• 沒有找到結果。

Simulated big sagebrush regeneration supports predicted changes at the trailing and leading edges of distribution shifts

N/A
N/A
Protected

Academic year: 2022

Share "Simulated big sagebrush regeneration supports predicted changes at the trailing and leading edges of distribution shifts"

Copied!
31
0
0

加載中.... (立即查看全文)

全文

(1)

at the trailing and leading edges of distribution shifts

DANIELR. SCHLAEPFER,1,  KYLEA. TAYLOR,1VICTORIAE. PENNINGTON,1KELLENN. NELSON,1,2 TRACEE. MARTYN,1CAITLIN M. ROTTLER,1,2WILLIAMK. LAUENROTH,1,2ANDJOHN B. BRADFORD3

1University of Wyoming, Department of Botany, 1000 E. University Avenue, Laramie, Wyoming 82071 USA

2University of Wyoming, Program in Ecology, 1000 E. University Avenue, Laramie, Wyoming 82071 USA

3U.S. Geological Survey, Southwest Biological Science Center, 2255 N. Gemini Drive, Flagstaff, Arizona 86001 USA

Citation: Schlaepfer, D. R., K. A. Taylor, V. E. Pennington, K. N. Nelson, T. E. Martyn, C. M. Rottler, W. K. Lauenroth, and J. B. Bradford. 2015. Simulated big sagebrush regeneration supports predicted changes at the trailing and leading edges of distribution shifts. Ecosphere 6(1):3. http://dx.doi.org/10.1890/ES14-00208.1

Abstract. Many semi-arid plant communities in western North America are dominated by big sagebrush.

These ecosystems are being reduced in extent and quality due to economic development, invasive species, and climate change. These pervasive modifications have generated concern about the long-term viability of sagebrush habitat and sagebrush-obligate wildlife species (notably Greater Sage-Grouse), highlighting the need for better understanding of the future big sagebrush distribution, particularly at the species’ range margins. The leading and trailing edges of potential climate-driven distribution shifts are likely to be areas most sensitive to climate change. Although several processes contribute to distribution shifts, regeneration is a fundamental requirement, especially for species with episodic regeneration patterns, such as big sagebrush. We used a process-based regeneration model for big sagebrush to simulate potential germination and seedling survival in response to climatic and edaphic conditions. We estimated current and future regeneration under 2070–2099 CMIP5 climate conditions at trailing and leading edges that were previously identified using traditional species distribution models. Our results supported expectations of increased probability of regeneration at the leading edge and decreased probability at the trailing edge compared to current levels. Our simulations indicated that soil water dynamics at the leading edge will become more similar to the typical seasonal ecohydrological conditions observed within the current range of big sagebrush. At the trailing edge, increased winter and spring dryness represented a departure from conditions typically supportive of big sagebrush. Our results highlighted that minimum and maximum daily temperatures as well as soil water recharge and summer dry periods are important constraints for big sagebrush regeneration. We observed reliable changes in areas identified as trailing and leading edges, consistent with previous predictions.

However, we also identified potential local refugia within the trailing edge, mostly at higher elevation sites.

Decreasing regeneration probability at the trailing edge suggests that it will be difficult to preserve and/or restore big sagebrush in these areas. Conversely, increasing regeneration probability at the leading edge suggests a growing potential for conflicts in management goals between maintaining existing grasslands and croplands by preventing sagebrush expansion versus accepting a shift in plant community composition to sagebrush dominance.

Key words: Artemisia tridentata; climate change response; germination; Greater Sage-Grouse; reclamation; restoration;

seedling survival; species distribution.

Received 30 June 2014; revised 2 October 2014; accepted 3 October 2014; final version received 15 December 2014;

published 15 January 2015. Corresponding Editor: H. Epstein.

Copyright: Ó 2015 Schlaepfer et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited. http://creativecommons.org/licenses/by/3.0/

 E-mail: dschlaep@uwyo.edu

(2)

I

NTRODUCTION

Big sagebrush (Artemisia tridentata Nutall, a long-lived shrub [Shultz 2006]) dominates semi- arid plant communities across western North America (McArthur and Plummer 1978, West 1983). A large part of the economy of this region depends on livestock grazing, crop production, energy extraction, and recreation on lands potentially covered by big sagebrush. Addition- ally, sagebrush-dominated ecosystems provide important services such as wildlife habitat, nutrient and water cycling, and carbon storage (West and Young 1999). These ecosystems, however, have been reduced in distribution and quality due to economic and agricultural activity, infrastructure expansion, invasive species, and altered fire regimes (Welch 2005, Meinke et al.

2009, Davies et al. 2011, Knick et al. 2011, Manier et al. 2013).

Many sagebrush-dependent species have ex- perienced declines following recent decreases in quality and area of big sagebrush habitat, prompting increased attention to sagebrush ecosystem management practices (Rowland et al. 2006, Hanser et al. 2011). Decisions on a number of sagebrush-relevant management ac- tions are currently expected from land managers and conservationists (Davies et al. 2011, Manier et al. 2013, US Fish and Wildlife Service 2013).

These include managing cheatgrass (Bromus tectorum) invasions, addressing conifer en- croachment, and promoting habitat for at-risk species, such as Greater Sage-Grouse (Centroce- rus urophasianus, Manier et al. 2013). The most prominent is the pending final listing decision in 2015 by the U.S. Fish and Wildlife Service on the potential inclusion of Greater Sage-Grouse (Aldridge et al. 2008, Manier et al. 2013) on the Endangered Species List (US Fish and Wildlife Service 2013). Previous studies have found strong potential responses of big sagebrush to future climate scenarios (Shafer et al. 2001, Schlaepfer et al. 2012c), potentially compound- ing the impacts of current threats (Neilson et al.

2005, Bradley 2010). Important conservation planning decisions in sagebrush ecosystems, and the Greater Sage-Grouse listing decision, in particular, require an understanding of climate change impacts and information on the future distribution and habitat suitability of

target species (Glick et al. 2011, Stein et al.

2014).

Species distribution models (SDMs) are wide- ly-used to estimate climate change impacts on the suitability of climate or habitat space for a particular species (e.g., Wiens et al. 2009, Franklin et al. 2013). SDM results are useful because large-scale spatial patterns of species distributions are dominated by climatic factors (Walther et al. 2002) and potentially certain types of biotic interactions (Arau´jo and Rozenfeld 2014). Traditionally, SDMs rely on climatic variables to derive a species’ suitability distribu- tion (Franklin and Miller 2009), although mea- sures that are directly related to the physiological requirements of plants can be better predictors of species distributions (e.g., soil water availability, Rickebusch et al. 2008, Schlaepfer et al. 2012c, Piedallu et al. 2013). Both range expansions as well as contractions have been observed under recent climate change (Thomas et al. 2006, Bell et al. 2014) and most SDMs tend to predict large geographic shifts in suitable areas under future climate scenarios (e.g., Shafer et al. 2001, Walther et al. 2002).

Despite their utility, SDMs have several limi- tations that decouple results from expected distributions and restrict their ability to appro- priately inform management decisions. Some of these limitations include failure to represent key biological processes, notably migration, biotic interactions, and physiological sensitivity of different life-stages and incorporation of system- atic biases by selection of spatial data, statistical modeling techniques, and assumption of equilib- rium conditions (Guisan and Thuiller 2005, Wiens et al. 2009). For example, some studies have not observed recent large-scale range shifts forecasted by SDMs (Zhu et al. 2012) and highlight the need to scrutinize SDM predictions with independent approaches (Dawson et al.

2011). Thus, a combination of approaches and data are suggested to confirm SDM predictions:

if multiple independent lines of study come to the same conclusion about species responses to climate change scenarios, then we can be more confident in the importance of those findings (Dawson et al. 2011).

Of particular importance for understanding climate change impacts are geographic shifts near the margins of a current species’ range

(3)

(Thuiller et al. 2008). Populations persisting near these range margins are expected to be particu- larly sensitive to changing conditions (Anderson et al. 2009) and will represent the initial locations of altered species distributions. Leading edges are current unoccupied areas where a species may expand its distribution in the future; while trailing edges are areas of contraction in a species’ distribution that are currently occupied but may not be in the future (Hampe and Petit 2005, Anderson et al. 2009). Because of the transient nature of range shifts, leading and trailing edges may be strongly influenced by non-climate controls such as recruitment or mortality due to climate fluctuations and other events, which are not captured by SDMs relying on assumptions of species distributions existing in equilibrium with mean climatic conditions (Jackson et al. 2009).

Successful plant population expansion at a leading edge will necessarily involve successful regeneration. Population contraction at a trailing edge will very likely include regeneration failure in addition to increased mortality. Regenerative stages often have a narrower environmental tolerance than adult stages, particularly of long- lived species (Grubb 1977) such as big sagebrush (Schlaepfer et al. 2014b) as demonstrated by the episodic nature of recruitment in big sagebrush stands (Perryman et al. 2001). Consequently, understanding regeneration is crucial for popu- lation dynamics (Harper 1977, Fenner 2002) and responses to climate change (Jackson et al. 2009, Russell et al. 2011, Mok et al. 2012). Checking expectations about regeneration responses will thus also assess the delineation of trailing and leading edges based on SDMs.

Our overall goal was to characterize the effects of climate change on the regeneration success of big sagebrush at the trailing and leading edges that were previously predicted using SDMs (Schlaepfer et al. 2012c). Using an established process-based model that represents sagebrush germination and first-year survival, we quanti- fied the probability of conditions suitable for sagebrush regeneration under current and future climates at the leading and trailing edges. These results allowed us to compare the correspon- dence between the SDM predictions and those of a process-based regeneration model (Schlaepfer et al. 2014b). This work provides insight into the

impact of climate change on a key life stage for big sagebrush in areas likely to experience changes in sagebrush suitability.

M

ETHODS

Study area and site selection

We based our study on a framework of big sagebrush SDMs that compared predictions of habitat suitability based on either climatic or ecohydrological explanatory variables for refer- ence (1971–2000) and ensemble future (2070–

2099) conditions under SRES A2 for the western United States of America (Schlaepfer et al. 2012c).

We sampled 100 random sites from the trailing and leading edge of future big sagebrush habitat suitability (Fig. 1). A site here represents a 10 km 3 10 km cell as defined by Schlaepfer et al.

(2012c). Our point model simulation defines the entire cell by summarizing conditions for the extent of that cell. We defined the trailing/leading edges as those areas where climatic and ecohy- drological SDMs agreed that habitat suitability will change from either suitable to unsuitable or from unsuitable to suitable in the future (cf. Fig. 3 in Schlaepfer et al. 2012c). To simplify presenta- tion of the results, we excluded suitability changes based along elevational gradients even though these are also potentially important changes. The latter criterion further narrowed the geographic extent of our sample population:

(1) trailing edge sites were drawn from E of 1178 longitude to exclude the White Mountains and the Sierra Nevada, from S of 408 latitude and from elevations lower than 2137 m to exclude mountain ranges in northern Utah and southern Rocky Mountains; (2) leading edge sites were drawn from E of 1108 longitude to exclude the central Rocky Mountains and from N of 458 latitude to exclude the southern Rocky Moun- tains (Fig. 1). We acknowledge that our study area definition implies a more homogenous leading edge, whereas sites from the trailing edge are expected to show a larger spread due to a larger geographic extent and a higher topo- graphic richness.

Process-based regeneration model

To assess regeneration probabilities at each of the 200 sites we used a process-based model of sagebrush regeneration, which simulates suitable

(4)

conditions for germination and seedling survival in response to soil conditions and weather (Schlaepfer et al. 2014a). The model was trained and tested for big sagebrush with a total of 1435 site-years of observations and explained 74% of variability of numbers of years with successful regeneration and achieved 60% overall accuracy predicting yearly regeneration success/failure (Schlaepfer et al. 2014a). The model uses daily forcing data of air and top-soil temperature, water potential in soil layers, and snow cover to simulate favorable periods for germination, time to germination, and germination success. Based on germination success, the model simulates days with seedling emergence, favorable and lethal periods for seedling growth, root elonga- tion of seedlings, and first-season seedling mortality or survival.

Daily forcing data were provided by SOIL- WAT, a daily-time step, multiple soil layer,

ecosystem water balance simulation model (Par- ton 1978, Sala et al. 1992), which has been adapted and tested for sagebrush ecosystems (Schlaepfer et al. 2012a, b, Bradford et al.

2014a, b). The first year of each 32-year simula- tion period was excluded to limit effects of initial conditions. We used cell-specific daily historical weather data for 1979–2010 (a 1/8-degree grid- ded product, Maurer et al. 2002), soil information (Miller and White 1998), and estimates of vegetation composition, biomass, rooting distri- bution, and seasonal phenology (details in Bradford et al. 2014b) to run SOILWAT. Details of data preparation for the 10-km grid are provided in Schlaepfer et al. (2012c). There were no obvious relationships between soil texture (sand and clay content) and climate among the sampled sites (Appendix: Fig. A2).

Fig. 1. Geographic map of sample sites at the leading (b, black dots) and trailing edge (c, black dots) based on published SDMs of big sagebrush (gray, current predicted suitability, red, decrease in suitability under SRES A2 for 2070–2099, blue, increase in suitability, Schlaepfer et al. 2012c). The combined area of no change (gray) and trailing edge (red) indicate the extent of suitable habitat for current (1970–1999) conditions. Black lines indicate U.S. state boundaries; dashed boxes indicate study area (a; see Methods for details). Map is Albers equal area projection.

(5)

Climate scenarios

We extracted monthly time-series for 2069–

2099 of 1/8-degree downscaled and bias-correct- ed products of the fifth phase of the climate model intercomparison project (CMIP5; Taylor et al. 2012) of 16 global circulation models (GCMs) for two representative concentration pathways (RCPs; Moss et al. 2010), RCP4.5 and RCP8.5, from the ‘‘Downscaled CMIP3 and CMIP5 Climate and Hydrology Projections’’

a r c h i v e a t h t t p : / / g d o - d c p . u c l l n l . o r g / downscaled_cmip_projections/ (data accessed on February 3, 2014, Maurer et al. 2007). We applied a hybrid-delta downscaling approach to combine historic daily data with monthly future predictions to obtain future daily forcing (Ham- let et al. 2010, Dickerson-Lange and Mitchell 2014). Because we did not include all GCMs that participated in CMIP5, we selected 16 GCMs based on a family tree of GCMs and their evaluation against observed data (Knutti et al.

2013). Specifically, we chose the best performing GCM (based on Fig. 3 in Knutti et al. 2013) from each group from the family tree that we cut into

16 branches (Fig. 1b in Knutti et al. 2013). This approach selected the most independent and best performing subset of GCMs (Appendix:

Table A1).

Analysis of simulation output

Regeneration probabilities were calculated as the fraction of years in each 31-year simulation period with germination and seedling success.

Because the regeneration model is deterministic, we performed an evaluation of model results, but no hypothesis testing (Simpson et al. 1997, White et al. 2014). We examined relationships between explanatory variables (Appendix: Table A2), regeneration probabilities, and changes of regeneration probabilities between future and current climate scenarios. We summarized rela- tionships with locally weighted polynomial regression smoothers (R function ‘lowess’) and the (non-linear) Brownian distance correlation (Szekely and Rizzo 2009) using 2000 bootstrap replicates as implemented in the energy R package (Rizzo and Szekely 2014) with Bonfer- roni adjustment.

We selected 21 variables to explain patterns in regeneration responses based on their impor- tance in our previous work on big sagebrush (Appendix: Table A2; Schlaepfer et al. 2012a, b, c, Schlaepfer et al. 2014a, b). To reduce the number of variables for further analysis, we performed a principal component analysis by singular value decomposition (R function ‘prcomp’ with rota- tion, scaling, and centering) of the 21 variables under current climate conditions for all 200 simulated sites. We first selected those variables that are more proximate to big sagebrush biology than overall climate and soil variables. Then we chose the least correlated variables, i.e., those with the strongest and most unambiguous loadings on the first five principal components (Appendix: Table A2 and Fig. A1). We used R version 3.1.0 (R Development Core Team 2014) for all simulations and analyses.

R

ESULTS

Climate and soil water

Mean annual precipitation (MAP) and mean annual temperature (MAT) climate spaces did not overlap between trailing and leading edges under current climate (Fig. 2). Because of the Fig. 2. Climate space of mean annual precipitation

(MAP) and mean annual temperature (MAT) of sample sites (dots) and 90% data cloud (lines) at the trailing (red hues) and leading edge (blue hues) under current climate 1980–2010 (fat dots, bold lines) and future climate conditions (small dots, thin lines) of 16 GCMs with RCP8.5 for 2070–2099. See Appendix: Fig.

A3 for RCP4.5 data.

(6)

similar overall pattern and direction of trends among the simulations driven by RCP4.5 and RCP8.5, we focus here on RCP8.5 (all results based on RCP4.5 are available in the appendices).

On average, the current climate of the trailing edge was 20 mm/yr drier and 4.28C warmer than at the leading edge. The spread among sites in MAP at the trailing edge encompassed the range at the leading edge, whereas sites at the trailing edge were consistently warmer than at the

leading edge (Table 1). With future scenario RCP8.5, leading and trailing edges both became 13% wetter while the spread among sites increased by 23% and 16%, respectively. For the leading and trailing edges, MAT increased by 5.48C and 5.28C and the spread decreased by 2%

and 3%, respectively (Table 1 and Fig. 2).

Seasonal patterns in changes at both edges included greater warming trends during winter and late summer than for the rest of the year. At Fig. 3. Changes of mean daily volumetric water content (VWC) in four aggregated soil layers between future climate conditions of 16 GCMs with RCP8.5 for 2070–2099 (black, median among sites; gray, range of all sites) and current 1980–2010 climate (red) for the trailing (a, c, e, g) and leading edge (b, d, f, h). Positive/negative values indicate an increase/decrease in VWC under future scenarios compared to current climate. See Appendix:

Fig. A6 for RCP4.5 data.

(7)

the trailing edge, increased precipitation was predicted by most GCMs during winter and late summer and a drying during spring/early sum- mer. At the leading edge, increased precipitation was predicted during late winter/early spring and a drying during summer (Appendix: Fig.

A4).

Seasonal dynamics of soil water in current habitats of big sagebrush at the trailing edge are characterized by winter to late spring recharge of the soil profile followed by a dry summer period (Appendix: Fig. A5; Schlaepfer et al. 2012b). At the leading edge, winters are dry and recharge is incomplete for deep soils under current condi- tions (Appendix: Fig. A5). Future climate condi- tions of most GCMs at the trailing edge increased volumetric water content (VWC) during winter and decreased during early to late spring with some delay at deeper soils; VWC was unchanged during summer periods (Fig. 3). Soil water dynamics kept the typical seasonal pattern albeit with an earlier onset of the dry warm period. At the leading edge, VWC increased in soils between 0–60 cm depth for most GCMs during fall to late spring while VWC of surface soils increased during winter/spring and decreased during summer (Fig. 3). At greater depths, the leading edge experienced only modest change with the exception of increased VWC during late spring to early summer.

Regeneration potential of big sagebrush

Under current conditions, germination prob- ability at both the leading and trailing edges was high, with a median and adjusted median absolute deviation (MAD) of 1.00 6 0.00 among

sites (Fig. 4). By contrast, the probability of seedling success was 0.39 6 0.19 at the trailing edge and 0.26 6 0.10 at the leading edge. Future climatic conditions at the trailing edge de- creased median probabilities of germination by

0.13 6 0.08 among GCMs (all but two GCMs agreed on the direction) and seedling success by

0.10 6 0.10 (all but four GCMs agreed). Future climatic conditions at the leading edge did not impact germination probability (0.00 6 0.00) but increased median probabilities of seedling suc- cess by 0.10 6 0.10 (all but one GCM agreed;

Fig. 4).

We found strong relationships between ex- planatory variables and germination or seedling success at both the leading and trailing edges (Fig. 5). Snowfree days with minimum temper- atures of less than 98C, while generally not related to germination, correlated negatively with probabilities of seedling success at both edges (Fig. 5c). For future scenarios, sites with a high number of such cold days also showed an increase in seedling success, whereas sites with a smaller number showed a decrease (Fig. 5d). Hot days with maximum temperatures of .348C were negatively related to germination probabil- ities at the trailing edge (Fig. 5e). This variable was also negatively related to seedling success at the trailing edge, but positively at the leading edge albeit with overall fewer hot days (Fig. 5g).

For future scenarios, sites with a high number of hot days also displayed a decrease in seedling success, whereas sites with a smaller number displayed an increase (Fig. 5h). Relative maxi- mum soil water recharge showed a strong positive relationship with all responses both Table 1. Climatic conditions of sample sites at the trailing and leading edges for current 1980–2010 and future conditions of 16 GCMs with RCP4.5 and RCP8.5 for 2070–2099. Percentiles are means among scenarios for 5%

and 95% of values.

Edge Condition

Mean annual precipitation (mm) Mean annual temperature (C)

Mean 5% 95%

Ratio of means 

Ratio of

90%à Mean 5% 95%

Difference

of means§ Ratio of 90%à

Trailing edge Current 309 181 442 1.00 1.00 9.84 7.75 12.15 0.00 1.00

RCP4.5 340 203 498 1.10 1.13 12.73 10.72 15.04 2.89 0.98

RCP8.5 348 208 511 1.13 1.16 15.06 13.07 17.35 5.22 0.97

Leading edge Current 329 295 356 1.00 1.00 5.67 4.40 7.10 0.00 1.00

RCP4.5 357 322 390 1.09 1.13 8.71 7.47 10.12 3.05 0.98

RCP8.5 372 333 407 1.13 1.23 11.10 9.86 12.50 5.44 0.98

 Ratio of means ¼ mean(future)/mean(current).

àRatio of 90% ¼ spread among sites ¼ [95%  5% of future]/[95%  5% of current].

§ Difference of means ¼ mean(future)  mean(current).

(8)

under current and future scenarios at the trailing edge, but few at the leading edge (Fig. 5i–l). The timing of maximum recharge mostly related to increased probabilities at the trailing edge;

relationships at the leading edge were less clear (Fig. 5m–p). Duration of the longest drought period during the growing season negatively correlated to germination probabilities for dura- tions of .100–150 d depending on soil depth (Fig. 5q, r, u, v). However, seedling success was primarily negatively related with drought dura- tion with a clear decrease after 50–75 d (Fig.

5s, w). Future conditions increased seedling success at the trailing edge for sites with the shortest drought periods at 0–20 cm soil depth, but not for sites with longer drought periods (Fig. 5t). At the leading edge, future climates influenced seedling success via drought periods primarily at greater depths (20–150 cm) with increases for sites with drought periods of ,100 d (Fig. 5x).

D

ISCUSSION

Regeneration at trailing and leading edges

Big sagebrush is a long-lived species (Ferguson 1964) that, under recent climate conditions, has persisted regionally with episodic recruitment (reviewed in Schlaepfer et al. 2014b). For instance, recruitment probability in Wyoming, an area with abundant big sagebrush, has been reported to be between 0.4–0.6 yr1(Perryman et al. 2001). Our results of seedling success under current conditions at the trailing edge were comparable (;0.4 yr1), whereas they were one- third lower at the leading edge, where currently no big sagebrush occurs. Our simulations under future conditions reversed this pattern: in the future at the leading edge, probability of seedling success was close to the values found in Wyoming, whereas at the trailing edge, they decreased to a level similar to what we found under current conditions at the leading edge.

These values suggest that regeneration of big Fig. 4. Potential regeneration probabilities (median, interquartile range, whiskers include all data) of germination and seedling success under current 1980–2010 climate and future climate conditions of 16 GCMs with RCP8.5 for 2070–2099 for the trailing (a, c; red hues) and leading edge (b, d; blue hues). Horizontal lines indicate the median of current conditions. See Appendix: Fig. A7 for RCP4.5 data.

(9)

sagebrush will be able to expand populations at the leading edge in the future; conversely, probability of regeneration will decrease at the trailing edge below what is currently observed, presumably limiting the ability of sagebrush to recover from disturbances. Our results suggest that regeneration may play an important role in defining the trailing and leading edge for big sagebrush.

We found strong relationships between regen- eration success and temperature and availability of soil water. We described relationships in our results using a set of variables based on previous research that attempted to describe physiological constraints on big sagebrush regeneration such as temperature or drought periods leading to direct mortality of embryos or seedlings (Appen- dix: Table A1). Therefore, we argue that the patterns in our results support (multi-) causal relationships. For instance, we found support that minimum and maximum daily temperatures as well as soil water recharge and summer dry periods are important constraints for big sage- brush regeneration. Seasonal dynamics of soil water in habitats that support big sagebrush are characterized by winter to late spring recharge of the soil profile followed by a dry summer period (Schlaepfer et al. 2012b). Soil water dynamics at the leading edge in our future simulations became more similar to typical current seasonal patterns for big sagebrush ecosystems; at the trailing edge, the increased winter/spring dry- ness represented a departure from conditions typically supportive of big sagebrush (Appendix:

Fig. A5). These changes in seasonal soil water dynamics correspond with expectations of in- creased and/or decreased suitability of big sagebrush habitat as well as regeneration pat- terns.

Even though we selected variables describing physiological constraints of big sagebrush, our results represented aggregation across years, which demonstrated only a portion of the reasons for annual success/failure because annual probabilities are the product of day-to-day survival/mortality dynamics by the model. For instance, the number of very hot days at the trailing edge was negatively related to regener- ation probability (as expected), but positively at the leading edge, possibly, because hot days are related to greater growing season length. Never-

theless, our results also correlate strongly with surrogate climate and topographic variables (Appendix: Fig. A8; Franklin and Miller 2009).

For instance, more available water increased regeneration success, be it measured as recharge, MAP, or aridity. Such climate and topographic variables can provide a summary of our insights into the expected regeneration patterns within the leading or trailing edges. For instance, elevation, which is related negatively to temper- ature and positively to precipitation, provides an integrative example (Appendix: Fig. A8c, d):

elevation correlates positively with regeneration at the trailing edge (where cooler, wetter high- elevation sites may become refugia), while elevation correlates negatively with regeneration at the leading edge (where the growing season is reduced by snow and colder/wetter temperatures at high elevation sites; Appendix: Fig. A8aa;

Schlaepfer et al. 2012a).

Regeneration modeling supports changes in habitat suitability from SDMs

Our results confirm previous SDM predictions by showing consistent changes in areas identified as trailing and leading edges by previous SDMs.

Several recent studies have reported differences in climate responses among life stages support- ing the expected higher sensitivity of regenera- tion vs. adults to climate change (Jackson et al.

2009, Russell et al. 2011). For instance, seedlings of tree species in forests of the western United States demonstrated mostly more severe range contractions than adults under recent climate change (Bell et al. 2014). In southern Spain, seedlings of cork oaks may profit from warmer and wetter winters whereas adults may only do so in sandy soils (Iba´˜nez et al. 2014). Low forests on subantartic Campbell Island (New Zealand) experienced decreased adult growth under warm and wet winters in the last 50 years, whereas regeneration is increased in warm and dry winters (Harsch et al. 2014). Similarly, regenera- tion processes of five eucalypt species in south- eastern Australia were linked to SDMs and this combined simulation resulted in divergent re- sponses among species and wet vs. dry ecosys- tems (Mok et al. 2012). These studies suggest that sensitivity of juvenile life stages to climate change are not easily generalized across species and ecosystems; they can, however, be markedly

(10)

different from adult life stages. While SDMs relate climate variables to a prediction of adult presence/absence of a species, our regeneration model relates weather and soil variables to a particular life history stage of recognized impor- tance for big sagebrush. Therefore, our study contributes to the notion that the predicted shifts

in habitat suitability of big sagebrush may indeed indicate geographic range shifts.

Studying the causes of range limits allows us to better understand a variety of important ecological topics including ecological niches, evolutionary processes, (meta-) population dy- namics, biodiversity patterns, and conservation Fig. 5. Relationships between potential regeneration probabilities of germination and seedling success and change and selected explanatory variables (Appendix: Fig. A1 and Table A2) under current 1980–2010 climate (bold lines/dots) and future climate conditions of 16 GCMs with RCP8.5 for 2070–2099 (simple lines/small dots) for the trailing (red) and leading edge (blue). Locally-weighted polynomial regression (lowess) lines indicate that the Brownian distance correlation was significant after Bonferroni adjustment (counts displayed in panel header).

Bars at top of each panel indicate range of current conditions; bars at bottom indicate range of minimum/

maximum among 5–95% of each GCM. See Appendix: Fig. A9 for RCP4.5 data.

(11)

and management issues (e.g., Anderson et al.

2009, Angert 2009). Cahill et al. (2014) reviewed the causes of range limits at low elevations and latitudes and found much stronger support for abiotic factors such as temperature and moisture than for biotic factors such as competition. While our study did not compare the relative impor- tance of biotic vs. abiotic factors, our simulation confirmed that temperature extremes and a lack of soil moisture are strongly related to regener- ation at the trailing edge. Similarly, D’Odorico et al. (2012) synthesized that shrubland replacing grassland (e.g., what we simulated as the leading

edge; often also called ‘shrub encroachment’) is driven by conditions favoring shrub dominance such as overgrazing and climate changes (e.g., reduced limitation of shrubs by increasing lower temperatures). Once those changes in plant functional types have occurred, they can be sustained by the absence of positive fire-grass and negative erosion-grass feedbacks and by the presence of positive microclimate-shrub and small-mammal-shrub feedbacks (D’Odorico et al. 2012). Because the trailing and leading edges are likely limited primarily by abiotic factors that are directly and immediately impacted by chang- Fig. 5. Continued.

(12)

es in climate, regeneration and big sagebrush distribution at the trailing edge may be especially sensitive to climate change.

Limitation on interference

Our study focused on a comparison between responses of simulations of physiological pro- cesses constraining germination and seedling survival to climate change with responses of correlative models of habitat suitability. Howev- er, our models have important limitations and regeneration is only one process, albeit a neces- sary one, contributing to species distributions, which are emerging patterns from the entire suite of population dynamic, ecological, and evolu- tionary processes in combination with land use and management actions (Wiens et al. 1985, Caughley et al. 1988, Holt 2003, Angert 2009, Sexton et al. 2009). Because big sagebrush is a long-lived species with episodic recruitment, field-based experiments and observations may not address our objectives, which simulation models can attempt to answer.

Our study examined the impacts of climate change on regeneration as a fundamental process contributing to distributional shifts. In the future, not only climate will be different, but CO2

concentration is increasing, disturbance regimes are altered, management actions and land use will be adjusted, and other species will respond to global change as well leading to altered competitive interactions in addition to interac- tions among such factors (Shugart and Wood- ward 2011, Moran and Alexander 2014, Piontek et al. 2014). Global change acts both on adult as well as on regenerative stages. However, a plant population may only expand in distribution and can only survive over the long term if regener- ation success is large enough.

Our study did not account for effects of disturbances and adult mortality. Big sagebrush mortality has been observed following insect outbreaks (Haws et al. 1990), fungal pathogens (Nelson and Krebill 1981), freezing damage associated with the absence of snow cover (Hanson et al. 1982), and other causes (Ca´rdenas et al. 1997). Such causes of mortality will likely become more important with climate change (Wheeler et al. 2014) and mortality due to fire in the sagebrush region will increase as well (West- erling et al. 2006, Westerling et al. 2011). Our

study also did not account for seed production.

Generally, seed production rates are high in big sagebrush, but they can vary drastically from year to year and with climate (reviewed in Schlaepfer et al. 2014b). Observed migration distances of big sagebrush are short because seed dispersal is limited from a few meters up to about 30 m (reviewed in Schlaepfer et al. 2014b).

Slow recovery rates in large burn areas or restoration efforts have been attributed to slow migration rates in combination with episodic recruitment (e.g., Ziegenhagen and Miller 2009, Nelson et al. 2014). Slow migration will inhibit a fast tracking of changing climate by big sage- brush, particularly over the very large spatial scales over which habitat suitability is predicted to change (Shafer et al. 2001, Schlaepfer et al.

2012c). Our study also did not incorporate effects of biotic interactions such as competition. Big sagebrush can be outcompeted by the exotic annual grass cheatgrass partially because cheat- grass increases fire frequency (Knapp 1996, Bradford and Lauenroth 2006, Balch et al. 2013).

Pinyon-juniper woodland may also encroach on sagebrush dominated areas in the future, partic- ularly at lower elevations (Bradley 2010). Fur- thermore, this study focused on latitudinal and not elevational distribution shifts, this is why we sampled sites from a 10-km grid; consequently, this study cannot draw conclusions about short- distance and upslope movement of vegetation.

These and others factors may help to explain why our regeneration model was not able to explain 26% of variation among year-to-year regeneration success during model testing (Schlaepfer et al. 2014a). This error could also have propagated to the results of this study;

however, the results and conclusions should be robust against such variation. Consequently, by not including biotic interactions and other factors discussed above in our study, it is possible that we overestimated the future distribution of big sagebrush. This will need to be addressed by future research.

Management-relevant conclusions for big sagebrush-dominated ecosystems

Our simulation results support previous SDM predictions about the distribution of future big sagebrush habitat suitability; furthermore, they provide insights into the contribution of regen-

(13)

eration to range shifts. Current literature related to the management of big sagebrush ecosystems generally does not incorporate climate change.

This is at least partially due to the previous degree of uncertainty and the long-term nature of climate change impact studies (Manier et al.

2013). While many management recommenda- tions concern vegetation treatments and are designed for short-term success (Meinke et al.

2009, Davies et al. 2011, Madsen et al. 2013), many management strategies require long-term investments, which can have enormous financial consequences. Such strategies include controlling landscapes against anthropogenic development (Davies et al. 2011) and identifying management zones for Greater Sage-Grouse conservation (US Fish and Wildlife Service 2013). In the past, priority areas for big sagebrush restoration have been identified based on criteria including favorable conditions for revegetation, wildlife habitat connectivity, important Greater Sage- Grouse areas, and cheatgrass invasions (Meinke et al. 2009). For instance, Meinke et al. (2009) identified much of Wyoming big sagebrush in Nevada and Utah as high priority areas for restoration. Our results indicate, however, that much of that area, which lies in what we identified as the trailing edge, may be unsuitable for longer-term goals of maintaining it as sagebrush habitat. This finding is supported by the establishment failure after seeding in drier and low-elevation burns in the Great Basin (Knutson et al. 2014). Therefore, long-term management actions need to include a climate change vulnerability assessment (Glick et al.

2011, Stein et al. 2014). Otherwise, potentially large, investments may be made in areas where big sagebrush will no longer be present in the future.

At the trailing edge, our results suggest a decrease in habitat suitability and a decrease in regeneration probability under most future sce- narios. A management plan that relies upon long-term maintenance of big sagebrush domi- nated ecosystems in the trailing edge is more likely to fail than a plan built upon maintenance of sites with stable climatic suitability. However, we also identified several variables that can account for variation in current and future regeneration within the trailing edge. These variables (for which elevation provides a reason-

able surrogate) may be useful for identifying small local areas that will remain suitable for big sagebrush. These potential refugia are areas with high soil water recharge, short drought periods during the growing season, and few hot days (Fig. 5), i.e., mostly sites at higher elevations. This finding is supported by higher chances of successful seeding at higher elevations, whereas low elevations experience establishment failure, after fire in the Great Basin (Knutson et al. 2014).

However, a scenario where low-elevations are invaded by exotic grasses that promote fire spread also into higher elevations could hamper big sagebrush at these potential refugia. Such areas may benefit from management actions that promote seedling survival, including prevention of fires, conifer encroachment, and exotic annual grass invasion. Maintaining sagebrush-dominat- ed ecosystems within the trailing edge may also benefit from targeting restoration at areas that are expected to be climatically suitable, although restoration of areas invaded by exotic annual grasses remains a challenge even under current climatic conditions (Davies et al. 2011). An alternative may be management to encourage native non-fire promoting shrubs and perennial grasses that will experience a distributional shift and will find their leading edges where big sagebrush’s trailing edge is.

At the leading edge, our results suggest an increase in habitat suitability and an increase in regeneration probability under most future sce- narios. Action plans at the leading edge may need to resolve potential conflicts between management goals to maintain existing grass- lands and croplands and prevent sagebrush expansion and those to accept and promote new areas with big sagebrush dominated plant communities (cf. D’Odorico et al. 2012). Removal of big sagebrush has been practiced extensively and much practical knowledge has been accu- mulated (e.g., Torell et al. 2005, Bastian et al.

2012, D’Odorico et al. 2012). Promoting regener- ation and restoring big sagebrush is currently difficult (Lambert 2005, Shaw et al. 2005, Davies et al. 2013, McAdoo et al. 2013, Knutson et al.

2014). Future climate conditions may not only promote big sagebrush in areas at the leading edge, but also potential competitive species (e.g., conifers) and agricultural practices. Conifer encroachment in these new sagebrush areas

(14)

may become a concern, but studies indicate that sagebrush can be managed at moister sites with prescribed burning and seeding (Davies et al.

2011, Davies et al. 2014, Nelson et al. 2014). Even though accepting sagebrush advancement in the future may not seem to require management actions in our present time, delineating land at the leading edge as future big sagebrush habitat to support sagebrush-dependent species may eventually warrant more detailed consideration.

C

ONCLUSIONS

Our study successfully applied a process-based regeneration model to evaluate future scenarios of big sagebrush germination and seedling survival. We used a set of 16 global circulation models with two representative concentration pathways to simulate changes in regeneration probabilities at the end of this century. Results among individual global circulation models agreed, but with some important variation depending on the region. Our results generally confirmed expectations that regeneration at the trailing edge would decrease and increase at the leading edge. Furthermore, our results supported earlier predictions of changes in habitat suitabil- ity based on correlative species distribution models. Shifts in the distribution of big sage- brush likely will involve changes in plant functional dominance between shrubs and grass- es, which can have important ecohydrological consequences, mostly in dry areas and where disturbances reduce biomass (Bradford et al., 2014b). Our study yielded important implications for long-term management of big sagebrush habitats, including identification of processes contributing to leading and trailing edges and recognition of potential refugia within the overall trailing edge.

A

CKNOWLEDGMENTS

We thank the U.S. Geological Survey North Central Climate Science Center (grant G12AC20504) and the University of Wyoming. Any use of trade, product, or firm names is for descriptive purposes only and does not imply endorsement by the U.S. Government. We acknowledge the World Climate Research Pro- gramme’s Working Group on Coupled Modelling, which is responsible for CMIP, and we thank the climate modeling groups (listed in Appendix Table A1

of this paper) for producing and making available their model output. For CMIP the U.S. Department of Energy’s Program for Climate Model Diagnosis and Intercomparison provides coordinating support and led development of software infrastructure in partner- ship with the Global Organization for Earth System Science Portals.

L

ITERATURE

C

ITED

Aldridge, C. L., S. E. Nielsen, H. L. Beyer, M. S. Boyce, J. W. Connelly, S. T. Knick, and M. A. Schroeder.

2008. Range-wide patterns of Greater Sage-Grouse persistence. Diversity and Distributions 14:983–

994.

Anderson, B. J., H. R. Akcakaya, M. B. Araujo, D. A.

Fordham, E. Martinez-Meyer, W. Thuiller, and B. W. Brook. 2009. Dynamics of range margins for metapopulations under climate change. Proceed- ings of the Royal Society B 276:1415–1420.

Angert, A. L. 2009. The niche, limits to species’

distributions, and spatiotemporal variation in demography across the elevation ranges of two monkeyflowers. Proceedings of the National Acad- emy of Sciences 106:19693–19698.

Arau´jo, M. B., and A. Rozenfeld. 2014. The geographic scaling of biotic interactions. Ecography 37:406–

415.

Balch, J. K., B. A. Bradley, C. M. D’Antonio, and J.

Go´mez-Dans. 2013. Introduced annual grass in- creases regional fire activity across the arid western USA (1980–2009). Global Change Biology 19:173–

183.

Bastian, C. T., D. E. Peck, M. A. Smith, and J. J. Jacobs.

2012. Sagebrush control: at what canopy cover is it economically justified? Rangeland Ecology & Man- agement 65:351–361.

Bell, D. M., J. B. Bradford, and W. K. Lauenroth. 2014.

Early indicators of change: Divergent climate envelopes between tree life stages imply range shifts in the western United States. Global Ecology and Biogeography 23:168–180.

Bradford, J. B., and W. K. Lauenroth. 2006. Controls over invasion of Bromus tectorum: the importance of climate, soil, disturbance and seed availability.

Journal of Vegetation Science 17:693–704.

Bradford, J. B., D. R. Schlaepfer, and W. K. Lauenroth.

2014a. Ecohydrology of adjacent sagebrush and lodgepole pine ecosystems: The consequences of climate change and disturbance. Ecosystems 17:590–605.

Bradford, J. B., D. R. Schlaepfer, and W. K. Lauenroth.

2014b. Shifts in plant functional types have time- dependent and regionally variable impacts on dryland ecosystem water balance. Journal of Ecology 102:1408–1418.

Bradley, B. A. 2010. Assessing ecosystem threats from

(15)

global and regional change: hierarchical modeling of risk to sagebrush ecosystems from climate change, land use and invasive species in Nevada, USA. Ecography 33:198–208.

Cahill, A. E., M. E. Aiello-Lammens, M. Caitlin Fisher- Reid, X. Hua, C. J. Karanewsky, H. Y. Ryu, G. C.

Sbeglia, F. Spagnolo, J. B. Waldron, J. J. Wiens, and W. Daniel Kissling. 2014. Causes of warm-edge range limits: systematic review, proximate factors and implications for climate change. Journal of Biogeography 41:429–442.

Ca´rdenas, A., J. Lewinsohn, C. Auger, J. L. Downs, L. L. Cadwell, and R. Burrows. 1997. Characteriza- tion of a sagebrush (Artemisia tridentata ssp.

wyomingensis) die-off on the Handford Site.

PNNL-11700. Pacific Northwest National Labora- tory, Richland, Washington, USA.

Caughley, G., D. Grice, R. Barker, and B. Brown. 1988.

The edge of the range. Journal of Animal Ecology 57:771–785.

D’Odorico, P., G. S. Okin, and B. T. Bestelmeyer. 2012.

A synthetic review of feedbacks and drivers of shrub encroachment in arid grasslands. Ecohydrol- ogy 5:520–530.

Davies, K. W., J. D. Bates, M. D. Madsen, and A. M.

Nafus. 2014. Restoration of mountain big sage- brush steppe following prescribed burning to control western juniper. Environmental Manage- ment 53:1015–1022.

Davies, K. W., C. S. Boyd, J. L. Beck, J. D. Bates, T. J.

Svejcar, and M. A. Gregg. 2011. Saving the sagebrush sea: an ecosystem conservation plan for big sagebrush plant communities. Biological Con- servation 144:2573–2584.

Davies, K. W., C. S. Boyd, and A. M. Nafus. 2013.

Restoring the sagebrush component in crested wheatgrass–dominated communities. Rangeland Ecology & Management 66:472–478.

Dawson, T. P., S. T. Jackson, J. I. House, I. C. Prentice, and G. M. Mace. 2011. Beyond predictions:

biodiversity conservation in a changing climate.

Science 332:53–58.

Dickerson-Lange, S. E., and R. Mitchell. 2014. Model- ing the effects of climate change projections on streamflow in the Nooksack River basin, North- west Washington. Hydrological Processes 28:5236–

5250.

Fenner, M., editor. 2002. Seeds: The ecology of regeneration in plant communities. Second edition.

CABI Publishing, Wallingford, UK.

Ferguson, C. W. 1964. Annual rings in big sagebrush Artemisia tridentata. University of Arizona Press, Tuscon, Arizona, USA.

Franklin, J., F. W. Davis, M. Ikegami, A. D. Syphard, L. E. Flint, A. L. Flint, and L. Hannah. 2013.

Modeling plant species distributions under future climates: how fine scale do climate projections need

to be? Global Change Biology 19:473–483.

Franklin, J., and J. A. Miller. 2009. Mapping species distributions. Cambridge University Press, Cam- bridge, UK.

Glick, P., B. A. Stein, and N. A. Edelson. 2011. Scanning the conservation horizon: A guide to climate change vulnerability assessment. National Wildlife Federation, Washington, D.C., USA.

Grubb, P. J. 1977. The maintenance of species-richness in plant communities: the importance of the regeneration niche. Biological Reviews 52:107–145.

Guisan, A., and W. Thuiller. 2005. Predicting species distribution: offering more than simple habitat models. Ecology Letters 8:993–1009.

Hamlet, A. F., E. P. Salathe´, and P. Carrasco. 2010.

Statistical downscaling techniques for global cli- mate model simulations of temperature and pre- cipitation with application to water resources planning studies. Chapter 4. Final Report for the Columbia Basin Climate Change Scenarios Project.

Climate Impacts Group, Center for Science in the Earth System, Joint Institute for the Study of the Atmosphere and Ocean, University of Washington, Seattle, Washington, USA.

Hampe, A., and R. J. Petit. 2005. Conserving biodiver- sity under climate change: the rear edge matters.

Ecology Letters 8:461–467.

Hanser, S. E., M. Leu, S. T. Knick, and C. Aldridge, editors. 2011. Sagebrush ecosystem conservation and management: ecoregional assessment tools and models for the Wyoming Basin. Allen Press, Lawrence, Kansas, USA.

Hanson, C. L., C. W. Johnson, and W. J. Ross. 1982.

Foliage mortality of mountain big sagebrush [Artemisia tridentata subsp. vaseyana] in southwest- ern Idaho during the winter of 1976-77. Journal of Range Management 35:142–145.

Harper, J. L. 1977. Population biology of plants.

Academic Press, New York, New York, USA.

Harsch, M. A., M. S. McGlone, and J. M. Wilmshurst.

2014. Winter climate limits subantarctic low forest growth and establishment. PLoS ONE 9:e93241.

Haws, B. A., G. E. Bohart, C. R. Nelson, and D. L.

Nelson. 1990. Insects and shrub dieoff in western states: 1986-89 survey results. Pages 127–151 in E. D. McArthur, E. M. Romney, S. D. Smith, and P. T. Tueller, editors. Proceedings of the Symposium on cheatgrass invasion, shrub die-off, and other aspects of shrub biology and management. Las Vegas, Nevada, April 5–7, 1989. USDA Forest Service, Intermountain Research Station, Ogden, Utah, USA.

Holt, R. D. 2003. On the evolutionary ecology of species’ ranges. Evolutionary Ecology Research 5:159–178.

Iba´˜nez, B., I. Iba´˜nez, L. Go´mez-Aparicio, P. Ruiz- Benito, L. V. Garcı´a, T. Mara˜no´n, and W. Thuiller.

(16)

2014. Contrasting effects of climate change along life stages of a dominant tree species: the impor- tance of soil-climate interactions. Diversity and Distributions 20:872–883.

Jackson, S. T., J. L. Betancourt, R. K. Booth, and S. T.

Gray. 2009. Ecology and the ratchet of events:

climate variability, niche dimensions, and species distributions. Proceedings of the National Acade- my of Sciences USA 106:19685–19692.

Knapp, P. A. 1996. Cheatgrass (Bromus tectorum L.) dominance in the Great Basin Desert: History, persistence, and influences to human activities.

Global Environmental Change 6:37–52.

Knick, S. T., J. W. Connelly, and Cooper Ornithological Society. 2011. Greater Sage-Grouse: ecology and conservation of a landscape species and its habi- tats. University of California Press, Berkeley, California, USA.

Knutson, K. C., D. A. Pyke, T. A. Wirth, R. S. Arkle, D. S. Pilliod, M. L. Brooks, J. C. Chambers, and J. B.

Grace. 2014. Long-term effects of seeding after wildfire on vegetation in Great Basin shrubland ecosystems. Journal of Applied Ecology 51:1414–

1424.

Knutti, R., D. Masson, and A. Gettelman. 2013. Climate model genealogy: Generation CMIP5 and how we got there. Geophysical Research Letters 40:1194–

1199.

Lambert, S. M. 2005. Seeding considerations in restoring big sagebrush habitat. Pages 75–80 in N. L. Shawn, M. Pellant, and S. B. Monsen, editors.

Sage-Grouse habitat restoration symposium pro- ceedings; 2001 June 4–7, Boise, Idaho. Proceedings RMRS-P-38. USDA Forest Service, Rocky Mountain Research Station, Fort Collins, Colorado, USA.

Madsen, M. D., K. W. Davies, C. S. Boyd, J. D. Kerby, D. L. Carter, and T. J. Svejcar. 2013. Restoring North America’s sagebrush steppe ecosystem using seed enhancement technologies. Pages 393–401 in D. L.

Michalk, G. D. Millar, W. B. Badgery, and K. M.

Broadfoot, editors. Revitalising Grasslands to Sustain our Communities. Proceedings of the 22nd International Grassland Congress, 15–19 September 2013, Sydney, Australia. New South Wales Department of Primary Industry, Orange, New South Wales, Australia.

Manier, D. J., D. J. A. Wood, Z. H. Bowen, R. M.

Donovan, M. J. Holloran, L. M. Juliusson, K. S.

Mayne, S. J. Oyler-McCance, F. R. Quamen, D. J.

Saher, and A. J. Titolo. 2013. Summary of science, activities, programs, and policies that influence the rangewide conservation of Greater Sage-Grouse (Centrocercus urophasianus): USGS Open File Report 2013–1098. http://pubs.usgs.gov/of/2013/1098/

Maurer, E. P., L. Brekke, T. Pruitt, and P. B. Duffy. 2007.

Fine-resolution climate projections enhance region- al climate change impact studies. Eos Transactions

AGU 88:504.

Maurer, E. P., A. W. Wood, J. C. Adam, D. P.

Lettenmaier, and B. Nijssen. 2002. A long-term hydrologically based dataset of land surface fluxes and states for the conterminous United States.

Journal of Climate 15:3237–3251.

McAdoo, J. K., C. S. Boyd, and R. L. Sheley. 2013. Site, competition, and plant stock influence transplant success of Wyoming big sagebrush. Rangeland Ecology & Management 66:305–312.

McArthur, E. D., and A. P. Plummer. 1978. Biogeog- raphy and management of native western shrubs: a case study, section tridentatae of Artemisia. Great Basin Naturalist Memoirs 2:229–243.

Meinke, C. W., S. T. Knick, and D. A. Pyke. 2009. A spatial model to prioritize sagebrush landscapes in the Intermountain West (USA) for restoration.

Restoration Ecology 17:652–659.

Miller, D. A., and R. A. White. 1998. A conterminous United States multilayer soil characteristics dataset for regional climate and hydrology modeling.

Earth Interactions 2:1–26.

Mok, H.-F., S. K. Arndt, and C. R. Nitschke. 2012.

Modelling the potential impact of climate variabil- ity and change on species regeneration potential in the temperate forests of south-eastern Australia.

Global Change Biology 18:1053–1072.

Moran, E. V., and J. M. Alexander. 2014. Evolutionary responses to global change: lessons from invasive species. Ecology Letters 17:637–649.

Moss, R. H., J. A. Edmonds, K. A. Hibbard, M. R.

Manning, S. K. Rose, D. P. van Vuuren, T. R. Carter, S. Emori, M. Kainuma, T. Kram, G. A. Meehl, J. F. B.

Mitchell, N. Nakicenovic, K. Riahi, S. J. Smith, R. J.

Stouffer, A. M. Thomson, J. P. Weyant, and T. J.

Wilbanks. 2010. The next generation of scenarios for climate change research and assessment.

Nature 463:747–756.

Neilson, R. P., J. M. Lenihan, D. Bachelet, R. J. Drapek, R. D. Sparrowe, and L. H. Carpenter. 2005. Climate change implications for sagebrush ecosystems.

Pages 145–159 in Transactions of the 70th North American Wildlife and Natural Resources Confer- ence. Wildlife Management Institute, Arlington, Virginia, USA.

Nelson, D. L., and R. G. Krebill. 1981. A sagebrush wilt disease of unknown cause. Great Basin Naturalist 41:184–191.

Nelson, Z. J., P. J. Weisberg, and S. G. Kitchen. 2014.

Influence of climate and environment on post-fire recovery of mountain big sagebrush. International Journal of Wildland Fire 23:131–142.

Parton, W. J. 1978. Abiotic section of ELM. Pages 31–53 inG. S. Innis, editor. Grassland simulation model.

Springer, New York, New York, USA.

Perryman, B. L., M. M. Aaron, A. L. Hild, and R. A.

Olson. 2001. Demographic characteristics of 3

(17)

Artemisia tridentata Nutt. subspecies. Journal of Range Management 54:166–170.

Piedallu, C., J.-C. Ge´gout, V. Perez, and F. Lebourgeois.

2013. Soil water balance performs better than climatic water variables in tree species distribution modelling. Global Ecology and Biogeography 22:470–482.

Piontek, F., C. Mu¨ller, T. A. M. Pugh, D. B. Clark, D.

Deryng, J. Elliott, F. d. J. Colo´n Gonza´lez, M.

Flo¨rke, C. Folberth, W. Franssen, K. Frieler, A. D.

Friend, S. N. Gosling, D. Hemming, N. Khabarov, H. Kim, M. R. Lomas, Y. Masaki, M. Mengel, A.

Morse, K. Neumann, K. Nishina, S. Ostberg, R.

Pavlick, A. C. Ruane, J. Schewe, E. Schmid, T.

Stacke, Q. Tang, Z. D. Tessler, A. M. Tompkins, L.

Warszawski, D. Wisser, and H. J. Schellnhuber.

2014. Multisectoral climate impact hotspots in a warming world. Proceedings of the National Academy of Sciences 111:3233–3238.

R Development Core Team. 2014. R: A language and environment for statistical computing. http://www.

R-project.org

Rickebusch, S., W. Thuiller, T. Hickler, M. B. Arau´jo, M. T. Sykes, O. Schweiger, and B. Lafourcade. 2008.

Incorporating the effects of changes in vegetation functioning and CO2on water availability in plant habitat models. Biology Letters 4:556–559.

Rizzo, M. L., and G. J. Szekely. 2014. energy: E- statistics (energy statistics). http://CRAN.R-project.

org/package¼energy

Rowland, M. M., M. J. Wisdom, L. H. Suring, and C. W. Meinke. 2006. Greater Sage-Grouse as an umbrella species for sagebrush-associated verte- brates. Biological Conservation 129:323–335.

Russell, B. D., C. D. G. Harley, T. Wernberg, N.

Mieszkowska, S. Widdicombe, J. M. Hall-Spencer, and S. D. Connell. 2011. Predicting ecosystem shifts requires new approaches that integrate the effects of climate change across entire systems. Biology Letters 8:164–166.

Sala, O. E., W. K. Lauenroth, and W. J. Parton. 1992.

Long-term soil-water dynamics in the shortgrass steppe. Ecology 73:1175–1181.

Schlaepfer, D. R., W. K. Lauenroth, and J. B. Bradford.

2012a. Consequences of declining snow accumula- tion for water balance of mid-latitude dry regions.

Global Change Biology 18:1988–1997.

Schlaepfer, D. R., W. K. Lauenroth, and J. B. Bradford.

2012b. Ecohydrological niche of sagebrush ecosys- tems. Ecohydrology 5:453–466.

Schlaepfer, D. R., W. K. Lauenroth, and J. B. Bradford.

2012c. Effects of ecohydrological variables on current and future ranges, local suitability patterns, and model accuracy in big sagebrush. Ecography 35:374–384.

Schlaepfer, D. R., W. K. Lauenroth, and J. B. Bradford.

2014a. Modeling regeneration responses of big

sagebrush (Artemisia tridentata) to abiotic condi- tions. Ecological Modelling 286:66–77.

Schlaepfer, D. R., W. K. Lauenroth, and J. B. Bradford.

2014b. Regeneration processes in big sagebrush (Artemisia tridentata). Rangeland Ecology & Man- agement 67:344–357.

Sexton, J. P., P. J. McIntyre, A. L. Angert, and K. J. Rice.

2009. Evolution and Ecology of Species Range Limits. Annual Review of Ecology, Evolution, and Systematics 40:415–436.

Shafer, S. L., P. J. Bartlein, and R. S. Thompson. 2001.

Potential changes in the distributions of western North America tree and shrub taxa under future climate scenarios. Ecosystems 4:200–215.

Shaw, N. L., A. M. DeBolt, and R. Rosentreter. 2005.

Reseeding big sagebrush: techniques and issues.

Pages 99–108 in N. L. Shawn, M. Pellant, and S. B.

Monsen, editors. Sage-Grouse habitat restoration symposium proceedings; 2001 June 4–7, Boise, Idaho. Proceedings RMRS-P-38. USDA Forest Service, Rocky Mountain Research Station, Fort Collins, Colorado, USA.

Shugart, H. H., and F. I. Woodward. 2011. Global change and the terrestrial biosphere: achievements and challenges. Wiley-Blackwell, Chichester, West Sussex, UK.

Shultz, L. M. 2006. The genus Artemisia (Asteraceae:

Anthemideae). Pages 503–534 in Flora of North America north of Mexico. Oxford University Press, New York, New York, USA.

Simpson, T., J. D. Peplinski, P. N. Koch, and J. K. Allen.

1997. On the use of statistics in design and the implications for deterministic computer experi- ments. Pages 1–14 in Proceedings of DETC’97:

1997 ASME Design Engineering Technical Confer- ences, September 14–17, 1997, Sacramento Califor- nia. American Society of Mechanical Engineers, New York, New York, USA.

Stein, B. A., P. Glick, N. Edelson, and A. Staudt. 2014.

Climate-smart conservation: Putting adaptation principles into practice. National Wildlife Federa- tion, Washington, D.C., USA.

Szekely, G. J., and M. L. Rizzo. 2009. Brownian distance covariance. Annals of Applied Statistics 3:1236–1265.

Taylor, K. E., R. J. Stouffer, and G. A. Meehl. 2012. An overview of CMIP5 and the experiment design.

Bulletin of the American Meteorological Society 93:485–498.

Thomas, C. D., A. M. Franco, and J. K. Hill. 2006.

Range retractions and extinction in the face of climate warming. Trends in Ecology & Evolution 21:415–416.

Thuiller, W., C. Albert, M. B. Arau´jo, P. M. Berry, M.

Cabeza, A. Guisan, T. Hickler, G. F. Midgley, J.

Paterson, F. M. Schurr, M. T. Sykes, and N. E.

Zimmermann. 2008. Predicting global change

(18)

impacts on plant species’ distributions: Future challenges. Perspectives in Plant Ecology, Evolution and Systematics 9:137–152.

Torell, L. A., K. C. McDaniel, and C. G. Ochoa. 2005.

Economics and optimal frequency of Wyoming big sagebrush control with tebuthiuron. Rangeland Ecology & Management 58:77–84.

US Fish and Wildlife Service. 2013. Greater Sage- Grouse (Centrocercus urophasianus) conservation objectives: final report. US Fish and Wildlife Service, Denver, Colorado, USA.

Walther, G.-R., E. Post, P. Convey, A. Menzel, C.

Parmesan, T. J. C. Beebee, J.-M. Fromentin, O.

Hoegh-Guldberg, and F. Bairlein. 2002. Ecological responses to recent climate change. Nature 416:389–395.

Welch, B. L. 2005. Big sagebrush: a sea fragmented into lakes, ponds, and puddles. USDA Forest Service, Rocky Mountain Research Station, Fort Collins, Colorado, USA.

West, N. E. 1983. Temperate deserts and semi-deserts.

Volume 5. Ecosystems of the world. Elsevier, Amsterdam, Netherlands.

West, N. E., and J. A. Young. 1999. Intermountain valleys and lower mountain slopes. Pages 255–284 inM. G. Barbour and W. D. Billings, editors. North American terrestrial vegetation. Cambridge Uni- versity Press, New York, New York, USA.

Westerling, A. L., H. G. Hidalgo, D. R. Cayan, and T. W. Swetnam. 2006. Warming and earlier spring increase western U.S. forest wildfire activity.

Science 313:940–943.

Westerling, A. L., M. G. Turner, E. A. H. Smithwick, W. H. Romme, and M. G. Ryan. 2011. Continued warming could transform Greater Yellowstone fire regimes by mid-21st century. Proceedings of the National Academy of Sciences 108:13165–13170.

Wheeler, J. A., G. Hoch, A. J. Cortes, J. Sedlacek, S.

Wipf, and C. Rixen. 2014. Increased spring freezing vulnerability for alpine shrubs under early snow- melt. Oecologia 175:219–229.

White, J. W., A. Rassweiler, J. F. Samhouri, A. C. Stier, and C. White. 2014. Ecologists should not use statistical significance tests to interpret simulation model results. Oikos 123:385–388.

Wiens, J. A., C. S. Crawford, and J. R. Gosz. 1985.

Boundary dynamics: A conceptual framework for studying landscape ecosystems. Oikos 45:421–427.

Wiens, J. A., D. Stralberg, D. Jongsomjit, C. A. Howell, and M. A. Snyder. 2009. Niches, models, and climate change: Assessing the assumptions and uncertainties. Proceedings of the National Acade- my of Sciences USA 106:19729–19736.

Zhu, K., C. W. Woodall, and J. S. Clark. 2012. Failure to migrate: lack of tree range expansion in response to climate change. Global Change Biology 18:1042–

1052.

Ziegenhagen, L. L., and R. F. Miller. 2009. Postfire recovery of two shrubs in the interiors of large burns in the intermountain west, USA. Western North American Naturalist 69:195–205.

(19)

S

UPPLEMENTAL

M

ATERIAL

A

PPENDIX

Table A1. CMIP5 GCM available from the ‘‘Downscaled CMIP3 and CMIP5 Climate and Hydrology Projections’’

archive at http://gdo-dcp.ucllnl.org/downscaled_cmip_projections/, their node number in Knutti et al.’s family tree (Fig. 1B, Knutti et al. 2013) after cutting at level 16, and performance against temperature and precipitation observations, i.e., normalized distance from observations (smaller is better) (Fig. 3, Knutti et al. 2013). We selected the best performing GCM from each node with available information.

CMIP5 GCM Selection in our study Node in family tree Performance

ACCESS1.0 1 0.76

ACCESS1.3 1 0.91

BCC-CSM1.1 2 1.11

BCC-CSM1.1(m) NA NA

BNU-ESM NA NA

CanESM2 10 10 0.91

CCSM4 11 0.67

CESM1(BGC) 11 0.66

CESM1(CAM5) 11 11 0.60

CMCC-CM 4 0.72

CNRM-CM5 2 0.73

CSIRO-Mk3.6.0 14 14 1.19

EC-EARTH 2 2 0.70

FGOALS-g2 15 15 0.97

FGOALS-s2 12 12 0.96

FIO-ESM 11 0.98

GFDL-CM3 5 5 0.76

GFDL-ESM2G 5 NA

GFDL-ESM2M NA 0.92

GISS-E2-H-CC NA NA

GISS-E2-R 7 7 0.92

GISS-E2-R-CC NA NA

HadCM3 NA NA

HadGEM2-AO NA NA

HadGEM2-CC 1 1 0.74

HadGEM2-ES 9 9 0.71

INM-CM4 8 8 1.32

IPSL-CM5A-LR 13 1.01

IPSL-CM5A-MR 13 13 0.96

IPSL-CM5B-LR NA 1.28

MIROC-ESM 16 16 1.26

MIROC-ESM-CHEM 16 1.27

MIROC4h NA 0.87

MIROC5 3 3 0.78

MPI-ESM-LR 4 0.71

MPI-ESM-MR 4 4 0.67

MRI-CGCM3 6 6 0.99

NorESM1-M 3 0.87

NorESM1-ME NA 0.88

參考文獻

相關文件

• To enhance teachers’ knowledge and understanding about the learning and teaching of grammar in context through the use of various e-learning resources in the primary

Wang, Solving pseudomonotone variational inequalities and pseudocon- vex optimization problems using the projection neural network, IEEE Transactions on Neural Networks 17

Comparison of B2 auto with B2 150 x B1 100 constrains signal frequency dependence, independent of foreground projections If dust, expect little cross-correlation. If

Define instead the imaginary.. potential, magnetic field, lattice…) Dirac-BdG Hamiltonian:. with small, and matrix

1: Invariant mass distribution and the leading lepton transverse momentum in data and simulation..

“Big data is high-volume, high-velocity and high-variety information assets that demand cost-effective, innovative forms of information processing for enhanced?. insight and

www.edb.gov.hk> School Administration and Management> Financial Management > Notes to School Finance> References on Acceptance of Advantages and Donations by Schools

Microphone and 600 ohm line conduits shall be mechanically and electrically connected to receptacle boxes and electrically grounded to the audio system ground point.. Lines in