Table 4 The data sources, variables and outcomes of the returned articles relating to Patrol effectiveness in preventing wildlife crime
From: Systematic review of situational prevention methods for crime against species
Publication | Data source/s | Dependent variable/s | Independent variable | Confounding variable/s | Statistical analysis | Results: effectiveness of patrols |
|---|---|---|---|---|---|---|
Hilborn et al. 2006 | Arrest records Recorded since 1957 Antipoaching activity records Only available for some years Species abundance Buffalo, Elephant and Rhino numbers | Abundance of species African Buffalo, Elephant, Black Rhino | Relative Poaching effort Ratio of Poaching Intensity (arrests per patrol) to Antipoaching effort (patrols per day) | Influence of community engagement mentioned, however the authors claim the decline in poaching effort occurred before the community programs were initiated, therefore the authors attribute the reduction in poaching to patrols | Population dynamics model Plotting the time series data on Abundance of Species and Relative Poaching Effort | The trends observed indicated that declines in species abundance correlated with periods when patrolling was limited due to park funding. As patrolling increased, species abundance was found to improve |
Leader-Williams et al. 1990 | Field records Provided information on:  Illegal activity   Total Poachers Encountered   Camps—indication of illegal activity   Fresh Carcasses (flesh covered rhino/elephant carcasses with trophies axed off)—indication of successful poaching activity  Patrol data Number of Patrol days | Illegal activity (poachers, camps, carcasses) Abundance of species Historical sightings of Rhino and Elephant by foot patrols used as a measure of population trends | Patrol Effort Encounter rates per effective patrol day | Month/Season Patrol Length Number of Scouts Area Year (fitted as categorical) Anti-Poaching Unit (APU) | Simple regression Rates of change in species abundance and patrol effort (in all areas in all years), to assess if animal abundance declined faster in areas where there was least effort Multiple regression Illegal activity, patrol effort and species abundance to establish if illegal activity was distributed in areas with least patrol effort and most quarry | Negative relationship between camps, carcasses and effort Increased patrol effort was associated with a reduction in finds of fresh carcasses and camps Overall, the results suggested that patrol effort reduced the number of finds of fresh carcasses and illegal activity |
Jachmann and Billiouw 1997 | Patrol reports Detailing routes and observations | Number of elephants killed Illegally (per annum) | Financial and Patrol variables:  Total law enforcement budget per km2 (US$)  Personal emoluments per scout per month (US$)  Transport expenditure per km2 (US$) km2 per scout  Km2 per carrier  Effective patrol days per km2  Effective investigation days  Number of bonuses paid: average bonus rates (US$) | – | Poisson Regression Model | Deterrence effects of both foot and investigative patrols were found to be equally important The predictor variables below were found to be significantly negatively associated with the number of elephants killed illegally:  Bonuses Paid  Number of Scouts  Law enforcement budget  Personal emoluments per scout per month |
Linkie et al. 2015 | Field data from ranger patrol logbooks  Snare Trap Occurrence  Patrol Frequency  Patrol Effort Biological data on tiger prey base collected using camera trap images from 2004–2006 and 2009–2011  Tiger Occupancy/Abundance | Snare Trap Occurrence Tiger Prey Occupancy /Abundance | Patrol frequency (number of times visited over previous 1 and 2 year periods) Patrol effort (number of kilometres patrolled in previous 1 and 2 year periods) | Spatial covariates related to accessibility:  Elevation  Slope  Proximity to nearest road  Proximity to forest edge  Proximity to nearest village & protected area status Additional covariate assessed between 2009–2010:  Tip Off—influence of intelligence-based patrolling | Markov Model Modelling the deterrence of poachers by patrols, due to the removal of snares leading to reduction in the success of poaching | Snare trap occurrence between 2000 and 2010 declined by 24%, but the result was not statistically significant. The authors identified that the analysis did not control for potential spatial effects created by the establishment of new teams in previously unpatrolled areas. These new teams would likely contribute to the overall number of snares detected in the study area. Highlighting the importance for local-level, individual patrol measurements |
Linkie et al. 2015 …Continued | – | – | – | – | – | Patrol Frequency was found to have the most influence over the detection of snare traps, followed by Patrol Effort (number of kilometers walked) Both variables were found to have an increased influence on the detection of snares if performed routinely over more time (e.g. 2 years compared to 1 year) Accessibility was found to be an important factor in relation to patrol team performance. More snare traps were recovered from sites considered more accessible due to lower elevation, and flatter terrain |