seaturtle.org : MTN : ARCHIVES : Sign In

I participated in two KRRT stakeholders meetings (April 2004 and February 2006) in Houston, Texas. At the first, input from stakeholders was solicited. At that meeting, and in an earlier Marine Turtle Newsletter guest editorial, I offered the following suggestions for consideration by the KRRT in revision of the recovery plan (Caillouet 2005):

1. Designate Texas beaches as known nesting habitats for Kemp’s ridley, and provide these beaches, nesters, eggs and hatchlings additional protection,

2. Provide additional protection to Kemp’s ridleys in Texas’ nearshore waters, which are important migratory, breeding and foraging habitats, and

3. Reevaluate the Kemp’s ridley head-start experiment’s accomplishments to date (see Caillouet 2000a), and fulfill prior commitments to determine the fate of head-started turtles released into the Gulf of Mexico.

Adding support to these suggestions, documented Kemp’s ridley nests on Texas beaches this year totaled 101, almost double the 2005 total (D. Shaver pers. comm.), and additional nestings probably were missed (see Pritchard 1990). Year 1985 marked the lowest reported annual nest count (740) at Rancho Nuevo, the primary nesting site in Tamaulipas, Mexico (see Table 1 in TEWG, 2000). It is remarkable that the Texas nest count has already reached 13.6 % of the lowest at Rancho Nuevo, and hatchling production from Texas nests could reach 7,000 this year (D. Shaver pers. comm.).

At the second stakeholders meeting, oral and visual presentations of draft portions of the revised plan were made by KRRT members. No copies of these materials were distributed to stakeholders before, during, or after the second meeting, nor had minutes of this meeting been posted on the recovery plan web site when this editorial was written. Stakeholders’ were invited to submit written comments to the KRRT within 2 weeks of this meeting; mine were submitted March 2006, based on recall and notes. For purposes of discussion, below are my further suggestions and their rationale, based on those comments:

4. Update annual hatchling and nest count data and estimates of parameter values or levels through the 2006 nesting season, and rerun the age-based model.

Rationale - The KRRT is using age-based modeling (TEWG 1998, 2000; Heppell et al. 2005) to guide revision of the recovery plan. Model estimates of population status and trends, rates of population recovery, effects of recovery actions, and time required to fulfill downlisting and delisting criteria, could influence establishment and prioritization of recovery actions. Although age-based models are very helpful tools, flexibility and caution should be exercised in their use.

The input variable in such models is annual production of hatchlings in Tamaulipas, and the output variable is annual number of nests laid in Tamaulipas, with a time lag related to estimated age to maturity. Model-estimated rates of increase in annual number of nests are sensitive to choices of data time series and parameter values or levels (TEWG 1998, 2000; Heppell et al. 2005). The TEWG (1998, 2000) and Heppell et al. (2005) employed single point estimates for some model parameters; e.g., a 1F:1M sex ratio was assumed, despite studies showing more female than male hatchlings are produced through beach corral hatchery operations in Tamaulipas, and sex ratios of older life stages are female-biased (Coyne in press; Landry et al. 2005; D. Shaver pers. comm.; reviewed by Wibbels in press). Note that the TEWG (1998, 2000) and Heppell et al. (2005) conducted model runs incorporating several different ages to maturity, but 10 years to maturity produced the best fit to the time series of data available to them.

The KRRT should rerun models using data time series updated through the 2006 nesting season, and incorporating ranges of parameter values, and include results in the revised plan, along with descriptions of the models, input time series of data, parameters, and outputs.

5. Estimate the relative contributions of annual hatchling production and post-1990 reductions in benthic stage mortality to the rate of increase in annual number of nests in Tamaulipas, and examine the cost-effectiveness of these two conservation approaches.

Rationale - One consequence of the observed exponential increase in annual number of Kemp’s ridley nests in Tamaulipas is enhanced interest in determining which of several conservation approaches produced this increase, as well as the relative cost-effectiveness of each. This was among topics explored by a project <http://www.nceas.ucsb.edu/fmt/doc?/frames.html> initiated in 2000 and aimed at developing new approaches to assess the Biggest Bang for the Buck in management alternatives for endangered species, including Kemp’s ridley (Selina Heppell pers. Comm.). I have seen an early draft of the assessment for Kemp’s ridley, which may eventually be published (Mark Plummer, pers. comm.). Its findings could be very useful to the KRRT.

Following the February 2006 stakeholders meeting, I developed a simple method to estimate relative contributions of (a) annual production of hatchlings in Tamaulipas, and (b) post-1990 reductions in benthic stage mortality, to the annual rate of increase in number of nests, using results from age-based modeling. Note that the TEWG (1998, 2000) and Heppell et al. (2005) incorporated a post-1990 multiplier, representing reduction in benthic stage mortality, to force model-estimated annual numbers of nests to track closely with actual annual numbers of nests in years following 1990. Note that reduction in post-1990 benthic stage mortality could encompass mortality reductions associated with all threats listed in Table 6-2 in CSTC (1990), and any additional ones that may be listed by the KRRT in the revised plan. Of primary interest among these is reduction in mortality due to turtle excluder devices (TEDs) in shrimp trawls. Contribution of the post-1990 mortality reduction to the estimated rate of increase in nests can be viewed as an upper limit or boundary for the influence of TEDs.

Two annual rates of increase (expressed in %) in nests can be estimated with an age-based model, one with and one without the post-1990 reduction in benthic stage mortality included. Heppell et al. (2005) estimated the mean annual rate of increase in nests during 1988-2003 at 14.9 % with the post-1990 multiplier included, but unfortunately, they did not report a rate estimated with the post-1990 multiplier excluded from the model. Earlier, Selina Heppell (pers. comm.) estimated the rate with the post-1990 multiplier included in the model to be 14 %, and the rate without the post-1990 multiplier included to be 5.7 %. Using these results, I calculated the relative contribution of hatchling production alone to be 5.7 %/14 % = 0.407, or 40.7 % (Table 1). The corresponding contribution of the post-1990 multiplier alone was 100 % - 40.7 %, or 59.3 %. The KRRT could use a similar approach on data updated through 2006. It should be especially interesting to see what effects female-biased sex ratios have on the magnitude of the estimated post-1990 multiplier.

I also decomposed the relative contribution of the post-1990 effect into its component parts, using as multipliers the geometric mid-points of mortality class intervals (Table 1), based on mortality categories listed in Table 6-2 in CSTC (1990). I simply multiplied the relative contribution of the post-1990 effect (59.3 %) by the geometric mid-point for each mortality class interval. The reduction in mortality associated with shrimping was the largest single contributor (51.6 %). The KRRT could use a similar approach with updated data and parameter ranges.

6. Caution should be exercised in imposing arbitrary limits on the number of nests protected annually in Tamaulipas.

Rationale - In 2005 and 2006, nests protected on Tamaulipas beaches exceeded 10,000. Annual reports to USFWS covering Tamaulipas operations for years 2000-2005 (provided by Jaime Peña, Gladys-Porter Zoo, Brownsville, Texas) stated: “Clearly we are going in the right direction but we cannot diminish our present effort if we are to succeed.” That “present effort” has been increasing exponentially with the exponential increase in nests laid on Tamaulipas beaches (Heppell et al. 2005), but resources supporting that effort are not limitless.

The TEWG (1998, 2000) and Heppell et al. (2005) considered arbitrary limits on nests protected in beach corral hatcheries, by simulating effects of such limits on annual rate of increase in nests in Tamaulipas. Nests not protected in beach hatchery corrals were assumed to be left in situ. Before arbitrary limits are imposed on numbers of nests protected in Tamaulipas, further studies should be conducted on hatchling production of nests left in situ. Currently, only 3-4% of the total nests per season in Tamaulipas are left in situ, and hatchling production from these nests appears very low. Eggs left in situ are still stolen by human poachers or destroyed by predators, parasites, tides, etc., as was the case before Mexico’s government began its beach protection operations in 1966. For four decades, standard practice has been to transfer eggs to beach corral hatcheries to protect them, and to protect hatchlings during their crawl to the surf. Leaving large numbers of nests in situ does not appear to be a good recovery option for Kemp’s ridley, unless in situ nest receive greater protection than they do at present. Leaving all or a large portion of nests in situ, could substantially diminish hatchling production and rate of recovery, and other conservation measures alone might not be adequate to recover and maintain the species.

An exponential increase in annual number of nests is observed even when no post-1990 multiplier is included in the age-based model (Heppell et al. 2005). This suggests that restored and enhanced annual hatchling production alone (through beach protection and corral hatchery operations in Tamaulipas) more than compensated the detrimental impacts of all threats before TEDs were required. Later, TEDs and other factors reduced mortality in large juveniles, subadults, and adults, thus accelerating the rate of increase in nests after 1990 (Heppell et al. 2005).

It is puzzling why there is reluctance to give credit for the reversal of the decline in the Kemp’s ridley population to beach protection and corral hatchery operations in Mexico (see Heppell et al. 2005). TED regulations were implemented in the late 1980s, yet signs that the downward trend in nests and nesters had slowed then reversed were evident earlier <https://seaturtle.org/mtn/special/MTN_Kemps.pdf>. This is not to say that additional efforts, such as the requirement for TEDS, that mitigate detrimental impacts from other major threats should be diminished or discontinued. Although major anthropogenic threats to Kemp’s ridleys have been assuaged, they have not been eliminated.

The growth curve for annual number of nests in Tamaulipas will eventually reach an inflection point, either before or after Kemp’s ridley is downlisted to threatened from endangered status. Peter Pritchard (pers. comm.) suggested that the inflection point in this growth curve will be reached at a much higher population level than would have occurred naturally, because (a) beach corral hatchery operations result in much higher production of hatchlings than in situ, and (b) most density-dependent destruction of eggs by nesters during future arribadas is avoided in beach corral hatcheries. Eventually, natural limitations on population growth will be imposed by environmental carrying capacity. To date, there is no evidence that density-dependent influences have slowed the rate of increase in nests. If limits are imposed on numbers of nests protected, whether arbitrary or driven by limitations of available resources, this could postpone recovery of the species, as compared to status quo conservation practices in Tamaulipas and elsewhere.

7. Proposed methods for ranking threats should be reconsidered and revised.

Rationale - There was an early focus by the KRRT on use of reproductive equivalents (REs) as multipliers in ranking threats. REs suggest that adults are relatively more important to recovery than eggs or hatchlings. Focus was shifted from REs to reproductive values (RVs) by the second stakeholders meeting in Houston. Although both REs and RVs provide indices of relative importance of various life stages, calculations of RVs and REs differ. The KRRT distinguished REs from RVs as follows: REs were scaled to adults, with 1 adult=400 hatchlings, and RVs were scaled to hatchlings, with 40-50 hatchlings=1 adult. The KRRT should reconsider use of RVs in ranking threats, because this risks giving a false impression that protecting eggs and hatchlings on nesting beaches is not as important as protecting larger life stages to recovery. Protecting all life stages is essential to recovery. The threats ranking approach presented by the KRRT at the February 2006 meeting involved age-based model-generated RVs multiplied by logarithms of the arithmetic mid-points between lower and upper limits of annual mortality class intervals for each threat. Stakeholders at different meetings voiced reservations about this approach (see minutes posted at http://www.fws.gov/kempsridley). RVs are based on mortality rates, and the logarithmic multipliers are derived from mortality ranges for threat categories, so multiplication of RVs by the logarithm of such arithmetic mid-points seems redundant. Also, “extremely skewed” sex ratios affect age-based model estimates of REs [and, likely, RVs as well], but the threshold level of skewness that affects such estimates was not defined by the KRRT (see KRRT meeting minutes, 24-25 October 2002, posted on the web site).

In any case, use of logarithms of mortality levels to rank threats to each life stage could be problematical for other reasons. First, the original mortality scale, expressed in annual numbers of deaths (e.g., Table 6-2 in CSTC 1990), better represents actual mortality than does a logarithmic transformation of deaths. Although logarithms provide a kind of “Richter scale” of threat severity, used as multipliers, they distort the actual numbers of deaths. Also, the arithmetic mid-point of annual mortality class intervals does not adequately represent the central tendency of mortalities in these intervals. Instead, the geometric mid-point is a statistically better estimator of the central tendency than an arithmetic mid-point. The antilog of the mid-point between logarithms of lower and upper limits of each mortality category is the geometric mid-point for the category (Table 1). If mortality-based multipliers are to be used in ranking threats, the geometric mid-points would be superior to logarithms as multipliers. If the mid-points between logarithms of mortality limits were used as multipliers, the threat due to shrimp trawling would not appear to be much greater than other threats (i.e., logarithms convert exponential scale to linear scale). Finally, I recognize that annual mortality levels shown in Table 6-2 in CSTC (1990) are no longer applicable, especially with regard to mortality due to shrimp trawling, because TEDs allow escapement of sea turtles caught in trawls, and shrimping effort and shrimping fleet size in the Gulf of Mexico are declining due to rising fuels costs and competition from imported shrimp.

The pre-TED reversal of the downward trend in nests in Tamaulipas prior to implementation of TED regulations should be emphasized in the revised plan, in the context of evaluating and ranking threats. TEDs and all other measures that reduce mortality in benthic stages do not produce eggs or hatchlings, so they contribute to future generations of Kemp’s ridleys only by saving some of the turtles produced from eggs. There are many threats to all Kemp’s ridley life stages, but they all pale in comparison to the impacts of reducing or discontinuing protection of nesters, eggs and hatchlings in Tamaulipas before the species has recovered.

8. Use of Hildebrand’s (1963) point estimate of the 1947 arribada as a benchmark for establishing recovery goals and criteria should be reconsidered.

Rationale - The recovery plan (USFWS and NMFS, 1992) established criteria for downlisting Kemp’s ridley from endangered to threatened status. Included among these criteria is a threshold of “10,000 females nesting in a season” that must be reached before Kemp’s ridley can be considered only threatened. According to Richard Byles (pers. comm., during TEWG deliberations in 1998, Miami, Florida), this threshold was calculated as 25% of an estimated 40,000-nester arribada (Hildebrand 1963), based on a movie made at Rancho Nuevo by Andres Herrera on one day in June 1947. This arribada size estimate has been used often as a benchmark from which to gauge population trends and status, without question as to its validity (Caillouet 2000b). Dickerson and Dickerson (2006) reexamined Herrera film images and concluded that the 1947 arribada “may have been overestimated due to the limited capabilities at the time for image analyses.” Their “best” estimate appeared to be 5,746 nesters in the 4-hr period during which Herrera filmed the arribada (Dickerson and Dickerson, 2006).

Carr (1977) presented “nesting arrival” (=arribada) numbers for 1970 (2,500 nesters) and 1974 (1,200 nesters), as well as for 1947. Frazer (1986) estimated “around 3,000” (actually 2,940 by my recalculation) Kemp’s ridley nesters in the 1947 season at Rancho Nuevo, using data from 1978-1985 (Woody 1985). Using the same data on nests, and 2.5 nests per nester per season, I recalculated nesters in 1947 to be 1,529. Given the extremely low nest numbers during 1978-1985, Frazer’s (1986) back-calculation likely underestimated the number of nesters in 1947.

As an alternative, I back-calculated nesters in the 1947 season, using annual numbers of nests (TEWG 2000) at Rancho Nuevo during the 1966-1977 seasons. This period encompassed the earliest available data and the rapid decline in nests that preceded implementation of the joint Mexico-U.S. restoration and enhancement program (Woody 1989). Coding the years 1947=0, 1948=1, 1949=3...1977=30, I fitted the following linear regression (r² =0.966) to ln(nests) versus coded year:

ln(nests)=12.0866-0.1769(coded year)

From this, I estimated the number of nests (= e^12.0866) in 1947 at 177,478, which is equivalent to 70,911 nesters (i.e., 177,478/2.5).

Table 1. Estimated percentage (%) contributions of annual hatchling production and reductions in benthic stage mortality to the rate of increase in Kemp’s ridley nests in Tamaulipas, Mexico

My back-calculated point estimate is of the same order of magnitude as Hildebrand’s (1963), but larger (as would be expected) because it represents the entire nesting season rather than a single arribada. All these back-calculations assume static rates of decline, an assumption that cannot be tested. Interestingly, the absolute value of the estimated annual rate of decline (17.7 %) in nesters during 1966-1977 is less than 3% higher than the mean annual rate of increase (14.9%) in nests estimated by Heppell et al. (2005) for years 1988-2003.

The KRRT should reconsider the Hildebrand (1963) arribada estimate for 1947, in the context of more recent estimates (e.g., Dickerson and Dickerson, 2006). Continued use of Hildebrand’s (1963) estimate as a benchmark from which to gauge Kemp’s ridley recovery, or for setting downlisting or delisting criteria is questionable.

Acknowledgments: I am grateful to Carole Allen, Michael Coyne, Benny Gallaway, André Landry, and Peter Pritchard who reviewed the manuscript and made many helpful suggestions. I also appreciated comments and suggestions from an anonymous reviewer and from Marine Turtle Newsletter Editor Matthew Godfrey.

CAILLOUET, C.W., JR. 2000a. Sea turtle culture: Kemp’s ridley and loggerhead turtles. In: R.R. Stickney (Ed.), Encyclopedia of Aquaculture, John Wiley & Sons, Inc., New York.

CAILLOUET, C.W., JR. 2000b. Letter to the editors: which Kemp’s ridley nest numbers are correct? Marine Turtle Newsletter 87:14.

CAILLOUET, C. W., JR. 2005. Guest Editorial: Wild and headstarted Kemp’s ridley nesters, eggs, hatchlings, nesting beaches and adjoining nearshore waters in Texas should receive greater protection. Marine Turtle Newsletter 110:1-3.

CARR, A. 1977. Crisis for the Atlantic ridley. Marine Turtle Newsletter 4:2-3.

CSTC (Committee on Sea Turtle Conservation). 1990. Decline of the Sea Turtles: Causes and Prevention. National Academy Press, Washington, D.C., 259 pp.

COYNE, M. in press. Population sex ratio and its impact on population models. In: P. Plotkin (Ed.). Biology and Conservation of Ridley Turtles. Johns Hopkins University Press, Baltimore, Maryland.

DICKERSON, V.A. & D.D. DICKERSON. 2006. Analysis of arribada in 1947 Herrera film at Rancho Nuevo, Mexico. In: Frick, M., A. Panagopoulou, A.F. Rees & K. Williams (Compilers) Book of Abstracts. Twenty Sixth Annual Symposium on Sea Turtle Biology and Conservation. International Sea Turtle Society, Athens, Greece. pp 290-291.

FRAZER, N.B. 1986. Kemp’s decline: special alarm or general concern? Marine Turtle Newsletter 37:5-7.

HEPPELL, S.S., D.B. CROUSE, L.B. CROWDER, S.P. EPPERLY, W. GABRIEL, T. HENWOOD, R. MÁRQUEZ, & N.B. THOMPSON. 2005. A population model to estimate recovery time, population size, and management impacts on Kemp’s ridley sea turtles. Chelonian Conservation & Biology 4:767-773.

HILDEBRAND, H.H. 1963. Hallazgo del área de anidación de la tortuga marina “lora” Lepidochelys kempi (Garman), en la costa occidental del Golfo de México. Sobretiro de Ciencia, México 22:105-112.

LANDRY, A.M., JR., D.T. COSTA, F.L. KENYON, II, & M.S. COYNE. 2005. Population characteristics of Kemp’s ridley sea turtles in nearshore waters of the upper Texas and Louisiana coasts. Chelonian Conservation & Biology 4:801-807.

PLOTKIN, P. (Ed.). In press. Biology and Conservation of Ridley Turtles. Johns Hopkins University Press, Baltimore, Maryland.

PRITCHARD, P.C.H. 1990. Kemp’s ridleys are rarer than we thought. Marine Turtle Newsletter 49:1-3.

TEWG (Turtle Expert Working Group). 1998. An assessment of the Kemp’s ridley (Lepidochelys kempii) and loggerhead (Caretta caretta) sea turtle populations in the Western North Atlantic. NOAA Technical Memorandum NMFS-SEFSC-409, 96 pp.

TEWG (Turtle Expert Working Group). 2000. Assessment Update for the Kemp’s ridley and loggerhead sea turtle populations in the Western North Atlantic. NOAA Technical Memorandum NMFS-SEFSC-444, 115 pp.

USFWS (U.S. Fish and Wildlife Service) & NMFS (National Marine Fisheries Service). 1992. Recovery Plan for the Kemp’s Ridley Sea Turtle Lepidochelys kempii. USFWS, Albuquerque, NM, and NMFS, Washington, D.C., 40 pp.

WIBBELS, T. In press. Sex determination and sex ratios in ridley turtles. In: P. Plotkin (Ed.). In press. Biology and Conservation of Ridley Turtles. Johns Hopkins University Press, Baltimore, Maryland.

WOODY, J.B. 1985. Kemp’s ridley continues decline. Marine Turtle Newsletter 35:4-5.

WOODY, J.B. 1989. International efforts in the conservation and management of Kemp’s ridley sea turtle (Lepidochelys kempi), In: C.W. Caillouet, Jr. & A.M. Landry, Jr. (Ed.). Proceedings of the First International Symposium on Kemp’s Ridley Sea Turtle Biology, Conservation and Management. Texas A & M University, Sea Grant College Program, TAMU SG 89 105, vi plus 260 pp.