As I mentioned in my last bird nerd post, I have quite a lot of data on the birds that visit my garden, and am keen to hear ideas for questions I could look at with it. Someone suggested that it might be worth looking at whether birds with similar feeding habits have similar patterns of visits over the year. So I gave it a go.
First I tried to work out how I could group my avian visitors, and settled on the following categories:
- those that feed from the seed feeder (house sparrows, great tits, chaffinches)
- those that feed from the suet pellet feeder (blue tits, coal tits, black caps, long-tailed tits)
- those that eat seed from the table (wood pigeons, collared doves, feral pigeons)
- those that eat suet pellets or mealworms from the table (starlings, magpies, jackdaws, crows)
- those that feed from the ground / other sources (blackbirds, song thrushes, pied wagtails, wren, dunnock, robin)
Of course there is a certain amount of overlap. For example, sparrows and bluetits will feed from both the seed feeder and suet pellet feeder, but do seem to have preferences.
I then created some simple line charts, using the average number of each species seen per observation day for each month of the year, based on data from June 2012 – May 2013. Here are the charts.
For most of them it looks like the average number seen per observation day is independent of feeding habits. But there may be some relationship for those that feed from the ground or suet pellets or mealworms from the table.
The patterns are unlikely to be driven by changes in the availability of food in my garden, as this is broadly steady throughout the year. However it could be linked to the availability of other food sources beyond my garden. It may be that birds in this category are more influenced by the weather than the other categories, so fluctuations are more in line with each other.
It’s not conclusive evidence, but it’s an interesting hypothesis. When I have time I will use data from the whole 3 years to draw up scatter plots for pairs of birds whose average numbers seem correlated. Can you think of other ways I should test for a relationship? Are there any other questions you think I should look at?