Responses of ground flora to coppice management in an English woodland – a study using permanent quadrats

Summary

Coppicing (cutting of certain tree species to near ground level every few years) used to be a main method of management of English deciduous woodlands. This traditional coppice management may be greatly beneficial to woodland ground flora which often exhibits a large positive response to the cyclical increase in light, and also to invertebrates which rely on such plants or require a more open woodland structure with open sunny glades. However, coppicing rapidly declined drastically during the 20th century, as a consequence of which many woodlands developed a closed canopy structure with greatly reduced light levels. Coppicing is now being reinstated in many woodlands managed for wildlife conservation, but there is little published evidence on the impact of coppicing on the ground flora and there is very little long-term monitoring.

In this study, changes in the ground flora in sweet chestnut Castanea sativa woodland in southeast England were studied over 11 years.

Study area: The study was conducted from 1986 to1997 in Stour Wood (National Grid ref. TM 190313), Essex, southeast England. The wood (58 ha) is a nature reserve owned by the Woodland Trust, and managed by the RSPB (Royal Society for the Protection of Birds). It is dominated by sweet chestnut coppice (with some areas of hazel Corylus avellana and small-leaved lime Tilia cordata), with oak standards, mainly Quercus robur with some Q. petraea. Permanent quadrats were established in February 1986 on three sites:

A - 0.73 ha, coppiced in November 1985;

B - 0.56 ha of overgrown coppice, cut in November 1988;

C - 0.53 ha with most trees and shrubs cut and stumps killed in the winter of 1985–1986 to widen a ride (running in a north–south direction) to about 30 m. Regrowth was periodically cut back.

Ground flora monitoring: Twenty randomly placed 0.5 m x 0.5 m quadrats (divided by a grid into 25, 10 cm x 10 cm squares) were located in each site. The number of squares in which each species were present was recorded to give a measure of percentage frequency, and the number of flowers of each species was counted. Surveys were made twice a month from March to July 1986–1989, thereafter monthly.

Light and temperature: Incident light radiation on the woodland floor in the quadrat was measured mid-morning monthly from 1986 to 1988 inclusive, thereafter in July each year; temperature was measured in the soil surface monthly from 1986 to 1989, thereafter in early August each year.

Temperature: Soil temperatures ranged from 0.75ºC (February) to 17.7ºC (August). Summer temperatures were generally 3–4ºC higher in the ride than in the coppice plots in the first 6 years of the study. They became up to 10ºC higher in the hotter summers of the last 5 years, when temperatures in the later, closed stages of the coppice cycle were also higher than in the earlier stages.

Light: Light reaching the woodland floor in summer fell to less than 1% of incident radiation in later years of the coppice cycle.

Species richness: Site A species richness peaked in the third spring following coppicing (38 species), then declined rapidly until the fifth year, remaining relatively stable for 4 years, before declining again (to 11 species) in year 12; Site B peaked in the second year after coppicing (25 species), then declined steadily; Site C peaked in the second year after widening, declining slowly after the fourth year.

Species frequency: At all sites percentage frequency peaked in the third year following coppicing. At site A cover fell sharply in the fifth year and more slowly thereafter, being 51% of the peak frequency in year 12. At site B there was a steady decline from the peak frequency to 39% in year 12. After an initial decline from the peak frequency in site C, there were 5 years of relative stability, and was 54% of the peak frequency in year 12. At all sites wood anemone Anemone nemorosa was the most frequent species and was little influenced by the coppice cycle (site A above 90% except in 1996, when 77%; site B ranged from 62 to 70%; site C 68–88%).

At site A lesser cleandine Ranunculus ficaria, moschatel Adoxa moschatellina, wood speedwell Veronica montana and enchanter's-nightshade Circaea lutetiana declined 4-5 years following coppicing. Wood spurge Euphorbia amygdaloides peaked in the third and fourth year then almost disappeared. Bramble Rubus fruticosus peaked in the fourth and fifth year, and declined from year 9. Grasses (creeping soft grass Holcus mollis and wood meadow-grass Poa nemoralis) appeared in the third year and their frequency then remained fairly constant.

At site B, honey suckle Lonicera periclymenum, R. fruticosus, wood sage Teucrium scorodonia and grasses all peaked in the second year then declined. A.moschatellina and C.lutetiana both declined after coppicing. Yellow pimpernel Lysimachia nemorum appeared in the second year but showed no clear response to coppicing.

At site C, T.scorodonia, R.fruticosus, L.nemorum, slender St. John's wort Hypericum pulchrum, soft rush Juncus effusus and grasses all peaked 3–4 years after the widening of the ride, then declined, despite no decrease in incident radiation. L.periclymenum increased steadily to year 8 and then remained relatively constant.


Note: If using or referring to this published study, please read and quote the original paper, this can be viewed at:

http://springerlink.metapress.com/content/m61n804655jj2vh6/fulltext.pdf