Citizen Science Takes Flight: Benefits and Challenges in Data Collection

The sun was blazing and the air was alive with insects when I visited a local nature reserve a couple of weeks ago in search of Chalkhill blue butterflies.

My uncle had been there a week before and returned with tales of vast clouds of butterflies, claiming you could barely place one foot in front of the other for fear of treading on one! My uncle's not one to exaggerate but even I had my doubts as to whether this could be true. Due to last year's wet summer, and this year's uncharacteristically cold spring in the UK, butterflies here are at an "historic low". But I went along anyway, camera in hand, hoping to glimpse at least a handful of Chalkhill blues - a beautiful, iconic species of chalky grasslands - and I wasn't disappointed...

These striking butterflies were present in their hundreds as I paced cautiously through the reserve, enchanted by the myriad flecks of blue that flickered up from the ground with every step like showers of living confetti...

__________

Not just Chalkhill blues but Marbled whites, Meadow Browns, Gatekeepers, Speckled Woods, Large Skippers, and plenty of day-flying moths too, like this Six-spot Burnet on the right. It was a real treat for the eyes, and for a budding wildlife photographer... but it was also SCIENCE.

Now I know what you're thinking - How can prancing about a meadow photographing butterflies be science? But I wasn't just there to see and photograph them, I was there to count them, as part of the Big Butterfly Count - one of the most prominent UK projects in the increasingly popular field of Citizen Science.

Broadly defined as "the involvement of volunteers in science"⁵, citizen science has grown rapidly over the past decade to become a mainstay of ecological research on biodiversity conservation. Many of the processes that ecologists wish to study occur on very large geographical scales, making them difficult to assess using ordinary research methods. Having a nation, or even several nations, of voluntary research participants therefore facilitates a vast geographic reach in data collection. It also enables the public to participate in, and gain a greater understanding of the scientific process and, in many cases, a greater connection to the natural world².

What can we study?

There's a huge range of research topics which can be addressed using citizen science data including, but certainly not limited to²:

The Big Butterfly Count - 20^th July - 11^th August 2013

The Big Butterfly Count is, unsurprisingly, a nationwide survey of UK butterfly species. The aim of the project, however, isn't just to help identify trends in butterfly species. Counting butterflies can be described as "taking the pulse of nature" because butterflies act as excellent biodiversity indicators, exhibiting rapid responses to changes in their environment. So trends in butterflies can also act as early warning signals for other potential wildlife losses, as well as helping to understand the effects of climate change and, of course, to help plan for the conservation of specific butterfly species.

___________________________________________________________________________________________________________

_

With 223,000 individual butterflies and day-flying-moths recorded across the UK in 2012, the Big Butterfly Count has quickly become one of the world's biggest surveys of butterflies.

The project's simplicity (as well as the wide appeal of its subjects) is most likely key to its success. All participants need to do is to count butterflies for 15 minutes during a sunny spell of weather. Volunteers can either record their observations online later, or, thanks to the wonders of modern technology, submit them straight away by keeping tally on the Big Butterfly Count app.

Counting butterflies, however, may not be quite as easy as you'd initially think. When I started looking into the challenges associated with the collection and analysis of citizen science data, I realized that I'd actually encountered several of these difficulties myself on my trip to the chalk pit reserve...

Challenges:

Participant skills:

Now I'm no lepidopterist, but I like to think I'm pretty good at recognizing the majority of common butterfly species in the UK. There's a difference, mind you, in spotting the occasional peacock butterfly on a bush in the garden, and being in a meadow surrounded by butterflies, moths and other insects and seeing, all of a sudden, a large orange-winged butterfly dart across your field of vision. Was it a Comma? A painted lady? A silver-washed fritillary? Alas, it was gone before I could blink so, sadly, whatever it was won't make it onto my butterfly count, even though it was there during my 15 minute slot.

Another problem I had was that female Chalkhill blues have brown upper-wings and actually look incredibly similar to Brown argus butterflies but with a dusting of blue near their bodies. The ones that I could get close enough to I could tell were females (which made sense given the number of males of their species around) but what's to say that one of the others wasn't a brown argus?

This highlights a few common issues regarding citizen science data: first, that the identification skills of participants are much lower and more varied than that of professional biologists, and also that they are less aware of when and where a certain species might occur². I wasn't sure, for example, whether I would have been likely to see a brown argus in a chalk grassland anyway. (I've since discovered that this, along with limestone grassland, is exactly where they like to hang out, making it even more troublesome to determine whether or not I saw one!). The biggest problem, however, is data biases which form due to a participant failing to detect or recognize a particular species. This results in part of the variation in presence or abundance of individuals being due to variation in detectability (how easy it is for volunteers to spot and identify) rather than the true presence of absence of that individual. This is why many programmes only rely on a selected list of the most common species - usually those which are easiest to detect and most abundant¹.

Variations in sampling effort

Now this one doesn't apply so much to me, being the dedicated naturalist that I am (*honest*), but in some cases, samples can become biased by participants either over-reporting rare species, or under-reporting common (seemingly "boring") species². They may also give up sampling altogether if there are no interesting organisms to be seen. (When we all know that a count of zero is equally as important as a count of 250...)

Abundance estimations

In my case, I had the opposite problem - there were SO MANY butterflies that I had no idea how to estimate their numbers. I'd learnt scaling up techniques before when guesstimating tropical fish species abundances on scuba diving surveys. You take a certain, smaller section of what you're looking at (i.e a shoal of fish or a cloud of butterflies), count the number of individuals in it, and multiply that to get an estimate of the number of individuals in the total area. That's pretty difficult even when you're looking side-on at a small shoal of fish, but try doing that in a meadow crammed full of fluttering insects...for 15 minutes...

I still haven't quite decided what number to record...(good job I have until the end of August to submit my count!)

A winning combination

It's clear, then, that citizen scientists have their flaws, indeed there is call for greater emphasis on data quality. One example would be to use quizzes and games to evaluate observer skill. This would both improve the participants ID skills, and allow data on observer quality to be incorporated into the database to help account for bias².

As always, the internet can help! The development of internet and cell phone technologies has been instrumental in the popularisation of citizen science, allowing instant transmission of data, the combination of electronic sensor data with observation logs, and the immediate validation of observations2. Good old twitter is actually a great platform for getting help IDing all kinds of critters, as I found when I shared a photo of some other insects I'd spotted with, as I termed it, "ridiculously long antennae". Within an hour some friendly insect experts had helped me identify them as Nemophora metallica, or "Brassy Longhorn" moths, from the Adelidae family - also known as fairy longhorn moths.

The field of citizen science will no doubt continue to grow, becoming ever-more refined and taking advantage of the latest technological advances and social media platforms to engage and involve ever expanding communities of citizen scientists. We've still got plenty of creases to iron out, but the future's looking bright for "the world's largest research team"³.

Have you ever taken part in a citizen science project? What difficulties did you encounter? How much do you think participant errors and biases affect large, nation-wide data sets?

Sedeer el-Showk has also written a post about butterflies over on the accumulating glitches blog looking at some curious patterns in butterfly mating and habitat choice.

__________________________________________________________________________________________________________

References/Further reading:

1. Devictor, V, Whittaker, R, & Beltrame, C. Beyond Scarcity: citizen science programmes as useful tools for conservation biogeography. Diversity & Distributions 16 354-362 (2010)
2. Dickinson, J., Zuckerberg, B. & Bonter, D. Citizen Science as an Ecological Research Tool: Challenges and Benefits. Annual Reivew of Ecology, Evolution and Systematics 41 149-172 (2010)
3. Irwin, A. Citizen Science: A Study of People, Expertise and Sustainable Development, by Alan Irwin. London: Routledge, 1995
4. Kremen, C., Ullmann, K. S., & Thorp, R. W. Evaluating the Quality of Citizen-Scientist Data on Pollinator Communities. Conservation Biology 25 607-617.
5. Roy, H.E. et. al. Understanding Citizen Science & Environmental Monitoring. Final Report on behalf of UK-EOF. NERC Centre for Ecology & Hydrology and Natural History Museum. (2012)

Photo credits:
All author's own, on Flickr