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.

Global Vegetation Dynamics: Concepts and Applications in MC1 model describes the creation in the mid 1990s, architecture, uses, and limitations of the MC1 dynamic global vegetation model (DGVM) that is being used by an increasing number of research groups around the world. The scientific foundation of most models is often poorly documented and difficult to access, and a centralized source of information for MC1, including the complete list of over eighty papers and reports with MC1 results will be useful to scientists and users who want to better understand the model and the output it generates.

Global Vegetation Dynamics: Concepts and Applications in MC1 model will be a valuable resource for students and researchers in the fields of climate change science, conservation science, biogeochemistry and ecology, as well as for land managers looking for a better understanding of the projections of climate change impacts and of the tools that have been developed to produce them.

Table of Contents

Part I: General Description of the Model MC1

1  History and General Description of the Dynamic Global Vegetation Model MC1- Dominique Bachelet

2  Historical Climate and Suppression Effects on Simulated Fire and Carbon Dynamics in the Conterminous United States-James M Lenihan and Dominique Bachelet

3  Challenges and Limitations of Using a DGVM for Local to Regional Applications- Dominique Bachelet, Brendan M Rogers, and David R Conklin

4  The Making of a Dynamic General Vegetation Model, MC1- Ronald P Neilson

Part II: Examples of Projects Using MC1 at Various Spatial Scales

5  A Brief Description of the VINCERA Project; Vulnerability and Impacts of North American Forests to Climate Change: Ecosystem Responses and Adaptation- David T Price, Daniel Scott, Mark R Lomas, Daniel W McKenney, Dominique Bachelet, Raymond J Drapek, James M Lenihan, Ronald P Neilson, F I Woodward, and Jonathan A Foley

6  Continent wide Simulations of a Dynamic Global Vegetation Model over the United States and Canada under Nine AR4 Future Scenarios- Raymond J Drapek, John B Kim, and Ronald P Neilson

7  Drivers of Future Ecosystem Change in the US Pacific Northwest: The Role of Climate, Fire, and Nitrogen- Brendan M Rogers, Dominique Bachelet, Raymond J Drapek, Beverly E Law, Ronald P Neilson, and John R Wells 91

8  Application of MC1 to Wind Cave National Park: Lessons from a Small Scale Study- David A King, Dominique Bachelet, and Amy J Symstad

9  Simulating Effects of Climate and Vegetation Change on Distributions of Martens and Fishers in the Sierra Nevada, California, Using Maxent and MC1- Wayne D Spencer, Heather Rustigian Romsos, Ken Ferschweiler, and Dominique Bachelet

Part III: Packaging MC1 Results to Increase Its Usability by Managers

10  Using a Dynamic Global Vegetation Model to Help Inform Management Decisions- Joshua S Halofsky, Jessica E Halofsky, David R Conklin, Dominique Bachelet, Miles A Hemstrom, Becky K Kerns, and Anita T Morzillo

11  Bringing MC1 Model Results to Data Basin to Facilitate Access, Distribution, and Interpretation- Dominique Bachelet and the CBI Data Basin team

Appendix: Publications and Reports Featuring MC1

Glossary

Index

The 1994 Northwest Forest Plan (NWFP) shifted federal lands management from a focus on timber production to ecosystem management and biodiversity conservation. The plan established a network of conservation reserves and an ecosystem management strategy on ~10 million hectares from northern California to Washington State, USA, within the range of the federally threatened northern spotted owl (Strix occidentalis caurina). Several subsequent assessments—and 20 years of data from monitoring programs established under the plan—have demonstrated the effectiveness of this reserve network and ecosystem management approach in making progress toward attaining many of the plan’s conservation and ecosystem management goals. This paper (1) showcases the fundamental conservation biology and ecosystem management principles underpinning the NWFP as a case study for managers interested in large-landscape conservation; and (2) recommends improvements to the plan’s strategy in response to unprecedented climate change and land-use threats. Twenty years into plan implementation, however, the U.S. Forest Service and Bureau of Land Management, under pressure for increased timber harvest, are retreating from conservation measures. We believe that federal agencies should instead build on the NWFP to ensure continuing success in the Pacific Northwest. We urge federal land managers to (1) protect all remaining late-successional/old-growth forests; (2) identify climate refugia for at-risk species; (3) maintain or increase stream buffers and landscape connectivity; (4) decommission and repair failing roads to improve water quality; (5) reduce fire risk in fire-prone tree plantations; and (6) prevent logging after fires in areas of high conservation value. In many respects, the NWFP is instructive for managers considering similar large-scale conservation efforts.

We used Maxent distribution models and MC1 to investigate effects of climate and vegetation on the distribution of martens (Martes caurina) and fishers (Pekania pennanti) in the Sierra Nevada, California, under current and projected future conditions. Both species are forest carnivores of conservation concern in California, where they reach their southernmost distributions. The species occupy similar ecological niches and may compete in the elevation band where their ranges overlap—but martens mostly occupy higher elevations with deep, persistent snow, and fishers occupy lower elevations with less snow. We systematically varied types of environmental variables (climate, vegetation, terrain, presence or absence of the other species) included in Maxent models and compared area‐under‐curve (AUC) values to determine what variables best predict current distributions. Terrain variables and presence or absence of the competing species did not add significantly to model fit. For fishers, models using both climate and vegetation variables outperformed those using only vegetation; for martens, there was no significant difference between vegetationonly, climate‐only, and vegetation + climate models. We then prepared climate + vegetation Maxent models using MC1‐derived variables that best approximated the variables used in the best current (benchmark) models, compared predicted distributions with benchmark models, and projected distributions to mid‐ and late 21st century using MC1 vegetation projections and an array of downscaled general circulation models (GCMs) and emission scenarios at three resolutions (10 km, 4 km, 800 m). The finest available GCM resolution (800 m) provided the best spatial congruence between MC1‐derived models and benchmark models. Regardless of GCM emission scenario, predicted marten distribution shifted to higher elevations, became more fragmented, and decreased in area by 40−85% (depending on scenario) compared to current distributions. Predicted changes in fisher distribution were more variable across GCM scenarios, with some increases and some decreases in extent and no consistent elevation shifts—suggesting high uncertainty in climate change effects on fishers. Management to benefit these species should consider ways of sustaining appropriate vegetation conditions within their preferred climate envelopes via adaptive management.

Aim: Forest regeneration data provide an early signal of the persistence and migration of tree species, so we investigated whether species shifts due to climate change exhibit a common signal of response or whether changes vary by species.

Location: California Floristic Province, United States; mediterranean biome.

Methods: We related Forest Inventory and Analysis (FIA) data from 2000−07 for 13 tree species to high-resolution climate and geographical data. Using methods from invasion ecology, we derived indices of species-specific regeneration overlap and central tendency change (range-wide global indicators) based on kernel density estimation of presence and absence of regeneration. We then built regeneration surfaces to identify areas of occurrence of high regeneration (regeneration
hotspots, local indicators) in both geographical and climate space for 13 common tree species.

Results: Differences between presence and absence of regeneration in forests varied in magnitude across species, with little evidence that tree regeneration is shifting to higher latitudes and elevations, the expected geographical fingerprint of climate change. We also identified potential topographic mediators of regeneration dynamics. Multiple regeneration hotspots were found for many species, suggesting the influence of non-climatic factors on regeneration. Differences between the presence and absence of regeneration in geographic and climate spaces were not always congruent, suggesting that shifting climate space and range area are not entirely coupled.

Main conclusions: The distributions of regeneration in Californian forests show diverse signals, not always tracking the higher latitudinal–elevation fingerprint of climate change. Local regeneration hotspots are common in our analysis, suggesting spatially varying persistence of forest linked to natural and anthropogenic disturbances. Our results emphasize that projections of tree range shifts in the context of climate change should consider the variation of regeneration drivers

Climate change has significant effects on critical ecosystem functions such as carbon and water cycling. Vegetation and especially forest ecosystems play an important role in the carbon and hydrological cycles.Vegetation models that include detailed belowground processes require accurate soil data to decrease uncertainty and increase realism in their simulations. The MC2 DGVM uses three modules to simulate biogeography, biogeochemistry and fire effects, all three of which use soil data either directly or indirectly. This study includes a correlation analysis of the MC2 model to soil depth by comparing a subset of the model’s carbon and hydrological outputs using soil depth data of different scales and qualities. The results show that the model is very sensitive to soil depth in simulations of carbon and hydrological variables, but competing algorithms make the fire module less sensitive to changes in soil depth. Simulated historic evapotranspiration and net primary productivity show the strongest positive correlations (both have correlation coefficients of 0.82). The strongest negative correlation is streamflow (0.82). Ecosystem carbon, vegetation carbon and forest carbon show the next strongest correlations (0.78, 0.74 and 0.74, respectively). Carbon consumed by forest fires and the part of each grid cell burned show only weak negative correlations (0.24 and 0.0013 respectively). In the model, when the water demand is met (deep soil with good water availability), production increases and fuels build up as more litter gets generated, thus increasing the overall fire risk during upcoming dry periods. However, when soil moisture is low, fuels dry and fire risk increases. In conclusion, it is clear climate change impact models need accurate soil depth data to simulate the resilience or vulnerability of ecosystems to future conditions.

Click here to see the spatial data on Data Basin

The potential evapotranspiration (PET) that would occur with unlimited plant access to water is a central driver of simulated plant growth in many ecological models. PET is in!uenced by solar and longwave radiation, temperature, wind speed, and humidity, but it is often modeled as a function of temperature alone. This approach can cause biases in projections of future climate impacts in part because it confounds the effects of warming due to increased greenhouse gases with that which would be caused by increased radiation from the sun. We developed an algorithm for linking PET to extraterrestrial solar radiation (incoming top-of atmosphere solar radiation), as well as temperature and atmospheric water vapor pressure, and incorporated this algorithm into the dynamic global vegetation model MC1. We tested the new algorithm for the Northern Great Plains, USA, whose remaining grasslands are threatened by continuing woody encroachment. Both the new and the standard temperature-dependent MC1 algorithm adequately simulated current PET, as compared to the more rigorous PenPan model of Rotstayn et al. (2006). However, compared to the standard algorithm, the new algorithm projected a much more gradual increase in PET over the 21st century for three contrasting future climates. This difference led to lower simulated drought effects and hence greater woody encroachment with the new algorithm, illustrating the importance of more rigorous calculations of PET in ecological models dealing with climate change.

To assess the genetic diversity and phylogeography of the blunt-nosed leopard lizard (Gambelia sila), we sequenced 1,285 base pairs (bp) of the mitochondrial cytochrome oxidase-b (cyt-b, 682 bp) and cytochrome oxidase III (CO3, 603 bp) genes from 33 individuals representing eight natural populations in central California. Phylogenetic analysis indicated that 17 observed haplotypes are partitioned into two major clades, which correspond geographically to where the lizards were collected. We also conducted a focused analysis of individuals collected from the canyons leading into the Cuyama Valley in Ventura and Santa Barbara counties, a geographic area with lizards possibly representing a remnant hybrid (with G. wislizenii) population. All lizards from the Cuyama Valley and adjacent canyons exhibited the mitochondrial haplotype of G. sila and were embedded within one clade. Our morphological analysis placed some leopard lizards collected from Cuyama Valley with true G. sila, whereas some individuals aggregated with G. wislizenii. This finding suggests that the quantitative morphological characteristics often used to distinguish between the two species are fairly labile and may be influenced by prevailing environmental conditions.

Background: Frequent outbreaks of insects and diseases have been recorded in the native forests of western North America during the last few decades, but the distribution of these outbreaks has been far from uniform. In some cases, recent climatic variations may explain some of this spatial variation along with the presence of expansive forests composed of dense, older trees. Forest managers and policy makers would benefit if areas especially prone to disturbance could be recognized so that mitigating actions could be taken.

Methods: We use two ponderosa pine-dominated sites in western Montana, U.S.A. to apply a modeling approach that couples information acquired via remote sensing, soil surveys, and local weather stations to assess where bark beetle outbreaks might first occur and why. Although there was a general downward trend in precipitation for both sites over the period between 1998 and 2010 (slope = −1.3, R2 = 0.08), interannual variability was high. Some years showed large increases followed by sharp decreases. Both sites had similar topography and fire histories, but bark beetle activity occurred earlier (circa 2000 to 2001) and more severely on one site than on the other. The initial canopy density of the two sites was also similar, with leaf area indices ranging between 1.7-2.0 m2·m−2. We wondered if the difference in bark beetle activity was related to soils that were higher in clay content at site I than at site II. To assess this possibility, we applied a process-based stand growth model (3-PG) to analyze the data and evaluate the hypotheses.

Context Many tree species will shift their distribution as the climate continues to change. To assess species’ range changes, modeling efforts often rely on climatic predictors, sometimes incorporating biotic interactions (e.g. competition or facilitation), but without integrating topographic complexity or the dynamics of disturbance and forest succession.

Objectives We investigated the role of ‘safe islands’ of establishment (‘‘microrefugia’’) in conjunction with disturbance and succession, on mediating range shifts.

Methods We simulated eight tree species and multiple disturbances across an artificial landscape designed to highlight variation in topographic complexity. Specifically,we simulated spatially explicit successional changes for a 100-year period of climate warming under different scenarios of disturbance and climate microrefugia.

Results Disturbance regimes play a major role in mediating species range changes. The effects of disturbance range from expediting range contractions for some species to facilitating colonization of new ranges for others. Microrefugia generally had a significant but smaller effect on range changes. The existence of microrefugia could enhance range persistence but implies increased environmental heterogeneity, thereby hampering migration under some disturbance regimes and for species with low dispersal capabilities. Species that gained suitable habitat due to climate change largely depended on the interaction between species life history traits, environmental heterogeneity and disturbance regimes to expand their ranges.

Conclusions Disturbance and microrefugia play a key role in determining forest range shifts during climate change. The study highlights the urgent need of including non-deterministic successional pathways into climate change projections of species distributions.