The development of dragonfly communities and the consequences of territorial behaviour: a 27 year study on small ponds at Woodwalton Fen, Cambridgeshire, United Kingdom

20 small ponds were constructed at the Woodwalton Fen National Nature Reserve in 1961 and 3 elsewhere in the reserve in 1974 and 1977. The Twenty Ponds were allowed to develop naturally except for scrub control on their perimeters. In the other (Heathfield) ponds emergent plants were planted immediately after their construction. The ponds were visited near noon on fine days as often as possible and the numbers of adult and teneral dragonflies were recorded. Of the 20 spp. breeding within 6200 m of the Twenty Ponds, 6 bred in all the ponds, 4 bred irregularly, 3 rarely, 5 visited the ponds but did not breed and two were never seen on the ponds. The dragonfly community of the Twenty Ponds went through a short pioneer stage in which Ischnura elegans and Sympetrum striolatum were virtually the only visitors and breeders, then a developmental stage of 11 yrs in which 6 other spp. colonised the ponds and in which Pyrrhosoma nymphula and Brachytron pratense appeared and disappeared, and finally a climax stage, which has lasted 15 yrs, in which 5 spp. breed regularly, 1 frequently and 2 occasionally. In the 3 Heathfield ponds the pioneer stage was eliminated by planting emergent plants following construction, but in 2 of the ponds a fourth senescent stage followed the climax owing to colonisation by reeds (Phragmites australis). All 23 ponds varied considerably in their simple aquatic vegetation but their odon. faunas were very similar. No one plant sp. was necessary for larval development and most spp. bred successfully when waterweed was totally absent. The partial dependence of the odon. fauna on surrounding populations was evident, notably after the exceptional drought of 1976. Records made throughout the 27 yr period showed that the value of highest steady density for each sp. was remarkably similar for all 20 ponds. The highest steady densities of the spp. of the Twenty Ponds were also very similar to those at the Heathfield ponds and at ponds at Arne in Dorset in the 1950s. The behavioural basis of highest steady density and its ecological consequences are discussed. The implications for conservation strategy of this study on suboptimal populations are outlined.