Update index.md

Author: creeas

methods/index.md

@@ -32,9 +32,13 @@ We used treePL 1.0 in Ubuntu 14.04[27], which implements a flexible rate-smoothi
 We added taxa without DNA sequence data to each of the 500 treePL-dated, molecular trees (“stage 1” trees) and then used a taxon-addition and polytomy-resolver algorithm (modified from Kuhn et al. 2011; details below) to generate a large distribution of fully resolved, taxon-complete candidate phylogenies. We used two taxonomic sources: the [Chondrichthyan Tree of Life](http://sharksrays.org) and the [International Union for the Conservation of Nature Red List](http://www.redlist.org). Our distribution of taxon-complete trees includes three types of species[31]. Type 1 species have genetic data and are represented in the stage 1 trees. Type 2 species have no genetic data (or were identified as rogue taxa), but have at least one congener represented in the stage 1 trees. Type 3 species have no genetic data (or were identified as rogue taxa), and have no congeners in the stage 1 trees. Type 2 and Type 3 species were allowed to populate particular clades using taxonomic information and the topology (via node identities) of the stage 1 trees. We outline the rules we used to include taxa without any sequence data below.
 
 Type 1 species were anchored relative to one another as resolved in the stage 1 trees, and we used 70% bootstrap-support (BS) as a threshold to topologically constrain inferred nodes during taxon addition. There were four scenarios in which the stage 1 trees needed to be modified to enforce genus, family, or order monophyly (Supplementary Table 4). 
+
 •	First, there were nine genera with relatively weak support (BS < 70%; range: 7-68%) for genus monophyly within a highly supported clade (BS ≥ 70%; Supplementary Table 4). For these nine genera instances we pruned ten type 1 taxa from the stage 1 trees, reducing its size from 620 to 610 sp. We reincorporated these ten pruned taxa subsequently as type 2 species by constraining them to their named clades. 
+
 •	Second, there were four families (Anacanthobatidae, Hemigaliidae, Somniosidae, and Triakidae) where there was weak support (BS < 70%; range: 33-62%) for family non-monophyly. In these four instances we collapsed the weakly supported nodes and subsequently reconstituted clades to enforce family monophyly[31] (Supplementary Table 4). 
+
 •	Third, there were 24 mixed-genus and/or mixed-family clades with strong evidence (BS ≥ 70%) against monophyly for at least one genus or family. These “mixed clades” took on a variety of forms, from simple paraphyly to complex interdigitation of sub-genera or sub-families. We enforced genus monophyly for any genus or family within these mixed clades, unless there was strong evidence (BS ≥ 70%) against monophyly. 
+
 •	Finally, for consistency between taxa with and without genetic data, we assumed that all genera, families, and orders were monophyletic unless there was strong evidence (BS ≥ 70%) against monophyly in the stage 1 trees. This rule affected one subgenus (Galeus minor clade), nine genera (Atlantoraja, Centrophorus, Chiloscyllium, Halaelurus, Mobula, Rajella, Rhinobatos, Sphyrna and Squalus), one mixed-genus clade (Dentiraja, Dipturus, Spiniraja together with Zearaja), one family (Pristidae) and one order (Orectolobiformes) in our stage 1 trees – each of these had only weak evidence (BS < 70%; range 32-69%) for monophyly. 
 
 After using these rules to modify the stage 1 trees, we imposed topological constraints on the placement of the remaining type 2 and type 3 species, including the 21 rogue taxa. Each type 2 species was restricted to its genus or its mixed-genus clade. There were four genera (Dasyatis, Galeus, Himantura, and Triakis) with strong evidence (BS ≥ 70%) against monophyly in the stage 1 trees that also required the addition of type 2 species. For these four genera, type 2 congeners were added to the largest candidate sub-clade for the genus. Each type 3 species was restricted to its named genus, and the entire genus was constrained in its placement among other genera according to higher-level (supergenus, family or order) taxonomic information (Supplementary Table 1). Twenty-three of the 198 recognized genera were not represented in the stage 1 trees, and these 23 genera were restricted to 6 of the 14 orders and 18 of the 60 families (Supplementary Table 1). There were 9 type 3 species not assigned to a family on http://sharksrays.org/. In these instances, we referred to the IUCN http://iucn.org/ for the original family-level designation (Supplementary Table 1).