This DATSETNAMEreadme.txt file was generated on 20200226 by Provini ------------------- GENERAL INFORMATION ------------------- Title of Dataset: Supplementary Material associated with the publication Provini, Pauline; Höfling, Elizabeth (2020), To hop or not to hop? the answer is in the bird trees. Systematic Biology Author Information: Principal Investigator: Pauline Provini, Universidade de São Paulo, pauline.provini@mnhn.fr Associate or Co-investigator: Elizabeth Höfling / Universidade de São Paulo, ehofling@ib.usp.br Date of data collection (approximate date): 2014-2019 Geographic location of data collection: Universidade de São Paulo, Brasil / Muséum national d'Histoire Naturelle in Paris, France -------------------------- SHARING/ACCESS INFORMATION -------------------------- Licenses/restrictions placed on the data, or limitations of reuse: public domain, without restriction on use Recommended citation for the data: Provini, Pauline; Höfling, Elizabeth (2020), To hop or not to hop? the answer is in the bird trees, v3, Dryad, Dataset, https://doi.org/10.5061/dryad.qjq2bvqc2 -------------------- DATA & FILE OVERVIEW -------------------- File list: SM1: Categories and Phylogenetic relationships of the 1004 bird species used in the ancestral character state reconstruction (.pdf and .csv files) - SM1a: Categories used for the ancestral character state reconstruction (available inside the SM1.pdf file and as a separate SM1a.csv file) - SM1b: Phylogenetic relationships of the total bird species (from http://birdtree.org/subsets/). The species included in the ancestral character state reconstruction of the present study are highlighted with a grey circle - SM1c: Phylogenetic relationships of the bird species used in this study. The species included in the morphological analysis are highlighted with an arrow SM2: Summary of the statistical tests (.pdf file) - SM2a: Best fitted evolutionary model results - SM2b: Analysis on multiple trees SM3a: Information about the 47 bird species used in the shape analysis (.csv files) - SM3a: Categories of the 47 bird species used in the shape analysis - SM3b: Collection number of the 47 bird species used in the shape analysis SM4: Anatomical description of the 3D landmarks of the three long bones of the hindlimb used in the geometric morphometrics analysis (.pdf file) - SM4a: pelvis - SM4b: femur - SM4c: tibiotarsus - SM4d: tarsometatarsus Data derived from other sources: In SM1 and SM3, locomotion categories are derived from data in the Internet Bird Collection (IBC: https://www.hbw.com/ibc), associated with the Handbook of the Birds of the World Alive (Rising et al. 2019). In SM1 and SM3, habitats categories are derived from EltonTraits 1.0 database (Wilman et al. 2014). In SM1 the phylogenetic trees are derived from the BirdTree online tool (http://birdtree.org/) In SM3, shape analysis is performed on specimens from the Coleção de Aves of the Departamento de Zoologia, Instituto de Biociências, of the University of São Paulo, Brazil, the Coleção de Aves of the Museu Nacional, in Rio de Janeiro, Brazil and the Collections ostéologiques d'oiseaux d’Anatomie Comparée of the Muséum national d'Histoire naturelle, in Paris France. -------------------------- METHODOLOGICAL INFORMATION -------------------------- Description of methods used for collection/generation of data: SM1: We chose 1004 species from 40 orders and 161 families, representing 83% of the total number of bird families. We tried to balance our sample as much as possible to obtain a diversity of orders, families, locomotion abilities and habitats, taking into account the constraints linked to the data availability. First, to categorize these species according to their locomotion habits, we watched videos available on the Internet Bird Collection (IBC: https://www.hbw.com/ibc), associated with the Handbook of the Birds of the World Alive (Rising et al. 2019). We distinguished two categories of terrestrial locomotion for each species of birds (SM 1a): species that mostly walk as “walking” and species that mostly hop as “hopping”. We considered species spending most of their time standing, without clear terrestrial locomotion as “hopping” birds. The resting posture of these species corresponds to the perching posture, usually adopted while hopping in branches. Moreover, the primary use of their legs would be to take off or land, which is closer to a hopping gait than to a walking gait. In addition, these birds are sometimes observed turning around on the branch using a flight phase and moving their legs symmetrically, which is similar to a hop. Second, we categorized the chosen species according to their general habitat. Because the foraging function is tightly linked to exploration and thus to locomotion, we grouped species according to their foraging stratum, referenced in the EltonTraits 1.0 database (Wilman et al. 2014). Species spending more than 70% of their time foraging in mid to high levels in trees, high bushes (2m upward), just above tree canopy, or well above vegetation were considered as “arboreal” (SM 1a). We categorized bird species foraging on ground or below 2m in understory in forest, forest edges, bushes or shrubs as “terrestrial”. Since semi-aquatic birds usually also spend time on the ground, we categorized bird species foraging near the water (on, below or around the surface) as “terrestrial”. We also combined similar habitats together, for example if the total percentage of time spent foraging in canopy and in mid to high levels in trees was superior than 50% we considered the species as arboreal. Finally, we grouped bird species according to their body mass. We computed the logarithm of the mass available in the EltonTraits 1.0 database (Wilman et al. 2014). We opted for a discrete categorisation of the body mass trait to remain homogenous with the locomotion and habitat traits. We sorted the result by ascending order and divided this data into three equal groups, defining three categories. Thus, species weighing less than 54.5 g were considered as “small”, species weighing from 55 g to 2980 g, were defined as “medium”, and bird species weighing from 2980 g to 111 kg, were categorized as “large” (SM 1a). We built 100 time-calibrated trees from the BirdTree online tool (http://birdtree.org/), using Ericson method as a backbone for phylogenetic reconstruction (Ericson et al. 2006; Jetz et al. 2012). All of our analyses were performed in R v3.6.1. We used the “phytools” library version 0.6-99 (Revell 2012) to build a rooted consensus tree with branch lengths, using the function “consensus.edges” and the least-squares method. To resolve multichotomies in our phylogenetic tree, we applied the “multi2di” function from the “ape” library version 5.3 (Paradis et al. 2004) (SM 1b, 1c). SM2: We determined which models best fitted the evolution of the three discrete traits we were studying (i.e. size, locomotion and habitat). To test for the best evolutionary model among the equal-rates model (ER), the symmetric model (SYM), and the all-rates-different model (ARD), we used the “fitDiscrete” function from the “Geiger” library version 2.0.6.2 (Harmon et al. 2008). As the best fitted model was the ARD model for all three studied traits (SM 2a), we estimated the ancestral character states by applying maximum likelihood with an ARD model, using the “make.simmap” function from the “phytools” library version 0.6-99 (Revell 2012), on 500 simulations on the consensus tree. We obtained the character change reconstruction for each node on the tree, for locomotion characters, habitat characters as well as for size range. To test the influence of the consensus tree topology on the results, we computed the same analysis on the 100 initial trees used to build the consensus, with 10 simulations per tree. The results were similar to the overall results (SM 2b). SM3: To investigate whether walking or hopping abilities were related to a specific morphology and influenced by a given habitat, we conducted a shape analysis on the lower appendicular skeleton of a sample of bird species. We selected 47 species of birds among the 1004 previously used in the ancestral state reconstruction, to obtain a diversity of orders and families, locomotion habits, habitat, and size (SM 3). We selected 47 species among 22 orders (55% of the total number of families) (SM 1c), available in two museum collections. One specimen of each species was chosen in the Coleção de Aves of the Departamento de Zoologia, Instituto de Biociências, of the University of São Paulo, Brazil, in the Coleção de Aves of the Museu Nacional, in Rio de Janeiro, Brazil and in the Collections ostéologiques d'oiseaux d’Anatomie Comparée of the Muséum national d'Histoire naturelle (SM 3). We verified that the chosen specimens showed no signs of abnormality, with all bones intact and fused epiphyses. The lower appendicular skeletons, including the pelvis, femur, tibiotarsus, and tarsometatarsus, were scanned. For specimens measured in Brazil, the bones were scanned at the micro-CT scan imaging facility of the Instituto de Biociências of the Universidade de São Paulo Brazil, with a SkyScan microtomographer (SkyScan 1176 BrukerMicroCT, Aatselaar, Belgium, 2003). Resolutions were adapted depending on the size of the specimen, ranging from 9 µm (50 Kv and 300 μA) for the smallest specimen to 40 µm (90 Kv and 260 μA) with multiple stacks acquisitions for the largest one. The images were reconstructed using NRecon software (SkyScan 2011, Version 1.6.6.0) and segmented using Avizo software (9.3 version). For specimens borrowed in France, the bones were scanned at the Surfaçus platform in the MNHN, with a Faro edge arm 450 nm, CDRH/IEC classe 2M, FaroBlu TM. Both acquisition campaigns provided the 3D shape of the pelvis and the long bones of one hind limb (femur, tibiotarsus, and tarsometatarsus) for each specimen. People involved with sample collection, processing, analysis and submission: Pauline Provini supervised, collected and processed the data associated with SM1, SM2, SM3. Elizabeth Höfling supervised and collected the data associated with SM1 and SM3. Marcos Raposo and Guilherme R. R. Britto provided access to the collection and the specimens from the Museu Nacional, Rio de Janeiro, Brazil (SM3). Christine Lefèvre provided access to the collection and the specimens from the collections ostéologiques d'Anatomie comparée of the Muséum national d'Histoire naturelle in Paris, France (SM3). Enio Mattos and Phillip Lenktaitis from the Departamento de Zoologia of the Instituto de Biociências, Universidade de São Paulo, Brazil conducted the CT-scan acquisition (SM3). Delphine Brabant performed the surface scans and the platform Sufaçus, numérisation surfacique 3D de la Direction Générale Déléguée aux Collections, MNHN, Paris (SM3).