Coughton Park, the Alne Wood complex and Coxmere, walked weekly by volunteer recorders since 2018. This report consolidates every weekly recording form into one dataset and tracks what it says about species diversity, Fisher's α and the habitat behind it.
Four standard biodiversity indices, computed from every weekly count and rolled up by season. Coughton Park is the only site walked in all eight years; the Alne Wood complex ran 2018–2023, Coxmere only since 2024.
A Theil–Sen slope and Mann-Kendall significance test for Fisher's α and species richness, computed independently for each survey with at least three seasons of data.
Coxmere (2 seasons) and Brandens Wood (1 season) don't yet have enough seasons for a trend test — shown descriptively only in the summary dataset.
Annual abundance trend (Theil–Sen slope, individuals/year) for every species with a non-zero record across Coughton Park's continuous 2018–2025 run — the site with enough seasons to say something about individual species.
Mean total abundance per recording week, pooled across every visited site-season. Week 1 opens on 1 April; the season runs 26 weeks to late September.
Sward-height surveys exist for the same transects in most years. Joined to the diversity indices above, two relationships clear conventional significance.
Historical daily weather for the study area (Open-Meteo's ERA5-based archive, grid point 52.23°N 1.88°W near Alcester) joined to every recorded week. A raw correlation would mostly just rediscover the calendar — both counts and warmth peak in midsummer — so each week's abundance and weather are first compared against the average for that week of the season, and it's the leftover anomalies that are correlated below.
Not every scheduled week was walked. Shading below is the share of that season's weekly visits that were actually walked, by site.
A site being walked and a butterfly count actually being submitted that week are different things. Most years the two track fairly closely; 2020 is a striking exception.
A one-line description of each; tap the i for the formula and how to read it.
Every source file, every transect's full run, and the complete species checklist — the reference tables the report is built from, for anyone who wants to check the working or reuse the data.
Three raw inputs, consolidated by the scripts alongside this report — see the data dictionary below for how each processed file relates to these.
The Alne Wood complex is shown as its four constituent transects here — each was walked and recorded separately.
The complete site-by-year table behind every chart earlier in the report — see the glossary on the Report page for what each index means.
The standard checklist used on every recording form, with how often each was actually seen across all seven sites and eight seasons.
Every file in processed/, what grain it's at, and what it's for.