Data Diaries #1 - Bird pellets & small mammals
Data is at the core of what we do as a Local Environmental Records Centre. This series shines a light on the breadth and variety of the records we receive and how we deal with them.
When we think of the typical process of making biological records, we typically imagine a recorder in the field – surveying a specific site or habitat, or heading out to seek a certain taxa, group, or species. We picture naturalists holding binoculars or hand lenses, pacing through grasslands, sweeping for insects, or lying level with the ground to identify a fungal fruiting body. Rarely do we picture someone making records without ever leaving the house, but that is the case with the first dataset in our Data Diaries series.
In September 2023 we received an unassuming stack of paper from a recorder which contained a gold mine of small mammal records from across the Yorkshire region. Rather than being observed by eye, caught in a small mammal trap, or photographed via camera trap, these records were all made by dissecting pellets from birds of prey.
From 2015 – 2023, the recorder was provided with samples of bird of prey pellets, usually Barn Owl, by contributors from different locations across Yorkshire. The recorder carefully dissected the pellets and made a note of the small mammal species present in each sample, based on examining the bone structures of each prey item. Because the samples contained differing numbers of pellets, the recorder also worked out the relative percentage of each prey item in the sample as well as recording the actual number of each prey species, allowing the composition of samples of varying size to be compared to one another. Up to eight species were recorded from each sample, including protected and priority species such as the Harvest Mouse and Water Shrew. Over 300 samples of pellets were provided to, and dissected by, the recorder, with each pellet having up to 9 species contained within it – a lot of data processing to get our teeth into!
An example of one of the page of records we received, with the recorder name, date, and locations removed. each row of the table was linked to a single location on a specific date. The acronyms in the headings refer to the eight different mammal species recorded. The top number refers to the number of prey items of that species in the bird pellet sample, and the number below is the % occurrence of that species in that sample.
Usually, we receive records where one entry pertains to one species, on one date, in one location. This time, each ‘record’ in the table provided related to the pellets themselves, the bird which they originated from, and the prey items of that bird.
Owing to the complexity of the data structure, the records were collated at two different levels. Firstly, the ‘base’ record of each pellet sample was made. This resulted in ~300 records of Barn Owl (or another bird species, if the group of pellets had come from another species) where the record type was listed as ‘Bird pellet’ and the number of pellets making up the sample was recorded as the Count. A unique reference number was added into the Comments field of each record, to enable these individual bird records to be linked to multiple subsequent small mammal records at the next level.
A comparison of bird pellets from different species, including Barn Owl. (c) Barn Owl Trust
Next, small mammal records were made based on the species present in each pellet sample. The date, location, and grid reference of each were copied from the relevant pellet sample (bird of prey) record created earlier, and a note in the Comment field provided the same unique reference. In this manner the data user can obtain both the record of the source species and the prey species and link the two occurrences to one another.
Naturally we have to make a number of assumptions with this data, including that the pellets came from the same, single bird and that the prey items contained within it originated from the same location (grid reference) that the bird of prey did. We can’t be sure that these assumptions are correct, but by entering the data in a structured manner and making the original format of submission clear, the user will be aware of the limitations. Whilst this is an artefact of the way in which the data was collected, this method still has a number of benefits compared to more traditional mammal recording. Typical methods of recording small mammals can be physically labour intensive (such as live trapping) or yield poor results (trail cams often don’t pick up very small animals). In combination with the small mammals’ often nocturnal nature, small size, and adeptness at avoiding human contact, this has resulting in them being a highly under-recorded group. As yet, we hold under 200 records of Harvest Mouse and under 100 for Water Shrew, for example. Different species may also require different methods of survey based on their life cycle and habitat.
Using pellets, meanwhile, avoids some of these pitfalls. Whilst still time and skill-intensive, it allows us to obtain data on multiple species at once, and by calculating their relative abundances within the same sample, we can still make some inferences about how common one species is compared to another in the same area. Aside from the valuable records obtained, this dataset submission also showcases the opportunities for recording outside of traditional contexts. It is a great example of just one of the ways in which those who might not be able to easily record in the field due to health, accessibility, or other commitments can contribute valuable data and build a speciality.
Thanks to the skills and dedication of this recorder we will now be able to add a large number of recent and highly valuable mammal records, including those of protected species, to our database.