Peter Arcese

The loss of tropical forest for agriculture is a principal cause of declines in species diversity in the Neotropics. Shade-grown coffee plantations, where coffee plants grow under a canopy of trees, have been found to support a higher diversity of both Neotropical migrant and tropical resident birds compared to agroecosystems without trees. However, how local and landscape factors interact to affect avian biodiversity remains poorly understood. I studied the influence and interaction of local and landscape attributes on bird richness and community completeness in 160 shade coffee plantations in two regions of Colombia differing in extent of regional forest: Antioquia, where forest covered 20% of the study area and Cauca, where only 1% of forest remained. By using generalized linear models, I determined how species richness and community completeness were influenced by the local scale structure and composition of shade-grown coffee plantations (i.e., tree richness, tree abundance, and coffee density) and landscape variables surrounding the plantation (i.e., plantation distance to forest, cover by intact forest, forest-agriculture mosaic, or the combination of the two). My findings suggest that the extent of regional forest cover affects the landscape-level contribution to bird richness and community completeness within plantations. In Antioquia, richness of bird communities increased with the proportion of all forest elements within the surrounding landscape. Community completeness was influenced by the interaction between the local and landscape scales; tree richness enhanced community completeness only when the forest-agriculture mosaic cover at the landscape scale was greater than 73%. In Cauca, higher richness and abundance of trees at the local scale were associated with greater bird richness and community completeness, while there were no landscape-scale effects. Within highly deforested regions, birds are most likely to benefit in the near-term from increasing the diversity of shade trees within farms. However, reforestation efforts at the landscape scale should be considered as a long-term strategy to provide critical resources that support more complete communities of birds in coffee plantations. In more forested regions, bird communities could benefit by protecting and/or increasing the forest cover in the landscape and increasing the diversity of trees within plantations.

View record

Human impacts can directly (e.g., bycatch, over-fishing, pollution), indirectly (e.g., climate forcing), or in concert affect marine systems and the species that rely on them. These effects can alter ecosystem function, productivity, and the abundance, distribution and demography of many marine predators. Marine birds are particularly sensitive to these impacts and given their ease of observation and diverse foraging tactics and habitat uses are outstanding indicators of ecosystem health. Conserving marine birds and their ecosystems requires robust predictions of species distribution and the identification of ‘hotspots’ of richness and abundance to help mitigate human disturbance in areas and at times when large aggregations of diverse species occur. I modeled seasonal variation in marine bird species diversity (defined by the Shannon Weiner Diversity Index) and in marine bird family occurrence to map the intensity and extent of hotspots in relation to season and climate in Canada’s Pacific Exclusive Economic Zone. To do so, I used 20 years of survey data from the North Pacific Pelagic Seabird Database (1997-2017) and remote sensing data describing marine conditions and local geography (sea surface temperature, chlorophyll a, bathymetry, and distance to shore) to illustrate how seasonality within years and the El Niño Southern Oscillation (ENSO) across years influenced spatial patterns in diversity. Hotspots were most persistent in Hecate Strait, off the west coast of Vancouver Island, and surrounding the Scott Islands in most seasons. Hotspots became less intense and/or shifted northwards as sea surface temperatures rose, reflecting the results of many studies on individual species responses to increasing temperatures and demonstrating such effects at the level of marine bird communities. I describe how my results could be applied to potentially mitigate the deleterious effects of humans on the diversity and distribution of marine birds. I also provide cautions which must accompany such predictions given sparse data for some areas and seasons. Overall, the maps I present provide a template for mapping marine species distribution for the purpose of identifying ‘hotspots’ of diversity, and thereby facilitate planning to minimize harmful impacts in highly diverse and dynamic systems.

View record

Biological invasions by non-native ‘ecosystem engineers’ can radically alter the ecological and socio-economic values of ecosystems in ways that may require decades to detect. The invasion of North American glacial refuges by non-native earthworms is a prominent but understudied example of a cryptic invasion by an ecosystem engineer. Non-native earthworms are known to reduce soil carbon, disrupt mycorrhizal networks, and homogenize plant communities in their role as seed predators, root foragers, and in nutrient cycling and redistribution. However, natural resource managers have struggled to discern the scale at which non-native earthworms influence plant species diversity across invaded biomes. With no effective methods to eradicate or control established earthworm populations, there is great need for preemptive strategies to identify high-value conservation areas at risk of invasion. Herein, I address two main questions with implications for forest management: 1) Can the influence of non-native earthworms on plant community assembly be reliably predicted using plant traits? 2) Can abiotic factors be used to identify and predict natural refuges from earthworms in heterogenous habitats? I found that the presence of earthworms contributed to the simplification of plant communities in experimental mesocosms and observational surveys of in-situ forest and meadow habitat. In general, earthworms were associated with plant communities dominated by species with large seeds and fibrous roots, whereas species with small seeds and taproots only persisted in multi-species mesocosms without earthworms. These findings suggest that earthworms shape community composition in the early stages of invasion by acting as ecological filters on morphological plant traits. Last, I constructed an ensemble species distribution model for non-native earthworms using data from 300 survey plots to identify the suite of environmental conditions needed to limit the dispersal and persistence of invading earthworms. This model showed that shallow and dry soils on steep terrain strongly limit the occurrence and abundance of non-native earthworms. My results show that earthworms reduce plant species richness in coastal forest and meadow habitats of southwest British Columbia and highlight the conservation value of shallow-soil habitats that limit earthworm distribution and persistence.

View record

Global declines in large-bodied terrestrial vertebrates have been widely linked to human disturbance and habitat loss. Consequently, identifying remaining opportunities to conserve habitats likely to maximize the persistence of such species remains a key challenge to conserving biological diversity globally. Andean bears (Tremarctos ornatus) are among the least-known wide-ranging large mammal species, but occupy montane ecoregions throughout the Northern Andes, a global biodiversity hotspot. Recent evidence suggests the Andean bear range includes forests at the fringe of human development, including the equatorial dry forests of Peru. If Andean bears are to be conserved in this region, a better understanding of their potential densities, use of habitats, and response to human influence is required. This thesis used spatial density models, existing field and remote-sensed data, and spatial planning tools to relate species detection to environmental variables and optimize conservation plans for Andean bears and the Northern Andes. I assessed habitat use by Andean bears in equatorial dry forest at a local scale in Northwestern Peru, using camera-trap data to construct spatially-explicit capture recapture (SCR) models. I compared models for resources thought to affect bear density by their association with threats and food availability; including elevation, slope, forest cover, and proximity to roads. I found that proximity to roads reduced the density of Andean bears, and the influence of factors other than roads varied seasonally. I identified potential areas of equatorial dry forest outside the IUCN range that could support Andean bears, but noted that unmapped roads and smallholder agriculture affected the reliability of results. I also employed systematic prioritization methods to identify configurations of land parcels that maximized biodiversity features at the least cost. I found that the Andean bear range performed better at capturing species richness than a random control feature. I also found that planning for multiple goals in a systematic planning framework greatly increased the area efficiency of the solutions, compared with planning for biodiversity features separately. Overall, this thesis highlights the importance of working at multiple scales for efficient conservation planning, and provides a common framework for conducting such analyses in the future.

View record

Identifying causes and consequences of variation in species life history should improve predictions about how climate and land use change will affect the demography and distribution of species in future. Sooty fox sparrows (Passerella iliaca unalaschcensis) were documented as obligate migrants, abundant in winter but with only three breeding records in coastal habitats of British Columbia and Washington prior to 1950. Because this subspecies has since established year-round resident populations in this region, I studied resident populations of sooty fox sparrows to test theory on how climate change and life history might affect the demography and distribution of a new partial migrant.I estimated demographic vital rates in one recently established resident population on Mandarte Is., BC, using color-banded birds. Annual fecundity (F) was higher than reported in migrant populations studied previously in Alaska and Newfoundland, supporting the hypothesis that residents invest more in reproduction on average than migrants within species. I also estimated high juvenile and adult overwinter survival (Sj = 0.32 ± 0.06, and Sa = 0.69 ± 0.05) and population growth (λexp = 1.61 ± 0.57), implying rapid population growth. I next tested the hypothesis that climate warming facilitated the establishment of resident populations by reducing the net benefit of migrating out of the wintering area to breed. Because resident sooty fox sparrows breed earlier and longer than migrants, I asked if climate warming during the pre-breeding period coincided with patterns of establishment by: 1) characterizing the pre-breeding climate niche of resident populations using species distribution models, occurrence data, and monthly climate records, and 2) testing if the emergence of the pre-breeding climate niche currently occupied by resident populations corresponds to the first reports of sooty fox sparrows breeding after 1920. Niche models suggest that the mild, near-shore climate niche now occupied by recently established resident populations expanded dramatically from 1920 to 2015 in a pattern matching early records of breeding by sooty fox sparrows within their historic wintering area from 1950.Whereas prior studies focus on climate warming affecting overwinter survival, my results suggest warming may also affect migration through improved fecundity and breeding niches.

View record

Nest success is a key factor affecting the dynamics and life–history evolution of avian populations. While multiple factors affect nest success, their relative influence remains unclear, partly due to the short time periods over which many studies take place relative to the scales of temporal variation in the environment, and the traits of birds that make up populations. I examined the effects of two intrinsic (female age, inbreeding coefficient), two abiotic (rainfall, temperature), and three biotic (breeding densities, cowbird parasitism rates, and brood parasitism) factors potentially affecting nest success (≥ 1 fledged young) in an insular song sparrow (Melospiza melodia) population over 39 years. I also compared the influence of these factors in three 13–year intervals representing the early, mid and late periods of the study to estimate temporal influence. Over 39 years, song sparrow nesting success first increased (1 – 3 years), and then decreased with female age (3+ years), and declined in relation to the degree of inbreeding in females, increased rainfall during the nest period, and by increased breeding densities. Nests that were parasitized by brown–headed cowbirds, or that experienced increased risk of cowbird predation, tended to fail more often. Parallel analyses of nest success in the early, mid and late periods of the study showed that only female age and breeding densities explained success in all periods, whereas the effects of inbreeding, cowbirds, and rainfall were episodic through time. This discrepancy was due to temporal variation in abiotic or biotic conditions that affected which factors were most influential of success. Many studies of nest success in passerine birds are limited in duration and the number of variables that can be considered due to limits on the amount or quality of data, preventing the comparison of many biotic and abiotic factors reported to affect success in the literature. I show that over 39 years, the intrinsic effects of inbreeding, abiotic effects of climate and biotic effects of brood parasites on nest success were each influential but varied through time, indicating that any ranking of their relative influence on demography will also vary temporally.

View record

Species invasion and range shifts are widely reported and facilitate novel interactions among potential competitors in plant and animal communities worldwide. However, predicting which novel interactions will result in the extirpation of subordinate competitors is challenging. Coexistence versus extinction as alternative outcomes of competition between resident and colonizing species may arise due to (1) variation in interaction strength, (2) change in other demographic drivers more influential than those linked to competition, or (3) differences in the extent to which resources are equitably partitioned between competitors, which may in turn depend on the spatial scale examined. To date, however, empirical studies suggest these factors rarely align to cause the competitive exclusion of native species. I used a combination of field experiments and demographic analyses to test the hypothesis that colonizing fox sparrows (Passerella iliaca) have caused the 0.6% per year decline of a song sparrow (Melospiza melodia) population resident on Mandarte Island, BC, Canada. Several lines of evidence indicate that interspecific competition for winter food has: a) reduced survival in juvenile song sparrows after colonization by fox sparrows in 1975, b) led to an inverse relationship between juvenile song sparrow survival and fox sparrow population size, c) excluded song sparrows from high-quality foraging sites in winter via consistent behavioural dominance by fox sparrows and complete overlap of fox and song sparrow preference for local seeds, despite d) no evidence of competition for breeding territories or nesting habitat. My results suggest that in the absence of rapid ecological or evolutionary shifts in niche dimension, song sparrows will likely be extirpated from Mandarte Is., thus demonstrating that competitive exclusion of native species can occur when interactions are strong and resources are not easily partitioned.

View record

Natural selection in the wild has been extensively researched, but few studies have identified the abiotic and biotic drivers of selection or quantified their influence. I characterized and quantified annual selection on laying date in the song sparrow (Melospiza melodia) population of Mandarte Island, British Columbia over 35 years. Selection was approximately linear for annual reproductive success (ARS) and over-winter survival based on cubic splines visualization of fitness surfaces. Fecundity selection (recruited offspring) was roughly 10 times stronger (-0.271 ± 0.031, mean ± SE) than viability selection (-0.028 ± 0.043, mean ± SE). Since opposing selection could constrain evolution, I used multiple measures of fitness, including fecundity measured at sequential offspring life stages, and female over-winter survival, to test if selection was complementary or opposing. Selection favored early breeding through all fitness measures and was therefore complementary. The strength of fecundity selection on recruits was 1.43 times stronger than selection on nestlings, indicating that the effects of early laying were additive and accumulated over time. Despite strong selection and moderate heritability (h² = 0.16), there has been no phenotypic advance of laying date in this population from 1975 to 2010. I also investigated several potential drivers of selection including density, spring temperature, precipitation and the intensity of brood parasitism by brown-headed cowbirds (Molothrus ater), and determined their relative influence on fecundity and viability selection. I found that density was the most influential driver of fecundity selection, and that selection favoring early breeding increased in magnitude at high population density and high intensities of brood parasitism. Because population density and the age and inbreeding structure of this population are correlated, I also included each of these variables as a covariate in calculation of selection differentials, and found weaker selection for some fitness measures. A climate index representing precipitation during the pre-laying period was the most influential driver of over-winter viability selection, with early breeding favored in winters preceded by warm, dry springs, and late breeding favored in winters preceded by cool, wet springs.

View record

Garry oak meadow ecosystems in British Columbia are fragmented, increasinglydegraded, and have been prioritized for conservation. While distribution maps of remnantmeadow patches have been developed, the ecological integrity of plant communities inmany of these remnants remains unknown. Modeling and mapping ecological integritycould inform conservation prioritization exercises in the region.The primary goal of this thesis was to develop distribution models of native and exoticplant species richness in Garry oak meadow remnants. Secondly, multiple independentdatasets were used to analyze the effects of sample size and sampling bias on theaccuracy and reliability of resulting predictive maps, which is an active area of researchin species distribution modeling. Finally, I investigated whether Terrestrial EcosystemMapping (TEM) – a publicly available GIS layer of plant community associations –provided a valid geographical extent over which to map predictions.In Chapter 2, different datasets were found to produce different models of speciesrichness. However, different native richness models produced similar distribution maps,while exotic richness maps based on different datasets were less similar. Theincorporation of spatial variables into models did not improve model fit, howeversignificant residual spatial autocorrelation at a broad scale was detected in some cases,suggesting that an important environmental covariate is missing from these models.Examples of potential missing covariates include deer density and disturbance history.Overall, this research demonstrates that multiple independent datasets are very importantiiifor validating species distribution models, especially in heterogeneous landscapes.Additionally, large sample sizes and sampling broadly across of the area of predictionresult in more robust models.The results presented in Chapter 3 suggest that mapping predictions exclusively overGarry oak ecosystem-classified TEM polygons is potentially overly conservative, asspecies richness of native meadow species was found to be high in other TEMclassifications as well. This suggests that Garry oak meadow plant communities do notexist solely in discreet meadow patches, and that they are dispersed throughout otherhabitat types including Douglas-fir – salal forests.

View record

The Vancouver Island marmot is a critically endangered sciurid endemic to British Columbia, Canada. By 1997, the species had collapsed to fewer than 150 individuals in total. Between 2003 and 2010, 301 captive-bred marmots were implanted with radiotelemetry transmitters and released to new sites and extinct colonies to supplement the wild population. I evaluated reintroduction success based on three short-term measures: first season fidelity to release site, active season survival from release to hibernation, and overwinter survival through early spring. I used generalized linear mixed models to evaluate the influence of sex and age, release practices and procedures, and the local and landscape-level attributes of release habitats on site fidelity and survival. Results suggested that poor overwinter survival has been the limiting factor in first-year reintroduction success. In all years, overwinter survival was lower for newly released captive-bred marmots than for their wild or previously-released counterparts. Release date was the variable most predictive of success, but was positively associated to site fidelity and active season survival and negatively associated to overwinter survival. Release date was the only predictor for overwinter survival, whereas site fidelity was also negatively impacted by the presence of resident females at the release site. Active season survival was highest for females, 2-year olds, and marmots released to sites with talus. I discuss these results in the context of current release practices for Vancouver Island marmots, and make recommendations for future release candidates, site characteristics, and release procedures.

View record