REVIEW several trends: increasing frequency, extent, and

severity of disturbances; growing recognition

of the profound effects of anthropogenic cli-
Temperate forest health in an mate change; presence of novel anthropogenic
stressors; and a burgeoning global human pop-

era of emerging megadisturbance ulation that imposes escalating demands on

forests (9). The focus of forest health has shifted
toward evaluating forest conditions relative to
Constance I. Millar1* and Nathan L. Stephenson2 supporting human needs—that is, the capacity
of forests to sustainably provide ecosystem ser-
Although disturbances such as fire and native insects can contribute to natural dynamics vices. These services include provisioning (e.g.,
of forest health, exceptional droughts, directly and in combination with other disturbance water), regulating (e.g., carbon sequestration),
factors, are pushing some temperate forests beyond thresholds of sustainability. Interactions supporting (e.g., biogeochemical cycling), and

Downloaded from [Link] at Amrita Vishwa Vidyapeetham (Amritapuri Campus) Kollam on February 06, 2024
from increasing temperatures, drought, native insects and pathogens, and uncharacteristically cultural (e.g., recreational) benefits (10). To the
severe wildfire are resulting in forest mortality beyond the levels of 20th-century experience. complexity of ecological dynamics affecting for-
Additional anthropogenic stressors, such as atmospheric pollution and invasive species, est health are thereby added the many ways in
further weaken trees in some regions. Although continuing climate change will likely drive which humans use and value forests (9).
many areas of temperate forest toward large-scale transformations, management actions can Recent forest research has thus focused on
help ease transitions and minimize losses of socially valued ecosystem services. the role of thresholds and ecological conversions

F
(changes in ecological state) (11–14). Whereas
orests not only are essential components Early definitions (7) described healthy forests in recent decades, promoting resilience has been
of the natural environment but also offer as those with trees growing at their optimal a widespread goal of forest management, the
profound spiritual and material benefits to capacity, free from serious effects from insects, increasing pressure of chronic and acute dis-
humanity. After centuries of exploitation, disease, or wildfire. Fires were suppressed, and turbances is pushing many temperate forests
there is much to celebrate in the resilience insects and pathogens controlled. Through the toward and over resilience thresholds. The con-
(ability to rebound after perturbation) of tem- mid to late 20th century, evolving understand- sequences of heat waves, extreme droughts, mas-
perate forests. Broad swaths of forest that were ing of ecological dynamics, as well as increas- sive wildfire, and widespread insect outbreaks
cut in recent centuries continue to regrow vig- ing focus on forests as including organisms demonstrate to the public and scientists alike
orously, absorbing a substantial proportion of beyond the trees, led to recognition that nat- that resilience can be exhausted and that major
anthropogenic carbon-dioxide emissions (1). De- ural disturbances—including fires, insects, and ecological transformations can result. Serious
spite deeply concerning declines of ancient trees diseases—were essential ingredients of ecosys- thresholds are crossed when forests convert to
in forests worldwide (2), large trees are increas- tem functioning (8). Combined opinion shifted vegetation types without trees and, as a result,
ingly abundant in areas of temperate forests that to a recognition that disturbance was inherent lose valued forest ecosystem services.
are regrowing after logging (3). In other regions, to forest dynamics and contributes to healthy Forest health can be considered in the context
air-quality regulations have reduced acidic depo- forest functioning and resilience. of disturbance effects. Four patterns of 21st-
sition and other air-pollution effects on forests, Most recently, however, researchers are seek- century forest response to cumulative distur-
providing improved conditions for forest growth ing to understand how much disturbance can bance range from resilient (healthy) to collapse
and sustainability (4). be tolerated before forest health and persist- (unhealthy) states, as megadisturbances increase
Despite some encouraging trends, 21st-century ence are threatened. This concern emerges from in frequency and extent (Fig. 2). Thresholds can
forests still face grave threats. For millennia, the
main threat to forests was overexploitation, but
recent research has identified a range of emerging
challenges to forest persistence and health. We fo-
cus on emerging “megadisturbances” that are cap-
able of driving abrupt tree mortality of a spatial
extent, severity, and frequency surpassing that
recorded during recent human history. Where these
occur, effects to ecosystems and society follow. Thus,
while acknowledging the resilience of many for-
ests, we highlight here the nature and consequences
of changing environmental conditions that in-
creasingly threaten widespread regions of tem-
perate forest. In particular, we describe the rise
of an especially potent threat to forest health that
has only recently begun to receive broad attention,
that is, persistent and recurring drought combined
with high temperatures (see Fig. 1).

Forest health, thresholds,

and megadisturbance
Concepts of temperate forest health have changed
substantially over the past several decades (5, 6).
1
U.S. Department of Agriculture Forest Service, Pacific
Southwest Research Station, Albany, CA 94710, USA. 2U.S.
Geological Survey, Western Ecological Research Center,
Three Rivers, CA 93271, USA. Fig. 1. Drought- and bark-beetle–induced mortality in high-elevation whitebark pine (Pinus albicaulis)
*Corresponding author. E-mail: cmillar@[Link] forests, northern Warner Mountains (Drake Peak), Oregon. [Photo by C. I. Millar]

SCIENCE [Link] 21 AUGUST 2015 • VOL 349 ISSUE 6250 823

 F O R E S T H EA L T H

occur within and between classes; ecological tures increase tree water stress by increasing The historically unprecedented severity of
functions of forests change; and capacities to the atmosphere’s evaporative demand for water some recent hotter droughts has, in turn, driv-
deliver ecosystem services are altered. This frame- (21). Such temperature-induced increases in evap- en unprecedented temperate forest mortality
work helps to distinguish a healthy amount of orative demand have transformed what would (15, 26, 27) (Fig. 3). Although the most severe
disturbance from profound transformations that otherwise be typical droughts (in terms of low examples come from semi-arid forests, as global
affect society in undesired ways. precipitation) into droughts of historically un- temperatures continue to rise most temperate
precedented severity (15, 17, 18). Additionally, forests may experience elevated forest mortal-
Rise of the “hotter drought” with warming temperatures, precipitation that ity during hotter droughts (28). Importantly, in
For millennia, drought has been a key distur- formerly fell as snow increasingly falls as rain addition to the acute effects of hotter drought,
bance agent in temperate forests. Over the past (22, 23). In historically snow-dominated forests, increasing temperatures can also result in long-
few decades, however, rising global temperatures the diminished snowpacks melt earlier and are term chronic increases in drought stress even
have contributed to droughts of a severity that thus unable to replenish soil moisture during when precipitation remains average or increases
is unprecedented in the last century or more the driest parts of the year (23–25), further in- (29). Such chronic temperature-induced increases

Downloaded from [Link] at Amrita Vishwa Vidyapeetham (Amritapuri Campus) Kollam on February 06, 2024
(12, 15–18). These exceptional droughts have been creasing water stress on trees. Finally, the direct in drought stress in the absence of declining
variously called “global-change-type droughts” effects of tree water stress can be exacerbated precipitation have been implicated in long-term
(19), “hot droughts” (20), or “hotter droughts” (14); by detrimental physiological effects of high tem- increases in background (noncatastrophic) tree
we use the last term because it best contrasts peratures (12). Hotter droughts can also affect mortality rates, such as in western North Am-
these recent droughts with the generally cooler forests indirectly by making them more vulner- erica (30).
droughts of earlier in the last century. Hotter able to attacks by insects and pathogens (de-
droughts have emerged as particularly powerful scribed in the next section). Both the direct and Compound stressors
drivers of temperate forest mortality (14). indirect effects of hotter drought can yield ab- Hotter droughts are emerging as novel drivers of
Hotter droughts affect forests both directly rupt and threshold responses in forest condition forest megadisturbance, but they do not act in
and indirectly (14). Directly, higher tempera- and process (Fig. 2) (14). isolation, and their effects are often compounded

Response 1 Response 2 Response 3 Response 4

Thresholds Not reached Approaching or crossed Crossed Crossed widely

Mega-
Absent or minimal Low to moderate Moderate Severe
disturbances
Forest
ecosystem Sustained Sustained Decline Collapse
services

Forest type sustained Forest converts to new Forest converts to new Forest converts
forest type (ecosystem forest type (ecosystem to non-forest type
services retained) services decline)

Forest1 Hotter Forest1 Forest1(2) Forest1(2, 3)

drought
Hotter Hotter Hotter
Fire drought Fire drought Fire drought
Fire Insects Insects Insects
Pathogens Pathogens Pathogens

Forest2 Forest3 Shrub or grassland

Pathogens Insects

British Columbia, Canada: California: oak replacing pine (57) South-central US:
maritime coniferous forests (55) Spain and Switzerland: oak replacing pine (58, 59) savannah and grassland
replacing oak forest (62)
Eastern and central US: American Southwest: piñon pine being lost
northern mid-successional from mixed pine-juniper forests (60) Australia: scrubland replacing
oak-hardwood forests (56) Argentina: Chilean cedar replacing false beech (61) eucalypt forest (11)

Fig. 2. Temperate forest responses to 21st-century disturbances. (Response 1) substantial enough that ecologic functions change and forest ecosystem services
Existing forests are resilient and capable of sustaining structure, composition, decline. Major structural changes may have occurred, or one or more dominant
function, and forest ecosystem services. (Response 2) As a result of distur- tree species are lost and may or may not be replaced by species not formerly
bances that affect species composition, forests convert to new forest types but present or present in minor amounts. (Response 4) Existing forests transform to
retain primary ecosystem functions and services. The transition results from nonforest types, such as shrublands or grasslands, losing forest structure, com-
changes in species abundances or the loss of one or more minor species. (Re- position, and function; forest ecosystem services severely decline. Thresholds
sponse 3) Existing forests convert to new forest types, and the changes are can occur both within classes and between classes. [Examples from (55–62)]

824 21 AUGUST 2015 • VOL 349 ISSUE 6250 [Link] SCIENCE

through interactions with other stressors. In 2
recent decades, outbreaks of insects and path-
0.45
A
1
ogens have resulted in millions of hectares of
forest defoliation, canopy dieback, declines in 0.4 0
NDVI FDSI
growth, and forest mortality in western North
-1
America and Europe (31–34). In many cases,
climate was a direct or indirect trigger for these r = 0.83 -2
0.35
other agents of megadisturbance or influenced
the severity and extent of outbreaks. In temper-
ate forests, warming temperatures can trigger 30
B piñon
population increases in many insect species, 25 -1
which serve as catalysts for widespread outbreaks. ponserosa pine
20
Indirectly, drought can weaken trees to a point % Dead Douglas-fir 6-yr FDSI

Downloaded from [Link] at Amrita Vishwa Vidyapeetham (Amritapuri Campus) Kollam on February 06, 2024
15 0
where insects and pathogens are able to over-
10
whelm tree defenses, further catalyzing wide-
spread outbreaks and epidemics. In some cases, 5 1
insect outbreaks are not directly related to climatic
events, and causes pertain instead to vagaries of
10000 C Bark-beetle area
regional context and forest type (34, 35). Indepen-
dently or in combination with other stressors, r = -0.81 -1
insect and pathogen disturbances can lead to 2 1000
km 2-yr FDSI
changes in forest condition similar in magnitude 0
to climate effects (36). 100
Hotter droughts interacting with other stress- 1
ors are also catalyzing major changes in fire
regimes (37, 38), and the term “mega” is most 10000
often applied to fires (11, 39). Of overriding con- D Wildfire area
cern is the increasing frequency of uncharac- 1000 -2
r = -0.83
teristically severe and large fires and longer fire
km2 100 0 FDSI
seasons in temperate forests globally (37, 38).
Megafires are more likely where conditions favor 10 2
build-up of dry fuel, either standing or on the
forest floor. Droughts directly affect fuel flam- 1 4
mability and structure through lowering mois-
1980 1990 2000 2010
ture contents and increasing tree mortality, and
indirectly through cumulative effects from in- Fig. 3. Effects of hotter drought in the American Southwest. The forest drought-stress index (FDSI)
sect epidemics and diseases that alter forest integrates the effects of warm-season water deficit (controlled mostly by high temperature) and cold-
conditions. As with insects and disease, mega- season precipitation. Declining values of FDSI correspond to increasing drought. In the Southwest, in-
fires often occur in atypically dense forest stands creasing drought has been accompanied by (A) declining vegetation greenness [normalized difference
(often the legacy of past management actions, vegetation index (NDVI)], a remotely sensed index of greenness; (B) increasing mortality of the three most
including fire suppression), homogeneous forest common conifers; (C) increasing area affected by bark-beetle outbreaks; and (D) increasing area affected
structure, and where large fires occurred previ- by wildfires. [Adapted by A. P. Williams (July 2015) with permission from Macmillan Publishers Ltd. (15)]
ously (39, 40). To the extent that these large fires
increase in the future, the potential for shifts to grassland and exclude restoration of native tree decades, temperate forests have provided a val-
new forest types and nonforest vegetation will species (45). uable ecosystem service by acting as a net sink of
accelerate (Fig. 2). atmospheric carbon dioxide (1), partly offsetting
Other anthropogenic environmental changes Effects on forest ecosystem services anthropogenic emissions. As megadisturbances
affect forest health, although not all detrimen- The effects on ecosystem services resulting from increase in frequency, extent, and severity, this
tally. In some cases, nitrogen and carbon dioxide hotter droughts and compounding stresses vary, service is likely to diminish. At the extreme, tem-
as atmospheric pollutants can act as fertilizers, based on how the ecological functions of forests perate forests could become net sources rather
improving tree growth, although effects are are affected and on the regional demands and than sinks of atmospheric carbon dioxide (9, 48).
highly varied and often transient (4, 41, 42). At needs of society (9). Extensive forest mortality can
chronically high concentrations, and in combi- impair water quantity and quality, forest products, Forests of the future: Easing transitions
nation with climate-change effects, air pollu- cultural and spiritual values, and recreation, with What do these changes portend for temperate
tants can defoliate and weaken trees, reduce concomitant effects on rural and urban econo- forests through the 21st century? In the short
forest growth, and contribute to forest mortality mies (33, 46). Forest fires, in particular, have term, some forests will likely continue to absorb
(4, 43, 44). Similarly, many nonnative invasive profound effects on human life, property, and eco- or rebound from disturbances, sustain a diversity
species—including insects, pathogens, plants, nomies. While megafires represent a small fraction of ecological functions, and deliver ecosystem
and mammals—interact with heat and drought of total wildfires, they account for a dispropor- services similar to those of past decades (Fig. 2).
to impair forest health (32, 33). In North America, tionately large percentage of suppression costs, Over the longer term, however, most temperate
for instance, alien pathogens widely transformed private property losses, natural resource dam- forests are likely to change in condition (49), with
chestnut and elm forests and increasingly threat- ages, and fatalities, representing some of the megadisturbances frequently catalyzing these ef-
en high-elevation pine ecosystems (32, 33). After worst civil disasters on record (11, 39, 47). fects (14). The changes could range from minor
wildfires in the Great Basin woodlands of west- Forests play a particularly critical role in the shifts in forest structure (e.g., tree density and ages)
ern North America, invasion by nonnative cheat- global carbon cycle, mediating climatic changes by and species compositions to major transformation
grass (Bromus tectorum) can alter fire regimes to providing feedback to atmospheric carbon diox- of vegetation types, some resulting in novel eco-
the extent that they favor persistence of invasive ide concentrations. Over at least the past several systems relative to recent centuries. In many cases,

SCIENCE [Link] 21 AUGUST 2015 • VOL 349 ISSUE 6250 825

 F O R E S T H EA L T H

18. N. S. Diffenbaugh, D. L. Swain, D. Touma, Proc. Natl. Acad. Sci.

1 2 3 U.S.A. 112, 3931–3936 (2015).
19. D. D. Breshears et al., Proc. Natl. Acad. Sci. U.S.A. 102,
15144–15148 (2005).
20. J. Overpeck, B. Udall, Science 328, 1642–1643 (2010).
21. D. D. Breshears et al., Front. Plant Sci. 4, 266 (2013).
22. G. J. McCabe, D. M. Wolock, Clim. Change 99, 141–153 (2010).
23. T. P. Barnett, J. C. Adam, D. P. Lettenmaier, Nature 438,
303–309 (2005).
Forest 24. P. W. Mote, A. F. Hamlet, M. P. Clark, D. P. Lettenmaier, Bull.
ecosystem Am. Meteorol. Soc. 86, 39–49 (2005).
services 25. I. T. Stewart, D. R. Cayan, M. D. Dettinger, J. Clim. 18,
1136–1155 (2005).
Time 26. J. Carnicer et al., Proc. Natl. Acad. Sci. U.S.A. 108, 1474–1478
(2011).

Downloaded from [Link] at Amrita Vishwa Vidyapeetham (Amritapuri Campus) Kollam on February 06, 2024
27. G. Matusick, K. X. Ruthrof, N. C. Brouwers, B. Dell,
1 2 3 G. S. J. Hardy, Eur. J. For. Res. 132, 497–510 (2013).
28. B. Choat et al., Nature 491, 752–755 (2012).
29. G. J. McCabe, D. M. Wolock, Clim. Change 2015, 1–13
(2015).
30. P. J. van Mantgem et al., Science 323, 521–524 (2009).
31. B. J. Bentz et al., Bioscience 60, 602–613 (2010).
32. R. N. Sturrock et al., Plant Pathol. 60, 133–149 (2011).
33. A. S. Weed, M. P. Ayres, J. A. Hicke, Ecol. Monogr. 83, 441–470
(2013).
Forest 34. A. Sallé, L.-M. Nageleisen, F. Lieutier, For. Ecol. Manage. 328,
79–93 (2014).
ecosystem
35. W. R. Anderegg et al., New Phytol. 10.1111/nph.13477
services (2015).
36. C. E. Flower, M. A. Gonzalez-Meler, Annu. Rev. Plant Biol. 66,
Time 547–569 (2015).
Fig. 4. Management practices can influence the nature of transitions between forest types. 37. M. Flannigan et al., For. Ecol. Manage. 294, 54–61 (2013).
38. W. M. Jolly et al., Nat. Commun. 6, 7537 (2015).
Numbers represent forest transitions through time. (Top) (1) Despite rapid directional environmental
39. S. L. Stephens et al., Front. Ecol. Environ 12, 115–122 (2014).
changes, managers strive to maintain forests within historical ranges of conditions and may initially 40. M. J. Jenkins, J. B. Runyon, C. J. Fettig, W. G. Page, B. J. Bentz,
succeed. (2) The forest may be inherently unstable in the new environment and, once a threshold is For. Sci. 60, 489–501 (2014).
exceeded, substantial mortality occurs, with an abrupt loss of most forest ecosystem services. (3) After 41. M. E. Fenn et al., Bioscience 53, 404–420 (2003).
42. S. Piao et al., Glob. Change Biol. 21, 1601–1609 (2015).
the die-back, recovery of forest ecosystem services is slow. (Bottom) (1) Managers anticipate an
43. M. Lorenz et al., in Forest and Society: Responding to Global
impending forest transition and facilitate it by reducing the probability of sudden die-back (e.g., by Drivers of Change, G. Mery et al., Eds. (IUFRO World Series,
thinning the forest to reduce competition for water) and by assisting establishment of species better Vienna, International Union of Forest Research Organizations,
adapted to future conditions. (2) The transition is gradual rather than abrupt, and ecosystem services 2010), pp. 55–74.
44. Y. Hoshika et al., Sci. Rep. 5, 9871 (2015).
are maintained at a higher (although reduced) level. (3) Forest ecosystem services more rapidly ap-
45. W. D. Billings, in The Earth in Transition: Patterns and
proach their original levels. Although some forest ecosystem services are eventually lost in both cases, Processes of Biotic Impoverishment, G. M. Woodell, Ed.
active management might facilitate a gradual rather than abrupt transition. (Cambridge Univ. Press, Cambridge, 1990), pp. 301–322.
46. I. L. Boyd, P. H. Freer-Smith, C. A. Gilligan, H. C. J. Godfray,
Science 342, 1235773 (2013).
drought-hardy species, species with physiologi- from compartmentalized to comprehensive strat- 47. J. Williams, For. Ecol. Manage. 294, 4–10 (2013).
cal plasticity capable of coping with compound egies, with attention to proactive methods (54). 48. J. A. Hicke, A. J. H. Meddens, C. D. Allen, C. A. Kolden, Environ.
stresses, and species with shorter statures might Although thresholds are likely to be approached Res. Ltrs. 8, 035032 (2013).
outcompete current species (50, 51). Native in- in the future, and changes are inevitable, the 49. P. Gonzalez, R. P. Neilson, J. M. Lenihan, R. J. Drapek, Glob.
Ecol. Biogeogr. 19, 755–768 (2010).
sects and pathogens may effectively act as inva- actions we take now in temperate forests can 50. X. Jiang et al., J. Clim. 26, 3671–3687 (2013).
sive exotics as they move beyond their historic ease and guide transitions, diminishing effects 51. N. G. McDowell, C. D. Allen, Nat. Clim. Change 5, 669–672 (2015).
ranges (52). to forest ecosystems and human societies. 52. A. J. Woods, J. For. Sci. 60, 484–486 (2014).
Minimizing the effects to society from these 53. B. E. Law, Glob. Change Biol. 20, 3595–3599 (2014).
54. C. I. Millar, C. W. Swanston, D. L. Peterson, in Climate Change
transitions is emerging as a primary goal for and United States Forests, D. L. Peterson, J. M. Vose,
RE FERENCES AND NOTES
forest management today (Fig. 4). A challenge T. Patel-Weynand, Eds. (Springer, Berlin, 2014), pp. 183–222.
1. Y. Pan et al., Science 333, 988–993 (2011).
to research will be to develop tools to assess the 2. D. B. Lindenmayer, W. F. Laurance, J. F. Franklin, Science 338,
55. R. A. Hember et al., Glob. Change Biol. 18, 2026–2040 (2012).
sensitivity of forests to thresholds from cumu- 56. M. C. Dietze, P. R. Moorcroft, Glob. Change Biol. 17, 3312–3326
1305–1306 (2012).
(2011).
lative disturbances and evaluate their vulnerability 3. P. E. Kauppi et al., Biogeosciences 12, 855–862 (2015).
57. P. J. McIntyre et al., Proc. Natl. Acad. Sci. U.S.A. 112,
to transformation. If we can identify in advance 4. W. deVries, M. H. Dobbertin, S. Solberg, H. F. van Dubben,
1458–1463 (2015).
M. Schaub, Plant Soil 380, 1–45 (2014).
the most vulnerable forests, in some cases man- 5. W. A. Warren, Soc. Nat. Resour. 20, 99–117 (2007).
58. A. Rigling et al., Glob. Change Biol. 19, 229–240 (2013).
agement intervention might be able to ease the 59. J. Carnicer et al., Glob. Ecol. Biogeogr. 23, 371–384 (2014).
6. A. Sulak, L. Huntsinger, J. For. 110, 312–317 (2012).
60. D. D. Breshears, L. López-Hoffman, L. J. Graumlich, Ambio 40,
transition to new and better-adapted forest states, 7. T. E. Kolb, M. R. Wagner, W. W. Covington, J. For. 92, 10–15
256–263 (2011).
minimizing losses of ecosystem services. Because (1994).
61. M. Suarez, T. Kitzberger, Can. J. For. Res. 38, 3002–3010 (2008).
8. K. F. Raffa et al., J. For. 107, 276–277 (2009).
the scope of the challenge is vast, triage exercises 9. D. Thom, R. Seidl, Biol. Rev. Camb. Philos. Soc. 10.1111/brv.12193
62. D. P. Bendixsen, S. W. Hallgren, A. E. Frazier, For. Ecol. Manage.
will almost certainly be necessary to identify the 347, 40–48 (2015).
(2015).
highest-priority forests and those where man- 10. Millennium Ecosystem Assessment, Ecosystems and Human
AC KNOWLED GME NTS
agement action might have the greatest effect. Well-Being: Synthesis (Island Press, Washington, DC, 2005).
11. M. A. Adams, For. Ecol. Manage. 294, 250–261 (2013). We thank M. Dettinger (U.S. Geological Survey), C. J. Fettig (U.S.
Success will depend on far more integrated 12. R. Teskey et al., Plant Cell Environ. 38, 1699–1712 (2015). Forest Service), J. Hicke (University of Idaho), S. Stephens (University
and coordinated efforts by institutions, agencies, 13. C. P. O. Reyer et al., J. Ecol. 103, 5–15 (2015). of California, Berkeley), and A. P. Williams (Lamont-Doherty Earth
and governments than presently exists (53). Dis- 14. C. D. Allen, D. D. Breshears, N. G. McDowell, Ecosphere 6, 129 Observatory) for reviewing the draft manuscript; A. P. Williams for
(2015). providing an update of FDSI trends for Fig. 3; and D. Delany (U.S.
tributed monitoring systems that observe changes Forest Service) for help with figures. C.I.M. and N.L.S. are supported
15. A. Park Williams et al., Nat. Clim. Change 3, 292–297
on multiple scales of forest health are essential; (2013). by the U.S. Forest Service and U.S. Geological Survey, respectively.
these will become increasingly reliant on remote 16. M. Lindner et al., J. Environ. Manage. 146, 69–83 (2014).
methods. Climate adaptation will likely move 17. S. M. Vicente-Serrano et al., Environ. Res. Lett. 9, 044001 (2014). 10.1126/science.aaa9933

826 21 AUGUST 2015 • VOL 349 ISSUE 6250 [Link] SCIENCE