Wildlife Cameras at Saguaro National Park's Tucson Mountain District, 2022

A gray fox (Urocyon cinereoargenteus). Credit: National Park Service.

The National Park Service's Sonoran Desert Network (SODN) monitors medium and large mammals in parks across the Sonoran Desert and Apache Highlands. Begun in 2016, the project aims to detect biologically significant changes in mammal communities and populations over time, giving park managers reliable information at a range of scales. It uses remote wildlife cameras that trigger themselves, with sampling and analysis designed around management needs.

2022 at a glance

• 59 cameras in Saguaro National Park's Tucson Mountain District (TMD), recording from January 11 to February 26, 2022.
• 2,411 detections (photographs of animals).
• 10 mammal species, plus two groups identified only to family or genus; three bird species, plus 48 detections of birds not identified to species.

Why cameras, and why occupancy

Habitat loss, climate change and other human pressures are forcing many species to shift where they live, and even large protected areas such as national parks are not immune. Understanding how species respond needs data over broad areas.

Remote cameras are non-invasive and record many species at once over large areas. Animals with unique markings, such as jaguars, can be tracked individually, but most mammals can be identified only to species. For them the network uses occupancy modelling:

• Occupancy is the proportion of an area a species inhabits. A species photographed is present; one not photographed may be absent, or present but missed. Occupancy models account statistically for that imperfect detection.
• Covariates help explain why a species is found where it is. Some, such as elevation, are fixed during sampling and likely affect occupancy; others, such as temperature or rain, change daily and may affect detection — animals may be less active in great heat.

At least five years of pooled data are needed for reliable multi-year models. The network had just reached that point, and was developing its trend methods.

A wildlife camera at Saguaro National Park's Tucson Mountain District. Image from the National Park Service's page.

How the cameras were run

• Placement. The 59 unbaited cameras stand at fixed monitoring locations paired with the network's long-term vegetation and soil plots, so vegetation data can be used to explain wildlife patterns. Cameras have gone out in the same places and season every year since 2017, except 2019, because of the government shutdown.
• Operation. Triggered by motion or a sudden temperature change, the cameras were set up and retrieved following standard operating procedures, with the same settings each year.
• Sorting. Photos without an animal ("false triggers") were removed; the rest — "detections" — were catalogued in the Colorado Parks and Wildlife Photo Warehouse Database and exported for analysis.

In 2022, five volunteers and six interns from various parks and partner organisations helped deploy and retrieve the cameras.

Field crews deployed 59 wildlife cameras in the Tucson Mountain District in 2022. Image from the National Park Service's page.

What the cameras saw

• Badgers were an encouraging find: they matter ecologically but are uncommon, even rare, to see in the district.
• Dogs are uncommon in the park but not unheard of, with homes nearby.
• Annual totals of photos with animals have ranged from 1,994 to 4,054. Why is unclear; occupancy modelling of environmental factors should help explain it.

An example: javelina and elevation

Single-season data cannot show trends, but they can still be informative. The network modelled javelina occupancy for 2022 alone (a single-season, single-species model), grouping detections into seven-day periods. Only mammals at least 43 centimetres long — the size of a rock squirrel — and detected at more than 30% of sites were analysed, which excluded small rodents and rarer species such as the American badger.

Javelina occupancy rose with elevation: in the district's lower areas the probability of an area being occupied was low (25–50%), but above 1,050 metres it rose to more than 75%. Why is not yet known; higher ground may offer cooler, more secluded habitat than the hotter lowlands, which see more disturbance from roads, trails and people.

Javelina occupancy probability by elevation, Tucson Mountain District, 2022. Image from the National Park Service's page.

Why it matters

Occupancy is often used as a stand-in for abundance, so estimates pooled over years can show whether mammal populations are increasing, stable or decreasing. Modelling the effect of temperature and rainfall can also show how climate change may affect species.

Sources

• National Park Service, Sonoran Desert Network, "Wildlife Monitoring at Saguaro National Park's Tucson Mountain District: 2022." https://www.nps.gov/articles/000/sodn_wildlife-monitoring-saguaro-tmd-2022.htm
• The report's text describes all 2,339 mammal detections as identified to 10 species; in its table that total also includes the detections identified only to family or genus, or not at all. The table is given here.
• The report cites the occupancy-modelling literature (MacKenzie and others, 2002, 2006, 2018) and other studies; see its literature list.
• Rewritten in hubnx's own words.