1 Final Report To Tembec Industries Inc. December 1, 2008 in fulfillment of the contribution agreement (RC08-1569) between Tembec and UNBC INVENTORY OF ARBOREAL-FORAGING ANT COMMUNITIES WITHIN WILLIAMSON’S SAPSUCKER NEST AREAS IN THE EAST KOOTENAYS Prepared by: Robert J. Higgins Thompson Rivers University & Professor B. Staffan Lindgren University of Northern British Columbia 2 Introduction The Williamson’s Sapsucker (WISA), Sphyrapicus thyroideus, is a cavity nesting sapsucker that is known to occur in three geographically distinct areas in southern British Columbia, including one population in western larch forests of the East Kootenays near the city of Cranbrook (Figure 1). This bird was assessed as endangered by the Committee on the Status of Endangered Wildlife in Canada (COSEWIC) in 2005 and added to Schedule 1 under the Species at Risk Act (SARA) by the federal government in 2006. This led to the commencement of recovery planning for the species. Preliminary research (Gyug et al. unpublished) has identified ants (Hymenoptera: Formicidae) as a critical food resource linked to WISA fledging success. This is consistent with work by Wiebe (2005) in which increased foraging on epigaeic ants by the Northern Flicker was positively associated with fledging success. In the case of the WISA, ants are obtained from tree boles (Crockett 1975) as the ants move up and down the trunk while foraging. Beal (1911) described the WISA as being more dependent upon ants during the breeding season than any other North American woodpecker. Ants of the genus Camponotus (carpenter ants) were identified as comprising the majority of the ant diet of the WISA (Crockett 1975), although Gyug et al. (unpublished) has also found some species within the genus Formica, in particular Formica aserva, that may also be a frequent food source. Both Camponotus species ants and Formica aserva are known to be dependent, in turn, upon coarse woody debris (CWD) as a nesting resource (Hansen and Klotz 2005, Higgins and Lindgren 2005). Horn and Hanula (2008) reported that the short-term removal of CWD from experimental plots reduced the overall availability of arthropods to bole-foraging birds. This 3 supports the findings of Gyug et al. (unpublished) who found that WISA fledgling success correlated with ant availability, which in turn, correlated with CWD volumes. The main objective of this project was to inventory the arboreal foraging ant communities at WISA nests sites of varying nest productivity in order to determine the relative availability of CWD-associated and other ant species to the WISA in the East Kootenays. This project inventory will contribute toward a more comprehensive examination of the degree of dietary dependence of the WISA upon CWD-associated ants in southern British Columbia (Meggs et al. 2007). The specific objectives of this project were to: 1. Compile an inventory of ant species/morphotypes available to WISA on the trunks of live trees; 2. Quantify the diversity and abundance of ant species/morphotypes available to WISA; 3. Categorise the species/morphotype list by known or assumed habitat association (e.g. CWD, soil or rock-associated ants) Methods Site inventory data relating to coarse woody debris volumes, live tree densities, and fledgling success were previously collected by Gyug et al. (unpublished.) using Resource Information Standards Committee (RISC) guidelines as outlined in the FIA Activity Standards Document for Williamson Sapsucker Nests and Territories (MoE 2006). Canopy cover was measured as per RISC guidelines during trap installation. Sampling for ants followed, where possible, the Inventory Methods for Terrestrial Arthropods (MELP 1998), although the requirement for sampling upon tree boles necessitated a unique sampling design. 4 Sampling methodology followed Meggs et al. (2007). In this study we used trunk traps (Figure 2), to intercept ants moving along the surface of tree boles. Sampling occurred within a 16 ha area centred upon the nest tree. The size of the sampling area was chosen to correspond to the known minimum foraging area for the WISA given that adjacent nests are normally no closer than 450 m to each other (Guyg et al. unpublished). Centred on each of the seven identified WISA nests (Appendix 1), twelve trunk traps were positioned every 30 degrees at a random distance between 10 and 225 m. Traps were not placed closer than 10 m to the nest tree to avoid oversampling the area immediately adjacent to the nest. Each trap was placed on a tree bole just above a height of 2.5 m to avoid damage from cattle and held in place by two 13 mm wood screws. Traps were put in place in mid-June 2008, and removed approximately two weeks later. This trapping period was selected to coincide with active foraging by the WISA for ants during the WISA chick-rearing season,.. Samples collected following the removal of the traps were cleaned of debris and the ants sorted, identified, and then stored in 95% alcohol. As carpenter ants (genus Camponotus) have been reported to be of significance in the diet of the WISA (Gyug et al, unpublished) these ants were identified to species where possible (Table 3). Ants of the genus Formica were identified to the level of species complexes except where anatomical features made identification to species readily possible (e.g., Formica aserva). Other ants were identified to the level of genus. Taxonomy followed Bolton (2005). The number of individual ants of each species or morphospecies (i.e., taxa) in a trap was counted, however presence/absence was used for analysis. This was necessary because the unequal distribution of ants within their foraging area generally results in unreliable estimates of abundance. Traps placed on trails used by ants to access resources will accumulate large 5 numbers of ants while a trap in close proximity may collect few making abundance based analysis of individual traps of little value. The relative abundance of ant taxa at each site (i.e., WISA nest area) was determined by calculating the frequency of occurrence of each taxa from individual trap data pooled by site. Individual trap data were then pooled across all sites to determine the frequency of occurrence of each taxa for WISA nest areas in the East Kootenay region. Results include all species of ants collected, not just those suspected of being used as a food source by the WISA. 6 Results and Discussion Background data are available for each WISA nest site relating to fledgling success, productivity class, coarse woody debris volumes, live tree density, and canopy cover (Table 1). Table 1. Fledgling success, productivity class, coarse woody debris (CWD) volumes, live tree density and canopy cover data for seven WISA nesting sites in the East Kootenays. Site code Fledged Fledged Fledged Product- CWD Merchantable Canopy ivity Young Young Young volume tree (>17.5cm) Cover Class density 2006 2007a* 2008b (m3/ha)c (%)d (H=high; (trees/ha)c L=low)*** WNE003 Successful 5 Unoccupied H 15.7 129.8 6.3 WNE010 Unoccupied 4 Unoccupied H 9.5 176.0 13.5 WNE011 Unknown 5 Reoccupied H 0.6 98.6 13.8 WNE012 Unknown 5 Unoccupied H 9.4 97.1 12.9 WNE013 Unknown 4(3) Unoccupied H** 33.1 64.2 13 but nest within 100m occupied WNE014 Unknown 4(2) Unoccupied H** 11.4 86.7 12.6 but nest within 100m fledged 3 young WNE015 Unknown 3 Unoccupied L 21.4 100.7 7 a Ohanjanian et al. (2007) b Ohanjanian et al. (2008) c Gyug et al. (unpublished) d Tembec – Cranbrook Division *Two concerns relate to these data (Ohanjanian pers. comm.). First, nest observations taken on June 18, 2007 found only the young in WNE010 andWNE011 were old enough on the day of observation to meet FIA standards of successful fledgling (i.e., 22 days old). Second, nest observations of WNE013 and WNE014 induced early fledging and anthropogenic mortality of some young. Number in brackets represent successfully fledged after anthropogenic-induced mortality. 7 **Each of these nests were originally designated as low productivity in this project based on the post-anthropogenic fledged number. ***Nests in which 4 or more young successfully fledged were considered high productivity while nests with less than 4 were considered low productivity (Gyug et al. unpublished). A few trunk traps were not recovered (Table 2). This may have been due to errors associated with GPS accuracy when recording trap locations, errors in GPS accuracy when relocating traps, or disturbance by human or animal activity. Table 2. The number of trunk traps set and relocated at seven WISA nest areas in the East Kootenays. Site Code Number of traps set Number of traps recovered Reason for failure to recover trap WNE003 12 10 Unable to relocate 2 traps WNE010 12 11 Unable to relocate 1 trap WNE011 12 12 WNE012 12 12 WNE013 12 10 Unable to relocate 2 traps WNE014 12 11 Unable to relocate 1 trap WNE015 12 12 In total, 13 species or morphospecies of ants were captured in trunk traps at the East Kootenay sites (Table 3). Based on this initial determination, we anticipate that a full species determination would result in the identification of 20-30 species. Of the 13 taxa identified, five were carpenter ants in the genus Camponotus. Several of these species require large diameter stumps or coarse woody debris for nesting (Table 4), as their nests can become very large with up to 50,000 workers (Hölldobler and Wilson 1990). Preliminary examination of WISA fecal sacs by Gyug et al. (unpublished), found carpenter ants to be the most common ant present, although they also 8 tentatively identified Formica aserva as a frequent prey item. While this latter species can nest broadly in differing habitats (Table 4), it is most commonly located nesting within CWD, especially in forests.