Patterns of bee diversity in the Madrean Sky Islands: elevational habitat gradients, isolation, and gene flow

Methods

Summary

Our project will target six mountain ranges: the Chisos, Guadalupes in Texas, the Chiricahuas, Huachucas, and Pinaleños in Arizona, and an undisclosed range in Mexico. Each range will have from 5-7 collection sites comprising high-elevation (>7,500 ft) meadows and open pine forest, mesic canyons, and springs. Site selection will be conducted in consultation with National Park Service and Forest Service staff and by using GIS and remote sensing data.

After the collection sites have been chosen, our sampling protocol will be relatively straightforward. While the specifics of data collection at each site is likely to be modified as we gather more preliminary information, we are currently planning to establish a series of 2-4 50 meter transects at each site. Field staff will collect bees by walking each transect for a total of 15 minutes of collection time (time spent labeling specimens, logging data, and identifying plants will not be counted). There will be two collection events a day at each site: early-mid morning and mid-late afternoon. Before collection begins in the morning and between collection events at mid day, flowering plant species will be identified and the amount of floral resources will be collected at 5 quadrats (10 meter intervals) along each transect. Identifying plant species at the time of sampling will ensure greater accuracy in recording floral host-plant for the bees collected. Plants that cannot be identified in the field will be collected and sent to an expert for positive identification.

Challenges

If this research is so important to bee conservation, you may ask, then why hasn't it already been done? The answer lies somewhere between the increased attention that bees and other pollinators have received in recent years and the very real threat of losing even more undocumented biodiversity to the effects of climate change.

But a very important factor to consider is also the challenging nature of working in the mountains of the southwest US. The remoteness and long distances between mountain ranges mean that preparation is even more critical for a successful research project.

Our preparation for field work will begin with training the volunteer field assistants before work begins so that we can address any potential inconsistencies in the interpretation of protocol and collection methods. Our time at each site will be limited and we want to ensure it is used most efficiently and with greatest potential for replication.

Coordination with staff at National Park Service units and Forest Service ranger districts will also give us critical insights in terms of potential collection sites, their accessibility, and timing of peak bloom periods. Of equal importance will be the use of historical records to assess the breadth and depth of sampling that has already taken place across the region so we can better target our efforts.

Finally, we also anticipate the omnipresent challenge of unpredictability in ecological field work (or the challenge of nature not behaving, as we have come to call it). Despite our best efforts, we cannot think of or plan for everything. To address this, we have built some redundancy into our sampling design, aiming to sample slightly more intensively so that a reduction in data due to unforeseen weather or accessibility issues, for example, won't mean a subsequent need to dramatically alter our analytical approach.