Sheffield et al. 



JESO Volume 139, 2008 



nesting tubes of each group were then placed into a wooden laminate nesting block with 81 

 tunnels (arranged in a 9 x 9 pattern), each tunnel approximately 0.85 cm in diameter and 

 15 cm deep (i.e., able to house the nest liners) (Fig. la). The nesting tubes were randomly 

 selected and placed into the inner 49 nesting tunnels (i.e., a 7 x 7 pattern) of each nesting 

 block until all were accounted for. 



FIGURE 1 . A) Nest used for Osmia lignaria overwintering studies, showing B) data logger 

 probe position. 



Winter Survival 



The nesting blocks were randomly assigned to the two wintering conditions, 

 resulting in six blocks from each population in each wintering site. Data loggers (Onset; 

 HOB07 H8 Pro Series model H08-031-08 and HOB07 H8 4-Channel model H08-006-04) 

 with external probes were used to record ambient temperature (°C) at each of the wintering 

 sites (six channels per site) and within the nesting tubes (six channels per site, with three 

 allocated to the Nova Scotia and Utah populations) (Fig. lb). 



In the spring of 2002, bees wintered inside were slowly warmed to 25°C, and 

 emerged bees from each block were collected. Following the emergence period, nesting 

 tubes were examined for dead un-emerged bees (i.e., adult bees which were still inside their 

 cocoons and presumably died during wintering). Bees that had chewed out of their cocoons 

 and died due to nesting tube blockage were included in the analysis as "live bees" as it was 

 apparent they survived the winter. The number of cocoons containing dead larvae or pupae 

 were subtracted from the starting total as death would have occurred prior to winter (these 

 made up a very low proportion (<1%) of the population). 



Bees wintered outside were allowed to emerge naturally to determine emergence 

 phenology of this species in Nova Scotia. Bees emerging daily from each nesting block 

 were collected with numbers and sex recorded. Following the emergence period, nests were 

 examined for un-emerged adult bees as above. 



This experiment was set up and analysed as a multi-factorial design; the three fixed 

 factors of interest were 1) wintering site (inside at 4°C, 50-70% RH versus outside ambient 



6 



