paper update

paper/paper.md

@@ -37,7 +37,7 @@ Landscape genetics is a statistical framework that parses genetic variation with
 
 # Statement of need
 
-Network approaches, particularly those graph-theoretic in nature, are increasingly being used to capture functional ecological or evolutionary processes (e.g., dispersal, gene flow) within/ among habitat patches [@Peterson2013]. In some cases (e.g., riverscapes) topological patterns are explicitly mirrored by the physical habitat, such that the network structure itself places constraints upon processes such as individual movement [@CampbellGrant2007]. It is no surprise then, that the importance of network properties such as topological complexity are increasingly implicated as driving evolutionary dynamics in dendritic habitats [@Chiu2020, @Thomaz2016].
+Network approaches, particularly those graph-theoretic in nature, are increasingly being used to capture functional ecological or evolutionary processes (e.g., dispersal, gene flow) within/ among habitat patches [@Peterson2013]. In some cases (e.g., riverscapes) topological patterns are explicitly mirrored by the physical habitat, such that the network structure itself places constraints upon processes such as individual movement [@CampbellGrant2007]. It is no surprise then, that the importance of network properties such as topological complexity are increasingly implicated as driving evolutionary dynamics in dendritic habitats [@Chiu2020; @Thomaz2016].
 
 Despite this, quantitative frameworks for modelling the relationships between evolutionary and ecological processes (e.g., through spatio-genetic associations) are predominantly focused on landscapes, and as such often involving mechanistic assumptions which translate poorly to networks. We address this limitation by providing a novel package, autoStreamTree, that facilitates network modeling of genome-scale data. It first computes a graph representation from spatial databases, then analyses individual or population-level genetic data to ‘fit’ distance components at the stream- or reach- level within the spatial network. Doing so within a network context allows the explicit coupling of genetic variation with other network characteristics (e.g., environmental covariates), in turn promoting a downstream statistical process which can be leveraged to understand how those features drive evolutionary processes (e.g., dispersal/ gene flow). We demonstrate the utility of this approach with a case study in a small stream-dwelling fish in western North America.