1. WALLACE RESOURCE LIBRARY
Module 02 – Ecological Survey Techniques
D04 – Camera trapping to assess large mammal populations in Amazonia

2. Why survey mammal populations?
Understand population dynamics
- e.g. calculating the carrying capacity of the environment for different mammals
Determine impacts of extreme events
- e.g. extreme high water levels of during flooded season
Check sustainability of hunting
- confirm hunting quotas are leaving sufficient animals for stable populations
Monitoring the population sizes of mammals is crucial
For understanding the effects of natural and man-made environmental changes
To ensure conservation strategies are working effectively.
There are over 100 species of mammal in the Pacaya Samiria Nature Reserve in Peru.
These include arboreal primates, for example howler monkeys, saddle-backed tamarins and brown capuchins, and diverse terrestrial and arboreal mammals such as jaguar, puma, opossums, armadillos, sloths, tapirs and peccaries.
As a result in seasonal changes in rainfall at the source of the major Amazon river systems, the Peruvian Andes, the water levels of the Samiria River typically differ by 8–10m between the peak of the high water and low water seasons.
As the water levels rise so the amount of available land decreases, condensing terrestrial animals into smaller areas.
For herbivores this means there is both less vegetation present and also greater competition for resources, though there may be a lag before seeing the effects of these circumstances reflected in the population size.
Whilst carnivores are also concentrated into a reduced area, they benefit from the tighter assemblages of their prey.
Community-based conservation methods means the people who have called the land of the Pacaya Samiria reserve their home for centuries are permitted to hunt certain mammals and birds for sustenance.
Quotas are set as to how many animals can be hunted, but as environmental conditions change, so do the numbers that can be sustainably hunted
Surveying the population sizes of these animals shows if the quotas are appropriate, or whether they need to be changed to ensure population numbers remain stable

3. Line transects Used for counting numbers of conspicuous animals
Observers walk along transect
Distance at which animals of interest are seen are measured
Software estimates population densities
Line transect surveys are commonly used to estimate population densities of mammals and birds, particularly those that are large, gregarious, conspicuous or that favour open habitats
Observers walk a fixed trail, recording any sightings and the perpendicular distance from the trail at which the animal was seen
The further away an animal is from the trail, the less likely it is to be seen
A software programme called DISTANCE then calculates the density of each animal species, based on the rate at which sightings decrease with distance from the transect

4. Camera traps Digital cameras triggered remotely by motion and heat
Placed along trails where mammals are easily photographed
Capture images of shy and elusive animals, and crepuscular and nocturnal animals
Many animals are not observed on transects, either because they are principally nocturnal or because they are well camouflaged and can sense approaching observers and evade detection.
Camera traps are digital cameras are equipped with infrared sensors that take images when triggered by heat or motion.
They are housed in waterproof casing, and are attached to trees at a height where terrestrial mammals will trigger the sensor.
Heat from small mammals will trigger the camera up to a distance of roughly 10m, although the limited extent of the light generated by the flash means clear pictures cannot be taken at night beyond a distance of about 6m.
Images (L to R): collared peccary, puma, Brazilian tapir and common opossum

5. Survey methodology
Cameras placed in pairs on trails of the Pacaya Samiria Nature Reserve, Peru, for 3 months
Cover an area of approximately 50km2
Checked weekly to collect images/change batteries
Counts calculated per camera trap days
Between the months of June and August in 2009 and 2011, forty camera traps were set in the periodically flooded Várzea forest of the Pacaya Samiria Nature Reserve along transects that covered an area of approximately 50km2.
The cameras were set along transects on trees at a height of c. 1m, and were checked once a week to download the images and to replace batteries as necessary.
Camera traps were set in pairs on either side of trails so that for mammals with unique markings, such as the coat patterns of jaguar, individuals could be identified.
The relative changes in the counts of camera trap images of various species can be used to establish populations are faring from year to year.
In order for the number of images of each species to be comparable between years, they must first be standardised to give the number of images per 1000 camera trap days.

6. Research questions
Based on the camera trap images, how did the sizes of the red brocket deer and ocelot populations change between 2009 and 2011?
How might any changes in red brocket deer and ocelot numbers be related to the extreme high water levels of 2009?
Based on the camera trap sightings, how did the numbers of red brocket deer and ocelot present in the Pacaya-Samiria reserve change between 2009 and 2011? Were any differences in camera trap capture rates for the two species over this time statistically significant?
How might the changes in the number of sightings of red brocket deer and ocelot be the result of the extreme high water levels during the flooded season of 2009?

7. Summary Conclusions Summary Conclusions
Captured images of red brocket deer fell significantly between 2009 and 2011
Captured images of ocelot increased significantly between 2009 and 2011
The high water levels may explain the population size changes
reducing food availability and increasing competition and predation for the deer
increasing prey density for the ocelot
Numbers of both red brocket deer and ocelot differed significantly between 2009 and 2011.
For the red brocket deer, the number of sightings decreased from sightings per 1000 trap days in 2009 to sightings per 1000 trap days in 2011.
The ocelot numbers showed the reverse trend; number of sightings increasing from a relatively low 9.09 sightings per 1000 trap days in 2009 to sightings per 1000 trap days in 2011.
Both Chi-square (χ²) tests produced p-values < 0.05, showing that there was a less than 5% probability that the observed differences between 2009 and 2011 occurred by chance. The changes in camera trap image capture rates of red brocket deer and ocelot between the two years were therefore both statistically significant.
The high numbers of red brocket deer recorded in 2009 likely reflect a large, stable population of animals being forced into decreasing areas of land as the water levels rose to unprecedented heights. The decrease seen in red deer sightings two years later can potentially be attributed to higher mortality rates from the extreme reduction in available vegetation, increased competition with other herbivores and an increased in predation from big cats such as jaguar and puma that would have reduced the size of the population.
In contrast it appears that the ocelot benefitted from the extreme high waters. The increased density in their prey concentrated into a smaller area potentially contributed to the increase seen in the ocelot population, reducing mortality rates and enabling more offspring to be reared.