Using species records and ecological attributes of bryophytes to develop an ecosystem health indicator

Abstract

• Bryological records and Ellenberg values were combined as Ecosystem Health Indicators. • We analysed change across Scotland (1960–2016) for each 10 km × 10 km grid square. • Ellenberg Nitrogen (N) showed a humped relationship indicating recent recovery. • Increasing trends were evident for January and July Temperature. • Bryological records can form the basis of an effective Ecosystem Health Indicator. Environmental policy decisions should be based on robust indicators of changes in the environment. In Scotland these have been formalised as Ecosystem Health Indicators. Indicators work best where there is a direct link between what is measured and environmental change; changes in indices of species diversity or abundance provide alerts to environmental change but do not in themselves provide evidence of mechanisms. An alternative approach is to replace species occurrence data with species habitat or attribute values to create an indicator that links changes in species occurrences to environmental drivers. A series of indicators based on biological recording data, Ellenberg habitat preference values and mean climate niche was created by taking all records for bryophytes in Scotland (522187 records, 1960–2016) and analysing changes in mean attribute values through time for each 10 km × 10 km grid square. The resulting mean attribute values were analysed at the country level and showed clear quadratic (humped) trends for Ellenberg Nitrogen (N) and Reaction (R), an index of environmental acidity. There are linear decreasing trends for Ellenberg Light (L) and Precipitation, and linear increasing trends for January and July Temperature. Ellenberg Moisture (F) showed a clear trend, but it was not interpretable and appeared influenced by recording effort. Disaggregating to region was partly successful, but reduced samples sizes meant that for many regions the national trend was not significant. Ellenberg N and the two Temperature indices appear to be useful indicators as they are interpretable and performed better in the downscaling. Changes in L were also interpretable, but it performed poorly in downscaling to regional values. Ellenberg R appeared to be driven by changes in N as the two values are highly correlated across species. The downward trend in Precipitation ran counter to increases in rainfall suggesting that temperature driven evaporation is more likely to be driving changes in bryophyte distribution and whilst the indicator performed well it is difficult to explain to non-experts. Opportunistically collected data, like plant records, have great potential to be employed as indicators, especially as citizen scientists are involved and state resources can be used for data archiving and analysis. Also, conversion to attribute values removes many of the biases brought by differences in recorder effort and it directly links changes in species occurrences to environmental drivers.