Utilisation of Molinia caerulea Tussocks by Overwintering Beetles at Epping Forest 115
The unflooded heath tussocks had 11 species including the only curculionid. Atheta spp., T.
dermestoides and Agonum obscurum were far less abundant than in site 1, whilst Agonum fuliginosum
occurred in greater numbers. Anthobium unicolor, Pterostichus rhaeticus and Stenus providus,
which occurred in low numbers in site 1, were absent from site 2.
The coleopterous fauna of the flooded tussocks (site 3) was distinctive in that carabids were all but
absent ( 1 individual P. diligens).
Total numbers of each species is also given in Table 2 J. hypnorum was the most common species
with 48 individuals in total. S. providus was the rarest species in the samples being represented by
only 2 individuals.
The numbers of individuals in 4 families represented in the sample is given in Table 3. The samples
in site 2 harboured all the 4 families although had fewer carabids and staphylinids than site 1. At
sites 1 and 2 carabids were slightly more abundant than Staphylinids and at site 3 the Carabidae was
only represented by 1 individual, the beetles here were otherwise exclusively staphylinids.
Table 3. Numbers of individuals of each Family from the 3 sites
One-way Analysis of Variance and Kruskal- Wallis tests were used to compare numbers of staphylinids
between sites. For whole cores and top core sections only, both tests result in an acceptance of the
null hypothesis (P>0.05). It can be stated therefore that there is no significant difference between the
numbers of staphylinids occurring at each site.
Results of t-tests comparing mean numbers of carabids per core and top core sections in sites 1 and
2 show an acceptance of the null hypothesis in both cases. (Site 3 was not included as carabids were
represented by 1 individual only). There is therefore, no significant difference between the means of
the 2 samples (P>().()5).
The non parametric distribution of the Trixagidae warranted the use of log 10x+l transformation
and the Mann Whitney U-test. Both results concur in that there is no significant difference (P>0.05)
between mean numbers of this family per core at the sites at which it occurs.
The Product Moment Correlation and the non parametric Spearman Rank Correlation Coefficients
were used to test for relationships between numbers of individuals per core and tussock dimensions
or water content. All but one test revealed no significant correlation (at P=0.05 level). At site 1
beetle core density was negatively correlated with tussock volume estimates (r = -0.67, P>0.05). It
was not possible to determine whether this reflected higher beetle numbers in smaller tussocks or
whether the same numbers were concentrated in smaller tussocks and therefore more likely to be
sampled by coring.
Essex Naturalist (New Series) 17 (2000)