Factors Affecting the Thermal Environment of Agassiz's Desert Tortoise (Gopherus agassizii) Cover Sites in the Central Mojave Desert during Periods of Temperature Extremes
Agassiz's Desert Tortoises (Gopherus agassizii) spend >95% of their lives underground in cover sites that serve as thermal buffers from temperatures, which can fluctuate >40°C on a daily and seasonal basis. We monitored temperatures at 30 active tortoise cover sites within the Soda Mountains, San Bernardino County, California, from February 2004 to September 2006. Cover sites varied in type and structural characteristics, including opening height and width, soil cover depth over the opening, aspect, tunnel length, and surficial geology. We focused our analyses on periods of extreme temperature: in summer, between July 1 and September 1, and winter, between November 1 and February 15. With the use of multivariate regression tree analyses, we found cover-site temperatures were influenced largely by tunnel length and subsequently opening width and soil cover. Linear regression models further showed that increasing tunnel length increased temperature stability and dampened seasonal temperature extremes. Climate change models predict increased warming for southwestern North America. Cover sites that buffer temperature extremes and fluctuations will become increasingly important for survival of tortoises. In planning future translocation projects and conservation efforts, decision makers should consider habitats with terrain and underlying substrate that sustain cover sites with long tunnels and expanded openings for tortoises living under temperature extremes similar to those described here or as projected in the future.Abstract
Conceptual diagram displaying potential effects of climate change on Desert Tortoise populations and responses that may occur. Potential increases or decreases are indicated by mixed dashed lines. Here we pose the possible question whether climate change could lead to local extirpations if available cover sites inadequately provide relief from increased drought conditions.
Map of Soda Mountains study plot, San Bernardino County, California, showing geological units and cover-site locations. Map base layer was a 1-m NAIP image downloaded from USGS National Map Viewer.
Diagram of cover-site characteristics (A,B) and temperature probes used (C–E) at the Soda Mountains study plot, San Bernardino County, California. Characteristic abbreviations are S: soil cover, H: opening height, W: opening width, and L: probe length. Three models of probes used: (C) HOBO Temp Series loggers measured temperatures inside the tunnel (Tin), (D) HOBO Pro Series temperature probe measured temperatures at opening (Tm) and inside and (E) HOBO Outdoor/Industrial 4-Channel External Series measured temperatures at four locations (T1–4) on the soil surface beneath the shrub canopy.
Average monthly maximum and minimum temperatures calculated from daily temperature extremes recorded by probes placed on the soil surface under a white-bursage shrub canopy (Ambrosia dumosa) at the Soda Mountains study plot, and collected by a NOAA weather station in Baker, San Bernardino County, California. Probes were placed at four cardinal directions (N, S, E, W). Gray boxes indicate periods of summer (July 1–September 1) and winter (November 1–February 15) extremes. Error bars represent ± standard error.
Regression tree models generated for (A) maximum summer and (B) winter minimum temperatures, and (C) summer and (D) winter temperature fluctuations recorded inside tunnels of 30 Desert Tortoise cover sites at the Soda Mountains study plot, San Bernardino County, California. The number of cover sites is shown in parentheses and r2 of the split is in bold.
Relationship between probe length and log10 temperature fluctuations ([A] summer and [B] winter), and temperature extremes ([C] summer maximum and [D] winter minimum) inside Desert Tortoise cover sites at the Soda Mountains study plot, San Bernardino County, California. Solid lines represent least-squares linear regression for all data points. The dashed and dashed-dotted lines represent 95% confidence and prediction intervals, respectively.
Contributor Notes
