- CEBRA: An optimised and standardised sampling protocol for BioBlitz
- Stop it people, the plural of anecdote IS data!
- That's no pollinator, that's a flower visitor
- Add a scale to your macro photos
- What to photograph when counting the wild
- Imagine counting the wild on EVERY street in a city!
- My wild counting workflow
- A butterfly flew through
- Why iNaturalist observations without photos can be research grade
- Why you shouldn’t use a spreadsheet for data entry
- All articles ...
Go beyond species lists in your surveys and BioBlitzes. Tweak Three: how much did you overlook?
One of four tweaks to surveys and BioBlitzes that make them useful for long-term monitoring.
written Jun 11, 2018 • by jonsullivan • Category: Wild Counting
This article is one in a series about four simple tweaks you can make to your surveys and BioBlitz events that will make them much more useful for long-term monitoring.
Tweak Three: how much did you overlook?
Tweak Three: how much did you overlook?
A lot of wild nature is pretty sneaky. Most wild species don’t become successful by sitting around in obvious places waiting to be seen. This makes counting the wild an interesting challenge. It also means that, for many species, you’re inevitably going to miss some of the individuals that you pass. Don’t panic though. With some care and cleverness, this need not be a problem.
Ecologists refer to your chances of missing something as the “detection probability”. Given a choice between a counting method with a high and a low detection probability, high is best, all else being equal. However, all else usually isn’t equal, and counting methods with high detection probability are often expensive and time consuming. Counts made with a low detection probability can still be very useful, so long as your counting method allows researchers to estimate what your detection probability was. That’s the big trick.
Tweak Two allowed you to estimate how many species you missed by keeping track of exactly where and when you were looking. That shows the areas, and times, at the site that were not surveyed, where additional species could have been present.
You could also overlook species even if they were where you were. Perhaps a species was hard to see, buried in vegetation, or perhaps it was obvious but you were just looking the other way when you walked past. Also, one person in your team might be better at finding things than someone else. For repeat surveys, it’s important to know what these detection probabilities were.
There are four simple ways to assess detection probability. If you haven’t tried any of these before, you’ll likely be amazed at how much you overlook, Even top professionals don’t see everything.
Perhaps the easiest option is to repeat some survey units during the survey. For example, walk the same section multiple times and honestly report the species you find on repeat walks, including those that you missed earlier. It’s also useful to send out different teams to walk the same routes or survey the same plots. This method works particularly well for species that stay put, like plants and mushrooms.
You can also get several members of a team to simultaneously and independently record what species they see or hear in the same time period at the same site. You can then compare notes afterwards, to see how the detection and accuracy of counts varied within the team. This method is especially useful for bird counts, since birds don’t stay still so repeat visits to the same sites are less helpful for estimating detection probability.
Another method, often used by plant ecologists, is to carefully look in subplots of a larger plot, once it has been surveyed. For example, once a team has finished surveying a 20 m by 20 m forest plot for all vascular plants, they then survey some random 1 m square plots inside that plot. If they’d found all of the species in their big plot, then they won’t find any more species when they carefully look in subplots. It can be surprising how often that is not the case.
The fourth method, most often used in professional bird monitoring, is distance sampling. I put this last because it’s the trickiest to do, and takes some practise to get good at. That makes it the least suitable of the methods for a community event like a BioBlitz. In distance sampling, you write down the distance between you and everything you count while you’re standing at a site, or walking a route. The further away things are, the harder they are to detect. If you can estimate how much harder, you can estimate how many things you missed. Researchers can do exactly that by applying mathematical cleverness to distance sampled counts. I’ll write more about distance sampling in a separate article.
Which of these four methods you choose to use depends on the species you’re surveying and how keen you are. The first two, repeat surveys and independent simultaneous counts, are particularly straightforward and easy to add to BioBlitz events or to monitoring done by community groups.