Genetic Monitoring for Managers


Type of GEM

Category 1a Applications

Abundance (population size):

The most common metric used for monitoring animal populations is abundance. With rare species, abundance indices, such as the number of animals detected, have proven relatively inexpensive to obtain through non-invasive genetic sampling. DNA-based abundance indices are often adjusted to remove biases associated with limited sample size or imperfect detection rates. These adjustments can be accomplished by applying asymptotic analyses or using models that account for covariates associated with capture, such as season, time of day, or habitat (White 2005, Frantz and Roper 2006).

Abundance can also be estimated through capture-mark-recapture (CMR) analyses. Traditionally, this has required physical capture and marking of a sample, followed by release and subsequent recapture of a proportion of marked animals (reviewed in Luckacs and Burham 2005). Because each individual has a unique, naturally occurring genetic code (typically revealed with microsatellites), non-invasive genetic sampling of hair, scat, feathers or other materials can be used to identify individuals, obviating the need to capture and mark individuals physically (Avise 2004). Consequently, non-invasive CMR analysis is usually used on species that are rare, difficult, dangerous, or expensive to capture (Proctor et al. 2005, Boulanger et al. 2004). For example, in Australia, microsatellite DNA and CMR analysis has provided a feasible and cost-effective strategy for monitoring long-term changes in brush-tailed rock-wallaby Petrogale penicillata abundance (Piggott et al. 2006).

As genetic monitoring is a relatively recent phenomenon, most studies using molecular tags to generate trends in abundance have asked whether the population has changed in size over only one or two time intervals. However, long-term data sets with multiple abundance estimates are generally required to distinguish meaningful changes in abundance from sampling error and environmental variability [64,70]. To learn how changes in abundance can be inferred through changes in genetic diversity see Category II: genetic variation.


Vital rates:

Vital rates, such as survival and recruitment, drive population change over time and are of fundamental importance to population ecologists. Traditionally, survival has been difficult to estimate, much less monitor, for rare and elusive species. The ability to use molecular data to track known individuals can either replace or augment traditional efforts to monitor survival (Prugh et al. 2005). For instance, naturally shed feathers of Eastern imperial eagles (Aquila heliaca) have been used in a microsatellite analysis to identify individuals, monitor population turnover and estimate annual survivorship over a four-year period (Rudnick et al. 2005) (learn more on this in "Examples"). The results showed that turnover rates varied annually and that cumulative survival was lower than expected; such information can not be obtained easily using traditional approaches.


Fate of reintroduced animals:

Although our focus is on monitoring wild populations, artificial propagation is being used increasingly as a conservation and management tool. Many captive programs release individuals into the wild, and informed management requires monitoring of the genetic consequences of these releases on natural populations. We expect that genetic monitoring of vital rates will become prevalent in reintroduction efforts, where post-release genetic monitoring is badly needed, but currently underused. This approach will be most compelling when genotypes of reintroduced animals can be determined before release (Flagstad et al. 2004) and the fate of each individual tracked by the presence or absence of its genotype in subsequent, non-invasive samples. These data can also be used for parentage analysis to monitor reproductive output.

Molecular markers can also be used in parentage analysis to assess reproductive success of captive individuals in the wild. Araki et al. (2007) monitored a supplementation program for steelhead (anadromous Oncorhynchus mykiss) in Oregon and found that the relative reproductive success of hatchery fish improved when the program phased out the non-native hatchery stock and began using local native fish for brood stock. Monitoring the genetic contribution of captive individuals to natural populations is necessary before one can assess how the captive program affects the Ne of the combined captive-wild system (Ryman et al. 1995, Hedrick et al. 2000). Genetic monitoring of captive propagation is common in fishes, especially salmonids (Allendorf and Ryman 1987), but is increasingly being applied to other taxa. Recent studies include evaluations of the effects of oysters transplanted from Louisiana on local Chesapeake Bay populations; spread of ectomycorrhizal fungi used to enhance forest tree growth after out-planting; diversity in translocated populations of wallabies in Australia; and the success of protocols to introduce an endangered annual plant to restored vernal pools in southern California (Selosse et al. 1998, Milbury et al. 2004, Sigg 2006, Ramp et al. 2006).

Category 1b Applications


Regular monitoring with molecular markers can provide early detection of hybridization, thus giving managers a wider range of options than is available if introgression is already extensive before detection. In recognition of this reality, genetic monitoring of hybridization has been proposed for many species, including Canada lynx-bobcat (Lynx rufus) (Schwartz et al 2004), coyote (Canis latrans)-red wolf (Canis rufus) (Miller et al 2003), and spotted owl (Strix occidentalis)-barred owl (Strix varia) hybridizations (Haig et al 2004). Genetic monitoring programs have already been initiated for several species; for example, native westslope cutthroat trout (Oncorhynchus clarki lewisi) in the Rocky Mountains are threatened with genomic extinction owing to widespread hybridization with introduced rainbow trout (O. mykiss) (Hitt et al 2003, Allendorf et al 2004). Diagnostic molecular assays applied to trout from the Flathead River system of Northwest Montana have demonstrated an increase in the proportion of streams with westslope cutthroat-rainbow trout hybrids between 1984 and 2001 (Hitt et al 2003).


Site occupancy and geographical range:

Changes in the proportion and distribution of sites where a species is observed is commonly monitored to elucidate effects of environmental change (Bailey et al 2004, Manley et a l2004). These occupancy statistics can be hindered by low probability of detection owing to the elusive nature of some species, and by false positives that occur when visual identification is cryptic or a species leaves signs that are easily confused with signs of other species. Because species identification through DNA is reliable and assays are inexpensive, we expect that the use of non-invasive collection devices will soon form the basis of many large-scale surveys monitoring changes in occupancy and geographical range. Several large-scale, DNA-based surveys have already been established to monitor the distribution of species over time (McKelvey et al 2006, Fernandez et al 2006). For example, the presence of Canada lynx (Lynx canadensis) has traditionally been monitored using unreliable metrics, such as pelt returns, snowtracks (which have high false positive rate) and incidental sightings. In 1998, a five-year monitoring effort was initiated using non-invasive genetic sampling for lynx presence on ~44 million ha of US national forest land (McKelvey et al 2006, Mills2002). Sampling was conducted using bait stations that attracted lynx to the site and elicited a rubbing response on a hair collection device. Lynx were detected in multiple locations from the hair samples using mtDNA. The success of this work provides a baseline to monitor long-term changes in the geographical range of the lynx and can already be compared with historical records (McKelvey et al 2006).


Presence, prevalence, transmission of pathogens and parasites:

Disease is one of the greatest threats to population persistence. Molecular genetic markers, including DNA barcoding markers (Herbert et al. 2004), have enormous potential to help monitor the presence, prevalence and transmission of pathogenic organisms. Whole genomes and gene sequences are available for increasing numbers of pathogens, making marker development relatively easy. For example, brucellosis and TB, both caused by bacterial pathogens, are among the most problematic diseases in ungulates and some carnivores (Joly and Messier 2005, Briones et al 2000); they cost millions of dollars through vaccinations, 'test-and-slaughter programs' and reduced numbers of economically valuable wildlife and livestock. Genomes were recently sequenced for both pathogens (Brucella abortus and Mycobacterium tuberculosis, respectively), facilitating development of molecular markers to diagnose the presence of different strains and to track their spread (Bricker et al 2005). Similarly, genetic monitoring provides new insights into West Nile virus (WNV) epidemics in humans. Over a five-month period, molecular markers were used to detect the presence of WNV-carrying Culex pipiens mosquitoes in the Washington DC and Baltimore, MD area of the USA; whether the mosquito was engorged with avian or mammalian blood; and the presence of WNV RNA infection in the mosquito. Temporal patterns in the molecular results, combined with temporal abundance data for mosquitoes and birds, suggests that the increase in human WNV infections during late summer is due to mosquitoes shifting their diet from their preferred avian host to humans (Kilpatrick et al 2006).