Conservation Evidence strives to be as useful to conservationists as possible. Please take our survey to help the team improve our resource.

Providing evidence to improve practice

Individual study: Early responses of the tree and field layer vegetation to pollarding and mowing during restoration of the elm forest of Loi, Sogn og Fjordane, southwest Norway

Published source details

Austad I. & Skogen A. (1990) Restoration of a deciduous woodland in western Norway formerly used for fodder production: effects on tree canopy and field layer. Vegetatio (now Plant Ecology). Plant Ecology, 1-20


Species rich deciduous forests were formerly widespread in Norway but large areas have been lost through conversion to arable land and other utilization. The elm forest of Loi in western Norway, the site of the experimental restoration efforts described here, is an example of old high-pollarded forest of great cultural-historical and ecological interest. Traditional uses included harvesting of leaves and twigs, pollarding, and hay-making and grazing in spring and autumn. This produced open woodland with conditions suitable for light demanding grassland and forest edge species. After cessation of traditional management, the woodland became increasingly dense due to invasion by grey alder Alnus incana and other woody species, and the development of a luxuriant ground layer. Efforts to restore the former open 'meadow woodland' in part of the area were undertaken through a combination of the removal of invading understorey trees, pruning of old elm Ulmus glabra pollards, coppicing of hazel Corylus avellana, and mowing the ground vegetation.

Study area: The experimental restoration was undertaken in deciduous woodland at Loi, Inner Sogn (61º 20'N) south west Norway. This site know n as 'The elm forest of Loi' is within one of the largest remnant areas (about 3 km long and up to 600 m wide) of rich deciduous forests in Norway and is situated on a steep, boulder-strewn hillside. A representative area of about 0.3 ha of woodland was selected, dominated by A. incana, with some old U.glabra, small-leaved lime Tilia cordata and C.avellana. The whole forest was surveyed, mapped and described prior to restoration management implementation.

Restoration of old pollards: This mainly involved pollarding the old elm trees using a chain saw early in spring. The aim was to prune 20-25 cm above the last point of earlier high pollarding or lopping, but for some trees it was necessary to depart from this procedure (e.g. if boughs broken or decaying). These trees were very large and had previously been cut 8-10 m above the ground; many were extremely old and the trunk wood partly rotten. Old T.cordata pollards were also cut back. Two experienced forestry workers cut back an average of five pollarded trees a day.

Topping young trees to form pollards: As many of the existing pollards were old and decaying, it was necessary to also undertake pollarding of some younger trees to ensure gradual renewal and maintenance of an open, pollarded, tree canopy. Selected young elm trees were cut at a height of 2.5-3.5 m.

Undesirable tree and bush removal: Invading A.incana, goat willow Salix caprea and bird cherry Prunus padus trees, young saplings of rowan Sorbus aucuparia, aspen Populus tremula and newly planted Norway spruce Picea abies, were removed. C.avellana, rose Rosa spp. and gooseberry Ribes uva-crispa and saplings of U.glabra and T.cordata were retained. Old C.avellana stools were coppiced. Cut material was removed.

Mowing of the field layer: Mowing was undertaken twice in 1985 (end of June and mid-August), and likewise in 1986. In 1987 and 1988 mowing was done once at the end of September. Grey alder and bird cherry seedlings were also removed. A short-handled scythe, sickle and leaf-cutting knife were used, and cut material was removed. One person completed the mowing in about one week. Permanent plots were established to record ground flora responses to management.

Responses to pollarding and coppicing: The old pollards developed very well after pruning; elms developed vigorous bundles of new twigs at the pruning points and were in good condition. The young elm pollards, not pruned previously, and coppiced hazel, rejuvenated satisfactorily.

Removal of woody species: The regeneration of grey alder, and to some extent bird cherry, from stumps, root suckers and seeds was vigorous, thus necessary repeated cutting and removal for some years is predicted.

Ground flora: Before restoration implementation, the ground layer was dominated by a few tall perennial herbs and raspberry Rubus idaeus. In general, after an expansion of several conspicuous nitrophilic herbs (e.g. common nettle Urtica dioica, touch-me-not Impatiens noli-tangere and R.idaeus), smaller sized herbs and grasses increased in frequency and abundance. Light-demanding and thermophilous forest margin plant species became more prominent. However, initially, strong insolation in cleared areas led to severe desiccation; desirable species such as wild strawberry Fragaria vesca, common dog-violet Viola riviniana, wonder violet V.mirabilis and other low herbs dried up. Thus tree canopy thinning in the first years did not favour those species which it was intended to promote. The moss layer also gradually became more dense and varied.

The average number of species in the permanent plots nearly doubled over the study period. The removal of the previous year’s growth reduced litter cover and improved conditions for germination and establishment of new plants.

Conclusions and discussion: The results to date indicate that it is possible to restore the forest, although much work is needed for further development towards an open forest meadow, as the vegetation is still far from that of a traditional species-rich meadow woodland. The restoration of the old pollards and pollarding of young trees has been very successful, but it will be necessary to repeat this at 5 to 6 year intervals. However, the financial cost of the management is high, exacerbated by the difficult working conditions - steep slopes on boulder/stone scree.

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