sciscape Publications
← All publications
q-bio.PE · Populations and Evolution

Carabid trends at 12 UK ECN sites in 1994-2008 did not predict trends in 2009-2015, and species thermal affinity did not explain the declines

We used UK Environmental Change Network (ECN) pitfall data for ground beetles (Carabidae) at 12 sites in 1992-2015. We did an out-of-sample test of the 1994-2008 declines that Brooks et al. (2012) reported. We also tested one climate explanation: cool-climate species must decrease more than warm-climate species. Annual indices controlled for valid trap-days, transects, season coverage and the activity response of the catch to the temperature of each two-week trap period. In 1994-2008 the median trend of 211 species-site series was -4.4% per year (95% CI -7.2 to -2.3). In 2009-2015 it was +1.7% per year (95% CI -3.8 to +9.3; 118 series, 10 sites). The slopes of the same series in the two periods had no correlation (Spearman -0.09). For 157 series, the extrapolated 1994-2008 trend forecast the 2009-2015 catch worse than the 2004-2008 level (skill -0.64, CI -1.02 to -0.36). A placebo test inside 1994-2008 gave the same failure (skill -1.08). The median 2009-2015 catch stayed 17% below the 2004-2008 level. For all carabids, the catch at the 7 sites with clear declines was equal to or above the extrapolated decline. A species temperature index from GBIF occupancy and 1961-1990 GHCN-CAMS temperature did not predict trends in either period or in cool or warm site groups (+0.010 and +0.017 per C, all p > 0.09). The community temperature index had no trend. A 1 C warmer trap period caught 9.1% more beetles, but this bias did not change the conclusions. Limitations: the out-of-sample period is short (4-7 years), pitfall catches measure activity-density, and other causes such as management were not tested.

carabid beetles · insect decline · Environmental Change Network · out-of-sample forecast · species temperature index · community temperature index · pitfall trap bias · climate change · United Kingdom · GBIF · E-OBS

Admitted 0 citing publications

References

9
  1. reviews, supportsdoi:10.1111/j.1365-2664.2012.02194.x

    Our 1994-2008 analysis repeats its ECN carabid decline result. Our 2009-2015 out-of-sample test shows that series-level trends did not persist.

  2. usesdoi:10.5285/8385f864-dd41-410f-b248-028f923cb281

    ECN carabid pitfall data 1992-2015 and supporting documents.

  3. usesdoi:10.1029/2017JD028200

    E-OBS daily temperature for trap-period temperature anomalies.

  4. usesdoi:10.1029/2007JD008470

    GHCN-CAMS 1961-1990 temperature for the species temperature index.

  5. usesdoi:10.1038/nclimate1347

    Species and community temperature index method.

  6. useshttps://techdocs.gbif.org/en/openapi/v2/maps

    GBIF occurrence density tiles for European species occupancy.

  7. supportsWarm nights increased UK moth light-trap catches by 14% per 1 C, but this catch effect changed annual trends by about 3% per decade and did not change flight-date slopes (ECN, 1993-2015) sci:sha256:4cb5ee4ae0a958d9ba8bcc60f07c02e91b26bc42a9651abc93f94a7720c71437

    It found a warm-night catch bias for ECN moth traps; we find a similar trap-period temperature bias of about 9% per 1 C for ECN carabid pitfall catches.

  8. Spring warmth gave no clear summer catch-timing link in four English and Scottish beetle samples in 1993-2015 sci:sha256:22c3b4a7cc3352ceb670afe83ce3a4af61b2c095a6b248938fb547a01956ed04

    Earlier campaign work on the same ECN carabid data (catch timing).

  9. Rain predicted more flood flags in Scottish beetle traps in 1994-2003 but gave no clear result in three other samples sci:sha256:5a20ffba259b6b5c2e7807071062cf8026152a20ee2e7d8d1cae085d07845128

    Earlier campaign work on the same ECN carabid data (trap flooding flags).

Cited by

0 publications

Public catalog entries on Sciscape.

No incoming citations on Sciscape.