Responses ofNesting Sandhill Cranes to Research Activities and Effects on Nest Survival

We examined the response of nesting greater sandhill cranes (Grus canadensis tabida) to research-related activities at Grays Lake, Idaho, during 1997–2000. Data were collected on \u3e500 crane nests from 53 fields that were exposed to variable levels of research activity, from fields subjected to only periodic searches for and monitoring of crane nests (crane-only fields) to fields subjected to periodic searches for and monitoring of all waterbird nests as well as breeding bird surveys, nocturnal amphibian surveys, and other research-related activities (multi-use fields). For each nest, we calculated the visit rate to the field during the period the nest was known to be active (hereafter field visit rate) as the number of observer visits to a field while the nest was known to be active divided by the number of exposure days of the nest. The median field visit rate for all years was 0.188 visits/day, which would equate to 5.6 visits over the course of a 30-day incubation period. Length of field visits for most activities averaged 93–155 min. Of known responses, most (83.7%) cranes were flushed, but actual frequency of flushing (considering unknown responses as not flushed) was likely ~30%. Frequency of flushing for known responses was highest during nest searches (89.4%) and nest checks (87.9%) and lower during bird surveys (68.1%) and other activities (54.3%). Half of cranes that were observed to flush did so when the investigator was between 20 and 80 m away (25 and 75% quartiles, respectively; median = 32.5 m). Median flushing distances tended to be greatest for nests located in very short vegetation, but the response differed between cranes nesting in multiuse fields and cranes nesting in crane-only fields. We found no relationship between flushing distance and number of field visits. We examined the effects of field visit rate and year on nest survival using logistic-exposure approach and Akaike’s Information Criterion to evaluate support for 3 models. The model including field and year was most supported by the data; we found no support for field visit rate as an important variable explaining nest success