Submitted by Erika Demachki (erikademachki@gmail.com) on 2014-10-29T14:42:08Z No. of bitstreams: 2 Tese - Marina Zanin Gregorini - 2014.pdf: 10952620 bytes, checksum: bc9f9431ac9ea502525179cb6f912f05 (MD5) license_rdf: 23148 bytes, checksum: 9da0b6dfac957114c6a7714714b86306 (MD5) / Approved for entry into archive by Luciana Ferreira (lucgeral@gmail.com) on 2014-10-30T10:07:47Z (GMT) No. of bitstreams: 2 Tese - Marina Zanin Gregorini - 2014.pdf: 10952620 bytes, checksum: bc9f9431ac9ea502525179cb6f912f05 (MD5) license_rdf: 23148 bytes, checksum: 9da0b6dfac957114c6a7714714b86306 (MD5) / Made available in DSpace on 2014-10-30T10:07:48Z (GMT). No. of bitstreams: 2 Tese - Marina Zanin Gregorini - 2014.pdf: 10952620 bytes, checksum: bc9f9431ac9ea502525179cb6f912f05 (MD5) license_rdf: 23148 bytes, checksum: 9da0b6dfac957114c6a7714714b86306 (MD5) Previous issue date: 2014-03-31 / Coordenação de Aperfeiçoamento de Pessoal de Nível Superior - CAPES / Nosso trabalho teve como objetivo investigar o efeito da perda e fragmentação de habitat sobre os felinos (Carnivora: Felidae), testando hipóteses relacionadas ao tema, bem como fazendo inferências para a conservação. Apresentamos aqui três capítulos no formato de artigo científico e uma breve discussão geral, que consiste na compreensão geral proveniente dos resultados dos três primeiros. Iniciamos essa tese com uma revisão sistemática e quantitativa da literatura sobre o efeito da perda e fragmentação do habitat sobre felinos. Esse trabalho consistiu no passo inicial dessa tese, pois permitiu identificar as lacunas de conhecimento, tendências gerais e metodologias eficientes a serem aplicadas nas outras etapas do trabalho. No entanto, nossa revisão se estende além da avaliação do “estado da arte”, pois testamos também hipóteses relativas a alocação do esforço de pesquisa. / Nosso trabalho teve como objetivo investigar o efeito da perda e fragmentação de habitat sobre os felinos (Carnivora: Felidae), testando hipóteses relacionadas ao tema, bem como fazendo inferências para a conservação. Apresentamos aqui três capítulos no formato de artigo científico e uma breve discussão geral, que consiste na compreensão geral proveniente dos resultados dos três primeiros. Iniciamos essa tese com uma revisão sistemática e quantitativa da literatura sobre o efeito da perda e fragmentação do habitat sobre felinos. Esse trabalho consistiu no passo inicial dessa tese, pois permitiu identificar as lacunas de conhecimento, tendências gerais e metodologias eficientes a serem aplicadas nas outras etapas do trabalho. No entanto, nossa revisão se estende além da avaliação do “estado da arte”, pois testamos também hipóteses relativas a alocação do esforço de pesquisa.

Urbanization is a key driver of habitat loss, ecosystem degradation and has a great impact on biodiversity. Exploitation of buffer zones surrounding conservation areas and green structures in urban environment can affect biodiversity through reduced total area of habitat, increased edge effects and lost connectivity on a landscape level. The previous regional development plans for the Stockholm region, had the purpose of leaving large green structures undeveloped to secure core areas of great biological value by focusing on a dense city core. However, the latest regional development strategy puts stress on the green wedges by shifting the focal areas of the development into suburban regional city centers, in many cases close to the green wedges. The purpose of this study was to map habitat loss and changes in the total area of the Järva green wedge, west of Stockholm, caused by development of housing areas in previously sparsley exploited buffer zones, during the period 2000 – 2015. The study uses theories of landscape ecology, remote sensing and GIS to map and quantify habitat loss between 2000 and 2015. Two different birds were used as surrogate species, one grassland habitat specialist – Corncrake (Crex crex) and one forest habitat generalist – Eurasian jay (Garrulus glandarius). The two different surrogate species were used to identify how loss of two nature types could influence biodiversity and connectivity for a group of species. Documents and development plans on regional and local scale were also used to map and predict further habitat loss and exploitation of the green wedge until 2030. The result of the study shows that grassland habitat lost nearly twice the area than forest habitat due to development of housing areas during the period 2000 to 2015. A total of 1.3 km² of grassland and 0.7 km² forest habitat were replaced by housing areas during the fifteen years covered in the study, and the Järva green wedge will have lost a total of 3.84 km² buffer zones by the year 2030. The Corncrake and other grassland specialist species is likely to get most affected when grassland suffered the greatest habitat loss in the area. Even though the Eurasian jay has a key ecological function for the Oak forest in the Järva green wedge and relies on forest habitat for successful breeding, the loss of forest habitat will probably not affect the habitat generalist species in the same way. On a regional scale, the study suggests that habitat loss and fragmentation may affect grassland specialist species more than forest generalist species. The overall connectivity in the Järva green wedge is likely to get affected by a shrinking total area caused by narrowing of the green wedge until 2030. The function of the Järva green wedge as a dispersal corridor for biodiversity in the Stockholm region will most certainly get affected by further loss of buffer zones caused by exploitation of land for housing areas.

Dans les paysages fragmentés, les mouvements entre les patchs d'habitats jouent un rôle primordial dans la persistance des populations en assurant les flux génétiques, la possibilité de recolonisation après extinction locale, la liaison entre différents habitats pour la reproduction ou l'acquisition des ressources, … Malgré l’importance de ces mouvements, la structure du paysage entre ces patchs (i.e. la matrice) est le plus souvent considérée comme un ensemble homogène de non-habitat d’intérêt mineur et les adaptations comportementales en réponse à la fragmentation du paysage sont rarement prises en compte. Pour étudier l'importance de l'influence du paysage sur les mouvements, trois approches complémentaires sont pratiquées. 1) Des études expérimentales sont menées pour évaluer les réponses comportementales et physiologiques du crapaud commun (Bufo bufo) juvénile et adulte soumis aux différents types d'occupation du sol. Ainsi, les types d'occupation du sol sont caractérisés par le coût que leur traversée représente pour l’espèce étudiée. 2) Ce travail s’attache également à décrypter les règles régissant les déplacements en paysage hétérogène et à mettre en évidence des différences comportementales selon le paysage d'origine (adaptations). 3) Basé sur la modélisation, le troisième volet doit permettre d’effectuer des prédictions à l’échelle des paysages / In fragmented landscapes, movements between habitat patches play a crucial role in population persistence by allowing gene flow, rescue effect possibilities, linking different habitats for reproduction or resources acquisition, … Although the importance of these movements, the landscape structure between patches (i.e. matrix) is too frequently considered as a homogeneous non-habitat of minor interest and behavioural adaptations in response to fragmentation are rarely taken into account. To investigate the importance of the landscape influence on movements, three complementary approaches are used: 1) experimental studies are performed to assess behavioural and physiological responses of common toads when they are confronted to different land uses. Land uses are thus characterised par the cost of their crossing for the focus species. 2) This work also aims to decipher the rules of movement in heterogeneous landscapes and to highlight behavioural differences according to the originated landscape (adaptations). 3) On the basis of modelling, the third part of this work intends to make predictions at the landscape scale

Orientador: Milton Cezar Ribeiro / Resumo: Modificações da cobertura do solo, perturbações antropogênicas e mudanças climáticas na região Neotropical têm sido associadas à crise da biodiversidade. Nesta tese, o objetivo principal foi investigar os efeitos que a estrutura da paisagem (incluindo a perturbação antropogênica) e a qualidade do habitat têm na distribuição espacial, nos padrões de movimento e na seleção de habitat da queixada - Tayassu pecari. Esta espécie é o único ungulado florestal Neotropical que forma grandes bandos e é uma espécie-chave nos ecossistemas florestais, pois sua extinção local pode causar perdas adicionais de biodiversidade por meio de efeitos em cascata. Ainda assim, pouco se sabe sobre uso do espaço e movimentos de queixadas em escala fina. Para tanto, estudamos a distribuição e movimentação da espécie em diferentes escalas e biomas do Brasil, investigando quais variáveis de paisagem influenciaram a mesma em diferentes escalas e ordens de seleção. Nossos resultados mostram fortes efeitos negativos da modificação da paisagem e fragmentação do habitat na distribuição de queixadas, restando apenas 47% do território brasileiro com áreas adequadas para a espécie. Verificamos que menos da metade de todas as áreas protegidas nos biomas Cerrado e Mata Atlântica têm um habitat adequado para a queixada. Para a análise do movimento, rastreamos 30 indivíduos com colares de GPS de julho de 2013 a outubro de 2016 no Pantanal e no Cerrado do estado do Mato Grosso do Sul. Nossos principais resultados mos... (Resumo completo, clicar acesso eletrônico abaixo) / Abstract: Land cover modifications, anthropogenic disturbances and climate change in the Neotropical region have been linked to the biodiversity crisis. This thesis proposal was to investigate the effects that landscape structure (including anthropogenic disturbance) and habitat quality have in the spatial distribution, in the movement patterns and in the habitat selection of the white-lipped peccary - Tayassu pecari (WLP). This species is the only Neotropical forest ungulate that forms large herds, and it is a keystone species in forest ecosystems because its local extinction may cause additional biodiversity losses through cascading effects. Still, little is known about WLPs space use and fine-scale movement. For that aim we studied the species distribution and movements in different scales and biomes of Brazil, investigating which landscape variables would influence the species in different scales and orders of selection. Our findings show strong negative effects of landscape modification and habitat fragmentation on white-lipped peccary distribution, with only 47% of the Brazilian territory with suitable areas for the species. We also found that less than half of all protected areas in the Cerrado and the Atlantic Forest biomes have suitable habitat for WLP. For the movement analysis we tracked 30 individuals with GPS collars from July 2013 to October of 2016 in the Pantanal and in the Cerrado of Mato Grosso do Sul state. Our main results showed a positive selection of WLP for fore... (Complete abstract click electronic access below) / Doutor

Aim: Identifying how climate change, habitat loss, and corridors interact to influence species survival or extinction is critical to understanding macro-scale biodiversity dynamics under changing environments. In North America, the ice-free corridor was the only major pathway for northward migration by megafaunal species during the last deglaciation. However, the timing and interplay among the late Quaternary megafaunal extinctions, climate change, habitat structure, and the opening and reforestation of the ice-free corridor have been unclear. Location: North America. Time period: 15–10 ka. Major taxa studied: Woolly mammoth (Mammuthus primigenius). Methods: For central North America and the ice-free corridor between 15 and 10 ka, we used a series of models and continental-scale datasets to reconstruct habitat characteristics and assess habitat suitability. The models and datasets include biophysical and statistical niche models Niche Mapper and Maxent, downscaled climate simulations from CCSM3 SynTraCE, LPJ-GUESS simulations of net primary productivity (NPP) and woody cover, and woody cover based upon fossil pollen from Neotoma. Results: The ice-free corridor may have been of limited suitability for traversal by mammoths and other grazers due to persistently low productivity by herbaceous plants and quick reforestation after opening 14 ka. Simultaneously, rapid reforestation and decreased forage productivity may have led to declining habitat suitability in central North America. This was possibly amplified by a positive feedback loop driven by reduced herbivory pressures, as mammoth population decline led to the further loss of open habitat. Main conclusions: Declining habitat availability south of the Laurentide Ice Sheet and limited habitat availability in the ice-free corridor were contributing factors in North American extinctions of woolly mammoths and other large grazers that likely operated synergistically with anthropogenic pressures. The role of habitat loss and attenuated corridor suitability for the woolly mammoth extinction reinforce the critical importance of protected habitat connectivity during changing climates, particularly for large vertebrates.

The development of genetic tools to study populations at the molecular level has been one of the most important contributions to understand demographic processes in wild populations of conservation concern. Habitat loss and habitat fragmentation are the main factors causing declines of birds populations, as in the case of the Worthen’s sparrow (Spizella wortheni), a Mexican endemic Emberizid restricted to scrub and grassland habitats in northeastern Mexico. Here, I present the first molecular tools developed specifically for the study of S. wortheni populations. Genetic analyses of seven remnant populations of S. wortheni showed that no genetic impoverishment is present. This unexpected result may be caused by the nomadic life style of S. wortheni, which makes it tolerant to habitat modification. The high levels of inbreeding I found can be attributed to the tendency of S. wortheni to move in gregarious groups, since non-random mating might be present. My analyses based on the study of DNA segments from five mitochondrial genes support the hypotheses that S. wortheni and S. pusilla (Field Sparrow), as well as S. breweri (Brewer’s Sparrow) and S. passerina (Chipping Sparrow) are sister taxa. A phylogenetic network analysis showed that conflicting relationships among Spizella species might be caused by possible hybridization events between ancestors of extant taxa. Based on the study of the phylogenetic community structure, patterns of phylogenetic attraction and repulsion, and on the estimation of marginality, specialization and niche overlap among species, I investigated the possible causes of rarity in S. wortheni. The phylogenetic community structure analysis suggested that Spizella species are clustered across vegetation communities. Assuming phylogenetically conserved ecological traits, the phylogenetically clumped distribution may indicate habitat filtering among Spizella species. Additionally, behavioral traits such as flocking in mixed-species flocks might contribute to explain the pattern of community structure I found, and suggest that the rarity of S. wortheni is not a matter of interspecific competition. According to the phylogeny, evolutionary age was also discarded as a cause of rarity of this species. Compared with its congenerics, S. wortheni is the most specialized of all Spizella species. A potential evolutionary change of S. wortheni behavior that results in niche specialization is more likely to be the cause of its rarity.

No description available.

I investigated the effects ofhabitat loss and subsequent restoration on invertebrate community structure and ecosystem functioning in a mined peat bog in the North Island, New Zealand. In an experimental trial, the impact of peat bog habitat loss and isolation on the invertebrate community associated with Sporadanthus ferrugineus (Restionaceae) was investigated. Potted S. ferrugineus plants were exposed to invertebrates at various distances up to 800 m from an intact habitat (the presumed source population) over 18 weeks. Invertebrates rapidly colonised the experimental plants, with all major Orders and trophic groups present on Sc ferrugineus within 6 weeks. However. with increasing distance away from the undisturbed habitat, there was a significant decrease in total richness and abundance of invertebrates associated with the potted plants. Additional tests showed that even a moderate degree of isolation (i.e. greater than 400 m) from the intact habitat caused an almost complete failure of 'Batrachedra' sp. to colonise its host plant, at least in the short-term, The density of eggs and larvae, and the average larval size of 'Batrachedra' sp. (Lepidoptera: Coleophoridae) colonising Si ferrugineus plants, as well as the proportion of Si ferrugineus stems damaged by 'Batrachedra' sp. herbivory, all decreased logarithmically with increasing distance from the intact habitat. Surprisingly, though, the rate of recovery of the insect-plant interaction following experimental habitat restoration was remarkably rapid (i.e. between 3Y2 and 6 years). After just 6 years there was no significant difference in insect-plant interactions between the intact peat bog sites and any of the experimentally restored sites up to 800 m away. These results suggest that the degree of isolation from undisturbed habitat has a major impact on the rate and patterns of restoration recovery in the invertebrate community and that some insect-plant interactions can recover rapidly from habitat loss with restoration management. Restoration of mined peat bogs in northern New Zealand is initiated by establishing a native vegetation cover to minimize further peat degradation. The effects of various restoration techniques on litter decomposition, microbial community activity and beetle community composition were investigated within an experimental trial, These treatments included translocation ofpeat bog habitat (direct transfer of islands), milled peat islands with no seed and milled peat islands with seed, and were compared with an unrestored mined site and an undisturbed peat bog. In all the response variables measured, the undisturbed peat bog sites had significantly higher decomposition rates and microbial respiration rates, and significantly higher abundance and species richness of beetles than any of the restoration treatments. Inaddition, the technique used to restore mined peatlands had a significant effect on the beetle community composition and litter decomposition processes. Despite a rapid initial change in the beetle community following habitat translocation, the direct transfer islands were still the most similar in beetle species composition to the undisturbed peat bog. Microbial activity and decomposition rates were higher in the direct transfer and mined peat surface after 6 months. However, even after 12 months, decomposition rates in the restored habitats were still far from reaching the levels recorded in the undisturbed peat bog. The results suggest that beetle community structure and ecosystem processes such as decomposition and microbial activity rates may be able to recover faster with certain restoration techniques, such as direct transfer of intact habitat islands. Subsequently, I examined long-term beetle community reassembly on islands that had been restored by creating raised areas ofprocessed peat with the addition of Leptospermum scoparium seed. Monitoring of different-aged restored islands representing the full range of restoration ages (up to 6 years) available at the peat mine, indicated that as the peat islands became older and the vegetation structure became more complex, the abundance, species richness and composition of the beetle community became increasingly similar to the community in the undisturbed peat bog. Despite this, distinct differences between the intact peat bog and older restored peat islands still persisted, even after 6 years, particularly at an individual species level. However, it is predicted that within 12 years the restored peat islands will share 100% ofbeetle species in common with the undisturbed peat bog. Taken together, these results indicate that restoration is effective in initiating the recovery of beetle assemblages and ecosystem processes (such as litter decomposition and microbial community activity) in cut-over peat bogs. However, it is estimated to take at least 12 years before pre-mining communities and functions are attained, and ongoing monitoring to develop an understanding of the longer-term dynamics of such ecosystems and processes is clearly required.

Projet de recherche réalisé avec Bernard Angers comme directeur de maîtrise, Denis Réale en tant que co-directeur et grâce à la collaboration active d'Emmanuel Milot. / La région métropolitaine de Montréal est formée de nombreuses îles à la jonction du fleuve Saint-Laurent et de la rivière des Outaouais, isolant ainsi les populations insulaires en fonction de distances respectives ainsi que des courants. Ce système offre un contexte idéal pour évaluer l’effet de la perte d’habitat liée à la pression d'urbanisation dans un paysage métropolitain insulaire ou en situation d’archipel. La présente étude a pour objectif de comparer l’effet de la perte d’habitat sur la diversité génétique de deux serpents très distincts, Storeria dekayi et Thamnophis sirtalis. Des analyses réalisées à l’aide de marqueurs microsatellites révèlent une plus importante structure génétique entre les populations de S. dekayi (FST=0,19) qu’entre celles de T. sirtalis (FST=0,07) dans la région montréalaise. Chez les deux espèces étudiées, la majorité des populations des habitats réduits présente une richesse allélique moyenne comparable à celle observée dans les habitats plus vastes. Néanmoins, certaines populations présentent des réponses différentes, dont des traces de goulots d’étranglement, une perte de richesse allélique ou encore une importante modification des fréquences alléliques. Au niveau régional, les résultats présentent une importante perte de diversité génétique chez les couleuvres se trouvant sur le continent alors que les populations insulaires de la région montréalaise constituent désormais un réservoir de diversité génétique. Les résultats observés auprès des populations insulaires démontrent que les effets de la perte d’habitat peuvent s’avérer très spécifiques à chaque situation et que la détection de traces génétiques d’un tel phénomène peut nécessiter un contexte logistique très particulier. Un nombre croissant de publications reportent une absence de signature génétique suite à la perte d’habitat chez des oiseaux et des mammifères. Il s’agit de la première étude témoignant de ce phénomène chez les reptiles. Une note est fournie en annexe à l’intention des gestionnaires au sujet de la conservation de la couleuvre brune, S. dekayi. / The Montreal metropolitan community includes numerous islands located at the confluence of the Saint-Lawrence and Ottawa Rivers. In such a fragmented landscape, dispersal of animals is limited by the distance between islands as well as the currents. This system offers an ideal context for the study of the effects of habitat loss on the genetic diversity of animal populations located on islands or archipelagos. This study seeks to assess the effects of habitat area by comparing the organization of genetic diversity of two highly distinct snake species, Storeria dekayi and Thamnophis sirtalis. Analysis realized with microsatellite markers reveals a much stronger genetic organisation in S. dekayi (FST=0.19) than in T. sirtalis (FST=0.07) in the Montreal area. For both studied species, most populations found in reduced habitats showed similar genetic diversity to what was observed in larger habitats. Nevertheless, some populations showed different responses to the loss of habitat, including traces of genetic bottlenecks, a loss in mean allelic richness or an important alteration of their allelic frequencies. This study also reveals an important loss of genetic diversity in the continental snake populations. At the regional scale, the results reveal an important loss of genetic diversity in the continental snake populations and that the insular populations of the Montreal area now constitute a reservoir of the remnant genetic diversity. Moreover, this study not only demonstrates that the genetic response to habitat loss can be very case-specific, but also that to detect traces of such a phenomenon can require a very particular framework. A growing number of publications based on birds and mammals have reported the absence of a genetic signature following a habitat loss. This is the first study to report this phenomenon in reptiles. A note intended for managers is provided about the conservation of the Dekay’s brown snake, S. dekayi.

Invasões biológicas representam atualmente a segunda maior ameaça à biodiversidade e dois fatores são considerados os mais importantes para o sucesso de invasões: pressão de propágulos e distúrbios. Um tipo de distúrbio antrópico que pode promover invasões, por mudar a quantidade de habitats alterados e a extensão de bordas entre eles e habitats nativos, é a perda de habitat. Apesar da reconhecida importância da pressão de propágulos e dos distúrbios, poucos estudos os investigaram simultaneamente e, os que o fizeram, apresentam limitações, como escalas espaciais pequenas e correlações entre os determinantes, dificultando a compreensão da importância relativa e interações entre eles. Cachorros são os carnívoros mais abundantes no mundo. Em áreas rurais, a maioria mantém comportamento de animal de vida de livre, interagindo com e afetando espécies nativas através de predação, transmissão de doenças e competição. Usando um banco de dados obtido através de armadilhas fotográficas e censo da população de cachorros em uma região de Mata Atlântica de 300,000 ha, avaliamos a importância relativa e interações entre pressão de propágulos e distúrbios para a invasão por cachorros. Selecionamos 12 paisagens de 2830 ha cada, variando de 10 a 50% de floresta nativa remanescente. Em cada uma, alocamos através de amostragem aleatória estratificada 8 pontos de amostragem em florestas nativas, onde uma armadilha fotográfica foi instalada por ∼42 dias consecutivos. Todos os domicílios em cada paisagem foram visitados para a contagem do número de cachorros. A pressão de propágulos foi quantificada como a densidade de cachorros criados e a média e mediana das distâncias entre os locais de criação e a floresta nativa mais próxima; e distúrbios, como a proporção da paisagem ocupada por floresta nativa e total (nativa e exótica) e a extensão de bordas entre florestas nativas e áreas abertas. Através da identificação de cachorros nas fotos e considerando cada paisagem como uma unidade amostral, nós comparamos por AICc modelos de abundância (N-mixture) para estimar a abundância de cachorros em florestas nativas, considerando a detecção imperfeita. O único modelo selecionado indica que a abundância de cachorros em florestas nativas é maior onde a densidade de cachorros é mais alta e a cobertura florestal total é menor (ωi=0.82). A densidade de cachorros criados foi mais importante que a distribuição espacial dos indivíduos, e a cobertura floresta total mais importante que a extensão de bordas, para a abundância de cachorros invasores. A abundância estimada de cachorros variou de 12 a 79 (30.9 ± 19.5), e a proporção de cachorros criados que invadem florestas de 6 a 21% (12 ± 6%), entre as paisagens. Nossos resultados indicam que a perda de habitat é tão importante quanto a pressão de propágulos para a invasão de florestas nativas por cachorros, mas seus efeitos são aditivos em vez de sinérgicos. Dado que cachorros frequentemente realizam movimentos longos em áreas abertas, nós levantamos a hipótese de que a capacidade de deslocamento é a causa do efeito desprezível da distribuição espacial dos indivíduos criados sobre a invasão, e que florestas representam barreiras a estes movimentos, tornando o efeito da cobertura florestal mais importante do que o efeito da extensão de bordas (mais relacionada a extensão de acesso a floresta). Além disso, o número e proporção de cachorros invasores são expressivos, colocando o cachorro na posição de carnívoro mais abundante em remanescentes florestais. Junto com os conhecidos impactos severos de cachorros sobre espécies nativas, estes números sugerem a urgência de planos de ação para controlar a invasão por cachorros. Além dos métodos tradicionais de controle populacional, o contexto da paisagem deve ser levado em conta nestes planos. Paisagens muito desmatadas devem ser priorizadas, e manter e restaurar florestas também devem ser valorizados pelos efeitos negativos sobre invasões biológicas. Por fim, dada a associação da invasão por cachorros com a perda de habitat e com a densidade de cachorros e da população humana, sugerimos que pelo menos parte dos efeitos negativos sobre mamíferos nativos usualmente atribuídos ao desmatamento e a caça podem ser causados pela invasão por cachorros / Biological invasions are currently the second main threat to biodiversity and two drivers are considered as the most important for invasions success: propagule pressure and disturbance. An anthropogenic disturbance that can promote invasions, by changing the amount of altered habitats and the extension of edges between altered and native habitats, is habitat loss. Despite the recognized importance of propagule pressure and disturbance, few studies have simultaneously investigated these factors, and those that did so present limitations, such as small spatial scales and correlations between drivers, impairing our understanding of the relative importance and interactions between these drivers. Dogs are the most abundant carnivores worldwide; in rural areas, most are free ranging, interacting and affecting native species through predation, disease transmission and competition. Using a camera trap dataset and censuses of dog populations obtained across a 300,000-ha Atlantic forest region, we evaluated the relative importance and interactions of propagule pressure and disturbance as drivers of dog invasion. We selected 12 2830-ha landscapes, ranging from 10 to 50% remaining native forest. Within each, we selected through a random-stratified sample 8 forest sites where a camera trap was set for ∼42 consecutive days. All households in each landscape were visited to count the number of dogs. Propagule pressure was quantified as the density of raised dogs, and mean and median distances between locations where dogs were raised and the nearest forest; and disturbance as the proportion of the landscape occupied by native forest and by total forest (native and exotic), and edge extension between native forest and open areas. By identifying individual dogs in the photos and considering each landscape as a sampling unit, we compared through AICc N-mixture models to estimate the abundance of dogs within forests, considering imperfect detection. The only selected model indicates that dog abundance in forests is higher where the density of raised dogs is higher and where total forest cover is lower (ωi=0.82). Density of raised dogs was more important than the spatial distribution of individuals, and total forest cover more important than edge extension, in determining the abundance of invading dogs. The estimated dog abundance varied from 12 to 79 (30.9 ± 19.5), and the proportion of raised dogs that invade forests from 6 to 21% (12 ± 6%), across landscapes. Our results indicate that habitat loss is as important as propagule pressure in driving the invasion of native forests by dogs, but their effects are additive rather than synergic. Given that dogs frequently make long movements in open areas, we hypothesize that dog vagility is the cause of the negligible effect of spatial distribution of raised individuals on invasion, and that forests represent barriers to these movements, making the effect of forest cover more important than the effect of edge extension (more related to the extension of access to forests). Moreover, the number and proportion of invading dogs are impressive, ranking dogs as the most abundant carnivore in forest remnants. Together with the known severe impacts of dogs on native species, these numbers suggest the urgency of action plans for controlling dog invasion. Beyond the traditional population control, landscape context should be taken into account within strategies to reduce impacts of dogs. Highly-deforested landscapes should be prioritized, and maintaining and restoring forests should be valued also by their negative effects on biological invasions. Finally, given the observed associations between dog invasion and both habitat loss and density of dogs and human populations, we suggest that at least part of the negative effects on native mammals currently attributed to deforestation and hunting can be caused by dog invasion