Measuring Stress Hormones in Wild Animals; Different Techniques and Their Comparative Advantages and Disadvantages

Measuring circulating levels of stress hormones (cortisol, corticosterone, etc.) in wild animals can provide valuable information for many types of studies. For example, conservation biologists may want to determine if high levels of snowmobile traffic are associated with increased cortisol levels in elk. Behavioral ecologists may want to know if white-faced capuchins (a type of monkey from Central/South America) have lower levels of cortisol if they come from a troop with a high level of social support. And physiological ecologists may be interested in whether Galapagos marine iguanas have a corticosterone rhythm driven by photoperiod or tidal cycle (*see bottom of page for more details).

In order to measure levels of corticosterone or cortisol (CORT), researchers need to obtain blood, feces, urine, saliva, feathers, or hair from their animals. Here’s some information about each method and their advantages and disadvantages:

Blood

This is the most common way to measure CORT concentrations. Because hormones circulate through the blood, this method provides researchers with the most functional measure of CORT. Since CORT has a very high concentration in the blood (compared to other hormones), blood samples do not need to be very big (for degus, 30uL of whole blood is usually enough to determine baseline CORT). The downside to using blood samples is that within 3 minutes of encountering a stressor, an animal’s CORT levels start to increase. So, if a researcher wants to determine baseline levels of CORT, they must get a blood sample within 3 minutes of capture.

Another downside of using blood samples is that each sample is only a snapshot of an animal’s stress response. In order to get a more integrated picture of an animal’s stress profile, researchers usually take a series of blood samples over the course of 1-2 hours. That way, they can see the rise from baseline, the peak of the CORT response, and then the gradual decrease caused by negative feedback.

After obtaining a blood sample, the samples must be kept chilled until they can be spun in a centrifuge (you want to spin within 24 hours of collection). Once the samples are spun, the plasma or serum (the clear layer on the top) needs to be drawn off and saved for analysis. The plasma or serum must be frozen at -20˚C and can be stored for several months. These requirements are fairly reasonable, but sometimes field sites are far away from cities. Last year I worked in a national park that was pretty remote, so I had to manually spin my samples with a hand-crank centrifuge and store my samples in a cooler full of ice. Luckily, steroid hormones like CORT are fairly sturdy, so lengthy processing times aren’t a big deal.

Finally, to process the plasma samples, researchers need to run a competitive binding assay called a radioimmunoassay (RIA). Basically, known amounts of radiolabelled CORT and CORT-antibody are added to the sample. The radiolabelled CORT and normal CORT from the sample will compete to bind with the antibody. The unbound radiolabelled CORT is then washed off, and the radioactivity is measured. The more CORT you have in your sample, the more it will bind with the antibody and displace the radiolabelled CORT. Therefore, the higher the CORT in the sample, the lower the radioactivity. RIAs are fairly straightforward but require some special equipment, so they usually have to be run in laboratories that frequently measure hormone concentrations (like my lab!). The enzyme immunoassay (EIA) is another way to measure hormone concentrations and requires less equipment, but is generally more expensive to run than a RIA.

As an additional step, researchers may also want to measure levels of corticosteroid binding globulins (CBG). CORT binds to CBG in the bloodstream, and it’s thought that only unbound CORT is “free” and able to bind to cell glucocorticoid receptors. Therefore, the functional measure of CORT is the “free” CORT, so some researchers also measure CBG levels to determine the relative amounts of unbound CORT. However, it’s still under debate as to whether only unbound CORT can bind to glucocorticoid receptors, and it has also been pointed out that without CBG, CORT would be unable to circulate through the whole bloodstream and reach all the cells in the body.

My blood samples are collected in little, glass microhematocrit tubes. I put them in plastic tubes with clay in the bottom so the bottoms of the microhematocrit tubes are plugged. I then put three plastic tubes in a conical tube which are stored in a cooler full of cold packs.

Loading the centrifuge with my samples.

After spinning the blood sample, it separates into hematocrit (the bottom red stuff) and plasma (the top, clear stuff). The plasma is what has the hormone, and that's what I draw off with a glass Hamilton syringe (shown on the table).

I pipette the plasma into a plastic eppendorf tube which then goes into a -20C freezer.

Saliva

            CORT can pass from the blood to the saliva, and salivary samples are actually a common way to measure cortisol levels in humans. While captive animals can be trained to lick, chew, or drool on something, obtaining sufficient salivary samples from wild animals in non-invasive manner is most likely impossible. However, it usually takes 20 minutes after a stressor for CORT levels to increase in the salvia, so taking salivary samples could give researchers more time to collect a sample after capture.

Feces/Urine

Another way to assess CORT levels is to measure the metabolic products of CORT in the feces. However, because this method measures the metabolites of CORT, it doesn’t have quite the same functional value that a blood sample does. There are several steps involved in breakdown of CORT, and CORT metabolite formation can be affected by sex, season, metabolic rate, and diet. Therefore, experimental power is lost when using fecal samples for CORT analysis because animals cannot be compared between different seasons, locations, or sexes.

The main advantage of using fecal samples, however, is that they’re fairly non-invasive. Researchers can follow individuals and collect feces, or in the case of small mammals, check traps after a certain period of time and collect any feces from the bottom of the cage. Using feces to measure CORT levels is also great for repeated sampling, since frequent blood sampling and other more, invasive measures can change an animal’s stress response.

Using fecal samples to determine stress hormone levels makes a lot of sense if you’re working with an animal that’s endangered or difficult to capture (like an arboreal monkey). Another advantage of using feces is that they represent an integrated measure of CORT exposure; if you know how long it takes for an animal to form a fecal pellet, then you can estimate their total CORT exposure over a period of several hours.

Collecting feces is usually pretty easy, but investigators should always try to obtain fresh samples from a known individual. If this isn’t possible because the study animal is hard to find or lives in an aquatic environment, then dog trackers can be used. The Wasser Lab at the University of Washington (my alma mater!) uses canine trackers to find feces from whales, tigers, wolves, and other, elusive endangered animals (http://conservationbiology.net/conservation-canines/#scat).

 Storing fecal samples can be tricky; temperature, storage liquid, autoclaving, and storage time can all affect fecal glucocorticoid metabolite (FGM) concentrations. Sample mass can also affect FGM concentrations, as very small samples have disproportionally higher FGM levels. Getting an adequate sample mass can be difficult when studying small animals like songbirds, mice, etc., so many studies have to combine several fecal samples in their assays.

After homogenizing the fecal samples, FGM concentrations can be measured via RIA or EIA. Another downside of using fecal samples is that the CORT-antibody may not bind with all the different CORT metabolites. Therefore, researchers usually need to do a validation experiment to determine that increased CORT levels in the blood correspond to increased FGM concentrations in the feces. One advantage of using feces, though, is that by measuring levels of metabolites the researcher is essentially measuring “free” CORT, so there’s no need to measure levels of CBG.

Like feces, urine also contains glucocorticoid metabolites (and some unmetabolized CORT). In the field, however, it’s very difficult to collect urine samples in a non-invasive manner, so urine collection for CORT analysis is rarely used outside of the laboratory.

Feathers/Hair

             Believe it or not, you can actually measure CORT in feathers. During molt (feather growth), each individual feather is vascularized, and CORT from the blood can be deposited in the feather. After the feather stops growing, the blood supply is cut of and, supposedly, no further CORT can be deposited in the feather. Therefore, feathers provide a good, integrative measure of CORT during the period of molt.

            Plucking a feather is easy to do and there’s no time limit, unlike blood sampling. Using feathers can also be non-invasive if the researcher obtains naturally molted feathers. And another big advantage of using feathers to determine CORT levels is that feathers from dead birds can also be assayed (like museum specimens!). Also, feathers do not need to frozen or stored in any specific manner.

            The big downside to using feathers for CORT analysis is that we’re still not too sure what form of CORT we’re actually measuring. Different antibodies have been found to have different levels of success during feather CORT analysis, which has made researchers unsure whether they’re measuring CORT, CORT metabolites, or other molecules similar in structure to CORT.

Another downside to using feathers is that, like feces, there’s a sample mass bias where smaller feather samples have disproportionally higher CORT. Getting a large enough sample can be difficult because researchers don’t want to compromise a bird’s flying ability, and there’s also the problem that different feathers are grown at different times, so pooling certain feathers together may not be appropriate. Feather mass requirements can also prevent researchers from examining portions of the feather (sections near the end of the feather are older, and thus represent a different time period than sections near the bottom), but a recent study found that different feather sections didn’t correspond to the CORT concentrations in the blood at the time of their growth, anyway. Feather CORT can be measured via RIA, but the preparation is a real pain (you have to mince the feather into little, tiny bits).

            CORT can also be deposited in hair during follicle growth. One advantage of hair samples is that they can represent CORT exposure over a very long period of time. The downside to hair samples is that it’s hard to collect a large enough sample in a non-invasive manner. And, like feathers, there are still a lot of questions about what the antibody is actually binding to in the assay. Another disadvantage to using hair is that the sample could be contaminated with other, CORT-containing secretions, like saliva or sweat.

            So, I’ve tried to give a basic, if somewhat extensive overview of the different ways to measure CORT in wild animals. After I was almost done writing this blog, I came across a great review paper that includes everything I mentioned and more: Sheriff, M.J. et al. (2011) Measuring stress in wildlife: techniques for quantifying glucocorticoids. Oecologia 166:869–887. Also, here’s an excellent review on fecal CORT: Goyman, W. (2012) On the use of non-invasive hormone research in uncontrolled, natural environments: the problem with sex, diet, metabolic rate and the individual. Methods in Ecology and Evolution 3, 757–765. And finally, my favorite feather CORT paper: Lattin, C.R. et al. (2011) Elevated corticosterone in feathers correlates with corticosterone-induced decreased feather quality: a validation study. Journal of Avian Biology 42, 247-252.

*(Photoperiod vs. tidal rhythms: most animals have daily stress hormone rhythms that are related to foraging opportunities. For humans, the morning is an optimal time to forage, so our cortisol levels often peak right before we wake up, and these increased cortisol concentrations help us go out, run around, and collect some food. For marine iguanas, corticosterone concentrations are influenced by time of day AND tidal cycles. This is because marine iguanas need to eat algae in the inter-tidal zone and the algae are most available during low tide, but because iguanas are ectotherms and need to bask in the sun before and after swimming in the inter-tidal zone, foraging is only possible during the day. However, this relationship still needs be teased apart, since the observed CORT peak during low tides could also be influenced by food intake.)