Assessing variations in water availability to vegetation and its consequences on the riparian forest of the arid southwestern USA in service of ecosystem conservation

As Earth’s climate changes, a solid understanding of ecosystems’ sensitivity and reactivity to climatic and environmental controls is critical. Vegetation is considered a key structural element of habitats and ecosystems. By studying changes in vegetation communities’ (specific species assemblages) distribution, health, and timing of main life events, compared with potential controls, such as water availability, it is possible to infer important information on the interactions between vegetation communities (an indicator of habitat integrity) and a changing climate. This comprehension of vegetation dynamics is also crucial for the conservation of species vulnerable to changes and their habitat, even more so in the case of species that are already considered endangered. To fulfil their conservation mission, natural resources managers of lands, regardless of ownership, need a regional-scale understanding of climatic and environmental controls on habitat distribution, condition and vulnerability to climate change. They also need to be able to monitor habitat condition and distribution efficiently and accurately with limited resources. This thesis examines the case of riparian habitats in drylands. Intermittent streams and their streamside vegetation act as moist and cool refuges, support a high species richness and buffer sensitive populations from drought. But they are also rare, isolated, and highly sensitive to changes in water availability. My goal is to reconstruct the complex links between water distribution, water availability to vegetation, and vegetation distribution across a diverse landscape covering a range of elevation, topography and geology, supporting diverse ecosystems. To achieve this landscape-scale overview, I use remote sensing to map the main vegetation communities distribution, their long-term evolution, health, and sensitivity to drought, and conductivity sensors to detect flow and open water presence in-non perennial streams. These methods provide high spatiotemporal resolution datasets that can cover wide areas. The resulting information on vegetation and flow is then compared to climatic and environmental factors, including local geology and rainfall, to better understand their potential vulnerability to drought. This work is focused on applied research and is set in a wider context of providing knowledge and tools for natural resources managers to keep track of rare and sensitive habitats’ condition and extent. The methods and tools used were chosen to test their suitability as accessible and efficient monitoring tools. This thesis focuses on an area in the Southwest USA, covering the upper basin of the San Pedro River (an intermittent river of the Colorado basin), the Huachuca Mountains (an isolated mountain range) and the semi-arid valley in between. My research was conducted on military lands and, as such, the natural resources management side of this work is seen through the lens of military installations and their specific approach to ecosystem conservation. However, the findings of my thesis, both on the ecosystem functioning and the management implications sides, can be more widely applied to monitoring of isolated ecosystems in drylands. This work provides an overall understanding of water availability to vegetation across a diverse landscape, and how this availability controls vegetation distribution and health, from rainfall-supported grassland to groundwater-fed dense riparian forests. I study the differences in long-term, interannual vegetation density variations between ephemeral and perennial reaches along a dryland river, showing how local geology and groundwater levels can buffer riparian ecosystems against drought. I also describe the spatiotemporal distribution of flow in mountain ephemeral streams, from perennial springs to rainfall-fed washes. I link flow permanence to underlying geology, consider how different reaches might be impacted by prolonged drought, and the consequences on local flora and fauna. Finally, I look at short-term, intra-annual changes in vegetation phenology along mountain streams and show how droughts, with higher temperatures and lower precipitation, might shorten the growing season and negatively impact vegetation leaf density. Throughout the thesis, I relate my findings to natural resources management questions and needs, with the goal of providing tools and conclusions useful for endangered and critical habitat monitoring