Small rodent dynamics at Lathkill Dale SSSI, Derbys, 1971- 2005

Numbers were estimated (May/June, November/December) by catch-mark-release live-trapping in an English ash woodland (0.45 ha) and fruit-fall (September-March/April) measured. December-March mean minimum daily temperature represented winter severity. Total marked adult mice/voles provided density indices validated against a model-generated population estimate; this allowed estimation of the structure of a time-series model and the demographic impacts of climatic/biological variables. Juveniles were similarly assessed in May/June. In two winters of insignificant fruit-fall 6.79 gm-2 edible heat-treated ash seed (as fruits) was distributed over 0.45 ha of similar nearby woodland live-trapped in parallel. Population data for bank voles Myodes glareolus and wood mice Apodemus sylvaticus were generated from live-trapping samples (90 Longworth traps at 30 (6 x 5) points in groups of 3) taken over two days of trapping (following 2 days of pre-baiting) carried out at approximately 6-monthly intervals from early December 1971 to late November 2005. Modelling: To explain between-year variations in bank vole and wood mouse reproductive and population growth rates, we applied a state-space model to 33 years of live-trapping data. Total marked adult mice/voles provided density indices validated against a model-generated population estimate; this allowed estimation of the structure of a time-series model and the demographic impacts of climatic/biological variables. Juveniles were similarly assessed in May/June. Daily Minimum Temperature records (December- March) from Buxton Met Station were supplied by BDS (contract to JRF). Mean minimum daily temperature from December to March inclusive was taken as an index of winter severity. Fruitfall of Fraxinus excelsior on the 0.45 has study area was measured annually (1971-2005) as g-2 dry weight edible seed over the period September-March or April (depending on whether fruit remained on the trees after March) by collections at about 6-week intervals. The annual total fruits collected in 20 regularly spaced dustbins (total area 3.32 m2) was divided for analysis into Sept-December and December-March/April sub-totals. Dustbins had internal conical nets, allowing the fruit to be protected from predators until removed. Fruits were separated from plant debris and edible (unparasitised)seeds were then removed from the pericarp (wing), counted, dried at 80degreesC to constant weight and combined as a total annual sample or two sub-totals as above as g (dry weight ash seed)m-2 . Experimental addition of supplementary ash fruit. From October to March 1981-82 and 1984-85, (years of negligible natural fruit-fall) 60 kg of heat-sterilized ash fruit (pericarps treated at 80 degrees C for 2 days) were distributed in the same monthly proportions observed in 1966-7 (Gardner 1977). Fruits were scattered fortnightly at the 5 x 4 (20) intermediate points 7.5 m from the nearest trapping points within a 6 x 5 point 0.45 ha grid, similar to, and circa 150m west of, the main (control) study area. Assuming the fruits were effectively scattered they provided 6.79 gm-2 of edible ash seed, just below the two highest fruitings recorded from 1971-81. This experimental grid was live-trapped at the same intervals as the main study area from December 1981 to June 1985 so that the dynamics with additional ash fruit (1981-82, 1984-85) and without it (1982-83, 1983-84) could be compared with the dynamics on the main (control) study area.