Protocol efficiency and performance comparisons
In BC, species accumulation as a function of hours of effort was indistinguishable between ARU-only protocols and one- and two-point count rounds plus ARU sampling (Fig. 3A). This was because point counts did not contribute the total survey cost but, as shown above, they also did not contribute novel species to the accumulation curve. Three-point count rounds and a mixed method that included point counts at this intensity were the least-efficient sampling protocol in BC due to the increased cost associated with a third site visit. Surprisingly, in BC, a single ARU count/site detected more species than two point counts/site in the subalpine and more than three point counts/site in the alpine and upper montane, for less effort (13 vs. 14 hrs and 15 vs. 24 hrs, respectively; Fig. 3A).
In Chile, ARUs alone were less efficient than point counts alone and less efficient than mixed methods due to the fewer species detections acquired by ARUs. This was particularly notable in the Chilean alpine, where a single point count/site detected more species than 10 ARU counts/site and two point counts/site detected more species than were detected at our maximum ARU effort of 15 counts/site, for less effort (4.5 vs. 22 hrs and 9 vs. 29 hrs, respectively; Fig. 3B). While species accumulation curves of mixed methods showed a large degree of overlap, a minimum of two point counts/site supplemented with ARUs appeared to be the best methodology for the subalpine and upper montane in Chile, and three point counts/site in the alpine boosted species detections enough to warrant the additional visitation cost (Fig. 3B).