1 Rusty Blackbirds in the Northern Forest: Breeding Season Status and Habitat Associations at Local and Landscape Scales Talk created by Stacy McNulty, SUNY ESF for Vermont Monitoring Cooperative meeting, Dec , Burlington VT Stacy McNulty, SUNY ESF

2 Natural History Breeds in forested wetlands and bogs across North American boreal forest Winters in wooded wetlands and bottomlands in the southeastern U.S. Feeds on aquatic invertebrates Photos credits: D. Cote, S. Buckley,

3 Rusty Blackbird DeclineDeclined by approximately 90% since the 1970’s based on North American Breeding Bird Survey, Christmas Bird Count Difficult to study (remote, inaccessible habitats) Causes of decline still not clear Checklist (Greenberg and Droege 1999, Greenberg and Matsuoka 2010) International Rusty Blackbird Working Group (IRBWG) formed in 2005 Nearly extirpated from Adirondacks – only 1to a few nests located in (McNulty et al. 2015; AEC unpub. data)

4 Rusty Blackbird Rangewide DeclineRusty Blackbirds recorded in the U.S. during Audubon’s Christmas Bird Count,

5 Possible Causes for DeclineLoss of winter habitat to conversion, fragmentation Climate change (water levels, phenological, other) Acidification of boreal wetlands, mercury contamination Habitat conversion in boreal forest (timber harvest, reservoir formation, energy development) “Ecological trap” – predation on breeding grounds, social factors This bird is not Neotropical, so we can’t blame changes further south…or can we? Much habitat conversion and fragmentation has occurred in the SE US. The birds seem to be doing well in Alaska/western range, so the problem appears to be eastern N. Am. Birds. Edmonds et al. (2010) found mercury concentrations in breeding Rusty Blackbirds from New England and eastern Canada exceeded published minimum levels for songbirds, and mercury bioaccumulation in this Acadian forest population is among the highest reported at sites without a nearby industrial source Photos credits: S. Buckley, Garth Lenz,

6 Objectives Identify predators of Rusty Blackbird nestsModel nest survival as a function of habitat variables at multiple spatial scales Examine the relationship between cone cycles, predator populations and nest predation Model habitat selection at multiple spatial scales Inform future work on summer foraging ecology, behavior etc.

7 Study Areas NH Study Area ME Study AreaCoos County, NH (Umbagog-Androscoggin River) and Moosehead Lake, Maine regions Mostly private timberlands, with Umbagog National Wildlife Refuge

8 Nest Monitoring and HabitatLocated and monitored nests in ME and NH in 2011 and 2012 Installed covert, passive infrared (PIR) cameras < m from nests Measured vegetation in 5-m radius plots around nests (post-nesting) Control plot 50 m from nest in random direction

9 Squirrel Surveys Broadcast surveys for squirrels at nests and in select mature stands Basal area measurements 90 m transects, passive/active broadcast squirrel calls with game caller

10 Landscape-Scale HabitatGIS: percent cover of forest and types within 500-m radius of nests, distance to nearest road Data on stand area, species composition, etc. from landowners Wetland data from National Wetland Inventory (NWI) Database For each study site, we generated a number of random control points within 750 m of a road (the maximum distance from a nest to a road) equal to the number of nests monitored. We calculated the percent area of different stand and wetland types within a 500-m radius of each nest and random point. We chose this distance because of its relevance to Rusty Blackbird spatial ecology during the breeding season. The mean home range size of 37.5 ha for Rusty Blackbirds (Powell et al. 2010b), if circular, would have a 347-m radius. Home range sizes were quite variable, however, ranging from ,10 ha (178-m radius) to over 150 ha (691-m radius; Powell et al. 2010b). Therefore, 500 m seemed a relevant scale at which Rusty Blackbirds might respond to landscape features.

11 Statistical Analyses Nest Habitat SelectionNest-Patch-Scale (5 m): matched pairs logistic regression (MPLR), odds ratios Landscape-Scale (500 m): logistic regression, odds ratios Program R Nest Survival Logistic exposure models, 27 exposure days Two spatial scales: Nest Patch Scale (5 m) Landscape Scale (500 m) Program MARK Comparison of Cone/Squirrel Abundance Chi-square tests Survival exposure models (Matsuoka et al. 2010, Powell et al. 2010) nonparametric Mann-Whitney U tests to compare habitat variables around nests in Maine vs. New Hampshire at each spatial scale. Nest (5mscale): MPLR 1:1 case–control design to test all univariate and multivariate combinations of the 3 habitat covariates, comparing nests to random points in the pooled NH + ME dataset (n = 72). Stand: n = 56

12 Nest Habitat SelectionLocated 72 nests in NH = 43 ME = 29 88% (n = 63) nests in harvested areas - wetlands and uplands 9 in unharvested areas wetlands only Over 90% of nests within a few meters of an edge with an open habitat Wetland Skid road Tree gap

13 Nest Habitat SelectionNest-patch-scale Basal area at nests = small, dense, young conifers Probability of selection increased by 65% with each 5 m2/ha increase in small softwoods Probability of selection decreased by 39% with each 10% increase in canopy cover Canopy closure measured with ocular estimation at 20 points within a 5m radius of the nest (or center of random point) See Table 5 in Condor paper

14 Nest Habitat SelectionLandscape-scale Probability of selection increased by 46% with each 10% increase in young softwood cover Probability of selection increased with each 10% increase in wetland cover by site: ME = 2.46 times higher NH = 5.47 times higher Table 5, Condor paper – model-averaged

15 Nest Habitat Selection: SummaryDifferent factors driving selection at different spatial scales, decoupled Foraging requirements (wetlands) at landscape scale Nest safety (dense conifers) at nest-patch scale Greater plasticity at landscape-level than nest patch-level - as long as certain key features present, matrix appears of little importance Broad (regional) application of habitat management plans Habitat result of interacting disturbances, dynamic in space and time - need to look beyond “harvest history”

16 Cone Cycles Photo credits: jfoto.ca, en.academic.ru

17 Predators and Cones Proportion of squirrel surveys with positive detections: ME Nests: ME Mature Stands: 2 = 6.50, p = 0.01 2 = 6.75, p = 0.009

18 Predators Monitored 29 nests with cameras(21 in ME 2011/12, 8 in NH 2012) 34,446 photos 8 predation events documented by 4 confirmed, 1 suspected species Red squirrels most frequent predator, but only in 2012

19 Nest Survival – Nest ScaleOverall survival: % (n = 65) Importance of total basal area No effect of timber harvesting Distance to road a factor in 2011, but not 2012

20 Nest Survival – Landscape-scaleIf there is an association between certain predators and roads, then it may be prudent to consider the potential effects of infrastructure associated with harvest operations, rather than focusing on stand treatments exclusively. 53 +/- 9% (n = 65) Year-dependent effect of distance to road 2011: failed nests were 35 m from road, successful nests were 170 m from road 2012: failed nests farther from roads No effect of timber harvesting Photo credits: S. Buckley

21 Conclusions Decoupling of habitat selection at different spatial scales Red squirrels primary nest predators, but not every year Nest survival likely fluctuates with predator populations, mast cycles Nest survival not related to harvest per se – roads? Relationship between timber harvesting and Rusty Blackbird nesting ecology complex Productivity not chronically low, not driver of decline Photo credits: S. Buckley, Richard Kent

22 Habitat and Species ConservationOther habitat factors Hydric soils Beaver influence “ Mappable” wetlands ? Ditches Vernal pools Seeps Puddles State status - RARE VT – State Endangered NH – Special Concern ME – Special Concern NY – High Priority SGCN From , Rusty Blackbird declined to 16% of sites occupied (Glennon; McNulty et al.) Disturbance can be good thing: windthrow remains an important form of local natural disturbance, and severe storms that can blow down large areas and create short, dense conifer regeneration are, if anything, becoming more frequent across the U.S. (Melillo et al. 2014).

23 Management RecommendationsMaintain stands of young (seedling/sapling) softwood, particularly around wetlands/hydric soils - can be small patches Retain snags as perches Sub-stand structure important

24 Current/Future ResearchForaging ecology Aquatic invertebrate emergence phenology and diet Habitat characteristics Occupancy analysis of wetland foraging sites Detectability analysis and methodology testing Binational cooperation via IRBTWG Habitat disturbance in core (boreal) breeding range Climate/hydrological change Could Northeast US become a migratory stopover region? Amanda conducted in min passive surveys for RUBLS at 60 sites; using North American Breeding Bird Survey (BBS) data, McClure et al. (2012) showed a northward retraction of >140km from the southern range limit over the past 40 years.

25 Acknowledgments Rusty Research TeamShannon Buckley Lüpold Thomas Hodgman Jonathan Cohen Carol Foss Amanda Pachomski Luke Powell Patti Newell Wohner Ray Ary Henning Stebbins Stefan Lüpold Sean Flint Field technicians: Linnea D’Amico, Sara Prussing, Joe Roy, Devon Cote, Kelsey Schumacher, Amasa Fiske-White, and Thomas Ruland Landowners: Plum Creek, Wagner Forest Management, Northwoods Management, Prentiss and Carlisle, Appalachian Mountain Club, Umbagog National Wildlife Refuge Funding and Logistical Help: Edna Bailey Sussman Foundation, Garden Club of America, Maine Outdoor Heritage Fund, USFWS State Wildlife Grant Program, New Hampshire Audubon, SUNY ESF

26 References Greenberg, R., and S. Droege On the decline of the Rusty Blackbird and the use of ornithological literature to document long-term population trends. Conservation Biology 13: Greenberg, R., and S. Matsuoka Rusty Blackbird: Mysteries of a species in decline. Condor 112: Powell, L., T. P. Hodgman, W. E. Glanz, J. D. Osenton, and C. M. Fisher. Nest-site selection and nest survival of the Rusty Blackbird: Does timber management adjacent to wetlands create ecological traps? Condor 112: Robertson, B.A., and R. L. Hutto A framework for understanding ecological traps and an evaluation of existing ecological evidence. Ecology 87:

27