TOPOGRAPHY AND BIOGEOGRAPHIC CHANGES OF MONTANE FORESTS IN THE “SKY ISLANDS” OF SOUTHEASTERN ARIZONA AND THE ASSOCIATED GENETIC DIVERGENCE OF SCAPHINOTUS PETERSI DURING THE QUATERNARY

Several subspecies of Scaphinotus petersi, a flightless beetle, are isolated from each other on separate mountain peaks > 2100 m.a.s.l. in southeastern Arizona; the beetles live in the cool, high, mesic montane forests, “sky islands,” separated by desert scrub. In prior genetic research on these and other sky island species, it has been hypothesized that the forest habitat was connected at lower altitudes during glaciopluvial conditions. The populations are thought to have diverged as warming and drying after the last glacial maximum (LGM, ~20 kya) or Younger Dryas (YD, ~10 kya) forced the forest upslope and caused habitat segmentation. We test these hypotheses by pairing: 1) molecular phylogenetics of Scaphinotus petersi subspecies and 2) a GIS analysis linking regional topography, forest type, lapse rates of temperature (T°) and precipitation (P), and paleoclimate records to model if or when forests were contiguous in the past.

Molecular clock analysis indicates that divergence between the beetle populations occurred between 25 and 260 Kya, depending on the particular combination of populations analyzed. Furthermore, the genetic data are consistent with genetic separation due to habitat segmentation as opposed to dispersal. A preliminary model relating forest hysometry to past and present T° and P indicates that LGM paleoclimate conditions were likely wet and cool enough for montane forests to coalesce at lower altitudes. However, the age of divergence of Scaphinotus populations is correlated with neither the distance of geographic separation (90-215 km), nor the forest biome altitude depression necessary to link populations (750-1000 m). Furthermore, common ancestors between some populations, and thus the maximum date of separation, long predate the LGM. The complex relationship between the paleoclimate and paleoecology record and the genetic divergence data indicate that 1) a simple model of upslope biome migration since the LGM or YD is not sufficient, 2) the paleobiogeography of beetle populations differs from their current forest habitat, and/or 3) there is a multi-stage pattern of localized beetle extinction and re-population throughout the Pleistocene and Holocene.