Search

Molecular phylogenetic studies that have included sirenians from the genera Trichechus, Dugong, and Hydrodamalis have resolved their interrelationships but have yielded divergence age estimates that are problematically discordant. The ages of these lineage splits have profound implications for how to interpret the sirenian fossil record--including clade membership, biogeographic patterns, and correlations with Earth history events. In an effort to address these issues, here we present a total evidence phylogenetic analysis of Sirenia that includes living and fossil species and applies Bayesian tip-dating methods to estimate their interrelationships and divergence times. In addition to extant sirenians, our dataset includes 56 fossil species from 106 dated localities and numerous afrotherian outgroup taxa. Genetic, morphological, temporal, and biogeographic data are assessed simultaneously to bring all available evidence to bear on sirenian phylogeny. The resulting time-tree is then used for Bayesian geocoordinates reconstruction analysis, which models ancestral geographic areas at splits throughout the phylogeny, thereby allowing us to infer the direction and timing of dispersals. Our results suggest that Pan-Sirenia arose in North Africa during the latest Paleocene and that the Eocene evolution of stem sirenians was primarily situated in the Tethyan realm. In the late Eocene, some lineages moved into more northern European latitudes, an area that became the source region for a key trans-Atlantic dispersal towards the Caribbean and northern-adjacent west Atlantic. This event led to the phylogenetic and biogeographic founding of crown Sirenia with the Dugongidae-Trichechidae split occurring at the Eocene-Oligocene boundary (~33.9 Ma), temporally coincident with the onset of dropping global sea levels and temperatures. This region became the nexus of sirenian diversification and supported taxonomically-rich dugongid communities until the earliest Pliocene. The Dugonginae-Hydrodamalinae split occurred near Florida during the early Miocene (~21.2 Ma) and was followed by a west-bound dispersal that gave rise to the Pacific hydrodamalines. The late middle Miocene (~12.2 Ma) split of Dugong from all other dugongines also occurred near Florida and our analyses suggest that the Indo-Pacific distribution of modern dugongs is the result of a trans-Pacific dispersal. From at least the early Miocene, trichechid evolution was based entirely in South America, presumably within the Pebas Wetlands System. We infer that the eventual establishment of Amazon drainage into the South Atlantic allowed the dispersal of Trichechus out of South America no earlier than the mid-Pliocene. Our analyses provide a new temporal and biogeographic framework for understanding major events in sirenian evolution and their possible relationships to oceanographic and climatic changes. These hypotheses can be further tested with the recovery and integration of new fossil evidence. [ABSTRACT FROM AUTHOR]

The objectives of this work were to compare reproduction and parturition traits of the Criollo breed Romosinuano as straightbred and crossbred cows with Angus and Brahman, to estimate heterosis and direct and maternal genetic breed effects, and to describe calf loss, cow removals from the project, the occurrence of calving difficulty, inadequate calf vigor at birth, and udder problems by cow breed groups. Cows (n = 404) were bom from 2002 to 2005. After their first exposure to bulls as young cows, in all subsequent breeding seasons crossbred cows were bred to bulls of the third breed, and straightbred cows were bred to bulls of the other two breeds. Calving records (n = 1,484) from 2005 to 2011 were used to create calving and weaning rate and calving interval (excluding the interval between 2 and 3 yr of age). Final models for these traits included sire breed-dam breed interaction, cow age within year, and random animal effects. Heterosis estimates for Romosinuano-Brahman calving and weaning rate were 0.06 ± 0.02 and 0.07 ± 0.03 (P < 0.05); those for Brahman-Angus were twice as large (0.13 ± 0.03 and 0.14 ± 0.03, respectively; P < 0.001). Estimates of Brahman direct effects on calving and weaning rate were -0.12 ± 0.04 and -0.14 ± 0.05 (P < 0.05); however, Angus direct effects were beneficial for both traits (0.1 ± 0.05, P < 0.05). The effect of heterosis was to reduce calving interval by -49.2 ± 9.9 and -37.2 ± 9.7 d for Romosinuano-Brahman and Brahman-Angus, respectively (P < 0.001). Romosinuano and Fj cows sired by Romosinuano and out of Angus dams had the most occurrences of difficult births as a proportion of cows that calved (0.028 and 0.025, P = 0.04). Angus-sired crossbred cows and Brahman cows had the most occurrences of udder problems as a proportion of lac-tating cows (0.14 to 0.21, P < 0.04). There were more Brahman-sired cows that died or were culled as a proportion of those cows that began the project (0.1 to 0.28, P < 0.02) than cows in the other breed groups. Romosinuano-Brahman and Romosinuano-Angus cow performance was acceptable, but for most traits, those pairs of breeds had lower heterosis than Brahman-Angus. [ABSTRACT FROM AUTHOR]