Lek ecology of male greater sage-grouse in Carbon County, Wyoming

Greater sage-grouse (Centrocercus urophasianus, hereafter �sage-grouse�) have experienced range-wide population declines for several decades, and as a result they were considered warranted for listing under the Endangered Species Act in 2010. Therefore, wildlife managers need to understand how sage-grouse breeding behavior influences long-term reproductive success, and should be able to accurately relate sage-grouse lek count data to population sizes. Behavior during the breeding season, such as how frequently males visit their leks or move among leks, could relate to a male sage-grouse�s ability to establish dominance at his lek and mate, so it is important to understand how these behaviors may change with environmental conditions. Additionally, some males are not seen by observers performing lek counts, and it is necessary to understand why the individual may not have been seen and counted so managers can improve lek count protocols to maximize detection and relate lek count data to population abundance estimates. We investigated factors influencing the probability of attending a lek (Chapter 1), daily probabilities of moving among leks (i.e. �interlek movements�, Chapter 2), and detection probabilities for males during lek counts (Chapter 3). We monitored GPS-PTT radio-equipped male sage-grouse (2011 n = 22, 2012 n = 36, 2013 n = 59, 2014 n = 55) throughout the spring breeding season to assess attendance and interlek movements, and included additional males equipped with VHF transmitters (n = 188) for assessing detectability. We performed lek counts on 57 leks in Carbon County, Wyoming and mapped the boundaries of all active leks (n = 20 � 33 per year). We examined attendance in three ways: at the hourly scale by 1) individual male lek attendance and daily by 2) daily probabilities of lek attendance and 3) visitation rates per lek. We used logistic regression to evaluate whether or not a male attended a lek (i.e. had a GPS-PTT transmitter location within lek boundaries) during an hour or day, and negative binomial regression to determine the number of locations a male had within lek boundaries during a day. We examined interlek movements by modeling transitions among leks as a function of their covariates using a multi-state mark-recapture model, assuming constant survival. We considered 3 states, including Lek, Lek�, and Dead, which allowed us to estimate survival and movements from an initial lek to any other lek. Finally, we estimated male detectability during lek counts using sightability surveys, in which two observers, one with access to telemetry equipment, located marked males during lek counts. We then compared males that were seen by both observers to males that were only seen by the observer with telemetry equipment. We used logistic regression to identify factors influencing detection, and applied lek specific detection rates to lek counts to estimate the number of males per lek, including males unseen by the observer. Average daily attendance rates per lek ranged from 16.1% in 2011 to 82.0% in 2014, with high annual variability. This challenges use of lek counts as an index to population size because a good index is temporally consistent, but we observed substantial changes from year to year in probabilities males that are not detected on leks because they are not attending leks. Males were 2-3 times less likely to attend a lek on days with 0.5 � 0.75 cm precipitation in all years and all metrics of attendance, and about 2 times less likely to attend a lek on days following 0.5 cm precipitation. Date also predicted attendance, with peak dates of attendance ranging from April 7 in 2012 to May 13 in 2011. Broad time-scale precipitation and weather patterns likely shifted the peak date of attendance, with males attending earlier in the season when conditions were dry and warm in 2012, and later in the season when conditions were wet with high snowpack in 2011. Attendance also generally decreased with higher wind speeds, but the effects were not as strong as for precipitation or date. Lek counts should be avoided during precipitation and the day following precipitation due to lower lek attendance. Males had a 1.04% - 2.22% daily probability of moving to a new lek on any given day, demonstrating high daily lek fidelity as expected for a lek-breeding bird. However, the yearly probability of moving to a new lek ranged from 38.6% - 69.9% per year, suggesting many males may make at least one interlek movement at some point during the breeding season. Males that attended a lek every morning were 28 times less likely to move than males that rarely attended leks, and males with greater mass were also 15 times less likely to move than males with less mass. Dominant males with high attendance and mass were less likely to move to a new lek possibly because they established themselves at their preferred lek and therefore would be more likely to mate. Males were 5 times less likely to move during a day with 0.5 cm precipitation than with no precipitation, but 2 times more likely to move on the day following precipitation. Males were 4-12 times more likely to move at the beginning of the lek season, and 12 times more likely to move to a high elevation lek than a low elevation lek. Males may display at low elevation leks and move to higher elevations in the beginning of the season as precipitation declines and snowpack melts at the higher elevation leks. Additionally, males may move down in elevation following precipitation at high elevation leks, which can potentially bias lek counts on the day following precipitation. Male sage-grouse were most likely to be detected on leks with shorter sagebrush and higher snow cover. The average detection rate across all leks was 87%, and there was little variation from lek to lek in their lek-specific detection rates (77 � 93%), suggesting the lek count is an appropriate index to population size from lek to lek because it is spatially consistent. Lek count protocols already in place are sufficient, and cannot be improved to maximize detection. When accurate population abundance estimates are necessary, sightability methods can be used to determine detection rates at leks. Throughout all chapters, lek ecology was strongly influenced by weather, and understanding factors affecting attendance, interlek movements, and detection allows managers to estimate abundance from lek count data. Attendance was lower with precipitation, and interlek movements and detection of males on leks also changed at multiple time scales with precipitation and snow. Daily attendance and interlek movement rates can be used to predict when males are most likely to be present and available for detection on a lek during a count, which could be combined with sightability detection probabilities for accurate abundance estimates incorporating availability for detection and detection during counts.

Name thesis:degree_name

M.S.

Level thesis:degree_level

Masters

Discipline thesis:degree_discipline

Fisheries and wildlife sciences (MU)

Grantor dc:publisher

University of Missouri--Columbia

Year dc:date.issued

2014

Author dc:creator

• Fremgen, Aleshia Lynn

Advisor dc:contributor.advisor

• Millspaugh, Joshua J.

Language dc:language.iso

eng, English

Handle dc:identifier.uri

https://hdl.handle.net/10355/48256

OAI identifier oai:identifier

oai:mospace.umsystem.edu:10355/48256