Full text of DOI:10.1093/forestry/73.4.381

Interactions between forests and
herbivores: the role of controlled
grazing experim ents
1
2
3
A.J. HESTER , L. EDENIUS , R.M. BUTTENSCHØN AND
A.T. KUITERS4
1
MLURI, Craigiebuckler, Aberdeen AB15 8QH, Scotland
Department of Animal Ecology, SLU, S-901 83 Umeå, Sweden
Ministry of Environment and Energy, Danish Forest and Landscape Research Institute, Kvak Møllevej 31,
2
3
DK-7100 Vejle, Denmark
4
Institute for Forestry and Nature Research (IBN-DLO), PO Box 23, NL-6700 AA Wageningen,
The Netherlands
Summary
The use of controlled grazing by large herbivores as a research and management tool in forest
systems is reviewed, with particular focus on exclosure-based studies of plant–animal interactions
within north-west Europe. Controlled grazing studies have revealed that large herbivores (wild and
domestic) have a substantial influence on forest composition and dynamics. We review current
knowledge and highlight the deficiencies and potentials for new research, in relation to specific areas
identified as of key importance in forest dynamics. Few attempts have been made to assess critical
thresholds of grazing intensity in relation to factors such as recruitment and maintenance of different
tree species. Thus it is not possible to define appropriate grazing management techniques for specific
aims, yet there is an urgent need for this type of knowledge, especially in countries with little
remaining natural forest. Gradients of primary productivity have been shown to affect grazing
choice and species responses to damage in non-forest vegetation, yet few studies have explored these
effects within forest systems. More research on plant–herbivore and herbivore–herbivore interactions
is required at a range of both plant and animal densities. The increasing awareness of biodiversity
issues has also highlighted deficiencies in our knowledge of forest systems. We conclude that
controlled grazing experiments have an important role to play in these key aspects of forest research,
and that their potential has not yet been fully utilized.
Introduction
on ground vegetation, soils and other fauna, her-
bivores can profoundly affect the structure and
Large mammalian herbivores are an integral part dynamics of these systems. Throughout Europe,
of closed and open forest systems in north-west forests are currently undergoing changes in form
Europe. By direct and indirect effects on seedling and intensity of management, with a concomitant
establishment and sapling recruitment, as well as desire to examine and promote the role of large
©
Institute of Chartered Foresters, 2000
Forestry, Vol. 73, No. 4, 2000

3
82
FO R EST RY
herbivores in the maintenance of traditional data summaries (Table 2) on five tree species
woodland pasture systems as well as high forest which are common to many European forests,
(
1
Gomez-Gutierrez et al., 1998; Hester et al., namely Scots pine (Pinus sylvestris), birch (Betula
998; Papachristou, 1998; Quelch, 1999). In pendula and B. pubescens), rowan (Sorbus aucu-
addition to forest management for wood prod- paria) and oak (Q uercus robur). Throughout the
ucts and herbivore grazing, the increasing import- text, Latin names are only used on first mention
ance of biodiversity issues in natural resource of the species unless more than one species of the
management necessitates a thorough understand- same genus are referred to (e.g. Betula).
ing of the wider impacts of herbivores within
different forest types (Council of European Com-
munities, 1992; Department of the Environment,
Thresholds of grazing intensity
1
994; Danell et al., 1998; Ratcliffe, 1998).
The objective of this paper is to discuss the role As large herbivores are an integral part of many
of controlled grazing experiments in the assess- forest ecosystems, their complete removal by
ment of herbivore–plant interactions, in particu- fencing is now considered undesirable in many
lar the effects of large herbivores on natural forests, except where used as a short-term
establishment, growth and performance of measure in heavily degraded or deforested areas
seedlings and saplings, as well as wider biodiver- (Mitchell and Kirby, 1990; Kirby et al., 1994;
sity issues. In this paper, we use the term ‘exclo- Putman, 1996; Buttenschøn, 1998; Kuiters,
sures’ to mean fenced areas within which 1998), or as protection for high output timber
herbivores have been controlled in some way (e.g. production (Gill, 1992b; Faber and Edenius,
density, season), or have been completely 1998; Hester et al., 1998; Ratcliffe, 1998). Thus
removed. We use the term ‘forest’ in a broad an understanding of the impacts of different
sense, encompassing closed and semi-open forest grazing intensities, rather than simply the pres-
as well as areas with scattered or fragmented tree ence or absence of herbivores, is crucial (Mitchell
coverage that may cover wide areas, often with and Kirby, 1990; Reimoser, 1998; Suchant et al.,
great potential for forest development under 1998).
appropriate management. We use a ‘key topics’
Results of a wide range of studies, including
framework to highlight those areas which we our case studies, indicate that herbivore density
consider, from our own experience, to be of prime and grazing impact is not linearly related (Beau-
importance in forest grazing research and mont et al., 1994; Staines et al., 1995; Stewart,
management, but which have either not been suf- 1996; Jorritsma et al., 1999), and that certain
ficiently well addressed to date or may be less thresholds of herbivore densities exist with
amenable to exclosure studies alone. The follow- respect to, for example, sapling recruitment
ing topics are addressed: (1) thresholds of grazing and/or sapling survival. Different thresholds also
intensity, (2) effects of herbivory in relation to appear to exist for different tree species as well as
primary productivity, (3) plant–herbivore den- different stages of their growth. In the Scottish
sities and timing of herbivory, (4) interactions uplands, for example, monitoring of the impacts
between different herbivore species, and (5) of reduced deer densities has revealed marked
impacts on biodiversity and whole-system increases in sapling recruitment (i.e. growth
dynamics.
above ground vegetation height) once deer densi-
–2
Each key topic is examined in relation to pub- ties have fallen below about 5 km , with pine
lished literature and data from four selected case- responding sooner than deciduous species (Beau-
studies within natural and semi-natural forest mont et al., 1994; Stewart, 1996). In the UK case-
areas in Denmark, the UK, the Netherlands and study (Tables 1 and 2), sapling recruitment of
Sweden. Forests in north-west Europe are rela- both rowan and birch was only suppressed above
tively diverse, differing in structure, area, cultural about 1 sheep per ha. In the Denmark case-study
history and grazing regime, and our selected (Tables 1 and 2), sapling recruitment of all species
studies represent a subset of forest types and her- except pine was suppressed at high cattle densi-
bivore species (Table 1). To facilitate comparison ties, but only oak and rowan were suppressed at
between the four studies, we have focused our low cattle densities. Results from the New Forest

IN T ER AC T IO N S BET W EEN FO R EST S AN D H ER BIVO R ES
383

3
84
FO R EST RY

IN T ER AC T IO N S BET W EEN FO R EST S AN D H ER BIVO R ES
385
(
Peterken and Tubbs, 1965) also indicated to manipulate, but numbers can be regulated by
narrow thresholds of grazing pressure for the culling to examine the effects of changes in den-
regeneration of oak. Such impacts on sapling sities in specific areas (e.g. Beaumont et al., 1994;
recruitment contrast with the generally positive Stewart, 1996).
impacts of herbivores on tree seedling establish-
ment (e.g. Table 2, Denmark and UK).
Identification of such thresholds is crucial for
Effects of herbivory in relation to primary
the successful management of herbivores within
productivity
forest systems, but many other factors also influ-
ence density–regeneration relationships, and indi- A review of grassland exclosure studies encom-
vidual site information is also crucial to establish passing a range of nutrient conditions (Milchunas
what these might be (Frelich and Lorimer, 1985; and Lauenroth, 1993) showed that the effects of
Huntly, 1991; Hobbs, 1996; Stewart and Hester, herbivory on plant species composition are con-
1
998). The Swedish case-study, with the widest ditional upon both grazing history and primary
range of sites, illustrates this particularly well, productivity, with the strongest effects (e.g. plant
whereby no overall density–regeneration species changes) often occurring on the most pro-
relationships were found due to variation in ductive sites (see also Olff and Richie, 1998).
factors such as previous disturbance (forestry or Whilst sites of moderate to nutrient-poor quality
fire), availability of other forage, or tree seed have been relatively well studied in forest grazing
sources (cf. the Denmark case-study).
research (Humphrey, 1998a; Table 1), herbivore
To summarize, grazing per se does not gener- impacts have more rarely been assessed at pro-
ally limit regeneration provided that sufficient ductive sites. Saplings growing in more produc-
seed trees are available (cf. Mitchell and Kirby, tive soils tend to tolerate much more severe
1
990). However, sapling recruitment, especially browsing than those growing in low-nutrient soils
of preferred tree species, is badly affected by (McNaughton, 1983; Danell et al., 1991a;
browsing such that even low herbivore densities Edenius et al., 1995a). On the other hand, plants
may be sufficient to prevent grazing-sensitive tree containing higher levels of nutrients are generally
species from attaining canopy status. Indications preferred by herbivores (Mitchell et al., 1977;
of herbivore density–regeneration relationships Coley et al., 1985; Danell et al., 1991b; Arm-
can be drawn from specific studies (e.g. Table 1), strong and Milne, 1995). Thus, heavier grazing
but we are still unable to predict thresholds of pressure may offset the greater resilience to
grazing intensity across different forest systems. damage of those saplings growing in higher nutri-
This problem has been further exacerbated by the ent soils. Furthermore, the relationships between
location of many grazing exclosures to date in plant ‘quality’ and accessibility are also import-
areas with high herbivore densities or in areas ant in forest systems (Danell et al., 1991b). For
particularly vulnerable to damage (e.g. Mladenoff example, during severe snow conditions in boreal
and Stearns, 1993), presumably because exclo- forests, moose tend to be spatially aggregated in
sures have generally been erected with the implicit winter ‘yards’ which are often located on nutri-
objective of demonstrating herbivore impacts! ent-poor riverine flood plains. Food plant selec-
Furthermore, the ‘impacts’ of herbivore removal tion is thus determined more by shelter and food
have often been inferred against uncon- accessibility rather than its inherent quality,
trolled/unrecorded background levels of her- resulting in greater damage to species such as
bivory (exclosure database: Humphrey, 1998a). Scots pine than would be expected on purely food
In summary, there is an urgent need for more quality grounds (Danell et al., 1991b; Edenius et
studies using controlled (and particularly low) al., 1995a; Table 2). Similar findings have been
herbivore densities rather than simply herbivore directly or indirectly reported elsewhere (e.g.
exclusion. This is relatively straightforward for Staines, 1976; Mitchell et al., 1977; Gill, 1992a;
domestic herbivores (e.g. Buttenschøn and But- Stewart, 1996).
tenschøn, 1985; Hester et al., 1996; Table 1),
There is clearly a need for more controlled
where season of grazing as well as density can grazing research encompassing a wider range of
easily be controlled. Wild herbivores are less easy nutrient gradients than has previously been

3
86
FO R EST RY
studied. Exclosures along productivity gradients certain density of seedlings, indicating a threshold
can also be used to analyse the dynamic relation- limit in the balance between available browse
ships between plant productivity and herbivory, a (seedlings) and browse offtake by cattle. This may
subject which is poorly understood in many Euro- well drive the direction of longer-term grazing
pean forest systems (e.g. Mitchell et al., 1996; impacts on this system (Buttenschøn, 1988; R.M.
Gough and Grace, 1998).
Buttenschøn and J. Buttenschøn, unpublished).
Herbivore–sapling density relationships have
also been examined in a range of models. The
model FORGRA, for example, developed in
Wageningen, uses scenario analysis to explore
the potential longer term impacts of her-
Plant–herbivore densities and timing of
herbivory
Models of plant–animal interactions predict the bivore–sapling density relationships on forest
occurrence of multiple stable states in dynamics on sandy soils (Jorritsma et al., 1999).
plant–animal dynamics, dependent on initial For the Isle Royale system in USA, coupled diffu-
plant density, density of animals, herbivore per sion models have been used to create novel and
capita consumption rate, and plant growth rate informative descriptions of spatially oscillatory
(
Noy-Meir, 1975). Probably the most compre- patterns of three-way moose–vegetation–soil
hensive studies of these interactions to date relate interactions (Pastor et al., 1997). Multi-way
to moose (Alces alces), predominantly in the USA plant–herbivore interactions of this sort are
and in Sweden. At Isle Royale, USA, suppression clearly of major importance in forest systems at
of balsam fir (Abies balsamea) recruitment by all levels of study, and there is great potential to
moose was found at several combinations of fir extend this research and modelling to other forest
and moose densities (Brandner et al., 1990). Simi- types and herbivore species (see Frank and Groff-
larly, in Sweden, intense selective browsing by man, 1998).
moose is thought to have a region-wide detri-
mental effect on recruitment rates of aspen
Timing of grazing
(Populus tremula), rowan and goat willow (Salix
caprea) to the forest canopy, which in turn affects Timing of grazing, as well as herbivore–sapling
subsequent tree densities, forest composition and density per se, can critically alter the balance of
moose densities (Angelstam, 1996). In our forest–herbivore dynamics, influencing not only
Netherlands case-study, differential browsing of sapling selection by herbivores, but also the
pine, beech, oak and silver birch also significantly physical and chemical responses of different
altered the densities and species composition of sapling species to that damage (Miller et al.,
regenerating saplings in different areas (Tables 1 1982; Danell et al., 1994; Hester et al., 1996;
and 2; Van Hees et al., 1996). Using exclosures, Stewart and Hester, 1998; R.M. Buttenschøn and
Augustine et al. (1998) provided evidence for two J. Buttenschøn, ongoing study; Table 2). Timing
alternate stable states in white-tailed deer of browsing of different species is also affected by
(
O docoileus
canadensis) interactions at high deer density, Miller and Cummins, 1982; Mitchell et al., 1996;
which were dependent on initial herb density. Stewart, 1996; Miller et al., 1999). Controlled
virginianus)–forb
(Laportea the presence/absence of alternative forage (e.g.
Major impacts of herbivores on forest canopy grazing experiments provide ideal tools to
composition can also be inferred from studies address these issues, and are particularly amen-
elsewhere in Europe (e.g. McVean and Ratcliffe, able to domestic herbivore manipulation, as illus-
1
1
962; Buttenschøn, 1988, 1998; Beaumont et al., trated by our UK case-study (Tables 1 and 2;
994; Stewart, 1996; Table 2). The influence of Hester et al., 1996; Mitchell et al., 1996). In
tree density on browsing patterns has also been relation to wild herbivores, many species are
demonstrated, but by a much more limited range known to use different parts of their range at
of studies (e.g. Andren and Angelstam, 1993; different times of the year and forest areas used
Shipley and Spalinger, 1995). In our Denmark in winter can have very different patterns of her-
case-study (Tables 1 and 2), sapling recruitment bivore impact than those used in summer (Danell
in the presence of cattle only occurred above a et al., 1991b; Edenius et al., 1995a; Stewart and

IN T ER AC T IO N S BET W EEN FO R EST S AN D H ER BIVO R ES
387
Hester, 1998). Judicious use of seasonal exclo- Furthermore, grazing of herbaceous vegetation
sures in studies of wide-ranging wild herbivores also affects conditions for seedling establishment,
may help to uncouple the various factors con- as well as competition between herbaceous and
trolling herbivore selection, such as shelter woody vegetation, both of which in turn affect
seeking, predator avoidance, as well as forage food recruitment/growth for browsers (e.g.
choice.
To summarize, species–density–timing interac-
Austin et al., 1994; Jorritsma et al., 1999).
Major changes in grazer–browser ratios within
tions do have significant effects on the dynamics a forest system will inevitably alter such processes
and species composition of developing forest of facilitation or inhibition by one herbivore
areas, but we cannot yet make predictions about species on another. In a study by Austin et al.
the importance of these interactions across a (1994) in North America, for example, spring
range of different forest types. More factorial- and early summer grazing by horses led to
type controlled grazing studies are needed, at enhanced shrub growth which in turn facilitated
different plant and animal densities and times of winter browse activity for mule deer (O docoileus
year, to enable us to understand and predict hemionus) (cf. Kie, 1996). In the Netherlands,
plant–herbivore density relationships for a range grazing by large herbivores (cattle, horses) has
of situations. This can feed into the development been shown to facilitate grazing by rabbits (P.A.
of predictive models for a range of forest systems Slim and A.T. Kuiters, unpublished data); the
(
Humphrey, 1998b), for use in hypothesis rabbits grazed small seedlings and thus prevented
development and testing as well as contributing the succession towards woody species in that
to practical decision support tools.
area. Interactions between different functional
types of herbivores can thus have significant
impacts on forest dynamics, but these impacts
have seldom been directly or systematically
addressed (Mitchell and Kirby, 1990; Young et
al., 1997; Table 1).
Interactions between different herbivore
species
Different herbivore species generally co-exist in
In some parts of Europe, browsers (e.g. deer)
forest ecosystems and often overlap in their diet are now replacing grazers (e.g. sheep, cattle) as
selection. Nevertheless, large herbivores can be the dominant herbivores (Bradshaw and Edenius,
ordered along a continuum of functional types 1998; Reimoser, 1998), and changes in the Euro-
from ‘grazers’ to ‘browsers’, according to food pean Common Agricultural Policy could result in
plants selected, diet composition and digestive this phenomenon becoming more widespread if
anatomy (Hoffman, 1985; Bodmer, 1990). animal headage-based financial support is
Although browsers by definition consume woody removed. Exclosure-based experiments allow
plants, grazers also consume some woody plant manipulation of absolute and relative herbivore
material. The main differences are that: (1) densities to examine the interactions between
browsers actively select woody material, whereas different grazing animals and grazing regimes,
grazers may eat it as part of the sward (seedlings) whilst controlling other variables. Exclosures
or sample it when it is conspicuous, e.g. saplings with semi-permeable fences in nested designs can
at low densities; and (2) grazers seem to be less readily be used to address the relative impacts of
selective of particular woody species than large and small herbivores within the same
browsers. From a functional point of view, in this experiment (e.g. Jefferies et al., 1994; Nolan et
respect seedlings are defined as those individuals al., 1994).
occurring in the herbage/grass layer height strata.
Therefore at the grazer end of the browser–grazer
spectrum, in view of the greater influence of
Impacts on biodiversity and whole-system
herbaceous vegetation on grazing/browsing pat-
dynamics
terns, saplings growing within a preferred vege-
tation type may be more heavily grazed simply Many exclosure studies to date have focused on
because of their location within this vegetation specific aspects of plant–animal interactions, e.g.
type (e.g. Mitchell et al., 1996; Stewart, 1996). regeneration/growth of specific plant species

3
88
FO R EST RY
(
exclosure database: Humphrey, 1998a). Wider heterogeneity and creating and maintaining
ecosystem effects, such as alteration of nutrient spatially shifting mosaics of different forest suc-
cycling regimes induced by changes in plant cessional stages (Mitchell and Kirby, 1990;
species composition due to grazing, have only Kuiters et al., 1996; Reimoser, 1998; Suchant et
relatively recently received direct attention al., 1998). However, more quantitatively based
(
Pastor et al., 1988; Pastor and Naiman, 1992; definitions of this concept of ‘moderate’ are
Hobbs, 1996; Suominen et al., 1998). The magni- needed for practical application to different forest
tude of these wider impacts will vary between systems (Humphrey, 1998b).
forest types, depending on the strength and direc-
tions of the feedbacks in the particular system.
Non-feeding impacts of herbivores, which can
Controlled grazing and forest research: the
be of fundamental importance for other life
way forward
forms, have therefore been rather neglected in
many past studies. For example, trampling, A collation of exclosure databases in central and
fraying and dunging can radically change the north-western Europe, conducted within the EU-
internal structure and spatial heterogeneity of Concerted Action ‘Grazing as a Management
forest systems (Mitchell and Kirby, 1990; Gill, Tool’ (see Humphrey, 1998a), demonstrates the
1
992a; Pastor et al. 1997). Feeding by large her- widespread application of animal exclosures in
bivores can also directly affect other phy- forest grazing research to date, but also the limi-
tophagous organisms. For example, Danell and tations in their use so far. The database demon-
Huss-Danell (1985) found increased abundance strates in particular the preponderance of
of ants, psyllids, leaf-galls and other leaf-eating exclosure use just to remove herbivores rather
insects following simulated winter browsing by than to manipulate densities or timing of grazing.
moose on birch. Herbivore impacts on forest suc- The potential of animal exclosures to explore the
cession have already been discussed but as well as more complex or indirect effects of herbivory has
driving canopy composition, grazing within a not been fully realised to date and we have high-
forest can also drive, deflect or accelerate succes- lighted what we consider to be the main areas
sion in the ground vegetation, e.g. away from less requiring attention, for example thresholds of
grazing-tolerant species towards more grazing- herbivore density for specific end requirements.
tolerant grasses and herbs (Hester et al., 1991). Animal enclosures can be valuable tools for
Similarly, season of grazing can also affect biodi- multi-factorial experimental control of herbivore
versity; e.g. under winter grazing species can densities within different forest systems, as is
flower and set seed in the summer, but this may demonstrated by the multitude of research publi-
be prevented under moderate or heavy summer cations on moose and forest dynamics which
grazing. Seasonal grazing is already widely used illustrate the complex of interactions which can
for grassland biodiversity management, but its drive such systems. Of the other issues touched
role in forest biodiversity has not been fully upon in this paper, examination of gradients of
explored (Putman et al., 1989; Mitchell and productivity, browser–grazer interactions and
Kirby, 1990; Hester et al., 1996; Schreiner et al., biodiversity issues may be less amenable to exclo-
1
996; Olff and Ritchie, 1998).
In view of the increasing awareness of biodi- role to play and there is ample room for novel
versity issues in forest research and management, experimental design.
sure-based experiments alone, but they have a
it is important to broaden the scope of exclosure
Scale issues are of fundamental importance in
studies in the future to encompass issues such as evaluating both the direct and indirect effects of
the relationships between grazing pressure, forest herbivory, and exclosure studies should not be
dynamics and species diversity (Humphrey et al., considered the sole solution to the assessment of
1
999). Complex interactions cannot be assessed herbivore–plant interactions, but rather as com-
simply by removal of herbivores by fencing. For plementary to other research methodologies such
example, a ‘moderate’ grazing pressure is prob- as wider-scale surveys, monitoring programmes
ably most beneficial for the co-existence of many and herbivore culling programmes (e.g. Beau-
plant and animal species, by increasing structural mont et al., 1994). However, exclosures clearly

IN T ER AC T IO N S BET W EEN FO R EST S AN D H ER BIVO R ES
389
have a greater potential contribution than has
generally been realised or utilized in the past. We
conclude, therefore, with a plea for the better
integration of exclosure-based experiments
within forest research to explore both spatial and
temporal aspects of herbivory as a structuring
factor in forest systems.
species in grassland conservation areas. Aspects
Appl. Biol. 16, 373–381.
Buttenschøn, J. and Buttenschøn, R.M. 1978 The effect
of browsing by cattle and sheep on trees and bushes.
Nat. Jutlandica 20, 79–94.
Buttenschøn, J. and Buttenschøn, R.M., 1985 Grazing
experiments with cattle and sheep on nutrient poor,
acidic grassland and heath: IV. Establishment of
woody species. Nat. Jutlandica 21, 117–140.
Buttenschøn, R.M. 1998 Forest grazing in Denmark:
past, present and future. In Grazing as a Manage-
ment Tool in European Forest Ecosystems. J.
Humphrey, R. Gill and J. Claridge (eds). Forestry
Commission Technical Paper No. 25, 20–23.
Coley, P.D., Bryant, J.P. and Chapin, F.S., III 1985
Resource availability and plant herbivore defense.
Science 230, 895–899.
Acknowledgements
We thank David Henderson and John Milne for com-
ments on the manuscript, and Kjell Danell for sharing
unpublished data.
References
Council of European Communities 1992 Directive
Angelstam, P. 1996 The ghost of forest past – natural
disturbance as a basis for reconstruction of biologi-
cally diverse forests in Europe. In Conservation of
Faunal Diversity in Forested Landscapes. R.M.
DeGraaf and R.I. Miller (eds). Chapman and Hall,
London, 287–337.
Andren, H. and Angelstam, P. 1993 Moose browsing on
Scots pine in relation to stand size and distance to
forest edge. J. Appl. Ecol. 30, 133–142.
Armstrong, H.M., and Milne, J.A. 1995 The effects of
grazing on vegetation species composition. In Heaths
and Moorland: Cultural Landscapes. D.B.A. Thomp-
son, A.J. Hester and M.B. Usher (eds). HMSO, Edin-
burgh, 162–173.
Augustine, D.J., Frelich, L.E., and Jordan, P.A. 1998
Evidence for two alternate stable states in an ungu-
late grazing system. Ecol. Appl. 8, 1260–1269.
Austin, D.D., Urness, P.J. and Durham, S.L. 1994
Impact of mule deer and horse grazing on trans-
planted shrubs for regeneration. J. Range Manage.
9
2/43/EEC on the Conservation of Natural Habitats
and of Wild Fauna and Flora. ‘The Habitats Direc-
tive’. EC, Brussels.
Danell, K. and Huss-Danell, K. 1985 Feeding by insects
and hares on birches earlier affected by moose. O ikos
4
4, 75–81.
Danell, K., Hjälten, J., Ericson, L. and Elmqvist, T.
991a Vole feeding on male and female willow
shoots along a gradient of plant productivity. O ikos
2, 145–152.
Danell, K., Niemela, P., Varvikko, T. and Vuorisala, T.
991b Moose browsing on Scots pine along a gradi-
1
6
1
ent of plant productivity. Ecology 72, 1624–1633.
Danell, K., Bergström, R. and Edenius, L. 1994 Effects
of large mammalian browsers on architecture,
biomass, and nutrients of woody plants. J. Mammal.
7
5, 833–844.
Danell, K., Willebrand, T. and Baskin, L. 1998
Mammalian herbivores in the boreal forests: their
numerical fluctuations and use by man. Conserv.
4
7, 8–11.
Beaumont, D., Dugan, D., Evans, G. and Taylor, S.
994 Deer management and tree regeneration in the
Ecol. 2,
9
(Available from the Internet:
http://www.consecol.org/vol2/iss2/art9.).
Department of the Environment 1994 Biodiversity: the
UK Action Plan. HMSO, London.
Edenius, L., Danell, K. and Nyquist, H. 1995a Effects
of simulated moose browsing on growth, mortality,
and fecundity in Scots pine. Can. J. For. Res. 25,
1
RSPB reserve at Abernethy Forest. In O ur Pinewood
Heritage. J.R. Aldhous (ed.). Forestry Commission,
Edinburgh, 186–195.
Bodmer, R.E. 1990 Ungulate frugivores and the
browser–grazer continuum. O ikos 57, 319–325.
Bradshaw, R.H.W. and Edenius, L. 1998 The
Fennoscandian perspective: grazing in boreal, hemi-
boreal and nemoral forest. In Grazing as a Manage-
ment Tool in European Forest Ecosystems. J.
Humphrey, R. Gill and J. Claridge (eds). Forestry
Commission Technical Paper No. 25, 2–10.
5
29–535.
Edenius, L., Ball, J.P., Bergström, R., Danell, K. and
Ericson, L. 1995b Influence by large mammalian her-
bivores on the performance of boreal tree species.
Abstracts from the symposium ‘Ungulates in Tem-
perate Forest Ecosystems’, Wageningen, The Nether-
lands, 23–27 April 1995.
Faber, W.E. and Edenius, L. 1998 Bark stripping by
moose in commercial forests of Fennoscandia – a
review. Alces 34, 261–268.
Brandner, T.A., Peterson, R.O. and Risenhoover, K.L.
1
990 Balsam fir on Isle Royale: effects of moose her-
bivory and population density. Ecology 71, 155–164.
Buttenschøn, J. 1988 The establishment of woody
Frank, D.A. and Groffmann, P.M. 1998 Ungulate vs.

3
90
FO R EST RY
landscape control of soil C and N processes in grass- Huntly, N., 1991 Herbivores and the dynamics of com-
lands of Yellowstone National Park. Ecology 79,
229–2241.
munities and ecosystems. Annu. Rev. Ecol. Syst. 22,
477–503.
2
Frelich, L.E. and Lorimer, C.G. 1985 Current and pre- Jefferies, R.L., Klein, D.R. and Shaver, G.R. 1994 Verte-
dicted long-term effects of deer browsing in hemlock
brate herbivores and northern plant communities: rec-
forests in Michigan, USA. Biol. Conserv. 34, 99–120.
iprocal influences and responses. O ikos 78, 193–206.
Gill, R.M.A. 1992a A review of damage by mammals Jorritsma, I.T.M., van Hees, A.F.M. and Mohren,
on north temperate forests: 1. Deer. Forestry 65, G.M.J. 1999 Forest development in relation to ungu-
45–169. late grazing: modeling approach. For. Ecol.
Gill, R.M.A. 1992b A review of damage by mammals Manage. 120, 23–34.
on north temperate forests: 3. Impact on trees and Kie, J.G. 1996 The effects of cattle grazing on optimal
1
a
forests. Forestry 65, 363–388.
foraging in mule deer. For. Ecol. Manage. 88,
Gomez-Gutierrez, J.M., Perez-Fernandez, M. and
131–138.
Soriguer, R. 1998 Grazing and woodland manage- Kirby, K.J., Mitchell, F.J.G. and Hester, A.J. 1994 A role
ment in the Dehesas: a silvi-pastoral system in central
Spain. In Grazing as a Management Tool in Euro-
pean Forest Ecosystems. J. Humphrey, R. Gill and J.
for large herbivores (deer and domestic stock) in
nature conservation management in British semi-
natural woods. Arboric. J. 18, 381–399.
Claridge (eds). Forestry Commission Technical Paper Kuiters, A.T. 1998 Ungulates and forest management in
No. 25, 63–71.
the Netherlands. In Grazing as a Management Tool
in European Forest Ecosystems. J. Humphrey, R. Gill
and J. Claridge (eds). Forestry Commission Technical
Paper No. 25, 11–19.
Gough, L. and Grace, J.B. 1998 Herbivore effects on
plant species density at varying productivity levels.
Ecology 79, 1586–1594.
Hester, A.J., Miles, J. and Gimingham, C.H. 1991 Suc- Kuiters, A.T., Mohren, G.M.J. and Van Wieren, S.E.
cession from heather moorland to birch woodland. I.
Experimental alteration of specific environmental
conditions in the field. J. Ecol. 79, 303–315.
1996 Ungulates in temperate forest ecosystems. For.
Ecol Manage. 88, 1–5.
McNaughton, S.J. 1983 Compensatory plant growth as
a response to herbivory. O ikos 40, 329–336.
Hester, A., Mitchell, F.J.G. and Kirby, K.J. 1996 Effects
of season and intensity of sheep grazing on tree McVean, D.N., and Ratcliffe, D.A. 1962 Plant Com-
regeneration in a British upland woodland. For. Ecol.
Manage. 88, 99–106.
munities of the Scottish Highlands. HMSO, Edin-
burgh.
Hester, A.J., Kirby, K.J., Mitchell, F.J.G., Gill, R., Milchunas, D.G. and Lauenroth, W.K. 1993 Quantita-
Latham, J. and Armstrong, H. 1998 Ungulates and
forest management in the British Isles. In Grazing as
a Management Tool in European Forest Ecosystems.
tive effects of grazing on vegetation and soils over a
global range of environments. Ecol. Monogr. 63,
327–366.
J. Humphrey, R. Gill and J. Claridge (eds). Forestry Miller, G.R., and Cummins, R.P. 1982 Regeneration of
Commission Technical Paper No. 25, 24–35.
Hobbs, N.T. 1996 Modification of ecosystems by ungu-
lates. J. Wildl. Manage. 60, 695–713.
Scots pine Pinus sylvestris at a natural treeline in the
Cairngorm mountains, Scotland. Holarct. Ecol. 5,
27–34.
Hoffman, R.R. 1985 Digestive physiology of deer— Miller, G.R., Kinnaird, J.W., and Cummins, R.P. 1982
their morphophysiological specialization and adap-
tion. In Biology of Deer Production. P.F. Fennesy and
K.R. Drews (eds). R. Soc. N.Z. Bull. 22, 393–408.
Humphrey, J.W. 1998a Grazing as a management tool
in natural and semi-natural woodland ecosystems.
Liability of saplings to browsing on a red deer range
in the Scottish Highlands. J. Appl. Ecol. 19, 941–951.
Miller, G.R., Cummins, R.P. and Hester, A.J. 1999 Red
deer and woodland regeneration in the Cairngorms.
Scot. For. 52, 14–20.
Final Report to the Commission of the European Mitchell, B., Staines, B.W. and Welch, D. 1977 Ecology
Communities, Contract No. AIR3-CT94-1965. EC,
of Red Deer. Institute of Terrestrial Ecology, Cam-
Brussels.
bridge.
Humphrey, J.W. 1998b Grazing as a management tool: Mitchell, F.J.G. and Kirby, K.J. 1990 The impact of
a synthesis of common research needs. In Grazing as
large herbivores on the conservation of semi-natural
a Management Tool in European Forest Ecosystems.
woods in the British uplands. Forestry 63, 333–353.
J. Humphrey, R. Gill and J. Claridge (eds). Forestry Mitchell, F.J.G., Hester, A.J. and Kirby, K.J. 1996
Commission Technical Paper No. 25, 79–90.
Humphrey, J.W., Hawes, C., Peace, A.J., Ferris-Khan,
R. and Jukes, M.R. 1999 Relationships between
Effects of season and intensity of sheep grazing on a
British upland woodland: browsing damage to
planted saplings. Bot. J. Scot. 48, 199–207.
insect diversity and habitat characteristics in plan- Mladenoff, D.J., and Stearns, F. 1993 Eastern hemlock
tation forests. For. Ecol. Manage. 113, 11–21. regeneration and deer browsing in the Northern

IN T ER AC T IO N S BET W EEN FO R EST S AN D H ER BIVO R ES
391
Great Lakes region: a re-examination and model
simulation. Conserv. Biol. 7, 889–900.
Nolan, A.J., Still, M.J., Bell, J.S. and Gauld, J.H. 1994
The Environmentally Sensitive Areas designated in
In Grazing as a Management Tool in European
Forest Ecosystems. J. Humphrey, R. Gill and J. Clar-
idge (eds). Forestry Commission Technical Paper No.
25, 11–19.
Scotland. The Breadalbane ESA Biological Monitor- Schreiner, E.G., Krueger, K.A., Happe, P.J. and
ing Report, years one to five, 1989–1993. Macaulay
Land Use Research Institute, Aberdeen.
Noy-Meir, I. 1975 Stability of grazing systems: an appli-
cation of predator graphs. J. Ecol. 63, 459–481.
Houston, D.B. 1996 Understory patch dynamics and
ungulate herbivory in old-growth forests of Olympic
National Park, Washington. Can. J. For. Res. 26,
255–265.
Olff, H. and Ritchie, M.E. 1998 Effects of herbivores Shipley, L.A. and Spalinger, D.E. 1995 Influence of size
on grassland plant diversity. Trends Ecol. Evol. 13,
61–265.
Papachristou, T.G. 1998 Forest ecosystems of Greece
and density of browse patches on intake rate and for-
aging decisions of young moose and white tailed deer.
O ecologia 104, 112–121.
2
and their relation to grazing animals. In Grazing as a Staines, B.W. 1976 The use of natural shelter by red
Management Tool in European Forest Ecosystems. J. deer in north-east Scotland. J. Zool. 180, 1–8.
Humphrey, R. Gill and J. Claridge (eds). Forestry Staines, B.W., Balharry, R. and Welch, D. 1995 Moor-
Commission Technical Paper No. 25, 72–78.
Pastor, J. and Naiman, R.J. 1992 Selective foraging and
ecosystem processes in boreal forests. Am. Nat. 139,
land management and impacts of red deer. In Heaths
and Moorland: Cultural Landscapes. D.B.A. Thomp-
son, A.J. Hester and M.B. Usher (eds). HMSO, Edin-
burgh, pp. 294–308.
6
91–705.
Pastor, J.N., Naiman, J., Dewey, B. and Cohen, Y. 1988 Stewart, F.E. 1996 The effects of red deer on the
Moose, microbes, and the boreal forest. BioScience
8, 770–778.
Pastor, J., Moen, R.A. and Cohen, Y. 1997 Spatial het-
regeneration of upland birch woodland in the Scot-
tish highlands. Unpublished Ph.D. thesis, University
of Aberdeen.
3
erogeneities, carrying capacity, and feedbacks in Stewart, F.E. and Hester, A.J. 1998 The impact of red
animal–landscape interactions. J. Mammal. 78,
040–1052.
Peterken, G.F. and Tubbs, C.R. 1965 Woodland
regeneration in the New Forest, Hampshire since
deer on woodland and heathland dynamics in Scot-
land. In Population Ecology, Management and
Welfare of Deer: Proceedings of the Third Deer Sym-
posium Manchester 1997. C.R. Goldspink, S. King
and R.J. Putman (eds). Department of Biological Sci-
ences, Manchester Metropolitan University, pp.
54–60.
Suchant, R., Turk, S. and Roth, R. 1998 Grazing prob-
lems in Germany: balance or imbalance between
wildlife and habitat. In Grazing as a Management
Tool in European Forest Ecosystems. J. Humphrey,
R. Gill and J. Claridge (eds). Forestry Commission
Technical Paper No. 25, 36–44.
1
1
650. J. Appl. Ecol. 2, 159–170.
Putman, R.J. 1996 Ungulates in temperate forest
ecosystems: perspectives and recommendations for
future research. For. Ecol. Manage. 88, 205–214.
Putman, R.J., Edwards, P.J., Mann, J.C.E., How, R.C.
and Hill, S.D. 1989 Vegetational and faunal changes
in an area of heavily grazed woodland following
relief of grazing. Biol. Conserv. 47, 13–32.
Quelch, P. 1999 Veteran trees in Scottish wood pasture.
In Grazing Management O ptions for Native Wood- Suominen, O., Danell, K. and Bergström, R. 1998
lands. A.J. Hester, R.V. Birnie and J. Lund (eds).
Macaulay Land Use Research Institute, Aberdeen,
pp. 65–66.
Moose, trees and ground-living invertebrates: indi-
rect interactions in Swedish pine forests. O ikos 84,
215–226.
Ratcliffe, P.R. 1998 Woodland deer management: inte- Van Hees, A.F.M., Kuiters, A.T. and Slim, P.A. 1996
grating the control of their impact with multiple
objective forest management in Scotland. In Popu-
lation Ecology, Management and Welfare of Deer:
Growth and development of silver birch, pedunculate
oak and beech as affected by deer browsing. For.
Ecol. Manage. 88, 55–63.
Proceedings of the Third Deer Symposium Manches- Young, T.P., Okello, B.D., Kinyua, D. and Palmer, T.M.
ter 1997. C.R. Goldspink, S. King and R.J. Putman
eds). Department of Biological Sciences, Manchester
Metropolitan University, pp. 67–72.
1997 KLEE: a long-term multi-species herbivore
exclusion experiment in Laikipia, Kenya. Afr. J.
Range For. Sci. 14, 94–102.
(
Reimoser, F. 1998 Forests and ungulates in Austria:
problems, management strategies and research needs. Received 7 June 1999