Do techniques have a bearing on species representation in data?

This year, I have made a serious attempt to make sure that I record everything that I see whilst in the course of daily walks. In addition to noting all visual records I have retained specimens of those species that I cannot do in the field. In this way I had hoped to replicate the efforts of several Facebook members who record on a daily basis and also John Bridges who has recorded by sitting and watching a given patch and then retaining specimens of his photographic subjects. I have still to work through the specimens John has collected, but this early analysis may help to demonstrate how the basic photographic database compares with detailed recording based on field data and retained specimens.

The charts (figure 1) show only those records where it has been possible to take individual specimens to a firm identification. I have presented data for only the 50 most frequently represented species because the tail of all three datasets is highly variable and is probably not significant in telling a story about what people most frequently see. Note that I have added one species to the photographic dataset because both the segregate and aggregate of Xanthogramma pedissequum appear in the photographic database.

The story is complicated because the general photographic dataset comprises records from across the country whereas the dataset that both I and John Bridges have assembled is much more closely aligned to a region. My data primarily cover a ten mile radius of Stamford at the junction of Northants, Cambridgeshire and Lincolnshire but include regular visits to London and short visits to both Somerset and to the Spey Valley. John's data are confined to a block of four ten km squares in County Durham.

Inevitably, some of the differences are geographical. Even so, it is clear from figure 2 that once one starts to retain specimens the Tribes that figure within the dominant data change substantially (figure 2).

Figure 1. The 50 most frequently recorded in 3 datasets from 2016: RM = Roger Morris; PH = Photographic Records; JB = John Bridges. The figures represent the rank order of the species within individual datasets.

Figure 2. Numbers of species within each Tribe represented in each of the datasets depicted in Figure 1.

There is a lot more analysis to perform, but even these basic charts show how the dataset differs markedly between the general coverage by photography and the coverage that may be expected from an experienced Dipterist. What becomes very apparent, however, it the way the photograpahic dataset can be augmented by specimen retention. John Bridges dataset aligns rather more closely with my own data than it does with the general photographic dataset.