Opportunities and Challenges in the Genetic Evaluation of Fiber Characteristics in U.S. Alpaca


The availability of Alpaca EPDs for U.S. producers provides exciting opportunities to enhance rates of genetic improvement in fleece and fiber characteristics. EPD-based genetic evaluations of economically important measures of animal performance are the global standard for all important domestic animal species and provides a mechanism to summarize large amounts of data on many animals in a way that provides unbiased predictions of future progeny performance.

Principles of EPD-based Genetic Evaluations

The principles of EPD-based systems are relatively simple in that they all:

  1. Utilize objective measures of animal performance, such as fleece weight and fiber diameter;
  2. Recognize that genetic merit, or breeding value, can only be definitively assessed by measuring and summarizing the performance of an animal’s offspring. The best example of this is in the dairy cattle industry where individual bulls, used in artificial insemination, can have literally thousands of offspring on several continents. The genetic merit of these bulls, derived from the milk production of those numerous daughters, is thus predicted with near-perfect accuracy and can be used to develop specific and individualized breeding objectives for individual breeders. While we will not generally have more than a few tens of offspring for alpaca males, we can compensate for this by the high heritabilities of fleece and fiber traits compared to milk production (see below);
  3. Correct the reported measures for obvious non-genetic influences such as the age and sex of the animal, the age at shearing (for yearlings), and the interval between shearings (for adults) using widely accepted adjustment procedures;
  4. Also corrects for differences in the production environment by first normalizing records to the average of the “contemporary group” or “cohort.” Thus in years when feed is plentiful and fiber diameters are perhaps increased as a result, animal records are first deviated from the mean of that year before they enter the EPDs system. The same holds for records from years when grass is short and fiber diameters are perhaps lower as a result. The actual records that enter the EPD system are thus not measurements per se, but measurements deviated from the contemporary group or cohort average. The underlying assumption is that animals with fine fleeces compared to their herd-mates in good years will also be the animals with fine fleeces compared to their herd-mates in poorer years. Extensive studies with other species confirm that in most cases this is an acceptable assumption. (Note that definition of the “cohort”, though not very sexy, is a critical part of an successful EPD system, and will merit additional attention in future articles.);
  5. Account for the different heritabilities of the individual traits to recognize difference in the predictive value of different measurements. Thus fiber diameter has a relatively high heritability (approximately 50%), and a single measure of fiber diameter for an individual is a reasonably good predictor of genetic merit. In contrast, fleece weight has a somewhat lower heritability (approximately 40%) and a single fleece weight is correspondingly a somewhat poorer indicator of genetic merit. And finally, reproductive and fitness traits have very low heritabilities (generally 10% or less) and the outcome of a single mating (pregnant versus not) is a very poor (though not worthless!) predictor of the genetic merit of the hembra (female alpaca) for fertility and reproductive competence (which explains why these traits will not be included in the initial ARI alpaca EPD program). For comparison, the heritability of milk production in dairy cattle is only about 20%, yet with large numbers of progeny, dairy bulls can achieve highly accurate predictions of breeding values;
  6. Utilize the principle that a measurement taken on an individual provides information about the genetic merit of all that animal's records. Thus if I find that an individual yearling alpaca has a very fine fleece, I can infer not only that that individual is somewhat superior to the average of the population in genetic merit, but also that his relatives (who share some of the same genes) deserve some credit as also potentially genetically superior. The beauty of EPD-based systems is that they combine measurements from ALL relatives to provide a single estimate of genetic merit. Each individual measurement is weighted by the extent of the genetic relationship to the animal in question, so that individual and progeny performance are most influential, and records of distant relatives (for example, a grandparent) receive progressively less weight in deriving the EPDs. Thus the EPD system is relatively simple in that it utilizes all available records on related animals and weights each of those records appropriately based on the closeness of the genetic relationship;
  7. Are continually updated and become progressively more accurate as new information is recorded.  Thus newborn crias get EPDs based on fleece records from their parents, previously recorded full- and half-siblings, cousins, etc. When a cria’s first fleece is evaluated, that information is added to the system (along with additional new fleece records from other relatives) and the EPD is updated with the new information. Future fleeces result in additional adjustments to, and greater accuracy of, the EPD. If the cria produces progeny, the EPD is further updated as those records enter the system, and if the cria becomes a sire, large numbers of progeny records are expected to result in a highly accuracy and increasingly stable EPD. So EPDs DO CHANGE, as more information accumulates (just as our subjective, visual assessment of an animal evolves as the animal matures). Changes in EPDs as more information accumulates are sometimes frustrating and sometimes gratifying, but in either case, should provide increasingly accurate assessments of genetic merit.

From the above points, the picture emerges that EPDs are predictions derived from available objective data on animals and their relatives. They are contingent on the data available in the system and on the specific animals reporting data. They will change as more information becomes available, becoming increasingly accurate and stable.

Challenges in the Development of EPD-based Systems of Genetic Evaluation

The primary challenges in developing successful EPD-based systems of genetic evaluation are operational rather than theoretical. The methods of deriving EPDs from fleece and fiber records are relatively straightforward and generally quite uniform among different countries or genetic evaluation centers.

Attention in program development thus must center on the critical issue of data acquisition and management in order to capture large volumes of accurate information. The U.S. alpaca industry is fortunate in that the histograms and other measures of fleece quality that are regularly obtained on both young stock and adult breeding animals are derived at a small number of highly respected laboratories using established and carefully validated methods. The resulting large volumes of data are literally a treasure-trove of genetic information.

A number of more mundane issues require attention. For example, methods of reporting fleece weight need to be standardized (whole fleece vs. blanket), contemporary or cohort groups must be clearly defined, adjustment factors for things like animal age and sex must be derived for each measurement, heritabilities and genetic correlations must be estimated, etc. These will have more to do with the success of the program than the specific software used to derive the EPDs. Simple and straightforward methods for submitting data (for example, directly from the fiber laboratories to the EPD system) will encourage participation. Careful data checking and data management protocols are essential.

It is also particularly important that we focus on the positive aspects of EPDs and not become paralyzed or overly negative about potential limitations of the system. We can raise concerns over many of the details of any genetic evaluation system. For example:

  1. Are differences among animals in fiber diameter in high-feed versus low-feed years, or in the Pacific Northwest versus the Southeast really consistent?
  2. Is “micron blowout” (the tendency for some animals to increase fiber diameter as they age while others do not) a heritable trait, and can we select against it?
  3. What is the relationship of body size to fleece characteristics in alpaca? Should we calculate body weight EPDs? If we do, how should we use them?
  4. Do dark-colored animals necessarily have coarser fleeces than their lighter-colored relatives? Are there any limitations to developing colored animals with extremely fine fleeces? How do we consider fleece color in genetic evaluation?

We don’t have much (if any!) data on these questions for alpacas. But I would argue from results in other species that these concerns are unlikely to invalidate the value of EPD as tools for genetic improvement. And, perhaps more importantly, establishment of an EPD system is probably the best possible strategy for developing answers to these questions.

Over the past 10 years, I have been actively involved in the development of EPD systems for the U.S. sheep industry. In that role, I have strongly promoted the concept that development of a successful EPD system is an ongoing collaboration among breeders, breed associations, educators, and others. We do not know, today, everything that we need to know about the genetic evaluation of alpacas. But answers to many of our questions will emerge from the data that we collect and accumulate in useful and readily accessible data sets. Experience with other species has shown that EPDs are a powerful positive force for genetic improvement. But equally important has been the potential of the EPD datasets to provide information for development of new and better EPDs and to allow for updating of protocols based on that new information. We cannot answer all the questions we might have because we do not have the data required in a readily accessible form. But we CAN begin to build such a data set.

In summary, the future for genetic evaluation of alpacas in the U.S. is incredibly bright. The traits of interest are highly heritable, so accuracies of evaluation will be high. Because of the relatively recent introductions of foundation animals from South America, genetic relationships exist among animal across many herds, making separation of true genetic differences from contemporary or cohort effects relatively easy. Sires are, in many cases, used widely across herds, also enhancing genetic relationships among the herds and allowing progeny to be evaluated in many production environments. Data are, in most cases, coming from highly reputable laboratories and can be imported directly into the EPD databases. These characteristics all contribute positively to the development of an EPD system for alpacas in the U.S.