Cally Ham
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Research Subjects
Cally Ham has a total of 51 co-authors. Here are the top co-authors by number of shared publications:
- Rosie Woodroffe
Zoological Society of London London UK
5 shared publications - Christl Ann Donnelly
Department of Statistics, University of Oxford, Oxford, United Kingdom
4 shared publications - Eleanor R. Scopes
Forest Research, Forestry Commission National Office, Bristol BS16 1EJ, UK
4 shared publications - Matt Guy
Forest Research Farnham, Surrey UK
3 shared publications - Catherine M. McNicol
British Association for Shooting and Conservation Marford Mill Wrexham UK
2 shared publications - Chloe Bellamy
Forest Research Northern Research Station Roslin UK
2 shared publications
Latest Publications
- Open Access
Grey Squirrel Density and Damage in Relation to the Presence of Predators
The grey squirrel (Sciurus carolinensis) is an invasive species in Great Britain that causes substantial economic and ecological damage through stripping bark from broadleaf trees. The relationship between grey squirrel abundance and tree damage remains poorly understood, as does the potential role of recovering native predators in mitigating these impacts. We conducted an observational survey of 22 woodland sites across England and Wales with differing occurrences of pine martens (Martes martes) and northern goshawks (Accipiter gentilis), estimating grey squirrel density and quantifying bark-stripping damage. Grey squirrel density varied across sites. No significant relationship was detected between squirrel density and observed tree damage when analyses were restricted to sites where squirrels were present. However, when sites with no detected squirrels were included, squirrel density was positively associated with tree damage. Predator occurrence was associated with significantly lower grey squirrel densities. In contrast, predator presence was not significantly associated with the proportion of trees exhibiting bark-stripping damage. These findings suggest that predators may contribute to grey squirrel population suppression but that bark stripping is influenced by factors beyond contemporary squirrel abundance. Forest management practices that facilitate predator presence should be encouraged to support reduction in grey squirrel density and therefore improve future tree health. Further research should focus on fresh damage assessments and predator-mediated behavioural responses to better understand the mechanisms linking predator recovery, squirrel abundance and tree damage.
Open AccessGrey Squirrel Density and Damage in Relation to the Presence of Predators
The grey squirrel (Sciurus carolinensis) is an invasive species in Great Britain that causes substantial economic and ecological damage through stripping bark from broadleaf trees. The relationship between grey squirrel abundance and tree damage remains poorly understood, as does the potential role of recovering native predators in mitigating these impacts. We conducted an observational survey of 22 woodland sites across England and Wales with differing occurrences of pine martens (Martes martes) and northern goshawks (Accipiter gentilis), estimating grey squirrel density and quantifying bark-stripping damage. Grey squirrel density varied across sites. No significant relationship was detected between squirrel density and observed tree damage when analyses were restricted to sites where squirrels were present. However, when sites with no detected squirrels were included, squirrel density was positively associated with tree damage. Predator occurrence was associated with significantly lower grey squirrel densities. In contrast, predator presence was not significantly associated with the proportion of trees exhibiting bark-stripping damage. These findings suggest that predators may contribute to grey squirrel population suppression but that bark stripping is influenced by factors beyond contemporary squirrel abundance. Forest management practices that facilitate predator presence should be encouraged to support reduction in grey squirrel density and therefore improve future tree health. Further research should focus on fresh damage assessments and predator-mediated behavioural responses to better understand the mechanisms linking predator recovery, squirrel abundance and tree damage.
- Open Access
- 13 October 2026
- Published by Wiley in Journal of Applied Ecology
- Vol. 63 (10)
Comparing the development of naturally colonised and planted temperate forests across a 60‐year chronosequence
Natural colonisation is proposed as a cost‐effective alternative to tree planting for combatting climate change and promoting biodiversity. However, in temperate regions, few studies have compared the development of planted and naturally colonised forests across the spatial or temporal scales needed for generalisable results to inform policy and practice. We identified a paired space‐for‐time chronosequence spanning >60 years of naturally colonised and planted forest patches (forest creation types), all adjacent to mature forest. We compared forest structure across forest ages and examined differences in tree community composition in naturally colonised sites. Field surveys were conducted using 10‐m radius plots at varying distances from the mature forests. All trees ≥7 cm diameter were identified and measured to calculate basal area. Using LiDAR data, we extracted tree canopy height, variability and cover as metrics of structural heterogeneity across the forest creation types. Both canopy height and basal area, predictors of above‐ground biomass, were consistently higher in planted forest patches, averaged across all age pairings at 8.56 m 2 ha −1 greater and 2.1 m taller than naturally colonised forests, respectively. Naturally colonised forests had more open canopies (19% on average), while canopy height variation was similar between forest creation types. A greater abundance of willow species ( Salix spp.) was associated with younger naturally colonised forests while the community composition of older naturally colonised forests was more comparable to adjacent mature forests. Synthesis and applications: During forest creation in temperate, agriculturally dominated systems, tree planting accumulates above ground biomass more quickly than natural colonisation whereas naturally colonised forests are more open and structurally heterogeneous. Differences in above‐ground biomass persist for decades where differences in structural heterogeneity reduce with age. Tree planting may therefore be preferred for carbon storage and timber production, whereas natural colonisation may be preferred for increasing forest biodiversity.
- 13 October 2026
- Published by Wiley in Journal of Applied Ecology
- Vol. 63 (10)
Open AccessComparing the development of naturally colonised and planted temperate forests across a 60‐year chronosequence
Natural colonisation is proposed as a cost‐effective alternative to tree planting for combatting climate change and promoting biodiversity. However, in temperate regions, few studies have compared the development of planted and naturally colonised forests across the spatial or temporal scales needed for generalisable results to inform policy and practice. We identified a paired space‐for‐time chronosequence spanning >60 years of naturally colonised and planted forest patches (forest creation types), all adjacent to mature forest. We compared forest structure across forest ages and examined differences in tree community composition in naturally colonised sites. Field surveys were conducted using 10‐m radius plots at varying distances from the mature forests. All trees ≥7 cm diameter were identified and measured to calculate basal area. Using LiDAR data, we extracted tree canopy height, variability and cover as metrics of structural heterogeneity across the forest creation types. Both canopy height and basal area, predictors of above‐ground biomass, were consistently higher in planted forest patches, averaged across all age pairings at 8.56 m 2 ha −1 greater and 2.1 m taller than naturally colonised forests, respectively. Naturally colonised forests had more open canopies (19% on average), while canopy height variation was similar between forest creation types. A greater abundance of willow species ( Salix spp.) was associated with younger naturally colonised forests while the community composition of older naturally colonised forests was more comparable to adjacent mature forests. Synthesis and applications: During forest creation in temperate, agriculturally dominated systems, tree planting accumulates above ground biomass more quickly than natural colonisation whereas naturally colonised forests are more open and structurally heterogeneous. Differences in above‐ground biomass persist for decades where differences in structural heterogeneity reduce with age. Tree planting may therefore be preferred for carbon storage and timber production, whereas natural colonisation may be preferred for increasing forest biodiversity.
- Open Access
- 13 August 2026
- Published by Wiley in Ecology and Evolution
- Vol. 16 (8) , e74175
Individual‐Based Models Reveal Population Recovery and Impacts of Woodland Creation on Pine Martens ( Martes martes ) in Britain
Individual‐based models (IBMs) predict distribution and abundance patterns resulting from individual life histories. They are useful for simulating spread, particularly in reintroductions, and testing multiple conservation actions. The accuracy of IBMs relies on the parameters used, and recently, calibration methods have been developed to integrate disparate information. We calibrate an IBM for the pine marten ( Martes martes ) in Britain using observed presence/absence data. We then use this model to investigate whether the multiple reintroduction projects in Southwest Britain are likely to coalesce into a single population, and whether woodland creation can be designed to facilitate dispersal. Our calibrated model reveals that some populations in Southwest Britain may remain isolated after 150 years, likely due to a lack of residential habitat between population centres under our definition of suitable habitat. However, they remain successful at a local scale, with populations persisting in all release zones. We find that none of our woodland creation scenarios, which vary in the area and configuration of new patches, increase the abundance of pine martens or their dispersal capacity. Instead, the biggest factor influencing outcomes is the landscape. Our results suggest that realistic tree planting, in particular patch size, is insufficient to benefit pine martens under our model definitions. However, the coarse scale of our model and our definition of suitable patches could underestimate the influence of small woodlands and hedges on population dispersal and viability. We show that the reintroduction of pine martens to England and Wales successfully restores parts of the former range into the near future. Woodland cover is an important factor in landscape carrying capacity, but significant woodland creation may be needed to enact meaningful change. Our calibrated model can support the selection of reintroduction sites with the greatest probability of facilitating population coalescence.
- 13 August 2026
- Published by Wiley in Ecology and Evolution
- Vol. 16 (8) , e74175
Open AccessIndividual‐Based Models Reveal Population Recovery and Impacts of Woodland Creation on Pine Martens ( Martes martes ) in Britain
Individual‐based models (IBMs) predict distribution and abundance patterns resulting from individual life histories. They are useful for simulating spread, particularly in reintroductions, and testing multiple conservation actions. The accuracy of IBMs relies on the parameters used, and recently, calibration methods have been developed to integrate disparate information. We calibrate an IBM for the pine marten ( Martes martes ) in Britain using observed presence/absence data. We then use this model to investigate whether the multiple reintroduction projects in Southwest Britain are likely to coalesce into a single population, and whether woodland creation can be designed to facilitate dispersal. Our calibrated model reveals that some populations in Southwest Britain may remain isolated after 150 years, likely due to a lack of residential habitat between population centres under our definition of suitable habitat. However, they remain successful at a local scale, with populations persisting in all release zones. We find that none of our woodland creation scenarios, which vary in the area and configuration of new patches, increase the abundance of pine martens or their dispersal capacity. Instead, the biggest factor influencing outcomes is the landscape. Our results suggest that realistic tree planting, in particular patch size, is insufficient to benefit pine martens under our model definitions. However, the coarse scale of our model and our definition of suitable patches could underestimate the influence of small woodlands and hedges on population dispersal and viability. We show that the reintroduction of pine martens to England and Wales successfully restores parts of the former range into the near future. Woodland cover is an important factor in landscape carrying capacity, but significant woodland creation may be needed to enact meaningful change. Our calibrated model can support the selection of reintroduction sites with the greatest probability of facilitating population coalescence.
- Open Access
- 13 August 2026
- Published by Springer Nature in European Journal of Wildlife Research
- Vol. 72 (5) , 106
More nuts, more squirrels, more damage? The covariation of beech (Fagus sylvatica) masting with grey squirrel (Sciurus carolinensis) abundance, and bark stripping damage, in the UK
Grey squirrels ( Sciurus carolinensis ) are an invasive non-native species in the UK which can strip bark from trees, detrimentally impacting biodiversity and forestry. Extensive control measures are undertaken to reduce grey squirrel populations each year; however, it is unclear if abundance influences the extent of damage. Predicting changes in abundance could be an important technique to increase the efficacy of control measures. We explore whether climate-based predictors of beech ( Fagus sylvatica ) masting, covary with an abundance index of grey squirrels. We find predicted masting years covary with the abundance of grey squirrels at all sites and those in broadleaf woodland the following year. We also find that predicted masting years covary with squirrel abundance at beech dominated sites in the same year as masting. These inconsistent patterns may be due to the limitations with the proxies used this analysis, but indicate potential hypotheses for future studies. We also investigate whether grey squirrel abundance index covaries with the prevalence of accumulated bark stripping damage with inconclusive results. More research is needed to identify the drivers of bark stripping to identify new preventative measures. Our exploratory results suggest potential influences of beech masting on grey squirrel abundance, although more research is needed before this could be used within management.
- 13 August 2026
- Published by Springer Nature in European Journal of Wildlife Research
- Vol. 72 (5) , 106
Open AccessMore nuts, more squirrels, more damage? The covariation of beech (Fagus sylvatica) masting with grey squirrel (Sciurus carolinensis) abundance, and bark stripping damage, in the UK
Grey squirrels ( Sciurus carolinensis ) are an invasive non-native species in the UK which can strip bark from trees, detrimentally impacting biodiversity and forestry. Extensive control measures are undertaken to reduce grey squirrel populations each year; however, it is unclear if abundance influences the extent of damage. Predicting changes in abundance could be an important technique to increase the efficacy of control measures. We explore whether climate-based predictors of beech ( Fagus sylvatica ) masting, covary with an abundance index of grey squirrels. We find predicted masting years covary with the abundance of grey squirrels at all sites and those in broadleaf woodland the following year. We also find that predicted masting years covary with squirrel abundance at beech dominated sites in the same year as masting. These inconsistent patterns may be due to the limitations with the proxies used this analysis, but indicate potential hypotheses for future studies. We also investigate whether grey squirrel abundance index covaries with the prevalence of accumulated bark stripping damage with inconclusive results. More research is needed to identify the drivers of bark stripping to identify new preventative measures. Our exploratory results suggest potential influences of beech masting on grey squirrel abundance, although more research is needed before this could be used within management.
- Open Access
- 16 September 2025
- Published by Wiley in Journal of Applied Ecology
- Vol. 62 (11) , 2930-2940
Validating habitat suitability models for pine marten ( Martes martes ) reintroductions to England and Wales
Habitat suitability models (HSMs) are used extensively in conservation research and practice, particularly in reintroduction projects to identify prospective locations for translocations to take place. However, HSMs are only useful if their predictive capabilities are accurate; otherwise, there is a risk of reintroduction failures and misdirecting conservation actions. We validate and refine a HSM for pine marten ( Martes martes ) in Britain using an independent dataset of new observational data from two reintroduction projects and natural re‐colonisation. We test whether additional data on roads and trees outside of woodland improves the HSM predictions of habitat suitability and present an updated final HSM integrating all the observation data. We find the existing HSM adequately predicts pine marten presences but has random fit when considering both presences and absences. This is likely due to biases in the absence dataset which solely derives from failed scat transects on forest roads. The HSM reveals woodland cover at the 3 km scale is the key environmental factor for pine marten habitat suitability. The addition of data on roads and trees outside of woodland did not improve the fit of the HSM to the independent dataset. The updated model we present is therefore based on the original environmental inputs but uses up‐to‐date pine marten presences. Synthesis and applications: We present a validated HSM for pine martens in Britain which can be used to identify areas for future reintroduction projects, such as Devon and Cumbria, as well as areas where woodland planting could be targeted to improve habitat suitability and connectivity. Our adaptive methodology, which uses the monitoring data from reintroduction projects to validate and refine the HSMs created in the feasibility stage, can be applied to other species as reintroduction projects continue to grow in popularity. This ensures a robust basis for ongoing reintroduction suitability mapping and population recovery.
- 16 September 2025
- Published by Wiley in Journal of Applied Ecology
- Vol. 62 (11) , 2930-2940
Open AccessValidating habitat suitability models for pine marten ( Martes martes ) reintroductions to England and Wales
Habitat suitability models (HSMs) are used extensively in conservation research and practice, particularly in reintroduction projects to identify prospective locations for translocations to take place. However, HSMs are only useful if their predictive capabilities are accurate; otherwise, there is a risk of reintroduction failures and misdirecting conservation actions. We validate and refine a HSM for pine marten ( Martes martes ) in Britain using an independent dataset of new observational data from two reintroduction projects and natural re‐colonisation. We test whether additional data on roads and trees outside of woodland improves the HSM predictions of habitat suitability and present an updated final HSM integrating all the observation data. We find the existing HSM adequately predicts pine marten presences but has random fit when considering both presences and absences. This is likely due to biases in the absence dataset which solely derives from failed scat transects on forest roads. The HSM reveals woodland cover at the 3 km scale is the key environmental factor for pine marten habitat suitability. The addition of data on roads and trees outside of woodland did not improve the fit of the HSM to the independent dataset. The updated model we present is therefore based on the original environmental inputs but uses up‐to‐date pine marten presences. Synthesis and applications: We present a validated HSM for pine martens in Britain which can be used to identify areas for future reintroduction projects, such as Devon and Cumbria, as well as areas where woodland planting could be targeted to improve habitat suitability and connectivity. Our adaptive methodology, which uses the monitoring data from reintroduction projects to validate and refine the HSMs created in the feasibility stage, can be applied to other species as reintroduction projects continue to grow in popularity. This ensures a robust basis for ongoing reintroduction suitability mapping and population recovery.
- 3 September 2025
- Published by Forest Research
Feral wild boar and deer in the Forest of Dean Public Forest Estate
The Forest of Dean maintains the largest population of feral wild boar (Sus scrofa) in England, as well as populations of roe deer (Capreolus capreolus), fallow deer (Dama dama), and muntjac deer (Muntiacus reevesi). The deer and wild boar populations in the Forest of Dean have been monitored annually since 2013 to help inform the public and support the management programme. Nationally, the deer population is increasing (Matthews et al., 2018), so monitoring long-term fluctuations in density provides important information to guide woodland and wildlife management. In this report, we provide results up to and including the latest survey, which was carried out in March 2025.
- 3 September 2025
- Published by Forest Research
Feral wild boar and deer in the Forest of Dean Public Forest Estate
The Forest of Dean maintains the largest population of feral wild boar (Sus scrofa) in England, as well as populations of roe deer (Capreolus capreolus), fallow deer (Dama dama), and muntjac deer (Muntiacus reevesi). The deer and wild boar populations in the Forest of Dean have been monitored annually since 2013 to help inform the public and support the management programme. Nationally, the deer population is increasing (Matthews et al., 2018), so monitoring long-term fluctuations in density provides important information to guide woodland and wildlife management. In this report, we provide results up to and including the latest survey, which was carried out in March 2025.
- Open Access
- 1 January 2025
- Published by Slovenian Forestry Institute
- p. 83
Using novel methods to prioritise grey squirrel control
- 1 January 2025
- Published by Slovenian Forestry Institute
- p. 83
Open AccessUsing novel methods to prioritise grey squirrel control
- Open Access
- 5 August 2024
- Published by Wiley in People and Nature
- Vol. 6 (5) , 1960-1973
Farmer‐led badger vaccination in Cornwall: Epidemiological patterns and social perspectives
In the United Kingdom, the management of bovine tuberculosis (bTB) challenges the coexistence of people and wildlife. Control of this cattle disease is hindered by transmission of its causative agent, Mycobacterium bovis, between cattle and badgers Meles meles. Badger culling has formed an element of bTB control policy for decades, but current government policy envisions expanding badger vaccination. Farming leaders are sceptical, citing concerns that badger vaccination would be impractical and potentially ineffective. We report on a 4‐year badger vaccination initiative in an 11 km2 area which, atypically, was initiated by local farmers, delivered by scientists and conservationists, and co‐funded by all three. Participating landholders cited controversies around culling and a desire to support neighbours as their primary reasons for adopting vaccination. The number of badgers vaccinated per km2 (5.6 km−2 in 2019) exceeded the number culled on nearby land (2.9 km−2 in 2019), and the estimated proportion vaccinated (74%, 95% confidence interval [CI] 40%–137%) exceeded the 30% threshold predicted by models to be necessary to control M. bovis. Farmers were content with how vaccination was delivered, and felt that it built trust with wildlife professionals. The percentage of badgers testing positive for M. bovis declined from 16.0% (95% CI 4.5%–36.1%) at the start of vaccination to 0% (95% CI 0%–9.7%) in the final year. With neither replication nor unvaccinated controls, this small‐scale case study does not demonstrate a causal link between badger vaccination and bTB epidemiology, but it does suggest that larger‐scale evaluation of badger vaccination would be warranted. Farmers reported that their enthusiasm for badger vaccination had increased after participating for 4 years. They considered vaccination to have been effective, and good value for money, and wished to continue with it. Synthesis and applications: Although small‐scale, this case study suggests that badger vaccination can be a technically effective and socially acceptable component of bTB control. A wider rollout of badger vaccination is more likely if it is led by the farming community, rather than by conservationists or government, and is combined with scientific monitoring. Read the free Plain Language Summary for this article on the Journal blog.
- 5 August 2024
- Published by Wiley in People and Nature
- Vol. 6 (5) , 1960-1973
Open AccessFarmer‐led badger vaccination in Cornwall: Epidemiological patterns and social perspectives
In the United Kingdom, the management of bovine tuberculosis (bTB) challenges the coexistence of people and wildlife. Control of this cattle disease is hindered by transmission of its causative agent, Mycobacterium bovis, between cattle and badgers Meles meles. Badger culling has formed an element of bTB control policy for decades, but current government policy envisions expanding badger vaccination. Farming leaders are sceptical, citing concerns that badger vaccination would be impractical and potentially ineffective. We report on a 4‐year badger vaccination initiative in an 11 km2 area which, atypically, was initiated by local farmers, delivered by scientists and conservationists, and co‐funded by all three. Participating landholders cited controversies around culling and a desire to support neighbours as their primary reasons for adopting vaccination. The number of badgers vaccinated per km2 (5.6 km−2 in 2019) exceeded the number culled on nearby land (2.9 km−2 in 2019), and the estimated proportion vaccinated (74%, 95% confidence interval [CI] 40%–137%) exceeded the 30% threshold predicted by models to be necessary to control M. bovis. Farmers were content with how vaccination was delivered, and felt that it built trust with wildlife professionals. The percentage of badgers testing positive for M. bovis declined from 16.0% (95% CI 4.5%–36.1%) at the start of vaccination to 0% (95% CI 0%–9.7%) in the final year. With neither replication nor unvaccinated controls, this small‐scale case study does not demonstrate a causal link between badger vaccination and bTB epidemiology, but it does suggest that larger‐scale evaluation of badger vaccination would be warranted. Farmers reported that their enthusiasm for badger vaccination had increased after participating for 4 years. They considered vaccination to have been effective, and good value for money, and wished to continue with it. Synthesis and applications: Although small‐scale, this case study suggests that badger vaccination can be a technically effective and socially acceptable component of bTB control. A wider rollout of badger vaccination is more likely if it is led by the farming community, rather than by conservationists or government, and is combined with scientific monitoring. Read the free Plain Language Summary for this article on the Journal blog.
- Open Access
- 1 July 2024
- Published by Wiley in Ecological Solutions and Evidence
- Vol. 5 (3)
Evaluating camera‐based methods for estimating badger (Meles meles) density: Implications for wildlife management
Accurate and precise assessment of population density plays a critical role in effective wildlife management, but reliable estimates are often difficult to obtain. Camera traps have emerged as valuable noninvasive tools for studying elusive species, offering cost‐effective solutions for both marked and unmarked populations. We evaluated the consistency of badger (Meles meles) density estimates obtained from the random encounter model (REM) and camera trap distance sampling (CT‐DS) with independent estimates from spatial mark‐resight (SMR) models and quantified the bias in CT‐DS arising from animals reacting to camera traps. Six camera trap surveys were conducted in Cornwall, UK, in 2019 and 2021, and data were used to estimate badger density using the REM and CT‐DS. Four sites were included in a badger vaccination research project, providing an opportunity to mark badgers with uniquely identifiable fur clips to facilitate resighting within a SMR framework. We found consistency in the density estimates across all methods, but results had wide confidence intervals. Density estimates derived from CT‐DS tended to be higher than those from the REM and were sensitive to the exclusion of reactive sequences, resulting in a twofold decrease in the estimated density in one case. The REM tended to be the most precise method; however, where badger density was low, precision was low using all methods. Practical implication: our findings suggest animal density can be assessed from camera traps in the absence of individual identification; however, it is important to account for reactive behaviours, especially where such behaviour is prevalent. In these circumstances, we recommend utilising the REM which offers a clear methodology for addressing bias arising from reactive sequences. In addition, we emphasise the need for improved precision to ensure the effectiveness of these methods in the context of wildlife management. We offer practical considerations to facilitate the broader application of these methods, drawing upon the example of disease control through badger vaccination.
- 1 July 2024
- Published by Wiley in Ecological Solutions and Evidence
- Vol. 5 (3)
Open AccessEvaluating camera‐based methods for estimating badger (Meles meles) density: Implications for wildlife management
Accurate and precise assessment of population density plays a critical role in effective wildlife management, but reliable estimates are often difficult to obtain. Camera traps have emerged as valuable noninvasive tools for studying elusive species, offering cost‐effective solutions for both marked and unmarked populations. We evaluated the consistency of badger (Meles meles) density estimates obtained from the random encounter model (REM) and camera trap distance sampling (CT‐DS) with independent estimates from spatial mark‐resight (SMR) models and quantified the bias in CT‐DS arising from animals reacting to camera traps. Six camera trap surveys were conducted in Cornwall, UK, in 2019 and 2021, and data were used to estimate badger density using the REM and CT‐DS. Four sites were included in a badger vaccination research project, providing an opportunity to mark badgers with uniquely identifiable fur clips to facilitate resighting within a SMR framework. We found consistency in the density estimates across all methods, but results had wide confidence intervals. Density estimates derived from CT‐DS tended to be higher than those from the REM and were sensitive to the exclusion of reactive sequences, resulting in a twofold decrease in the estimated density in one case. The REM tended to be the most precise method; however, where badger density was low, precision was low using all methods. Practical implication: our findings suggest animal density can be assessed from camera traps in the absence of individual identification; however, it is important to account for reactive behaviours, especially where such behaviour is prevalent. In these circumstances, we recommend utilising the REM which offers a clear methodology for addressing bias arising from reactive sequences. In addition, we emphasise the need for improved precision to ensure the effectiveness of these methods in the context of wildlife management. We offer practical considerations to facilitate the broader application of these methods, drawing upon the example of disease control through badger vaccination.
- Open Access
- 27 November 2023
- Published by Wiley in Methods in Ecology and Evolution
- Vol. 15 (1) , 91-102
Sherlock—A flexible, low‐resource tool for processing camera‐trapping images
The use of camera traps to study wildlife has increased markedly in the last two decades. Camera surveys typically produce large data sets which require processing to isolate images containing the species of interest. This is time consuming and costly, particularly if there are many empty images that can result from false triggers. Computer vision technology can assist with data processing, but existing artificial intelligence algorithms are limited by the requirement of a training data set, which itself can be challenging to acquire. Furthermore, deep‐learning methods often require powerful hardware and proficient coding skills. We present Sherlock, a novel algorithm that can reduce the time required to process camera trap data by removing a large number of unwanted images. The code is adaptable, simple to use and requires minimal processing power. We tested Sherlock on 240,596 camera trap images collected from 46 cameras placed in a range of habitats on farms in Cornwall, United Kingdom, and set the parameters to find European badgers (Meles meles). The algorithm correctly classified 91.9% of badger images and removed 49.3% of the unwanted ‘empty’ images. When testing model parameters, we found that faster processing times were achieved by reducing both the number of sampled pixels and ‘bouncing’ attempts (the number of paths explored to identify a disturbance), with minimal implications for model sensitivity and specificity. When Sherlock was tested on two sites which contained no livestock in their images, its performance greatly improved and it removed 92.3% of the empty images. Although further refinements may improve its performance, Sherlock is currently an accessible, simple and useful tool for processing camera trap data.
- 27 November 2023
- Published by Wiley in Methods in Ecology and Evolution
- Vol. 15 (1) , 91-102
Open AccessSherlock—A flexible, low‐resource tool for processing camera‐trapping images
The use of camera traps to study wildlife has increased markedly in the last two decades. Camera surveys typically produce large data sets which require processing to isolate images containing the species of interest. This is time consuming and costly, particularly if there are many empty images that can result from false triggers. Computer vision technology can assist with data processing, but existing artificial intelligence algorithms are limited by the requirement of a training data set, which itself can be challenging to acquire. Furthermore, deep‐learning methods often require powerful hardware and proficient coding skills. We present Sherlock, a novel algorithm that can reduce the time required to process camera trap data by removing a large number of unwanted images. The code is adaptable, simple to use and requires minimal processing power. We tested Sherlock on 240,596 camera trap images collected from 46 cameras placed in a range of habitats on farms in Cornwall, United Kingdom, and set the parameters to find European badgers (Meles meles). The algorithm correctly classified 91.9% of badger images and removed 49.3% of the unwanted ‘empty’ images. When testing model parameters, we found that faster processing times were achieved by reducing both the number of sampled pixels and ‘bouncing’ attempts (the number of paths explored to identify a disturbance), with minimal implications for model sensitivity and specificity. When Sherlock was tested on two sites which contained no livestock in their images, its performance greatly improved and it removed 92.3% of the empty images. Although further refinements may improve its performance, Sherlock is currently an accessible, simple and useful tool for processing camera trap data.