Indigenous fire management and cross-scale fire-climate relationships in the Southwest United States from 1500 to 1900 CE.

Prior research suggests that Indigenous fire management buffers climate influences on wildfires, but it is unclear whether these benefits accrue across geographic scales. We use a network of 4824 fire-scarred trees in Southwest United States dry forests to analyze up to 400 years of fire-climate relationships at local, landscape, and regional scales for traditional territories of three different Indigenous cultures. Comparison of fire-year and prior climate conditions for periods of intensive cultural use and less-intensive use indicates that Indigenous fire management weakened fire-climate relationships at local and landscape scales. This effect did not scale up across the entire region because land use was spatially and temporally heterogeneous at that scale. Restoring or emulating Indigenous fire practices could buffer climate impacts at local scales but would need to be repeatedly implemented at broad scales for broader regional benefits.

Forest restoration treatments increased growth and did not change survival of ponderosa pines in severe drought, Arizona.

We report on survival and growth of ponderosa pines (Pinus ponderosa Douglas ex P. Lawson & C. Lawson) 2 decades after forest restoration treatments in the G. A. Pearson Natural Area, northern Arizona. Despite protection from harvest that conserved old trees, a dense forest susceptible to uncharacteristically severe disturbance had developed during more than a century of exclusion of the previous frequent surface-fire regime that ceased upon Euro-American settlement in approximately 1876. Trees were thinned in 1993 to emulate prefire-exclusion forest conditions, accumulated forest floor was removed, and surface fire was re-introduced at 4-years intervals (full restoration). There was also a partial restoration treatment consisting of thinning alone. Compared with untreated controls, mortality of old trees (mean age 243 years, maximum 462 years) differed by <1 tree ha-1 and old-tree survival was statistically indistinguishable between treatments (90.5% control, 92.3% full, 82.6% partial). Post-treatment growth as measured by basal area increment of both old (pre-1876) and young (post-1876) pines was significantly higher in both treatments than counterpart control trees for more than 2 decades following thinning. Drought meeting the definition of megadrought affected the region almost all the time since the onset of the experiment, including 3 years that were severely dry. Growth of all trees declined in the driest 3 years, but old and young treated trees had significantly less decline. Association of tree growth with temperature (negative correlation) and precipitation (positive correlation) was much weaker in treated trees, indicating that they may experience less growth decline from warmer, drier conditions predicted in future decades. Overall, tree responses after the first 2 decades following treatment suggest that forest restoration treatments have led to substantial, sustained improvement in the growth of old and young ponderosa pines without affecting old-tree survival, thereby improving resilience to a warming climate.

Trends of forest and ecosystem services changes in the Mescalero Apache Tribal Lands.

Forests are critically important for the provision of ecosystem services. The Sacramento Mountains of New Mexico, USA, are a hotspot for conservation management and the Mescalero Apache Tribe's homeland. The multiple ecosystem services and functions and its high vulnerability to changes in climate conditions make their forests of ecological, cultural, and social importance. We used data from the Mescalero Apache Tribal Lands (MATL) Continuous Forest Inventory over 30 yr to analyze changes in the structure and composition of ecosystems as well as trends in ecosystem services. Many provisioning, regulating, cultural, and supporting services were shared among the MATL ecosystems and were tied to foundational species dominance, which could serve as a reliable indicator of ecosystem functioning. Our analysis indicates that the MATL are in an ongoing transition from conifer forests to woodlands with declines in two foundation species, quaking aspen and ponderosa pine, linked to past forest management and changing climate. In addition, we detected a decrease in species richness and tree size variability, amplifying the risk of forest loss in a rapid climatic change. Continuous permanent plots located on a dense grid (1 × 1 km) such as the ones monitored by the Bureau of Indian Affairs are the most detailed data available to estimate forests multiresource transitions over time. Native lands across the USA could serve as the leading edge of detecting decadal-scale forest changes and tracking climate impacts.

Wildfire-Driven Forest Conversion in Western North American Landscapes.

Changing disturbance regimes and climate can overcome forest ecosystem resilience. Following high-severity fire, forest recovery may be compromised by lack of tree seed sources, warmer and drier postfire climate, or short-interval reburning. A potential outcome of the loss of resilience is the conversion of the prefire forest to a different forest type or nonforest vegetation. Conversion implies major, extensive, and enduring changes in dominant species, life forms, or functions, with impacts on ecosystem services. In the present article, we synthesize a growing body of evidence of fire-driven conversion and our understanding of its causes across western North America. We assess our capacity to predict conversion and highlight important uncertainties. Increasing forest vulnerability to changing fire activity and climate compels shifts in management approaches, and we propose key themes for applied research coproduced by scientists and managers to support decision-making in an era when the prefire forest may not return.

The Fire and Tree Mortality Database, for empirical modeling of individual tree mortality after fire.

Wildland fires have a multitude of ecological effects in forests, woodlands, and savannas across the globe. A major focus of past research has been on tree mortality from fire, as trees provide a vast range of biological services. We assembled a database of individual-tree records from prescribed fires and wildfires in the United States. The Fire and Tree Mortality (FTM) database includes records from 164,293 individual trees with records of fire injury (crown scorch, bole char, etc.), tree diameter, and either mortality or top-kill up to ten years post-fire. Data span 142 species and 62 genera, from 409 fires occurring from 1981-2016. Additional variables such as insect attack are included when available. The FTM database can be used to evaluate individual fire-caused mortality models for pre-fire planning and post-fire decision support, to develop improved models, and to explore general patterns of individual fire-induced tree death. The database can also be used to identify knowledge gaps that could be addressed in future research.

Fire regimes shape diversity and traits of vegetation under different climatic conditions.

Changes in climate and land use are altering fire regimes in many regions across the globe. This work aims to study the influence of wildfire recurrence and burn severity on woody community structure and plant functional traits under different environmental conditions. We selected three study sites along a Mediterranean-Oceanic climatic gradient, where we studied the fire history and burn severity of the last wildfire. Four years after the last wildfire, we established 1776 1-m2 plots where the percentage cover of each woody species was sampled. We calculated (i) structural parameters of the community such as total cover, alpha species richness, evenness and diversity (Shannon diversity index), and (ii) vegetation cover of each functional group (differentiating life forms, eco-physiological traits and regenerative traits). Focusing on community structure, results showed increases in species richness and diversity as wildfire recurrence increased, but this effect was partially counterweighted in the areas affected by high severity. In relation to functional groups, we found that increases in recurrence and severity fostered transition from tree- to shrub-dominated ecosystems. Non-arboreal life form, high specific leaf area, N2-fixing capacity, resprouting ability and heat-stimulated germination were advantageous traits under high recurrences and severities, and low seed mass was advantageous under high recurrence situations. We suggest that the strength of the effects of wildfire recurrence and burn severity on vegetation structure and traits might vary with climate, increasing from Oceanic to Mediterranean conditions. In the Mediterranean site, recurrence and severity were strongly related to traits associated with germination (seed mass and heat-stimulated germination), whereas in the Oceanic site the strongest relationships were found with a resprouting-related trait (bud location). This study identifies changes in vegetation structure and composition in scenarios of high recurrence and severity, and provides useful information on plant traits that could be key in enhancing vegetation resilience.

Diné kinship as a framework for conserving native tree species in climate change.

Climate change affects all ecosystems but despite increasing recognition for the needs to integrate Indigenous knowledge with modern climate science, the epistemological differences between the two make it challenging. In this study, we present how Indigenous belief and knowledge system can frame the application of a modeling tool (Climate-Forest Vegetation Simulator). We focus on managing forest ecosystem services of the Diné (Navajo) Nation as a case study. Most Diné tribal members depend directly on the land for their livelihoods and cultural traditions. The forest plays a vital role in Diné livelihoods through social, cultural, spiritual, subsistence, and economic factors. We simulated forest dynamics over time under alternative climate change scenarios and management strategies to identify forest management strategies that will maintain future ecosystem services. We initialized the Climate-Forest Vegetation Simulator model with data from permanent plots and site-specific growth models under multiple management systems (no-management, thinning, burning, and assisted migration planting) and different climate scenarios (no-climate-change, RCP 4.5, RCP 6.0). Projections of climate change show average losses of basal area by over 65% by 2105, a shift in tree species composition to drier-adapted species, and a decrease in species diversity. While substantial forest loss was inevitable under the warming climate scenarios, the modeling framework allowed us to evaluate the management treatments, including planting, for conserving multiple tree species in mixed conifer forests, thus providing an anchor for biodiversity. We presented the modeling results and management implications and discuss how they can complement Diné kinship concepts. Our approach is a useful step for framing modern science with Indigenous Knowledge and for developing improved strategies to sustain natural resources and livelihoods.

Fire regime on a cultural landscape: Navajo Nation.

Fire has played an important role in the evolutionary environment of global ecosystems, and Indigenous peoples have long managed natural resources in these fire-prone environments. We worked with the Navajo Nation Forestry Department to evaluate the historical role of fire on a 50 km2 landscape bisected by a natural mountain pass. We used fifty 5-ha circular plots to collect proxy fire history data on fire-scarred trees, stumps, logs, and snags in a coniferous forest centered on a key mountain pass. The fire history data were categorized into three groups: All (all 50 plots), Corridor (25 plots closest to Buffalo Pass drainage), and Outer (remaining 25 plots, farther from pass). We assessed spatial and temporal patterns of fire recurrence and fire-climate relationships. The landscape experienced frequent fires from 1644, the earliest fire date with sufficient sample depth, to 1920, after which fire occurrence was interrupted. The mean fire interval (MFI) for fire dates scarring 10% or more of the samples was 6.25 years; there were 13 large-scale fires identified with the 25% filter with an MFI of 22.6 years. Fire regimes varied over the landscape, with an early reduction in fire occurrence after 1829, likely associated with pastoralism, in the outer uplands away from the pass. In contrast, the pass corridor had continuing fire occurrence until the early 20th century.Synthesis. Fires were synchronized with large-scale top-down climatic oscillations (drought and La Niña), but the spatially explicit landscape sampling design allowed us to detect bottom-up factors of topography, livestock grazing, and human movement patterns that interacted in complex ways to influence the fire regime at fine scales. Since the early 20th century, however, fires have been completely excluded. Fuel accumulation in the absence of fire and warming climate present challenges for future management.

Forest restoration as a strategy to mitigate climate impacts on wildfire, vegetation, and water in semiarid forests.

Climate change and wildfire are interacting to drive vegetation change and potentially reduce water quantity and quality in the southwestern United States, Forest restoration is a management approach that could mitigate some of these negative outcomes. However, little information exists on how restoration combined with climate change might influence hydrology across large forest landscapes that incorporate multiple vegetation types and complex fire regimes. We combined spatially explicit vegetation and fire modeling with statistical water and sediment yield models for a large forested landscape (335,000 ha) on the Kaibab Plateau in northern Arizona, USA. Our objective was to assess the impacts of climate change and forest restoration on the future fire regime, forest vegetation, and watershed outputs. Our model results predict that the combination of climate change and high-severity fire will drive forest turnover, biomass declines, and compositional change in future forests. Restoration treatments may reduce the area burned in high-severity fires and reduce conversions from forested to non-forested conditions. Even though mid-elevation forests are the targets of restoration, the treatments are expected to delay the decline of high-elevation spruce-fir, aspen, and mixed conifer forests by reducing the occurrence of high-severity fires that may spread across ecoregions. We estimate that climate-induced vegetation changes will result in annual runoff declines of up to 10%, while restoration reduced or reversed this decline. The hydrologic model suggests that mid-elevation forests, which are the targets of restoration treatments, provide around 80% of runoff in this system and the conservation of mid- to high-elevation forests types provides the greatest benefit in terms of water conservation. We also predict that restoration treatments will conserve water quality by reducing patches of high-severity fire that are associated with high sediment yield. Restoration treatments are a management strategy that may reduce undesirable outcomes for multiple ecosystem services.

Tree growth-climate relationships in a forest-plot network on Mediterranean mountains.

In this study we analysed a novel tree-growth dataset, inferred from annual ring-width measurements, of 7 forest tree species from 12 mountain regions in Greece, in order to identify tree growth - climate relationships. The tree species of interest were: Abies cephalonica, Abies borisii-regis, Picea abies, Pinus nigra, Pinus sylvestris, Fagus sylvatica and Quercus frainetto growing across a gradient of climate conditions with mean annual temperature ranging from 5.7 to 12.6°C and total annual precipitation from 500 to 950mm. In total, 344 tree cores (one per tree) were analysed across a network of 20 study sites. We found that water availability during the summer period (May-August) was a strong predictor of interannual variation in tree growth for all study species. Across species and sites, annual tree growth was positively related to summer season precipitation (PSP). The responsiveness of annual growth to PSP was tightly related to species and site specific measurements of instantaneous photosynthetic water use efficiency (WUE), suggesting that the growth of species with efficient water use is more responsive to variations in precipitation during the dry months of the year. Our findings support the importance of water availability for the growth of mountainous Mediterranean tree species and highlight that future reductions in precipitation are likely to lead to reduced tree-growth under climate change conditions.

Slow carbon and nutrient accumulation in trees established following fire exclusion in the southwestern United States.

Increasing tree density that followed fire exclusion after the 1880s in the southwestern United States may have also altered nutrient cycles and led to a carbon (C) sink that constitutes a significant component of the U.S. C budget. Yet, empirical data quantifying century-scale changes in C or nutrients due to fire exclusion are rare. We used tree-ring reconstructions of stand structure from five ponderosa pine-dominated sites from across northern Arizona to compare live tree C, nitrogen (N), and phosphorus (P) storage between the 1880s and 1990s. Live tree biomass in the 1990s contained up to three times more C, N, and P than in 1880s. However, the increase in C storage was smaller than values used in recent U.S. C budgets. Furthermore, trees that had established prior to the 1880s accounted for a large fraction (28-66%) of the C, N, and P stored in contemporary stands. Overall, our century-scale analysis revealed that forests of the 1880s were on a trajectory to accumulate C and nutrients in trees even in the absence of fire exclusion, either because growing conditions became more favorable after the 1880s or because forests in the 1880s included age or size cohorts poised for accelerated growth. These results may lead to a reduction in the C sink attributed to fire exclusion, and they refine our understanding of reference conditions for restoration management of fire-prone forests.

Average Stand Age from Forest Inventory Plots Does Not Describe Historical Fire Regimes in Ponderosa Pine and Mixed-Conifer Forests of Western North America.

Quantifying historical fire regimes provides important information for managing contemporary forests. Historical fire frequency and severity can be estimated using several methods; each method has strengths and weaknesses and presents challenges for interpretation and verification. Recent efforts to quantify the timing of historical high-severity fire events in forests of western North America have assumed that the "stand age" variable from the US Forest Service Forest Inventory and Analysis (FIA) program reflects the timing of historical high-severity (i.e. stand-replacing) fire in ponderosa pine and mixed-conifer forests. To test this assumption, we re-analyze the dataset used in a previous analysis, and compare information from fire history records with information from co-located FIA plots. We demonstrate that 1) the FIA stand age variable does not reflect the large range of individual tree ages in the FIA plots: older trees comprised more than 10% of pre-stand age basal area in 58% of plots analyzed and more than 30% of pre-stand age basal area in 32% of plots, and 2) recruitment events are not necessarily related to high-severity fire occurrence. Because the FIA stand age variable is estimated from a sample of tree ages within the tree size class containing a plurality of canopy trees in the plot, it does not necessarily include the oldest trees, especially in uneven-aged stands. Thus, the FIA stand age variable does not indicate whether the trees in the predominant size class established in response to severe fire, or established during the absence of fire. FIA stand age was not designed to measure the time since a stand-replacing disturbance. Quantification of historical "mixed-severity" fire regimes must be explicit about the spatial scale of high-severity fire effects, which is not possible using FIA stand age data.

Slow lifelong growth predisposes Populus tremuloides trees to mortality.

Widespread dieback of aspen forests, sometimes called sudden aspen decline, has been observed throughout much of western North America, with the highest mortality rates in the southwestern United States. Recent aspen mortality has been linked to drought stress and elevated temperatures characteristic of conditions expected under climate change, but the role of individual aspen tree growth patterns in contributing to recent tree mortality is less well known. We used tree-ring data to investigate the relationship between an individual aspen tree's lifetime growth patterns and mortality. Surviving aspen trees had consistently higher average growth rates for at least 100 years than dead trees. Contrary to observations from late successional species, slow initial growth rates were not associated with a longer lifespan in aspen. Aspen trees that died had slower lifetime growth and slower growth at various stages of their lives than those that survived. Differences in average diameter growth between live and dead trees were significant (α = 0.05) across all time periods tested. Our best logistical model of aspen mortality indicates that younger aspen trees with lower recent growth rates and higher frequencies of abrupt growth declines had an increased risk of mortality. Our findings highlight the need for species-specific mortality functions in forest succession models. Size-dependent mortality functions suitable for late successional species may not be appropriate for species with different life history strategies. For some early successional species, like aspen, slow growth at various stages of the tree's life is associated with increased mortality risk.

Stand-replacing wildfires increase nitrification for decades in southwestern ponderosa pine forests.

Stand-replacing wildfires are a novel disturbance within ponderosa pine (Pinus ponderosa) forests of the southwestern United States, and they can convert forests to grasslands or shrublands for decades. While most research shows that soil inorganic N pools and fluxes return to pre-fire levels within a few years, we wondered if vegetation conversion (ponderosa pine to bunchgrass) following stand-replacing fires might be accompanied by a long-term shift in N cycling processes. Using a 34-year stand-replacing wildfire chronosequence with paired, adjacent unburned patches, we examined the long-term dynamics of net and gross nitrogen (N) transformations. We hypothesized that N availability in burned patches would become more similar to those in unburned patches over time after fire as these areas become re-vegetated. Burned patches had higher net and gross nitrification rates than unburned patches (P < 0.01 for both), and nitrification accounted for a greater proportion of N mineralization in burned patches for both net (P < 0.01) and gross (P < 0.04) N transformation measurements. However, trends with time-after-fire were not observed for any other variables. Our findings contrast with previous work, which suggested that high nitrification rates are a short-term response to disturbance. Furthermore, high nitrification rates at our site were not simply correlated with the presence of herbaceous vegetation. Instead, we suggest that stand-replacing wildfire triggers a shift in N cycling that is maintained for at least three decades by various factors, including a shift from a woody to an herbaceous ecosystem and the presence of fire-deposited charcoal.

Simulating post-wildfire forest trajectories under alternative climate and management scenarios.

Post-fire predictions of forest recovery under future climate change and management actions are necessary for forest managers to make decisions about treatments. We applied the Climate-Forest Vegetation Simulator (Climate-FVS), a new version of a widely used forest management model, to compare alternative climate and management scenarios in a severely burned multispecies forest of Arizona, USA. The incorporation of seven combinations of General Circulation Models (GCM) and emissions scenarios altered long-term (100 years) predictions of future forest condition compared to a No Climate Change (NCC) scenario, which forecast a gradual increase to high levels of forest density and carbon stock. In contrast, emissions scenarios that included continued high greenhouse gas releases led to near-complete deforestation by 2111. GCM-emissions scenario combinations that were less severe reduced forest structure and carbon stock relative to NCC. Fuel reduction treatments that had been applied prior to the severe wildfire did have persistent effects, especially under NCC, but were overwhelmed by increasingly severe climate change. We tested six management strategies aimed at sustaining future forests: prescribed burning at 5, 10, or 20-year intervals, thinning 40% or 60% of stand basal area, and no treatment. Severe climate change led to deforestation under all management regimes, but important differences emerged under the moderate scenarios: treatments that included regular prescribed burning fostered low density, wildfire-resistant forests composed of the naturally dominant species, ponderosa pine. Non-fire treatments under moderate climate change were forecast to become dense and susceptible to severe wildfire, with a shift to dominance by sprouting species. Current U.S. forest management requires modeling of future scenarios but does not mandate consideration of climate change effects. However, this study showed substantial differences in model outputs depending on climate and management actions. Managers should incorporate climate change into the process of analyzing the environmental effects of alternative actions.

Testing a pyroclimatic hypothesis on the Mexico-United States border.

The "pyroclimatic hypothesis" proposed by F. Biondi and colleagues provides a basis for testable expectations about climatic and other controls of fire regimes. This hypothesis asserts an a priori relationship between the occurrence of widespread fire and values of a relevant climatic index. Such a hypothesis provides the basis for predicting spatial and temporal patterns of fire occurrence based on climatic control. Forests near the Mexico-United States border offer a place to test the relative influence of climatic and other controls in mountain ranges that are ecologically similar and subject to broadly similar top-down climatic influence, but with differing cultural influences. We tested the pyroclimatic hypothesis by comparing fire history information from the Mesa de las Guacamayas, a mountain range in northwestern Chihuahua, with previously published fire data from the Chiricahua Mountains, in southeastern Arizona, approximately 150 km away. We developed a priori hypothetical models of fire occurrence and compared their performance to empirical climate-based models. Fires were frequent at all Mesa de las Guacamayas study sites through the mid-20th century and continued uninterrupted to the present at one site, in contrast to nearly complete fire exclusion after 1892 at sites in the Chiricahua Mountains. The empirical regression models explained a higher proportion of the variability in fire regime associated with climate than did the a priori models. Actual climate-fire relationships diverged in each country after 1892. The a priori models predicted continuing fires at the same rate per century as prior to 1892; fires did in fact continue in Mexico, albeit with some alteration of fire regimes, but ceased in the United States, most likely due to changes in land use. The cross-border comparison confirms that a frequent-fire regime could cease without a climatic cause, supporting previous arguments that bottom-up factors such as livestock grazing can rapidly and drastically alter surface fire regimes. Understanding the historical patterns of climate controls on fire could inform the use of historical data as ecological reference conditions and for future sustainability.

A predictive model of community assembly that incorporates intraspecific trait variation.

Community assembly involves two antagonistic processes that select functional traits in opposite directions. Environmental filtering tends to increase the functional similarity of species within communities leading to trait convergence, whereas competition tends to limit the functional similarity of species within communities leading to trait divergence. Here, we introduce a new hierarchical Bayesian model that incorporates intraspecific trait variation into a predictive framework to unify classic coexistence theory and evolutionary biology with recent trait-based approaches. Model predictions exhibited a significant positive correlation (r = 0.66) with observed relative abundances along a 10 °C gradient in mean annual temperature. The model predicted the correct dominant species in half of the plots, and accurately reproduced species' temperature optimums. The framework is generalizable to any ecosystem as it can accommodate any species pool, any set of functional traits and multiple environmental gradients, and it eliminates some of the criticisms associated with recent trait-based community assembly models.

Fire regime in a Mexican forest under indigenous resource management.

The Rarámuri (Tarahumara) people live in the mountains and canyons of the Sierra Madre Occidental of Chihuahua, Mexico. They base their subsistence on multiple-use strategies of their natural resources, including agriculture, pastoralism, and harvesting of native plants and wildlife. Pino Gordo is a Rarámuri settlement in a remote location where the forest has not been commercially logged. We reconstructed the forest fire regime from fire-scarred trees, measured the structure of the never-logged forest, and interviewed community members about fire use. Fire occurrence was consistent throughout the 19th and 20th centuries up to our fire scar collection in 2004. This is the least interrupted surface-fire regime reported to date in North America. Studies from other relict sites such as nature reserves in Mexico or the USA have all shown some recent alterations associated with industrialized society. At Pino Gordo, fires recurred frequently at the three study sites, with a composite mean fire interval of 1.9 years (all fires) to 7.6 years (fires scarring 25% or more of samples). Per-sample fire intervals averaged 10-14 years at the three sites. Approximately two-thirds of fires burned in the season of cambial dormancy, probably during the pre-monsoonal drought. Forests were dominated by pines and contained many large living trees and snags, in contrast to two nearby similar forests that have been logged. Community residents reported using fire for many purposes, consistent with previous literature on fire use by indigenous people. Pino Gordo is a valuable example of a continuing frequent-fire regime in a never-harvested forest. The Rarámuri people have actively conserved this forest through their traditional livelihood and management techniques, as opposed to logging the forest, and have also facilitated the fire regime by burning. The data contribute to a better understanding of the interactions of humans who live in pine forests and the fire regimes of these ecosystems, a topic that has been controversial and difficult to assess from historical or paleoecological evidence.

A century of increasing pine density and associated shifts in understory plant strategies.

We analyzed one of the longest-term ecological data sets to evaluate how forest overstory structure is related to herbaceous understory plant strategies in a ponderosa pine forest. Eighty-two permanent 1-m2 chart quadrats that were established as early as 1912 were remeasured in 2007. We reconstructed historical forest structure using dendrochronological techniques. Ponderosa pine basal area increased from an average of 4 m2/ha in the early 1900s to 29 m2/ha in 2007. Understory plant foliar cover declined by 21%, species richness declined by two species per square meter, and functional diversity also declined. The relative cover of C4 graminoids decreased by 18% and C3 graminoids increased by 19%. Herbaceous plant species with low leaf and fine root nitrogen concentrations, low specific leaf area, high leaf dry matter content, large seed mass, low specific root length, short maximum height, and early flowering date increased in relative abundance in sites where pine basal area increased the most. Overall, we observed a long-term shift in composition toward more conservative shade- and stress-tolerant herbaceous species. Our analysis of temporal changes in plant strategies provides a general framework for evaluating compositional and functional changes in terrestrial plant communities.

El Niño-southern oscillation effect on a fire regime in northeastern Mexico has changed over time.

The El Niño Southern Oscillation (ENSO) is a climate-forcing mechanism that has been shown to affect precipitation and the occurrence of wildfires in many parts of the world. In the southern United States and northern Mexico, warm events (El Niño) are associated with moist winter conditions and fewer fires, while cool events (La Niñia) tend to favor dry winters and more fires. We tested this relationship in a region of northeastern Mexico by characterizing the historical fire regime and climatic influences: Fire regimes were reconstructed from fire-scar samples collected from 100 trees in three high-elevation sites on Peña Nevada in southern Nuevo Le6n. The sites were approximately 25 ha each, and the site centers were approximately 1 km apart. The earliest recorded fire occurred in 1521 and the time period we used for analysis was 1645-1929. The sites were characterized by frequent surface fires before the 1920s. In the three sites, mean fire intervals ranged from 8.6 to 9.6 years (all fires) and 11.9 to 18.6 years (fires that scarred > or = 25% of recording trees). The per-tree mean fire return interval was 17 years, and all three sites burned in the same year seven times between 1774 and 1929. After 1929, fires were nearly eliminated in all sites, likely due to human causes. We found a temporal change in the association between ENSO events and fires; before the 1830s La Niña events were significantly associated with fire years, while after the 1830s this association was not significant. In 1998, when the most severe El Niño event of the past century occurred, the three sites experienced severe, stand-replacing fires that killed many trees that had survived multiple surface fires in the past. Prior to the 1830s, fires tended to occur during dry La Niña years, but since then both La Niña and El Niño have been associated with dry years in this region, especially during the last three decades. This result suggests that ENSO effects have changed over time in this location and that phases of ENSO are not consistent indicators of precipitation, fire occurrence, or fire behavior in this area of northeastern Mexico.