Analysis of clutch size variation for loggerhead sea turtles (Caretta caretta) nesting on Bald Head Island, NC USA

1
Analysis of clutch size variation for loggerhead
sea turtles (Caretta caretta) nesting on Bald
Head Island, NC USA

• Melissa Hedges1,2 Jim Berkson2
• 1 Bald Head Island Conservancy, 2NOAA Fisheries
  RTR Unit at Virginia Tech

Regression Results
Introduction
ANOVA Results
Loggerhead sea turtles exhibit variation in the
number of clutches laid per season as well as
number of eggs per clutch within a single season.
The following analyses explore variables that
have been shown to have significant effects on
clutch size in the literature. A regression
analysis explores clutch size (number of eggs per
clutch) as a function of year and straight
carapace length (SCL). The change in clutch size
as the season progressed was also analyzed to
determine if clutches 1-3 were significantly
different in size. Significant results will help
to understand the fluctuation in nesting behavior
of loggerhead sea turtles. The two analyses
presented here will be included as parameters in
a loggerhead sea turtle population model. The
data are significant because Bald Head Island
lies at the northern range of the northern
nesting subpopulation of loggerhead sea turtles
in the USA.
The small number of observations of turtles
laying more than 3 nests in a season is not
sufficient to a statically robust ANOVA test.
For this reason, the analysis was limited to
observations of females laying three or fewer
nests in a season.
Figure 1 displays the distribution of clutch size
for the 342 observations used in the regression.
Each observation represents one randomly selected
clutch for each tagged female nesting on Bald
Head Island from 1991-2005. The normal
distribution of eggs indicates a a good fit for
use of clutch size as a response variable in the
regression.
Figure 4.
Figure 1.
The ANOVA was limited to 743 observations. All
observations for all turtles were included.
Outliers result from nests that suffered
degradation during incubation, resulting in
incomplete egg counts during excavations. There
is no significant difference between clutch size
for the first three observed clutches in a season
(F1 0.12, p .7333).
The regression of SCL and year to clutch size is
significant (F29,312 6.53, plt.0001,
r2 0.378). While SCL and year explain 37 of
the variation in clutch size, there is still a
great deal of unexplained noise found in the
data, as can be observed in Figure 2. The
positive trend of increasing clutch size with
increasing SCL is clearly evident. Straight
carapace length has a significant effect on
clutch size (F1,312 81.87, plt.0001). The
significant correlation of carapace length with
clutch size has also been identified for females
nesting on Cyprus (Broderick et al. 2003),
Cephalonia, Greece (Hays Speakman 1991), and
Little Cumberland Island, Georgia USA (Frazer
Richardson 1986).
Map of study site, Bald Head Island, North
Carolina, USA. The island has 9.5 miles of beach
on which sea turtles nest.
mapquest.com
Methods
Data Collection Data were collected on Bald Head
Island, NC USA from 1991 to 2005 during nightly
beach patrols, daily nest monitoring and nest
excavations. Individual females are identified
by Inconel flipper tags (National Band and Tag
Co.). Yearly averages of SCL were taken for
females that were measured multiple times a year.
The dataset is not a complete census of nesting
females, their nests, or egg counts. The average
sighting rate of nesting females from 1991-2005
was 78 and lower for nesting data due to nest
destruction during incubation, prohibiting an
excavation. Regression of the effect of year and
SCL on clutch size Observations included in the
regression analysis are those for which both the
straight carapace length (cm) and total number of
observed eggs are available. To eliminate female
effects, one random observation was chosen for
each individual female if she laid more than one
clutch within the 15 year period. Clutches of
eggs with less then 40 observed eggs were removed
from the analysis. All observations with less
than 40 observed eggs per clutch suffered a
disturbance during incubation that resulted in a
loss of eggs these observations are not
representative of the total number of eggs
laid. ANOVA model of observed clutch within year
and clutch size Data used to perform the ANOVA
were limited to turtles laying three or fewer
nests. There are not enough females laying four
or more nests to perform a balanced ANOVA. All
observations were used in the ANOVA because it is
reasonable to assume observations for individual
females nesting in different years are
independent. A one-way ANOVA was performed using
observed clutch number as the variable.
Figure 5.
Figure 2.
Survival rates The preliminary survival rates
presented in the abstract were obtained using
program MARK. The authors believe that
multistate models represent a more biologically
sensible model to obtain survival and recapture
rates for loggerhead sea turtles. Rivalan et al.
(2005) recently applied multistate modeling to
leatherback sea turtle capture data and
preliminary runs show that the multistate model
also fits the capture data from the Bald Head
Island study. Survival and recapture rates will
be calculated in the near future.
Year has a significant effect on clutch size
(F14,312 5.96, plt.0001). The abnormal data
from 1998 is a result of a high destruction rate
of nests and limited data. The interaction
effect of yearSCL was not significant (F14,312
1.71, p .053). Clutch size variation is a
function of numerous factors as can be seen from
only a moderate amount of variation explained by
the two variables analyzed here. Long time
series of data are not particularly common for
loggerhead turtles and only a limited number of
variables are available for analyses.
Literature Cited Broderick, A.C., F. Glen, B.J.
Godley, G.C. Hays. 2003. Variation in
reproductive output of marine turtles. Journal
of Experimental Marine Biology and Ecology
28895-109. Frazer, N.B. and J.I. Richardson.
1986. The relationship of clutch size and
frequency to body size in loggerhead turtles,
Caretta caretta. Journal of Herpetology
2081-86. Hays, G.C. and J.R. Speakman. 1991.
Reproductive investment and optimum clutch size
of loggerhead sea turtles (Caretta caretta).
Journal of Animal Ecology 60455-462. Rivalan, P,
A. Prevot-Julliard, R. Choquet, R. Pradel, B.
Jacquemin, and M. Girondot. 2005. Trade-off
between current reproductive effort and delay to
next reproduction in leatherback sea turtle.
Oecologia 145564-574. Acknowledgements We would
like to thank Penelope Pooler (Virginia Tech) for
statistical consulting and data interpretation
and committee members Nancy Thompson and Marcella
Kelly.
Figure 3.
Special thanks to Penelope Pooler for
statistical consulting