284
The Biology of Sea Turtles, Vol. II
Recent studies have highlighted the possibility of shifts in growth rates that do
not occur in conjunction with previously defined ontogenetic habitat shifts. Chaloupka (1998) analyzed age data from a study by Zug et al. (1995) and found
evidence of polyphasic growth within the pelagic stage of loggerheads in the Pacific
Ocean. Similarly, with the Kemp’s ridley, Chaloupka and Zug (1997) found evidence
of polyphasic growth, with the first peak in growth rates occurring around 15 cm
SCL, which is consistent with the size at which they begin to appear in coastal
waters (Turtle Expert Working Group, 2000). However, this is difficult to interpret
because the data set did not include any pelagic animals and the growth rate from
hatchling to the first benthic animal in the sample was modeled as essentially linear,
making the growth rates appear to slow after this period. The second peak in growth
rates occurred at approximately 45 cm SCL (Kemp’s ridleys mature at about 60 cm
SCL). This observation is supported by Schmid (1998), who found that although
not significant, average growth rates in the 40–50 cm SCL size class were higher
than in 30- to 40-cm and 50- to 60-cm SCL size classes. These growth shifts do not
relate to ontogenetic habitat shifts, but may be indicative of additional ontogenetic
shifts, whether they be diet, habitat, or physiological in nature, that result in changes
in growth rate.
The predominant method used to infer age-based growth rates in sea turtles
has been the comparison of growth in carapace length between captures–recaptures
of tagged individuals. This information is used in the interval forms of the von
Bertalanffy and/or logistic growth equations to produce a size-at-age growth curve
(Fabens, 1965; for example, see Frazer and Ehrhart, 1985). Because of the inaccessibility of all life stages, these curves are often prepared from data that span
only a portion of the life stages. The stage most commonly not included is the
pelagic (Mendonca, 1981; Frazer and Ehrhart, 1985; Frazer and Ladner, 1986;
Frazer et al., 1994; Turtle Expert Working Group, 2000). As discussed previously,
there is likely a shift in growth rates following the pelagic stage, and estimating
pelagic growth rates from benthic juvenile growth rates is inappropriate. Growth
rates from the adult life stage have also not been included when age-to-maturity
has been estimated (Mendonca, 1981; Frazer and Ladner, 1986; Ehrhardt and
Witham, 1992). Recently, authors have become more aware of this oversight and
are estimating only the length of time it takes an animal to grow through the size
classes for which they have data (Bjorndal and Bolten, 1988; Bjorndal et al., 1995;
National Marine Fisheries Service [NMFS], 2001; Braun-McNeill et al., in
review). Another potential problem with this application of the von Bertalanffy
growth curve is that this growth function implies a monotonically declining relationship with growth and age (Chaloupka and Musick, 1997). As discussed previously, this is contradictory in some studies because their data indicate a nonmonotonic relationship (Chaloupka and Limpus, 1997; Limpus and Chaloupka,
1997; Chaloupka and Zug, 1997).
Skeletochronology uses growth marks found in bone tissue to estimate age.
Numerous studies have applied this technique to sea turtles (Zug et al., 1986; 1995;
1997; Klinger and Musick, 1992; Zug and Parham, 1996; Parham and Zug, 1997;
Bjorndal et al., 1998; Zug and Glor, 1998; Coles et al., 2001; Zug , 2002; Snover
and Hohn, in review). Klinger and Musick (1992), Coles et al. (2001), and Snover
1123 book.book Page 284 Tuesday, November 12, 2002 7:43 AM
The Biology of Sea Turtles, Vol. II
Recent studies have highlighted the possibility of shifts in growth rates that do
not occur in conjunction with previously defined ontogenetic habitat shifts. Chaloupka (1998) analyzed age data from a study by Zug et al. (1995) and found
evidence of polyphasic growth within the pelagic stage of loggerheads in the Pacific
Ocean. Similarly, with the Kemp’s ridley, Chaloupka and Zug (1997) found evidence
of polyphasic growth, with the first peak in growth rates occurring around 15 cm
SCL, which is consistent with the size at which they begin to appear in coastal
waters (Turtle Expert Working Group, 2000). However, this is difficult to interpret
because the data set did not include any pelagic animals and the growth rate from
hatchling to the first benthic animal in the sample was modeled as essentially linear,
making the growth rates appear to slow after this period. The second peak in growth
rates occurred at approximately 45 cm SCL (Kemp’s ridleys mature at about 60 cm
SCL). This observation is supported by Schmid (1998), who found that although
not significant, average growth rates in the 40–50 cm SCL size class were higher
than in 30- to 40-cm and 50- to 60-cm SCL size classes. These growth shifts do not
relate to ontogenetic habitat shifts, but may be indicative of additional ontogenetic
shifts, whether they be diet, habitat, or physiological in nature, that result in changes
in growth rate.
The predominant method used to infer age-based growth rates in sea turtles
has been the comparison of growth in carapace length between captures–recaptures
of tagged individuals. This information is used in the interval forms of the von
Bertalanffy and/or logistic growth equations to produce a size-at-age growth curve
(Fabens, 1965; for example, see Frazer and Ehrhart, 1985). Because of the inaccessibility of all life stages, these curves are often prepared from data that span
only a portion of the life stages. The stage most commonly not included is the
pelagic (Mendonca, 1981; Frazer and Ehrhart, 1985; Frazer and Ladner, 1986;
Frazer et al., 1994; Turtle Expert Working Group, 2000). As discussed previously,
there is likely a shift in growth rates following the pelagic stage, and estimating
pelagic growth rates from benthic juvenile growth rates is inappropriate. Growth
rates from the adult life stage have also not been included when age-to-maturity
has been estimated (Mendonca, 1981; Frazer and Ladner, 1986; Ehrhardt and
Witham, 1992). Recently, authors have become more aware of this oversight and
are estimating only the length of time it takes an animal to grow through the size
classes for which they have data (Bjorndal and Bolten, 1988; Bjorndal et al., 1995;
National Marine Fisheries Service [NMFS], 2001; Braun-McNeill et al., in
review). Another potential problem with this application of the von Bertalanffy
growth curve is that this growth function implies a monotonically declining relationship with growth and age (Chaloupka and Musick, 1997). As discussed previously, this is contradictory in some studies because their data indicate a nonmonotonic relationship (Chaloupka and Limpus, 1997; Limpus and Chaloupka,
1997; Chaloupka and Zug, 1997).
Skeletochronology uses growth marks found in bone tissue to estimate age.
Numerous studies have applied this technique to sea turtles (Zug et al., 1986; 1995;
1997; Klinger and Musick, 1992; Zug and Parham, 1996; Parham and Zug, 1997;
Bjorndal et al., 1998; Zug and Glor, 1998; Coles et al., 2001; Zug , 2002; Snover
and Hohn, in review). Klinger and Musick (1992), Coles et al. (2001), and Snover
1123 book.book Page 284 Tuesday, November 12, 2002 7:43 AM