BR. J. ENT. NAT. HIS I., 7: 1994 



Table 4. Comparison of numbers of 'delicate' compared to 'robust' moths taken in yew woodland 

 and chalk grassland. Only data from those species which did not give a significant result in 

 the analyses given in Table 3 are included. 



Yew Woodland Chalk Grassland Totals 



Delicate 129 85 214 



Robust 149 194 342 



Totals 277 279 556 



Heterogeneity Chi-squared= 15.24; d.f. = 1; P< 0.001. 



light-bodied, large-wing-area species, as characterized by the geometrids. On the other 

 hand, for the more robust build characteristic of many noctuids, shelter from wind 

 may not be of such great importance. Of course, not all geometrids are 'delicate' and 

 not all noctuids 'robust'. To consider this further, all species were categorized as either 

 delicate (D) or robust (R). The categories are given in Table 2. In a small number 

 of cases the category that a species should be placed in was not obvious, and these 

 species have been omitted from the analysis. The number of moths of species 

 which showed no significant bias to either habitat in either of the analyses given in 

 Table 3, and all those taken that were not analysed due to insufficient numbers taken, 

 were totalled under the assigned classes D and R for each habitat. The results are 

 given in Table 4. A heterogeneity chi-squared test, comparing the ratios of the classes 

 between the two habitats shows that overall the 'delicate' species are taken in 

 significantly higher numbers in the yew woodland than the grassland, the reverse being 

 the case for the 'robust' species. 



Discussion 



Consideration of the species taken in large enough numbers for individual analysis, 

 suggests that, with some exceptions, those species with large wing area to body weight 

 ratios, i.e. the more delicate species, tend to be caught more commonly inside the 

 yew woodland than in the grassland. The reverse is true for the more robust species. 

 These deductions are endorsed by the general comparison between delicate and 

 robust moths (Table 4). One interpretation of these findings is that delicate, less 

 strongly flying species may habitually seek shelter from the wind in dense woodland 

 such as the yew woodland in this study. However, that is not to say avoidance of 

 wind buffeting is the only factor producing the biases observed, and it is pertinent 

 to consider each species, showing a significant bias to one habitat or the other, 

 individually. This is done in Table 5 in which, for each of the relevant species, the 

 type of habitat bias, whether they are classed as delicate or robust, their larval food 

 plant, their roosting behaviour, and a tentative deduction of the principal factor 

 influencing the habitat bias they show, is given. 



In the majority of cases seemingly sensible reasons for the habitat preferences 

 observed can be given. In some cases, such as P. rhomboidaria, Deileptenia ribeata 

 and A repandata it is probable that all three factors under consideration contribute 

 to the behaviour. For the majority of the 'delicate' species, using the yew woodland 

 as a sheltered flight corridor is probably the principal cause of the observations, and 

 in many cases over-rides larval foodplant. There are three species which buck this 

 trend. It is possible that in each of the three, Cyclophora linearia, Colostygia peetinataria 

 and Hydriomena furcata, proximity to larval foodplants takes precedence. This 

 is almost certain in the case of C. linearia, for it was taken most often in a trap 



