P. Dennis et al., THE RESPONSE OF EPIGEAL BEETLES (COL, CARABIDAE, STAPHYLINIDAE) TO VARIED GRAZING REGIMES ON UPLAND NARDUS-STRICTA GRASSLANDS, Journal of Applied Ecology, 34(2), 1997, pp. 433-443
1. The effect of different livestock grazing regimes on the insect fau
na of an upland, semi-natural grassland was measured in 1993 and 1994
by a survey of the epigeal Carabidae and Staphylinidae within an exper
iment established in 1991. Grazing by sheep, or sheep and cattle, to a
chieve two different inter-tussock sward heights, provided four treatm
ents. In addition, a further treatment was ungrazed from 1992 to test
the impact on beetles of a short-term cessation of grazing, trampling
and dung inputs. 2. Pitfall traps sampled Carabidae and Staphylinidae
within the Nardus stricta-dominated grassland of the experiment. Data
on these epigeal Coleoptera were collected from April to October in 19
93 and 1994; years three and four of the experiment. 3. The epigeal Co
leoptera species were ranked by decreasing abundance in traps, where t
he captures in traps were accumulated for both seasons. The responses
to the grazing regimes were analyzed using ANOVA, applied to the most
abundant species (that together represented 99% of the two seasons' ca
tch). There were significant experimental effects of grazing regime on
five of these 32 Coleoptera species, namely Carabus violaceus, Othius
angustus, Pterostichus strenuus, Xantholinus linearis and Olophrum pi
ceum. 4. The ordination technique, Canonical Correspondence Analysis (
CCA), was applied to the data on the Coleoptera assemblage. Variables
measured to represent the experimental treatments (mean vegetation hei
ght, stocking rate and botanical diversity) and environmental covariab
les (altitude and aspect) were entered in the direct gradient analysis
procedure of CCA. This application of CCA partialled out the effects
of altitude and aspect of each plot and revealed the significant effec
ts of vegetation structure, botanical species composition and stocking
density on a larger number of Coleoptera species than suggested from
ANOVA. 5. Twenty-four of the 32 most abundant Coleoptera species corre
lated with the effects of different grazing regimens imposed on Nardus
grassland. Greater abundances of C. violaceus, O. angustus, X. linear
is and T. corticinus were indicative of the typical upland grassland a
nd heathland Coleoptera assemblage. These species could be monitored t
o balance the impact of grazing management on arthropod biodiversity w
ith the need to restrict the dominance of N. stricta in drier upland g
rasslands, achieved in this instance, by summer grazing sheep and catt
le to maintain an average, between-tussock sward height of 6-7 cm. How
ever, the results from the direct gradient analysis suggest that the g
razing regimes should be varied in rotation over time to achieve a mos
aic of structurally different grassland patches (0.70-4.73 ha) because
this encourages a larger overall number of beetle species.