Spatial patterning of fuels and fire hazard across a central U.S. deciduous forest region

Information describing spatial and temporal variability of forest fuel conditions is essential to assessing overall fire hazard and risk. Limited information exists describing spatial characteristics of fuels in the eastern deciduous forest region, particularly in dry oak-dominated regions that historically burned relatively frequently. From an extensive fuels survey of unmanaged forest lands (1,446 plots) we described fuel loadings and spatial patterns of fine and coarse fuels. We attempted to explain the variability in fuel loading of each time-lag fuel class using landscape and seasonal variables through a multiple regression modeling approach. Size class distributions of woody fuels were generally homogeneous across the region except in the glaciated portions of Illinois where loadings appeared lower. Temporally, litter depths progressively decreased from leaffall (November). A fire hazard model that combined seasonal changes in litter depth and fuel moisture content depicted the degree of regional spatial variability during the transition between extreme dry and wet conditions. In the future, fire hazard indices could be paired with ignition probabilities in order to assess spatio-temporal variability of fire risk within the region.     

Using Neutral Models to Identify Constraints on Low-severity Fire Regimes

Climate, topography, fuel loadings, and human activities all affect spatial and temporal patterns of fire occurrence. Because fire is modeled as a stochastic process, for which each fire history is only one realization, a simulation approach is necessary to understand baseline variability, thereby identifying constraints, or forcing functions, that affect fire regimes. With a suitable neutral model, characteristics of natural fire regimes estimated from fire history data can be compared to a “null hypothesis”. We generated random landscapes of fire-scarred trees via a point process with sequential spatial inhibition. Random ignition points, fire sizes, and fire years were drawn from uniform and exponential family probability distributions. We compared two characteristics of neutral fire regimes to those from five watersheds in eastern Washington that have experienced low-severity fire. Composite fire intervals (CFIs) at multiple spatial scales displayed similar monotonic decreases with increasing sample area in neutral vs. real landscapes, although patterns of residuals from statistical models differed. In contrast, parameters of the Weibull distribution associated with temporal trends in fire hazard exhibited different forms of scale dependence in real vs. simulated data. Clear patterns in neutral landscapes suggest that deviations from them in empirical data represent real constraints on fire regimes (e.g., topography, fuels). As with any null model, however, neutral fire-regime models need to be carefully tuned to avoid confounding these constraints with artifacts of modeling. Neutral models show promise for investigating low-severity fire regimes to separate intrinsic properties of stochastic processes from the effects of climate, fuel loadings, topography, and management.     

Influence of forest management alternatives and land type on susceptibility to fire in northern Wisconsin,USA

We used the LANDIS disturbance and succession model to study the effects of six alternative vegetation management scenarios on forest succession and the subsequent risk of canopy fire on a 2791 km² landscape in northern Wisconsin, USA. The study area is a mix of fire-prone and fire-resistant land types. The alternatives vary the spatial distribution of vegetation management activities to meet objectives primarily related to forest composition and recreation. The model simulates the spatial dynamics of differential reproduction, dispersal, and succession patterns using the vital attributes of species as they are influenced by the abiotic environment and disturbance. We simulated 50 replicates of each management alternative and recorded the presence of species age cohorts capable of sustaining canopy fire and the occurrence of fire over 250 years. We combined these maps of fuel and fire to map the probability of canopy fires across replicates for each alternative. Canopy fire probability varied considerably by land type. There was also a subtle, but significant effect of management alternative, and there was a significant interaction between land type and management alternative. The species associated with high-risk fuels (conifers) tend to be favored by management alternatives with more disturbances, whereas low disturbance levels favor low-risk northern hardwood systems dominated by sugar maple. The effect of management alternative on fire risk to individual human communities was not consistent across the landscape. Our results highlight the value of the LANDIS model for identifying specific locations where interacting factors of land type and management strategy increase fire risk.This revised version was published online in May 2005 with corrections to the Cover Date.     

Edge effects on fire spread in a disturbed Northern Wisconsin landscape

The effect of area-of-edge influence (AEI) on fire size and movement was simulated by considering the distribution of single and multiple edges in the Chequamegon-Nicolet National Forest in Northern Wisconsin, USA. Six hypothetical landscapes with different delineations of AEIs were created for simulating fire spread using FARSITE to evaluate the influence of edges on the rate and direction of fire spread. The burned area differed significantly among the six landscapes. In the three scenarios with buffered edges, the burned area increased by 35% with high loading fuel in AEIs, while it decreased by 21 and 46% with medium and low fuel loading in the AEIs, respectively, as compared to the no edge scenario. In two scenarios we delineated the area-of-multiple-edge influence (AMEI) and placed more than one high loading fuel within it. This increased the burned area by 5% from the high buffered edge scenario and by 40% from the control. When the depth-of-edge influence (DEI) was doubled to 60 m using AMEI with high fuels, the burned area increased by 20% from the high buffered edge scenario and by 60% from the control. We found that low and medium fuel loading slowed the fire spread and over time, caused the fire front to change direction. In high fuel loading scenarios, AEIs acted as corridors facilitating the fire spread by providing a contiguous patch of fuel which allowed fires to increase in size and pulled the fire front in the same direction.     

Landscape heterogeneity and fire behavior: scale-dependent feedback between fire and grazing processes

Fire and grazing are ecological processes that frequently interact to modify landscape patterns of vegetation. There is empirical and theoretical evidence that response of herbivores to heterogeneity is scale-dependent however the relationship between fire and scale of heterogeneity is not well defined. We examined the relationship between fire behavior and spatial scale (i.e., patch grain) of fuel heterogeneity. We created four heterogeneous landscapes modeled after those created by a fire–grazing interaction that differed in grain size of fuel patches. Fire spread was simulated through each model landscape from 80 independent, randomly located ignition points. Burn area, burn shape complexity and the proportion of area burnt by different fire types (headfire, backfire and flankfire) were all affected by the grain of fuel patch. The area fires burned in heterogeneous landscapes interacted with the fuel load present in the patch where ignition occurred. Burn complexity was greater in landscapes with small patch grain than in landscapes with large patch grain. The proportion of each fire type (backfire, flankfire and headfire) was similar among all landscapes regardless of patch grain but the variance of burned area within each of the three fire types differed among treatments of patch grain. Our landscape fire simulation supports the supposition that feedbacks between landscape patterns and ecological processes are scale-dependent, in this case spatial scale of fuel loading altering fire spread through the landscape.     

Bottom-up control of a northern Arizona ponderosa pine forest fire regime in a fragmented landscape

Fire regimes often vary at fine spatial scales in response to factors such as topography or fuels while climate usually synchronizes fires across broader scales. We investigated the relative influence of top-down and bottom-up controls on fire occurrence in ponderosa pine (Pinus ponderosa) forests in a highly fragmented landscape at Mount Dellenbaugh, in northwestern Arizona. Our study area of 4,000?ha was characterized by patches of ponderosa pine forest in drainages that were separated by a matrix of pinyon?Cjuniper woodlands, sagebrush shrublands, and perennial grasslands. We reconstructed fire histories from 135 fire-scarred trees in sixteen 25-ha sample sites placed in patches of mature ponderosa forest. We found that, among patches of ponderosa forest, fires were similar in terms of frequency but highly asynchronous in terms of individual years. Climate synchronized fire but only across broader spatial scales. Fires occurring at broader scales were associated with dry years that were preceded by several wet years. The remarkable level of asynchrony at finer scales suggests that bottom-up factors, such as site productivity and fuel continuity, were important in regulating fire at Mount Dellenbaugh. Understanding where bottom-up controls were historically influential is important for prioritizing areas that may best respond to fuel treatment under a warming climate.     

Modeling fire and landform influences on the distribution of old-growth pinyon-juniper woodland

Expansion of Pinus and Juniperus species into shrub steppe in semi-arid regions of the western United States has been widely documented and attributed in part to fire exclusion. If decreased fire frequency has been an important cause of woodland expansion, one would expect to find age structures dominated by younger trees on more fire-prone sites, with old-growth pinyon-juniper woodland limited to sites with lower fire risk. We compared current old-growth distribution with spatial models for fire risk in a 19-km² watershed in central Nevada, USA. Multiple GIS models were developed to represent fire susceptibility, according to abiotic factors representing fuels and topographic barriers to fire spread. We also developed cellular automata models to generate fire susceptibility surfaces that additionally account for neighborhood effects. Rule-based GIS models failed to predict old-growth distribution better than random models. Cellular automata models incorporating spatial heterogeneity of site productivity predicted old-growth distribution better than random models but with low accuracy, ranging from 58% agreement at the single-pixel (0.09-ha) scale to 80% agreement for 20-pixel neighborhoods. The best statistical model for predicting old-growth occurrence included the negative effect of topographic convergence index (local wetness), and the positive effects of solar insolation and proximity to rock outcrops. Results support the hypothesis that old-growth woodlands in the Great Basin are more likely to occur on sites with low fire risk. However, weak relationships suggest that old-growth woodlands have not been confined to fire-safe sites. Conservation efforts should consider the landscape context of old-growth woodlands across a broad landscape, with an emphasis on conserving landscape variability in tree age structure.     

Use of artificial landscapes to isolate controls on burn probability

Techniques for modeling burn probability (BP) combine the stochastic components of fire regimes (ignitions and weather) with sophisticated fire growth algorithms to produce high-resolution spatial estimates of the relative likelihood of burning. Despite the numerous investigations of fire patterns from either observed or simulated sources, the specific influence of environmental factors on BP patterns is not well understood. This study examined the relative effects of ignitions, fuels, and weather on mean BP and spatial patterns in BP (i.e., BP variability) using highly simplified artificial landscapes and wildfire simulation methods. Our results showed that a limited set of inputs yielded a wide range of responses in the mean and spatial patterning of BP. The input factors contributed unequally to mean BP and to BP variability: so-called top-down controls (weather) primarily influenced mean BP, whereas bottom-up influences (ignitions and fuels) were mainly responsible for the spatial patterns of BP. However, confounding effects and interactions among factors suggest that fully separating top-down and bottom-up controls may be impossible. Furthermore, interactions among input variables produced unanticipated but explainable BP patterns, hinting at complex topological dependencies among the main determinants of fire spread and the resulting BP. The results will improve our understanding of the spatial ecology of fire regimes and help in the interpretation of patterns of fire likelihood on real landscapes as part of future wildfire risk assessments.     

Spatial models for inferring topographic controls on historical low-severity fire in the eastern Cascade Range of Washington,USA

Fire regimes are complex systems that represent an aggregate of spatial and temporal events whose statistical properties are scale dependent. Despite the breadth of research regarding the spatial controls on fire regime variability, few datasets are available with sufficient resolution to test spatially explicit hypotheses. We used a spatially distributed network of georeferenced fire-scarred trees to investigate the spatial structure of fire occurrence at multiple scales. Mantel’s tests and geostatistical analysis of fire-occurrence time series led to inferences about the mechanisms that generated spatial patterns of historical fire synchrony (multiple trees recording fire in a single year) in eastern Washington, USA. The spatial autocorrelation structure of historical fire regimes varied within and among sites, with clearer patterns in the complex rugged terrain of the Cascade Range than in more open and rolling terrain further north and east. Results illustrate that the statistical spatial characteristics of fire regimes change with landform characteristics within a forest type, suggesting that simple relationships between fire frequency, fire synchrony, and forest type do not exist. Quantifying the spatial structures in fire occurrence associated with topographic variation showed that fire regime variability depends on both landscape structure and the scale of measurement. Spatially explicit fire-scar data open new possibilities for analysis and interpretation, potentially informing the design and application of fire management on landscapes, including hazardous fuel treatments and the use of fire for ecosystem restoration.

Effects of weather, fuel and terrain on fire severity in topographically diverse landscapes of south-eastern Australia

The effects of weather, terrain, fuels on fire severity were compared using remote sensing of the severity of two large fires in south-eastern Australian forests. The probability of contrasting levels of fire severity (fire confined to the understorey vs. tree canopies consumed) was analysed using logistic regression. These severities equate to extremes of fire intensity (<1,500 vs. >10,000 kW m^?1), consequent suppression potential (high vs. nil) and potential adverse ecological impacts on vertebrate fauna and soils (low vs. high). Weather was the major influence on fire severity. Crown fire was absent under non-extreme weather and but more likely under extreme weather, particularly on ridges in vegetation unburnt for >10 years. Crown fire probability was very low in recently burnt vegetation (1–5 years) and increased at higher fuel ages. In all cases, fire severity was lower in valleys, probably due to effects of wind protection and higher fuel moisture in moderating fire behaviour. Under non-extreme weather, fires are likely to be suppressible and burn heterogeneously, due to the influence of topographic position, slope and fuel load. Under extreme weather, fires are influenced only by fuel and topographic position, and probability of suppression on accessible ridges will be low except in recently burnt (i.e. 1–5 year old) fuels. Topographically imposed variation may mitigate adverse ecological effects on arboreal fauna and soil erosion potential.     

Spatial and temporal heterogeneity of agricultural fires in the central United States in relation to land cover and land use

Agricultural burning is an important land use practice in the central U.S. but has received little attention in the literature, whereas most of the focus has been on wildfires in forested areas. Given the effects that agricultural burning can have on biodiversity and emissions of greenhouse gasses, there is a need to quantify the spatial and temporal patterns of fire in agricultural landscapes of the central U.S. Three years (2006?C2008) of the MODIS 1?km daily active fire product generated from the MODIS Terra and Aqua satellite data were used. The 2007 Cropland Data Layer developed by the U.S. Department of Agriculture was used to examine fire distribution by land cover/land use (LCLU) type. Global ordinary least square (OLS) models and local geographically weighted regression (GWR) analyses were used to explore spatial variability in relationships between fire detection density and LCLU classes. The monthly total number of fire detections peaked in April and the density of fire detections (number of fires/km²/3?years) was generally higher in areas dominated by agriculture than areas dominated by forest. Fire seasonality varied among areas dominated by different types of agriculture and land use. The effects of LCLU classes on fire detection density varied spatially, with grassland being the primary correlate of fire detection density in eastern Kansas; whereas wheat cropping was important in central Kansas, northeast North Dakota, and northwest Minnesota.     

Anthropogenic influences on potential fire spread in a pyrogenic ecosystem of Florida, USA

Fire has historically been an important ecological factor maintaining southeastern U.S. vegetation. Humans have altered natural fire regimes by fragmenting fuels, introducing exotic species, and suppressing fires. Little is known about how these alterations specifically affect spatial fire extent and pattern. We applied historic (1920 and 1943) and current (1990) GIS fuels maps and the FARSITE fire spread model to quantify the differences between historic and current fire spread distributions. We held all fire modeling variables (wind speed and direction, cloud cover, precipitation, humidity, air temperature, fuel moistures, ignition source and location) constant with exception of the fuel models representing different time periods. Model simulations suggest that fires during the early 1900's burned freely across the landscape, while current fires are much smaller, restricted by anthropogenic influences. Fire extent declined linearly with patch density, and there was a quadratic relationship between fire extent and percent landscape covered by anthropogenic features. We found that as little as 10 percent anthropogenic landcover caused a 50 percent decline in fire extent. Most landscapes (conservation or non-conservation areas) are now influenced by anthropogenic features which disrupt spatial fire behavior disproportionately to their actual size. These results suggest that land managers using fire to restore or maintain natural ecosystem function in pyrogenic systems will have to compensate for anthropogenic influences in their burn planning. This revised version was published online in May 2005 with corrections to the Cover Date. This revised version was published online in July 2006 with corrections to the Cover Date.     

Spatial and non-spatial factors: When do they affect landscape indicators of watershed loading?

The percentage of a watershed occupied by agricultural areas is widely used to predict nutrient loadings and in-stream water chemistry because water quality is often linked to non-point sources in a watershed. Measures of the spatial location of source areas have generally not been incorporated into such landscape indicators although empirical evidence and watershed loading models suggest that spatially explicit information is useful for predicting loadings. I created a heuristic grid-based surface-flow model to address the discrepancies between spatially explicit and non-spatial approaches to understanding watershed loading. The mean and variance in loading were compared among thousands of simulated watersheds with varying percentages of randomly located source and sinks. The variability in loading among replicate landscapes was greatest for those landscapes with ~65% source areas. This variance peak suggests that considering the spatial arrangement of cover types is most important for watersheds with intermediate relative abundances of sources and sinks as the wide variety of different spatial configurations can lead to either very high or very low loading. Increasing the output from source pixels (relative to the amount absorbed by sink pixels) among different landscapes moved the peak in variance to landscapes with lower percentages of sources. A final scenario examined both broad- and fine-scale heterogeneity in source output to disentangle the relative contributions of spatial configuration, percentage of source covers, and heterogeneity of sources in governing variability in loading. In landscapes with high percentages of source pixels, fine-scale heterogeneity in source output was responsible for a greater portion of the total variability in loading among different watersheds than was spatial arrangement. These results provide several testable hypotheses for when spatial and non-spatial approaches might be most useful in relating land cover to water chemistry and suggest improvements for the spatial sensitivity analyses of eco-hydrologic watershed models.     

Using stochastic simulation to evaluate competing risks of wildfires and fuels management on an isolated forest carnivore

Natural resource managers are often challenged with balancing requirements to maintain wildlife populations and to reduce risks of catastrophic or dangerous wildfires. This challenge is exemplified in the Sierra Nevada of California, where proposals to thin vegetation to reduce wildfire risks have been highly controversial, in part because vegetation treatments could adversely affect an imperiled population of the fisher (Martes pennanti) located in the southern Sierra Nevada. The fisher is an uncommon forest carnivore associated with the types of dense, structurally complex forests often targeted for fuel reduction treatments. Vegetation thinning and removal of dead-wood structures would reduce fisher habitat value and remove essential habitat elements used by fishers for resting and denning. However, crown-replacing wildfires also threaten the population’s habitat, potentially over much broader areas than the treatments intended to reduce wildfire risks. To investigate the potential relative risks of wildfires and fuels treatments on this isolated fisher population, we coupled three spatial models to simulate the stochastic and interacting effects of wildfires and fuels management on fisher habitat and population size: a spatially dynamic forest succession and disturbance model, a fisher habitat model, and a fisher metapopulation model, which assumed that fisher fecundity and survivorship correlate with habitat quality. We systematically varied fuel treatment rate, treatment intensity, and fire regime, and assessed their relative effects on the modeled fisher population over 60 years. After estimating the number of adult female fishers remaining at the end of each simulation scenario, we compared the immediate negative effects of fuel treatments to the longer-term positive effect of fuel treatment (via reduction of fire hazard) using structural equation modeling. Our simulations suggest that the direct, negative effects of fuel treatments on fisher population size are generally smaller than the indirect, positive effects of fuel treatments, because fuels treatments reduced the probability of large wildfires that can damage and fragment habitat over larger areas. The benefits of fuel treatments varied by elevation and treatment location with the highest net benefits to fisher found at higher elevations and within higher quality fisher habitat. Simulated fire regime also had a large effect with the largest net benefit of fuel treatments occurring when a more severe fire regime was simulated. However, there was large uncertainty in our projections due to stochastic spatial and temporal wildfires dynamic and fisher population dynamics. Our results demonstrate the difficulty of projecting future populations in systems characterized by large, infrequent, stochastic disturbances. Nevertheless, these coupled models offer a useful decision-support system for evaluating the relative effects of alternative management scenarios; and uncertainties can be reduced as additional data accumulate to refine and validate the models.     

A large-scale fire suppression edge-effect on forest composition in the New Jersey Pinelands

Broad scale ecological edge-effects are most likely common in urbanized landscapes prone to wildfire, but most edge-effect studies have focused on fine scale processes such as shade tolerance and seed dispersal. Evidence has suggested a shift from pine dominated to oak dominated forests at the interface of developed land and natural areas in the Pinelands of New Jersey with the presence of a large edge-effect due to fire suppression. The goal of this study was to assess the location, magnitude and mechanism of the shift from pine to oak cover focusing on distance to human-altered land as the driver of fire suppression and forest composition changes. Overall, fire frequency and upland pine cover decreased sharply closer to human-altered land and affected up to 420 m of adjacent upland forest. Other factors, such as prescribed fire and wetlands configurations may play a role in the interior forest dynamics, but trends toward lower upland pine forest cover and higher upland oak cover near human altered were dominant. The areal summations of distance from altered land and the use of percent change thresholds for determining the scale and magnitude of large scale ecological edge-effects could be useful to managers attempting to maintain or restore forest types in areas of high wildland–urban interface.     

The relative impacts of vegetation,topography and spatial arrangement on building loss to wildfires in case studies of California and Colorado

Context

Wildfires destroy thousands of buildings every year in the wildland urban interface. However, fire typically only destroys a fraction of the buildings within a given fire perimeter, suggesting more could be done to mitigate risk if we understood how to configure residential landscapes so that both people and buildings could survive fire.

Objectives

Our goal was to understand the relative importance of vegetation, topography and spatial arrangement of buildings on building loss, within the fire’s landscape context.

Methods

We analyzed two fires: one in San Diego, CA and another in Boulder, CO. We analyzed Google Earth historical imagery to digitize buildings exposed to the fires, a geographic information system to measure some of the explanatory variables, and FRAGSTATS to quantify landscape metrics. Using logistic regression we conducted an exhaustive model search to select the best models.

Results

The type of variables that were important varied across communities. We found complex spatial effects and no single model explained building loss everywhere, but topography and the spatial arrangement of buildings explained most of the variability in building losses. Vegetation connectivity was more important than vegetation type.

Conclusions

Location and spatial arrangement of buildings affect which buildings burn in a wildfire, which is important for urban planning, building siting, landscape design of future development, and to target fire prevention, fuel reduction, and homeowner education efforts in existing communities. Landscape context of buildings and communities is an important aspect of building loss, and if taken into consideration, could help communities adapt to fire.

     

Modeling acoustic diversity using soundscape recordings and LIDAR-derived metrics of vertical forest structure in a neotropical rainforest

We determined the relationship between acoustic diversity and metrics of vertical forest structure derived from light detection and ranging (LIDAR) data in a neotropical rainforest in Costa Rica. We then used the LIDAR-derived metrics to predict acoustic diversity across the forest landscape. Sound recordings were obtained from 14 sites for six consecutive days during dusk chorus (6?pm). Acoustic diversity was calculated for each day as the total intensity across acoustic frequency bands using the Shannon index and then averaged over the 6?days at each site. A 10?m radius around each site was used to obtain several LIDAR-derived metrics describing the vertical structural attributes of the forest canopy. Multiple linear regression (MLR) with Akaike information criterion was used to determine a top-ranked model with acoustic diversity as the dependent variable and the LIDAR metrics as independent variables. Acoustic diversity was modeled for forested areas (where canopy height was?>20?m) at 20?m resolution using coefficients obtained from the MLR, and a hotspot analysis was conducted on the resulting layer. Acoustic diversity was strongly correlated (R ² ?=?0.75) with the LIDAR metrics suggesting that LIDAR-derived metrics can be used to determine canopy structural attributes important to vocal fauna species. The hotspot analysis revealed that the spatial distribution of these canopy structural attributes across the La Selva forest is not random. Our approach can be used to identify forest patches of potentially high acoustic diversity for conservation or management purposes.     

Human influence on the abundance and connectivity of high-risk fuels in mixed forests of northern Wisconsin,USA

Though fire is considered a natural disturbance, humans heavily influence modern wildfire regimes. Humans influence fires both directly, by igniting and suppressing fires, and indirectly, by either altering vegetation, climate, or both. We used the LANDIS disturbance and succession model to compare the relative importance of a direct human influence (suppression of low intensity surface fires) with an indirect human influence (timber harvest) on the long-term abundance and connectivity of high-risk fuel in a 2791 km² landscape characterized by a mixture of northern hardwood and boreal tree species in northern Wisconsin. High risk fuels were defined as a combination of sites recently disturbed by wind and sites containing conifer species/cohorts that might serve as ladder fuel to carry a surface fire into the canopy. Two levels of surface fire suppression (high/current and low) and three harvest alternatives (no harvest, hardwood emphasis, and pine emphasis) were compared in a 2×3 factorial design using 5 replicated simulations per treatment combination over a 250-year period. Multivariate analysis of variance indicated that the landscape pattern of high-risk fuel (proportion of landscape, mean patch size, nearest neighbor distance, and juxtaposition with non fuel sites) was significantly influenced by both surface fire suppression and by forest harvest (p > 0.0001). However, the two human influences also interacted with each other (p < 0.001), because fire suppression was less likely to influence fuel connectivity when harvest disturbance was simultaneously applied. Temporal patterns observed for each of seven conifer species indicated that disturbances by either fire or harvest encouraged the establishment of moderately shade-tolerant conifer species by disturbing the dominant shade tolerant competitor, sugar maple. Our results conflict with commonly reported relationships between fire suppression and fire risk observed within the interior west of the United States, and illustrate the importance of understanding key interactions between natural disturbance, human disturbance, and successional responses to these disturbance types that will eventually dictate future fire risk.This revised version was published online in May 2005 with corrections to the Cover Date.     

Untangling the positive and negative effects of shrubs on herbaceous vegetation in drylands

Woody vegetation, as an ecosystem engineer, can modulate the landscape such that the levels of resources in its vicinity undergo positive and negative changes as far as the herbaceous vegetation is concerned. To better understand how these processes play out in a semi-arid ecosystem, we examined resource modulation by woody vegetation, and the response of herbaceous vegetation to that modulation, at a fine spatial scale. Experimental manipulations were employed to separate the positive and negative effects of water, light and seed dispersal in determining herbaceous species density and biomass in three patch types within and adjacent to the shrub (core, periphery and open). We synthesized our results into a multilayered landscape diversity (MLLD) model. Woody vegetation creates distinct multilayered resource patches at its core and periphery which do not correspond to the dichotomous structural pattern of shrub canopy versus intershrub background. The combined effect of these multilayered resource patches had higher herbaceous species density (8.2 vs. 4.0 species 400?cm^?2) and herbaceous biomass (5.4 vs. 1.0?g 400?cm^?2) in the periphery than in the core (3-yr averages). The periphery??s net positive effects are due to enhancement of soil properties (water infiltration depth of 11.1?cm at periphery vs. 8.1?cm at core), while the core??s net negative effects are due to modulation of seed (seed abundance per seed trap of 44.2 at periphery vs. 3.0 at core) and light availability (PAR transmittance of 41.9?% at periphery vs. 16.5?% at core) by the shrub canopy. Thus, when examined at this fine spatial resolution, woody vegetation has both net positive and net negative effects on herbaceous vegetation. Analysis of our results by means of the MLLD model emphasizes the importance of examining the landscape at the spatial scale of the modulated resources and of recognizing different patch types and their differing effects on herbaceous vegetation.