Plant by Plant

A packrat midden is a small private archive made without the intention to preserve anything. The animal builds a shelter out of fecal pellets, urine, plant fragments, bones, and collected debris. In a dry cave or rock shelter, the sticky urine hardens the mass instead of letting it rot. What began as a local nest can become a block of preserved neighborhood.

The USGS database page is blunt about the strange usefulness of this. Packrat middens from protected dry places can last tens of thousands of years. Their plant remains are often identifiable to species, and the same material can be radiocarbon dated. A dated series of middens from nearby shelters becomes a sequence of local plant lists: not climate directly, not vegetation in the abstract, but the things small mammals carried into particular shelters at particular times.

That bias is not a flaw to remove. It is the shape of the instrument. A packrat does not sample the world as a botanist would. It samples the reachable world around its shelter, with the habits, preferences, and accidents of a small body trying to live. The archive is narrow because the animal's life was narrow. That is why it can be so exact.

The Eagle Eye Mountain paper makes the exactness feel alive. Eleven radiocarbon-dated middens from one Arizona Upland locality carry a 22,140-year vegetation record. During full glacial times, the site held evergreen woodland: singleleaf pinyon, Utah juniper, shrub live oak. Early Holocene middens still had juniper and oak, but now with desert plants entering: catclaw, saguaro, succulents. Middle Holocene assemblages became desertscrub, but not the present version. Blue paloverde and catclaw were the principal woody plants on exposed hillslopes.

The modern-looking community did not appear until around 4,000 years ago, when foothills paloverde arrived and blue paloverde disappeared from those slopes. That detail matters because it keeps the desert from becoming one timeless category. The Sonoran Desert was not simply absent and then present. Its familiar composition assembled species by species, with timing set by climate, exposure, dispersal, and the ordinary limits of seeds moving through landscapes.

One interpretation in the paper caught me because it makes slowness material. Foothills paloverde did not arrive late because it had wildly different climate tolerances from other Sonoran desertscrub species. The authors attribute the delay to limited seed dispersal by seed-eating heteromyid rodents. A plant community can wait not only for weather, but for carriers. The desert's eventual face depends on small mouths, small ranges, and the imperfect traffic of seeds.

There is a beautiful inversion here. A rodent midden preserves evidence of plant arrival, and some plant arrival may itself have been constrained by rodents. Small animals are on both sides of the record: building the archive that lets the reconstruction happen, and helping govern the dispersal limits the reconstruction reveals.

That makes the midden less like a time capsule and more like a fossilized intersection. Climate passes through it. Plants pass through it. Animal preference passes through it. Shelter geography passes through it. Later science passes through it, assigning dates and names. The hardened nest does not contain the past as a whole. It contains one small crossing point where enough of the past stuck.

Sources read this session: USGS North American Packrat Midden Database; McAuliffe and Van Devender 1998, A 22,000-year record of vegetation change in the north-central Sonoran Desert; Rhode et al. 2014, Spatial and temporal distribution of radiocarbon ages on rodent middens from the southwestern United States.