Mining bees and their cleptoparasites

Geoff Trevis.

Whilst idly mulling over the factors affecting the distribution and status of mining bees I wondered if the cleptoparasites found on a given site would be good predictors of the species of mining bee that should be present. Clearly the cleptoparasites depend on the host bees but the converse is not true – the mining bees will indicate which cleptoparasites could be there but not which should be there. As a model I decided to use Nomada bees which mainly parasitise the nests of Andrena mining bees and occasionally Melitta species. Nomada are particularly useful because many are black and yellow and thus easily seen and collected whilst many Andrena are small or brown and camouflaged or both and hence are not so readily seen.

The initial port of call was to look up the host Andrena(s) for each Nomada species. This provided the first hitch. The information for many species is difficult to find, even on the internet, and when data are available there seems to be considerable uncertainty. In many cases the nomadas are noted as parasitizing a couple or so species definitely and a few others probably or possibly. There is a good area of research here for anybody with the time and patience!

To narrow down the predictions of which species should be present a number of other factors had to be considered. Of these, the most important and obvious was the habitat and the presence of flowers on which the mining bees are known to forage. Andrena lapponica, for example, is known to forage on bilberry hence it will not be found on sites without this plant. At the moment its only known Worcestershire location is the Wyre Forest though I have hopes of finding it elsewhere e.g. on the Malvern Hills where there are extensive patches of bilberry. However, other factors to be considered are climate and the geographical distribution of the species in question. The bilberry on the Malverns is exposed to the elements at altitude and this might not suit A. lapponica – certainly I have yet found it despite some searches. Habitat too is important as many andrenas are fairly restricted in their habitat choice. For instance, the common A. chrysosceles is found in woodland clearings and along rides whilst A. fuscipes is an ericaceous heath specialist.

It would be tedious to list the results for all 47 sites for which I analysed the data but the results for a few illustrate the predictive value of cleptoparasites and the difficulty in accumulating data. Some sites have received a lot of attention and are well recorded but many others have received scant attention despite their potential for hymenoptera (and other species!). Again it would be tedious to list all the sites but it is worth noting those for which fairly comprehensive lists are available. These are Beckford Gravel Pit, Rosemary Winnall’s land at Bliss Gate, Caddicroft Farm near Drakes Broughton, meadows at Drakes Broughton, Hartlebury Common, Kemerton Lake reserve, Kempsey Common, Rough Hill Orchard near Defford, Tiddesley Wood near Pershore, Windmill Hill near the Littletons, and Wyre Forest. The sharp-eyed might notice the absence of Devil’s Spittleful but I will come back to that later. I will not list the other sites but I have selected Burlish Top and Grafton Wood to illustrate those with intermediate levels of recording and Burley Dene Meadows and the Knapp and Paper Mill Reserve (surprisingly) as ones with poor levels.

Nine species of Nomada bees and twenty four species of Andrena have been recorded at Hartlebury Common. However, four of the Andrena have not been recorded there since the mid-nineteen nineties. Comparison of the lists for the two groups shows that potentially a further seven species of Andrena should be there, these being Andrena carantonica, A. thoracica, A. clarkella, A. apicata, A. minutula, A. semilaevis and A. subopaca. A. carantonica is accorded universal status; A thoracica is a heathland specialist; A. clarkella and A. apicata (Na) are early species mainly associated with sallow; and A. minutula, A. semilaevis and A. subopaca are very small species and are the hosts for Nomada flavoguttata. Despite the lack of records for the last three species N. flavoguttata is present.

Tiddesley Wood has records of eight species of nomada and fifteen species of Andrena. Nonetheless a further six Andrena species should be found and two more, A. apicata (Nb) and A. varians (Nb) could also be present. The six relatively common species indicated by the Nomada cleptoparasites are A. nigroaenea, A. angustior, A. flavipes, A. semilaevis, A. clarkella and A. helvola. A. nigroaenea, A. angustior and A. varians and all are widely distributed and polylectic. As noted above, A. clarkella and A. apicata are early species on sallow whilst A. semilaevis is common and particularly associated with hogweed. The last of the expected species, A. helvola, is the only woodland specialist.

Burlish top, an intermediately recorded site, has three species of Nomada and four species of Andrena. One Nomada, N. flavopicta, parasitises Melitta species rather than Andrena and a suitable host, Melitta leporina, has been recorded. Undoubtedly more Nomada remain to found but even with those already noted another four Andrena could be present. These are A. bimaculata, A. tibialis, A. fuscipes and A. denticulata. A. bimaculata and A. tibialis are uncommon (Nb and Na respectively) but A. bimaculata is relatively common and seems to favour sandy soils of the type found on Burlish Top. A. fuscipes is a fairly common heather heath specialist and A. denticulata is a widespread but local species foraging on yellow asteraceae.

There are records of four Nomada species at Grafton Wood but only two records of Andrena. Potentially A. carantonica, A. nigroaenea, A. minutula, A. semilaevis, A. subopaca, A. bicolour and A. angustior could be present and undoubtedly some of these are, to support the Nomada population.

The poorly recorded sites have few records either of Nomada or Andrena species. Burley Dene Meadows has received one recording day visit and the only cleptoparasite noted was Nomada flava. The hosts for this species are A. carantonica and A. nigroaenea. A. carantonica was found but A. nigroaenea might also be there. The variety of habitats at Knapp and Papermill should have a wide range of mining bees and their parasites. However, only three species of Nomada, N. flava, N. favoguttata and N. marshamella have been recorded accompanied by four Andrena species, A. helvola, A. flavipes, A. fulva and A. haemorrhoa. Based only on the limited range of cleptoparasites found the Andrena mining bee list could be extended to include A. carantonica, A. nigroaenea, A. minutula, A. semilaevis, A. subopaca and A. trimmerana. These and most of the other under recorded sites would amply repay further surveys.

Devil’s Spittleful is an interesting case. As one of our premier heathland sites, and hence also one of our best sites for hymenoptera, it should be in the well recorded category with a long list of species. Indeed at first glance this would appear to be the case. However, a look at the dates of the records indicates a different picture. Several rare species of Andrena have been found there but for most of these the latest record is in the mid 1990’s when surveys were being carried out by Dr. Michael Archer and Colin Plant. Since then little work has been done, which is unfortunate as the opportunity to assess the impact of recent management work has been lost – we have no baseline.

Overall, this brief review of the Andrena mining bees and their Nomada cleptoparasites shows that using the Nomada species to identify missing Andrena is of only limited value, even on well recorded sites. This is because some Nomada bees may have several hosts and because even with sustained recording Nomada species are missed. However, it is worthwhile having some indications of species that may be present and are worth looking for.

Another question was raised by looking carefully at the Andrena/Nomada pairs. Why is one of our rarest nomad bees, N. signata (RDB2) associated with one of our commonest mining bees, A. fulva?