Crane fly project

A s part of our overall project investigating blanket bog management impacts on ecosystem services across three UK upland sites we also determined crane fly (tipulid) emergence (traps) and abundance (transects). This happened at all three sites and for both, the burnt and mown catchments. Comparisons over time (years) and between replicated catchments of different blanket bog management (burn versus mow) will reveal main factors impacting on cranefly numbers and thus bird populations depending on this food source. Notably red grouse, but also important upland birds such as golden plover and dunlin depend on this protein rich food source for chick survival and growth.

We worked by a modified method as used by Matthew Carroll (PhD at York as part of a CASE studentship with the RSPB, see Carroll et al., 2011). Using crane fly emergence traps of 0.11 m2 (i.e. upside down basket with sticky traps inside - see below pictures) and also walk several 10 m transects (assessed for flying or sitting craneflies 2 m either side of the transect) across each catchment to count adult crane fly abundance.

The work also involve modelling impacts on upland birds depending on tipulids by predicting crane fly abundance based on environmental data (i.e. soil moisture). This was done in collaboration with the BTO (James Pierce-Higgins) and RSPB (David Douglas).

Setup of crane fly traps in the field Baskets used for trapping emerging flies From a distance - looks just like sheep

Finally, all traps set ... time to wait and then come back and count. The first counting was in May 2014 and showed a difference between sites (highest counts at Nidderdale, then Mossdale and Whitendale) but also an indication of treatment (T; mown) supporting larger numbers of emerging crane flies than on control (C; burnt) plots, further enhanced counts were observed on brash removal (Br) plots.

Images for crane fly larvae (left) and adult crane fly (right).

Above: We observed a strong positive relationship between emergence of crane flies and soil moisture 2014 for all sites (trap data only). Crucially, the relationship was simialr to that observed by Carroll et al., (2011) in areas of the Peak District. Our relationship is shown in the above graph; note that the wettest Mossdale site seems to have supported a similar population to the driest site Nidderdale but at lower soil moisture and the intermediate site, Whitendale, showed lowest Tipulid counts. Overall analyses (a generalised linear model with negative binomial distribution) for the 2014 data showed:

      • Significant positive relationship between Tipulid count and soil moisture.
      • Significant difference in Tipulid abundance between treatment and control catchments (greater in mown than burnt).

Below: However, although the overall positive relationship held true over three years (2014-2016; left), there was also a notable decline at soil moisture above about 95%, indicating possible drowning of larvae or any other factor related to water logging. Moreover, soil moisture was higher (wetter) in mown than in burnt catchments (middle), particularly at the wettest site (Mossdale), which related to lower cranefly emergence (right) in mown catchments observed at Mossdale overall (i.e. an indication that mowing made the peat too wet for craneflies in very wet years, particularly 2014/15). Importantly, at the drier sites, particularly at Nidderdale, mowing with leaving brash resulted in higher moisture and higher cranefly numbers, indicating a buffering against drying out compared to burnt catchment areas.

Cranefly emergence vs SM (three sites)
SM versus catchment (mown vs burnt)
Cranefly emergence versus SM versus site