Remove, control or exclude vertebrate herbivores

How is the evidence assessed?
  • Effectiveness
    not assessed
  • Certainty
    not assessed
  • Harms
    not assessed

Study locations

Key messages

  • Ten studies evaluated the effects on butterflies and moths of removing, controlling or excluding vertebrate herbivores. Three studies were in the USA, two were in the UK, one was in each of Mauritius, the Netherlands, Canada and Japan, and one was a global systematic review.

COMMUNITY RESPONSE (6 STUDIES)

  • Richness/diversity (6 studies): Two of four replicated studies (including three controlled studies and one site comparison study) in the USA, Mauritius and Canada found that forest plots fenced to exclude, or reduce the density of, non-native pigs and deer (in one case along with weeding of invasive plants) had a greater species richness of butterflies and macro-moths than unfenced plots. The other two studies found that forest plots fenced to exclude elk had mixed effects on the species richness of butterflies and arthropods including moths depending on fire intensity and year. One of these studies also found that grassland plots fenced to exclude elk had a similar species richness of butterflies to unfenced plots in all years. One global systematic review found that reducing or removing grazing or browsing by wild or domestic herbivores in temperate and boreal forests did not affect the species richness of butterflies and moths.

POPULATION RESPONSE (10 STUDIES)

  • Abundance (9 studies): Five of eight studies (including five controlled studies, one before-and-after study, and two site comparison studies) in the UK, the USA, Mauritius, Canada and Japan found that forest and grassland plots fenced to exclude, or reduce the density of, deer, sheep, pigs and large herbivores (in one case along with weeding of invasive plants) had a higher abundance of butterflies, moths, caterpillars, rare macro-moths and New Forest burnet moths than unfenced plots. One of these studies also found that the total abundance of macro-moths was similar in fenced and unfenced plots. Two studies found that forest plots fenced to exclude elk had mixed effects on the abundance of butterflies and arthropods including moths depending on fire intensity and year. One of these studies also found that grassland plots fenced to exclude elk had a similar abundance of butterflies to unfenced plots in all years. The eighth study found that a forest fenced to exclude sika deer had a similar abundance of all moths, but a lower abundance of tree-feeding moths, than unfenced forest. One global systematic review found that reducing or removing grazing or browsing by wild or domestic herbivores in temperate and boreal forests increased the abundance of butterflies and moths.
  • Survival (1 study): One paired, controlled study in the Netherlands reported that all Glanville fritillary caterpillar nests survived in grassland fenced to exclude sheep, compared to 88% in a grazed area.
  • Condition (1 study): One paired, controlled study in the Netherlands found that fewer Glanville fritillary caterpillar nests were damaged in grassland fenced to exclude sheep than in a grazed area.

BEHAVIOUR (0 STUDIES)

About key messages

Key messages provide a descriptive index to studies we have found that test this intervention.

Studies are not directly comparable or of equal value. When making decisions based on this evidence, you should consider factors such as study size, study design, reported metrics and relevance of the study to your situation, rather than simply counting the number of studies that support a particular interpretation.

Supporting evidence from individual studies

  1. A replicated, paired, controlled study in 1991–1992 in eight pinewoods in the Scottish Highlands, UK (Baines et al. 1994) found that excluding red deer Cervus elaphus increased the abundance of moth and butterfly caterpillars. In plots where deer were excluded, or present at lower density, the abundance of caterpillars (30–440 individuals/plot) was higher than in forest where deer were present at higher density (10–325 individuals/plot). Two of the four most common species (July highflyer moth Hydriomena furcata and twin spot carpet Perizoma didymata) were more abundant in exclosures with no deer (July highflyer: 32–44; twin spot: 43–57 individuals/plot) than in forest with deer (July highflyer: 10–11; twin spot: 13–16 individuals/plot), but the abundance of the other two most common species was similar between plots (winter moth Operophtera brumata (no deer: 5–20; deer: 4–10 individuals/plot) and grey mountain carpet Entephria caesiata (no deer: 2–5; deer: 1–4 individuals/plot)). In each of eight forests, three pairs of plots with different deer densities were monitored. In five forests, deer exclosures (no deer) were compared to open forest (11–16 deer/km), but in the other three forests, plots were compared between sporting estates with different deer management policies (low density: 3–10 deer/km; high density: 11–20 deer/km). Between mid-May and early June 1991 and 1992, caterpillars were sampled by sweep-netting once/plot (125 sweeps covering 10 m2).

    Study and other actions tested
  2. A replicated, controlled study in 1997–1998 in a mixed forest in Arizona, USA (Bailey & Whitham 2002) found that aspen Populus tremuloides stands where elk Cervus canadensis were excluded had a higher abundance and species richness of arthropods (including moths) following intense fire, but a lower abundance and species richness following intermediate severity fire. After intense fire, the abundance and species richness of arthropods (including moths) was higher in aspen stands where elk were excluded (abundance: 6 individuals/plot; richness: 4 species/plot) than in browsed stands (abundance: 2 individuals/plot; richness: 1 species/plot), but following intermediate severity fire arthropod abundance and species richness was lower in elk-excluded (abundance: 5 individuals/plot; richness: 3 species/plot) than browsed stands (abundance: 8 individuals/plot; richness: 5 species/plot). The abundance of the most common moth, aspen blotch miner Lithocolletis tremuloidiella, did not differ significantly between elk-excluded (2–6 individuals/plot) and browsed stands (0–4 individuals/plot). Following a wildfire in 1996 which burned at high and intermediate intensity, in 1997 two 75-ha elk exclosures were constructed within a mixed ponderosa pine Pinus ponderosa and aspen forest. In summer 1998, arthropods (e.g. insects and spiders) were surveyed visually on the tallest aspen shoot in each of six 1-m2 plots in each of 12 aspen stands (three inside and three outside the exclosures in each of the high and intermediate intensity burned areas).

    Study and other actions tested
  3. A before-and-after study in 1990–2004 in a grassland in western Scotland, UK (Young & Barbour 2004) reported that after fencing excluded sheep, a population of New Forest burnet moth Zygaena viciae increased. Results were not tested for statistical significance. After seven years of complete sheep exclusion, 264 adults/transect were recorded, and the population was estimated at 8,500–10,200 individuals, compared to 0.1–1.2 adults/transect (estimated population 10–24 individuals) before and in the first six years of fencing (with occasional sheep grazing due to fence damage). The authors reported that this increase followed the spread of the hostplant meadow vetchling Lathyrus pratensis across the site. In early 1991 a 1-ha grassland, where 12 moths were found in 1990, was fenced to exclude sheep. In early 1994 and 1996, some sheep entered the site following damage to the fence, but from 1997–2004, sheep were completely excluded. In 1990, an intensive search for the moth was conducted. In July 1990–1991 and 1994–2003, moths were surveyed 1–15 times/year along a 5-m-wide, 300-m-long transect across the site, and were separately caught, marked and recaptured to estimate population size.

    Study and other actions tested
  4. A replicated, randomized, paired, controlled study in 1997–2002 in a grassland and forest reserve in New Mexico, USA (Kleintjes Neff et al. 2007) found that forest areas where elk Cervus elaphus were excluded had a higher abundance and species richness of butterflies in one out of four years, but the number of butterflies was similar in the remaining years and in all grassland areas. In forest sites where elk were excluded, the abundance (91 individuals/site) and species richness (17 species) of butterflies was higher than in browsed sites in one of four years (abundance: 42 individuals/site; richness: 13 species), but was not significantly different in the other three years (excluded: 2–9 individuals/site/year, 5–6 species/year; browsed: 2 individuals/site/year, 4–6 species/year). In grassland sites, the abundance and species richness of butterflies was not significantly different between exclusion (abundance: 3–141 individuals/site/year; richness: 5–16 species/year) and browsed sites (abundance: 3–85 individuals/site/year; richness: 4–13 species/year). In 1997–1998, four areas of ponderosa pine Pinus ponderosa grassland and five areas of mixed forest were selected, and a 60 × 60 m exclosure was constructed randomly on half of each site. Exclosures were 3 m high with 10-cm wire fencing to exclude elk. Three grassland sites and one forest site were deliberately or accidentally burned during the experiment. In June–August 1999–2002, butterflies were surveyed 2–5 times/year along a 360-m-long zigzag transect through each exclosure and browsed site. Grassland sites were not surveyed in 2002. Skippers (Hesperiidae) were not identified to species.

    Study and other actions tested
  5. A replicated, randomized, paired, controlled study in 1999–2004 in a pine forest in Arizona, USA (Huffman et al. 2009) reported that Fendler’s ceanothus Ceanothus fendleri shrubs protected from large herbivores with exclosures had a higher abundance of moths than unprotected shrubs. Results were not tested for statistical significance. On protected shrubs, 0.03–0.20 individual moths/plant, from three families, were recorded, compared to no moths on unprotected shrubs. In 1998–1999, trees <36 cm diameter were thinned in three experimental units (14–16 ha), and sixty Fendler’s ceanothus Ceanothus fendleri shrubs/unit (1–25 upright stems, covering <2 m2) were located. In 1999, thirty shrubs/unit were randomly selected, and had 4-m2, 1.4-m-high exclosures built around them. Exclosures had a large mesh (5 × 10 cm) on the sides, and open tops. In June 2002–2004, insects including moths were sampled by sweep netting (five sweeps/shrub, 20–50 cm above ground) through a subset of 30–52 shrubs/year (see paper for details), and identified to family level.

    Study and other actions tested
  6. A replicated, paired, site comparison study in 1998 in eight lowland forest sites in Mauritius (Florens et al. 2010) found that areas which were fenced to exclude non-native pigs and deer, together with removal of non-native plants, had a higher abundance and species richness of butterflies than unfenced sites where invasive species control had not been conducted. In fenced sites where invasive plants had been removed, both the abundance (5.9 individuals/100 m) and species richness (9 species) of native butterflies was higher than in unfenced sites where no weed removal had been conducted (abundance: 0.3 individuals/100 m; richness: 3 species). From 1986–1996, eight Conservation Management Areas (0.4–6.0 ha) were fenced to exclude non-native pigs and deer, and were regularly hand-weeded (1–3 times/year) to remove invasive plants, primarily strawberry guava Psidium cattleianum, rose apple Syzygium jambos, Ossaea marginata and Christmas berry Ardisia crenata. From April–June 1998, butterflies were surveyed on point counts along four to six 100-m transects in each weeded plot and in adjacent, non-weeded plots with an equivalent number of native canopy trees.

    Study and other actions tested
  7. A paired, controlled study in 2009–2010 on a calcareous grassland in the Netherlands (van Noordwijk et al. 2012) found that fewer Glanville fritillary Melitaea cinxia caterpillar nests were damaged in a fenced, ungrazed area than in a grazed area. After 10 days of autumn grazing, fewer caterpillar nests had signs of damage in a fenced area (2/24 nests damaged) than nests in a grazed area (15/25 nests damaged). Two months later, the number of nests with signs of damage was similar in fenced (6/24 nests damaged) and grazed areas (6/25). All 24 nests in the fenced area survived until spring, compared to 22/25 surviving in the grazed area (statistical significance not assessed). In July–August 2009, a grazed 4-ha grassland was searched three times for caterpillar nests. Half of the area with the highest density of nests was fenced to create a 0.15-ha ungrazed area. Twenty-four pairs of the largest, equally sized nests (>1 m apart) in each area were selected, and their location marked on GPS. In September 2009, the unfenced area was grazed by 114 sheep over 1.23 ha for 10 days, after which an expanded 1.76-ha area was grazed by 15 sheep for 50 days. In October and December 2009, nests were checked for damage, and in March 2010 the survival of each nest was recorded.

    Study and other actions tested
  8. A replicated, controlled study in 2001–2007 on a forested island in Quebec, Canada (Brousseau et al. 2013) found that reducing invasive white-tailed deer Odocoileus virginianus density increased the total species richness of macro-moths, and the abundance of rare nocturnal macro-moths, but not total macro-moth abundance. The total species richness of macro-moths, and the abundance of rare species, in areas with no deer (richness: 34 species/exclosure; abundance: 84 individuals) or reduced deer density (7.5–15 deer/km2: richness: 36 species/exclosure; abundance: 86–113 individuals) was higher than in areas where deer were not controlled (richness: 21 species/exclosure; abundance: 12 individuals). However, the total abundance of macro-moths did not differ significantly between sites (no deer: 113; 7.5 deer/km2: 139; 15 deer/km2: 122; uncontrolled: 87 individuals/exclosure). In 2001, fenced deer exclosures were built at three sites across a 7,943-km2 island. From 2002–2007, at each site, all deer were removed from a 10-ha exclosure (0 deer/km2), and three deer were stocked in both a 40-ha (7.5 deer/km2) and a 20-ha (15 deer/km2) exclosure from early spring to late autumn. An adjacent area with uncontrolled deer (26–57 deer/km2) was also monitored at each site. Within each exclosure, 70% of the area was harvested for timber just prior to construction in 2001. From June–August 2007, moths were sampled over five 3-day periods, using two Luminoc® traps/exclosure (>100 m apart). Traps were placed 3 m high and fitted with a 1.8 W blue light tube and Vapona® strips.

    Study and other actions tested
  9. A site comparison study in 2014 in a deciduous forest in Hokkaido, Japan (Iida et al. 2016) found that forest where deer were excluded with fencing had a similar abundance of all moths to forest with deer present, but had fewer tree-feeding moths. In an exclosure with no deer, the abundance of moths (320 individuals) was similar to the surrounding forest with deer present (322 individuals), but higher than in an enclosure with high deer density (280 individuals). The abundance of herb- and shrub-feeding species in the exclosure (19 individuals) was also similar to the surrounding forest (20 individuals), but higher than in the enclosure (17 individuals), whereas the abundance of tree-feeding species was lower in the exclosure (51 individuals) than in the surrounding forest (57 individuals) or enclosure (62 individuals). From 2004, sika deer Cervus nippon were excluded from a 1.5-ha fenced exclosure (0 deer/km2), and deer density was maintained at 20 deer/km2 within a 16.4-ha fenced enclosure. The remaining forest contained approximately 10 deer/km2. In June, July and September 2014, moths were sampled once/month using three light traps/site. Traps contained a 4 W fluorescent light and a 4 W UV light. Half of the moth species were classified as either herb- and shrub-feeding species, or tree-feeding species.

    Study and other actions tested
  10. A systematic review in 2018 of 13 studies in temperate and boreal forests from across the world (Bernes et al 2018) found that reducing or removing grazing or browsing by wild or domestic herbivores increased the abundance of moths and butterflies, but did not affect species richness. Forest plots where grazers and browsers were excluded or where herbivore density was reduced had a higher abundance of moths and butterflies than more heavily grazed forest, but species richness was not affected (data presented as model results). A total of 144 studies were included in the review, 13 of which reported data on moth and butterfly abundance, and three on species richness. The majority of the 144 studies came from North America (75), Europe (53) and Australia/New Zealand (14). Experimental plot size within studies ranged from 0.5 m2 to 2,428 ha. The majority of studies were controlled, and some included before-and-after measurements. Studies that were unreplicated, or did not include suitable comparisons, were excluded from the review.

    Study and other actions tested
Please cite as:

Bladon A.J., Bladon, E. K., Smith R.K. & Sutherland W.J. (2023) Butterfly and Moth Conservation: Global Evidence for the Effects of Interventions for butterflies and moths. Conservation Evidence Series Synopsis. University of Cambridge, Cambridge, UK.

Where has this evidence come from?

List of journals searched by synopsis

All the journals searched for all synopses

Butterfly and Moth Conservation

This Action forms part of the Action Synopsis:

Butterfly and Moth Conservation
Butterfly and Moth Conservation

Butterfly and Moth Conservation - Published 2023

Butterfly and Moth Synopsis

What Works 2021 cover

What Works in Conservation

What Works in Conservation provides expert assessments of the effectiveness of actions, based on summarised evidence, in synopses. Subjects covered so far include amphibians, birds, mammals, forests, peatland and control of freshwater invasive species. More are in progress.

More about What Works in Conservation

Download free PDF or purchase
The Conservation Evidence Journal

The Conservation Evidence Journal

An online, free to publish in, open-access journal publishing results from research and projects that test the effectiveness of conservation actions.

Read the latest volume: Volume 21

Go to the CE Journal

Discover more on our blog

Our blog contains the latest news and updates from the Conservation Evidence team, the Conservation Evidence Journal, and our global partners in evidence-based conservation.


Who uses Conservation Evidence?

Meet some of the evidence champions

Endangered Landscape ProgrammeRed List Champion - Arc Kent Wildlife Trust The Rufford Foundation Save the Frogs - Ghana Mauritian Wildlife Supporting Conservation Leaders
Sustainability Dashboard National Biodiversity Network Frog Life The international journey of Conservation - Oryx Cool Farm Alliance UNEP AWFA Bat Conservation InternationalPeople trust for endangered species Vincet Wildlife Trust