“Bob said I should call you because it would be good to do some genetics on this mammal population, how long will that take…?”
“Alice and I collected a few tissue samples when we were in the field earlier this year and now we’d like to talk to you about collaborating on some DNA analysis…”
I’m paraphrasing of course, but I’ve heard variants of these numerous times over the last decade or so. I’ve also met many people with a passion for ecology and conservation who have wanted to learn more about genetics as it applies to their research. Which is great! DNA can be a wonderful tool with great potential to contribute to wildlife conservation and management. As DNA sequencing technologies advance, genetic analysis has become both more affordable and more accessible to people who study non–model organisms (i.e. most of the world’s wildlife!).
However, as with anything, a little knowledge can be a dangerous thing. Now I certainly don’t claim to know everything there is to know about genetics, but I’ve spent enough time in the lab to see some of the same mistakes made over and over. As difficult as it may be to believe, I’ve even made some of them myself! So I thought it might be worth compiling a list of five key things you should think about if you’re planning a wildlife genetics study. Of course most of these are also relevant to any well-organised research project, so in that sense there’s nothing special about genetics. And when it comes down to it, this is just a list of my top pointers: genetics can be complicated and there are many other ways to stuff up your work! But these are all points that I try to impress on my students and if you’re interested in learning more about wildlife genetics then this is not a bad place to start…
1. What is your question?
You wouldn’t start a trapping or spotlighting survey without deciding beforehand on your key research questions and designing your sampling and data collection strategies to address those questions. Genetics is no different, and you need to have a good idea of what you want to learn before you start collecting samples. Otherwise you might collect the wrong samples. For example, if you want to use paternity analysis to study the mating system of a deer population (do all males get to father offspring, or only a few successful males?), you might aim to sample many individuals within that population, including lots of mother-offspring pairs and as many of the adult males as possible. In contrast, if you want to study patterns of individual movements among several deer populations in relation to landscape features (do rivers or highways act as barriers to gene flow?), you might aim to sample a representative number of animals from each population, deliberately avoiding closely related individuals such as mother-offspring pairs because these might skew your results. If you’re not sure if you can use DNA to address your burning question, or how you might go about doing that, it’s best to find a collaborator who has the relevant experience and discuss your ideas before you start fieldwork. Then your sampling efforts won’t be wasted.
2. Are there genetic markers available for the species you’re interested in?
The poor availability of genetic markers for non-model species is becoming less of a problem now that deepsequencing methods are widely available (e.g. see Ekblom and Galindo 2011, Seeb et al 2011, Thomson et al 2010). But these approaches may still be too costly for many researchers to access and some research questions may require different types of marker, which may need to be developed for each taxon in turn. Indeed for many species there are few if any reference DNA sequences available for any part of the genome, let alone species-specific markers. So before you get too excited, find out whether there are suitable resources already available for your species of interest. Of course a lack of genetic markers is not in itself an insurmountable obstacle, but you may need to explore newer methods, such as genotype-by-sequencing and population genomics, or else commit some time and resources to marker development.
3. Collect sample data. Lots of data!
I know, field work is hard work. You’ve walked 35 kilometres in 3 hours up a steep muddy hill carrying twice your body weight in gear whilst being eaten alive by mosquitoes and leeches and now you have finally reached your sample site it is raining and your lunch is soggy. The last thing you want to do is spend an extra 5 minutes per sample crouched under your umbrella filling in additional fields on the data sheet and taking photographs just because some gel jockey (who is probably sipping a latte* in a cosy cafe as you write) asked you to. But… your geneticist colleagues probably asked for those data for a good reason. And surely you want to get as much return as possible for that 35km uphill trudge? Sometimes we don’t truly realise the value of a complete dataset until much much later, when we’re trying some shiny new analysis that the clever theoretical types have come up with. And then it hurts, physically hurts, to have to exclude a hard-won sample because we don’t have a full record of the GPS coordinates, or the GPS datum used, or the sex of the individual, or a photograph to verify the phenotype / species identification that turns out to be really important to interpretation of our DNA sequences. So if you really want to make the most of your research project, grin and bear it and collect the extra data, then get the gel jockeys to buy you a latte when you get home…
4. Who is going to do the lab work?
So you’re an ecologist and you’re a world expert in lizard-wrangling, but you’ve never held a pipette in your life. How are you going to turn your samples into genetic data? You’ve got a trusted collaborator in a genetics lab and they / their technician / their postdoc will do the labwork once you send them the samples? You’ve got lots of money and you can pay for a service provider to do it all for you? That’s great, proceed to point 5! Otherwise… there’s nothing to stop you from learning the necessary skills if you have access to the facilities you need, but you really need to find a mentor who can spend a decent amount of time in the lab with you and help you to troubleshoot when things go wrong. Which they almost certainly will at some point (that’s all part of the fun, honest!). Just don’t underestimate the time you might need and be prepared for plenty of lab frustration along the way. And while we’re at it… the same point holds true for data analysis. Deciding on the most appropriate analyses and software for your project can be a minefield if you don’t understand the assumptions and limitations of the different options. And there are many options out there! Depending on the scale of your project, it will probably be worth talking to a bioinformatician from the outset as well.
5. Contamination, contamination, contamination!
If you talk to a lot of geneticists, especially those who work with ancient DNA, environmental DNA and non-invasive samples (like hairs, shed skins and scats), you might be forgiven for thinking we’re a paranoid bunch, especially when it comes to sample handling and contamination. But we’re like this for good reason – contamination is much easier to prevent than it is to get rid of! And nothing can de-rail a project timeline like persistent contamination. Possible sources of contaminants include unwanted DNA in laboratory reagents, equipment and consumables, cross-contamination among samples during sample collection, and laboratory contamination with DNA or PCR products from previous projects conducted in the same facility. This is especially true if you’re working with ancient samples or using mammalian primers that might also amplify human DNA.
Fortunately, if you followgoodlaboratory and sampling practices, it can be fairly straightforward to minimise the risks of contamination. In the field, it may be important to wear a clean pair of gloves to collect each sample, to only open a single sample tube at a time etc. Once you get your head around it, it becomes second nature. Likewise, if you’re doing lab work, make sure you understand what is expected of you in each laboratory. All of the work I do with degraded environmental samples at the University of Canberra takes place in a dedicated Trace DNA Laboratory – similar facilities exist elsewhere in labs that handle trace or forensic samples, or ancient DNA. There are strict rules about which rooms samples and DNA can be taken into and which order the work can be done in. DNA extraction from samples that are particularly sensitive to contamination is done in a laminar flow hood that can be UV irradiated after each use. Every batch of DNA extractions and PCRs is monitored for contamination. If you’ve been in the post-PCR labs, you have to leave, shower and change into clean clothes before you can re-enter the Trace Lab. It may seem extreme, but it means that we rarely have to discard samples or repeat DNA extractions and PCRs because of contamination. Which means we can analyse more samples before we run out of money!