Getting the most from data

In previous posts, I have discussed the importance of recording all species and not just species considered to be of conservation importance (Why Record Common Species). In this post I thought it might be useful to discuss the importance of extracting the most information one can from a biological record.

When I log records posted on the UK Hoverflies Facebook page, I make a serious effort to extract ALL relevant data. Location, full date and a site name are of course essential. in addition, I record the gender of the animal(s) depicted (as separate lines), flower visits (i.e. flowers visited and some evidence of nectaring or taking pollen). Further effort means that any behavioural notes or other observations (e.g. whether in copula or territoriality described by the observer) are also noted. This is starting to turn the database of photographic records into a very powerful tool for better understanding the ecology of individual animals and of hoverflies in general.

One of the obvious benefits of this system has been that I was recently able to supply several thousands of records to a University project looking at pollinators at Ivy. I hope to be able to do more myself, but sadly my plant ID skills are pretty rudimentary so I fail miserably on garden plants even though I'm generally OK at commoner native plants.

Another great benefit is to be able to look at the phenology of both males and females. Many people might assume that all hoverflies follow a similar pattern; but they don't. A nice example is provided by the two commonest Eristalis: E. pertinax and E. tenax. Both are very abundant at ivy in the autumn, and are again seen in the spring. But that is where the similarities cease.

Phenology of Eristalis pertinax in 2016

Phenology of Eristalis tenax in 2016

In these two species we see two entirely different life strategies.  

Eristalis pertinax largely disappears in the winter. It is possible that a few individuals over-winter, but I think it is more likely that small numbers of animals emerge early, depending upon prevailing conditions. Records suggest that males vastly outnumber females in the Spring, but I suspect that this really reflects behavioural traits, with males holding territory in the spring and being far more visible, whilst females are possibly more abundant closer to breeding sites (a nice little ecological project for somebody to take a look at). In the Autumn females outnumber males, probably because males are of little ecological value once they have mated and they die off earlier in the year whilst females are busy egg-laying.

Eristalis tenax follows an entirely different pattern. Male and female abundance seemingly mirror each other in the Autumn, but by January males have largely disappeared. Females, meanwhile, over-winter and emerge in late February or March to oviposit. The phenology of this species is striking because both males and females largely disappear in late April, and only start to figure again in the records towards the end of May. This clear separation in the generations is interesting because it is not terribly well reflected in the overall phenology data emerging from the Hoverly Recording Scheme database.

The phenology of Eristalis tenax is also interesting because it might be a useful pointer to assist record validation.  In the past I have wondered about some datasets because they seemed to have a very high number of records of E. tenax, when my own observations had already led me to conclude that it was substantially absent for much of the Spring. That can now be shown to be the case and it inevitably allows validators to ask more searching questions about the reliability of data.

The case for better recording

Neither of these examples would be possible if one were to rely on the majority of records submitted to the recording scheme. Most data on iRecord lacks this important element. What is more, most data lacks details of flower visits or is potentially misleading because there may be an entry saying 'on ivy' but upon further investigation it is found that the entry means 'on an ivy leaf' rather than 'feeding at ivy flowers'.

Both the case of flower visits and phenology analysis illustrate how detailed recording can be of critical importance to our better understanding of individual species' ecology and of the ecology of the wider countryside. I think they justify the time and effort that is made to extract data from photographs. I know there are shortfalls in some aspects of my ID skills, and that I am not infallible; but there is now a substantial database that has a constant provenance of having been identified and recorded by the same person. What I hope to see is an extensification of this approach in general recording. These sorts of data may be of great importance as we try to understand what is happening to our wonderful wildlife.