.

Garrott, R. A., B. B. Ackerman, J. R. Cary, D. M. Heisey, J. E. Reynolds II, P. M. Rose, and J.
R. Wilcox. 1994. Trends in counts of Florida manatees at winter aggregation sites. J.
Wildl. Manage. 58: 642-654.


51








Harris, R. B., S. G. Fancy, D. C. Douglas, G. W. Garner, S. C. Amstrup, T. R. McCabe, and L.
F. Pank. 1990. Tracking wildlife by satellite: Current systems and performance. U.S.
Fish & Wildlife Service, Fish and Wildlife Technical Report 30. 52 pp.

Hartman, D.S. 1974. Distribution, status, and conservation of the manatee in the United States.
U. S. Fish and Wildlife Service, National Fish and Wildlife Laboratory Contract Report
No. 14-16-0008-748. National Technical Information Service, #PB81-140725,
Springfield, VA. 246 pp.

Hartman, D. S. 1979. Ecology and behavior of the manatee (Trichechus manatus) in Florida.
American Society of Mammalogists. Special Publ. No. 5. 153 pp.

Irvine, A. B. 1983. Manatee metabolism and its influence on distribution in Florida. Biological
Conservation 25: 315-334.

Kinnaird, M. F. 1985. Aerial census of manatees in northeastern Florida. Biological
Conservation 32: 59-79.

Lefebvre, L.W., B.B. Ackerman, K.M. Portier, and K.H. Pollock. 1995. Aerial survey as a
technique for estimating manatee population size and trend problems and prospects.
Pp. 63-74 in O'Shea T.J., B.B. Ackerman and H.F. Percival, eds., Population Biology of
the Florida Manatee (Trichechus manatus latirostris). National Biological Service,
Information and Technology Report 1. 289 pp.

Lefebvre, L. W. and R. K. Frohlich. 1986. Movements of radio-tagged manatees in southwest
Florida, January 1985 March 1986. Unpublished report, Sirenia Project, U.S. Fish and
Wildlife Service and Florida Department of Natural Resources. 87 pp.

Lefebvre, L.W., T.J. O'Shea, G.B. Rathbun, and R.C. Best. 1989. Distribution, status, and
biogeography of the West Indian manatee. Pages 567-610 in Biogeography of the West
Indies; Past, Present, and Future (C. A. Woods, ed.). Sandhill Crane Press, Gainesville,
FL. 878 pp.

O'Shea, T.J. 1988. The past, present, and future of manatees in the southeastern United States:
realities, misunderstandings, and enigmas. Pp. 184-204 in Proceedings of the Third
Southeastern Nongame and Endangered Wildlife Symposium (R. R. Odom, K. A.
Riddleberger, and J. C. Ozier, eds.). Georgia Department of Natural Resources, Game
and Fish Division, Social Circle, GA. 253 pp.

O'Shea, T. J. and H. I. Kochman. 1990. Florida manatees: Distribution, geographically
referenced data sets, and ecological and behavioral aspects of habitat use. Pp. 11-22 in J.
E. Reynolds, III and K. D. Hadad (eds.). Report of the Workshop on Geographic


52


______ I~ II~








Information Systems as an Aid to Managing Habitat for West Indian manatees in Florida
and Georgia. Fl. Mar. Res. Publ. No. 49.

O'Shea, T.J. and C.A. Langtimm. 1995. Estimation of survival of adult Florida manatees in the
Crystal River, at Blue Spring, and on the Atlantic Coast. Pp. 194-222 in O'Shea, T.J.,
B.B. Ackerman, and H.F. Percival eds., Population Biology of the Florida Manatee
(Trichechus manatus latirostris). National Biological Service, Information and
Technical Report 1. 289pp.

Provancha, J. A. and M. J. Provancha. 1988. Long-term trends in abundance and distribution of
manatees (Trichechus manatus) in the northern Banana River, Brevard County, Florida.
Mar. Mamm. Sci. 4: 323-338.

Rathbun, G.B., R.K. Bonde, and D. Clay. 1982. The status of the West Indian manatee on the
Atlantic Coast north of Florida. Pp. 152-165 in Proceedings of the Symposium on
Nongame and Endangered Wildlife (R. R. Odom and J. W. Guthrie, eds.). Georgia
Department of Natural Resources, Game and Fish Division, Technical Bulletin WL5.
179 pp.

Rathbun, G.B., J.P. Reid, and J.B. Bourassa. 1987. Design and construction of a tethered,
floating radio-tag assembly for manatees. National Technical Information Service,
#PB87-161345/AS, Springfield, VA. 49 pp.

Rathbun, G. B., J. P. Reid, and G. Carowan. 1990. Distribution and movement patterns of
manatees (Trichechus manatus) in northwestern peninsular Florida. Fla. Mar. Res. Publ.
No. 48. 33 pp.

Reid, J. P. 1995. Chessie's most excellent adventure: The 1995 East coast tour. Sirenews 24:
9-11.

Reid, J. P., R. K. Bonde, and T. J. O'Shea. 1995. Reproduction, mortality and tag loss of
telemetered and recognizable individual manatees on the Atlantic coast of Florida. Pp.
171-191, In: Population Biology of the Florida Manatee (Trichechus manatus latirostris).
Edited by T. J. O'Shea, B. B. Ackerman, and H. F. Percival. National Biological
Service, U. S. Department of the Interior. Information and Technology Report 1.

Reid, J. P. and T. J. O'Shea. 1989. Three years operational use of satellite transmitters on
Florida manatees: Tag improvements based on challenges from the field. Pp. 217-232
in: Proc. 1989 North American Argos Users Conference and Exhibit. Service Argos,
Inc., Landover, Md.


53









Reid, J.P., G.B. Rathbun, and J.R. Wilcox 1991. Distribution patterns of individually
identifiable West Indian manatees (Trichechus manatus) in Florida. Mar. Mamm. Sci.
7(2):180-190.

Reynolds, J. E., III. 1981. Behavior patterns in the West Indian manatee, with emphasis on
feeding and diving. Florida Scientist 44: 233-242.

Reynolds, J. E., III and J. R. Wilcox. 1994. Observations of Florida manatees (Trichechus
manatus latirostris) around selected power plants in winter. Mar. Mamm. Sci. 10: 163-
177.

Schwartz, F. J. 1995. Florida manatees, Trichechus manatus (Sirenia: Trichechidae), in North
Carolina 1919-1994. Brimleyana 22: 53-60.

Service Argos, Inc. 1990. User Manual. Landover, MD.

Shane, S. H. 1983. Abundance, distribution, and movements of manatees (Trichechus manatus)
in Brevard County, Florida. Bull. Mar. Sci. 33:1-9.

Sirenia Project. 1992. Study Plan: Determination of Florida manatee movements and areas of
significance on the Atlantic coast. Unpublished U.S. Fish & Wildlife Service proposal.
Work Unit #406, Study/Task #3.

Sirenia Project. 1993. Atlantic Coast Manatee Telemetry: 1986-1993 Progress Report. Vol. I.
Unpublished interim report, National Biological Survey, U.S. Department of the Interior.

U. S. Fish and Wildlife Service. 1996. Florida Manatee Recovery Plan Second Revision. U. S.
Fish and Wildlife Service, Atlanta, Georgia. 160 pp.

Weigle, B. L., R. K. Bonde, S. J. Marcus, and P. A. Carbonara. 1987. Coordinated manatee
aerial surveys on the east coast of Florida during 1986. Seventh Biennial Conference on
the Biology of Marine Mammals. Dec. 5-9, 1987. Miami, FL. Abstract.

Weigle, B. L. and R. O. Flamm. 1995. GIS analyses for estimation of manatee habitat
utilization and migratory corridors. Eleventh Biennial Conference on the Biology of
Marine Mammals. December. Orlando, Florida. Abstract.

White, G. C. and R. A. Garrott. 1990. Analysis of Wildlife Radio-tracking Data. Academic
Press, Inc. New York. 383 pp.

Zoodsma, B. J. 1991. Distribution and behavioral ecology of manatees in southeastern Georgia.
Master's Thesis. Univ. of Florida.


54


__ ___1^1_1











REPORT APPENDICES


55


__ _I LI _~____




















OOO)COOOOOOOOOOONOOC) NOCOTm0)
cD CM CM





m0 u- rw 'T 0 0 m 0 0 C) 0 0 C) 0 0 v 0 o m v u- N c0 m) 0 m)
- - - -66J


c >

a)).
r-1
0 u



o 0





o
0)
C
*H 3
- 0
0 LH

z -


o
>1,


Co

00














*-*
oo



0\




I-
CO





-o
Cu


o
O





Cu

CO






D4





0
C.-














CO
-o



4)
-o













4-


O CM CM CM C C C CO) L)O O C CM CM C) CU C) CM )O CM C0 C0 CM CO CM CO CM CM








- N N C C C N N 0 O 0 C N M) M M M C- N M M M NN 02 -









<<< < <<<<<<<< <<<<<<<<<0





o~~O I C)C )0C )C )C )~C ) )C )C )C :C )< )a


CD


z
c


>O
0

0
0
T-


N
a
a
CO
0
N
cL



<


0
N
oc


N
a
CO

-
N
r-
CO4


N
CM


N



a-
CC
<
M
N
r-

CO


N
rD
a

o



z

-)
0


-


r-

OD




co
z

CO
-





-


I^-
aD
>
0
z

r-
N
>
0
z
0
-


I-
CO

z
-
-
0)
r--



-


r,-
CD
r>
0
z
N
-

t-

0
z
N
-


N
0
O





|
z
N
-
N

0
z
N
-


N N NCO 0 0

0) 0) 0C 0) C

- T- M M

000 <0


zz <
- OM
0) 0)0) 0
--- 02 N


CD
0C
a
a-
CM


aw
a

N
(C

CM


00
aM
a<

CO

N
a
a

N
N


aD
OD
z

N
0





N
CO

CO


0D
CO
z

N
0


m
LU
u-
ol


0)

a-
m

0
0)


CL
oo


<
o


0)
CD
a
a-
<


0
CL
a<
a-


0)
O

z
0
0)
-


0
z


N



CO
0)
a


0



0


0

0)
N
0



3
a
o


0
C)
0


CD

02



N


S a)
o

4-m
SCU
0) 0



a
C) 4-



. C
0) a)
() Cu

a


a)
a)
Vn 3


co
0
0) 4-1
< CB

- a)
al
0) 0)
4-' -1
ca a)

0 cc
OL

) 4-ca
0 0



,-,



X1
0


Lu
UJ
C)
0 -J a Lu -I 0
L ao m <
ou mo


O
0

m
N


CD

0
m


r-
0
I-


O
0


C)
0



C)
0

m
I--


z
<
u_


< UJ

5.J UJ L
CC^20
o 1m U C
_UJ OC U
CT 1f 3


z
Lu
Z
m


O


0
LU
-J


z
U)
oc
<


0 -; N 0 O CO N CD 0)
- O -O -o -



- N N I- I- I
66666666m-~-~-~


c:

0


m
F-


LU
-I
1-
E




a


O
0






m
F--


a:

-J
Ic
1-1


CO


m


>-







0
m


0
C)





O
m


J
UJ aC>-

wU Lu 1-1 <
CL 2 -4 0
mose
wws
n-wJOc


(D
CU
03
m
F-


r0

F-
o


CO
O4


0)
CM
0)


c0 0 0 N 0 -
2 r- CM c Ct q CM N CM
CM


c
0)
C
+-
.J


a




4-c


N D O CD 0 ) O N N -(
SO on ~ C- N
N -


zzzzzzzzzzzzzzzzzzzzZ>->-ZZZ>-- >-



U-. LL LL LL. LL LL LL U- U- LL. LL- m UL- LU- L- U- LL. LL m LL- m UL- 2


-o



00









-o
II








cj










-d





01
(S


a0


0
O

0


0
o)


(D

0


0
CO



0
0)
C)


0_
a-
Lo


w
0
CO

CL

0
O


0
0

0
CO
0

(0)


0
o2


LU
Lu X-


co c
Z 0N


C\

t- -


4a
a)
C
mu


CO
0

0
0)


0

a
0)
N















0 0 C O O CO C 0
0o 0o 0 00 0D 0 0
OD N-


I N 0 0 N M N O0
CD U) 0 0 N 0) 0) 0) 0 0
Q~NO~q UQ) N 0


0
C


M0)000000001000
oCo OOOO O O O O O


c>
*H
C >
.r *d
a-


0)
C
0ca
-c 0
u -H
C L
L a

do
(0
ca a)


0 L
z I-


0 (0
a
0)

r0)
C
a,











) 0)
Sa)


o L
co
Cc +1
-I c









a
L)
C ID






o L




0


m
3) +-'
< C




S0)
0 c
a0
a) ca




C 0
OL



CO .

a)
UX)
a-
0)




mI,


0
m
0)
I--

0
o
0
o

0
O
O
Vo


0
0
0
0


0
I-
0
0

-
0)
O
5,


m



N

0)
>
r




CM
CM


Va



O
>D


0)
-J




0
0)
>


0
z
0
CM


m
>1
0



0)
0


CM



N
a
0)


0C
LU
O
0


CV



00
N
CM
Co

0)

0
,-I


CM
0


CD
N

0)
-J
I
o,
r


w
0)
z




O


(D
CD
N


0)

0
,-
CD


<
m
O


I^-
0
0

O

r-
O
Fq-

0

-


0
0
IL


<


-
N
U)
0


-


W)
0

O


-0
-J


0
3
-
5


0
0


0
I)
C,
N


m
0
0)

0

N

0)

0>


0
0
CD




0
CL
N
a-


0)
CM


M




0)
LU


CfD
-)


0


O
t-
0
0
-

0


C-
0)


O
0)




m
0




F-
0
0
0)
0)


0

a
0




LU
0
-



CO
N


0
0)
>
0
Z
N
0
0
-J


mr
0)
-J
,-.
N

0)

OC

0


0)







CM
N


-J


N


0)
o,

0o


-
O

0
-


N 0) LO LO WD
m co 0 m m


0 (0 0 0 N
N N -

0) 00 0) 0) 0)


S- OO-
N N w- -


0

LU
co
-



>-
0
N


N
0
-


m


N
0


0>

0)
C-
0
0
0)
0
0


L- LL M M L- L- LL LL LL 2 2 2 2 2 2 L L- L-L LL- 2 LL. L L


E
LUZ



J, LU


LU
z
LU
W


>
LU
-j
F-4
M
mI


00) ) 0000
- 000000
SCmm C C C C


wom
Ozz
sis
LU C0 a.


0
CO
m


N
0
0
C


I

CO
0
0
I-


J




cO

I
m


LU
. W M
1) CDc
S0 LU CO
<0X0


0


I--


0
m
I-


CM


C--


0
CO

t-


0
0






m
I-


m
I-
0


I.

CI
0
L




0


>0
z


N

0
m


>-
Z




0

IL
U-
h-


z


I-

0

F-


I-1
-J

-rJ
0

CO
0

LL-
I-


0





1
I
I-


-J
Uj LU


S0 Z i00
C Z CZ 2 C-
O flCOT>$S220N


0

a-
LL
I--


0

CL
w-
1-


0

IL-
LL


L)
0

IL
I-
I--


cD
0

IL
0-
1-


-
0
.


N 0) 0
000


0

I--
-


CM,
0

--)
I-
o


o -00 0 CD C\ 0 N C 0 0 O 0 0 0 0 0 CD CD 0 0 0 Nc r- C ) 0 0 0 r- 0 LO UO 0
0N0)0ONC0NNC000UN)0000O0CD)NU) 000 0
O r- C C O O O-- O------ r- O O O OO





N 0 N M M N N M M M N N qCOO omm
0) O UO N 0 CO O m 0 N N N D O) 0 M) 0 NO 0 O D C O M M D
aNC Cf )) C Co C ) -- N- M coN co0 e






o0 U) 0 0) 0 0 M CD 0 0 V- ) 0 V- UO UO OD UO 0 0 U) Uf) 0 0 LO 0 0 UO V 0 '-T U)
c0 0 C) OD 0O (D CD 0 0 CoD N V cOD N O (D 0) COD CoD 0 ) ) c0) CN co 0) 0 0) rN co V
C CO CM C \ C C\ C\ C CM J CN C\O OJ C C N O M V) N V) V ) CO ) ) O C) C c C\







OU) O N N O) *0-- t o CD 0 o U) U) o o- o M) U)
cO O 0 o O 0 0 CO r- O O O NO M O O M NO 0 0 M
N O mC 4 J C C C C CM J CO C\ CM CM N M O C cr CO CO CM COO CO CM N
I I 1 1 1 1
0000 0 0 00000 0 0 0 0 0 0







0O00 <000<000<000 000000000000000>
<<<






QQQcc< Q acQQQ <<<0o0 << 0 <<<0oooo<<<<<<<<<<<<<<<<


0)

c
CU


0)
LU


-i
0
X
LU



0
o

--















00o0000000000000000o









SO O O OONOOO OOON O NO)
*a 0 0 ^- 0 0 0 M 0 0 0 oo 0 0 m i^ o cM


S>
+jl +j

C -4l
C Q0
0 CL
C 0

m 0
co
e-
-d 0
o -H
La

dP


ca
o -,
0
O Cu
0 l
Z I-




a)
C O
c i
r-
S4-'

E
0)





..J CI
a )
0 l
n0)
cm









Ca C
0 CU
u ca
a)


0) 4-C
a0)
Ia

C x
C4 4-'

0
C.)




co





OL





0-.
a)

Q.4
W>


Co U) Un) 0) CF)
CD (D 0 CD 'a-
o '- U)


* 0000 00 0- t 0 0 0 U M 0








* 0 0 0 0 N 0 O U) 'M I
c') 0 '- '- U) U N N c) U) CD - N c') 0 -

I I


<<<<<<<< C

D C D CD CD C C CD 0m C 0C CD CD
CD c C m rrC m z Cm


--- LL) N ) U)





- -)) C D CD CD CD N N N C) N U)
^- -.I-NC C N----------'^----- '^- M T-


0)
oD
C)
o


I0




w
0)


CD
o



C1
0
0)




CM
-J


CM



OD

LU
0)

Lu
CO)
T-


>- Z ZZZZZZZZZZZZZ >-




2 Ul- 2 U- 2 LLL U- U- L L LL U- U


XUJ
0 2


E m Er o < co m wf
o O LUJ MLU<< O
Z m U- O -- Ac

C C N C0 ) M U)
00000000


-


co
0
>O
O

z
I-


LU
I-'







C)
z
4--


LUJ
WW LU

0 2 0 UJ Cl)
0m< -


CDO
0


I-


0)
0

z
I-


0


I-


0
I--


CM
z
.--


I.-

O




UJ
0


I-


-J
-J





CLj

0

CL
I-


ui
-1





CO
0



ui
a-
I--


C)



0

U.J
C-
s


z
>
z
CO

0


C-
I-


CO C
0 (









00
CO N4


O C N NU C C C C 0 Vq M M
O U) 0 m C C C C D co 0) I N-
CD i- r- (?)


a)

-c
'4-





V cr
'-41 -









090
-4











A 0C
.-I

-H c a)
4- -q














L4-' -
D a)L








O 4





E
-o4
o) 0
eO L






CL









E C
= -.,

o o o

















- e4
CO -o,"
Cr











Ga a
SL
,'-)



0 ) 4
> L



OO
OCO

4f-

a) (D
c- -a

04 4a
9 co

- 4cc

CO

) C
+-
** a) <






-2- c0




CO



a
o


o>
0
o-

0

0

-J


CM

LU
CO
0



0)
CU
o


a,


0


a)
4)


a
cu
(0









Appendix lb. Radio-tagged manatee identity database for the Atlantic coast of Florida and
Georgia, from May 1986 through May 1996, showing capture and release locations and
method of capture. SAS database name = tagmanid.sd2. N = 83 individuals.


Manatee
ID

TBC-01
TBC-02
TBC-03
TBC-04
TBC-05
TBC-06
TBC-07
TBC-08
TBC-09
TBC-10
TBC-11
TBC-12
TBC-13
TBC-14
TBC-15
TBC-16
TBC-17
TBC-18
TBC-19
TBC-20
TBC-21
TBC-22
TBC-23
TBC-24
TBC-25
TBC-26
TBC-27
TBC-28
TBC-29
TBC-30
TBC-31
TBC-32
TBC-33
TBC-34
TBC-35
TBC-36


Manatee
Name

DIXIE
TRIXIE
MOON
ECLIPSE
GYRO
MOE
BILL
ANGELA
C-COW
FRAN
GLORIA
HEIKE
IRENE
JEROME
KAREN
LARRY
MADONNA
LEROY
SHARON
RUTH
WILLIE
MAGOO
HILLARY
BETTY
D-COW
PEEWEE
MEL
LIBERTY
CASEY
MARIA
FREDDIE
TOMASINA
ERNIE
MILEY
E-COW
SUNSHINE


59


Capture Location

BANRIVUP:MSS
BANRIVUP:MSS
BANRIVMID:CCS
BANRIVMID:CCS
BANCRK:HWY3
BANCRK:HWY3
BANCRK:HWY3
BANRIVUP:MSS
BANRIVMID:CCS
BANCRK:HWY3
BANCRK:HWY3
BANCRK:HWY3
BANCRK:HWY3
BANCRK:HWY3
BANCRK:HWY3
BANCRK:HWY3
BANCRK:HWY3
BANCRK:EBASIN
BANRIVUP:MSS
BANRIVUP:MSS
BANRIVUP:MSS
BANRIVUP
HALIFAXRIV:OAKHILL
BANRIVMID:CCS
BANRIVMID:CCS
BANRIVMID:CCS
INDRIVMID:MELBOURNE
BANRIV:SYKESCRK
PORTEVERGL:PPINTAKE
BANCRK:HWY3
BANCRK:HWY3
BANRIVMID:CCS
INDRIVUP:HAULOVERCNL
BANCRK
BANRIVMID:E528CWY
BANRIVMID:CCS


Release Location

BANRIVUP:MSS
BANRIVUP:MSS
BANRIVMID:CCS
BANRIVMID:CCS
BANCRK:HWY3
BANCRK:HWY3
BANCRK:HWY3
BANRIVUP:MSS
BANRIVMID:CCS
BANCRK:HWY3
BANCRK:HWY3
BANCRK:HWY3
BANCRK:HWY3
BANCRK:HWY3
BANCRK:HWY3
BANCRK:HWY3
BANCRK:HWY3
BANCRK:EBASIN
BANRIVUP:MSS
BANRIVUP:MSS
BANRIVUP:MSS
BANRIVUP:MSS
BANRIVUP:MSS
BANRIVMID:CCS
BANRIVMID:CCS
BANRIVMID:CCS
BANRIVUP:NASACWY
BANRIVUP:NASACWY
INDRIVUP:PORTSTJOHN
BANCRK:HWY3
BANCRK:HWY3
BANRIVMID:CCS
BANRIVUP:NASACWY
BANCRK:HWY3
BANRIVMID:E528CWY
BANRIVMID:CCS


Capture
Method

NETCAPT
NETCAPT
FREETAG
NETCAPT
NETCAPT
NETCAPT
NETCAPT
NETCAPT
FREETAG
NETCAPT
NETCAPT
NETCAPT
NETCAPT
NETCAPT
NETCAPT
NETCAPT
NETCAPT
NETCAPT
NETCAPT
NETCAPT
NETCAPT
CAPTREL
CAPTREL
FREETAG
NETCAPT
FREETAG
CAPTREL
CAPTREL
CAPTREL
NETCAPT
NETCAPT
FREETAG
CAPTREL
CAPTREL
NETCAPT
FREETAG










Manatee
ID

TBC-37
TBC-38
TBC-39
TBC-40
TBC-41
TBC-42
TBC-43
TBC-44
TBC-45
TBC-46
TBC-47
TFK-01
TFK-02
TFK-03
TFP-01
TFP-02
TFP-03
TFP-04
TFP-05
TFP-06
TGA-01
TGA-02
TGA-03
TGA-04
TJX-01
TJX-02
TJX-03
TMI-01
TMI-02
TMI-03
TNC-01
TNC-02
TNC-03
TNC-04
TNC-05
TNC-06
TNC-07
TNC-08
TNC-09
TNC-10
TNC-11


Manatee
Name

PAMELA
ADAIR
LYDIA
DANISE
SCOTT
CHESSIE
ROBBIE
MOOSE
HARVEY
FOSTER
INDY
MANNY
STAN
OLLIE
HUTCH
ROSS
NATALIE
SOPHIA
LANI
VANNA
MARY
TORY
MARMONTEL
MERCURY
CONNIE
PATIENCE
FRECKLES
BOB
FERGIE
CLOCKWORK
DIANE
JESSE
TAMMY
PAT
ROSEANNE
NANCY
WHITE
GEORGE
JANATEE
MARCH
VALE


Capture Location

STLUCIE:PPINTAKE
INDRIVUP:FPLPP
CAPTIVE-BORN
INDRIVMID:MELBOURNE
RIVIERABCH:PP
CHESAPEKBAY:CHESTRIV
BANRIVMID:W528CWY
HALIFAXRIV
CAPTIVE-BORN
CAPTIVE-BORN
CAPTIVE-BORN
FLKEYS:KEYLARGO
FLKEYS:TAVERNIER
FLKEYS:TAVERNIER
STLUCIE:HUTCHINSONIS
PORTEVERGL:PPINTAKE
STLUCIE:PPINTAKE
STLUCIE:JENSENBCH
JUPITER:SOUND
FTPIERCE:PPDISCH
KINGSBAY,GA
BRUNSWICK:GAPAC
BRUNSWICK:GAPAC
BRUNSWICK:GAPAC
STJOHNRIV:DRSINLET
STAUGUST:GUANALAKE
STAUGUST:GUANARIVPK
PORTEVERGL:PPINTAKE
BISCBAY:DEERINGBAY
MIAMI:LITTLERIV
FERNBCH:CONTCORP
FERNBCH:CONTCORP
FERNBCH:CONTCORP
FERNBCH:CONTCORP
FERNBCH:CONTCORP
FERNBCH:CONTCORP
FERNBCH:CONTCORP
FERNBCH:CONTCORP
FERNBCH:CONTCORP
FERNBCH:CONTCORP
FERNBCH:CONTCORP


60


Release Location

BANRIVUP:MSS
BANRIVUP:MSS
BANRIVUP:MSS
BANRIVUP:NASACWY
BANRIVUP:NASACWY
BANRIVUP:NASACWY
BANRIVMID:W528CWY
BANRIVUP:NASACWY
BANRIVUP:NASACWY
BANRIVUP:NASACWY
BANRIVUP:NASACWY
FLKEYS:KEYLARGO
FLKEYS:TAVERNIER
FLKEYS:TAVERNIER
STLUCIE:JENSENBCH
HOBESOUND
STLUCIE:PP
STLUCIE:JENSENBCH
JUPITER:SOUND
FTPIERCE:PPDISCH
KINGSBAY,GA
BRUNSWICK:GAPAC
BRUNSWICK:GAPAC
BRUNSWICK:GAPAC
STJOHNRIV:ORANGEPK
STAUGUST:GUANALAKE
STAUGUST:GUANARIVPK
BISCBAY:MATHESONHAM
BISCBAY:DEERINGBAY
MIAMI:LITTLERIV
FERNBCH:CONTCORP
FERNBCH:CONTCORP
FERNBCH:CONTCORP
FERNBCH:CONTCORP
FERNBCH:CONTCORP
FERNBCH:CONTCORP
FERNBCH:CONTCORP
FERNBCH:CONTCORP
FERNBCH:CONTCORP
FERNBCH:CONTCORP
FERNBCH:CONTCORP


Capture
Method

CAPTREL
CAPTREL
CAPTREL
CAPTREL
CAPTREL
CAPTREL
NETCAPT
CAPTREL
CAPTREL
CAPTREL
CAPTREL
CAPTREL
FREETAG
FREETAG
CAPTREL
CAPTREL
CAPTREL
FREETAG
FREETAG
FREETAG
FREETAG
NETCAPT
NETCAPT
NETCAPT
CAPTREL
RESCUE
RESCUE
CAPTREL
RESCUE
FREETAG
NETCAPT
NETCAPT
NETCAPT
NETCAPT
NETCAPT
NETCAPT
NETCAPT
NETCAPT
NETCAPT
NETCAPT
NETCAPT











Capture Location

FERNBCH:CONTCORP
PORTEVERGL:PPDISCH
PORTEVERGL:PPDISCH
PORTEVERGL:PPDISCH
PORTEVERGL:PPINTAKE
RIVIERABCH:PPDISCH


Release Location

FERNBCH:CONTCORP
PORTEVERGL:PPDISCH
PORTEVERGL:PPDISCH
PORTEVERGL:PPDISCH
INDRIVUP:PORTSTJOHN
RIVIERABCH:PPDISCH


NOTE: Meaning of Capture Method. CAPTREL = Tag and release a captive rehabilitated
manatee. FREETAG = Attach belt harness and radio-tag assembly in water on an
unrestrained wild manatee. NETCAPT = Capture and restrain a free-ranging manatee
with net; tag and release on site. RESCUE = Same as NETCAPT but purpose was to
rescue the manatee (e.g., to remove line entanglement and assess condition); tag and
release on site.


61


Manatee
ID

TNC-12
TPE-01
TPE-02
TPE-03
TPE-04
TRB-01


Manatee
Name

MOSSIE
SPOT
FIREBALL
SICKLE
SUSAN
SONNY


Capture
Method

NETCAPT
FREETAG
FREETAG
FREETAG
CAPTREL
FREETAG









Appendix 2. Description of the types of sensor data collected from satellite-monitored PTTs,
including technical information on the manner in which the data were formatted in the
transmitted message.


TYPES OF SENSOR DATA

1. PTT Temperature: The temperature sensor yields "counts" that are converted to
temperature using a calibration curve provided by Telonics. We determine a 4th-order
polynomial through five points (from 0 40 C, at 10-degree intervals) to make the conversion.
Each PTT has a unique calibration curve and polynomial equation; this changes only if the
PTT's motherboard is replaced, which is rare (B. Burger, Telonics, pers. comm. to J. Reid).
# Bits (maximum value): 16 (65,535), 14 (16,383), or 8 (255). A scaling factor needs to be
multiplied times the raw sensor counts before fitting to the temperature calibration
curves: 1 (16-bit); 4 (14-bit); or 142.222 (8-bit).
Note: The sensor is located within the tag housing; therefore, it may be affected by solar
radiation and is likely to show a time lag in temperature change.
Note: The maximum value for an 8-bit variable yielded maximum temperatures of only about
270 32 C for four new PTT's deployed in March 1995. This problem needs to be
investigated.

Accuracy: According to Bill Burger of Telonics, the sensor is only accurate to about 2 0C.
Precision: At the poorest level of precision used (8-bit value), my "eyeball" estimate of
precision is about 0.2 oC. Precision improves when a larger number of bits are used to
carry the information.

2. Low Battery Voltage Flag: A non-zero value (1 or 3) indicates that the battery voltage
is low and the PTT will stop transmitting in several days (- 4 9 days, J. Reid and R. Bonde,
pers. comm.).
# Bits: 1 (values 0 or 1) or 2 (values 0 3).
Note: For a 2-bit specification, the values of 1 and 2 are erroneous.

3. Short-term (1-hr) Activity Index: This counter gives the number of minutes in the
preceding 60-minute period in which a motion-sensitive switch inside the PTT housing was
triggered at least once. The switch is triggered by a tip of 90 or more from the vertical.
# Bits (max. value): 6 (63).
Note: The maximum value, by definition, cannot exceed 60.

4. Long-term (12-hr) Activity Index: This counter gives the number of seconds in which
a motion-sensitive switch inside the PTT housing was triggered in the preceding 12-hr period
(this period is specified by the user). The switch is triggered by a tip of 90 or more from the
vertical.
# Bits (max. value): 10 (1023).


62









Note: The 12-hr period begins with the cut-on time (typically between 0700-0900 EST), so it
approximates activity in daytime and nighttime periods. Daytime transmissions give the
nighttime activity value until 12 hrs from the cut-on time; the daytime activity value is
transmitted starting in the evening and through the night.
Note: There are 43,200 sec per 12-hr period. So a maximum tip count occurs when the sensor
is triggered an average of once per 42 sec (or 1.43 tips/min). The maximum tip count could be
increased by changing the Activity Count Scaling factor (currently = 1) to a larger number.

5. Saltwater Switch (SWS) Fail-safe Flag: This flag indicates that the SWS is
functioning properly (0) or has failed (1) and is being overridden. The failsafe timeout period
was set at 2 hrs for the four PTT's deployed in March 1995. If the switch senses conductivity
(i.e., presence of saltwater) for a continuous period of greater than 2 hrs, the flag changes to a
value of 1; then the SWS function is overriden, allowing the PTT unit to operate according to
programmed duty cycle.
# Bits (max. value): 1 (1)

6. Average Dive Time (in seconds) over "AVGINT" hours: The average duration that the
PTT cap was underwater during the preceding "AVGINT" hours (chosen to be 4 hrs). The raw
value must be multiplied by the chosen Dive Time Scaling Factor (5) to obtain the average dive
time in seconds.
# Bits (max. value): 6 (63) ==> Max. Aver. Dive Time = 315 sec (5.25 min)
Note: Dive durations less than a specified value (chosen to be 5 sec) are-excluded from the
calculations.
Note: Dive times refer to the behavior of the PTT, not the manatee!

7. Dive Count over "AVGINT" hours: The number of times that the PTT became
submerged during the preceding "AVGINT" hours (chosen to be 4 hrs). Must multiply the raw
value by the chosen Dive Count Scaling Factor (5) to obtain total dive count.
# Bits (max. value): 6 (63)
Note: Dive lasting less than a specified value (chosen to be 5 sec) are excluded from the total
count.
Note: Dive times refer to the behavior of the PTT, not the manatee!
Note: Dive Count X Average Dive Time = Total Time PTT Submerged
(Total Time/AVGINT in seconds) X 100 = % Time PTT Submerged


63


~_~_r~___~D~~1_I









SENSOR DATA FORMATS


The data stream following the PTT identification code is comprised of 32 bits (4 bytes).
Argos specifies that sensor data must arrive in 32-bit blocks. We use all 32 bits, in various
combinations for the different sensors. Five sensor formats have been used. The first one
(Format 0) used with Beauregard in 1985 included two 8-bit temperature words (combined in a
formula to calculate PTT temperature) and two activity indices. Sensor data (Format 1) for East
coast manatees included PTT temperature, short-term activity, and long-term activity. The next
sensor format (Format 2) was a "software upgrade" by Telonics that added a low-battery voltage
flag. The last programmed format (Format 3) incorporated a saltwater switch to yield
information on diving behavior and to increase the efficiency of transmissions (only transmitting
when at the surface), and dropped short-term activity. Another sensor format (Format 4) was
generated in the dataset when Telonics upgraded the software from Format 1 to Format 2 but
Argos was not immediately informed of the change. Consequently, Argos reported a 16-bit
value that was actually 2 bits (usually a value of 0) for the low voltage flag followed by 14 bits
for temperature. Sensor data from some PTT's therefore follow Formats 2 and 4.

No. Bits for each Sensor

Format 0 Format 1 Format 2 Format 3 Format 4 *
No. Sensors 3 3 4 6 4

Low Battery Flag 2 1 [2]

Temperature 8 + 8 16 (1) 14 (4) 8 (142.222) [14 (4)]

Short-term Activity 6 6 6 6

Long-term Activity 10 10 10 10 10

SWS Failsafe Flag 1

Aver. Dive Time 6 (5)

Dive Count 6 (5)

* Same as Format 2, except Argos reported the low battery flag and temperature as one 16-bit value.
1 Two temperature words were combined in a formula to calculate temperature (C):
[ (A 256)+B]- 33,531 6
808.11
Note: Scaling Factor is given in parentheses, if not equal to 1. Multiply raw sensor value times
the scaling factor to calculate the actual value.
Note: The sensor order in the Argos output datasets is as shown above for formats 1, 2 and 4.
Variables are listed in the following order for Format 3: Temperature, Low Battery
Voltage Flag, SWS Failsafe Flag, Average Dive Time, Dive Count, and Long-term
Activity. The meaning and order of the Format 0 activity indices are unclear.


64









Format 4 applies to five PTT's over the following dates of operation, up until 1245 hr
EST on 26Jan95 (except for PTT No. 9993, changed on 02Dec95); then Format 2 is in effect.

PTT ID No. 9648 19 July 1993 Software Upgrade (from Format 1 to 2)
21Oct93 05Dec93 Fergie
20Dec93 12Apr94 Spot
29Apr94 04May94 Hillary
25May94 25Aug94 Vanna
30Aug94 15Sep94 Moose (in pen)
26Jan95, 1245 hr Argos changed sensor format to correspond to Format 2.

PTT ID No. 9990 19 July 1993 Software Upgrade (from Format 1 to 2)
190ct93-04Dec93 Stan
09Dec93-16Feb94 Stan
(Unit was sent to Telonics on 30Mar94 and was not repairable.)

PTT ID No. 9993 19 July 1993 Software Upgrade (from Format 1 to 2)
11Feb94-26Aug94 Clockwork (Format 4)
02Dec95 Argos changed sensor format to correspond to Format 2.

PTT ID No. 9995 3 August 1994 Software Upgrade (from Format 1 to 2)
07Oct94- 18Oct94 Chessie
12Jan95 >26Jan95 Chessie
26Jan95, 1245 hr Argos changed sensor format to correspond to Format 2.

PTT ID No. 9999 19 July 1993 Software Upgrade (from Format 1 to 2)
04Jan94 17Jan94 Sonny
24Oct94 Ongoing Sonny
26Jan95, 1245 hr Argos changed sensor format to correspond to Format 2.

Sensor data reported by Argos for the following PTT was incorrectly changed from
Format 1 to Format 2 on 26 January 1995 (1245 EST). This mistake was corrected on 12 July
1995 (1000 EST). Since this problem only affects a short period (28 Jun 1995 12 July 1995)
for one animal (TBC-43, Robbie), and would require conversion of the values back to binary
first, these temperature data will just be deleted.

PTT ID No. 9997
27Dec88 26Jan95 Format 1 (16,6,10)
26Jan95, 1245 hr Argos changed sensor format to correspond to Format 2.
26Jan95(1245EST)-12Jul95(1000EST) Format ??
Counted as 16,6,10, but reported as 2,14,6,10.
12Jul95 31Dec99 Format 1 (16,6,10)


65









The sensor data are reported by Argos for each message that is received, resulting in
several or many messages per location. The compression index indicates the number of identical
consecutive messages received. In order to condense the sensor information into one
observation per fix, the following protocol will be used:

Temperature -- median count, weighted by compression index. Temperature can change from
one message to the next.
Short-term Activity -- modal count, weighted by compression index. This changes at the end of
each hour.
Long-term Activity -- modal count, weighted by compression index. This changes at the end of
each 12-hour period.
Dive Count and Average Dive Time -- modal count, weighted by compression index. This
changes at the end of each 4-hour period.


66









Appendix 3. Changes in PTT duty cycles and turn-on times used for radio-tracking manatees
along the east coast of Florida and Georgia from May 1986 to May 1996. The periods
over which different duty cycles were employed overlapped because the cycles could
only be changed by technicians at Telonics. Likewise, changes in turn-on times did not
occur instantaneously because they could only be made by Sirenia Project staff after the
PTT units were retrieved from the animals.


Years: 1986- 1987
Duty Cycle: 1 hr ON 4 OFF 3 ON 3 OFF 3 ON 10 OFF
Period On per 24-hr Cycle: 7 hr
Turn-on Time: 0200 hr EST
Notes: The first year of tracking on the Atlantic coast used the same duty cycle as was used for
the three manatees tracked in southwest Florida in 1986.

ON TIMES (EST)/[GMT] OFF TIMES
0200 0300 [0700-0800]
0300 0700
0700- 1000 [1200-1500]
1000- 1300
1300- 1600 [1800-2100]
1600 0200



Year 1987-1988+:
Duty Cycle: 2 hr ON 5 OFF 2 ON 2 OFF 2 ON 7 OFF 2 ON 2 OFF
Period On per 24-hr Cycle: 8 hr
Turn-on Time: 0630 hr EST
Notes: Some PTT's in 1987 (and maybe 1986?) had this programmed duty cycle and at least
one PTT continued with this duty cycle until July 1993.

ON TIMES (EST)/[GMT] OFF TIMES
0630-0830 [1130-1330]
0830- 1330
1330- 1530 [1830-2030]
1530- 1730
1730- 1930 [2230-0030]
1930 0230
0230- 0430 [0730-0930]
0430 0630


67









Years 1987 1996:
Duty Cycle: 2 hr ON 4 OFF 2 ON 3 OFF 2 ON 6 OFF 2 ON 3 OFF
Period On per 24-hr Cycle: 8 hr
Turn-on Time: Variable (0630 0900 hr EST); 0700 hr in 1987 and 1989.
Notes: During 1987-1988 some PTT's used this duty cycle, as well as those listed above. The
turn-on time has been changed many times over the course of the study in order to better
match PTT transmission periods to times of satellite overpasses. Only the most recent
changes in turn-on times are listed below. See the PTT log records for turn-on times for
each tag in different years.

Time Period: < Spring 1994 to August / Fall 1994
ON TIMES (EST)/[GMT] OFF TIMES
0730- 0930 [1230-1430]
0930- 1330
1330- 1530 [1830-2030]
1530- 1830
1830-2030 [2330-0130]
2030 0230
0230- 0430 [0730-0930]
0430 0730
-,-----.------.-----.--------------------------------"-------------------^----------------.-

Time Period: August / Fall 1994 to Winter / Spring 1995
The turn-on time was changed on 9 August 1994, starting with the captive-release animal
"Scott" & the PTT Accuracy Experiment. Other tagged manatees were changed later in the year
as PTTs were switched. The change in timing was necessary due to drifting of the NOAA-11
satellite such that the current duty cycle was not optimal for the overpass times. Service Argos
offered free multi-satellite processing during this period, adding a third satellite (NOAA-9) in
addition to NOAA-11 and -12. We changed the turn-on time from 0730 hr EST to 0900 hr EST
to cover the NOAA-9 and -11 overpasses. The programmed duty cycle remained as above.

ON TIMES (EST)/[GMT] OFF TIMES
0900- 1100 [1400-1600]
1100-1500
1500 1700 [2000-2200]
1700- 2000
2000- 2200 [0100-0300]
2200 0400
0400- 0600 [0900-1100]
0600 0900


68









Time Period: Winter / Spring 1995 to Spring (March) 1996
On 1 Feb 1995, Argos replaced the NOAA-11 satellite which had been drifting out of
orbit with NOAA-14. This reinstated a "normal" pattern of overpasses. Since we had altered the
turn-on time last year to hit NOAA-9 and -11, there was a large reduction in late afternoon
(1600 hr) and early morning (0500 hr) locations after the switch in satellite operation. The turn-
on time was changed to 0700 hr on 10 and 12 March 1995 for 5 manatees tagged by GA DNR
(Brunswick, GA & Fernandina Beach, FL). Subsequently, the same was done for tagged
manatees when PTTs were switched. The programmed duty cycle remained as above.

ON TIMES (EST)/[GMT] OFF TIMES
0700 0900 [1200-1400]
0900- 1300
1300 1500 [1800-2000]
1500- 1800
1800- 2000 [2300-0100]
2000 0200
0200- 0400 [0700-0900]
0400 0700


Time Period: Spring (March) 1996 Present
ON TIMES (EST)/[GMT] OFF TIMES
0630-0830 [1130-1330]
0830- 1230
1230- 1430 [1730-1930]
1430- 1730
1730- 1930 [2230-0030]
1930- 0130
0130- 0330 [0630-0830]
0330 0630


69















WC N ;; 1t 0 M) 0r) M CD 0 CY c0) 0 (D v co (D 0 0 le N, O z 0) 0)
..-. co c) 0)- ) 0c M c C V r co t CD


4-
0












cqs

(>
Q.)
0

















co
(L>

0















to
e




























.2 .
oJ --r





















C3 (U
-o ^





















,-
















>1 0
:a m
C3 S




4- 4-

















.4n
W
'n 0

co
C.)^
* 0~



C^ C.)






















oci
4- o








U)S










o DLo



-O


| o
*E. .2


z
UJ




0c
UE 0
LU LU


I- I-


LJ LU




0 U
N -

0 0
UJ .J



co co
z
' .


-0


z
0
CD

U)
0
0







cr--
--z
0



Go
co
0

0
I-














z 0
0) N



U) '-
-^ 0)


F-
, CL I--
U LUJ U
X. c ccc
000
CO O




,> Z
00 <

- 0


CN CM N


OO


co (D

0m)


>



0O
O
IL

"t-

h--


LUJ
-1 X
L U
x X-
Q.

N 0
0

co LL.
> -
00
QO (
z
C co
NO <

0 I-
0 I

... o ... ..-
X XIX X
0



N N

no on o
IL 0) 00.L
U) < LU LU
w -i cc =.


0>0
C -






o Go co
U) > 0 z
0 0 UJ <
0 = 0 -3
CMj to CMj CO
0 0 -


Sz
D


C -D
o z
- Z






0
CU Z
C
cc cc
l) 0



co J

0

I- N
I-














N o


-J

IL

LUJ Z Z
> cc


5J
Ni)
0.0
U -
F- Y Z
IL =D <
0
* 0
NOZ

F- Z
; -1 D
oi LLu



m x -i
o a



o) **J *
U) >< X
0
-N x




UO UO Cc
< V
000










CDL Z? CD 0
ooco




<010 01 LU
= ) <> UJ
0) ) 0) c
0 0 N


F-

0
LU

0)
UJ

LU
(N
0
0

-J





co
a)
0



0)





LU
IL
m0
cN
CM


0
LUJ
LU

U)
z

z
z
0







co







0
0


U)


>-


LU
x X X CD X X


LU LU lU LUI LU LU


0U)0000
0 CO 0 0 0 0

M CM U) 0) i- '-





00 N'- O
0) 0) 0) 0) 0) 0)

CD c LO Cf

0 0 C0M 0 -
000000
CM N CM ) C0) C0)
CN CM N CN CM CN


c0) 0 c0) m) LO O
0 0 0 0 N N


IL

z
U.


co
co


N I-- LL L -
0.0..>>>
v CD 0 co m)
v~ a) v~ F) tN
0) 0) r- r to
Cm m~


ooCooN 0o 00ooo0-0ooo


0o 0) 0)
000
000



CN CN M N


0

0

CD

CM


z





0
0



f--
,,


o
0



N~


0

0




CM


0 0

co
00



0 -
c) c)


I-


0
z


C)



0
U




C)


R


0



F-
CL
0
cr-



C0
0
C)






0
0


LI-
U-)





0
F-




LU
U)
0
0







CD


C)



C)

0


io 10 n in 10l
U) U) U) UO U) U)
-T v ~ v I I


--



0
LU

CD
>-



0



-J
NM

U)


00
CO O

U I-
N I-
> CL

0D CM
Uo 0


0 -


F-
1-
0.
0

I-




co










z
0



D,-


0
LU
0


LJ








0

CI


0O
e


x
I-

a.


LO



CM
N

U)


0

0
0
CO


0 0 0
UJ LU






0 "- cO
M CO 0






< Q
N 0) C0
0 '- CN


N CN


x

CL
X



cc
0



CO
0o
co





0)
N


0
UJ






0
co

co
O




z

CM

0



CL
0
c)


C








I--
I-
a.
0
















N3


I--
LJ
z





CD



0

o
C)
0
o









0
o



F-
10





0








o
0











I-
L-


LU



0



CD

CL
UJ
o

0




0






oL
0









z


co
0
0


>-
v0


no
LU LU


CL r- I,





UO N- N
CD CD ND

0 -0<
0 Z -7
0 N 0
0-0
0 U) U)








aL > NL

N 0 CN
0 0

-0 CM


0000


SCU) 0) 0


cc
L"
Qi
LUJ
X
I-

-J

CL
U)



z










LU

z
CD
N
Nw

0S
0


IU.



0
CO


> > cL

0 0 O



o 0


U- I-
> I-

0 -
No 0)
Sm


>)




U)
co
co

O
CO
Tf


I--
h-
Cr


O
0)




0






o



I.--


LU
x




0
I-


x




--
0


o
0

0)


co


8-


z z



oo

Ch- -


o
0)


N-


0

U)


o

F--
C0


0


co


U)

0
IN


o


N-


0


0

N
,N














0! 0) CO CO 0 -

0 0 M ; 0
CMi N00) 0


C- O-no -o
CDN0CDe-t)sComCD
meo N U).- e


D V CO V) CD
SCD )


CD 0 C 0 0! 0 CD CD 1
0) CQ 'W 0 V C COD CM V C
I)D CO nCD co v CMN


cc
LU
N
I-
I--





z


LU
C-



0
I-






N





O
7
a-
or-



LU
CD







LU
CC
0)


I I
N- 0
F- <2z
i- o =)
CL U-
0-0
0 0<
F- -J 2

NNN
a- a- a-
a. CL CL
LU LU LU
a: a: a:
0 U) CD
N -
000

-N

NNN
CM CD CD

LU a- 0;
N CD N
U. < -


F-
I-




I-

a-


0
N

0



CD



0


N
{1-
N
m
3:
o0


X
I-

oL
LU


U)
CD

-


N
CD


0
Z
CD
M


Q
LU

N a- a- a- a- a-
CL C. L CL CL IL CL
LU. LU w. LU LU LU LU

0 0 CD CC0 CO







co i-ocooo
U)-C-N--CV
-n -o -~ r- i- -

N N N N N N N
CM CM 0M CD CD CD CM
0m 0 0m 0m 0m 0m 0m
Z 0M 0M a: -J C0 >
3- U- < -13 < z
0 0 N CD N CD CD

0 UO U) U) U) U)

>- N- N- N- N N N
N N N N N N N



N N N N N N N

U- LLN N-- LL N I-
M =XF- F- X F I- -
> > aL a- > a-L -
N 0 N V N CD
- CD N t CD 0 M
CO N~ 0 CD CD i- CD
0 0) NCD0

0 0NCM O0 *T


z
0
0o

o

m

CM
0
0-


z
0



0
|
CO
N


CO
0


0
mO
*O

cD
N
CD


z


0


C0
co
N

U')


z


0


0O
CD
o
CD


z
LU
0
acc

-J

CO
I- I-



-J
I- *-
LU




F- S- h-



CL0 C0 0 *
(N ,-
N- I I a-



a- n a-
7J (R -a>C

o 0 0O








M CM M CM


<
C-




LO
N



-
n
o.



0)

0






>-
0










>-
iF-
CL
It-

0
(0


LU
)O
COL
z z Z .-

0 0 0 0j

0000


F- F- F- I
N CD CD 0
CO CO CO


CD

4-


o
LU
z
0
Co


-

0
0
CM
m

N


z
LU
0
.I



-.



CO

-I
CO




x



I--






co
0
z






>-
I-
0O


0










CM


I-









a-
LU





o

-
M,


-,


.M



F-
0
LU
ar
N
CD
00



F-
U3
<



0
0

N







CO
co



M
CD
0"
0


a-
I-




X X


LU LU
: a:

-0
CM4
0 0
0 0
N Uo

- -

< 0
M0 cD


Q
LU

Q<
LU 4: LU

N a- N N
F- CL F- CL
LU Lu Lu LU LU
cc cc: z cc z
0 U) 0 U) N
0 C -
NM CD CD 0 CD

N N CD N CD
Z M > Z M:
< wU 0 m> g
CD U) 2D N CD
0 n co z 0n
-M -M NN M


Nv N N N N


>- >- >-




-o


0 O 0

'-ON


LU
rj)


CO
CL
a-
0
LU



0
CD
N-


LU

-i
0
CO
LU


0
0
o4
F-
CN
Nr


I-
a-



0



a-
0








0



CD
U-
0
0


cD
N
CD


a:
N
a-

0
0.
O
0
CM
0
N
04


CD


<


s
4:




N
N
LU





z
LO
CD



a-
4:
o-
<
0
N
NM
in


_-J


N

0
LU
Ir
0)
OD
>-
m I-
0
0
.- F-
0"
.-I X
I-w


>-
< LU

- 0
CMj






N
- U
in 7*
0)0


m 0
< <
CD CD
CM M
w N
N
>-
LUI
U-
<
V)

0
LU
XO

a- N
LU LU
m m:
0 O
0-
- -

cw 0
> Xc

CM CS

U0
N N


CC
> > 0-



0

0 0 ,-


0 0
wC w LU
>- >1 Q


000
000
C0 CDC
0 10 (0
V LO 0
m U) CD


LU


0

0
N
NQ
h-

hO-


z

CO



0


<

0

-J
Ir

0
ar


N 0 N
N LU co
a- S a-
LU LU
0 w CDC
LU 0 0 v
a: CDNI

0) i 03
cc 2 -1 0
a- z -a -a
< XN0 <
0 1--0 X
N-
OS-1

0 0
-01 w w w -I

cD a- 0 0 CD
4:LUwLULU4:
w a: w: <:
NI CDCD CD 0 _
M- n(0 Mo o
CM
U) 0 W) 0 0

C 0 0

C 0U0 N N
wOO D
< LU LU. 0) 0
E0 0 -) -n
CM V 0 o WM
N 0 0 N-

>-

LU
4<
U)


LU C3 CD
XL 0X 0
I- C F N N I-
a- N- N. N-
1U i Ui 11 111 LU
a: a: a: z z

- C CO N 0


- o r-
0 0 0 NO N
CW > 0 CD CD
<0 U ML 'L
z 2Q < <

M CM 0 M CM

N CD N N CN


LLU I-- IU-


0 CO


CD

1U-
>

0
N.


U-

>

CD


CD
N
U--
v
LL
M

>
N
CD
U)


<

.-
<
0


z





.1 m
x
3::






a.-
34:







*X* I-- -
CO
V) 0 cc


0
U
0 0
0 0 .-
-i -1 X
I-,-
I- F-
CO CO
< wN

0 0 W
o o -
000
000-





NM 0

> c- -


0 < =)
z 2. n,
N, C N N


C-
ui
LU
LU
LL
0










O

--
U-
CD














co


LU


o


CD
CD
4:
N





<
LU
a:
0
CD



cD


CD
-


4n
Co
N

N


<
0
C,
N

a-
CL
ar-
0LU
COD



LO
01


--o

--0
I .-

wc <
a- CD





0 C
N -
U M
o Mo
0 0)


w wn
0 U)
w- c
- N


0 0
LU LU

CD CD


I-
a-









o
0






In
a-
LU


M
0


0 F
LU. LU

N a- a- N3
F- IL 0-
LU LU LU LU
a: a: a: a:
0000
0000
co co c 0)
cocooo c
>0 I 0 2
- -N -)





em <
in -")c i
CD CD CD CD
U) N CD LO

N N N N

N N N N


I-U-LLU-
>m >- >-



a>>



0


>-

a-
CO

C)
01


>-

>

0
ph.


se
LU

-J


z
a:
z N
0 N

LL I--
a-



< 0


<
F- N o
F-a-
m 0U
~C0
a- O-
2S CL V





0 L
- N-




N
m m
a1: a: .

M U) -
I-
- Q- CO
LU 0U 0

CD in --
--00


o
F-
a-
LU
a:
0
CD
CD


o
m





w


M
U.
LL
01
0
U)
N
N



N
CM



a-
CM


0

C)


a-
LU

0

0

C)


L-


m




>
CD


M





N


I--
CD



I-





o
0


cD
CO
-





0
N


B:
,-

I-
I- -
a- LU
cc
0
LU 0}
SCO
h- < wU
F- -I Q
IL < 0M
m 0
0 LU
Ow
I- cc *-
0
X X -I
F- F-
i-
I- 0- CD
LU LU <

0 0 -a






0> 0> 0

CO O- CO
S>-
000
I-






1:1


X X
0 0 C







LU LUL
NC.D F-

CD CD CD

0 00








CO 0 0-
SCM) 01
CO00WG

000
U--







022

c >(>
CM 0 0*

W CMOC
N a- a-
LU L LO


NN ND


0 w
U)U)U)

NU-
I- N F-
U) > CD



CD 0)


F-
N



CL
N


0)
N
a-


CD


0 oN000 0 o- o 0 N oCDo 0 o0UC


LU


(0

0

F-
MD
N
CD


CD

(0
_q


0
CM
N
01
O*


LL


(0

0
CQ

0
10


C0
0
CD
N


U)


<
_1 LU0
-1 CO 0
mn - Z


0 0 0

I- C- -

NM N
iN CD -i


0
0

0
0
0
I-

in


0
0


0
0
6


00
U')
I-


0
0




0
F-

N


6
0
C.)


0


M
m
CD
0o
0n


000
0 0 0


000
000
Co CD C

0 0 0

I- I-- F-
M 0 (-
000o c


0
0

0


O
C-

CD


0
6


oC
0

CD
CO

CD
C-


0
6
01


0
m
CD
N
CD


0
0


0
0

1--
10
co
U')
CD


C.4


0


0


0






C)
C.
&.


0
0
z
0


ui-
1--
z
0

LU

N
0


0
ac
LU
N
LU
N
-I
LU
m
LU
CWaL
4:>-
N N
0


I
CD
<





L4
N

N
0-













o C9 0 N 0) N -- C 0 0) 0) 0) U) CD 0 o0 0! C '- 0) 0 0) CN 0 0) C
o U N C 0 0 0) N -0 N M) 0) N ) 0 N M) UO 0) -0 M D U 0 MD LO N
oCT ooor-me-so-ow a CO 0 co oM
O M L CDC oLO O -w -V NNt-NCD


ra:
LU
X
I-
LU
I-


-41
0-
O -
F-
I- a-
LU

c-

0)


- CN
D N







-4 -J

CM
AN








0) 01



0 0

CM CM


m-
a-
UJ
cc


I-
0



CL
I--




o
CL






0
i f-
0
o0--
I-
-CO
< UJ
aj -

o 0

CO




0) 0)

4:4:C


-4



w
0)
CD
LUJ
-I U-


f- I-
0 a-
I- NO
LU X CL UJ
U- F- Ca
< 0
U) JL; I- l
0 0
*- 0 **- 0
I- I-
I-- I-
CL Cl) IL co
LUJ < UJ <
ac .J Ca -4
0)00
- coM C

ooo o
0 N N 0)


0) 0) C0)
0 Cm
0 Ca I- CD
LU < 0 UJ

0)-0
0 S0 U)-

>- I-
f- -4
LU CD
U-
< X

CO LLJ

4: ><
CDoI-
0 0

In IR 0R In

LU LU
I- a- f- a-







CO 0
LM LU LU LU




0 --
0) 0) 0 0
0 0 >-f-


N U)
0M O ) -U)
- U) C) U)
CN M CN CN M


0)0)-
in n C

a- a- a-
F- F-
CL 0- CL
0) 0) 0)
U) .- i- r
N Nl N


X

I-
z


0
I-

-4
CD
LU



U
0



X
>-
0
LU
2:
0
1C



0
0





z


I-







I--
0
-4
z
=:




I-
X


I-
a_
LU
1CD
0
0

r~-



-
CO


0

C)



-4 z
<

f-
i I-
0 0
z z
LL. U
-J -4



SF-I--




LUJ J LU U

- -* **






CC CD co
z m >-


t-
0
-4
X
I-



I-
n-



CO
LU
cv


0




0


-4




Q
Co)
a


-4
0D


0
-4


0


I-









2
CC
F-



I-
a-

0
LU



cc










I.


0


000 0
<< I- F- I- I-f-
I- C L- a- I- a- CL
ZZ Z z a
c 0 0 0 0 0oo
U U) 0 0 0 0
vr N V, ND CD NO
-^ i- -o -o -
o o~oooj
N N- N, CO r- CD CD
> > CD CD CD CD CD




UO UO U) U) 0 0 0
CM CN '- -
CM CM CN N CNM


>-
CM
CO
U.

>
CD


>-
0
N


a-


CO


cD
<0
m
CO



CO
wq


co
C0

U-
O
0
U)


0)




OD
U0


v-


0) 0)

UO- U


0 h-


oo---- -o- --ooM-Oc- -oc-o-o-o-- -o-o--------O


o4
<-


0
f-I
0


0


o-
0-
0)


LU









0


LU




CM
N
F--

CD
01


zzz

U) U) U)


N- M M-
000
CD CD CD


01 0) 0)


F-
>-

0



- C9 CO 0
0C N 0 N
U) cD-V


a-
<
0
I-
LU
z
m

n
UM
-
U)

U)
0)
cc
a-
CM
<
0

CO
n
U)


-4


0
LU

N
F- 0)
< < LU
CD CD 0
I- -
0O

I-



UJ UJ <





- M CD

0 0 N
000


0 0 -


-4


I-
I-
a-
0
Lu

co
0)
m
LU
U-





I--
F-
-i


OU)

0
o





0)
LU
LL
M)


0
,LL
LU
0



Lc
I--
LU
F-
UJ

U-i

0


U-

u-
u0


U
c)






U


0
o



.<


-4





LJ
cc



0
0
-

COD
CO



-


C-
0


LI-
0
N

f-
X
a-
LU

0



CMj
m
U)

0)
w
N
0
0
CO
0
0
U)


FI-.- h-



n-f-f-
f- IF- F-
Sm mm
CD C0 CD
000
-J -1 -


I-f- I-



U) U) U)




CU C0 0

- ,- -)
0) 0) 0)


M o 0M
0 co 0


XXX
fI--f--
a- a- a-
IL LL iL
a: a: a:

moo
0 0 C
N N 0
-T -n
0V 0V 01
0) 0 0)

:LU
ocoo
o o
U) U) U)
U) U)w


I-
I-
OL
0
I-


f-
a-
IU
a:
cc
0
CMJ
0
0



CMj
0)
m
z

UM
m
***


0

Q
LU


0
0



<
F-
LU
cc



0
-



o
0)
z


0
0
C)
N


cc
UO
0)
L-


CD
CC


<
m
I-

3:

0
-J
**



x
F-
a-
LU)
a:
Q-
U)


0

CD
o
o



c)
0)


N
=>


X
I-











a--
0
0





U-
N







0)
I-
I-
a-
U)

0
Vr


LU LU
Q> Q

0 CD -


U) 0 0
moo
- 0 U)



0C 01 00


0 -3 -n
co n **



-^ -^ U)

moo
Ur) N N
CN CN CN
NM CMl CM


X
I--

LU










N
c)

-In




0
0
N
CO






0)
U-
0
0





0)
u-



in)


0
0
z
0

LU


z









uJ
-
0
LJ



0


F-

-J
LU

LU

I- -
0


I-
I-


0
UJ







I-NO
4:
CC G CC







moo
a-a-fo
F- CM C
LU LU LU




U) 0 0
o00
CV UJ U
- N

Nl N N~
0) M) 0)
= LU LUI
< Q 0






4:00
U) cD 'cr
0 0 0
000
c) c)
CMj CMj



01 0) 0)

F- F- =
c- IL >

0 1w CO
o *- f-U


<- CO CO
Ca N-


X
I-
al-
a:
0
0
N
-


0)
z

U)
N
0
C)




0)
I-
I-
a-
IL

UT


X
I-
a.


0
0
I

0)
z




LO




0)
f-
I-




r-


co
I-
I-
a-

0)
0)
0)
OD


U) U) U)
co co Go
u- F- U-
0)



T- a) 01
C m V
01


co
>


0
UO


UO
U-
co


>
0)
co)
co


n) UO

U--
e o-

N N
- N
CDO 0
0


C-





a0







=
*o
J


4



0

O)
I-
C-
a-
U)


000
0 0 0
0 0 0






0 rC oo


0

0
0
0
o


o0


0
0


0
o)
0
o
I-

N-


0
0


0
o0


I-

N^


0

0
0
o0


I-


0
0

0


o
0)


000

0 0 0

000


I- O- I-
N N N


0

0
0
o


I-
N
N


0
0


0
o


F-
CD


0
0


0


oI-


0
0


0


I-
0
CO


0

0

0



D


000
000

000

000

I- I- I-
N U) V
CD CD C


0
0
0
0)
0

0
f-


0


0)
0
0


CD


0
0





I-
0N


0
0


0)

0

I--
c
CD


0

0

0



,--
o
8

6
0)


000

000
0 0 0


000

oDC CD
.- -- h-

0) 010














PD0 0 C D CD rD -
CD NN C Djc)Cc


z

CD
CO

I-
z






I-

0






O
0.
(N


I-
I-

z
0






LO





c-
00









CD
U-





N0


a.
-Jo
N-
CL



JOa
X X


LU LU.
-0



0 U
Ia I


0 0

CI 0
CMC




0<
0 0
00
ar ^
0o0


0
LU



CC
0
0




IL



0
co
CM


>-


O
U.



r-


. . o 9, .


02



L J
0-


Q<
CO

0


LU
I--
O




= CD
0




I-
UJ


N o
X --



0-

C-C
0C C




0-








-0


_J _J




00
LU ILU

00

z z

N -



-iJ -J
o 2
S-j
0 O




0 1

00


001

z


0
0
LUJ

u. O-




00

CD 0


00
ON
0 0










o r
0) 0
Q


0
UJ
CD

I-
0

LU
cc
CC


0
In
0)

0

CD



CD


co
0-
0)
co
N1


CM


CM
>C




CD
co
O
A-


0O O


LUI
0
CE
c-3
LU
-o






0
0
0


z







I-

O


z
ui
LU




U0
0


z
LU
aC


0


I--

co


2 2 2



U) U) U)

000)

coo
CO COCD
't- U) CD
000
ooo


>-



CD
0

i-


1--
LU


CN
0O



CD





co



>-
0
OC



CD


















I.

0




N


a.

i--












i-





Or
o


-J


V-

m
*x F
0


X-
I-
0
-) I-
0
o0*-
IXH


I-
I-
< U
001



co co
Go CD
C-





LU
00







on
0
N C


< 0





CDCD
CD Co


ca i



N N
N-
I-
LU


0L


co
UJ
CD
0

NI- t
Ul LU
z a:
CD 0
T CD




CD C
<2
NO
co o


NM N


Cl'
5o
F-
CO
CO







X



1-
N




0
z





u-
I.-




cc
CD

UO

OD

OD
0-
LU
CD
CM
N
N


-J

SCL



aL.
0
CO






LU<
CC -
aJ X




0

I-



CD 0

0)
O W

z z
< m



O
-O- 01









0




LJ
U-
CO
00
UJ UJ
0 CD

1- F
co o

1m

UD LO

7: a:




0c CD
1010




C0 CM
Oz
ON




NN


X
I-
0.
LU


0
0
CD

co
-j
CD
N)
U)

in


0a a: 0a
000


000 D 0)

OCD C


01




CM
rU
N


U-


0
CO
T


0
0)
U-


co
CO
vD


CL

0
I-






O
.o
C0
0

N






0


I-
0
U-



X
--


-J

QE


O.


I-
a.


r


2
W
0


-J






X
I-



0-
LU
CD
-O




CD



LO

0
>C


0


LU






0
I-







L
NQ






X -J
I-



<- 0
CDO 0




a. CM
U0


00

0 0
I-



-0



oNo
COwO


LU
a x

N a



0) U)






<
oN N



O N
CM CM
CM N


a: a:
0 0
01 01
U- F-
2 i-
>a 0

CO 0
0
U) C


00000-OOONO


000




N-viO


0



CD
i-
C0


D
N
a:

0


N
0
CD


CI-


O
a0





co


2
E

O
I-
a:


0
CD
N
CD


O


N
I-



CD


I-
aL
0


X
I-




0
CN
0
CO



co
0o
CL


0


z a




C- 0
I-- In






CD cD


oN
o 0


CM N


I-



CL
0




cI-
a.
LU

0

0



COD
CM

U-


0
UJ
CD

CC
0
LUI


0




m
uj
LO
O
0
N

0



CM
LM



N


Cm CD CD
U N- LL.
2I-I

- to CD
CM CMj 'c"
LO 0 0CD
0

0-0




JU uJ U

-1 -I -1



CM CM CM



00 0 0
CM CM m


000
No N

CO 00


N C C


-I-

-J
00





-z
0 Z
aE



UJ
0
0 Z






O) CL
0 "
LU 1-





-JX



cc

00
0I0
0 I-






0: CD
I m
Suj
O N
0 -


X
i-
LU


0



0
0"


CD
UN

N0
NM


-J
LU

CL
s
a:
C.


CO

0
co
z

N

0

CD
0
N


a 00
cc CO CD







01
C\O


LJ
-J


_~j
-J




CD


CO


0
0


0
CMj

CM


CD


I0

CD


0
z
0

LUJ
1-
2
0

I--


0


C)
CC)
CC

z
C)


0
-C)
C)

.2

0

0


-C)
C)
CD5


ac
I-
0


I- i-
-J
LUL
m







IU
I-
0







F-<


0


C.)


*5~














0000 0U)o0CoNN0o0o0o0o0CD.-o0N00'o
O 9 O 0) K r. m 0 0 w N r 0 0 m 0 Ot 0 U' 0 <0 C 0 CM CN O 1 0
w* mo r, o o ci Oc'3 CM m '' -: co o o; cM V; c4 m c r- C ci n m\ 0 m) C 0 V- Ln- '- r--
01 tD LO -^ n

z z






o w
t- I-
z z




- I-






I-- -
<
Q 0

n =
xx










0




0- >

03 o


-J




LLO












>"
V-
LU








Ni




O1
l3







0 0
0




I-- I--





0 O 0
0U0

O 0O 0



0 >-






--
VU
L1E
0


_J


I-.

0
LU

0o
OC
0
UJ
0
N

*--









CO




0
10
C0


>>. l

u- -




00
C)


0
UJ
LU








0
cc


z

D


a.

LU
cc
0.


Ur)
0
C-
LU
0







0
0
-j

C.l
-J
to

0)
U)

0

m
0
=0


-J

I-

O-
.-
0


0
0)
z
0



0O



-I





0
0
--




.C
0
co



z
-


I_

0



I-
a.

0

0
N
0

U')




n
LO


--i -i




0O 0
0 U ,,-

L 0 u0
0 0










S0 0)
S0 0









0-C
to0o00
<0<0
('0i -
- -





0-N


_J


I-


F-
F-
0

;-
I--

CL
0


0

..
0
0

0
m
LL
N


_-J


I-
I-
z a. I-
0 r
00,
I- w


>- 10
0 0


-
S0

X -I







0) 0

I- I-





uJ N







OF-
0 01 0


,w w? =.
< -1 ->








'00


-0

0-








LO

- -
It I


>-
0

LL


O0
LU


cc

- LO

I- <
O 0



.o
F-
0- V



0-

0 0

Nco


I--




-J

F1
-
I-
;a


-
IL
>-
OL

U)
OR

0
0



0


x



1-
0




CL
Oc

0

0

0


-J
I-
C-


OL

L




0
0
0



co

0


IL
0
0



0
U)
m
LU
CM
'-


LO 0 M 0 OM
U) U) U) U) U) U) U)

NL u- N N U- N-
F- = =- F-i -
aL > CL a. aL > a.
0 M 0 M 0 0M
0 N 0 0 0M 0
0N 0) 01 0 0 0


U-
0
IO




10
V*


F-

m


F-
CO
J-



-J
I-





0

01.














I-
aL
UJ
0
O
0
O
0
0





N

N~


Q)
ca





0





0


>- >-


I-- uL U-

a o>>
V00
CD N
0)


0 0 0 0 00 N 0 0 0 V O 0 0 0 N 0 0 0 0 N 0O V 0 0 0 0 0 0 w 0 0


>- >-
00
-I .J
-J -J


co c
= =


0 O
I- I-
*- NM
in i)


N
N
L-





CO
0





F-

0
-


I-
u]
N




*f--
0
I-
10


F--
LU





I--
-
0
0


U-

0





-


F-






"-
D
0*
NM


0 0 0
I- I- I-
0C 00





U) U) CD


I'-




>-
UJ

*



P

CD


F--

LJ





-


F-
UJ

0
-


CD


I- I- I--
SI- -
00 w






I- t-
AA in c

N N N


0

-





CO


X



0
0
0
CD




-
-I

N
0
-


I-
I-





F-
a.




0
0




z


0 0 0 00
< LU U LUL UJ
- X (0 X0 0 C
UJ < <<<
WJ CL I- 0L I-
LUa. NC a. NC NC N
U- 0 0 000 0
0000U000
0 0 0 0 LO 0 0
0o 0 cD U) 01 N NM

0 0 0 0 0 0 0

cczzz)om
< 7) z3 05Z U-
iO wo (D mn cocD
0 0 0 0
o oioino 0 0
00----
0 0 UO U) 0 0 U)
N N N M -
N N N N m N N


I,
>-
0


0
OC
LL






t-


I-

LL,
0

LU
N





LL.
L-
0


z
0
0






0
I-



0


cU
I-
Iu


-J


F-



OD






0
Lo


z
cc


I-
_3



F-
a-
CL


U)
CV

0

..T


0

0

0


z
=)
0
M




1-

-J




O


0
..




z
CO


X
x
a.
OL

0


0
0
CO
L
a-
w
0
N
N
0
N
N


0


I-
I-
L
N

01


x
1-
0



i-
.C



0

0



0
a.
Ln


x

CL
UJ
0


0


O
M


CD

0
CN

N





O
0


F-
Q.
OL
0
I-




IJ
CL

0

0


N
0

0
CL


CN


X

IL
I.-



0
0
OU
CC


N
0

O


0



CM
0
C-
0
0






0


0 0Z




I--
U-




J <
0 X a


UJ UJ UJ



-O 0 0



z c z


0 -


x
k-
a.


0O



N
0







,--
o)





0
0


F
CL
CM
OC


0 0
FM 0 -
00I- U IU
<
a. N- a. N a


00000


COCO 00C
0 0 U) 0 0
0 0 0 0 0

- C CD -.-
M0 0-
CD 0) 0 ;;
>>0a.N
0 0cCLU.
ZZ N 0 0 0 -
0 M 0
0 0 0 U)O M
N N N 0 N
N N N N N


-J
LU x X
CC I- I-
OL OL CL
< U" UJ
0 cc
000
0 0 0

- N N



=oo
NW
00'-




0 0 0
000
000





00X
rV 0 0


0


u.-
>
C0
0


I-
OL
N
0
0
0


U-



>
0
Nr


,- U- -


i CDN -


U.
U>



in


-
UJ
0
LU
a .
<>-
N- -
0

I
C5
<

M
0
00


LUJ

m
<


0


u-




10


-o





0
C.)




:6


>-

I--
0






0


>- >- >-


-J -J -J






I- I- -
000 Wt-
XXXi-





I- I-N

0l 10 (0


0
-j



-
x
No
0M

0
N-
No


<
>-
-J




m
O
-M
0
NO
0o


>-
-J
-.

No
0M


N


<
>-

-J



F-f
0
O
N



0)
0o


<
>-

-J_1



F--
-

0
N


00

0


>-
0
-J
-_
x
CO
O
0M

oJ


-i-i -i
>F >- >-








-- I- --
-i j-I-
h-( -4 -
XXX
0o 0o 0
NM NM NM
000
0Q 0D 0
NI-N
0O ^t U)
.f Tl. .^


<
>-

-J



-




CO


<
>-
-J



F-(
CO

-
N
0M

0
No
0<


I
-J
-J



O


0-
0*a


>-





F--
v
O
I-

r'-,
0T


<
-J
-J


0





-)
co

I--

g


M













C' ) Q- Q f-o)0 0o
- N N O 0) CD 0) N J
-0)000)00)' N


0

Tf-


0N0) 0 -0
4. ) TO 4. 0 CO C)
O CN- C 0) CO


C 7 9 9oc csio om ot cMi oc r
N, V cW C') N 0)M NCM M) 0 0 (0
M )U0)) M M M MN N NCDCD0C
C'0


I- F-
NN
NN
a- a-



wLL
00
N-I-



a- a-
LL
CJ CJ
00

00
0 C






z22
(0 r

00)
0II
22 i
U)oU


<
m

C-
L
cc
i-
0


0)







0
-J






in

<*


LO
CM
01
03
4.
o


I-
-J
0)

N N N
n x
OL CL CL


0 0 0

0) co

- N
0) 0) 0)
-I > >-
O 2
-0 0




OCr Cc Cc







oCD ( c
c' 4. om
0) rN


Oc



m
Cr
0

F-
-J
0c
I-





CM
-J

LO




01
CD



LO
U)
N



LU
IL
4T
C')


CL
a-
LU

U0



N
0)
0
2
0

N



0)



0-
10


XJ
h-



co
O




N C
N 0

X I X




0. .0 0D


tno (1D
00
I-






U) 0 U


N
N
0-


r-


I-

LUJ
Cr
0




0
2
n



V)
0

CN
NO








N
a-

N


I
0


x

0





0_
a-







0



0)



2


m
0
0
LL








oc
ULL
0




F-
UJ


UL

Ui
0


UJ
N



LU
0L
0


F-
N
;E
CD
4:
0)
-J



0L
LU



C)





>
0
Z
2
CM


X

I-
LU
CC
0
4.









0)
CD







I-
N
0)
0)
0)


4C


LU








z








F--
-J




I-1
N







-O
0)
c-
CL

<)
N.









QN


X


F I- -

-- ..

C. CL 0L


-J 0 0)


CM 0 LO-
2NN









00 a00




cD m >-
= LUJ <
< UL S
Cr CO CM
-00M

00U
004.

N0 -
0)00


X


0
JL




0

00
CN 0


O 0
IL Q
LUJ U



0 0



00




N o

- 0)
CO 0)
0)


0
LU
0


0
CC
0

CO



0
cc


0
c)


0
4:
<

rM



u-
N
C)
N

0


01
U-
=

0
0)
r-


x
X-
CL
LU



u
Cr
10





V
o)
>
0
2
N









I-


N
a-
0)
4.
C')


U-


X
LU







0 0



OU U




O oc
-- I-

S- C.J
Ma 0


0



0 0
0







C>Z

o 0



00)
J > z
<0 <
0) CC00


z
LUJ
0
0
i-
-i
UJ


I-
-I


i-
0
cr
0

0




0)
0M
21


x
I-
CL
UJ

0



0



CMj
0)



N
LO



0
CM4



0

LL

>
0)
CD
co


0
CD
0
0
I-

N
LU



X
CC






0
0
CL
0
co
M
r~-

0
a-
ClM


Nr



01
0)
01
0)


v
0)
Lb
=

0)
Mo
co


I-


m
-I
0)



I-

OLL
cU
CC
0
0





0)
N


U')
10l
CM4

N

-

LL
=
>
0
m







0
C'




U)


F


I-



L

O
N
0
CM
0
O
0




C)
LU
U-
0


>-
I-







N
LU
L>
<








V3
0
LUJ
X 0








NO
0--
LU LUJ
Cr Cc
00
0>0





0 <
z2 -
CO 0
CN CO)
U) U)
CM4 CMj

CV CMj
>- >-
CV CMj

U-
N N
NO o
v* CMl
0 01
CD


22

xTx
I -
00
Cr CC
NM CO)
2 -


m m
= -I






00
00

< <
CM tJ



-N O
0 0

0) 0)
Z-j
=D
-j


x

I-
LU
Cc
0
0







0
UO
CN






I-
0-
0)
0)
C)
N-


0
LUJ
(D


LO
0

0
4:




I-
LU





CC
10





UO
It
(0






0
N.
4.




I-
CM
U)

C')
10


X
h-


0
0





_J
CC



.-




0




0





0


01
CD



0)
0


x



0~
LL
CC
0

<*
o.

O
0
cc
V

0
z
<
0
m'
UO


>-
CMl


IL
1-
V
N


c)
P,


0
LU


LU









2 ..J
0 0
c- I-
0




C-






01 cN
0 0







0 0)
O- N Cc







U-


c c1 0

>-



CO













N) N 0)


CMj
C-

I-

CL
r-.
C')
N


N
>-


U-




w-


0
0)
CO
Lb
=
>
0
cM
UO


-J


I-

0


0U

0





CM


X
a-

LU

0

N






co
0
0
CM
aS




0
0







0
-
co


C)
CO


C)


0

C)



0
CO
.4.


0
0)
OD
Lb
=

0
D
N


ON0)0)0)0-NOC')4.U)CDOOO-00-NC')04.0U)CDON0)0000


N
N
LU
m

0-
-
W
4.
N
0
0)

CD


N N N

LU LU LU


F-- I--


^- ^- -
M CM M


4. 4. 4.


0) 0 <-
CD N^ N~


m


F-
-
4.
N


>-

LU
m





..

I-


I-
I-
LU

.



I-
4.
N^


>-
I-
LU
0)
4.

N


I^-


FN-
LU LU
UJ UJ
m N



00
0) 0)
I- N-
CD N
N N


1-
Lu
m
I-
LU
0)
*4

N
0)
NO
0)
No


u
m
I-
LU
m
0*
4M
0
I-
0)
N~


>-

UJ





-
0)


0

o



0
-


Q
0
0
0



03
0)
co
N
N
0)


000
000
000



000


>- I-
O)0)0)
N o N
C') 4 U
0) ) 0


U

L
CD

0
-

0)


w
U
LU
w

CL




-


w
U
LU
w
LU





a-
CD

0
0)


w
U

m
L



a-
CO
N

0
0)


UJ

UJ
LU


-


0


wwww
LU LU LU UJ
LU U LU LUJ
2222
LU LU LU LU
CL CL IL CL




a- I- a- a-
0) 0) 0) 0)
I- N
-N0)4
0) ) 0 0


LU

m
LU




a-


w
w
L

CD

CD
N
0
0-
N


LU

LU
2
w
LU

r-
CD
N
0



0
N
N
C0
.-
8


w
LU

LU2
6 -j -j
LU LU UJ
CL m 2
CDNN



- N
000



0) 0) 0
i- *- CM


-J
m



N
NI


0<



I
>-
<
0
0
z


F-
W
LU
Lb


<
LU
S0 N






0) C) -



LU U <
CD 0 CD
0 Cm 0

N CN LO







I- =N
m > a-
- 0 U)
v N v.


0

0


2
I-
2
0

LU
4:
2
0


cr
LU

LU
S-
-1
LU
0)
tU
< a-
4:-

0
0




0
0)


LU

z



0-


d)




0
0



I'4














009--CDO
0 0 i -0 CM -
01 UO ) CM
*- CN


t- V O 0 0 N M C N 0 U) U- 0 N CD NO U) wo wD N 0 0
ICDCMDM N D0 M 0" OD r- co N
N--


0 _1
<
z



o <
-CC C)


0 =
0 LLZ
I- =

8 x

0) 0
o *- *- o
F- h-

< uj >u <
ED
--CD-





0 0 -

CL
0000w


0 CM 0 0


0
LUJ -I 1



4 <
C 0oo






aC)C
000
0 N CD






N 0 0
-0-

N N K

aCD
CD ZZ
0 NC 0


U- U- U-


= = =


x

a.
I--




o
LU



cc
0





m
0









-
0
Co





z
0
N
-
CL




u,.


CO
.4.


4

I--

z
Lo










0
N
U)
>-


0
0
0
0
CD





U-
=

0
U


z

0)



I-


0
--


I-




=D


-

0
o


CO





0
-j




0
CM
>-




4:


-4

-1





I-



-J








a-
x

LU





0
U0
CC
(D




0


cD
4:
cD
O
0


CD

a0
0
0


C0
Ln

=



ID


0000000o00000000-CM-O- 00ON0 o000-0oO


----ww
CO N CD 0 0-




- N- -
N N N N N Nu


zz



0 N CD wi




00 CD 00 CD

CO N- cD 0)
CM M C C
CM CMCM


-J




0



0
N


CO



--


mmm


000
F-~I
CD CD CD
Nma

CM CM CM
ocer,
cucuc


I-
<
0









I- -- ,x
m








z x
0

-






So
- --

moo
I-- o i.-



PCDC
No











- C CM
---


N N_





>C)C


-J




I--

x
a
0



I--

C)
0






o
-J



-J

0




CM
-
0
0
N
N


Ll

0C




0





0
O)
F-


CD




0

0
0

LUJ




1C -I
a:


--o




CM



00
o 0
00
>( ,


-

x-


o o


a 0
N -
0 0
O.N
0 IL CN
co N
0
-- N
0 0
-j -o
-o-






I00
-i cc -i
S0 -







0) 0
< --)
V w

No-
mf-0 0

10CO 0
I- CM 1
** C

CM T


-1

.-
x




-


0





CM
0



<
N



o
0


0
-j
U-


OL





>-
0
m


w







0
=
I-











L






<
o
a
C


o


ac
0
F-


< -J
SLU <
o C.

-I -1 I-
m m
OmF-


I- I- I- I-


c cca: c a: cc
C0 0 0 CM 0
- -- -

o0o 0 0
U= in) N 0 (o


0 0- 0M 0 0


<< 0 0
n) o o




u-


aC
LU
=r
I-
LU

L-

z
0
I-

x
I-

a.
LU



O
cc
0


0
0




z


N


X
x


LU
a
0
O
o
-

z
0
O

Z
3u


z


a:
1=
LUJ
I-
I-
LU
m
0


x
I-

L




c
a
0






N



0


0
LU




cc
0
o
03



z
=a




N
N
2
in
CM
CM


x


LU

0
1-

x

CL
LU

0

0
in
Co

-
0



CM
<


0
LU
CD
a







-
UJ










o
0
0
z


0









u-




.4.


.J



--













-i
CC
0






-I
=)
0






--j
o




-
-J



0



co
0
-J


0
LU
Cc




0
0
o
o


0
-





0
o
CM


LUJ

OC
ui
0
CO
z

z
0






a:
0


0
D
I--







0


x


LU
o
0



-
CO





m
<
0


m




-
--
a.
O
0
Uo


a.



03


-1





LUJ
cc
a:











=:


x








-
a.
a:
O







<
in



cD



co
o0

cu


1--
I-


R:





CL
LU




0
m
O

-

0

0
CO


x

a.
CL


0
o
m

-


-1

D
-
0
N


000 -)

U LLu- LL
o10 o M

o






00
!2 <



000 C







- -
0)- 0) CM


000
- N
00 0


x

m
I-

LU
aI

0
CO
0
o


-
m


0
>-



o




m
In
o


-
CL
0


I-
CL
0



a-
X

L
LU




0





a-
a
0


0




oL


o
0


o
0
0

o



m
N
CM

>-

-


0


0
o


z


I-
0

LU



z


F--


0
o
-j




-i










m
0


'L.




r
-


LU
0



cc








LU
0
0
>T3

0






0
0
O



o


U-
>


CO


<


I--


0
0
-J
oV


-i
N
co
-J



0
m





a:
4:
N


0
LU


Cc


O
--
a
U)









,
0
o
in




o
0


-
=


.4
N


I-
a.
>
o




IL
LU

o
cc


0
CM


0
o


m
0



o


0
UJ
C9
0

I--
LUJ

0
o
C
N

m






o
0
0.
<

o
CM
0
0


m
0


-
U.
=


0


-J -J
0- a: a:
LUX J
c) 0 0.
.- 0 0

" 0 C


- -
O*L 0) 0
CU 00
z >- C

0 N -


00




0 CM N
co
LL. U-
SCom


CDCD~


0
LUJ




-
cc





0
0
N
CV
0
U-
o


In
o



U-







.4.


a:
=J

LU


x

aI
X

hc

0
-



z

m
0
CO

t
r-







o
IL
=>
a




0


X C X
0





an o ca


LC O. I0-



0 0 >




00 0
0a








0 0
N CN N
00>








> > >

cea(D
ewr> o


<
CD
4:

L--
uJ
z

0




,0)
Fr



0

0


CO
CD
0
N
N


o
0


-


XX-
F- W
LUJ
IL a: U.
O UDJ OC


oo o
o co c


0 0 03
- -



0< <
CO 0 N
U) U) 0
nCM CM


t- I- -
---o







03 0 N


N- N rN N
cucuu m


U.
>
0
CO


U.
>
0
r.


-o





0
CI




P)
a


a


U.
>

u,


U.



0


SI-
-J
LU
00


Fa




:-
I-
CD


o



>a.

re


co
-J



0
m
CO






CM


. -I
L L





ma
I- F-
N CO
0 0
N N


0
C)
0


m
I-
0


CDC-ODCM
0000
0 0 0 0

0000


o o oN


LU LU LU

000

0 0 0





0 CN
rCMM


0

4-
0


CM


-


0





L-
CM




4*
-
mD


<


t-
0



'--


<




o
I.-
0
CM


N
m

iN


4
z




F-



N
N

o
CD


4



0

N


m

r
N
N


z






M
o



0
m
LO
N

cu
U)i


0
Co









C
C-













C N CI N- CI C CM M CM 0 N 0 0 WD CD U) CM N CM N CNM 0M 0 m
CD 0i CD o C CD U CD CD U) N CD CD CD C6 U 0 0' CD U 0 C' -, 0
N 0UCNCC M '-N 1 MM -crCW CDc) ocD 0 co -


-4



N-
0
LU 0
a: LU
cc
CO
I-- CL 03 )

0 0
r- _|

0 0
0 0*-
X-1 -I X
N-- N-
a. CF CD o.
LU < Uj
a: -j -I a:
LO i- *- 1O
U)-U

0 0 0 10)
0 0 0 V,

- CN
Cm CD I N
Z CDL x 0
=3U < UJ
(CD UO CO
0 0 -


I-
N-


CN
N.-


U.
>
0
CM
N


--
0


X --
uJ CD
N-
0
0 LU
UJ ..
cc
CM-
co 00

03 C



00




-4 -4
*- N-

0 0
0 0

C N-.

CD N
CD CD
mc m


0-


F- W-U U-
N- N- = =
a. a.> >
CD N- N-
CD CD CD
CD CD U) CD
0CD03


Z
LU
0
m
a:
m
N-
-i



N-
_-
CD
X
N-
0
LU
cc
0

0
0
03


03
CD
CL
<
a:


-4 -4



- -
N-
0 0


CD-
~LU 0
03 0 1--





-J _J 0J
NM U. CD
-0-

000
000
CO C C)


CM 0
F- 4 -




co !2 co
<<4<



00N


N -
030 03


Z CD C-
m UJ 0
3U-0
i- o -


LUJX
"-> h-
0
C0D Z E
LLI LU c-c .-
LUJUJ a: 1-
cc cc
5 UJ
cl c) =>
C--0) 0) -- 0
H" z =r -j 09
< < L. N- UJ UJ
CD < I- < _j
CN Co CL
0 -I-
*_ _j*- *-_N--
00 m m
0- o*-
X -I .J X X X
--CL co co --L 0
UJI <: < LU UJ LU

O o o o) -
N- -N
LU0 0 LU LU LU
UM 0 0 U)j U) U
C- 0- 0 -
m m o m m0 m
CD NM *-
NM NM CO CD (D CD
> 0 a: z m M:
0 LU CL < LU a.
= 0 < L) U <
0M cD N N N-
N CM N

N-
LUJ
U.
<
U) I.-
a -o
0 0 CD
LU _1 LU
0 X LUJ 0 X X
Ng a. 0. I- a.
( F- CL IL >-CL IL
UJ LU < LU LUJ LU
a: a: 0 a: a: a:
0 U) 0 0 U) U)
CM CM CM CM N -

N- CD C- C- N 0

CN CD Cm CD CD C0
0 0a 0 0m 0m 0
05 > a: z z CD
=> 0 OL< < U
< Z < L3L3U
N- 03 N- CD -
'- 0 N CD -
0 0 0 N N CD
v v CMl CMJ 1O
N CN
>- >- >- >-
ULU. U- U. >-
CmCm CO CO 03
cD c -o


U.
=
>
CD
U)


-4
XLUJ
N- =>
0 co
LU LU
a: a:

0) 03
> CD
0 <
z -D
0 CM


0 0


CO) CO)

*< <
1 0





V3 03
-- -
N N-
0
U) 0
U) N
N N

co co







-j sU
UJ 0
CC QD
OLI
< UJ
0 CC
0 0

-< -^

O N


m Cm

10 0
0




C0 0
LL LU
co CD









v. v
- --
oa:*V


>-
N



a.
CO


0
N..
CD
N-


_-j
LU
a:
aL
<
0
0
0


CD

N-
03
I-
0
0
r-
0
CO
U')


>-

N-
M

I-
F-
I.-
CD
CD
C0


CD I 0 U U) 0 0 0 -


a: a:
< <


C3 CD

0<0
m m
N- N-
CDN-
NNc


a:
0
<


6

m
C-
03
N-
CM
N


t<
<




ao
4:




03
Nq-
CM
N


<




O
-4


m
0
CM


<
.J

-4
03


to

CDj
N-

U)
NM


UJ



-I- N- CD

03 0 CM


CD CD 00C
I- 1 1-
CM VOq 10
10 1O LO 1O
CM CD CD CD


..J




-cc


LO
03
0
a :



)-
N-
(N















O 5]
CV
0
0
0
-4-x
N-

10 CD



















O CO
< LU

>-Ia
0-

0Nr




LOUD
U) U)
U) C
CD uC
< LU
N-
CNO


LU



1-CD
OLLI
0
LU



~00
xN CD
a. N
LU LUI

0 0
00




0O CD


COi CD
LL.L
N- 0
- N1

CD N


CD >- >-



- U- -
>- h-
N- = N-
a. aL > CL

U) U) CD 0D
0 M C 0
N0C 0 03
CMl


LU LU U LU
-*~ ~ 0 JU >- > -
CO D5C CD3 L- -
wms mw w) >> >-z
LU LU LU LU wD wD 0LL:a:

NO NO NO NM CD co D cc C) (D co o-

C00000000000C
CD CD m m CD CD CD CD CD CD CD CD L1L


D tn CD C D (0 No CD U) U) U) U CD > Ci CD CDj CD CD CDJ CD CD


>-

0
0
Z


c o- N UO 0

c D 0 N- 0
N- CD-CDo


0)

*-


UJ LU UJ

aL aL a.
< < <
000
0 CD 03

CD N


0m 03 03
CD Z CD

a: z a.
< CtO
N- CD -
0 0

U') UO


oc
a:
U.
0
0
CD

I-
F-

r'
N-

N-


I-

N-


0
-_
X"
N-


a.
LU
a:
UO

0
'3r

I-



0

CO


N-
I-
aL
0


X
N-

CL


a.
LU




=
0
CMj
0
CD
N-

N
CD
0
z

a


0
0
U-
U-
0



cc
LUJ
N-
LU


U-
LL
UA
0
LU
<
U-
u-
0


0

LU
0


I-
I- I-
-I
LU CD


LU c :
=D L

U 0 a


MD X < z
I- 1. Uo
UJ CD

Io 0 0 0)



00
0 I--
x i _i
I- N-
a. 0 ClD
LU LU <
cc cc -4 -i
0M 0 0 0
- V~.~
1O 0 0 0
*- CP 0 CM



CD CD CD U)
01 0) 01 0
CD CD Z >
U U < 0
LL U. -n Z
in Mo Mo*
U)C C


0 0
LU
O- CL *-
X~ 2
I- Z N-



t-
0 0






I J UJ mUJ1
L UI LU LU
cc OCC a
0 X MCD







03 0 03CY
0 0 O 0
0 0








X -I Xi >
0 0
F- N-











CL _O S C
UJ< UJ 0 N- N- 0
WM U<
0- I 0)03










zr cc o CM
0 0 N- CD











>0 V0 10 M
03 03 03 0
Z~ -I -I z










0) C) =) 0)
LU
cc








< I' UJ (3
0 wC C

Z~ CC4 0
0OD 0 C 03
m -m 4-O O
CMON-a CO


0 a: 10 a:
U) 03 CD 03D

0M M O- CD
NO0 0 0 CD
U) U) U)3 U) 1


0
0
z
0

LU

L-

0

LU


0
0


<
<; 2
X
0
BN --


0
-4
I--4
N-
N-
CL CD
UJ <
CC -I
a0 1O
n- cD
O 0
-0














,cO ,-
T CD
CMj 0



NO




O -4
LU><

a: IL
< LU
0 a:
0 U)


U) -)
CL- .c.


CDJ




0 0
UL UO


00
CO
cz

x

Cf



U
CO


o)


-1
UJ X

CL CL
< LUJ
0 a:
0 U)
N V

CM CD


N-

03 03
N N


a: a:
L UJ
E0
CM T0

00a
mo m

N~ 04
cc. a:
U-LL.

1- -t


-1
LUJ
Ea
0C
<
0




0)
U)






CM
N


N


U-
CM

CD


a:
UJ
N-
Ii]


CD 0l
< LUJ
I- X X X C0
N-N CD
ULJ <
WU a. CL a. N-
a: W LU LU LU
LL Ca ac cc ac
0 0 U) 0 0
CD CD 0 0
0 N- CD -

N N N CD v
03 03 03 03 03
CDC CDCDCD

UZ) 0 C C
-- 0 -

0 0 0 0 CD
v~ v~ v~ v~ U')
CN N CM CNN
>- >. >- >
U- U. U. U. U.
IT 'T V,* v*^

N- N- N- N-
W 1- F- I--
> CL aL a. CL
CD V N N- N
N- NM 0 ^ 0
CD 0 T- CD NM
v N N- 0

S VC


-J
UJ
CC
IL
<
0
0





CO
0



0)
03
=
<
0


0


-4 -
LJ LU
aJ a:
a. a.

< <
00
00
00

CO 0


U) UF)
0 0

C CD

00
U) LO


x
N-
OL
LU
cc

CD
10
o-I


03

0)
CD
CD
LL
10
U-

0
UO


I--
-4
UJ
CD
0U

N- N-
0






z
N-


0
co


LU

z


F- 1F I-
CL CDCD CL

CD CD CD CD
0C N- 0D N-

- N N


N-

N--
3a.
N0
CD


-O


0J

0




a)
,=
=o


C3
N-
N-.
h..


N-
I-
a.
N-
03
03
03


an
N-
N.

U.



(0


C0
-D
LL

=


>-


N-

N

N-
N
v


ac cc


C CM -


*- -O
CL CL (.

v N- U
O-NI- -

0CD
N O N-


>-


N-

N
-
IL

N

0


LUJ
z
Z
CO

CO


cc
U.
CD


L-
N-


U)
C0
03
03


UJ
z

CD
z
CO


>-
U.
CD

N-
N-
CD
Dl-
CO
03


UJ LUJ
z =
--<
--4
Z Z
CDC D LU
1z z1
X X aI


LU
z
t-r

=
CO
C(O
cD

0
CD
N-
(0
CD


<
-4


CL
a.



0


N-
CV
N,
N


CD CO CO

000


N- CD CD
rO co 03
CD CD CD


a:
0

CD
CD

0
-


co
M-
CD


<4:
< <





co m

V- to
N^ N


CD


m
0
CD
N-

0
Nq


N-~
CD
0
CD
N-
N

CMj


>-
UL.
CD
N-
CO
N-
CM


z




0

LL.
N-

CD
Co
N















o O o o -
TO O 0 CO Co CO ) CO M 0 0 0 10
- CD (0 CMCM-O1N. *N- 1C O CO CONCO 0


I-











X-
Co
,,.1




I-
-J




0
a


Co




L.J














r O





















Co0
0
-I-

























CO CO
0 0




m o
-0




















a- C
o U
co
S-






S0
aU)<





o in



0co
B -e



COC0






0


I-N




CO Co
0C CO


>




0

--
;r
M


-J





I-
-J
0

0

U.
M


-J






I-
.a. <




Vm om
< L
a:s

I--
0o0













I- .
C N C)



cc>- c


I-




O QC
<

L -I
occ

LU oj
0-W










) CM) 0
0 0





CO CO CO


fr-
U)in
i-
a-


Co
01


z
I-
co

0

U-


o


I--
co
U-


CM


z


0o
0

I-
M

N


z



co
0

U-
I-
U)
r-
N~


IF-

CO
0

U-
Il-

cli
M
N


a:
CLO 0




>L


)0
0N0
NCO
r o


Zo



I-
00


U. U.

N f-
NC C


I-

m

0
I

X
I-






O
Co




>-


cu


z


I-

LU
Co
-J




X
I-





CM
0





0
O




P,.
o


F-
I-




m
I-



Co

UJ


Co
0



















co
0
0

0






0


I-
-J
I-




L.
LU
a:


in
in
^r


I-

m

0

I-
i-
a-





0
IL





CO
o
o
(0


.J

oB

SJ


O CC 0
o in o




0 > a:
0 0- C



O CO O
0--









cc > z
ONO





->-
-: -: (


I--

o
0
0O
0O
(J


Co
Co
0




u.
N
h&-
r>-
CO


z
Co-
Co



Z
t-



I-



Z <
D m
z











LU LUJ LU
ca Ca a:



000
0C 0 0
m M CO



,J r

o CO
0, -
o.....


X
I--
aC
LU
w

0
0
CO





CO






Z


I- I-N- -
C C CL COL COL


(co 0C CO 0) 0
cacaLLA


Co




Co
u_
O
aI


0
Co
Co
0
Oa



a.

I-
Co
Nu


Co
Co
0
N
O
0
u.
--


co
N


co
O





c
0


C.

N


0
Co

0

u-
0M
U.

N
.5
Co


CM CM C r 0 7 0
0(0C-OONCM


-J







< -
m 01




0
*- :- I-
0 0
00*-
-I -I X





N
ND 0
- N


C 0O CO




> > CO
0 0 UJ
z Z U-
N rN- co
0 CMJ -


.J


I-



x


0
-j
F-


O
0
-J
I-
Co





.-
co

CO
01


I- I- N-
L a. CL.
LUJ LUJ LU




CO











0.-I- --
O- -.



NO O


ZZZ
NNN


I-I-


'a N


000


000

LL LLU-

C CC N
N N


x
I-


CC
0









N
a
CO
CI

>-


CM


-j
LJ

CL
0
0
0
0







CO
L-J
N
m


a-
Z
N


NI-N


CO Co 'r N.
Co co o) CO
CONmC
maoWh
aWoaY
(ooG(o
oe ~o


Co
0
O




I-
c.

Co
co


o
0


0


CO

CL
LL

CI


CV

C)
0

N
0
a-
U.
IL
0
CO
N


J


N

I


oL
au
U.


a x
-I 0
0 0 X


000
COOO








- N (N
> CODC
0 LUW U
Z U. U-
N.. rN Co
NO-
NOO



ZZZ
N N N


0 O
Nh
>O C
rP


-J -U


ZZ

o o
o o
00 CL
LL. L


CO CO
N N


LU








IL
U-
N
Z



Co
0




CM
N)


JJ


XXX

000
UJ LU LU

C00 0
a) a) Oa

CO O O
JO O

0--

000





S0o -
000










CON
- N

aooo


20--


0
UJ
0


LU
CC
0
0




CO
0
o
0


N

<--


O
I-
a.



NM


LUJ
-J


z
o
0

U.
IL


CO


O-
-j
X



0
LU


p-

0
0




0)


U.
co
tl





CO

0




a-


I-



(D
a-


UJ
1-J






co
CO





CO
N


Z
WU
0
Co
v
z -

=>
W 0
CD

z LLU


XI-

x LU
X I-
XO
X U
X CC
X
CM
X 0





I-
0- I-
C.-
-






Co CM


a
I-
I--


CO










'-


a-
0-







Co
0


.-
t-
CL



CO
N-


N N N
U) CN CM



-00





Coo
Z 0
LU
1 0 x9




000
C(O






0) 0) 0)


cc C.-
0 -

CO Co Co

co cc cc





I- I. I-
I-LL
CL > IL


Nto
(*- r-





S0 C



4CL IL I
000
coo

000
aL IL 9.
LLu. LL. U.

0C 0-
0) 0
CC c)


co (D
COC


_J





I-

-J


0
0
0


0
0

0

LJ





I--
0


0


a:

I)

LU


I.--I--
-
LU
Co
LUJ
0


0

I-










z
0












0


c)
Co.
C)

0

C)

0
e-


C)
CO

CO


I-





















I-
a.

o
















CO
a-
0


0


I-


co


0
U


14

4.)















V 0 NV tt (N w- N- N0- ( N
LI) 04 LO)N tN) oU


CO M0 N
T- (D -
CMJ


co o Co 9 N-- 0 0 Co Co N- 0
M) MO aO 0 0) -t .r 0) (N 0 qr
Co N- CO CD 0 C) C( C( 0 CO Co
amm nst omo


z
cc


I-
LU







I--
O




0 <
J E-j







LO (N
00





-0
m< O


0 0













- (n




TCM


0 0
LU LU
Q. Co l4 Co
a- I- a. I-
UJ LUJ LU LU
a: a: a: ca
0 0 0 0
0000




0000
f- C(N Co C(

cro co) V< LO
0) 0) 01 0)
CO Co Co Co
LUJ LU <'



0L 0 1
(Nj (l (N






NOMO
U. U) U) LL
r-. P, P, r-~


U- U- U- UL
N- N- N- X-

N-- I-N- =
IL IL 0- >
oa 0) v ao
N- Co Co 0
0) m Co N
0C t- 0)


03
LU
0



,-t-

Oi


O
LU

0





.-I
N-
0
Oi
-J







N-
Co
-j
0)

U)




Co
0)
0)


N-
a-
CL
W
U)
CC
N-

0


co

U)
c)
C.



C-

I-
IL
Co
CO


-J

N-
I.-
C-

LU


Co
P





a.
0
O










0







a-
..4















Cl
a-
Co







(N


I-

_.4

LU




_--
0
cc





U-
-j








CD



N-



0
LU



















0
0
o


SO


X4
2


: 0
UJ -J


O O<
P- X
a. I-




Co iC






I- --
OO to
a0) 0)

*-- I- -



o 0
0 -0
I-
CO) W



,< u <
-I cc -J

NO o




-










0) ao a

002



I- Co Co




0 0)C







000
CD W0 Co


a: Co
It m
Co N N- N- N o
0o Co Co Co 0)o<4.


I-
I-



0
CV


- (N 0) 0 U) '- (M 0O 0 .4 U) Co 0 0 -m (


<




0
aL
U-
I-
0
0)


>c



0
<
CI
I-
(N
0)


>-

-



o
P-
(N


>-





<
0


co
F-

CM
0)
(N


I-
I-
IL

0

X
O-
a.
LU

0
CM
0
0
o





r-.
N
(N


X
1-
ao
LUJ

0
0
N-

0
0
LU
0)
0

C(
CMi


0

0







0
-
ccl
ir
0
h-
-J
Cm


0C
z
I- -


U J U-






O
OC -O
03
-- X





0-
a 0
Co



























mo
0 0U
a.-



1 .. -




oC1



-




< L-



>-
LL



- 0 U)


Co U) (N






- MN (N
- 0- 0)
< U
U-
LO 0 C
CMCMC


-I

tl-
I--

<


.-
a



1-



o
-


-o
0)


-J
cc
IL


0
0
(0





0
>-



(N


-J



I-
0




m
LU
U-
0



-J
I--
U.
-j



m
0)


0m


U-


X
I-
CL


U)
o
o
0


-


0

0)


-J
B:


O-
0
LU

CM
0)
Co

0
C


0
-j
I-
Co
-j



CO
0




(n
a-

0


>- >N >-
w w w x w w w w
cermr e ^ mem .-^


F- U- F- lF t l U- LL F- F- l
=- I- I-I- -=-N- I-- -f-N-
a. > a. aL aL aL > a. > CL a. a. a.
Co o cM Co o 0 Co Co c( Co N-
0'- V 0 Co Co Co V. Co V' N4
0 N (o N C Co 0) C Co 0
C CM o o C
(N (N(N t -OCMOC t MC


ui uj

=ZW
N-I-

z z





(N 0) 0)
000
<<'4
co UCC

CM* U) Co
(N (N co
0) ) 0


-J
LU
I-



O


0

m


co
N-


-I
N-
z





CO





CD


a:
D
0
w




<

0
a:



w
0
OC
UJ




01


z
0
0



p2
0
co
0)


z
z
0



x

-
co
co


z z z
WWW

z z z
000
000

000
XXX


(N 0) '4'
0) 0) 0)
0) 0) 0)


LU
0



a.
N-

X


U
0

N-1




.


CD


0
Z






.
t-
CL


X

N-
N-
r0
co


-o


C0
g





C.)
.=
*S
J


um


m_
*t


- CO f- o CO
c0 ) N Io C N-
0) N^ f- C)


z
cc


I-
-J


--





F-

UJ
Ir

CO

0.
0)





a:

0c)
CM


-J

I- 0
L--
a. I-
0 uJ
UJ


u- m
>-Co



C ..-*


I- I-
000






F-a
LU < <

CO CO
c-0










-^ 0 T


-J
B4

I-

0
LU






0


-i
N-





0
_J


I




0
0


-j
0
Oi
-I
o


-J

>4
X-

0
LU
Co
I-Co
U- 0)

<
LO
:N

0
I-
.-0
X -I
N-
CL W
LU <
CO -I
U),-

o o
o o


Vq- v<
01 0)
LO O
>Co

(n -


I-
N-
L-
0

I-

CL

LU
O



**




0)
Co
UC
C0
0


D
LU
0

N-
m


Co




>n
I-

0
w



m


LO




I:
0

(N








LU
5 O

D C








w
0
.a

0
a 8
0 0







M
Co





m

m
U-




LU

CO -

o
0
0
Co
Co


0

0)


CoJ
0)

0)

0


CM
LU
0
0)


I-

z

O




0)
Co





(D

CO


x
N-
a-
OL
Er
UV




0)
(1

CM





-J
(N

co
(N





N-


0)
0)
0)
01


I-
CL
0
N-


I-
.



0




CO
0

N-




U-
IV*
CM


0
LU


0







<
cc
Co



CO


0
CD
LU
U)
0
CV
0
EC




LL



U-
0





>
0
U-
N-


0
z
0



I-t

0

LU
I-

0
z
0
UJ

1


M
0

LUJU
CD>4
o on
<



CM C

(N 0)


m Ca
co co







C) CM
U- -
= -



0) le
(N C(
CB 0
Coo
.4.








00


cc
LUJ

LUJ


0 0

F-N


0
o o



0) 00

0 0


Z Q
0-
0 U)
SN
CMo





0 0

0 CO

N-


LU
Co
LJ
CoO.


0
z

N-
I-


<


Q
1-


<

IL
0)
CL




o
m.
a-
I-r
0
C)


<
'-4
a-



o
CL
U.

o0
I-
U')
0
C)


z
-1

0

CL
a.
U>-
F-
N
0
0)


0
-*

in
o
L



0)
co
o


I-I -I
z



0 0



oo
CL CL
w- U-
F- F-
01 0
0 -
0)0)


'-4
z
<
-J
U)
0
a.
OL

-
0-
5


<
z



0
m
CL
U-

-
0o
0)


z




L
F-
-J




-


C


<
z


Co
0
C.
L
I-

.-
CO


<
z



0
CL
U-

-

0)


< z z z
zzZ



000
aL aL CL
LL U- 1L

I---
CO N- Co
0) 0) 0)


<
z
z



0

IL

t-
01














* CO ": C9 7 C'J CM 1: CM 7- CO 7- C
-q- v N m CMi oC) c O CD m f
COj vM a9)* C\j N1 LO to 03 CMi
**-0NNN 0 C0


i'- Ci
4: U


0n V o o c 0 V N o ) to )o O
NQ -- --


z
LU
0
0
cc
0
1-
-J
0


I-
I-
0-

I-

I-

iL
LU
oc
0
CMl
0


0




0
0
0
CM


s
-4
0
I-
X I-
F- 9.

LU LU
0 0


1- 0 01
I->-
<<>
S0- N
R: CM -
0 0
-J z- F-

3 -4 < -4
I- I-
o- I-


0~00.0
CL CO CL M
UJ < UJI <

UO 0 0 -
- CM
m) 0 o 0
00-0
CM D O 0


0) a)0
cc >- 2 a:
0- < < <
< 2 --. 2
co co) Nr r-
C CM C 0


-4 -4
uj x x LLU

CL IL OL OL
< UJ UJ <
0 a:CIE 0
0 IO CI 0




0) 01 m 0)
I- 0r OC)
0 0 < UJ
U) U) 4o 0
000Q

1- 1- C)J
COVO
0 0 a 0
T- M NM *-
49: 4: 4 0
QMO

u- ou



00 >-
v: v:4 r-.
U-)

x I- 1- =
> CL 0L >
CO co T (o
N v 0 -V
CO c) i- to
N CM


X
1-
CL
w
Uc






0
<

CM
0



>-

1-


O
a-
4:
0


F-
I-

0
I-


I-

LU
m
0
C
0






LU
U.
CM


0
0
I--



0c
0






U-
0
-


0



U-
0
0
*-




0
N
N


F-

M<


IL




4U
m0


0



01
01





m
0







M
U)
0.
U)

0
2
O2
0


x
1-
OL
LUJ
cc
0
In


m0
U)
0


w
0)
LU
N
-
CD
LO


UL
0


1-
n-
-a.

N
0
v
o
O


I-







X
I-

OL
LU
0



0
-4






0)


0



CM


-4 -4

0
I. F--
C)
C )
-iLU
LU x

F- 01 I-
0 > I I--
Z= C O m<
LLO 0
9- 00
10


X- I-I-X
-

0. CO 0. 0.


- U) U)

0000
U) 0 U) -0

UO ) co c0


0 0 0 0
w >- C-
2220
so-8
N 0 -
C ) -r M


0
XXX<

CL L IL F--
U LU UJ LU

0 0 0 CO
U) U) U) U)






SO 0
-O N 0
01 01 0 0





0 0 0 N
4- 4- 4- -:


1-1-

CN 0)
o0 v
'- U)
CM r-


1-
OL
1r-

r-
N


0-
U)

r-


X
1--

a-
LU

O










0
U)
co

0
CL




I0
CM



N
P--


_1



C-
0
UJ


a,
LU





OC
0



-J
1-

-



0
0
0




0
0
0
4:
0


x
a-
LU


U)
cc


w


0
CM

0
0
-
0
0



0
00
Cr

NC
cc
1-
I-
0-
CL
CO
0
4:
03
0


0 M o U N O


-4 -4
00

m u
Lo o
00

O0


-4
0
LJ


0



0


u-
x:
U)


w
0


0
M

2
1-

4:
co


m
0

M


0
2

1-
No


w
0
CO

5
0
M

2
1-

U)
4:
U)


mam
00 0
000
000
000

222
W M M
I-I- h-
1- v v
4: U) 0
4: 4: oo


m
0
M




0
2
1-


m
0




2
1-

0
4:
U)


LU
0
0
UJ
u.



0

2
1-
I--
01
4:
co


LU
o
M

U-






to
0




Uo


m
0


C
0

2
1-


oo


0
0
I,-


I--


I--
..J


Oc



O
0


CC
0




-1

UJ
LO
<
-4
0

0
0

0


0
0
4:
N


I-
U.






0
0





m
m
0

11-


F-
1-
0
1-

X
I--

L


0
0

0

0



>
0
2
0


I-


1-
0









c)
cc


I-
R

B0
0
-4

.-
x




CL
cc
10
LO





to
w




-1

M
0
N


I-
2


0
4

0*-
4 i=
t- 0


0 0


I- I C
0=1-




0.-
I- S -
I- <

0 M

M -
0 0

o-





w m


- 0 0


XX03
LU
-I- 0
OL 0-IL
UJ LUJ U
0100
U) i- CM


0COO
0 0 0
co o c


>-I0)
w w
0 m> U
2-0
to O 0

U)10 0
N CM C
Q N N


03.L0. CL
L- 4: 4 4-

1- u21- 1-
I- = F- I
CL > CL m.
U) 03 ca tr
0 U) N N
0> (O 0 0
0 N0

OCO
0 00

-0- N







o5 5
0222

o<<<




10 10 0 10
04- : 0-
0000
9- C C C
2 2 2
1- 1 I

N U 4 U


C.

--
1-
o-
oL
0

01


2
0
cc
2

I-
-4
U
m

I-.
-i
0
X
I-
CL

0
CM4
-)








0


-4


I-
I-

LJ
C
0 -
I-
C- IL
Er o
C-
C- CC


o m
55
00



CM 0
-4 -4

F S
< <
-4 -4



v: 0
RO
4C0
@0
V. .D



zo
D =
n4<
N M)
- N


I-


X
-4






1-
c0





LO




0
0








CL

0
01


<


X

*-j
I-

0)
0
-J







0
CJ




OD
>-


0
-J


01 <
-


0 0


S-00
I-
F- F-
0 ") CO
UJ <
Cc i-q -I
N **- O
C)
000
000



C0 0) C


< o-


z



>
I-



C)

z

0
0
'-



CD

F-







0
I-
-4









0
1-



0

0o
>2

5)


000000
000 0 0
1- 1- 1- 1- 1- 1-


F t- I.-1 -
LUJ LU LU U


- v 0 v
Irn

N-mem
N w 0 0


w wm
co 0 01 c

0000
JOQo CCO
LUJ 2u. 2 u.
*- U) NM U)
i- M i- CM
U) U) 0 U')
- CN 0-
N N Nl


IL
U-
>- n.03


0.0)
cD r- co
) N -N



v< 0 0

0
CM


>-
co
0
u.


0
LO
O0


w
1-
z
m



0"







o4
C)



-




0
0)
r-


V 0 0 0 0 0 0 0- 0 0


LJ
2
<
0




tF-
0


1-
0
U)


z

0




i
0



Uo
U)


U,
2
<
0
o




I-
0
U)


UJ
2
<
0




L-
0
U)


U Wu >-
<02
HU<
0 F-

000




0 0 0
U.) U) U


2
2
<



0

I-
Uo
O
oo


F1



0


I-
O



U)


-4



1-- 0
a.
CLUX cc
n
0 0


0 UJ









F- F- I- F--
.- .J = )1
04 < <




*i _j _j -j







00 0





z 0 >- IL
-3 =l 4
0000
-4-4-4-4*
1-I-I -


-J


Xz



0
-4
O
0








0
CO






IE
0

0
0
N






0


z
u


0
C
&






<.


I-
I-
CL
0
I-

X
1-
OL
UJ
cc
0
CMJ
LO

0

0
m



>-
o
0

0



4:
2
U)


a
LU





o

0
0


-



U-

M

u-



r-.
=

0
r-


2
0
0
0

LU

2

;E
I-
2

0
LU


z
I-


4:
0
0c
0






F-
LU








=
-j
LU






m
LUI
0
LU

00

< >-



-
0
2





m
4:





1
0

0

LU


00
X<<
1-1- 1-
OL 1- 1-
LU LU LUJ
0zz2
U) 0 0

mom
-0
000
CCZ2




-00
-) -
N 0N U)
0 O N
> coc


z
1-





m
U9

2
00
10
M



0-



0
-
4:j


C)
8-
i
u

ee



Q.


cc





0)
0)
0)
0
0


CM 00

00



2o
v o


I--
0
4





0



01



o


0

0




N
t-


co





0
-


CM
0






ui
F-



0
0




r-
N0


UJ
2z
2 L
< >>
O Z Z Z
00000





X Z 22 2
I-I- 1-I- I-

0 0 0 0 0
co o o c c


0
0
0
o



0


h-
0














N- N- 0 (0 C9 ..

o N- T N- -


- 0- c N O Ln 0 M 0. 0 0 c o

CM 0 NM CM0 TNI 'cr N)0 t--- U IcDT Un i- C0 i M N, cD


.J

U.

I- F-
0- u
n--




0
0 0)
) >-


O CD
0

.- 0) 0
*-0 0

I- I-
0- O ()
,U < <

000
0. .I ..


0 0. 0



0 0) 0
< Oi <




T- CMJ C\J


0
LU
0


LU
BC


CD
0



w
CM



;r
01
UJ
0
0t

U)






v
CM
0)
u.
X

(0
U)


I-

0
LJ
cc
CD
0
0
0



2
CD


0




0
-J

0N
-J



10
0



0)


co


0
LU
X X N
0- i- N-



00
o
CL CL I-


uJ U) UJ


S0 0
000


N 1 01

<- 0




NN-N


0- 0- 0-




Tr V, 0)
m 0 0


LU

I-1
cc
I-
m4


0






0
-J
N-

N-

0
-J



-J


0)
0 0
0 u



(0 0
0 0


0
LU
(0


(0

0
CO
CD





5O
10



CMj
10






CO
U)
U.



0
CD
4.-


-J


--
c-
L-


0
cc
CL

0



-J
0-
0




Co
-j
0






CO


0o


0


cc.


cD


co



0
i--


_-





LU



< z






I-


L :
00
.- mr)
UJ<










N


-CO


N-
3:
0






LU

O0
CJ









0
CD


-J



0L




0
U,
0

I-
I-
0-
I-
0 (-
L



...I x X

t0
N-


0 C 0





0) 0 0)
N -





I- I-
UJ 0 CD

-n m
00
eN




N l-"




0 0 0
0- CM CM

- 000-
z0 0 1



CM M CM


X
N-
0-
LU
CD
cc




CD
en
0
0








LO
10





I--
0:
N-




CO
Uo


N-
(0
01


CD CO CO
LO U) U)


0- U -
> > 0
0 NM1
CMJ 'cr
Ns-(0CO
01


U-


F-
I-


O
I-




X
--





41
I-


N-
0-


N-


01





0

Cr


F,


0- Lj
LU C
CD U.
0 0
U) C.




0) CD





. N






N-U
I ->



0


*-
X-

c-
0
-j

0














co
0
O-

-J


-

I)
0
LU




LU
m
N- U.

CL (0
0

0

>-
0- CD





CO
0) 0
- M
0 0
0* 01
< UJ


U-
0



Lo
LU










01
CD
0
U)
In








0
0
0



CMl


CL
CMl
0
U-


0
U.


0I
UO

U.



U0
cO
vo


z
cc

0
0
0L

I-
0
z
U.
-J


X
I-




0-


Uo
N-~

0
CO
0




U-


LU


0- -
UJ LU

0 )



0-
7O




00
zz





N N


0 0O
SN
00
N-U-
0- >
0 (0
4. 0)
0)


F-
f-
N-


0




0


LJ
1C













0
4W
N-




0
-I








.-


0
0 -













,CB
N-


oC
05

LU ]








UN
10


C.
0
I-
N-
0-
N-.
0
0
0


0 0 NCD400 0 0 0U)0 0 0 N- C 0 o0 0 0n 0 0 0- 0-0 0N


NL
0
Cl.

0
LU


N-
--


CO
0


0




CD
Nr-
CD


N.-
0
0-




LU
0-

0
N.
co


000
0- Q- 0-




0 C O


0-N
Cm C. C0


N-
0-








co
0
0




CO
CD


CO
0


0

C-
N9-
co
co


N-

CL
0

LU
0-
N-

CO
Co


0



L
0u
N-


cm


N-
0
I-





N-
N>-
O
CD
oo


N-
0


0


LU


0-


a- C-




LU L
0- 0-
N-N-
0) 0
Co 0
C CD


N-

CO



L
0

N-

0


0
0
CD



LU
Q.
0-
N-

0
CO


o
I-
0


0
CO
.
LU
0-
N-
0
cD


0
0I



LJ
C.


co
CD


N-



CL

L0
N-

0)


CL
- CJ
m




00
LLU U
0- 0-

CN-
0 0
co cD


-J
UJ
0

uJ



CL
0

I.-
CD
0
co


-J
.J

UJ



CM
0

0.
NL

0
0
co


-J

U-
L

0
W
1-1

0

N-
u.



l-


.J
<

c-
0
L
I.

0

0.
N-


4.


LI LU LU
.J -, -J
000



000

0 CO 0


000
4I v


-J

0


uD
0-

LO
0
N-
0

4.
t-
10

o<


CO
CM


(0
F-

0
0


I-
I-
D-
0
F-

X
N-
0-
LJ

0
CM
0
CO
(0



z
4:


x
0-

IU


CDj
0




0)
ED

01
LU
U.
COD

CN
U)
CM
CM


0
0
U.
U-
0


2
cc
LUJ
Nr-
LU

N-(
L1.
LL
0
LU
N-
4:
0
u.
U-
0


0
0



z
0
LU

N-t


z
I.-
0

LU


0


cc
LU

LU


a (0
X <<


LUI LU LUI
cc Z






0-CD
I- '- CD













(00-
0 O D

N- CO CD-









LO o O
0m 01 0)

CO Cc












00@
CD CO N


0 (0 0




0. 0-a0.
1O -^ r-
0 NM CM
CMl CMl 0
NM CMJ -
CMJ C0 -






x m



00(0




000
0- 0- 0-
I- F- 1-
0 N-~ NO
CO CD CO


0
LU
CQ

a 5
N- N-
0





z
0
4:
N-
N-
1
0



LU





0


0s

.0
Ni-














o o 0 CO 0 0) CO
CN -- LN CM 1 0 -


Ca
UJ
=
F-
LU





a.
X


LU
CO
0)


CO
c<


I--




I--
OL
0

X






N
U)
CD




N
't


0)

01
~3
*V


X
4-
a-
CL
O

0





0
CO



CO


I,-
4:






-i X
I-
I-N









oo
o0o


0z






z

--
O

-4
t-







I-



a-
cO


co
0


N 0 CO
0)0) 0)
< U CL

NC 0


X
.-




CL
OL


-.
CD


0)


-

0


0
LU
NO X



-O O
0 0
oo
- CO


N N CO
0) 0) 0)
uN Nj CL
0 Q<

--0
- o


oo U) N N- 0 0 0- M 0)o 0 i 0 ~ o 0)N0 ro
0 N Nm O)N M 0 CON No 0 m CO Co U) r 0) CD 0
NNc OCONiCO 0)03 cocojco0 coN


LU
0
CO
LU
CO
cc
0









I--
0

0


0)
0)
-.1

CO










UJ
LU
cc
U-
0
0
-


-J


x
I-

0
LU
CO
0
0)

CO

CO











0
-J
9-

-\J


0
I-
_i

x

C-

Q-




0




-J

r-


X
I-

O
oc
0


-




Co

0


0
0)
CO
<
CD

0
N
N~


-J




I--
O

5)
0






I-
I-

.-I C

- (D
Co
CO 0
CO CN
-1 .-


_i
I--



X


UJ




co
r)
mO


F-




I--




X.





0
1U







CO

0
a-







CO


0
I-

C/)

<
LJ
DC

<-


N CN
0- 0) 0)

0<<
inCM co


x
I--

LU
CO
0
o)




-J

-

0
CM


0
LU
0

0
U.J <


U OU

SCO
oo U)


>-
u.1
LU



I-
0-
LU
X




o
0

-.


N N
0) 0i 0)
a-O 0
CO UI.4
CO CO CN
C *- *-
* CN CN
CM CM


cc: cc
>>>> >- L
cL a. cL cL a. aL a- N CO CO CO CO CO ;r r-
NN N N N ----- COCO
0000000--------- -- ---------


U-

>
0
-a-
N^


u.



CO


u.
=
>
0
-a.
N


0)
CO
0)


. I--
> L
00
m O


F-
I-.


COl
r-


U.


O


P-
CL
CO
0
0)
0)


F-
a-

I-.
0)
0)


I-
a-


0)
0)
a)


I-
o-

0)
0)
0)


I- h U.


N- T 0
N N.
Lo L>


I--


CM
N
oJ


N
0M
0
0
-a

0


cc
U.
Ia.
cc)
CO



1-
'a-

CO
LU-


I-
a-

0
I--

X
h-
LU

O
0

0

N


0)
m

CV



I-
LU


a-
I-
a-


0
o
CO


0)
z
-3
N
0


i>


0
cO


O
0-




0<0
a-.
NI--h -








-- -- --

LU U.J UJ


0



*-a
z


0)
<
a


a.
U.
Cc
0
CO


CO


z


N
0


CM
U-

t-.


CL

0
vn


0



-
0)
z
<


0



0)
-


o
CL

M


I-


CO




0
N-

I-

0)
0

CM


a. a. CL


0oo.

- CO -


-r W-
0)m0
z z
<<

0 -
00


V
CL
a-
4


0


L U.- Ur






0)0
0M V 0
-- I-v N-


F-





X
I-








0
-i






0-
>-







U)


CO
o:
in


CL

U)


O


-
V
0
CC







0


0
0


F-
I--
CL
0
0c
0)


I-






"-




0
-J




a-

0O


0)
F--

0
CM)


0
D



0)

CL



CD


cc
-.
CO



F-
0

r,


0 0o U o M- N CO U) o M, 0 N 0 V C N CO Mo


LU
-1
0


CO


-.
N


-I)


O
CO
-.1
IL
I-


LU
-1
0

CO
CO
0
LU
a-
N


-I -LU -I




CO CO CO
000

a. CL CL
NI-N


1-1
CL
-J


O

CO
0

a.
N-


I-



0
-J




I-




CO
In




CD
0)
0
<
0


in
Ia-

CO
LU







0
0)
I--








CD
0L
0

0
CO
0-.


ZT V V V V Z - - -
<<<<<


000000000000000000
LU LU LU LU LU LU CO CO CO CO CO CO CO CO CO CO CO CO
. a. a. a. C O C O C O CO CO CO CO CO CO CO
N- -- N- N N- N N N N NN


SN 0 0) 0 CO V U ) r- CO M N C ) C CO a) 0
0000----------- ---- --- --NNNN NNNNNMCO
-a. '- -a '- -- -a ^- a- a- a- a- a- a- a- a- a- at a- ,a- a- -a a-


o
a> 4-1
0 a)






-4 a
a) -








D O 0
m) 0 )


II
"U




000


-4
e0
a)
-4
a)
1-4


F-
ca
>


.0
v
,,r

x
*-


<
Q)



42)
Qa
,.-


cc
40














C.)
a)
a)
ca
V




"D
0)






4-=
.)I




4. >, 4
a = a)

a ) -0

U.) .
a~a|









0 -
C)0
Soc



1 a) 4-I
-4 C c




" II I


5 0
LL. 1C U.
U. LU U.
01N0


cc
0




0)
a)




O









0
4-'





I
ca



a)
0
=
0




tm
tm
4-'










01
a)


4-'
0







Z
.0

0)
l-






0


0)
0)
CO
4-'

U-
> S
L.
0
4-
tL O A
.0

E II-
C 3
a)



0 -
*S N
C J 0
o 0
*4 0
4-' L- L.
cc ) 0
C.) J V -4 'o
-4 E 0 ()
4- ZU, U J= >-
1- CX- 4- 0
4- V 4m _q
a)~ LL 0 0 0
ocoao- a
-4 r- J0 0) > a) 41'
a N V) CD
a)a< )C E 0)-1 C
(D a) cm 39 -4 c d
4-1 4. c a)
CIS cc 0
a) a a 0 a4 a) a)
C: N C 1 3 L NOl*' -
cc cc cgco 16 IL a)c
a E t--to 0-
II II II II 11 11 11
LU
0L LU
0 >* U.1 l
UJ M m 1- x
1-i z in iNi-m 0


0








C..
<.


0
0
z
0

UJ

I-
O




0

ILU
N
4:




0
0
1'--


O
LU
I-


-J








LU
m
O
LU
oa-
N N-


0

I-







LU



4:
1-









Appendix 4b. Explanation of TERM Codes and Off Comments in the tagging history lookup
table (Appendix 4a) for the causes for tagging bout termination.


Term
Code
0


1
2
3
4

5
6

7


10


Fate and Cause of End to Tagging Bout
Manatee Recovered Dead Wearing Transmitter
Recovered Transmitter
Weak-link broken
Weak-link broken by human
Weak-link broken by alligator
Other at transmitter or tether;
eyebolt broken, tether broken
Other at belt; buckle swivel failed
Other human caused; small connector
unscrewed, safety failed
Boat strike
Recovered Tag (belt and transmitter)
Belt broken


Replaced Transmitter
15 Low battery life remaining
16 Malfunction with housing; antenna
missing, leaked, barnacles
17 Malfunction with electronics; batteries
died prematurely, poor range, failed
18 Boat strike
19 To better unit
20 To PTT
Replaced Tag (belt and transmitter)
25 Belt worn out or swivel failed
26 Removed Tag (belt and transmitter)
Transmitter Lost
30 Missing at weak-link
31 Missing at weak-link; assume boat strike
35 Transmitter Missing; Manatee Not Resighted
40 Manatee Rescued
77 PTT Accuracy Test Unit
99 PTT Unit still on Manatee


Off Comment
DEAD


WL
WL (HUMAN)
WL (ALLIGATOR)

(various)
(various)

(various)
BOAT
REC TX BLT
BELT BROKEN
REP TX
LOW.BATT

MALFUNCT (various)

MALFUNCT (various)
BOAT
(various)
TO PTT
REP TX BLT
(various)
REM TX BLT
TX LOST
WL
WL; BOAT
TX MIA
RESCUE
PTT TEST


ON


83









Appendix 5. Definitions of manatee activity categories and database codes for VHF radio-
telemetry data collection and data entry.


ACTIVITY


DESCRIPTION


Rest


Surface Rest




Bottom Rest


Resting at or below the surface; stationary. Two
subcategories of resting (A, B) are given below. Includes
comfort movements (e.g., pectoral scratch, back rub).

Resting (perhaps sleeping) at the surface with back
exposed; tail & flippers hang down; generally breathing at
4+ minute intervals.
[Termed "suspended resting" by Hartman (1979).]

Resting (perhaps sleeping) under the surface; includes
resting on the bottom or in the water column (subsurface);
generally breathing at 4+ minute intervals.


M Mill Slow non-directed travelling in same general area.
[Formerly referred to in database as "Idling" (code "I") by
DEE, RKB & ST, as "Loitering" (code "L") by JPR, or as
Milling by BZ.]


Travel


Slow Travel




Fast Travel


Steady movement in one general direction at slow,
moderate or fast speed.

Steady movement in one general direction at a slow speed;
tag stays at surface.
[Classified as "Idling" by J. Reid; slow swimming (2-3
km/hr) called "Idling" by Hartman (1979).]

Steady movement in one general direction at a moderate to
fast speed; tag is pulled below surface frequently or for
considerable periods of time.


84


R


* B


T


* U


,A, rT-7t TV











Drink


Feed


Human Feed


Drinking fresh water; note source of water (e.g., hose, A/C
pipe, sewage effluent, stream) on tracking map.

Feeding on the bottom (e.g., muddy water displaced), on
the surface (e.g., pulling down floating vegetation,
flotsam), or on the shoreline (e.g., grass, overhanging
bushes). Head sometimes raised above surface with
vegetation visible in mouth or chewing during breath;
generally breathing at 1 3 minute intervals.
Document vegetation type and feeding mode.

Consuming food given by humans (excluding researchers);
record type & amount of food.


Socialize


Cavort


Mating Herd


Interacting with other manatees, including nuzzling,
"kissing", rubbing, and other "low-energy" behaviors;
excludes cavorting. Includes most mother-calf interactions.
[Also known as "Mildly cavorting".(Zoodsma 1991).]

Intense socializing with one or more other manatees,
including grabbing with flippers (embracing), rolling,
splashing, chasing and/or much physical contact. Includes
sexual activity (heterosexual & homosexual ).

Animal is part of a mating herd, defined as follows:
a group of two or more manatees (presumably males)
persistently following and interacting with a single "focal"
animal (presumably an estrous female) over a sustained
period of time (typically days to weeks).
General activity of the group tends to mirror that of the
focal animal (e.g., group rests/mills when female bottom-
rests; courtship activity resumes when female becomes
active). Female may "strand" in very shallow water,
apparently attempting to evade male escorts.
Actual activities may include resting, milling, travelling,
feeding, socializing, cavorting, or copulating. These
activities should also be recorded. For females, the mating
herd code is used only if she is the focal animal.
[Termed estrouss herd" by Hartman (1979)].


85


D


F


* H


S


C









Tagged cow nursing her calf or tagged calf suckling from
its mother. Calf below surface with head touching and
under mother's axilla.

Tagged female gives birth to calf. Restricted to actual
observation of birth, or if apparent that female has just
given birth (e.g., blood or placenta in water) within the
hour.


Other Activity




Unknown


Any other activity; currently used only for researcher
tagging interaction (i.e., initial capture, tag & release;
retag; unsuccessful attempt at retagging), indicating
possible disturbance.

Activity unknown or not recorded. For example, an animal
is stationary but could be feeding or resting.


* Denotes new or modified activity codes.

Observers: BZ = Barbara Zoodsma, DEE = Dean Easton, JPR = James Reid,
RKB=Robert Bonde, ST = Sharon Tyson.

Sources: Hartman (1979); Bengtson (1981); Zoodsma (1991); Puerto Rico captive manatee
activity data sheet (R. Bonde, pers. comm.).


86


N


* G


Nurse


Birth


0


X









Appendix 6. Preliminary assessment of the precision (i.e., repeatability) associated with
digitizing manatee point locations. This was based on a small number of points in the
Sharpes quadrangle, close to the FPL Canaveral power plant in Brevard County. Since
this small, non-random sample is not likely to be representative of the VHF database as a
whole, caution is necessary in interpreting these findings.

The preliminary conclusions are as follows:

(1) Locations may be digitized with approximately equal precision and accuracy on hardcopy
maps and on-screen in Arcview; the two methods differed by an average of about 20-25
m (1.0 mm on a 1:24,000 scale quad map).
(a) For a best case scenario, S. Butler and C. Deutsch each digitized five points that were
located at easily discernible landmarks (e.g., tip of FPL breakwater) and found that the
mean absolute differences ( SD, min-max) were 10 m (9, 2-22) and 14 m (7, 6-24)
for X and Y UTM coordinates, respectively. There was no directional bias, with the
mean differences (including + and signs) close to zero. The mean euclidean distance
between locations was 18.5 m.
(b) For a more realistic scenario, we each digitized five actual locations in the FPL area.
The mean absolute differences ( SD, min-max) were very similar to the above result:
18 m (16, 4-42) and 14 m (5, 6-19) for X and Y UTM coordinates, respectively.
Again, there was no apparent directional bias, with the mean differences (including + and
signs) close to zero. The mean euclidean distance between locations was 25.4 m.

(2) The repeatability (or precision) with which VHF locations are digitized may be poorer than
we anticipated. For four locations in March 1988 that S. Butler redigitized, the mean
absolute differences ( SD, min-max) were 130 m (66, 43-192) and 48 m (34, 20-91)
for X and Y UTM coordinates, respectively. Two of the X coordinate pairs differed by
+170 and -192 m, while the Y coordinate pairs differed by at most +61 and -91 m, so
there was no obvious directional bias such as might be caused by improper map
registration prior to digitizing. The original digitization was not done by S. Butler. This
comparison needs to be repeated with a much larger sample size over a number of areas
to estimate digitizing precision with a higher degree of confidence.


87









Appendix 7. Report on tagged manatee use of the Cocoa Beach / Thousand Islands area (mid-
Banana River, Brevard County) from 1992 to 1995 presented to the Bureau of Protected
Species Management, Florida Department of Environmental Protection.


Tagged Manatee Use of the Cocoa Beach / Thousand Islands Area

Data Summary by Dr. Charles J. Deutsch
Sirenia Project, National Biological Service
11 April 1996

Study Period included in Analysis: 1 Jan 1992 to 30 Sep 1995
Data Type: Platform Transmitter Terminal (PTT) locations determined via satellite
(Service Argos). VHF data were not included.
Location Class: 2 & 3 only.
Note about accuracy: Approximately 2/3 of given locations are within 350 m (LC2) and
150 m (LC3) of the actual locations (i.e., one error SD), according to Service Argos.

Procedures and Qualifications
I restricted the analysis to all locations of the 25 manatees that were recorded at least
once in the Banana River during the study period. The resulting database contained 19,277
locations (LC2 & LC3); this was condensed to only one location per day, for a total of 7922
"manatee-days." The attached tables give the number of manatee-days in the areas delineated by
you: MinuteMan (MM) (the proposed watersports area); Cocoa Beach / Thousand Islands (CB);
Banana River (BR); and total days tracked throughout the entire range. Note that the values
presented for these areas are inclusive; for example, the number of manatee-days in the Cocoa
Beach area includes the number of manatee-days in the MinuteMan area. I included data for the
entire BR in order to put the CB data into perspective. The data are pooled across sexes
(10M:15F), age classes (19AD:6SU) and rehabilitation status (15 Wild : 5 Short-term Rehab :
5 Long-term Rehab), but wild-caught adult females make up the single largest group in the
sample (n=l 1). The mean ( SD) duration of satellite-tracking (i.e., no. of days with LC2 or
LC3 locations) was 317 ( 325) days, but this varied greatly among individuals (range, 12 -
1091 days). Some manatees (e.g., TBC-09, TBC-24, TRB-01) were tracked for over 1000 days,
or nearly continuously during the 3 /2-year time period, while others were tracked for short
periods. The latter included three captive-born manatees (TBC-45, -46, -47) released from the
acclimation pen in the upper Banana River in late August 1995, one month prior to the end of
the current database; other tagged animals made only brief visits to BR (e.g., TBC-37, TFP-06,
TMI-03) or lost their PTTs soon after tagging (e.g., TBC-43). Six manatees included in the
database were located on less than 60 days over the study period; excluding data from these
animals, the mean tracking duration was 407 ( 324) days. An additional four manatees were
located in BR on less than 60 days. The statistics presented below are based on the entire sample
of 25 manatees; for some analyses, these are followed by statistics in brackets [ ] for the sample
of 19 manatees with at least 60 days of locations and by statistics in brackets { } for the sample
of 15 manatees with at least 60 days of locations in BR.


88








Overall Use of Cocoa Beach / Thousand Islands Area
For all animals and months pooled, 57.4% of total manatee-days were spent in the
Banana River (Table p. 7). On an individual basis, 64% (16/25) of tagged manatees spent at
least half of their total days tracked in the BR [53% (10/19)] {67% (10/15)}.
For all animals and months pooled, 10.6% of BR manatee-days were spent in theCocoa
Beach / Thousand Islands zone (Table p. 7). For comparison, the area of CB (25.76 km2) is
7.7% of the area of the entire BR (336.12 km2). On an individual basis, 64% (16/25) of tagged
manatees were located at least once in CB [74% (14/19)] {73% (11/15)}. 40% (10/25) of the
animals spent at least 10% of their days in the BR in the CB zone [47% (9/19)] {40% (6/15)},
and 8 [7] {7} of these were located in CB on 20 or more days (max = 167 days).
For all animals and months pooled, 1.4% of BR manatee-days were spent in the
Minuteman area that has been proposed as a watersports zone (Table p. 7). The area of MM
(5.76 km2) is 1.6% of the area of the entire BR. On an individual basis, 36% (9/25) of manatees
were located at least once in MM [42% (8/19)] {47% (7/15)}. 24% (6/25) of the manatees
spent at least 2% of their days in the BR in the MM zone [26% (5/19)] {27% (4/15)}, and 5
[4] {4} of these animals were located in MM on 6 or more days (max = 17 days).

Seasonal Use of Cocoa Beach / Thousand Islands Area
Tagged manatee use of the Banana River varied seasonally, with low use in winter
months (Dec Feb) and high use during most of the rest of the year (Apr Nov) (Table, p. 6 --
data for all individuals pooled into manatee-days). Most of the animals spent the winter in South
Florida or in the Indian River near the two power plants (FPL & OUC); they typically moved
into the Banana River in March and April. Restricting the analysis to just those manatees with at
least 10 days of tracking data in a given month, the following percentages of tagged manatees
were present at least one day in the BR by month:
Jan 23% (n=13) May 79% (n=14) Sep 79% (n=19)
Feb 21% (n=14) June 89% (n=18) Oct 93% (n=14)
Mar 60% (n=15) July 79% (n=19) Nov 77% (n=13)
Apr 79% (n=14) Aug 69% (n=16) Dec 42% (n=12)

Use of the Cocoa Beach area likewise showed marked fluctuations across months, with
distinct peaks in the spring (Apr, May, June) and fall (Sep, Oct, Nov) (Table, p. 6 -- data for all
individuals pooled into manatee-days). Restricting the analysis to just those manatees which
spent at least 10 days in the BR in the given month, the following percentages of animals were
present at least one day in the CB area by month:
Jan 0% (n=l) May 67% (n=9) Sep 40% (n=15)
Feb 50% (n=2) June 55% (n=11) Oct 58% (n=12)
Mar 57% (n=7) July 46% (n=13) Nov 60% (n=10)
Apr 40% (n=10) Aug 60% (n=10) Dec 25% (n=4)

Note that the small numbers of manatees using the BR regularly in winter (Dec-Feb) mean that
the corresponding percentages of animals using CB during this season have little meaning. Of
the 193 manatee-days logged in the BR for the three winter months, only 3 (1.6%) were located


89








in the CB zone (Table, p. 6). These 3 represent only 0.6% of the 483 manatee-days recorded in
CB. It is clear that our sample of tagged manatees rarely used CB during the winter period.

Use of the Minuteman zone followed a similar pattern to that of the larger CB area, with
higher use occurring in spring (Apr, May, Jun) and fall (Oct, Nov) (Table, p. 6 -- data for all
individuals pooled into manatee-days). Restricting the analysis to just those manatees which
spent at least 10 days in the BR in the given month, as above, the following percentages of
animals were present at least one day in the MM area by month:
Jan 0% (n=l1) May 33% (n=9) Sep 0% (n=15)
Feb 0% (n=2) June 36% (n=11) Oct 25% (n=12)
Mar 29% (n=7) July 8% (n=13) Nov 20% (n=10)
Apr 30% (n= 10) Aug 0% (n=10) Dec 0% (n=4)

The tagged manatees were not located in the MM zone in winter (Dec-Feb) and only once in
July, August and September. The small overall sample size (total of 62 manatee-days),
however, makes conclusions about seasonal usage patterns more tentative.


Conclusions
Approximately 2/3 to 3/4 of the tagged manatees in our sample were located in the Cocoa
Beach / Thousand Islands area, as delineated on the attached map. The area was used to a
considerable extent during the periods of spring/early summer (April to June) and fall
(September to November). These roughly correspond to the end and the beginning of the spring
and fall migratory periods, respectively. Tagged manatees rarely used the area in the winter
months (December February) and, for unknown reasons, there was an apparent lull in use in
the mid-summer (July August). Qualitatively, we know that the residential canals and other
protected waterways in this area (especially to the north in the Ten Thousand Islands part) are
used by manatees for resting, and that they often move offshore to feed on seagrass beds; a
favorite foraging ground is in open water along the southern edge of the Cocoa Beach zone.

The small size of the proposed watersports zone south of the Minuteman causeway
makes firm conclusions more difficult. About 1/3 to V of the tagged manatees were located in
this area over the study period, with a seasonal pattern similar to that of the Cocoa Beach area as
a whole. The Minuteman area did not appear to be a hotspot of tagged manatee use, but it is
likely that manatees passed through or near this zone when travelling between important
foraging and resting areas to the north and south; further more detailed analyses would be
required to examine potential travel routes. The radio-tag is usually underwater when a manatee
is travelling, and so the tag generates few (or poorer quality) locations in this situation.
Therefore, the telemetry data cannot directly be used to assess whether manatees are travelling
through the small zone of interest, although manipulation of the data with cost-path analyses (as
is being developed by FDEP researchers at FMRI) can produce models of potential travel
corridors. Aerial survey data would also be helpful to determine frequency of use by active
manatees in this area.


90






uO > > g '
til (" l 0 wK
r) < -3 10 ID F.4 u w! ^6l


H
o( )


HI- H 00
0 0 o- H w wM 0 0 t


SW>o on Hm w U 0>









W& ummm mw Ln
H 0 M 0 0 4 m w. -j L H0 C









M -H -3 -. -H L -f -.3 0 0 MH 0









o1 1'| |' 0^ 0 0i' H 1 -3 0 0 0 w












o H0 0 0 0 O 0 0 0 H












H 10 J O. Ul O LO0 Ul 01 0l O











o) M co


0


H-

0
ri.






0
A)
0




D
























H
r)






























0





H
I-.-







*1
0







C)'


DI


Ct


01
(l
w





0













CD
0
0
O3


C4
(D

(l
01
L<
H

10





M
C'.

(I)

(l-



H
10
10
U,


or
U,


0
o
-C





Ot








Dn 0.


I












H


to
0
At


H
o
LJ


o

iF,


Z


H
0


H
H
0
LJ


H
H
I
0
)


1-1
z
H
I
0
H


o-3

0
On


H
>-)

A
-.2


a)
00
>fc
en


H

I
U>
Ln


H
03
I
i
e
U)


H
I
U)
A
w


H
t0
I
A
K)


A


0
CO
(4


H
I
-4
w
a*


03
I
U)
(O
LJ
cn


H
I
B0
to
C)


H
0
I
M
U)


H
03
I
M
U)


W
0
o
u>


0
to
C)


W i 0 0 0 0 0 00
0 OD W 0 H 0 H 0 0 0 0 0 H 0 0 H-L, H-) ) W 0 oo


A
k? 1.) H Fl'I- tO I-4 -4 tQ M
'. w r0 U0 w w H w w t( H H m w m w OHD HL AJ 0) Ul f
-32 (J 0 W) -3 OD 00 W 0 -< H UH l ( 0 0 .' O W G) 0) 0 0 4 O t) ID 0 A






H HH A
0 K)AHHAUJU H HHHKA (4ooA'.
Ui W D H H. H o A O H M NW) 0 H W A M K
OD Kj b) A Li W ) w A -2 K) 0. A K) m) H Ut H W A L ) lA Al H O )


) H H
oinoo
0000


O ) K H H-
0 U1 0 0 0 om) 0 -o wO 0 0
. . 0 O.O O O O OO


H Hoo
0 0 O 0 K 0 H 0 0 0 0 0 0 0 0 0 0 0 0 w 0 )w 0 0 O
0. 0 0 0 0 U, 0 U0 0 00 0 A 0 0 0


OD F C O o omA OOOOO ooMMo o
I4 ( 0 ) 0 0 0 0 0 0 0 0 H 0 0 0 K K H '0 0 H 0










U, 0 H K) U L) A -i 0 K 0 V O A K -J H A 0 0 A 14 ( A n


CQ
fl


(DI
rt
(D








0< 3-
Qnrt





a o
0-

0D

w0


II
K)
0.'


0
()


H-









ft)
H-


0


Di
1<
Cl)




0


Di
Cl)


M






(D
4O


0)
rC









t-4




0

or


ta


0
O 1


0


(D





CD








Ix)
H-

tC
1'-1




0
M-







H


C4

(D

-10







co
CT





CD
Fn


H
til







til





t4
0




H







0
a































(D
a.


< 0

0
















\ 0
H



















WO

































N
w+ E
S


0


1


2


3 Miles


Cocoa Beach / Thousand Islands Study Area.
Two zones mentioned in the text are delineated. The proposed watersports area lies
within the Minuteman zone.




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

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