1
Introduction

Long lifespans and wide-ranging migrations make sea turtles difficult to monitor and susceptible to many sources of mortality, including direct and incidental “takes” (basically any potential impact on a turtle or its behavior1) that occur in human activities. All six of the species that occur in U.S. waters2—loggerhead (Caretta caretta), green (Chelonia mydas), hawksbill (Eretmochelys imbricate), Kemp’s ridley (Lepidochelys kempii), olive ridley (Lepidochelys olivacea), and leatherback (Dermochelys coriacea)—are listed as endangered or threatened under the Endangered Species Act, thereby prohibiting their direct harvest. (The seventh sea-turtle species is the flatback [Natator depressus], which is only found in the waters around Australia, Papua New Guinea, and Indonesia.) However, permits are available for some activities, such as shrimp fishing, dredging, and sand replenishment, that allow a specified number of incidental takes (i.e., a number of individuals that may be accidentally killed before the activity must stop). Therefore, accurate assessments are necessary to evaluate the status and trends of populations.

Regulatory decisions, such as allowing incidental takes, are best implemented with estimates of absolute population numbers, but these are unavailable because of the broad oceanic distribution of sea turtles and the very small proportion of each population that comes to land (nest-

1

50 CFR 17.3.

2

U.S. waters not only refers to waters around U.S. states but also waters around U.S. territories, such as American Samoa, Puerto Rico, Northern Mariana Island, Guam, the U.S. Virgin Islands, and Palmyra Atoll.



The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement



Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.

OCR for page 11
1 Introduction Long lifespans and wide­ranging migrations make sea turtles dif­ ficult to monitor and susceptible to many sources of mortality, including direct and incidental “takes” (basically any potential impact on a turtle or its behavior1) that occur in human activities. All six of the species that occur in U.S. waters2—loggerhead (Caretta caretta), green (Chelonia mydas), hawksbill (Eretmochelys imbricate), Kemp’s ridley (Lepidochelys kempii), olive ridley (Lepidochelys olivacea), and leatherback (Dermochelys coriacea)—are listed as endangered or threatened under the Endangered Species Act, thereby prohibiting their direct harvest. (The seventh sea­ turtle species is the flatback [Natator depressus], which is only found in the waters around Australia, Papua New Guinea, and Indonesia.) However, permits are available for some activities, such as shrimp fishing, dredg ­ ing, and sand replenishment, that allow a specified number of incidental takes (i.e., a number of individuals that may be accidentally killed before the activity must stop). Therefore, accurate assessments are necessary to evaluate the status and trends of populations. Regulatory decisions, such as allowing incidental takes, are best implemented with estimates of absolute population numbers, but these are unavailable because of the broad oceanic distribution of sea turtles and the very small proportion of each population that comes to land (nest­ 1 50 CFR 17.3. 2 U.S. waters not only refers to waters around U.S. states but also waters around U.S. terri­ tories, such as American Samoa, Puerto Rico, Northern Mariana Island, Guam, the U.S. Virgin Islands, and Palmyra Atoll. 

OCR for page 11
 ASSESSMENT OF SEA-TURTLE STATUS AND TRENDS ing adult females) (Turtle Expert Working Group, 2000, 2009). Current assessment models in the United States are based on good census data on nests and nesting females, but they lack key demographic information for extrapolating the counts to total population size accurately (e.g., Turtle Expert Working Group, 2007). With a paucity of data and of analysis of growth rates, annual survival, and reproductive frequency, current models used by the agencies can provide only relative or probabilistic numbers and trends under often limiting assumptions. For example, population A is larger than population B, population A is likely to decrease in the future, or population A is larger than it was five years ago (Heppell et al., 2003; Conant et al., 2009). Thus, they can only demonstrate population trends for segments of the population or make general predictions about the effects of disturbances on population persistence and recovery. Because sea turtles migrate across whole ocean basins, population assessments require an international context. Global activities, such as development on nesting beaches, killing of turtles for food, and incidental capture in commercial fisheries, can contribute to sea­turtle declines and affect populations found in U.S. waters (e.g., Conant et al., 2009). Management efforts appear to have slowed or reversed declines in some populations, such as Kemp’s ridley (Turtle Expert Working Group, 2000) and Hawaiian green turtles (National Marine Fisheries Service and U.S. Fish and Wildlife Service, 2007a), but the status of many populations is still unknown or poorly understood (Table 1.1), and none have reached their recovery goals. According to the 2007 five­year status updates for each species (National Marine Fisheries Service and U.S. Fish and Wildlife Service, 2007a, b, c, d, e, f), there are many uncertainties in population structure, in productivity trends, and in the nonbreeding population of most species. However, data needed for accurate assessments of most populations are not available, prohibiting diagnostic evaluations that can benefit management. There have been recommendations for improved data collection and analysis for status determination and assessment modeling in nearly every report and status review document published by the two federal agencies responsible for sea­turtle management—the National Marine Fisheries Service (NMFS) and the U.S. Fish and Wildlife Service (USFWS). As stated in the green turtle review, “the paucity of information regarding these [demographic] aspects continues to inhibit effective modeling of populations and prevents a full understanding of which nesting concen­ trations are most at risk” (National Marine Fisheries Service and U.S. Fish and Wildlife Service, 2007a). The reports repeatedly state a need for addi­ tional information on genetic relationships among nesting populations, effects of coastal and pelagic fisheries, identification of foraging areas, and identification of threats at foraging areas as key data needs for assessment

OCR for page 11
 INTRODUCTION TABLE 1.1 Current Endangered Species Act Listing Status and Trends of Sea Turtlesa Reported Trend of Subpopulations or Nesting Aggregations Listing ↑ ↓ Species Geographic Area Status − ? Green turtle Florida Endangered 1 0 0 0 Other western Atlantic Threatened 3 0 2 0 Western Pacific Threatened 2 0 1 1 Central Pacific (U.S.) Threatened 1 0 0 0 Eastern Pacific Endangered 1 0 1 0 Eastern Pacific Threatened 0 0 1 1 Hawksbill turtleb U.S. Virgin Islands, Puerto Endangered 3 0 0 1 Rico Other Caribbean Endangered 5 9 0 12 Central Pacific (U.S. and Endangered 1 2 01 holdings) Central Pacific (other) Endangered 0 3 0 1 Kemp’s ridley turtle Gulf of Mexico Endangered 1 0 0 0 Olive ridley turtle Eastern Pacific (Mexico) Endangered 5 0 4 0 Eastern Pacific Threatened 1 2 1 8 Leatherback turtle Florida, U.S. Virgin Islands, Endangered 3 0 0 0 Puerto Rico Other Caribbean 5 1 4 9 Western Atlantic Eastern Pacific 05 0 0 Western Pacific 02 0 4 Loggerhead turtle U.S. western Atlantic 13 1 0 North Pacific (Japan) 3 12 0 0 a Listed here are the reported number of increasing (↑), decreasing (↓), stable (−), or unknown (?) subpopulations or nesting aggregations that nest in the United States or U.S. territories or that commonly occur in U.S. waters. Trends based on numbers of nests or nest ­ ing females. Data from National Marine Fisheries Service and U.S. Fish and Wildlife Service (2007a, b, c, d, e, f). b Based on “recent trend” (in last 20 years). and management. Likewise, the Turtle Expert Working Group has regu­ larly highlighted inadequacies of its assessments for determining popula­ tion size, trends (except for nesting females), maximum take levels, and evaluation of the success of various management strategies (Turtle Expert Working Group, 1998, 2000, 2007, 2009; Table 1.2). Some recent incidental take statements—required as part of an incidental take permit—have made clear how important it would be to have that information (National Marine Fisheries Service, 2005; Merrick and Haas, 2008).

OCR for page 11
 ASSESSMENT OF SEA-TURTLE STATUS AND TRENDS TABLE 1.2 Summary of Reports by the Turtle Expert Working Group and Other Loggerhead Assessmentsa Geographic Document Species Area Reference Year Methods LDMb, PBRc, Loggerhead North Pacific Bolten et al. 1996 VORTEX PVAd, RAMAS Stage PVAe, TURTSIMf Kemp’s Western Turtle Expert 2000 Trend analysis; LDM and LSMg fit ridley North Working Group Atlantic to nest number Loggerhead Western Turtle Expert 2000 Trend analysis North Working Group Atlantic Loggerhead Western National Marine 2001 Trend analysis North Fisheries Service (nests); LDM Atlantic Southeast Fisheries Science Center

OCR for page 11
 INTRODUCTION Status Conclusion Conclusion Quotes Recommendations Existing incidental “Although the workshop “Develop and implement a mortality would not was an excellent exercise comprehensive quantitative have a significant effect, in population model framework for marine assuming a stable integration, more turtle recovery management population; maximum research is required to including…robust allowable removal, further develop a suite procedure for monitoring 28–800, depending on life of analytical tools robust turtle populations and stage affected to shortcomings in measuring progress toward biological knowledge and recovery goals.” data on human­caused mortality.” Population increasing; “It is clear to the TEWG 1. Obtain key vital rates, recovery goal achievable that continued work especially survival and life­ by 2020; cannot estimate towards developing stage duration. acceptable removal rates estimates of take which 2. Provide adequate do not negatively impact observer coverage to recovery is limited in statistically evaluate take meaning without a clear throughout the species understanding of the range. status and condition of these stocks. We are confident that future assessment teams can make progress as more data become available.” South Florida stable or “No method for 1. Obtain key vital rates, increasing; northern setting strandings especially survival and life subpopulation recovery limits was completely stage duration. goals unlikely to be satisfactory to all Group 2. Provide adequate members.h Significant met; cannot estimate observer coverage to acceptable removal rates data gaps exist which evaluate take statistically limit the pursuit of throughout the species complete age­specific range. assessments.” 3. Define subpopulations and rates of mixing in foraging areas. Northern subpopulation “It is unlikely that any “It is recommended that stable; Florida loggerhead nesting actions to reduce juvenile subpopulation increasing; subpopulation under mortality be identified 150–1,200 turtles killed in the status quo will be through research and longlines each year extirpated over the next implemented as soon as few years.” feasible.” continued

OCR for page 11
 ASSESSMENT OF SEA-TURTLE STATUS AND TRENDS TABLE 1.2 Continued Geographic Document Species Area Reference Year Methods Leatherback Western National Marine 2001 Trend analysis North Fisheries Service (nests) Atlantic Southeast Fisheries Science Center Leatherback Atlantic Turtle Expert 2007 Trend analysis; Working Group Bayesian state space analysis of trends Loggerhead Western National Marine 2009 LDM North Fisheries Service Atlantic Southeast Fisheries Science Center DA with SQE,i Loggerhead Western Conant et al. 2009 North LDM (probabilistic growth rates)j Atlantic

OCR for page 11
 INTRODUCTION Status Conclusion Conclusion Quotes Recommendations Population increasing “While the longline “It is recommended that in Florida and northern fishery and the U.S. research begin immediately Caribbean; decreasing in trawl fishery may not to identify and quantify the French Guiana; 150–530 be the immediate cause rate of mortality from the kills in longlines annually in declines in nesting longline fishery, as well as in French Guiana, they mortality rates from other could be contributing to fisheries.” these declines.” Adult population stable, “Nest numbers could “Analyses should be increasing in some fluctuate considerably interpreted with caution areas; 10,000–31,000 due to individual due to high parameter and adult females, excluding variance in remigration data uncertainty; efforts unknown nesting in intervals, clutch number, should be made to develop Africa and the reduced site a collaborative international fidelity in leatherbacks.” research plan on population dynamics and stock structure; need to estimate demographic parameters.” 1. Adult female 1. “This model cannot 1. Devote more time population, 20,000– effectively address any and resources to the 40,000+; total population specific question of what development of improved highly uncertain. the effect of mortality stock assessment models of 2. Any reductions in in a given fishery might sea turtles. mortality will improve be without making very 2. More in­water capture– recovery potential, but large assumptions that recapture and telemetry even elimination of some are difficult to justify.” studies…to improve anthropogenic mortality 2. “Predicting future estimates of survival and sources may not be populations of growth. sufficient to prevent loggerhead sea turtles extinction. is very uncertain due in part to large uncertainty in our knowledge of loggerhead life history.” Nine distinct population “This approach (LDM)… N/A segments (DPSs) produced a wide range of identified globally; three­ results.” fifths of DPSs with good time series show high risk of extinction; some DPSs show increasing trends, but all have possibly unsustainable anthropogenic mortality and extinction risk. continued

OCR for page 11
 ASSESSMENT OF SEA-TURTLE STATUS AND TRENDS TABLE 1.2 Continued Geographic Document Species Area Reference Year Methods Loggerhead Western Turtle Expert 2009 Nesting trend North Working Group analysis; juvenile Atlantic size and abundance trends aThis is not an exhaustive list but presents examples of the methods, conclusions, and recommendations in assessment reports. b LDM = linear deterministic matrix. c PBR = potential biological removal. d VORTEX = individual­based stochastic simulation program for population viability analysis (PVA). e RAMAS Stage = stochastic matrix PVA. f TURTSIM = length­based model developed by Weatherall at the Pacific Islands Fisheries Science Center. The fundamental theme underlying this report is that abundance assessment is essential, but abundance information alone is insufficient to diagnose the causes of trends in sea­turtle populations or to predict them. That is particularly true because abundance estimates in the United States are generally restricted to nesting females, which probably make up less than 1% of total population size (Crowder et al., 1994; Turtle Expert Working Group, 2000; Chaloupka, 2002a; Heppell et al., 2003). In addition to reliable abundance estimates of multiple segments of each population, understanding key demographic processes, such as annual survival and breeding probabilities, is essential. WHAT IS AN ASSESSMENT? Population assessments seek to measure the current status, evaluate trends over previous years, and predict the status of populations under various management scenarios by quantitatively evaluating population abundance and assessing such demographic parameters as productivity and survivorship (called “vital rates” that indicate the potential for change in a population). Population assessments are required when not all mem ­ bers of the population can be counted accurately—the case with almost

OCR for page 11
 INTRODUCTION Status Conclusion Conclusion Quotes Recommendations All nesting “We have no time series Fundamental life­history subpopulations in of any demographic and census data should be decline; increase in large parameters that are collected and evaluated. neritic juveniles; low appropriate to examine juvenile recruitment these hypotheses (for decline) quantitatively. We have bits and pieces of information, but lack the specific census and mortality data necessary to characterize and monitor trends.” g LSM = linear stochastic matrix. h This was also the case for the 1998 Turtle Expert Working Group assessment for these species, where potential biological removal (PBR) and strandings trend analysis were sug ­ gested as methods for setting limits on strandings to trigger management action. i DA with SQE = diffusion approximation with susceptibility to quasi­extinction (Snover and Heppell, 2009). j Results presented as probability of population decline given current estimates of anthro­ pogenic mortality. all animal populations except small populations of visible, individually identifiable animals, such as California condors (Gymnogyps californianus), most of which are tagged. The habit of most sea­turtle species to congre­ gate in relatively small areas ashore to deposit egg clutches (i.e., the eggs produced and laid at a single time) provides an opportunity to count animals, but these animals constitute only a small part of the total popu ­ lation. That feature of sea­turtle biology is shared by anadromous fish, such as salmon (family Salmonidae), which return as adults to specific spawning areas in freshwater. Like turtles, salmon except pink salmon (Oncorhynchus gorbuscha) have overlapping generations, but sea­turtle reproduction is more complex because their adult lifespan is long, and females do not breed every year. General life­history traits of sea turtles are provided in Box 1.1. When more is known about a population, including age, spatial dis­ tribution, and genetics, more sophisticated models can be used for assess­ ment wherein productivity can be evaluated for specific age groups and birth years. The value of a more sophisticated model is that, in theory, more of the uncertainty in life­history processes and vital rates can be evaluated explicitly. Ideally, a population assessment will reflect cur­ rent population status and productivity accurately and can be used to

OCR for page 11
0 ASSESSMENT OF SEA-TURTLE STATUS AND TRENDS Box 1.1 Some Distinctive Features of Sea-Turtle Life Historiesa • Long-lived with delayed maturity (at least 10 years, maximum of 30 years or more) • Iteroparous (nest more than once but not every year) • Life history in marine (foraging and mating) and terrestrial (nesting) habitats • Overlapping generations • Undertake long migrations and disperse widely • Nesting populations on beaches, consisting only of adult females and their eggs • Usually deposit several egg clutches in a breeding season (the number of clutches produced by a female in a season is termed clutch frequency) aNot all species have the following traits to an equal degree. predict the effect of future management practices on the population. As in almost all marine species, population assessments of sea turtles are challenging because of a lack of critical data or a difficulty in accessing data. Box 1.2 discusses how sea­turtle assessments compare with fisheries assessments. The term assessment is used somewhat generically to describe an evaluation of data to determine the status and trends of a population relative to its condition in the past or its potential condition. The results of assessments are used to address management questions, such as the maximum human­induced mortality that a population can absorb with­ out declining substantially. Key components of the assessment procedure include independent evaluation of data quality, model suitability and robustness, and development of biologically reasonable reference points for status evaluation and management (National Research Council, 1998). A thorough population assessment needs to include a description and evaluation of change over time and space in the following areas: • population structure (e.g., species, subspecies, distinct population segments; see Chapter 2) • population lifecycle and demography (e.g., life stages, rates of survival, reproduction; see Chapters 3 and 5) • population abundance and trends (e.g., evaluation and extrapola ­ tion of population indexes; see Chapter 4) • population ecology and behavior (e.g., habitat, distribution and movements, predators and prey, disease, parasites, contaminants)

OCR for page 11
 INTRODUCTION Box 1.2 Sea-Turtle Assessments and Fisheries Assessments A population (stock)a assessment is an evaluation of the status and trends of a population of organisms. It is usually motivated by a concern for the effects of human activities on those organisms. NMFS has a large repository of assessment tools that have been rigorously evaluated and applied to fisheries management, but it does not have a standardized framework for data evaluation and modeling of sea-turtle populations. Similarities of some characteristics of biology, data, and management needs in sea turtles and marine fish make the application of some fisheries assessment methods to sea turtles possible, but the two kinds of popula- tions also differ in some key respects. With respect to biology, sea turtles are similar to long-lived, slow-growing fish, but time lags from birth to reproductive maturity (decades) are much longer than most fish. Fish with similarly long lives include some Pacific rockfish (Sebastes spp.), dogfish (order Squaliformes), and sturgeon (family Acipenseridae). Sea turtles are highly migratory and occur in different habitats over their lifetimes. The population structure of sea turtles is highly complex; natal homing (the process by which animals return to their birthplace to reproduce) by females creates genetically distinct nesting units, as with salmon. However, although some males exhibit natal homing, there is genetic mixing through males that have an oppor- tunity to mate with females from different units, a pattern that is less common among fish species. Fishery-independent data, an important part of fish-stock assessments, on turtles exist in many forms (nesting beach surveys and in-water surveys) but are not always collected with comprehensive or standardized methods that allow their incorporation into population assessments; data from many excellent sources are proprietary and unavailable for evaluation. Because available fishery data on catches of turtles are based on bycatch from more than one kind of fishery and observer coverage of many U.S. fisheries is low and of many international fisheries is absent, fishery-dependent data for estimating stock abundance, which can be important for commercial species, are not as effective for estimating turtle abundance. Finally, length distributions are available from some bycaught animals, but age distributions are not—in contrast, the ages of most fish can be determined reliably. In fisheries management, assessment models are used to predict rates of change in biomass and productivity of a population to set harvest limits. In sea- turtle management, assessment models are used to evaluate the status of the population relative to recovery goals, to compare relative efftects of different human activities and natural stressors on populations, and to determine whether human activities that result in turtle mortality will impede recovery or increase extinction risk. Assessment of sea-turtle population status and trends is conducted according to the requirements of the Endangered Species Act and through expert working groups, recovery plan teams, and biological review teams convened by NMFS Fisheries Science Centers. Worldwide, fish-stock assessments usually are prepared by fishery agencies and—in the United States—stock-assessment teams associated with NMFS and regional management councils. Assessments continued

OCR for page 11
 ASSESSMENT OF SEA-TURTLE STATUS AND TRENDS Box 1.2 Continued of fish stocks undergo rigorous review, and reports on turtle populations produced by the above-mentioned groups and teams have recently also undergone external scientific review. a A “population” is usually defined as a group of organisms whose members interbreed and are subjected to processes that result in a common birth, mortality, and growth rate. All members of a species can potentially interbreed, and some migration occurs among popula- tions. An example of a population of sea turtles might be all the turtles that breed on a particular beach. “Stock” (synonymous with population in this case) refers to a group with common vital rates and is often used by fisheries scientists to identify a population that they seek to manage. For a detailed discussion, see Chapter 2. • population size (e.g., numbers of individuals, age structure, sex ratio) • current and projected threats (e.g., human­caused injury or mortal­ ity, habitat destruction, climate change) • sources of variability (e.g., genetic, demographic, environmental, catastrophic) Assessments of sea­turtle populations conducted by NMFS have included all those elements but to varied degrees of detail and quan­ titative evaluation (Table 1.2). To be useful in decision­making, assess­ ment requires more than simple description of trends; the large and dif­ fuse nature of sea­turtle populations makes extrapolation of trends over time, space, and generations difficult at best and potentially misleading. Observed and potential changes in sea­turtle populations through time need to be assessed with age­structured models to determine population­ wide status accurately and to diagnose causes of population change. Like­ wise, heuristic evaluation of possible futures under data­poor conditions has limited utility because management often requires “high­resolution” results—accurate and precise predictions of effect so that it can set take regulations and evaluate the outcomes of targeted management actions. ASSESSMENT CASE STuDIES To illustrate the importance of having demographic information, as well as abundance estimates in assessing sea­turtle populations, the com­ mittee briefly outlines here a comparative case study of two of the most

OCR for page 11
 INTRODUCTION important loggerhead sea­turtle populations in the world—the genetic stock that nests along the Atlantic coast of Florida (Ehrhart et al., 2003) and the genetic stock that nests along the Pacific coast of Australia (Limpus and Limpus, 2003a). The assessment of the Florida turtles was severely hampered by the lack of demographic information, but demographic information available on the Australian population allowed a much more thorough evaluation of hypotheses. The loggerhead sea turtle is considered to be a globally endangered species (International Union for Conservation of Nature, 2010). It has some major nesting populations that are in decline, such as in the north­ western Atlantic (Witherington et al., 2009), and other major nesting pop ­ ulations that are increasing, such as in the Pacific (Chaloupka et al., 2008a) and southwestern Atlantic (Marcovaldi and Chaloupka, 2007). Increases in loggerhead nesting populations are usually attributed to conservation measures (Marcovaldi and Chaloupka, 2007), and declines are usually attributed to climate change (Chaloupka et al., 2008a) or exposure to anthropogenic hazards, such as pelagic (open ocean) fisheries (Lewison et al., 2004) or coastal fisheries (Peckham et al., 2007). But often the data that would support confidence in those attributions are lacking. Most assessments of loggerhead sea­turtle population trends have been based on long­term monitoring of the seasonal beach nesting activ­ ity of adult females (Marcovaldi and Chaloupka, 2007; Chaloupka et al., 2008a; Witherington et al., 2009). However, monitoring only female nest­ ing activity provides insufficient information for population assessment because adult females usually skip one or more breeding seasons, and nest counts provide no information on demographic structure because immature, adult male, and non­breeding female components are not sampled. Robust assessment of the status and trend of a loggerhead sea­ turtle population suitable for population assessment and conservation­ management planning requires additional information and depends on sampling of the entire demographic structure of a population resident in the foraging grounds and on deriving a range of estimates of key demo­ graphic parameter of the population. The spatial and temporal variation in nesting activity of the north­ western Atlantic loggerhead population that nests along the Atlantic coast of Florida has been monitored for more than 20 years. These nesting pop ­ ulations have declined substantially over the last 10 years (Figure 1.1), but the causes remain elusive because of a lack of demographic parameters to help to diagnose the trends (Witherington et al., 2009). As a result, man­ agement agencies have not been able to predict the effectiveness of con ­ servation strategies. A recent Turtle Expert Working Group (2009) review of the status of the loggerhead population nesting along the U.S. Atlantic coast clearly recognized that limitation: “Our ability to assess the current

OCR for page 11
 ASSESSMENT OF SEA-TURTLE STATUS AND TRENDS FIGURE 1.1 Annual total nest counts for loggerhead sea turtles on Florida beaches, 1989–2006. The trend line was estimated by fitting a three­knot restricted cubic spline curve to the total counts via negative binomial regression. (Reprinted from Witherington et al., 2009; with permission from Ecological Society of America.) status of all segments of the Western North Atlantic loggerhead subpopu­ lations is limited. We have bits and pieces of the information but lack the specific census and mortality data necessary to characterize and monitor trends for these populations.” In this case, long­term abundance estimates without accompanying estimates of key demographic parameters were not sufficient to diagnose the cause(s) of the decline in nest numbers and to design suitable risk­mitigation or population­recovery strategies. The southern Great Barrier Reef (SGBR) loggerhead is one of two Pacific populations, and much better information is available on its popu­ lation trends. Loggerheads in this population nest on coral cays in the SGBR region and along the adjacent Australian mainland (Limpus and Limpus, 2003b). The SGBR loggerhead nesting population has been moni­ tored extensively for more than 30 years (Limpus and Limpus, 2003a; Chaloupka et al., 2008a), and several foraging­habitat aggregations of the population have been extensively monitored for decades with a com ­

OCR for page 11
 INTRODUCTION prehensive capture–mark–recapture program (Chaloupka and Limpus, 2001; Limpus and Limpus, 2003a). The tagging program is coupled with laparoscopic examination of female and male loggerheads of all ageclasses residing in nearby coastal habitats. The assessments of reproductive con­ dition support sex determination and direct estimates of breeding rates. Not only is the spatial and temporal variation in SGBR loggerhead nesting abundance well known but so are key demographic parameters, such as sex­specific and ageclass­specific survival probabilities, sex­specific breed­ ing rates, and trends in sex­specific and ageclass­specific foraging­habitat abundance estimates (Chaloupka and Limpus, 2001, 2002; Chaloupka, 2003a). The sex­specific and ageclass­specific abundance and demographic parameter estimates derived for the SGBR loggerhead population have provided a sound foundation for assessing the relative risks posed by exposure over the last 30 years to various anthropogenic hazards, such as coastal fisheries, pelagic fisheries, feral­animal predation of nests, coastal development effects on nesting habitat, and long­term climate change (Chaloupka, 2003a; Chaloupka et al., 2008a). For this population, it is pos­ sible to determine, for example, whether per capita fecundity (i.e., indi ­ vidual reproductive output) has changed, whether survival probabilities have declined, or whether the proportion of mature females has changed. It was then possible to diagnose the declining ageclass­specific abundance during the 1980s and 1990s as attributable to predation by foxes on the coastal nesting beaches and to incidental capture in coastal trawl fisheries (Chaloupka, 2003a). Both hazards have now been mitigated by federal and state government conservation agencies, and this has resulted in an apparent recovery of the stock (Chaloupka et al., 2008a). Some of the factors contributing to the ability of the SGBR loggerhead program to make those critical determinations were (1) a long­term research program maintained by a single agency with dedicated personnel, (2) a spatially extensive capture–mark–recapture program on the nesting beaches, and (3) additional capture–mark–recapture efforts in the coastal foraging habi­ tats coupled with laparoscopy to assess both sex and breeding status (see review in Limpus and Limpus, 2003a). The need to combine abundance trends with demographic parameters is important for all species and has been recognized for several sea­turtle species, including leatherbacks (Dutton et al., 2005), green turtles (Solow et al., 2002; Seminoff et al., 2003; Bjorndal et al., 2005), and loggerheads (Chaloupka and Limpus, 2001). These authors based their conclusions on a variety of assessments, and this committee agrees with them. For this reason, the committee has not provided a detailed review of a large number of assessments but instead has focused on methods for improving the collection of necessary data.

OCR for page 11
 ASSESSMENT OF SEA-TURTLE STATUS AND TRENDS THE PRESENT STuDY In light of the above concerns, NMFS requested advice from the National Research Council’s Ocean Studies Board on methods for improv­ ing sea­turtle population assessments. See Box S.1 for the committee’s full charge. This report is intended to help NMFS and USWFS to improve population assessments of sea turtles. NMFS is responsible for the man ­ agement of sea turtles in the water, and USFWS is responsible for sea turtles on land. The shared responsibility means that cooperation between the agencies in the management of sea­turtle populations is critical. The two agencies have a history of cooperation, as in the codevelopment of recovery plans mandated by the Endangered Species Act (e.g., National Marine Fisheries Service and U.S. Fish and Wildlife Service, 2008). The committee was asked to evaluate current and emerging popula­ tion assessment techniques being applied to provide advice to managers of sea turtles in the United States. Methods for conducting population assessments vary widely from simple regression­based approaches to the use of nesting­beach trend data to more mechanistic population­dynamics models. The choice of appropriate assessment approaches depends heavily on the management question being addressed, and the ability to answer a question is often limited by the available data. This report describes a variety of assessment types and techniques, including beach sampling, in­water surveys, genetic analyses, demo­ graphic analyses, use of bycatch (incidental take) information, and aerial surveys; reviews assessment methods; identifies information gaps; and suggests improvements for data collection. Its review of the methods used in assessments (see Table 1.2) has led the committee to conclude that most of the modeling and analysis that have been done is a valiant effort to compensate for a debilitating lack of data. Assessment methods that can be successful in fishery biology are less successful for turtles because the data generally are not as complete as they are for many commercial fish species. In addition, fishery models are focused on one main source of fish mortality—fishing—which has not been quantified for sea turtles and is only one of the anthropogenic sources of their mortality. Filling the large gaps in the available data has far greater promise for improving sea­turtle assessments than refinement of analytical methods (Heppell et al., 2003; Turtle Expert Working Group, 2000, 2007). The com­ mittee therefore decided that its most effective approach was to focus on the data problem, and it concluded that the agencies need to do so as well. Developing a rigorous process for assessment of sea­turtle populations also has high priority. Once better data that can be evaluated in a transpar­ ent framework of scientific review are available, it will become profitable to focus more on refinement of analytical techniques.

OCR for page 11
 INTRODUCTION This report does not revisit the 1990 National Research Council report Decline of the Sea Turtles: Causes and Prevention or any other report on the current status of sea­turtle populations or causes of sea­turtle declines. The committee felt that it was beyond its charge to discuss major stresses on sea­turtle populations, such as interactions with fisheries, and the potential effects of environmental conditions or external stresses; to detail environmental conditions or regime changes; and to assess the costs of its recommendations. Additionally, this report does not review specific assessments comprehensively, except as illustrative examples of methods and data gaps but does provide a summary of methods used. The com­ mittee was not asked to conduct its own assessments of sea­turtle popula­ tions. As a result, this report does not provide information on the status of sea­turtle populations. The committee recognizes the importance of taking an ecosystem approach to managing sea­turtle populations, but its report focuses on population assessments of a single species. Before agencies can undertake ecosystem­based approaches to assessments of sea­turtle populations, substantial information at the single­population or single­species level will be needed, as described in this report. Because the report was prepared in response to a request from NMFS, it is directed primarily at the biologists and managers in that agency. However, the committee expects it to be useful for biologists and man­ agers in other government agencies that have responsibilities for sea turtles and for academic and other researchers. The report also focuses on questions asked frequently of managers, on the current and emerging analyses that can be applied to address the questions, and on the sorts of data that are required to fuel these analyses. REPORT ORGANIZATION Chapter 2 describes the units of assessment. Typically, assessments do not cover an entire species but instead focus on populations (or stocks) or even smaller units delineated by geographic distribution or genetic information. The chapter describes the array of techniques available and in need of development for those assessments. Chapter 3 provides a conceptual model of sea­turtle life history that provides an intellectual framework for understanding survey needs and developing assessment methods. Chapter 4 focuses on methods of estimating abundance and trends in abundance and is centered on land­based and ocean­based methods. Chapter 5 discusses demographic parameters of sea turtles and what is known about them and methods and research needs. Chapter 6 discusses the importance of and methods for integrating demographic information with abundance estimates. Chapter 7 addresses a variety of

OCR for page 11
 ASSESSMENT OF SEA-TURTLE STATUS AND TRENDS issues that cut across many aspects of population assessments, includ­ ing data management, education and training, the permit process, and opportunities for coordination at various levels. Chapter 8 provides the committee’s major conclusions and recommendations.