Detecting inundation thresholds for dryland wetland vulnerability

Abstract

Dryland wetlands receiving periodical floods are of key importance for ecological resilience. The inundation regime (i.e., frequency, duration, depth, and timing of inundation), is one of the major factors that determine the suitability of local conditions for specific wetland species. During droughts, inundation conditions can reach a threshold after which wetland vegetation could transition to dryland vegetation. This study analyses the response of vegetation to hydrologic variability in an arid wetland in Australia over a period of 22 years (including the Millennium drought) in order to identify inundation thresholds for transitions. We use numerical modelling, field observations and remote sensing data to relate continuous detailed simulations of the inundation regime with the response of patches of Common reed, Water couch and River red gum, three key vegetation associations in the study site. We focus in patches that were affected by the drought and presented dryland vegetation invasion as well as reference patches that remained healthy throughout the drought. On each patch, we compare annual and inter-annual simulated inundation regimes to the minimum inundation conditions that can support the specific vegetation, and we compute the percentage area of the patch that verifies minimum inundation for each year. We define this area percentage as minimum inundation index. This index is analysed in conjunction with Landsat derived information on green vegetation coverage (green Seasonal Fractional Cover) for the selected patches. We found that the minimum inundation index and inter annual frequency are able to describe the vegetation dynamics of the patches, which can be characterised by two distinct response modes that depend on a threshold value of the minimum inundation index. Inundation below the threshold noticeably leads to degraded vegetation, but the vegetation can recover quickly if this threshold is later maintained for one or two years. Values below the threshold for more extended periods (drought) result in a gradual decrease of wetland vegetation to almost complete disappearance after four years and subsequent dryland vegetation invasion.

Resilience to drought of dryland wetlands threatened by climate change

“…Healthy non-woody wetland associations occupy the understory of good and intermediate condition woody wetland patches, while the terrestrial association occupies the understory of poor condition woody patches. We use detailed information on a small number of vegetation patches to determine transition conditions 33 and then apply those results to the rest of the wetland to simulate the extent of the vegetation over the period of analysis (see "Methods") as a function of changing hydrological conditions.…”

“…We use the Minimum Inundation Index (MII) 33 as a descriptor of the inundation experienced by a vegetation patch in a year. The MII is defined as the percentage area of the patch with annual inundation conditions that meet minimum water requirements of the specific vegetation in terms of water depth and flood duration.…”

“…River Red Gum can survive anywhere from one to ten consecutive years without minimum inundation 65 , however, health conditions will gradually deteriorate. Based on the literature 33 we assumed that more than two years below the MII threshold will result in transition, from good to intermediate conditions and more than 6 years below the threshold will trigger transition to poor conditions (Supplementary Table S1). The surveyed vegetation distributions 32 can be used to define patches of each vegetation, but cannot be used to detect transitions, due to their low temporal resolution.…”

“…This database provides seasonal averaged fractions of green, non-green and bare soil cover with a resolution of 30 × 30 m, from which we select autumn green SFC values as representative of vegetation health at the end of the growing season. The selection of the green SFC allows for an assessment of the decay of the vegetation due to consecutive years with inundation values below the MII thresholds 33 . By analyzing patches that remained healthy during the drought and patches that transitioned according to the 1991 and 2008 surveys, SFC analysis allowed to pinpoint the transitions within that period, which included transitions of non-woody wetland vegetation to terrestrial vegetation and different levels of deterioration of woody vegetation 32 .…”

Accretion, retreat and transgression of coastal wetlands experiencing sea-level rise

“…Here, we investiga te how accretion and migration processes affect wetland response to SLR using a computational framework that includes all relevant hydrodynamic and sediment transport mechanisms that affect vegetation and landscape dynamics, yet it is efficient enough to allow the simulation of long time periods. The framework consists of a fast-performance quasi-2D hydrodynamic model (Riccardi, 2000;Rodriguez et al, 2017) that we have extensively tested in wetlands (Rodriguez et al, 70 2017;Saco et al, 2019;Sandi et al, 2018;Sandi et al, 2019;Sandi et al, 2020a;Sandi et al, 2020b) and a sediment advection transport model (Garcia et al, 2015) that we couple with vegetation formulations for preference to tidal conditions to obtain rea listic predictions of wetland accretion and migration under SLR. Our framework incorporates two vegetation species, mangrove and saltmarsh, and accounts for the effects of manmade features like inner channels, embankments and flow constrictions due to culverts.…”