Search USGS⌕ Search

USGS · 70030846

Interactions across spatial scales among forest dieback, fire, and erosion in northern New Mexico landscapes

Abstract

Ecosystem patterns and disturbance processes at one spatial scale often interact with processes at another scale, and the result of such cross-scale interactions can be nonlinear dynamics with thresholds. Examples of cross-scale pattern-process relationships and interactions among forest dieback, fire, and erosion are illustrated from northern New Mexico (USA) landscapes, where long-term studies have recently documented all of these disturbance processes. For example, environmental stress, operating on individual trees, can cause tree death that is amplified by insect mortality agents to propagate to patch and then landscape or even regional-scale forest dieback. Severe drought and unusual warmth in the southwestern USA since the late 1990s apparently exceeded species-specific physiological thresholds for multiple tree species, resulting in substantial vegetation mortality across millions of hectares of woodlands and forests in recent years. Predictions of forest dieback across spatial scales are constrained by uncertainties associated with: limited knowledge of species-specific physiological thresholds; individual and site-specific variation in these mortality thresholds; and positive feedback loops between rapidly-responding insect herbivore populations and their stressed plant hosts, sometimes resulting in nonlinear “pest” outbreak dynamics. Fire behavior also exhibits nonlinearities across spatial scales, illustrated by changes in historic fire regimes where patch-scale grazing disturbance led to regional-scale collapse of surface fire activity and subsequent recent increases in the scale of extreme fire events in New Mexico. Vegetation dieback interacts with fire activity by modifying fuel amounts and configurations at multiple spatial scales. Runoff and erosion processes are also subject to scale-dependent threshold behaviors, exemplified by ecohydrological work in semiarid New Mexico watersheds showing how declines in ground surface cover lead to non-linear increases in bare patch connectivity and thereby accelerated runoff and erosion at hillslope and watershed scales. Vegetation dieback, grazing, and fire can change land surface properties and cross-scale hydrologic connectivities, directly altering ecohydrological patterns of runoff and erosion. The interactions among disturbance processes across spatial scales can be key drivers in ecosystem dynamics, as illustrated by these studies of recent landscape changes in northern New Mexico. To better anticipate and mitigate accelerating human impacts to the planetary ecosystem at all spatial scales, improvements are needed in our conceptual and quantitative understanding of cross-scale interactions among disturbance processes.

Explore related subjects

Keep this discovery

Explore connections, maps & timelines

BibTeXRIS

Craig D. Allen. 2007-06-20. Interactions across spatial scales among forest dieback, fire, and erosion in northern New Mexico landscapes. https://doi.org/10.1007/s10021-007-9057-4

Cite the original work for its findings. Save a collection to share your selection of sources.

KEEP EXPLORING

Related USGS reports

Loss of ecosystem resistance to invasion after wildfire facilitated by a pulse of soil nitrogen availability: Experimental evidence in the burned sagebrush steppe

The impacts of climate shifts on ecosystems dominated by long-lived perennials are likely most pronounced during community reassembly after disturbances such as fire and confounded by interactions between disturbances and plant community composition. A 30-year-experimental hydroclimate manipulation of multiple sagebrush steppe communities was completely consumed by a 2019 wildfire, providing an opportunity to evaluate effects of precipitation deficit on ecosystem recovery. Ambient precipitation was doubled for 23 years via irrigation in winter or summer in grassland or shrub–steppe communities until 2016. Plots that had received irrigation thus experienced drought for three years preceding and continuing after the fire. These landscapes are vulnerable to invasion by exotic annuals such as Bromus tectorum L. (cheatgrass) that promote wildfire occurrence, which favors even greater invasion levels. Thus, we asked whether patterns of invasion after the compound disturbance of drought and fire related to the long-term pre-fire climate and plant community structure. Established theory led to the prediction that plant communities developed under wetter climates would have greater resistance to invasion. The most resistant plots were the most arid (that is, never irrigated control plots with no drought) which had the least pre-fire canopy cover of shrubs and nitrogen-fixing forbs and greater proportional cover of perennial bunchgrasses. Plots that developed under winter irrigation had greater cover of shrubs and N-fixing forbs, corresponding to pulses of plant available soil nitrogen that were 8.2-fold greater than pre-fire levels, compared to a 0.20-fold post-fire reduction in soil nitrogen observed in the ambient plots. Nitrogen pulses and invasion were most evident in the inter-canopy bare-soil patches (‘interspaces’) and were least evident where perennial grasses were most abundant. Long-term hydroclimate altered pre-fire plant community composition in ways that affected post-fire resistance to invasion such that the combined effects of fire and water deficit led to greater than expected invasion in wetter regions that are conventionally considered resistant to invasion and resilient to wildfire.

Ecosystems↗

Identifying headwater streams across the conterminous United States

Headwater streams play critical roles in hydrologic and biogeochemical processes and functions, yet their spatial distribution and land cover context remain poorly understood at continental scales, and no dedicated geospatial dataset exists. Building from a high-resolution conterminous United States (CONUS) hydrography network dataset, we quantified the spatial extent, density, and upstream catchment characteristics of headwater stream segments across the CONUS. We identified approximately 8.4 million kilometers of headwater streams, finding that 77% of the total stream network consists of headwaters, nearly double the total length represented in prior estimates. Stream density varied fivefold across regions, from < 1 km·km −2 in arid basins to > 5 km·km −2 in humid, forested areas. Over 73% of the CONUS landmass drains from headwater streams. The majority of headwater stream length occurred in forested and cultivated catchments across the CONUS, while substantial regional differences were evident for headwater stream distribution in other land cover classes (for example, wetlands, urban areas, shrublands, and herbaceous-dominated catchments). The dedicated and novel geospatial dataset, HELiOS (HEadwater streams and Low-Order Systems) is introduced for management and research use. The HELiOS dataset provides the first continental-scale, high-resolution characterization of headwater streams, offering new insights and opportunities for hydrologic modeling, ecological assessments, and environmental policy.

conterminous United States↗

Variation in soil organic carbon across a latitudinal chronosequence of mangrove poleward expansion

The critical carbon sink provided by coastal wetlands, known as blue carbon, can be affected by multiple aspects of climate change. One important example is warming-induced mangrove poleward expansion, which is shifting dominant plant cover across tropical–temperate transitional zones and altering ecosystem structure and function. We examined how mangrove expansion affects soil organic carbon (SOC) quantity and source, using measurements of SOC density and isotopic signatures (δ 13 C and δ 15 N) at 15 sites across Florida’s west coast (USA). The sampled sites represent examples of three expansion stages: a latitudinal chronosequence of mangrove expansion, spanning mature mangroves in the south, former ecotones at mid latitudes, and current ecotones in the north. Our analyses of soil core data indicate that mangrove expansion stage is a significant predictor of SOC density, δ 13 C, and δ 15 N, but not C:N ratio. Current ecotones exhibited significantly lower SOC density but higher δ 13 C, suggesting a greater contribution of preexisting C 4 salt marshes, while no difference was found between former ecotones and mature mangroves. SOC density, δ 13 C, and δ 15 N were found to vary with mangrove aboveground biomass, stage, and sedimentary setting along the latitudinal gradient. For all three mangrove expansion stages, SOC density decreased with depth, but δ 13 C showed no vertical trend, suggesting that mangroves contributed organic carbon to the entire 20-cm soil profile. The observed regional trend of SOC across mangrove expansion stages highlights the ecological impacts of warming-driven vegetation shifts in coastal wetlands, though further evidence is needed to determine the primary drivers and mechanisms, while also considering local and regional environmental factors.

Florida↗