Background

The Countryside and Rights of Way Act 2000 (CRoW), which came into effect in 2005, has created a statutory right of access to open country and registered common land in England and Wales, extending the public’s ability to enjoy the countryside by opening up previously out-of-bounds areas. Scotland has similarly formalised access recently through the Land Reform (Scotland) Act 2003. However, public access (both in the UK and elsewhere) may have potentially deleterious impacts on habitats and associated flora and fauna, including species of conservation concern. Within the CRoW Act there is provision for the relevant authority to exclude or restrict access for the purpose of conserving flora, fauna, geological or physiographical features of the land in question. One particular concern and the focus of this systematic review, is the impact of human disturbance on breeding success of ground-nesting and cliff-nesting birds. It was considered that a systematic review would assist in evidence-based decision-making regarding the restriction of access for conservation purposes.

Objectives

To assess the impact of public access on foot (including associated activities i.e. dog- walking, picnicking, birdwatching, cross-country running, climbing, angling, mountain biking and horse riding) on breeding success of ground-nesting and cliff-nesting birds.

Search strategy

Relevant studies were identified through searches of the following 12 electronic databases: Agricola, Copac, Digital Dissertations Online, Directory of Open Access Journals, English Nature‟s “Wildlink”, Europa, Index to Theses Online (1970-present), ISI Web of Knowledge, JSTOR, Science Direct, Scirus and Scopus. Searches were undertaken on conservation and statutory organisation websites: Agricultural Development and Advisory Service (ADAS); Countryside Council for Wales (CCW); Department of Agriculture and Rural Development (DARD); Department of Environment, Food and Rural Affairs (DEFRA); English Nature (EN) (now Natural England); Joint Nature Conservation Committee (JNCC); National Trust (NT); Royal Society for the Protection of Birds (RSPB); and Scottish Natural Heritage (SNH). Bibliographies of traditional literature reviews and articles accepted into the systematic review at the full text stage were examined for studies that had not yet been identified by any other means. Subject experts were contacted.

Selection criteria

The criteria which studies had to meet for inclusion into the final stage of the systematic review were:

1. Subject: breeding ground-nesting and cliff-nesting birds;
2. Intervention: human activities of walking (including dog-walking), picnicking, birdwatching, crosscountry running, angling, climbing, mountain biking and horse-riding;
3. Outcome: primary outcomes were changes in breeding abundance/density and population effects (population size increase or decrease);
4. Type of study: any field/empirical study.

Main results

The searching of electronic databases and the internet produced 14,717 articles (Table 1). After duplicates were removed, a total of 4,904 unique articles remained for assessment at title and abstract stage, of which 173 were potentially relevant and required full text assessment against the study inclusion criteria. Full text assessment yielded a total of 85 articles that were relevant for inclusion within the systematic review. Of these, 27 had quantitative data (regarding the impact of public access and associated disturbance on breeding success of ground-nesting and cliff-nesting birds) suitable for meta-analysis but only 20 provided comparable data. A total of 42 independent data points presented data regarding the subject area of which 38 could be used for meta-analysis. Four data points were dropped due to heterogeneity in the scope of study, lack of comparators and lack of estimates of variance.

Of the meta-analyses that could be undertaken sample sizes were severely limited, primarily due to the lack of comparable quantitative data presented in studies. Those undertaken indicate that hatching success and pre-fledgling survival are both significantly reduced by human disturbance (although the latter outcome was skewed by findings of one study and thus must be interpreted accordingly). There is no significant overall effect on chick weight or fledgling success.

There is significant unexplained heterogeneity between studies and species for all the outcomes examined. The period in the breeding cycle when disturbance occurs; the type and intensity of disturbance; increased predation when attending adult birds are driven from eggs and/or young; the habitat, and degree of habituation to people have all been suggested as potential reasons for variability in avian response to disturbance. Small sample sizes confounded attempts to derive quantitative relationships between these explanatory covariates and effect.

Taking into account findings of many other (mostly observational) studies subject to review but which could not be incorporated into the meta-analysis, there is limited evidence for reduced hatching and fledging success due to human disturbance on foot (including with associated pet dogs) for some species.

Narrative synthesis shows that the few studies that have investigated the effects of human disturbance on breeding densities of ground-nesting bird species, indicate that breeding density (e.g. of common ringed plover Charadrius hiaticula, Eurasian golden plover Pluvialis apricaria, dunlin Calidris alpina, European nightjar Caprimulgus europaeus and woodlark Lullula arborea) is substantially reduced by recreational disturbances. Such reduced breeding density, or lack of breeding success within otherwise potentially suitable habitat (the latter as yet unproven), may be the main consequence of human disturbance.

There is little quantitative evidence to draw any firm conclusions regarding cliff-nesting species, but those few studies looking at impacts of human disturbance on breeding success of cliff-nesting birds suggest a negative impact on breeding success.

Reviewers’ conclusions

Implications for conservation

Evidence from quantitative studies (i.e. that subject to meta-analysis) for the impact of public access on breeding success is ambiguous (primarily due to small sample sizes). Qualitative/observational evidence derived from many other studies suggests that human disturbance through access on foot can be detrimental to the breeding success of ground- nesting birds at all stages of the breeding cycle from territory establishment to fledging. There are exceptions however, e.g. several species of penguins (Spheniscidae) and two species of tern (Sternidae). The design and reporting of the qualitative/observational studies leaves this evidence highly susceptible to bias.

A small number of mostly observational studies suggest that responses to a walker with a dog tended to be stronger than a person approaching without one; displacement of incubating or brooding birds led to increased predation risk from opportunistic predators, especially larger gulls Larus spp. and corvids Corvus spp.

The level of impact is highly variable between species and dependent upon locality and the disturbances involved. As such, proposed restrictions on access must take into account sensitivity and vulnerability to disturbance on a species by species basis, and site characteristics. There is insufficient evidence to draw any firm conclusions specifically regarding cliff-nesting species, but that which is available suggests a negative impact of human disturbance upon breeding success of a small number of species for which studies have been undertaken.

Implications for further research

There is much scope for further investigation into the effects of human disturbance on the breeding success of ground-nesting and cliff-nesting birds. Although there have been many studies looking at the effects of disturbance, few have robust quantitative data regarding impacts on breeding success and populations. Evidence given frequently stems from ad hoc observations rather than from rigorous, structured field research. A number of studies simply infer that a detrimental impact is presumed, or evidence presented is anecdotal. Some lack suitable controls, whilst in others disturbance levels (or treatments) are not well assessed or are poorly defined. Especially useful would be longer term studies investigating the consequences of disturbance impacting at the population level, of which to date very few ecologists have tackled. The role of recreational disturbance in reducing breeding bird densities (and potential site-scale desertion) could usefully be further addressed. It is evident that for some species the degree of habituation to people may be an important factor governing breeding success; this could also therefore be a topic of further research.