The following supplement accompanies the article
Population characteristics, age structure, and growth dynamics of neritic juvenile green turtles in the
northeastern Gulf of Mexico
Larisa Avens1,*, Lisa R. Goshe1, Craig A. Harms2, Eric T. Anderson2, April Goodman Hall1,
Wendy M. Cluse3, Matthew H. Godfrey3, Joanne Braun-McNeill1, Brian Stacy4, Rhonda Bailey5, Margaret M. Lamont6
1
NOAA Fisheries, Southeast Fisheries Science Center, Beaufort Laboratory, 101 Pivers Island Road, Beaufort, North Carolina 28516, USA
North Carolina State University College of Veterinary Medicine, Center for Marine Sciences and Technology, Morehead City, North Carolina 28557, USA
3
North Carolina Wildlife Resources Commission, 1507 Ann St., Beaufort, North Carolina 28516, USA
4
NOAA/National Marine Fisheries Service/Protected Resources, University of Florida, College of Veterinary Medicine, Large Animal Clinical Sciences, Gainesville, Florida 32608, USA
5
Florida Fish and Wildlife Conservation Commission, Fish and Wildlife Research Institute, 100 8th Avenue SE, St. Petersburg, Florida 33701, USA
6
Department of Wildlife Ecology and Conservation, University of Florida, Gainesville, Florida 32611, USA
2
*Email: [email protected]
Marine Ecology Progress Series 458: 213–229 (2012)
Supplement 1. Additional figure
Fig. S1. Chelonia mydas. Graphical summary of the generalized additive models (GAMs) and generalized additive mixed models (GAMMs) used
to characterize the influence of different covariates on growth rates of green turtles stranded dead during the January 2010 cold stun event in St.
Joseph Bay, Florida. (a, b, c, d) GAMSCL, (e, f, g, h) GAMAge, (i, j, k) GAMMSCL, (l, m, n) GAMMAge. SCL is the mean straightline carapace
length for the growth interval, BCI is body condition index calculated as Fulton’s K, Sex is female or male, as determined through examination of
gonads during necropsy, FP is the fibropapilloma score of each turtle at necropsy with 0 indicating no FP symptoms and 1 denoting observation of
FP tumors, and Year is the estimated calendar year for each growth increment. Statistical output for each model is summarized in Table 6 in the
main paper. For each factor plotted, the solid line at the center represents the mean growth response (on a response scale centered around 0) for the
covariate conditioned upon the other covariates included in the model, and the dashed lines represent the extent of the 95% Bayesian credible
interval. The ‘rugs’ or short, vertical lines above the horizontal axes represent the numbers and/or distributions of covariate values
GAM response function
Growth rate (cm yr-1)
20
0.4
0.2
0.0
-0.2
GAM response function
Growth rate (cm yr-1)
GAM response function
Growth rate (cm yr-1)
1
0
-1
-2
-3
-4
-5
GAM response function
Growth rate (cm yr-1)
a)
b)
30
40
50
Female
Sex
60
70
c)
Male
1
0
-1
-2
-3
-4
-5
80
0.8
1.0
SCL(cm)
(cm)
SCL
1.2
0
1.4
BCI
CI
d)
0.4
0.2
0.0
-0.2
-0.4
1
FP
2
GAM response function
Growth rate (cm yr-1)
0
0.4
0.2
0.0
-0.2
GAM response function
Growth rate (cm yr-1)
GAM response function
Growth rate (cm yr-1)
e)
0
-1
-2
-3
-4
GAM response function
Growth rate (cm yr-1)
1
f)
5
10
Female
Sex
15
g)
Male
1
0
-1
-2
-3
-4
20
0.8
-0.4
1.0
Age (yrs)
1.2
0
1.4
BCI
h)
0.6
0.4
0.2
0.0
-0.2
1
FP
3
GAMM response function
Growth rate (cm yr-1)
GAMM response function
Growth rate (cm yr-1)
3
2
1
0
-1
-2
-3
GAMM response function
Growth rate (cm yr-1)
i)
j)
20
30
40
50
Female
60
70
SCL (cm)
80
3
2
1
0
-1
-2
-3
1998
2000
2002
2004
2006
2008
Year
k)
0.10
0.00
-0.10
Male
Sex
4
GAMM response function
Growth rate (cm yr-1)
GAMM response function
Growth rate (cm yr-1)
2
0
-2
-4
GAMM response function
Growth rate (cm yr-1)
l)
m)
5
10
15
Female
20
Age (yrs)
2
0
-2
-4
1998
2000
2002
2004
2006
2008
Year
n)
0.10
0.00
-0.10
Male
Sex
5