Open data · University of Illinois · 1971–1997
The Getz Vole Trapping Dataset
For twenty-five years, Lowell Getz and his crews live-trapped wild prairie voles and meadow voles every month in alfalfa, bluegrass, and tallgrass prairie fields outside Urbana, Illinois. Every animal was individually marked, so each record says who was caught, in which trap, and when. This page lets you explore those records directly.
How the study worked
Voles were never kept in enclosures. The study fields were open and unfenced, so animals were born, moved in, moved out, and died on their own. Getz's crews sampled them by setting live-traps on a fixed grid, marking every animal, and releasing it where it was caught.
Example: "15 captures in 6 sessions" means the animal was trapped 15 times across 6 different months, about 2–3 times in each three-day session.
Where and when: the 19 grids
Grids sat in two places 6 km northeast of Urbana: the university's Phillips Tract and Trelease Prairie. Alfalfa fields were replanted as they aged, so the five alfalfa grids are mostly successive fields, not parallel ones. Some grid numbers re-use the same ground: bluegrass grid 6 was later split into 9, 10, and 11. Solid bars are the unmanipulated controls that make up the boom-and-bust chart; outlined bars are experiments. Click a bar to open that grid.
Every session on every grid
Each cell is one grid in one month. Brighter cells mean more animals were caught, and blank cells mean the grid wasn't trapped that month. Click a cell to open that grid and month.
Twenty-five years of boom and bust
Population density (animals known alive per hectare) in the unmanipulated grids, stitched into one monthly series per habitat. Each panel has its own y-scale. Click anywhere on a panel to open the grid behind that month.
Arrivals and departures
Populations rise when new animals arrive (born on the grid or moving in) faster than known animals disappear (dying or moving away). Trapping can't tell births from immigrants, or deaths from emigrants, but it shows how fast animals turn over.
How long animals stay
Share of animals still known alive (caught again later) at each month after their first capture.
When new animals appear
Share of the animals caught in each calendar month that were being caught for the first time.
Getz's experiments
Alongside the long-term record, Getz ran field experiments. Some grids got supplemental food (rabbit chow at every station). On others, one species was removed every month, to see how the remaining species did without its competitor. Each panel compares a treated grid with a nearby unmanipulated control over the same months.
Inside a trapping grid
Each row below is one marked animal. Its line runs from first to last capture, and each mark along it is one capture. Play through the sessions to see who was caught where. Click any animal, in the rows or on the map, to follow just that animal around the grid.
Caught together this session
Social life: who shares a trap
Getz's wooden traps could hold several voles at once, so two voles found in the same trap on the same check were physically together. Prairie voles are famous for male–female pair bonds, while meadow voles are usually described as solitary. Here the two species are side by side, with males in blue and females in red.
How often voles are caught with the opposite sex
Share of animals each session found in the same trap, on the same check, as a vole of the opposite sex, plotted against population density so that crowding alone can't explain the difference.
Who is in the trap together
Traps holding exactly two voles of the same species. If voles entered traps at random, about half of these pairs would be male + female (tick mark).
How many partners
Breeding animals in each session: how many different voles of the opposite sex they had as partners (as chosen above).
Heavier males, more females?
Breeding males by body mass: share with at least one female partner in a session (solid) and with two or more (dashed). Captures per session barely change with mass, so heavier males aren't simply caught more often.
Do pairs stay together?
Male–female pairs where both animals were caught again the next month: how often they were caught together again. They are compared with neighbors (pairs that shared a trap but weren't caught together) and with any male–female pair on the same grid.
Coming back to the trap
Being live-trapped should be stressful, so you might expect voles to learn to avoid traps. They didn't. They came back to the same traps, and were caught more often, not less, as they gained experience. Small-mammal trappers call this being "trap-happy"; a dry wooden box with cracked corn inside is a good deal for a vole. Only grids with exact station layouts are used here.
Where the next capture happens
Distance from each capture to the same animal's previous capture.
Do they learn to avoid traps?
The same animals followed through their first five months of being caught: average captures per three-day session.
Back in last month's trap
Share of months in which an animal's first capture was in the same trap as its last capture the month before.
Breeding calendar
Each cell is one month on the unmanipulated grids: the share of females caught that were pregnant (judged by palpation), and the share of males in breeding condition (testes descended). Blank cells had fewer than five animals of that sex.
Body mass
Every capture on the monthly grids was weighed to the nearest gram. Each animal counts at most once per session, using its average mass that session, so frequently trapped animals aren't over-counted. These are descriptive summaries of the raw records, not fitted models.
Through the year
Median mass by calendar month for males (blue) and females (red). Bands hold the middle 50% of animals; with one sex selected, a lighter band adds the middle 80%.
Share in breeding condition
By reproductive condition
Distribution of mass for each condition recorded at capture: males with testes descended or females with an open vulva count as reproductive, and pregnancy was judged by palpation.
Growth after first capture
Each faint line is one animal's mass at each session after it was first weighed; bold lines are the median for males (blue) and females (red). Animals that enter the grid small grow fast for about two months and then level off.
About the data
Trapping protocol in detail
Each grid had one wooden multiple-capture live-trap per station, 10 m apart. Each month, traps were prebaited with cracked corn for two days, then set in the afternoon and checked at about 08:00 and 15:00 on each of the next three days. Voles were toe-clipped at first capture for identification, and at every capture the crew recorded species, animal number, station, sex, reproductive condition, and body mass. Some grids were experimental: one species was removed every month, or the grid got supplemental food (rabbit chow in a bottle at every station). The twice-weekly nest trapping (Alfalfa II–III, March 1982 to May 1987) put four or five traps around each known nest from 06:30 Monday to 09:30 Tuesday and again Thursday to Friday. Getz estimated it caught more than 99% of the animals present.
How this explorer processes the files
- Source: Getz's 38 "original field trap records" files (one per species per grid), which he had already cleaned, plus the two twice-weekly Alfalfa data files. The only further changes are noted below.
- An individual is a species + grid + animal number. Getz re-used number sequences over the years, so if the same number goes uncaught on a grid for more than 300 days, it is treated as a new animal.
- Nest captures are linked to a grid animal with the same number when they fall within 60 days of that animal's grid captures. Recorded sex agrees for 95% of linked animals. Unlinked numbers become "nest-only" animals, marked with an n in their ID.
- Capture dates more than 30 days from the rest of their session (about 50 typos) are replaced with the session's median date.
- Density is the minimum number known alive. The 25-year means it gives match Getz et al. (2001): bluegrass prairie voles 19.6/ha vs. 18 published, bluegrass meadow voles 13.5 vs. 14, tallgrass prairie voles 8.0 vs. 8.
- Animals taken off removal grids are counted in the totals but not drawn as lifelines.
- Trap checks are coded morning (AM) or afternoon (PM). A third, undocumented code appears only in Alfalfa II and III from 1982 to 1987, the years of the social-behavior study, whose grid sessions added a 20:00 check. It is shown here as "Eve", probably that evening check.
- Grid positions are rebuilt from Getz's station codes and scanned grid maps. Alfalfa and the Phillips Tract grids are exact. The Trelease Prairie grid (an irregular L shape) and the BG Control grid, which changed layout twice, are approximate.