DNA Analysis Provides Dietary Clues

Scientists are exploring a number of theories explaining the mysterious decline of Steller sea lion populations in Western Alaska. One possibility suggests that a long-term change in the distribution of fish stocks might be depriving sea lions of high-quality prey such as Pacific herring.

Testing this theory requires some gastric sleuthing, and the easiest way to determine a sea lion’s most recent meal is to examine its feces. By analyzing undigested bones and other “hard parts” in fecal samples, scientists can identify the types of prey that have been consumed, indicating possible dietary trends.

This method, while useful, may not accurately show how much of a particular prey has been eaten relative to other types of prey. Some scientists have suggested that analyzing the mitochondrial DNA (mtDNA) of soft tissue remains in feces may provide a more complete dietary picture.

A new study by Drs. Bruce Deagle (University of Tasmania) and Dominic Tollit (University of British Columbia) assesses whether a quantitative real-time PCR approach could be useful on a broad scale. The study was recently published in the journal Conservation Genetics.

Differential Digestion
Deagle and Tollit assessed the use of quantitative real-time PCR by testing levels of mtDNA in a ground up sample of undigested fish tissue.  They also tested the levels of mtDNA present in actual fecal samples (collected from two captive Steller sea lions fed known amounts of different fish at the Vancouver Aquarium). The researchers found that the mtDNA in the undigested tissue sample corresponded reasonably well to the actual amounts of fish fed to the sea lions (50% Pacific herring, 36% surf smelt and 14% sockeye salmon by mass — see Figure 1a). However, Real-time PCR results based on the fecal samples did a poorer job of replicating the sea lions’ actual diet (Figure 1b), suggesting the flesh of certain prey types are digested at different rates than others.

Figure 1:  Box plots showing a summary of the percentage DNA composition data (median, range and upper/lower quartiles) based on quantitative real-time PCR estimates: (A) results from tissue mixture (n=10); (B) results from faecal samples (n=23). Closed triangles show percentage of each fish prey species (by mass) in the diet. Data for two independent replicate runs and the mean values are shown.

The authors report that the DNA analysis overestimated the proportion of salmon and underestimated the proportion of smelt mtDNA in the fecal samples, and highlighted that measuring prey DNA recovery rates in captive feeding trials may be useful in reducing this bias. Despite this bias in DNA recovery, Deagle and Tollit feel that quantifying DNA amounts in fecal samples can still provide some informative data.

Future for Feces?
The authors note that their quantitative DNA-based fecal analysis method could also be used to examine the diet of species whose feces do not contain hard parts (such as seabirds or cetaceans) and could be modified to quantify prey DNA in stomachs of invertebrates where few conventional approaches to studying diet have been explored.

DNA analysis remains costly and technically challenging, and the authors acknowledge that its use may likely be restricted to situations where traditional methods of diet analysis have proven inadequate. The approach also requires a basic knowledge of the diet under study, and particularly the prey of interest. Finally, they note that a large number of fecal samples would help to produce the most accurate results.

Based on the results of this feasibility study, the authors are optimistic about the utility of this DNA technique in future studies.  They note that the use of genetic analysis of feces for studying predator diet is likely to increase because of its ability to improve taxonomic identification of prey and address questions beyond the scope of traditional methods of diet analysis.

24 August 2007

Publication:
Quantitative analysis of prey DNA in pinniped faeces: potential to estimate diet composition? Deagle, B.E. and D.J. Tollit. 2007. Conservation Genetics 8:743-747. abstract Recent studies have shown prey DNA can be consistently recovered from faeces and effectively used to provide dietary information. We investigate the possibility of using the relative amounts of DNA recovered from different prey in faeces to obtain quantitative diet composition data. Faecal samples were obtained from captive Steller seas lions (Eumetopias jubatus) being fed a fish diet consisting of 50% Pacific herring (Clupea pallasii), 36% surf smelt (Hypomesus pretiosus) and 14% sockeye salmon (Oncorhynchus nerka) by mass. Quantitative real-time PCR was used to measure the amount of mtDNA from the three fish species in: (i) a blended tissue mix representative of the sea lion diet and (ii) the sea lion faecal samples. The percent composition of fish mtDNA extracted from the undigested tissue samples corresponded reasonably well to the mass of fish in the mixture. In the faecal samples (n = 23) the absolute amount of fish mtDNA recovered varied 100-fold, but the percent composition of the three fish was relatively consistent (57.5 ± 9.3% for herring, 19.3 ± 6.6% for smelt and 23.2 ± 12.2% for salmon). Differences between the mtDNA proportions in the tissue samples compared to the faecal samples indicate there are prey-specific biases in DNA survival during digestion. These biases may be less than those commonly observed in the conventional analysis of prey hard remains. Further investigation of this approach is warranted.