Spruce Beetle Outbreak Increases Streamflow From Snow‐Dominated Basins in Southwest Colorado, USA

Abstract

Disturbance of mountain forests, which supply critical streamflow to downstream water users, remains an active area of research (NRC, 2008) with substantial implications on large-scale water availability and carbon budgets (Williams et al., 2016(Williams et al., , 2022Zhang et al., 2017). Recent insect outbreaks have caused widespread forest mortality in snow-dominated basins in the Western US (Goeking & Tarboton, 2020;Meddens et al., 2012). Bark beetles, in particular spruce beetle (Dendroctonus rufipennis) and mountain pine beetle (Dendroctonus ponderosae), have affected approximately 58.8 million acres of coniferous forest in the region since 2000 (USDA Forest Service, 2020). Bark beetle population growth over the last several decades has been attributed to increased reproduction rates (Mitton & Ferrenberg, 2012) and warmer air temperatures (Pettit et al., 2020). Bark beetles and associated secondary fungal infections restrict water uptake and kill host trees over the course of a growing season (Frank et al., 2014;Hubbard et al., 2013). Dead host trees drop their needles over one to 3 years, but in contrast to trees affected by logging and stand-replacing fire, they can remain standing for many years (Rhodes et al., 2020).

The Music of Rivers: The Mathematics of Waves Reveals Global Structure and Drivers of Streamflow Regime

“…We speculate that precipitation, temperature, and catchment size may regulate near-universal hydrological processes that are responsible for the highly compressible components of streamflow regime (those captured by the 7 PCA metrics) (Giano, 2021;N. LeRoy Poff et al, 1997), and that the more complicated ecohydrological interactions introduced by the myriad possible land use regimes and geological factors are responsible for streamflow properties that were harder to identify with PCA analysis (Bladon et al, 2014;Manning et al, 2022;Tague & Grant, 2004;Wu et al, 2021)-though we note that land use is likely correlated with climatic and geological factors. Said differently, our results imply that a simple, emergent physics may exist at the catchment scale, where a handful of mean catchment properties accurately predict flashiness, timing, and volume of flow at the basin's outlet, to the extent that biological interactions remain simple (Sposito, 2017;Zhou et al, 2015).…”

“…We further suggest that as biogeochemical datasets increase in temporal and spatial scales, data-driven descriptors based on low-dimensional structure are key for several reasons: 1) they provide sanity checks of intuitive notions or concepts common in sub-disciplines that are otherwise quantitatively unfalsifiable (Kipper, 2021). For example, in hydrology the notions of "semi-arid watersheds" and "snow-driven watersheds" are commonly employed in the literature (Arheimer et al, 2017;Cosh et al, 2008;Manning et al, 2022;Poon & Kinoshita, 2018). We suggest that these intuitive concepts represent informal, expert-driven versions of dimension reduction (Bates, 2020;Wolff, 2019).…”

The Music of Rivers: The Mathematics of Waves Reveals Global Structure and Drivers of Streamflow Regime

“…And while individual land use characteristics were not particularly important in isolation, machine learning models consistently ranked this group to be as important as catchment area for predicting variability in flow at all timescales longer than a few months (Figure 10). We speculate that precipitation, temperature, and catchment size may regulate near‐universal hydrological processes that are responsible for the highly compressible components of streamflow regime (those captured by the seven PCA metrics; Giano, 2021; Poff et al., 1997), and that the more complicated ecohydrological interactions introduced by the myriad possible land use regimes and geological factors are responsible for the streamflow properties that were harder to identify with PCA analysis (Bladon et al., 2014; Manning et al., 2022; Tague & Grant, 2004; S. Wu et al., 2021), corroborating the need for a large number of non‐generalizable parameters that frequently occurs in modeling scenarios (Horton et al., 2022). Said differently, our results imply that a simple, emergent physics may exist at the catchment scale, where a handful of mean catchment properties accurately predict flashiness, timing, and volume of flow at the basin's outlet, to the extent that biological interactions remain simple (Sposito, 2017; Zhou et al., 2015).…”

Spatio-temporal Snow Variability in a Sub-Alpine Forest predicted by Machine Learning and UAV-based LiDAR Snow Depth Maps

“…This sub-alpine altitudinal band accounts for 25% of the total area of the European Alps and is thus more widespread in terms of area than the alpine (>2000m altitude) altitudinal band (15%). It is in these sub-alpine environments, that forest and water management strategies are needed to counteract the mentioned climate change impacts and to preserve a functioning eco-hydrologic system (Barnhart et al, 2016;Manning et al, 2022;Niittynen et al, 2018). Snow cover and its spatio-temporal variability are controlled by vegetation, topography and meteorology (e.g.…”