Louise writes: I used to be a biochemist studying human immune system malfunction whilst being a part-time naturalist and conservationist. Then I converted to being an environmental data geek, which is what I do part of the time in a vague attempt to pay the bills I have been a birder since childhood, and am now the Cambridgeshire county bird recorder, and am also a butterfly and moth enthusiast, with an interest in several other taxon groups including lichens, ants and molluscs, and when not in front of maps or a database can usually be found in woodlands carrying out vital management work, or surveying farmland birds.
Volunteers in the conservation sector: Part 3: Being a county recorder – volunteers with a spiralling data problem?
I took on the role of butterfly county recorder around a decade ago. As someone working in the environmental records centre network and an active county naturalist, it seemed the obvious thing to do at the time when the post became vacant. A few spreadsheets of submitted records, transect records to co-ordinate and check through for oddities. Not too onerous a task. A couple of years later, I was also approached by the county bird club, having spent 6 years co-ordinating the county and national atlas work [BTO atlas 2007-2011] to become county recorder, a role I was very pleased to accept. A similar role, getting in records from county-based birders, directly, through our club sightings pages and increasingly through Birdtrack. Not too bad, with probably 100,000 records per year. It does take a proportion of one’s time; data management, dealing with poor field behaviour of individuals putting the welfare of wildlife at risk, and all the other things like attending meetings, as alluded to in my section on the time cost of volunteering.
And then came the citizen science revolution, apps and online recording, and the realisation that getting anyone and everyone to participate in projects such as the Big Butterfly Count was a great way to engage new audiences. I agree, however, I don’t think any of us bargained for the burgeoning popularity, and promotion by conservation charities, of recording popular groups. What was, a decade ago, an interaction with enthusiastic naturalists in your county in person some times and usually by email or phone has become a faceless interaction with a spreadsheet and a web validation interface. With 10 times the amount of data, maybe more, surely that’s great, more records, giving more information across the county. Well, it would be nice if that were true. The sheer volume of data means that we have to take at face value the ID of any common species…. I would never, and do not have time to, query the ID of a Blackbird or a Robin or a Black-headed Gull, or a Speckled Wood butterfly or a Small Tortoiseshell. But, always at the back of your mind is the doubt over whether the common stuff IS correctly identified, and the reason for this follows. On occasions when a link to a twitter page or a photo is given, there is a low percentage of mis-identification of species which seem to me, and to many other naturalists, straightforward to ID, so how can it be wrong. Yet, in that link is the proof that is HAS been identified and/or entered into the system wrongly.
With the new GDPR regulations, it also means that, as a county recorder, or verifier as I think we have become, we have no direct means to contact the individual if it is on a national data site through which we have been granted access to verify county data – there may be a flag to raise a query, but in my experience they then usually go ignored by the user, who never responds. Those who still use direct record submission are obviously much more likely to interact if there is a query, purely by dint of their initial means of contact.
This raises the big question then. What IS the point of gathering huge amounts of weekly lists of birds from a site, with no counts made or no additional comments eg singing, or nesting, or the presence of young, added to the record? I increasingly struggle to know. Personally, my recording will be a list of things seen, heard, identified, with counts for many species and comments where appropriate. Because that’s what I was taught to do, by everyone who encouraged me as a child and a young adult. The ‘record it through the app’ with no further mentoring which is what tends to be the norm these days does indeed lead to spiralling numbers of records being made, and whilst the number of records for me to wade through for butterflies and birds has grown 10-fold, the quality has probably gone down at least 5-fold, although this is an incredibly difficult metric to assess……
Harking back to earlier, the amount of time taken to be a recorder, or a field naturalist is a thing under pressure in everyone’s lives, and I am sure that the web-based systems are designed to simplify some things. I, and others who perform similar roles, feel, however, that with increasing digital recording, constraints on how we can view and verify that information and now an increasing number of online platforms for recording (even birds have 3 different choices, Bird Track, e-bird and trektellen, although moves are underway to make all viewable through one platform; this is definitely not the case with other online systems), that the task has become near impossible, and increasingly, onerously, a daily chore rather than a pleasure.
So, when a volunteer decides they have had enough, of being a county recorder, of walking a transect as part of a monitoring scheme, of being on a committee, or recruiting the next generation, where indeed will the next generation come from? If anyone wants a decent work-life balance and a sane existence then I suspect there will NOT BE a skilled next generation and I certainly wouldn’t blame them. Right, I’m off out to enjoy some wildlife and not record it now.
[registration_form]
The whole idea of app based data gathering has roots, at least for birds, from the Christmas Bird Count in North America where a large number of records over a large geographical area and over several years were used to infer/deduce changes in bird populations. The whole idea is that outlying records such as mis-identifications are simply averaged out by the volume of data. There is a rapidly development body of statistics for such large scale data gathering which supplants more detailed, lower scale, lower resolution, but more precise surveys. The aim is continental-scale monitoring of populations. Whether this works at a smaller scale is probably open to question. The major advantage of this approach is that other large scale datasets could be integrated to create environmental models that map population changes on a scale that makes it hard for policy makers to ignore.
The other advantage of apps and the associated websites is that it provides great feedback to the observers. Information on what to see and where to go is easily available. That in turn encourages more people to participate. Annecdotally, the number of birders here in Portugal has exploded since eBird became popular. Making access to nature more democratic can only be a good thing.
In the end there is definitely a trade-off between quality and quantity. I’m hopeful that quantity will be more effective at driving the political changes needed to save the environment and everything and everyone in it.
Yes, those things are all true, but I will ask – are you a county recorder or other data manager? If not, then yes, these are the advantages f the system, but everything has to be checked and verified, and althoughthere aregreat auto-verify routines for the commoner things, lots of records still ave to be waded through, and the lackof loggin g number or breeding activity for most species seen is still at an unsatisfactorily high level. I know bTO are adressing this by repeatedly askign birders to add breeding codes and number seen but the reality is most wont. And with ther taxa, hoverfiels, lepidoptera, countyand counrty-wide online systems for all taxa, then the ID issues and regular non-response from observers on follow-up is high and makes many record useless. Things have to be recorded, but then those things have to be verifed, and although I DO welcome the growth of online recording portals, the increase in lower-quality data is something we as a community need to watch otufor and be aware of.
This is a debate that is raising its head in relation to recording a variety of different taxa. It is clear that recording apps have encouraged many people to take a closer interest in the wildlife they encounter and to appreciate a wider range of taxa and this is undoubtedly beneficial. But, as Louise makes clear, it does come at a cost. The ease with which records can be made and submitted means that the sheer volume of records generated is becoming enormous and verifiers are at risk of being overwhelmed (if they aren’t already). In virtually all cases the verifiers are unpaid volunteers and the risk is that they will give up in the face of a thankless task.
The quality of the records submitted is also an issue with incorrect identifications common, particularly in taxonomic groups such as many invertebrates where critical features may not be visible in a photograph. It may be possible to iron out these inaccuracies when carrying out statistical analyses across a national data set and I recognise the benefits of this, as outlined by Stuart above, but records are used for other purposes too. A planning decision might turn on the presence or absence of a particular species, for example, so we do need to have confidence in the information held by local biological records centres.
I don’t really know the answer to this. I would certainly not wish to discourage anyone from recording the wildlife they see and seeking to share their records and in any case these apps are here to stay, but the people behind the apps do need to do something to address the issues Louise raises.
“This raises the big question then. What IS the point of gathering huge amounts of weekly lists of birds from a site, with no counts made or no additional comments eg singing, or nesting, or the presence of young, added to the record?”
Fair point, but as I’m sure you’ll are aware, collecting a complete list of birds seen at a site for BirdTrack generates a ‘reporting rate’ for any given species over a year timescale, which can act as an accurate early warning system for any declines/increases.
I agree, counts are preferable, particularly from a county recorders perspective, from which I assume is the angle that you are referring to.
As an ex county recorder I have almost exactly the same views as Louise. If it was not for the enormous workload created by trying to keep up with BirdTrack, eBird, and, in my county at least, Naturespot records I would probably have continued beyond the ten years I served. Another thing that is time-consuming is trying to follow up records that you only see on Birdguides and RBA: if the news people have any contact details for observers only about half bother replying and as far as I was concerned that meant the record was lost as it could not be verified.
My county (Leics/Rutland) has a perfectly good recording system but unfortunately it is not directly compatible with all of the new options available to observers
The ‘reporting rate’ can also be really useful with migration timings……I find the birdtrack maps invaluable for that – being able to compare this year to the usual.. I think the biggest concerns are not in avian data but, as has been flagged as well by some of the espondees, in othe rtaxa eg fungi, lichens, invertebrates.
Let me take the liberty of summarising the responses above as asking the question: How can we use this new data in old ways?
You shouldn’t and probably can’t. Let’s take the example of birding sites in Lisbon, Portugal, https://ebird.org/region/PT-11-LI/hotspots?yr=all This is a fundamentally biased view as my patch, Jardim do Cabeço das Rolas is the most visited place in the city. It’s awesome for migrants in Autumn – I bird there as often as I can.
If you look at list for any site or for the city as a whole then there are some quite dodgy records. It the data useful for creating a species map for the city, yes. Is it systematic and scientific? No. However if this page is shown to the city council and the parks departments (not been done as far as I know) then there could be an interesting conversation about all the wildlife that can be found in the city’s green spaces – Griffon Vultures flying over the centre of Lisbon – it happens surprisingly often. That makes policy decisions on promoting Green Lisbon much easier and in turn it boosts the cities image with all the benefits for inhabitants and tourists and all the knock-on effects that result.
Going back to Jardim do Cabeço das Rolas and the reason I included it. It’s the only park in hte city, as far as anyone knows, that is almost exclusive filled with native plants. That attracts a lot of birds. If the decisions taken when the park was created, were replicated as park department policy across the city then there would be even more wildlife to see.
So, is the new source of data, large and noisy with some rubbish thrown in? Definitely. Should you try to clean it up? Probably not. Does it matter? Probably not. Is it suitable for small scale decisions? Definitely not. Is it useful for large-scale decisions and policies? Most definitely.
Interesting comments and some resonate with the situation here in Lothian, though we “only” have c. 100k online p/a (80k BirdTrack/20k ebird). A tiny fraction of those records have that key info on breeding status, number, sex etc. Positive development that trektellen now embedded in BirdTrack, that used to be a few hours work and I struggle to find a few hours. In terms of common birds I am baffled by comments from other recorders about spending time to *remove* then from BirdTrack archive, to me those are highly valuable records but only at a global level, not one by one. Scott Mayson has given some nice examples of this in his recent BB eye “Using BirdTrack in annual bird reports”, BB 112, 242 (May 2019). I’d like to be able to include that kind of thing but I’m not actually report editor and I don’t have time to assist, but certainly one for the future that has potential to greatly enhance annual reporting of what is actually the majority of our bird life, but which does not merit listing of individual records.
In terms of data processing it’s a mixed blessing – data coming in from online systems generates perhaps 5 times the volume we had historically and draws in literally hundreds of new observers, it is wonderful to download a single file that has a consistent format, hence needs no individual editing or manipulation (contra direct submissions in spreadsheets which may take perhaps an hour on average to convert to common naming conventions and column orders used in local database) – however, unless you use BirdTrack as the actual repository for the county/club records archive there is a huge new job which is converting all your other records to the same format. Sounds trivial but unless you can sort your records by a species code (we use EURING) you are sunk and you can’t extract for analysis/writing accounts, etc. Unfortunately the BirdTrack format and names have changed many times, it seems to be a new column order every year (this year we have addition of “Geometry type” = Point/Gridref/Polygon – all good stuff but needs to be stored somewhere!) and more importantly new species name formats (how many times have Brent geese and alba Wagtail conventions changed now?!) with a new “BTO species code” for sorting, which is not EURING! It is not a trivial job for me to convert over 1 million records in the existing database, sorted with EURING and in some cases historic versions of species names, into a format consistent with latest BirdTrack, which will probably change again next year. Macros written to convert American names in ebird to BTO Common names all now need updated again. This is many hours work. It would be highly beneficial to me if BTO/Cornell could sort the BirdTrack/ebird mutual exchange at a global level, there had been discussions a while back but I think the practicalities intervened. I don’t know how we solve the general problem in the short term but I think in the long term our best option will be getting all of our historic data into BirdTrack, so that any changes they make do not require any action from me on a separate local archive…for this I would need a volunteer to do the work, and/or wait till I retire, simply impractical right now…
Stephen Welch, Lothian SOC bird recorder
We have nearly 1000 contributors of records in Hampshire. I am sure it was more fun when there were 150 observers who all knew their stuff and kept detailed notebooks which they transcribed at the end of the year. Each person would write their sightings for a particular species on a separate line and the County Recorder would cut and paste these (using scissors and paste) onto another sheet beneath other records of that species. We still have these sheets and an interesting challenge is to now get all of that typed into a database before it all disintegrates. That will either take a long time, lots of people and/or lots of money! Today’s observers often consult Google before sending their sightings and once Google has reinforced their beliefs there really is no turning back. Two recent records that spring to mind are a Chough in a Southampton garden that called remarkably like a Blackbird, and a Corncrake that had nested in a field for at least 15 years and could be identified by the way it hovered. I try to treat everyone with kindness as we have all made such tall claims in our early birding days. It reminds me of a Long-eared Owl that I confidently identified when I was 10 based solely on its bark-like call. It was at least 15 years later that I decided I had to scrub that record from my notes and change it to “presumed dog barking”!!
Keith Betton
Chairman & County Recorder
Hampshire Ornithological Society