Eco-hydrological modelling of channel network dynamics-part 1: stochastic simulation of active stream expansion and retraction.

Dynamic changes in the active portion of stream networks represent a phenomenon common to diverse climates and geologic settings. However, mechanistically describing these processes at the relevant spatiotemporal scales without huge computational burdens remains challenging. Here, we present a novel stochastic framework for the effective simulation of channel network dynamics capitalizing on the concept of 'hierarchical structuring of temporary streams'-a general principle to identify the activation/deactivation order of network nodes. The framework allows the long-term description of event-based changes of the river network configuration starting from widely available climatic data (mainly rainfall and evapotranspiration). Our results indicate that climate strongly controls temporal variations of the active length, influencing not only the preferential configuration of the active channels but also the speed of network retraction during drying. Moreover, we observed that-while the statistics of wet length are mainly dictated by the underlying climatic conditions-the spatial patterns of active reaches and the size of the largest connected patch of the network are strongly controlled by the spatial correlation of local persistency. The proposed framework provides a robust mathematical set-up for analysing the multi-faceted ecological legacies of channel network dynamics, as discussed in a companion paper.

Eco-hydrological modelling of channel network dynamics-part 2: application to metapopulation dynamics.

Temporal variations in the configuration of the flowing portion of stream networks are observed in the large majority of rivers worldwide. However, the ecological implications of river network expansions/retractions remain poorly understood, owing to the lack of computationally efficient modelling tools conceived for the long-term simulation of river network dynamics. Here, we couple a stochastic approach for the simulation of channel network expansion and retraction (described in a companion paper) with a dynamic version of a stochastic occupancy metapopulation model. The coupled eco-hydrological model is used to analyse the impact of pulsing river networks on species persistence under different hydroclimatic scenarios. Our results unveil the existence of a climate-dependent detrimental effect of network dynamics on species spread and persistence. This effect is enhanced by dry climates, where flashy expansions and retractions of the flowing channels induce metapopulation extinction. Survival probabilities are particularly reduced in settings where the spatial heterogeneity of network connectivity is pronounced. The analysis indicates that accounting for the temporal variability of the flowing river network and its connectivity is a fundamental prerequisite for analysing in-stream metapopulation dynamics.

Emergent dispersal networks in dynamic wetlandscapes.

The connectivity among distributed wetlands is critical for aquatic habitat integrity and to maintain metapopulation biodiversity. Here, we investigated the spatiotemporal fluctuations of wetlandscape connectivity driven by stochastic hydroclimatic forcing, conceptualizing wetlands as dynamic habitat nodes in dispersal networks. We hypothesized that spatiotemporal hydrologic variability influences the heterogeneity in wetland attributes (e.g., size and shape distributions) and wetland spatial organization (e.g., gap distances), in turn altering the variance of the dispersal network topology and the patterns of ecological connectivity. We tested our hypotheses by employing a DEM-based, depth-censoring approach to assess the eco-hydrological dynamics in a synthetically generated landscape and three representative wetlandscapes in the United States. Network topology was examined for two end-member connectivity measures: centroid-to-centroid (C2C), and perimeter-to-perimeter (P2P), representing the full range of within-patch habitat preferences. Exponentially tempered Pareto node-degree distributions well described the observed structural connectivity of both types of networks. High wetland clustering and attribute heterogeneity exacerbated the differences between C2C and P2P networks, with Pareto node-degree distributions emerging only for a limited range of P2P configuration. Wetlandscape network topology and dispersal strategies condition species survival and biodiversity.

Stochastic dynamics of wetlandscapes: Ecohydrological implications of shifts in hydro-climatic forcing and landscape configuration.

Wetlands are embedded in landscapes in fractal spatial patterns, and are characterized by highly dynamic, interlinked hydrological, biogeochemical, and ecological functions. We propose here a stochastic approach to evaluate and predict the spatiotemporal hydrologic variability of wetlands at landscape scale (100 km2). Stochastic hydro-climatic forcing (daily rainfall and evapotranspiration) and the landscape topographic setting (spatial structure of wetlands within the landscape) are key drivers of wetland eco-hydrologic functionality. The novelty of our approach lies in the quantification of the hydrological dynamics for all wetlands distributed in a given landscape, and in linking stochasticity of hydroclimatic forcing and ecologically meaningful wetland network metrics. We applied the modeling framework to investigate daily hydrologic dynamics in six landscapes across the U.S. that span gradients of hydroclimate and abundance of wetlands. We assess landscape-scale patterns using four key wetland hydrological attributes that have significance in terms of aquatic habitat suitability and dispersal: (1) Abundance (2) Diversity (3) Persistence, and (4) Accessibility. We observe that the hydrologic responses of each of the six landscapes are driven by the interactions between regional stochastic hydro-climatic forcing and landscape topographic setting. Despite differences in these features, similar scaling relations define diversity (area distributions) and accessibility (separation-distance distributions). Persistence of hydrologic regimes, defined by duration of inundation above thresholds, was least in more-arid settings, and higher in humid settings, consistent with intuitive understanding. These results can support assessments of the spatiotemporal variability of ecohydrological attributes in diverse wetlandscapes, including aquatic species dispersal and habitat suitability for unique flora and fauna.