The overarching goal of the research strategy described in this report is to examine the evolutionary relationships among ancient red wolves, the extant red wolf populations (captive and managed), and the Gulf Coast canid (GCC) populations. Understanding these relationships is central to determining the taxonomic status of the red wolf and other canids occurring within the historical range of the red wolf. Attaining this goal will require a better understanding of the genetic ancestry of all relevant canids in the extant red wolf and GCC populations and of the phylogenetic relationships among these canids and the other North American canids to which they are related. The data needed include genomic data, morphological data, and behavioral data, along with informative ecological data. All of these data require careful interpretation in light of the roles of hybridization and admixture in the evolutionary history of these animals, particularly with respect to how the canids of eastern North America originated as individual species.
The five questions described in Chapter 1 (see Figures 1-4, 5-1, and 5-2), taken together, address the problems outlined in the statement of task. Each question addresses the relationships between particular pairs of canids (among three canid groups for Question 1) within a specific period (Figure 5-1). The answers to the first four questions require integrating new genomic data with new morphological data (Figure 5-2). The answer to the fifth question requires a further analyses of some of the data collected to answer the other questions. The composite results will describe the relationships among the key canid populations with respect to historical continuity and contemporary relatedness.
This strategy has three important features. First, different questions for canids in different periods require different sampling schemes (Figure 5-1). The answer to Question 1, for example, requires data from ancient specimens of three canids, red wolves, gray wolves, and coyotes. These data should come from locations where the three groups were sympatric (overlapping in their ranges) or abutting each other’s range in that period and from locations where coyotes and gray wolves were found without red wolves. The same geographic sampling scheme is necessary to
answer Question 2, but these data must come from historical specimens. Questions 3–5 require data from modern specimens collected across their present ranges. A sampling scheme that is haphazard in choices of period and location or that attempts to use all specimens in a single analysis will not provide compelling answers to these distinct questions.
Second, different methods of data analysis are required for different questions (Figure 5-1). Some methods, such as the ABBA–BABA test, are appropriate for answering more than one question, but no single method is optimal for answering all five questions. Applying all conceivable analyses to every question is likely to increase confusion and obfuscate the patterns in the data. Therefore, care should be taken to identify the analytic approach best suited for each question. The methods described in the report are among those showing promise for resolving the research questions. Other methods may also be suitable; the important point is to deploy the best method for the specific question and type of data brought to bear on the answer.
Third, although genomic data are necessary for answering Questions 1–4, morphological data add relevant information (Figure 5-2). In some cases, morphological data can provide an approach to one of the questions that is independent of an approach based on genomic data. In other cases, morphological data may be the only data available. More importantly, the two types of data can complement each other. This is why the strategy is emphasizing the collection of morphological and genomic data from the same specimens. For example, if genomic data carry a strong signature of hybridization, a signature of the disruption of development that often accompanies hybridization may be observable in the morphology. Such a connection could facilitate identifying contemporary hybrid animals in nature. Answering Question 5 requires only genomic data.
Understanding the relationships among extant and extinct canids will require a team of collaborators. The data required for a full understanding of those relationships range from DNA from ancient and historical specimens to morphological analyses to observations of mating behavior. No single individual or even a single laboratory will have all of the expertise necessary. In addition, whenever possible, morphological and genomic data ought to be collected from the same individual specimens, whether museum specimens or live wolves. This will require close cooperation among teams with different areas of expertise, strong overall management of the research, and considerable planning and preparing ahead of time.
The close relationships among different canid populations and the history of admixture indicate that new genomic data are required. Further efforts at acquiring genetic data via ddRad-seq or other genetic methods that sample a subset of the genome will not resolve the historical relationships or offer further insights into the history of admixture. Whole-genome data are essential. In addition, when new genomic data are collected, the DNA sequencing must use the same libraries as were used for existing genomic data to maximize the collection of data from precious samples.
Specimens to be studied should be chosen based on their geographic origin and on the period when collected, not on the particular species designation associated with the specimen. This is particularly important for fragmented museum specimens such as teeth, whose taxonomic identification carries an unavoidable degree of uncertainty. The sampling of ancient and historical specimens of red wolves should give priority to specimens collected from the southeastern United States, which is the center of the original range posited for red wolves (Figure 1-5). For one reason, there is some dispute over the identity of specimens collected from the northern reaches of the red wolf range. For another, one of the limitations of existing genetic data is that most of them are from the western edge of the species range, where historically the range of red wolves abutted that of coyotes. Data from specimens collected in the southeastern United States would provide a clearer characterization of the historical red wolf.
The amount of data necessary for answering each of the research questions laid out in Chapter 1 is discussed in the individual chapters devoted to each type of data. Chapter 2 describes morphological data, and Chapters 3 and 4 describe genomic data. Given the limitations of sampling, monetary constraints, and the desirability of acquiring the most information from the limited samples that are available (Appendixes B, C, and D), it will be necessary to have a careful balancing of research effort between having larger numbers of specimens from single locations and periods and having as extensive a coverage of historical specimens as possible.
All data collected for this research effort should be publicly available to ensure their maximum utility and to engage diverse researchers to become involved in answering the questions. The data that should be made publicly available include the raw data, not simply summary statistics or final products of statistical analysis. Biologists interpreting these data should do so with the knowledge that hybridization and admixture have characterized the history of North American canids. The critical issue is not whether there has been admixture, but in which period, and among which canid groups, and to what extent such admixtures took place.
Beyond the specific recommendations in each chapter about the types of data and analyses best-suited for the questions, the committee developed four overarching recommendations.
Recommendation 5-1: The U.S. Fish and Wildlife Service (USFWS) should adopt the proposed research strategy. The USFWS should encourage researchers to form collaborations that include experts in the extractions and analyses of ancient DNA, the collection and analyses of genomic data, the collection and analyses of morphological data, and the collection and analyses of behavioral and ecological data. The call for such a collaborative effort reflects the committee’s conviction that the integration of multiple types of data is needed to answer the questions in a compelling manner.
Recommendation 5-2: The USFWS should support the research strategy via funding. The funding should be sufficient to allow teams of researchers to perform the sampling necessary to answer each question. While data from some individual samples can contribute to answering more than one question, any attempt to economize by attempting to use one set of samples for all questions is unlikely to resolve any of the individual questions.
Recommendation 5-3: The USFWS should be open-minded about refining the strategy as initial results emerge. Research strategies are templates based on knowledge in hand. As new knowledge is gained, the emphases of a strategy may change, new elements may be appropriate, and old elements may appear unnecessary or at least far less important. The USFWS should make the execution of this strategy an iterative process in which researchers can reconsider priorities as a sufficient amount of new data becomes available.
Recommendation 5-4: The USFWS should work with the research community, including professional societies, to create opportunities for larger community involvement in the synthesis and dissemination of research results. In particular, having a shared space for dialogue among researchers with different interpretations of the existing data may produce improvements in the research strategy or in the plans for its individual components. It can also lead to a robust conclusion about the taxonomic status of the canids identified now as red wolves. Species delimitation remains a difficult and often controversial topic, especially when taxa have recently diverged and there is evidence of past admixture. An open-minded approach to all of the data is the best approach, and open-mindedness is aided when the community of experts are engaged in interpretation, discussion, and debate.
This page intentionally left blank.