Much of the work we do as environmental professionals, and efforts in the conservation and protection disciplines, revolve around fish. And as part of that the two most fundamental questions are: (i) what species are present in a waterbody and (ii) how abundant are they? Sounds like it should be straightforward to answer these, right? It is only two questions. To address these, however, requires robust and effective fish sampling programs. Over three decades I have seen a great variety of fish sampling programs, some achieving those metrics of robust and effective; others not meeting that bar. So how do we build a solid fish sampling program? Some suggestions are provided here. These propositions are not derived from academic or theoretical constructs of required sample numbers or optimal sampling theory – these are based on lessons learned participating in both very well-designed programs and others that appeared to have no coherent design to the process.
Know your objective: In order to determine an effective and efficient manner to sample a waterway for fish it is critical to know well the objectives of the work. Is it to establish presence? Determine relative abundance? Absolute abundance? Community structure? Fish health? Something else? Until this purpose is clear in the mind of the person designing the program, effective, efficient, and worthy methods and strategies cannot be selected. The tools, approaches, and skills for an absolute abundance survey are entirely different from those seeking presence or trying to understand the fish community. The project objectives dictate what we need to collect and, therefore, with some reflection, how to collect it.
Appreciate the difficulty of reliably capturing fish: Wild fish are difficult to capture. They are exquisitely attuned to their environment, using senses of pressure changes and sometimes electroreception to which we are naively blind. They have avoided predators for their lifetime while also meeting all of their own needs of feeding, resting, and reproduction. The water is their home; we are a casual visitor. These animals flee and seek shelter away from us the moment we cast a shadow on the water surface, step into the stream, or drop that gear into the bottom of the boat. Begin your project realizing the deck is stacked against you in terms of catching your quarry.
Not only are the fish difficult to physically capture, but our reliance on truck beds of equipment makes it more challenging. Equipment breaks down at the worst time; changing streamflows may carry our gear away; mink and otter target our gear for the free lunches.
Understand the species you are after: In order to capture fish to meet project objectives, the investigator must have a sound knowledge of the biology and ecology of the species of interest. One cannot age sculpin, sturgeon, eel, or lamprey by scales (they don’t have scales); hydroacoustic methods don’t work for some fish, such as eulachon (these fish lack swim bladder). Employing the wrong gear or tactics for your species of concern is a waste of time and money – do the research to ensure that your species is susceptible to your capture and sampling methods.
Allow sufficient time: One of my greatest disappointments in life was realizing the frantic pace at which professional field biology is conducted. We run around, completely out of sync with natural pacing of ecological patterns, and try to force nature to yield secrets on our schedule. She will not do so. Nature is slow. If we are to understand her in a meaningful way, we must spend time with her. This means ensuring our fish sampling has sufficient time to accurately reflect conditions. This may include multiple trips to a site or extended time at one location sampling over time. Time and again I have seen the biased and misleading results produced by short term (1-3 day) sampling periods. They do not capture species that are hard to catch or those that seasonally use the area. Certainly not the species at risk which are, by definition, at low abundance. A well-designed study, one in which some confidence may be placed, covers a reasonable temporal period; we suggest at least three seasons and preferably three seasons in each of two years. That would be data I would trust.
Use multiple methods (Always): Every fish sampling method has biases that prevent us getting an accurate picture of conditions in the water. Minnow traps are size selective, not including small- or large-bodied fish. Electrofishing effectiveness is highly dependent upon skill of the fishing crew. Stationary nets work only on mobile species. Because no gear is perfect, we always deploy more than one method so that the advantages of one method compensate for the disadvantages of another. The best designed program I ever worked on was in the Susitna River of Alaska, a large glacial river system with a grand variety of habitat types from headwaters to ocean. In the fish work of this river each field crew carried with them five different sampling gears (minnow traps, Fyke nets, snorkeling equipment, electrofisher, beach seine) and for each location to be sampled the crew made a field decision of the three best methods for that location. They had to use three methods, but they got to choose the best three of five. An expensive program? Yes! And effective program? One of the best I have ever seen. (Full disclosure, I was involved with the project after the study design stage so cannot claim ownership of this great study design).
Select methods consistent with previous work (when they align with objectives): Often a bit of research will show you what work has been done in the area before. When possible it is valuable to use the same methods as were used previously, as long as they will meet the needs of your purpose, so that you can compare with previous work. This allows examination for changes over time, over space, and possibly even trends if considerable work has been done previously. If you choose to use different methods from those previous you will be unable to make these valuable comparisons.
Impose quality control on fish identification: Accurate fish identification can be quite challenging. There are geographic differences in colour, many cohabiting species can appear superficially very alike, and we can often have hybrids between two species. Inaccurate identification is, unfortunately, ubiquitous. Any sampling program requires a quality control process on identifying the catch. A few suggestions:
- Provide field crews with species list of known species in the drainage they are sampling; if they identify something other than on the list, they better have very good documentation on their identification process.
- Photo-document examples of the species that the crew is unsure of. Submit those photos to colleagues for their identification.
- Train field staff in lateral line counts and fin ray counts – these meristic features are more consistent than colour and markings.
- Do not guess – it is far better to record a species identification as tentative rather than to imply certainty that is not there.
- For difficult to identify species, train field staff to record why they think it is that species. Which features did they use to draw their conclusions.
- Consider including voucher specimens in your sampling. Retaining one or two individuals of hard to identify fish, properly preserved in the field, would go far to ensuring your field crews are identifying fish accurately.
I trust some of these suggestions may be valuable to some of you facing fish sampling design tasks. To build upon these ideas, NRTG offers a MicroCourse – Advanced Fish Sampling – for those interested in greater discussion about fish sampling and effectively using our daily tools to capture fish.
Sean Mitchell