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: Exclusion of introduced Sika deer Cervus nippon, enhances mountain beech Nothofagus solandri cliffortioides seedling growth and recruitment, Kaimanawa Ecological Region, North Island, New Zealand

Published source details

Husheer S.W., Robertson A.W., Coomes D.A. & Frampton C.M. (2005) Herbivory and plant competition reduce mountain beech seedling growth and establishment in New Zealand. Plant Ecology, 183, 245-256


The tree Nothofagus solandri is endemic to New Zealand; var. cliffortioides (mountain beech) is an upland form that grows up to the treeline. In central North Island, intense browsing by introduced sika deer Cervus nippon appears to have induced shifts in ground cover composition towards browse-tolerant, turf-forming communities (herbs, bryophytes and ferns), whilst suppressing mountain beech regeneration through browsing of seedlings. The compact, deer-induced turf communities may also prevent beech regeneration through competitive exclusion of seedlings. Two experiments were undertaken. The experiment summarised here tested whether deer could prevent mountain beech from regenerating through browsing of their seedlings, or indirectly by maintaining competitive turf communities, by monitoring the responses of naturally occurring seedlings to the removal of deer herbivory (by using exclosures).

The second experiment used manipulations on transplanted and naturally occurring mountain beech seedlings (turf-removal, nutrient addition, and clipping to simulate deer browsing) to assess the effects of browsing and competition from deer-induced turf communities (for a summary see:

Study area: The study was undertaken in the Kaimanawa Ecological Region, a series of hills and mountains in central North Island (39ºS, 176ºE), New Zealand. Tweleve study sites (average 1.8 km apart) were selected in open beech dominated forest near the treeline (1,300 - 1,430 m). Two 10 x 10 m plots were established at each site, and one of the pair was fenced to exclude deer.

Exclusion of deer from naturally occurring seedlings: Naturally occurring mountain beech seedlings were monitored for 3 years at seven of the sites, in the paired fenced and unfenced plots. Each plot was divided into 16, 2.5 x 2.5 m sub-plots (all with >10 beech seedlings). In each sub-plot the heights of all established beech seedlings (5-135 cm high) were monitored for 3 years after initial measurements in the austral summer of 1998/1999. An aluminium tag was attached to each seedling stem and its 'pull-up height' was measured (to the end of the previous season's growth). In the austral summer of 2001/2002, all living seedlings were relocated and their heights remeasured. New seedlings that had reached 5 cm or more since the 1998/1999 survey were also measured to calculate recruitment.

To determine whether the removal of deer herbivory induced compositional changes in turf and shrub communities, all vascular plants, bryophytes and lichens growing in the soil were identified, and estimates of cover were made for each sub-plot, in two tiers (<15 cm; 15 -135 cm). Percent bare ground was also estimated.

Comparability of fenced and unfenced plots for key environmental characteristics measured in 2001/02 showed that paired plots had similar aspect, slope, total transmitted light, plant ground (<5 cm tier) and bare-ground cover, understorey composition and soil productivity.

In 1998 prior to fencing, nearly all mountain beech seedlings had been browsed in all plots (97%). Initial seedling counts were similar in fenced and unfenced plots. Beech seedlings continued to be browsed and their growth retarded by deer between 1998 and 2002. In unfenced plots, 408 browsed established seedlings (5 -135 cm) had significantly lower relative growth rates (average RGR 0.01) than the 23 seedlings that showed no signs of browsing (average RGR 0.05), suggesting that deer browsing was inhibiting growth. Established seedlings (5 -135 cm) in the exclosures had higher average RGRs than those in unfenced plots However, there was no significant difference in mortality or survivorship of established seedlings between fenced and unfenced plots.

Deer browsing also inhibited recruitment of small seedlings (<5 cm high); in 2001 there were many more small seedlings in fenced plots (average 196) compared to unfenced plots (average 17).

There was no obvious compositional response of turf communities following deer exclusion over the study period.

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