L’anthropisation est un phénomène mondial qui affecte plusieurs taxa qui se retrouvent dans des milieux variés, autant terrestres qu’aquatiques. Même si, pour quelques espèces, ces changements sont bénéfiques, pour la plupart d’entres-elles ils entrainent des conséquences désastreuses pouvant mener au déclin de leurs populations. L’avifaune insectivore aérienne champêtre n’est pas épargnée par ce phénomène puisque la détérioration des aires de reproduction par l’intensification des pratiques agricoles a participé, de manière non équivoque, au déclin de cette faune. En effet, plusieurs études ont montré des effets négatifs de l’intensification agricole sur certains traits d’histoire de vie de ses oiseaux (p. ex. : diminution de la croissance et du succès reproducteur chez les oiseaux de milieux intensifs vs milieux extensifs). Afin d’expliquer ces effets, l’hypothèse la plus souvent suggérée se base sur le fait que dans les milieux agricoles intensifs où des insecticides sont utilisés, la disponibilité en insectes est inférieure aux milieux naturels. Donc, par l’entremise d’une diminution en abondance et en diversité d’insectes disponibles dans les milieux agricoles, la diète des oiseaux est affectée ce qui peut potentiellement participer au déclin des insectivores aériens. Or, peu d’études ont tenté de déterminer la diète d’un insectivore aérien champêtre en déclin dans les milieux agricoles ou ont tenté de déterminer les variables qui peuvent influencer sa sélection de proies, changeant potentiellement la composition même de la diète. Sachant que 1) la diète insectivore est influencée par la disponibilité de proies dans le milieu; que 2) la disponibilité de proies est influencée par l’environnement (paysage, conditions météorologiques et temporelles, etc.); et que 3) les stratégies de quête alimentaire de l’insectivore aérien (fréquence et durée des quêtes, nombre de proies capturées, stratégies de sélection et d’évitement de proies, etc.) fluctuent et dépendent de la disponibilité de proies dans l’environnement et de l’environnement lui-même, je me suis intéressée, dans ce mémoire, aux fluctuations dans la sélection de proies d’un insectivore aérien, selon différentes variables. Plus particulièrement, je me suis intéressée à l’effet de l’intensification agricole, de la disponibilité de proies dans les milieux agricoles, de la condition morphologique de la femelle, de l’état de sa niché et des conditions environnementales et temporelles sur la sélection de proies d’un insectivore aérien en déclin. Ainsi, dans cette étude, des becquées alimentaires ont été récoltées chez l’Hirondelle bicolore (Tachycineta bicolor), passereau insectivore aérien en déclin dans le Nord-Est de l’Amérique du Nord depuis plus de 25 ans, puis comparées à des échantillons d’insectes disponibles, tous deux récoltés le long d’un gradient d’intensification agricole dans de Sud du Québec. Mes résultats montrent d’abord que les diptères représentent l’ordre le plus important dans la diète de l’Hirondelle bicolore et supportent l’idée d’un effet négatif de l’intensification des pratiques agricoles sur l’abondance en arthropodes disponibles dans le milieu. Par ailleurs, j’ai également montré que les associations entre les composantes du paysage agricole et certains taxa d’insectes étaient différentes entre les becquées et les pièges à insectes, suggérant une potentielle sélection de proies de la part des hirondelles. J’ai finalement déterminé les variables qui affectent le plus fortement la sélection de diptères par l’Hirondelle bicolore, c’est-à-dire la disponibilité de diptères dans le milieu, l’intensification agricole et la date de récolte des échantillons et ce, contrôlé pour des variables météorologiques et temporelles confondantes Cette étude est, jusqu’à présent, la plus développée et complète sur ce sujet tant au niveau de la couverture spatio-temporelle qu’au niveau de l’ampleur de l’échantillonnage qui aboutit en une base de données inégalée De plus, seule l’approche statistique complexe réalisée dans le cadre de cette étude permet d’adéquatement caractériser les déterminismes de la sélection de proies qui visent à comprendre une réponse comportementale potentielle de l’insectivore aérien face à des pressions au niveau des proies lui étant disponibles dans les milieux agricoles causée par l’intensification des pratiques agricoles.

The eastern whip-poor-will (Antrostomus vociferus: hereafter whip-poor-will) has been declining from historical population levels throughout its range in the northeast. Although whip-poor-wills have been reported to use a variety of habitats, most recent studies have associated whip-poor-wills with open habitat, such as early-successional habitats or forest edges. Population declines of other early-successional bird species have been attributed to the loss of early-successional disturbance-dependent habitats in the northeast, and it has been suggested that habitat loss is a significant factor in whip-poor-will population declines, as well. However, there remain substantial gaps in our understanding of whip-poor-will habitat associations, and quantitative habitat data in the literature are lacking. As forest management plays an important role in creating and maintaining habitat for many disturbance-dependent bird species, further characterization of whip-poor-will habitat preferences is necessary to determine whether management efforts may benefit this species as well. In order to derive quantitative estimates of habitat requirements, I studied whip-poor-will habitat associations at Fort Drum in upstate New York. In 2015 and 2016, whip-poor-wills were surveyed at night at randomly-selected point count locations and vegetation measurements were collected in the point count radii to relate whip-poor-will occupancy with structural and compositional habitat variables. Whip-poor-will occupancy was strongly related to intermediate amounts of basal area, with values that generally correspond to forest denser than most shrublands, but more open than closed-canopy forest. Occupancy was also related to lower understory height values, which supports evidence that whip-poor-wills may prefer habitat with a relatively open understory. In 2016, I also measured habitat at locations where whip-poor-wills were foraging, roosting, and nesting, to investigate the theory that whip-poor-wills require open habitat for foraging, but more closed habitat for nesting. Ten adult whip-poor-wills were tracked using radio telemetry and vegetation measurements were collected at a subset at these points where the birds were either foraging or roosting during the day, as well as at any identified nest sites. Comparisons of the vegetation measurements revealed that foraging habitat was significantly more open than roosting habitat, as foraging habitat had lower tree density, basal area, and understory height. Contrary to conventional thought, the few nest sites found in this study were in areas that had low basal area, similar to the habitat at foraging locations. The results suggest that while creating more open-canopy habitat may benefit whip-poor-wills by providing suitable foraging habitat, and potentially nesting habitat, maintaining denser forest within proximity to these open areas may also provide valuable cover for roosting whip-poor-wills. In conclusion, I suggest that land owners looking to create or maintain suitable habitat for whip-poor-wills apply forest management treatments that create openings but still maintain intermediate levels of basal area, such as shelterwood or group tree selection. Foraging habitat for whip-poor-wills appears to be generally more open than roosting habitat, both in terms of lower basal area and a more open understory, so having areas where tree and understory removal is concentrated in proximity to areas that are denser may also benefit this species.

Human-driven climate, habitat, and land use changes often co-occur in ecological communities. We must consider the multiple components of global change acting on individual species and assemblages to document biological responses to environmental change and determine the mechanisms underlying these responses. Here, I examine climate, habitat, and land management impacts on a model aerial insectivore, the Tree Swallow (Tachycineta bicolor), and its insect prey. Both groups are undergoing population decline and phenological shifts in many parts of the world; however, the magnitude and mechanisms of these shifts are not well understood. I first document the impacts of temperature and precipitation on fall flight times of 20 butterfly species with varied life histories in Massachusetts. I find many butterfly species are flying later into the fall now than they were over 20 years ago; however, the response of butterflies to fall climate is complex and often mediated by life history characteristics, like number of broods per season. I then examine the effects of climate, habitat, and insect prey abundance on Tree Swallow reproduction to determine if anthropogenic changes in the breeding habitat result in declines in reproductive performance that contribute to population decline. I find that climate and foraging habitat impact egg laying phenology, clutch size, hatching success and fledging success of Tree Swallows. For example, reproductive phenology is delayed in rainy springs and fledging success is increased in nests with open water in their foraging radius. However, I find no evidence to indicate a change in insect abundance or anthropogenic changes, including climate and land management, are driving decreases in reproductive success of Tree Swallows over time. Lastly, I examine the effects of artificial nest management on Tree Swallow reproduction across Mass Audubon conservation areas. I find habitat, density, and predation of artificial nests to be strong and often overlooked determinants of Tree Swallows reproductive performance; for example, fledging is increased in nests placed in open habitat, far from forests and developed areas. This work provides novel evidence for the impacts of local-scale nest habitat and management on Massachusetts Tree Swallows, a threatened population of aerial insectivores.

Understanding the patterns and processes that create differences among individuals in components of fitness, like the probability of survival or reproductive rates, is essential to our knowledge of population dynamics and for informing conservation efforts. For organisms in seasonal environments, early-breeding individuals regularly attain higher fitness than their late-breeding counterparts. Two primary hypotheses, related to quality and date, have been proposed to explain lower reproductive success of late breeders, but the veracity of these ideas has not been fully resolved. I tested predictions associated with these hypotheses to assess the effects of indices of parental and environmental quality on nestling quality and survival in an insectivorous passerine, the tree swallow (Tachycineta bicolor), at two widely separated breeding locations in western Canada. I combined experiments and statistical modelling of observational data to evaluate two mechanisms proposed to contribute to seasonal decline in environmental quality: an increase in nest parasite abundance and a decrease in food abundance with later breeding dates. A parasite reduction experiment revealed a disproportionate benefit of parasite removal on length of primary feather for early-hatched nestlings, suggesting greater energetic constraints early in the breeding season. Furthermore, late-hatched nestlings from parasite-reduced nests had longer head-bill lengths than their control counterparts, and developed head-bills of similar length to those of early-hatched nestlings. Other than these findings, there were few detectable effects of parasites on nestling size, growth and immunity, as has been reported from several previous studies. Indeed, negative effects of parasites were only apparent when food (i.e., insect) biomass was considered. In a second series of experiments in which parental quality was controlled, I also tested whether food abundance declined during the breeding season, as predicted if environmental conditions deteriorate seasonally (i.e., date). Reduced reproductive success of late-breeding individuals was causally related to a seasonal decline in environmental quality. Declining insect biomass and enlarged brood sizes resulted in nestlings that were lighter, in poorer body condition, had shorter head-bills, shorter and slower growing ninth primary feathers and that were less likely to survive to fledge. Next, I asked whether results obtained from long-term mark-recapture data corroborated findings of short-term manipulations. I examined seasonal variation in first-year apparent survival to investigate the relative influence of large-, small- and individual-scale factors associated with the quality and date hypotheses. Although parental quality was an important predictor of first-year apparent survival of tree swallows, my results further suggested that quality of parents was not the primary factor influencing seasonal variation in first-year apparent survival. Rather, findings were most consistent with the date hypothesis. The relationship between apparent survival and a direct measurement of environmental quality indicated that annual variation in moisture had important consequences for first-year apparent survival of tree swallows in Saskatchewan. First-year apparent survival probabilities were higher during wet years and wetter conditions are generally linked to greater insect abundance. In British Columbia, nestlings from larger broods were less likely to survive, possibly as a result of receiving less food. Apparent survival probabilities were also higher when food was more abundant. I demonstrated that both parental and environmental quality influenced seasonal variation in fitness-related traits of tree swallows. However, the strongest evidence suggests that environmental quality, and in particular food abundance, had the greatest effect on seasonal variation in nestling quality, reproductive success and first-year apparent survival in tree swallows. My results highlight the importance of considering regional precipitation trends when projecting effects of climate change on demography of aerial insectivores.

We reviewed the taxonomy of the Trogon rufus species-complex under the premises of the Biological Species Concept. Putative taxonomic units, breaks and transition zones, were visualised by heatmaps and isophenes (phenotypic contour lines) of the colour, barring, morphometric and song characters and tested by discriminant function analyses. Colourmetric data were obtained via spectrometry and barring patterns analysed via high quality digital photographs. We found four distinct biological species. Trogon chrysochloros Pelzeln 1856 from the Atlantic Forest with its denser and blacker undertail and wing covert barring, larger size and faster, generally higher song with more notes. Its upperparts vary from bluer to more coppery-green with increasing altitude. The bill is also relatively smaller and more serrated, linked to a diet that consists almost exclusively of large arthropods, making it the most insectivorous new world Trogon species yet known, which may account for its relative rarity compared to other Trogonids with which it is sympatric. Trogon tenellus Cabanis 1862, from Central America, and Trogon cupreicauda Chapman 1914 from the Chocó-Magdalena provide a classic case of typical biological species, coming into contact in the extreme NW Chocó Province, Colombia, but without intermediate forms. T. tenellus is identified by its blue to blue-green uppertail, blue or grey eye-rings, grey tarsi and song with 2-4 notes, longer note duration and greater change in peak and high frequencies between the intro note and loudsong. This contrasts with the shiny olive-green to coppery green uppertails, yellow eye-ring, usually olive tarsi, brown wash on the undertail of females and song with 6-8 notes of shorter duration and little change in frequency between the intro note and loudsong of T. cupreicauda. T. cupreicauda varies clinally from generally bluer- to more coppery-green plumage and from thicker to thinner black bars in a gradient from the Pacific coast on the border with Ecuador to the Magdalena Valley. The greater difference in colour and barring relative to T. tenellus in the region they come into contact provides possible evidence of character displacement as a result of the competitive exclusion between these two species, maintaining their parapatric distributions. The Amazonian population belongs to a single species, Trogon rufus Gmelin 1788, but with two highly distinct forms that we designate as Trogon rufus rufus in the Guiana Shield and Trogon rufus sulphureus in S & W Amazonia, for which Todd\'s amazonicus is synonymised. They are morphologically and, to a lesser extent, vocally distinct across the lower Rio Negro and matrix of highland and open habitats of the Rio Branco basin but show limited character exchange between the 52-58th parallels west on the southern bank of the Amazon, centred around the Rio Arapiuns on the left bank of mouth of the Tapajos. We postulate that this is the result of secondary contact as a consequence of shifts in the course of the main channel of the Amazon River at times of lower sea levels during the Plio-Pleistocene. T .r. sulphureus is identified by a typically coppery uppertail with subterminal tailband of greener hue, yellow eye-ring, low barring density and broad black bars of the undertail and wing-coverts barring with and lack of a pectoral band. They are also sometimes distinguishable in song by a higher frequency introduction note and/or more pronounced descending modulation across the loudsong. This varies clinally on a west-east gradient, from strong-coppery to shiny olive-green uppertails with more to less distinct subterminal tailbands, diminishing black bar widths with corresponding increasing density and decreasing intro note low frequency. T. r. rufus have green uppertails, blue eye-rings, presence or absence of a white pectoral band and denser undertail and wing panel barring with thinner black bars. These characters were shown to change as a function of geographic distance between specimens of sulphureus and rufus, connected via the \'Arapiuns contact zone\', suggesting isolation by distance. This is reminiscent of a ring species pattern and two specimens with a possible mixture of characters were indeed found from the upper Rio Negro and in Pantepui, where T. r. rufus and T. r. sulphureus would be expected to come into contact, effectively \'closing the ring\'. Whether Trogon rufus constitutes a valid ring species requires further testing, preferably including molecular characters, but this clearly illustrates that the distinction between clinal variation and ring-species is a matter of degree, not kind, with the formation of the ring-species necessarily passing through a clinal stage with no overlap between terminal taxa. We therefore propose the concept of a loop species, where the terminal forms do not overlap but are connected via a series of intergrading populations. It seems likely that such patterns are more widespread in Amazonia than presently known due to the propensity for clinal variation and parapatric speciation lended by its massive geographical extent and abundance of biogeographical semi-permeable barriers. With regards to the population from the Pernambuco Center of Endemism, the few records suggest that it is a valid taxonomic unit. It has the unique combination of a song very similar to T. r. sulphureus due to the high introduction note frequency and pronounced descent in frequencies across the loudsong, with a corresponding widening range but moderately large size, serrated bill and blue eye-ring but this certainly requires confirmation. This requires urgent attention, as the remnant population is very small and localised, recorded only from the Murici municipality, Alagoas. / Nós revisamos a taxonomia do complexo Trogon rufus sob o conceito Biológico de Espécies. Unidades taxonômicas possíveis, quebras e zonas de transição taxonômicas foram definidas usando mapas de calor e isofenas (linhas de contorno de fenótípo) baseados em caráteres de cor, barramento e morfometria. Esses possíveis táxons foram testados pelas análises de discriminantes. Dados de coloração foram obtidos por meio de espectrometria, e os padrões de barramento por meio de fotos de alta qualidade. Nós encontramos quatro espécies biológicas distintas Trogon chrysochloros Pelzeln 1856, da Mata Atlântica, diagnosticável pelo barramento mais escuro e denso na face inferior da cauda e coberteiras da asa, maior tamanho corpóreo, canto mais rápido, com mais notas e de frequência mais alta. Suas partes superiores (cabeça, dorso e cauda) variam de azul a verde acobreado com o aumento da altitude. O bico também é relativamente menor e mais serrilhado, o que está ligado a uma dieta que consiste quase exclusivamente de grandes artrópodes, fazendo desta espécie o Trogon mais insetívoro do mundo, o que deve ser a razão de sua relativa raridade quando comparado a outros Trogonidae com os quais vive em simpatia. Trogon tenellus Cabanis 1862, da América Central, e Trogon cupreicauda Chapman 1914, do Chocó-Magdalena, formam um caso clássico de espécies biológicas, entrando em contato no extremo noroeste da província de Chocó, na Colômbia, sem a presença de formas intermediárias. T. tenellus é identificável pela face superior da cauda azul ou azul esverdeado, anel perioftálmico azul ou cinza, tarso cinza, e voz com de 2 a 4 notas, maior duração das notas e maior mudança entre frequência de pico e frequência alta entre a nota introdutória e nota principal. Essas características contrastam com a cor verde-oliva brilhante da face superior da cauda, anel perioftámico amarelo, tarso geralmente oliva, presença de marrom claro na face inferior da cauda das fêmeas e canto com entre 6 e 8 notas, de menor duração, pouca mudança na frequência entre a primeira nota e o canto principal de T. cupreicauda. Este varia clinalmente de azul para uma plumagem mais verde acobreada e de barramento preto mais espesso para mais fino em uma gradiente da costa do Pacífico, do noroeste do Equador até o Vale do Magdalena. A maior distinção de estados de caráter relativa a T. tenellus na região onde os dois grupos se encontram provê uma possível evidência de deslocamento de caracteres como resultado de exclusão competitiva entre estas duas espécies, mantendo suas distribuições parapátricas. A população amazônica pertence a única espécie biológica Trogon rufus Gmelin 1788, mas com duas formas altamente distintas que designamos como Trogon rufus rufus do Escudo Guianense, e Trogon rufus sulphureus no sul e oeste da Amazônia, com a qual amazonicus de Todd é sinonimizada. Estas são morfologicamente, e em menor escala, vocalmente distintas nas duas margens do baixo Rio Negro e áreas abertas e/ou montanhosas da bacia do Rio Branco, mas apresentam troca de caracteres limitada nas longitudes entre 52 e 58 Oeste na margem sul do Rio Amazonas, centrado nos arredores do Rio Arapiuns, na margem esquerda da foz do Tapajós. Nós postulamos que isto é um resultado de contato secundário, como consequência de mudanças no curso principal do Rio Amazonas em tempos de níveis mais baixos do mar durante o Plio-Pleistoceno. T. r. sulphureus é identificado pela coloração tipicamente acobreada da face superior da cauda com uma banda sub-terminal de tonalidade esverdeada, anel perioftálmico amarelo, barras negras espessas e de baixa densidade na face inferior da cauda e nas coberteiras das asas e pela ausência de uma faixa peitoral. Em certos casos eles também podem ser distinguíveis pela voz com uma nota introdutória de maior frequência e/ou uma modulação descendente mais pronunciada no canto principal. Este grupo varia clinalmente em gradiente de oeste para leste, de uma coloração cobre forte até verde oliva na face superior da cauda com faixa sub-terminal mais ou menos distinta, anel perioftálmico azul, presença ou ausência de uma faixa peitoral branca e barramento na face inferior da cauda e coberteiras de menor espessura e, consequentemente, maior densidade. Nossos dados apontam que estes caracteres mudam linearmente entre sulphureus e rufus de acordo com a distância ao longo de uma gradiente clinal em forma de laço do Oeste Amazônico até o Escudo das Guianas, conectado pela zona de contato de Arapiuns, sugerindo isolamento por distância. Isto sugere um remanescente de um padrão de espécie em anel. Além disso, dois espécimes com possível mistura de caracteres foram de fato encontrados no alto Rio Negro e Pantepui, onde se espera que T. r. rufus e T. r. sulphureus entrem em contato, fechando o anel efetivamente. Ainda é necessário testar se Trogon rufus constitui uma espécie em anel válida, preferencialmente usando dados moleculares, mas este caso ilustra claramente que a distinção entre variação clinal e espécie em anel é uma questão de grau e não de tipo, com a formação de espécie em anel necessariamente passando por um estágio clinal sem sobreposição entre os grupos terminais. Nós assim sugerimos o conceito de espécie em laço, onde as formas terminais não se sobrepõem, mas são ligadas através de uma série de populações onde há fluxo gênico. Parece provável que estes padrões são mais amplamente distribuídos na Amazônia do que é sabido atualmente devido a uma propensão à variação clinal e especiação parapátrica causada por sua grande extensão geográfica e abundância de barreiras geográficas semipermeáveis. Em relação a população do Centro de Endemismo Pernambuco, os poucos registros sugerem que esta é uma unidade taxonômica válida. Este grupo apresenta uma combinação única de canto muito similar ao de T. r. sulphureus, devido à alta frequência da nota introdutória e pronunciada modulação descendente ao longo do canto principal, com uma maior amplitude da frequência, combinada a maior tamanho, bico serrilhado, e anel perioftálmico azul, mas isso requer confirmação. Este caso demanda atenção urgente, já que a população remanescente é muito pequena e de distribuição restrita ao município de Murici, em Alagoas.

We reviewed the taxonomy of the Trogon rufus species-complex under the premises of the Biological Species Concept. Putative taxonomic units, breaks and transition zones, were visualised by heatmaps and isophenes (phenotypic contour lines) of the colour, barring, morphometric and song characters and tested by discriminant function analyses. Colourmetric data were obtained via spectrometry and barring patterns analysed via high quality digital photographs. We found four distinct biological species. Trogon chrysochloros Pelzeln 1856 from the Atlantic Forest with its denser and blacker undertail and wing covert barring, larger size and faster, generally higher song with more notes. Its upperparts vary from bluer to more coppery-green with increasing altitude. The bill is also relatively smaller and more serrated, linked to a diet that consists almost exclusively of large arthropods, making it the most insectivorous new world Trogon species yet known, which may account for its relative rarity compared to other Trogonids with which it is sympatric. Trogon tenellus Cabanis 1862, from Central America, and Trogon cupreicauda Chapman 1914 from the Chocó-Magdalena provide a classic case of typical biological species, coming into contact in the extreme NW Chocó Province, Colombia, but without intermediate forms. T. tenellus is identified by its blue to blue-green uppertail, blue or grey eye-rings, grey tarsi and song with 2-4 notes, longer note duration and greater change in peak and high frequencies between the intro note and loudsong. This contrasts with the shiny olive-green to coppery green uppertails, yellow eye-ring, usually olive tarsi, brown wash on the undertail of females and song with 6-8 notes of shorter duration and little change in frequency between the intro note and loudsong of T. cupreicauda. T. cupreicauda varies clinally from generally bluer- to more coppery-green plumage and from thicker to thinner black bars in a gradient from the Pacific coast on the border with Ecuador to the Magdalena Valley. The greater difference in colour and barring relative to T. tenellus in the region they come into contact provides possible evidence of character displacement as a result of the competitive exclusion between these two species, maintaining their parapatric distributions. The Amazonian population belongs to a single species, Trogon rufus Gmelin 1788, but with two highly distinct forms that we designate as Trogon rufus rufus in the Guiana Shield and Trogon rufus sulphureus in S & W Amazonia, for which Todd\'s amazonicus is synonymised. They are morphologically and, to a lesser extent, vocally distinct across the lower Rio Negro and matrix of highland and open habitats of the Rio Branco basin but show limited character exchange between the 52-58th parallels west on the southern bank of the Amazon, centred around the Rio Arapiuns on the left bank of mouth of the Tapajos. We postulate that this is the result of secondary contact as a consequence of shifts in the course of the main channel of the Amazon River at times of lower sea levels during the Plio-Pleistocene. T .r. sulphureus is identified by a typically coppery uppertail with subterminal tailband of greener hue, yellow eye-ring, low barring density and broad black bars of the undertail and wing-coverts barring with and lack of a pectoral band. They are also sometimes distinguishable in song by a higher frequency introduction note and/or more pronounced descending modulation across the loudsong. This varies clinally on a west-east gradient, from strong-coppery to shiny olive-green uppertails with more to less distinct subterminal tailbands, diminishing black bar widths with corresponding increasing density and decreasing intro note low frequency. T. r. rufus have green uppertails, blue eye-rings, presence or absence of a white pectoral band and denser undertail and wing panel barring with thinner black bars. These characters were shown to change as a function of geographic distance between specimens of sulphureus and rufus, connected via the \'Arapiuns contact zone\', suggesting isolation by distance. This is reminiscent of a ring species pattern and two specimens with a possible mixture of characters were indeed found from the upper Rio Negro and in Pantepui, where T. r. rufus and T. r. sulphureus would be expected to come into contact, effectively \'closing the ring\'. Whether Trogon rufus constitutes a valid ring species requires further testing, preferably including molecular characters, but this clearly illustrates that the distinction between clinal variation and ring-species is a matter of degree, not kind, with the formation of the ring-species necessarily passing through a clinal stage with no overlap between terminal taxa. We therefore propose the concept of a loop species, where the terminal forms do not overlap but are connected via a series of intergrading populations. It seems likely that such patterns are more widespread in Amazonia than presently known due to the propensity for clinal variation and parapatric speciation lended by its massive geographical extent and abundance of biogeographical semi-permeable barriers. With regards to the population from the Pernambuco Center of Endemism, the few records suggest that it is a valid taxonomic unit. It has the unique combination of a song very similar to T. r. sulphureus due to the high introduction note frequency and pronounced descent in frequencies across the loudsong, with a corresponding widening range but moderately large size, serrated bill and blue eye-ring but this certainly requires confirmation. This requires urgent attention, as the remnant population is very small and localised, recorded only from the Murici municipality, Alagoas. / Nós revisamos a taxonomia do complexo Trogon rufus sob o conceito Biológico de Espécies. Unidades taxonômicas possíveis, quebras e zonas de transição taxonômicas foram definidas usando mapas de calor e isofenas (linhas de contorno de fenótípo) baseados em caráteres de cor, barramento e morfometria. Esses possíveis táxons foram testados pelas análises de discriminantes. Dados de coloração foram obtidos por meio de espectrometria, e os padrões de barramento por meio de fotos de alta qualidade. Nós encontramos quatro espécies biológicas distintas Trogon chrysochloros Pelzeln 1856, da Mata Atlântica, diagnosticável pelo barramento mais escuro e denso na face inferior da cauda e coberteiras da asa, maior tamanho corpóreo, canto mais rápido, com mais notas e de frequência mais alta. Suas partes superiores (cabeça, dorso e cauda) variam de azul a verde acobreado com o aumento da altitude. O bico também é relativamente menor e mais serrilhado, o que está ligado a uma dieta que consiste quase exclusivamente de grandes artrópodes, fazendo desta espécie o Trogon mais insetívoro do mundo, o que deve ser a razão de sua relativa raridade quando comparado a outros Trogonidae com os quais vive em simpatia. Trogon tenellus Cabanis 1862, da América Central, e Trogon cupreicauda Chapman 1914, do Chocó-Magdalena, formam um caso clássico de espécies biológicas, entrando em contato no extremo noroeste da província de Chocó, na Colômbia, sem a presença de formas intermediárias. T. tenellus é identificável pela face superior da cauda azul ou azul esverdeado, anel perioftálmico azul ou cinza, tarso cinza, e voz com de 2 a 4 notas, maior duração das notas e maior mudança entre frequência de pico e frequência alta entre a nota introdutória e nota principal. Essas características contrastam com a cor verde-oliva brilhante da face superior da cauda, anel perioftámico amarelo, tarso geralmente oliva, presença de marrom claro na face inferior da cauda das fêmeas e canto com entre 6 e 8 notas, de menor duração, pouca mudança na frequência entre a primeira nota e o canto principal de T. cupreicauda. Este varia clinalmente de azul para uma plumagem mais verde acobreada e de barramento preto mais espesso para mais fino em uma gradiente da costa do Pacífico, do noroeste do Equador até o Vale do Magdalena. A maior distinção de estados de caráter relativa a T. tenellus na região onde os dois grupos se encontram provê uma possível evidência de deslocamento de caracteres como resultado de exclusão competitiva entre estas duas espécies, mantendo suas distribuições parapátricas. A população amazônica pertence a única espécie biológica Trogon rufus Gmelin 1788, mas com duas formas altamente distintas que designamos como Trogon rufus rufus do Escudo Guianense, e Trogon rufus sulphureus no sul e oeste da Amazônia, com a qual amazonicus de Todd é sinonimizada. Estas são morfologicamente, e em menor escala, vocalmente distintas nas duas margens do baixo Rio Negro e áreas abertas e/ou montanhosas da bacia do Rio Branco, mas apresentam troca de caracteres limitada nas longitudes entre 52 e 58 Oeste na margem sul do Rio Amazonas, centrado nos arredores do Rio Arapiuns, na margem esquerda da foz do Tapajós. Nós postulamos que isto é um resultado de contato secundário, como consequência de mudanças no curso principal do Rio Amazonas em tempos de níveis mais baixos do mar durante o Plio-Pleistoceno. T. r. sulphureus é identificado pela coloração tipicamente acobreada da face superior da cauda com uma banda sub-terminal de tonalidade esverdeada, anel perioftálmico amarelo, barras negras espessas e de baixa densidade na face inferior da cauda e nas coberteiras das asas e pela ausência de uma faixa peitoral. Em certos casos eles também podem ser distinguíveis pela voz com uma nota introdutória de maior frequência e/ou uma modulação descendente mais pronunciada no canto principal. Este grupo varia clinalmente em gradiente de oeste para leste, de uma coloração cobre forte até verde oliva na face superior da cauda com faixa sub-terminal mais ou menos distinta, anel perioftálmico azul, presença ou ausência de uma faixa peitoral branca e barramento na face inferior da cauda e coberteiras de menor espessura e, consequentemente, maior densidade. Nossos dados apontam que estes caracteres mudam linearmente entre sulphureus e rufus de acordo com a distância ao longo de uma gradiente clinal em forma de laço do Oeste Amazônico até o Escudo das Guianas, conectado pela zona de contato de Arapiuns, sugerindo isolamento por distância. Isto sugere um remanescente de um padrão de espécie em anel. Além disso, dois espécimes com possível mistura de caracteres foram de fato encontrados no alto Rio Negro e Pantepui, onde se espera que T. r. rufus e T. r. sulphureus entrem em contato, fechando o anel efetivamente. Ainda é necessário testar se Trogon rufus constitui uma espécie em anel válida, preferencialmente usando dados moleculares, mas este caso ilustra claramente que a distinção entre variação clinal e espécie em anel é uma questão de grau e não de tipo, com a formação de espécie em anel necessariamente passando por um estágio clinal sem sobreposição entre os grupos terminais. Nós assim sugerimos o conceito de espécie em laço, onde as formas terminais não se sobrepõem, mas são ligadas através de uma série de populações onde há fluxo gênico. Parece provável que estes padrões são mais amplamente distribuídos na Amazônia do que é sabido atualmente devido a uma propensão à variação clinal e especiação parapátrica causada por sua grande extensão geográfica e abundância de barreiras geográficas semipermeáveis. Em relação a população do Centro de Endemismo Pernambuco, os poucos registros sugerem que esta é uma unidade taxonômica válida. Este grupo apresenta uma combinação única de canto muito similar ao de T. r. sulphureus, devido à alta frequência da nota introdutória e pronunciada modulação descendente ao longo do canto principal, com uma maior amplitude da frequência, combinada a maior tamanho, bico serrilhado, e anel perioftálmico azul, mas isso requer confirmação. Este caso demanda atenção urgente, já que a população remanescente é muito pequena e de distribuição restrita ao município de Murici, em Alagoas.