Atlantic Co-Operative Wildlife Ecology Research Network

LOCAL AND LANDSCAPE EFFECTS OF INDUSTRIAL FORESTRY ON THE REPRODUCTIVE ACTIVITY OF FOREST SONGBIRDS IN NORTHWESTERN NEW BRUNSWICK, CANADA

John S. Gunn Ph.D. Candidate

This project is part of ACWERN and the Canadian Sustainable Forest Management Network of Centres of Excellence (SFM NCE). The research is being conducted in co-operation with Fraser, Inc. on Freehold Land west of the Tobique River in northwestern New Brunswick. Co-advisors for my work are Marc-André Villard of the Université de Moncton and Tony Diamond of the University of New Brunswick.

Project Background

The impact of human activity on the landscape is hardly more visible than in the industrial forests of New Brunswick. With nearly 85% of the Province covered in forest, products from the industrial forests play a major role in the economy of the region. The forest products industry in New Brunswick was responsible for about 27,000 jobs in 1993, and nearly $2 billion in exports in 1994 (CFS 1996). Harvest volume in 1994 alone was 9.2 million m³ (or 92,702 ha harvested) (CFS 1996). A long history of an economy with a dependence on forest products has left a forest landscape with very few places untouched by the activities of timber harvesting. Continued economic pressure on the forest requires us to understand how these harvesting practices are affecting ecological processes such as the population dynamics of forest birds. Birds are an extremely diverse and highly visible component of forest ecosystems. Associations with virtually every strata and development stage of the forest make birds a useful tool to study the effects of forest management on ecological processes. This research is an investigation of the response of forest birds to the patterns on the landscape created by the harvesting and regeneration forest stands.

Below is a general abstract as well as an abstract for each primary chapter (4) from my dissertation. I will be defending this dissertation on December 12th 2003.

Chapter 1 - PLAYBACKS OF MOBBING CALLS OF BLACK-CAPPED CHICKADEES AS A METHOD TO ESTIMATE REPODUCTIVE ACTIVITY OF FOREST BIRDS

Chapter 2 - FACTORS INFLUENCING THE VARIABILITY OF ANTI-PREDATOR MOBBING BEHAVIOR IN FOREST BIRDS

Chapter 3 - IS DENSITY A MISLEADING INDICATOR OF HABITAT QUALITY? THE RELATIONSHIP BETWEEN FOREST SONGBIRD ABUNDANCE AND REPRODUCTIVE SUCCESS

Chapter 4 - LANDSCAPE AND LOCAL EFFECTS OF INDUSTRIAL FORESTRY ON THE REPRODUCTIVE SUCCESS OF EIGHT FOREST SONGBIRD SPECIES IN NORTHWESTERN NEW BRUNSWICK

In the northern hardwood forest of New Brunswick, industrial forest management affects within-stand vegetation and the landscape structure of the forest mosaic. Understanding the effects of industrial forestry on songbird populations requires the investigation of songbird reproductive success, in addition to abundance, on a landscape scale (i.e., in a mosaic of forest stands). I present a method to efficiently assess the reproductive success of multiple songbird species across a large area (e.g., > 300 ha). The method uses systematic playbacks of a recording of black-capped chickadee (Poecile atricapilla) mobbing calls to attract species to an observer. Once birds are attracted to the observer, reproduction-related activities can be recorded to infer successful pairing, hatching, or fledging. Trends in reproductive activity of two focal species obtained using this method were similar to their actual reproductive success as measured using intensive nest monitoring in the same locations. We studied the mobbing response of forest birds at different times of the year and in the presence or absence of potential avian predators. Mean duration of mobbing time was significantly longer when confronted with potential predators, but mobbing intensity was not significantly different. Mobbing group size and overall intensity were greatest early in the breeding season before the initiation of egg laying. We concluded that the variability in avian anti-predator mobbing is based on the proximate (individual safety) and ultimate (safety of offspring) risks of participation. I explored the relationship between reproductive activity and songbird abundance using Spearman correlation coefficients. Reproductive activity and songbird abundance were not significantly correlated in 56% of the comparisons (7 or 16). The lack of a consistent relationship emphasizes the importance of studying fitness parameters in addition to abundance or density. I then used canonical correspondence analysis to assess the relationship between forest management and reproductive activity of eight species of forest birds (Vireonidae, Paridae, and Parulidae) in three study grids of varying silvicultural intensity. I predicted landscape effects would become significant as silvicultural intensity increased. Reproductive activity, local vegetation, and landscape structure data were collected on one 6 x 8 systematic grid and two 8 x 8 systematic grids with stations spaced 250 m apart. Basal area of American beech (Fagus grandifolia) and sugar maple (Acer saccharum) within 100 m of the sample locations were the most significant variables associated with reproductive activity on two of the three study grids (most and least intensively managed). In the third study grid (moderately managed), the amount of tolerant hardwood forest within 1 km of the sampling locations was most significant. The relatively minor effect of landscape structure probably reflects the high proportion of suitable habitat remaining in the study area. Landscape effects on songbird populations may become more important if the proportion of suitable habitat declines.

Community-level indices of reproductive success are useful for measuring or monitoring demographic effects of habitat alteration on birds. We present a time-efficient method to estimate the relative reproductive activity of the forest songbird community. A recording of mobbing calls of Black-capped Chickadees (Poecile atricapillus) was broadcast at pre-selected stations during the breeding season. These calls attracted individuals of many bird species present in the vicinity, allowing visual detection of reproductive activity (e.g., adults carrying food or presumed pairs). In mature deciduous forests of northern New Brunswick, 50 bird species responded to the playbacks. Playbacks significantly increased the probability of visual observations of birds compared to silent observations conducted before broadcasting mobbing calls. In coniferous forests of central Québec, playbacks attracted 24 species and also provided a significantly greater opportunity to make visual observations of individual birds. In New Brunswick, mobbing playbacks facilitated more observations of reproductive evidence relative to point counts. Observation periods were brief and a 306-ha plot (1.75 x 1.75 km, 64 points spaced 250 m apart) could be surveyed by foot in less than 32 observer-hours. The proportion of individuals of a given species showing evidence of reproductive activity was used as an index of reproductive success. Black-throated Blue Warblers (Dendroica caerulescens) and Ovenbirds (Seiurus aurocapillus) had a reproductive index consistent with their true nesting success as derived from intensive nest monitoring on the same plots.

Avian anti-predator mobbing is a paradoxical behavior whereby an individual confronts a potential predator and puts itself in harm's way to prevent a potential attack. We provide one of the few empirical tests of hypotheses put forward to explain the variability of observed mobbing behavior in the context of its adaptive significance. We used playbacks of black-capped chickadee (Poecile atricapilla) mobbing calls to incite mobbing behavior in forest birds in New Brunswick, Canada. We studied the mobbing response of forest birds at different times of the year and in the presence or absence of potential avian predators. Fifty species responded to the playback of mobbing calls. Mean duration of mobbing time was significantly longer when confronted with potential predators, but mobbing intensity was not significantly different. Mobbing group size and overall intensity were greatest early in the breeding season before the initiation of egg-laying. Resident species tended to mob with greater intensity in the early spring and autumn when migrant territoriality was least established. Overall mobbing intensity tended to be higher in larger groups. We concluded that the variability in avian anti-predator mobbing is based on the proximate (individual safety) and ultimate (safety of offspring) risks of participation.

Atlantic Cooperative Wildlife Ecology Research Network (ACWERN), and Department of Biology, University of New Brunswick, Box 45111 Fredericton, NB, E3B 6E1 Canada

Atlantic Cooperative Wildlife Ecology Research Network (ACWERN), University of New Brunswick, Box 45111 Fredericton, NB, E3B 6E1 Canada

The abundance of a species is often assumed to be an adequate indicator of habitat quality. When the relationship between abundance and habitat quality is decoupled, interpretation of species abundance data becomes misleading. Habitat quality for songbirds is often measured in terms of the reproductive success achieved by individuals. It is important to understand what types of species and environmental conditions produce a decoupled relationship between abundance and reproductive success. We tested the prediction of Van Horne (1983) that the decoupling is more likely to occur in an environmental condition where usable habitat is patchy. We compared abundance data for eight species of songbirds to an index of reproductive success in patchy and relatively contiguous landscapes. Fewer species (0 of 8) showed high concordance (> 0.5 Spearman Correlation Coefficient) between abundance and the reproductive index in the contiguous landscape than in the patchy landscape (5 of 8, Chi-square test, Williams Correction, p = 0.003). This result was the opposite of the predicted relationship. In addition, reproductive success was not different between the landscapes for seven of the eight species, indicating little difference in habitat quality. The lack of a predictable decoupling between abundance and reproductive success may be related to a temporal pattern rather than a spatial pattern. Creation of the patches generally occurred greater than ten years prior to the study. This may have been sufficient time to allow a "settling" of the local abundances to typical densities. The unpredictable results emphasize the difficulty in relying entirely on abundance data to study the effects of land-use practices on songbird populations.

¹Atlantic Cooperative Wildlife Ecology Research Network (ACWERN), and Department of Biology, University of New Brunswick, Box 45111 Fredericton, NB, E3B 6E1 Canada

Current Address: Hancock Land Company, PO Box 299, Casco, ME 04015 USA, jlgunn@hancockland.com

Industrial forest management affects the local vegetation and landscape structure of the mosaic of forest stands in the Acadian forest of New Brunswick. The relative effects on forest bird reproductive success of changes on the landscape versus changes in local vegetation structure and floristics are largely unknown. We used canonical correspondence analysis to assess the impacts of forest management on a forest bird assemblage in three landscape contexts of varying silvicultural intensity. Local vegetation factors were the most important variables determining reproductive success on two of the three landscapes. In the third landscape, the amount of tolerant hardwood forest within 1 km of the sample locations was most important. However, the significance of this variable was confounded by the presence of spatially structured environmental data. Of the local variables sampled, it appeared that the basal area of American beech played a role in determining reproductive success of the bird species studied. When reproductive success was compared between grids, the lack of beech on the most intensively-managed study grid did not appear to reduce the likelihood of successful reproduction for most of the species. The lack of a landscape context effect further illustrated the importance of local factors in determining reproductive success in the forest birds we studied. If local factors are indeed most important, then significant information exists to provide forest managers with the tools required to manage for species of concern.