Foundation Species: Consequences for the Structure and Dynamics of Forested Ecosystems Author(S): Aaron M

Loss of Foundation Species: Consequences for the Structure and Dynamics of Forested Ecosystems Author(s): Aaron M. Ellison, Michael S. Bank, Barton D. Clinton, Elizabeth A. Colburn, Katherine Elliott, Chelcy R. Ford, David R. Foster, Brian D. Kloeppel, Jennifer D. Knoepp, Gary M. Lovett, Jacqueline Mohan, David A. Orwig, Nicholas L. Rodenhouse, William V. Sobczak, Kristina A. Stinson, Jeffrey K. Stone, Christopher M. Swan, Jill Thompson, Betsy Von Holle and Jackson R. Webster Source: Frontiers in Ecology and the Environment, Vol. 3, No. 9 (Nov., 2005), pp. 479-486 Published by: Ecological Society of America Stable URL: http://www.jstor.org/stable/3868635 . Accessed: 19/09/2014 10:31

Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at . http://www.jstor.org/page/info/about/policies/terms.jsp

. JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship. For more information about JSTOR, please contact [email protected].

Ecological Society of America is collaborating with JSTOR to digitize, preserve and extend access to Frontiers in Ecology and the Environment.

http://www.jstor.org

This content downloaded from 158.135.136.72 on Fri, 19 Sep 2014 10:31:20 AM All use subject to JSTOR Terms and Conditions REVIEWS REVIEWS REVIEWS

Loss of foundation species: consequences I for the structure and dynamics of forested ecosystems

Aaron M Ellisonl*, Michael S Bank1, Barton D Clinton2, Elizabeth A Colburnm,Katherine Elliott2, Chelcy R Ford2, David R Foster1, Brian D Kloeppel3, Jennifer D Knoepp2, Gary M Lovett4, Jacqueline Mohan1, David A Orwig1, Nicholas L Rodenhouse5, William V Sobczak6, Kristina A Stinson', Jeffrey K Stone7, Christopher M Swan8, Jill Thompson9, Betsy Von Holle1, and Jackson R Webster10

In many forested ecosystems, the architecture and functional ecology of certain tree species define forest structure and their species-specific traits control ecosystem dynamics. Such foundation tree species are declining throughoutthe world due to introductionsand outbreaksof pests and pathogens,selective removal of individual taxa, and over-harvesting. Through a series of case studies, we show that the loss of foundation tree species changes the local environmenton which a varietyof other speciesdepend; how this disruptsfun- damental ecosystem processes,including rates of decomposition, nutrient fluxes, carbon sequestration,and energy flow; and dramatically alters the dynamics of associated aquatic ecosystems. Forests in which dynamics are controlled by one or a few foundation species appear to be dominated by a small number of strong interactions and may be highly susceptible to alternating between stable states following even small perturbations. The ongoing decline of many foundation species provides a set of important, albeit unfortunate, opportunities to develop the research tools, models, and metrics needed to identify foundation species, anticipate the cascade of immediate, short- and long-term changes in ecosystem structure and function that will follow from their loss, and provide options for remedial conservation and management. I Front Ecol Environ 2005; 3(9): 479-486

WV e are living in an era of unprecedented and rapid life on Earth (Eldridge 1998). The loss of a common or ecological change (Reid et al. 2005). Through abundant foundation species (sensu Dayton 1972; see habitat conversion, over-consumption of resources, and Panel 1), which by virtue of its structural or functional worldwide introductions of pests and pathogens, humans attributes creates and defines an entire ecological com- are causing species extinctions at a record rate: the sixth munity or ecosystem, can have dramatic effects on our extinction crisis in the billion-year history of eukaryotic perception of the landscape and broad consequences for associated biota, ecosystem function, and stability. In a nutshell: Foundation species differ from keystone predators (Paine * In manyecosystems, a singlefoundation species controls pop- 1966) in that the former usually occupy low trophic levels ulationand communitydynamics and modulatesecosystem whereas the latter are usually top predators.They are also processes distinct from core (Hanski 1982) in that foundation * species The lossof foundationspecies acutely and chronically impacts are not abundant and com- fluxesof andnutrients, foodwebs, and bio- species only locally regionally energy hydrology, mon but also create stable conditions diversity locally required by * Humanactivities, including logging and the introductionof many other species. They also serve to stabilizefundamental exoticpests and pathogens, often functionally remove founda- ecosystemprocesses such as productivityand water balance. tion treespecies from forests Trees are most to be foundation in * likely species Foundationspecies that are currentlybeing lost fromNorth forested as their architecture and functional Americanforests include eastern hemlock, Port-Orford cedar, ecosystems, andoaks and physiological characteristics define forest structure and alter microclimates, while their biomass and chemi-

'HarvardUniversity, Harvard Forest, 324 NorthMain Street, Petersham, MA 01366 *([email protected]);2USDA Forest Service SouthernResearch Station, Coweeta Hydrologic Lab, Otto, NC, 28763; 3Universityof Georgia,Institute of Ecology,Athens, GA 30602; 4Instituteof EcosystemStudies, Millbrook, NY 12545;5Wellesley College, Department of BiologicalSciences, Wellesley, MA 02481;6College of theHoly Cross, Department of Biology,Worcester, MA 01610;70regon State University, Department of Botanyand Plant Pathology, Corvallis, OR 97331;8University of Maryland,Baltimore County, Dept of Geography& EnvironmentalSystems, Baltimore, MD 21250;9University of PuertoRico, Institute for Tropical Ecosystem Studies, San Juan, PR 00931; 'VirginiaPolytechnic Institute and State University, Department of Biology,Blacksburg, VA 24061

www.frontiersinecology.org © The Ecological Society of America www.frontiersinecology.org

This content downloaded from 158.135.136.72 on Fri, 19 Sep 2014 10:31:20 AM All use subject to JSTOR Terms and Conditions I Loss of foundation species AM Ellison et al.

Figure 1. (a) An old-growtheastern hemlock (Tsuga canadensis) stand, (b and inset) a stand decliningfollowing 10 years of infestationby the introducedhemlock woolly adelgid (Adelges tsugae), and (c) denseregeneration of blackbirch (Betula lenta) saplingson a siteformerly dominated by easternhemlock in southernConnecticut. Nearly all hemlocktrees in this 150-hectareforest in southernConnecticut were killed in themid-1 990s by hemlockwoolly adelgid.

cal makeup contribute substantially to ecosystem canadensis)resulting from an introduced insect and pre- processes. Foundation tree species are declining through- emptive salvage logging; the local extirpation of white- out the world due to a number of factors, including intro- bark pine (Pinus albicaulis)caused by interactions among ductions and outbreaks of nonindigenous pests and a nonnative pathogen, a native insect, and human alter- pathogens, irruptions of native pests, over-harvesting and ation of fire regimes; and the functional removal of high-intensity forestry,and deliberate removal of individ- American chestnut (Castanea dentata) by an introduced ual species from forests. We use three examples from pathogen. Our examples focus on trees in systems we North America to illustrate consequences for both terres- know best, but they are broadly representative of a wide trial and aquatic habitats of the loss of foundation tree range of foundation species and illustrative of their role in species: the ongoing decline of eastern hemlock (Tsuga forests throughout the world (Panel 2).

Panel 1. The many definitions of foundation species * The rise and fall of eastern hemlock Followingnomenclatural priority, we adopt Dayton's(1972) ter- minology and general definitionof a foundationspecies: a single Majestic hemlock groves (Figure 1) evoke reverence, speciesthat definesmuch of the structureof a communityby creating affection, and poetry (Frost 1923). Eastern hemlock stableconditions other and and sta- locally for species, by modulating (Tsuga canadensis),one of the most long-lived, shade-tol- bilizingfundamental ecosystem processes. erant trees in North America, dominates about 1 x 106ha Subsequentauthors, working in differenthabitats and appar- to southern ently unaware of historical antecedents, have suggested terms of forest from the southern Appalachians with some or all of the attributesof foundationspecies, including: Canada and west to the central Lake states (McWilliams Core species (Hanski 1982) are locallyabundant and region- and Schmidt 2000). In the north, hemlock typically ally common;associated satellite species are sparse and rare. occurs in stands with species-poor understo- An associated model core-satellite nearly pure metapopulation (the ries. In the south, hemlock in mixed stands in nar- hypothesis) explains relationships between a species' local grows abundanceand its regionaldistribution. row riparian strips and moist coves, often with dense Dominant species (Grime 1984) competitively exclude understories of rhododendron (Rhododendronmaximum). subordinate species by garneringa disproportionate share of In hemlock-dominated stands, the combination of deep resources and contributingmost to productivity. shade and acidic, slowly decomposing litter results in a Keystone predators (Paine 1966) preferentiallyconsume rates of dominant and enhance local cool, damp microclimate, slow nitrogen cycling, competitors biodiversityby prevent- al. of ing exclusion of weaker competitors.Holling's (1992) extended and nutrient-poor soils (Jenkins et 1999). Canopies keystone hypothesis posits that all terrestrialecosystems are evergreen hemlocks have a higher leaf area index and controlled and organizedby a smallset of keystone species. lower transpiration rates per unit leaf area than canopies Structural species (Huston 1994) create physicalstructures of co-occurring deciduous trees (Catovsky et al. 2002). of environments,produce variabilityin physicalconditions, pro- hemlocks have much whole-tree vide resources,and create habitatfor interstitial Although greater respi- species. in and when deciduous trees Ecosystem engineers Uones et al. 1994) cause physical ration rates the spring fall, state changes in biotic or abiotic materialsand modulate avail- are leafless, during the summer hemlocks transpire about abilityof resources to other species. Class 5 auLoTenic eco- 50% of the total water released by deciduous trees (J system engineers are directly analogousto Dayton'sfounda- Hadley unpublished). These characteristics of hemlock, tion species. along with its high snow-interception rates, mediate soil

This content downloaded from 158.135.136.72 on Fri, 19 Sep 2014 10:31:20 AM All use subject to JSTOR Terms and Conditions AM Ellison et al. Loss of foundation species moisture levels, stabilize stream base-flows, and decrease 2002), and there are currently no effective biological or diel variation in stream temperatures.As a result, streams chemical controls of the adelgid in forested ecosystems. I flowing through hemlock forests support unique assem- The insect's impact is further exacerbated by pre-emptive blages of salamanders, fish, and freshwater invertebrates salvage logging, in which hemlock, which has modest that are intolerant of seasonal drying (Snyder et al. 2002). economic value, is cut in anticipation of future infesta- Hemlock stands also shelter deer and other wildlife. tion (Orwig et al. 2002). Populations of eastern hemlock have declined precipi- Hemlock could functionally disappear from eastern tously three times since the Pleistocence glaciation: forests in the next several decades. This species generally approximately 5500 years ago, coincident with regional does not re-establish following adelgid-induced mortality climate change and an outbreak of an insect similar to (Figure 1), but is replaced throughout its range by hard- the extant eastern hemlock looper (Lambdinafiscellaria; wood species, including birch (Betula spp), oaks (Quercus Bhiry and Filion 1996); about 200 years ago, following spp) and maples (Acer spp) (Orwig et al. 2002). In the forest conversion to agriculture, increases in fire, and southeastern United States, hemlock is replaced by yel- extensive logging for timber and tannin (McMartin low poplar (Liriodendrontulipifera; J Vose et al. unpub- 1992); and from the mid-1980s to the present, due to an lished) when Rhododendronis absent. Decline of hemlock introduced insect, the hemlock woolly adelgid (Adelges may lead to the local loss of its uniquely associated ants tsugae; Figure 1). This rapidly spreading insect kills trees (Ellison et al. 2005) and birds (Tingley et al. 2002), cause of all sizes and age-classes within 4-15 years of infestation regional homogenization of floral and faunal assemblages (Orwig et al. 2002). Hemlock has no apparent resistance (Ellison et al. 2005), change soil ecosystem processes to the adelgid; it rarely recovers from attack (Orwig et al. (Jenkins et al. 1999; Figure 2), and alter hydrological

Panel 2. Additional examples of foundation species from forests around the world

Bald cypress (Taxodium distichum) dominates deepwater swamps of southeastern North America (Sharitzand Mitsch 1993). Its presence and density affect the water table and flow of sediment and nutrients,and control structure and composition of associated plant and animalcommunities (Sharitz and Mitsch 1993).Intensive logging and removalof bald cypress dramaticallyalter hydrologyand nutrientcycling, reduce primaryproductivity, and increase sedimentation(Sun et al. 2001).

Douglas fir (Pseudotsuga menziesii) dominates young and old-growthforests at low and mid-elevationswest of the Cascade Range and at higherelevations in the interior of the PacificNorthwest of North America.Live trees, snags,and fallenlogs provide uniquehabi- tats for wildlife,including endangered and rare species such as the spotted owl (Strixoccidentalis).The evergreen foliage controls lightlev- els, microclimate,and gas exchangefrom the forest floor to the canopy(Parker et al. 2004). Loggingalters C and N cycling,wildlife abundance, and plant successional dynamics(Halpern et al. 2005). Unlikethe other foundationspecies discussed in this paper,Douglas fir is not currentlythreatened, as it is strongly favored by current forest managementpractices. However, many old-growth stands in the PacificNorthwest have been lost to loggingover the past decades. High-intensityfires resultingfrom long-termfire suppressionprac- tices, introducedpests or changes in the ecological dynamicsof nativepests, or changesin forest managementthat pose mortalityrisks to old-growthDouglas fir stands could have importantecological impactsin the future.

Fraser fir (Abies fraseri) is a locally abundant endemic species that occurs in six discrete, high-altitudeareas in the southern Appalachians(Hollingsworth and Hain 1991).There, Fraserfir defines high-elevationspruce-fir communities, with tightlyassociated ani- mal and plant species. Fraser fir has been decliningsince the balsam woolly adelgid (Adelgespiceae) was introduced in the 1930s (Hollingsworthand Hain 1991).Its loss increasesthe susceptibilityof its co-dominant,red spruce (Picearubens), to windthrow,and both species are sufferingadditional effects of climatewarming and air pollution(Hamburg and Cogbill 1988).

Jarrah is a uniqueAustralian forest type comprised mainlyof Eucalyptus marginata. This species experiences mass collapse and sud- den death followingwaterlogging, which increases infection of jarrahroots by zoospores of Phytophthoracinnamomi (Davison and Tay 1987),a soil-born pathogenicfungus introduced into Western Australiain 1921 that affects-2000 of the 9000 extant plantspecies there (Wills 1992). Followinginvasion by P cinnamomi,richness of woody perennialspecies in the jarrahunderstory declines significantly, whereas richness of monocots, herbaceous perennials,annuals, and geophytes are largelyunaffected (Wills and Keighery1994).

Port-Orford cedar (Chamaecyparis lawsoniana) is endemic to southwestern Oregon and northern California,grows on ultramafic and non-ultramaficsoils, in riparianand uplandsites, and occurs in the most diverse plantassociations in the region.On ultramaficsoils, Port-Orfordcedar is often the only ripariantree species. It is a foundationspecies for both terrestrialand aquatichabitats: it recycles cal- cium to surfacesoils, providesshade, and stabilizessoil and stream banks(Hansen et al. 2000). Its highlyrot-resistant wood provideshabi- tat heterogeneity and alters hydrology.The non-native,water-dispersed, and generally lethal root pathogen Phytophthoralateralis has spread into virtuallyall naturalforest stands from nursery plantsinfected in the early 1920s (Hansenet al. 2000).

Mangroves (Rhizophora spp) form dense, often monospecificstands in estuarine and coastal forests throughoutthe tropics;these forests have some of the highest reported net primaryproductivity of any ecosystem on the planet (Ellisonand Farnsworth2001). Removalof mangrovesleads to rapidbuild-up of acid sulfides in the soil, increasedshoreline erosion and sedimentationonto offshore coral reefs, and collapse of intertidalfood webs and inshore fisheries (Ellisonand Farnsworth2001). More than 2%of mangroveforests are lost annually,as forests are cut for fuel,coastal development,and wood fiber used to produce rayon.

www.frontiersinecology.org © The Ecological Society of America www.frontiersinecology.org

This content downloaded from 158.135.136.72 on Fri, 19 Sep 2014 10:31:20 AM All use subject to JSTOR Terms and Conditions Loss of foundation species AM Ellison et al.

(a) Northeast hemlock forest (b) Southeast deciduous forest with riparian (c) Northeasternand southeastern I (closed canopy year-round) hemlock and Rhododendron understory deciduous forest (fullyto partiallyclosed canopy year-round) (growingseasons: closed-canopy; winter:open-canopy) A A IA A AA A AA IA~~~

I I y -TI I Soils have Highlight A seasonallyleads Climateis Lowlight Low light Climate is seasonal inputs to species rich cool in summer, Climateis warmerin leads to cool in supports leaf litter, understoriesand warmin winter, Soils have summer, Soilshave summer, lower in-stream species-poor warmin winter, subcanopy DOC, higher morestable diel year-round understoryand low-quality Rhododendron colderin winter, lowerC:NP, periphyton and annual inputsof low- morestable diel hemlockand but with net lower lowin-stream andannual understory, higher nd metals biomass temperatures qualityacidic periphyton Rhododendronlimits in-stream dieland annual litter,nutrient- biomass temperatures litter,seasonal periphyton thermalvariation poorsoil and inputsof higher biomass waterwith high qualitydeciduous C:N:P,high litter,C:N:P ratios I Aland Hg andmetals intermediate betweennortheast Associated streams hemlockforests higherevapotranspiration during anddeciduous growingseason, seasonally Associated streams forests have low round higherhydrologic variability, year flashier to evapotranspiration,stable base flows, response precipitation, flowextends less far perennialflow extends fartherup t perennial up watershedthan in deciduousforests Associated streams watershedthan in hemlock have evapotranspirationrates and hydrologicalcharacteristics intermediatebetween northeastern hemlockforests and deciduous forests.The magnitudeof the differencedepends on the extent of riparianhemlock Figure 2. Conceptualmodel of shifts in terrestrialand aquaticecosystem processes following loss of easternhemlock from (a) northemand (b) southernforests, and (c) conversionto hardwood-dominatedstands.

regimes (Figure 2). ity in nearby aquatic systems (Kizlinski et al. 2002; C The effects of adelgid-induced hemlock mortality on Swan unpublished). stream ecosystems will be extensive. For example, hemlock streams support significantly more taxa of aquatic * The shifting mosaic of whitebark pine invertebrates than paired mixed-hardwood stands, and nearly 10% of the taxa are strongly associated with the Whitebark pine forms extensive contiguous stands in presence of hemlock (Snyder et al. 2002). Hemlock death high elevation forests of the Rocky Mountains of may result in a rapid pulse of large amounts of wood that Wyoming, Montana, Idaho, and Alberta, and smaller dis- decays more slowly than coarse woody debris from hard- junct populations in eastern and southwestern Oregon, woods. Large hemlock logs in streams retain sediment California, and Nevada. This dominant late-successional and organic matter and create novel habitat types. In species (Figure 3) grows as dense krummholz (stunted general, large hemlock logs are abundant in streams trees growing at or just below treeline at higher elevations draining forests where hemlock is an important riparian forming a very low, cushiony mat) at its upper elevational species. Although logs from adelgid-killed hemlocks may limit, whereas at lower elevations and less extreme sites, persist in streams for decades to centuries, eventually the it grows in association with other conifers and its domi- loss of hemlock will reduce in-stream wood, leading to a nance is maintained by periodic fire (Aro 2001). decline in sediment retention and productivity. Whitebark pine has occupied its current range for Logging of hemlock initiates more rapid and greater approximately 8000 years. In western North America, ecosystem changes than the adelgid because of the abrupt extensive forests of whitebark pine, spruce (Picea spp) vegetation and environmental changes, removal of wood, and poplar (Populus spp) developed after glacial retreat. soil scarification, and the presence of extensive slash left As warming continued from 8000-4000 years ago, white- by logging operations (Kizlinski et al. 2002). Nitrogen bark pine became restricted to high elevation sites availability and nitrification rates are significantly higher (MacDonald et al. 1989). in cut forests than in adelgid-damaged ones, increasing Whitebark pine cover at upper elevations retards the threat of nutrient losses and changing food availabil- snowmelt and modulates runoff and stream flows (Fames

This content downloaded from 158.135.136.72 on Fri, 19 Sep 2014 10:31:20 AM All use subject to JSTOR Terms and Conditions AM Ellison et al. Loss of foundation species

a, 0) OD I cD (D

F O 0 c 0 LO m n D. i 0 Figure 3. High-elevationstands of whitebarkpine (Pinus albicaulis). This speciesis transformedfrom (a) healthystands to (b) deadstands through the interactionof fire suppression,the introducedpathogen Cronartium ribicola that causeswhite pine blister rust (inset), and thenative bark beetle Dendroctonus ponderosae.

1990). At lower elevations, post-fire mid-successional important resources for wildlife and humans, and locally whitebark pine stands provide shade and cool soil, facili- exerted a strong influence on ecosystem structureand func- tating establishment of diverse plant communities and tion (Paillet 2002). Chestnut blight, caused by the canker associated cryptogams, invertebrates, and microbes, while pathogen Cryphonectriaparasitica, was introducedfrom Asia its seeds serve as a major seasonal food source for many in the late 19th century.The blight was first noted in New species of mammals and birds (Mattson et al. 2001). Yorkin 1904, spreadrapidly (-37 km yr'l) across the range Throughout its range, whitebarkpine is declining due to of chestnut, and within 50 years had converted this stately the combined effects of an introducedpathogen, Cronartium tree to a rarely flowering understoryshrub across approxi- ribicola,a native bark beetle, Dendroctonusponderosae, and mately 3.6 million ha (Anagnostakis 1987). fire-suppressionpolicies (Kendall and Keane 2001). The Chestnut has a rapid growth rate and sprouting ability, pathogen C ribicola,which causeswhite pine blisterrust, was wood with an extremely high tannin content, and leaves introduced from Eurasia into western North America in with a relatively low C:N ratio. Therefore, fundamental 1910 on imported white pine (Pinus strobus) seedlings forest ecosystem processes, including decomposition, planted near Vancouver,British Columbia (MacDonald and nutrient cycling, and productivity, probably changed sub- Hoff 2001). After its introduction, C ribicolaspread in a stantially following chestnut's replacement by other series of episodic pulses throughout western North America species. Decomposition of chestnut wood is much slower and by the late 1930s was established throughout the west, than other co-occurring hardwoods and its high tannin where it devastated pine stands (MacDonald and Hoff concentrations could restrict the mobilization of nutri- 2001). Fire exclusion allowed furtherreplacement of white- ents in soils. Additionally, chestnut's fast growth rate barkpine by more shade-tolerantspecies and at lower eleva- (Jacobs and Severeid 2004) might have resulted in rapid tions promotedthe growth of dense stands of lodgepole pine sequestration of carbon and nutrients. (Pinuscontorta). In turn, lodgepole pine supportshigh popu- Chestnut dominated a wide range of environments and lations of D ponderosaebeetles that disperse into adjacent its decline is thought to have altered both terrestrial and whitebark pine stands when beetle populations irrupt.In a aquatic processes. There is evidence to suggest that the positive feedback loop, drought- and disease-stressedwhite- abundance of chestnut along riparian corridors of the barkpines are furthersusceptible to beetle attack. southern Appalachians was due to production of allelo- Loss of whitebarkpine alters watershed hydrology imme- chemicals that prevented establishment of what we now diately as flashiness of streams increases, and changes the consider "typical"riparian shrub and tree species, includ- dynamics of wildlife populations and succession over ing eastern hemlock and rhododendron (Vandermast et longer time scales. Cone crops of whitebark pine have al. 2002). Ironically, therefore, the loss of one foundation declined due to interactions between white pine blister species - American chestnut - may have facilitated the rust, fire exclusion, and bark beetles. Carryingcapacities of establishment of another - eastern hemlock - which in species dependent on whitebark pine seeds have also turn is now threatened. declined with the cone supply of this irreplaceablespecies In most forested headwater streams, autumn leaf inputs (Mattson et al. 2001). serve as the predominant energy base for aquatic ecosystems. Where chestnut was replaced by oak, relatively * The shrub that was a tree: American chestnut rapidly decaying chestnut leaves with high nutritional quality for aquatic macroinvertebrates were replaced by American chestnut was once a foundation species in east- more slowly decaying oak leaves with lower nutritional ern North American forests (Figure 4). Chestnut and oak quality (Smock and MacGregor 1988). As a conse- were co-dominants in the southern Appalachians for quence, leaf-processing and consumption rates would nearly 4000 years and reached the northeast from have declined, decreasing growth rates and adult body 2500-1500 years ago (Paillet 2002). Chestnut provided mass in macroinvertebrate shredder communities. Many

This content downloaded from 158.135.136.72 on Fri, 19 Sep 2014 10:31:20 AM All use subject to JSTOR Terms and Conditions Loss of foundation species AM Ellison et al. I

Figure 4. Americanchestnut. (a) Chestnuttimber, Great Smoky Mountains of westernNorth Carolina (circa 1910), a foundation speciesthat was transformedin the mid-20thcentury to an understoryshrub (b; smallshrub in center)by the introducedpathogen Cryphonectriaparasitica, which causes chestnut blight (c; chestnutin theBlue Ridge Plateau killed by theblight, circa 1946).

stream macroinvertebrates have life cycles closely syn- landscape (eg Halpern et al. 2005), whereas epidemiolog- chronized to the dynamics of detrital decay, and this ical models of plant pathogens or species invasions indi- change in detrital quality undoubtedly affected the cate that changes in forest structure occur in wave-like macroinvertebrate assemblage, although there are no patterns (Johnson et al. 2004). Such studies suggest that data to support this supposition. Furthermore, slowly where complex spatial and temporal patterns of species decomposing chestnut wood persists for decades in stream loss occur, the effects at any particular location are channels, altering channel structure and providing habi- unlikely to be a linear function of area altered or changes tat for fish and invertebrates. For example, in an in species' dominance. Indeed, threshold responses, Appalachian headwater stream sampled in the late 1990s, including transitions to new types of ecosystems, should Wallace et al. (2001) found that 24% of the large (> 10 be expected where key dependent variables, such as mast cm diameter) woody debris still consisted of American production, herbivore or detritivore abundance, or adult chestnut that had died over 50 years earlier. survival, result from a complex web of indirect relationships (eg Ebenman and Jonsson in press). * Functional loss versus total loss * Responding to the loss of foundation species As foundation species decline, their control of ecosystem structure and processes may wane long before the species Because foundation tree species tend to be common, itself disappears completely. For example, as hemlock abundant, and large, our responses to their loss often come stands decline, tree death opens the canopy, drastically late and are conducted at inappropriatescales. For exam- altering the understory microclimate and causing the loss ple, the ongoing attempt to recreate the American chest- of unique habitat. Similarly, shrubbychestnut contributes nut by backcrossing the few remaining fertile individuals little to leaf area, wood production, or nuts, so that while with resistant species from Europe and Asia holds out the it is still present in many forests, the American chestnut promise of specimen trees in suburban lawns but is tree is functionally extinct. unlikely to reforest four million hectares with hybrid The potential effects on ecosystem function and com- chestnuts. Similarly, chemical control of the hemlock munity composition caused by the loss of foundation woolly adelgid requires injecting trees annually and can species can be either exacerbated or ameliorated by pat- only target isolated single trees or small groves. Biological terns of decline in time and space. For example, logging control of the adelgid using non-native, generalist, preda- and diseases such as chestnut blight or white pine blister ceous beetles is being explored with uneven regardfor the rust have resulted in rapid loss of foundation species over long history of unexpected impacts that can accompany broad areas. This contrasts with the slow death of indi- the importation of exotic insects (eg Howarth 1991; viduals over decades or partial loss of a species through Boettner et al. 2000). Although several million beetles are removal or death of only one age or size class, as in beech released every year, there have been no systematic bark disease (Griffin et al. 2003). Similarly, whether the attempts to determine whether self-sustaining populations spatial pattern of individual deaths occurs in mosaic fash- have become established or how effective they are at actu- ion or as an advancing wave influences the timing and ally controlling the adelgid in the field. Overall, we would magnitude of loss of a foundation species (Holdenreider et be much more likely to conserve foundation species and al. 2004), and perhaps the ultimate outcome. Forest frag- the systems they create if we set aside very large reservesof mentation often occurs in mosaic patterns across the intact forests and adopted techniques that preserve ecosys-

This content downloaded from 158.135.136.72 on Fri, 19 Sep 2014 10:31:20 AM All use subject to JSTOR Terms and Conditions AM Ellison et al. Loss of foundation species tem integrity in managed forest stands (Foster et al. 2005). out hemlock, hemlock forests cease to exist, and no other native conifer possesses the same suite of structural and I * Conclusions functional characteristics that simultaneously define its position in the system and control system-wide dynamics There is no sign that the currently increasing rates of and processes. Many recognized foundation tree species resource extraction, climate change, or global movement (Panel 2) that have been identified are conifers, but it of pests and pathogens will slow any time soon. Foundation remains an open question whether conifers are dispropor- species have disappearedbefore and they will continue to tionately represented among foundation species. We need disappear. Despite nearly half a century of research on new researchtools, models, and metrics that will allow us to foundation species (Panel 1), our understanding of the identify foundation species a priori and to anticipate the consequences of their loss is based on a small number of cascade of immediate, short- and long-term changes in case studies, as we usually identify foundation species only ecosystem structure and function that follow their loss. after they have declined dramatically.Our examples illus- Community viability analysis (Ebenman and Jonsson in trate how foundation tree species can control both terres- press) may provide some of these tools, but its utility awaits trial forest processes and the dynamics of aquatic systems empiricalevaluation. Ongoing declines of many foundation within their watersheds.However, detailed information on species (Panel 2) provide timely, though unfortunate, the importance of foundation species for key ecosystem opportunitiesto develop such tools and models. processes are scarce. Likewise, the impact on water quality from the loss of foundation could be substantialand species * Acknowledgements merits furtherstudy. Long-term monitoring can reveal how losses of founda- We thank A Barker-Plotkin, P Dayton, E Farnsworth, S tion species alter rates and trajectoriesof succession, leading Jefts, J Jones, J Malloway, B Mathewson, R McDonald, T in some cases to novel forest types (such as black birch Spies, and F Swanson for useful discussion and construc- forests in New England) with unexpected dynamics. tive comments on the manuscript. This work was sup- However, monitoring is not enough. Ecologists have long ported by the HarvardForest, NSF grants DEB 00-80592, appreciated the complex nature of interactions among DEB 02-18001, DEB 02-18039, DEB 02-36154, and species, and we encourage direct, experimental approaches DEB 02-36897, and the US Forest Service, and is a cross- that use current losses of foundation species as an opportu- site contribution of the Andrews, Baltimore, Coweeta, nity to determine how the removal of a single species can Harvard Forest, Hubbard Brook, and Luquillo Long-Term have immediate and profound effects on other species and Ecological Research Programs. ecosystem processes. The dynamics of communities and ecosystems shaped by * References foundation species are dominated by a small number of AnagnostakisSA. 1987.Chestnut blight: the classicalproblem of strong interactions (Figure 2). Such systems are relatively an introducedpathogen. Mycologia 79: 23-37. fragile and susceptible to switching between alternative sta- Arno SE 2001. Community types and natural disturbance processes.In: TombackDF, Aro SF, and KeaneRE (Eds). ble states following even small and perturbations(Dudgeon Whitebark pine communities, ecology and restoration. Petraitis 2005). At the same time as many forested systems Washington,DC: Island Press. are losing their foundation species, they are simultaneously BhiryN and Filion L. 1996. Mid-holocenehemlock decline in and synergistically threatened by climate change, atmos- easternNorth America linked with phytophagusinsect activ- pheric deposition, drought, and invasion of exotic species, ity.Quaternary Res 45: 312-20. BoettnerGH, ElkintonJS, andBoettner CJ. 2000. Effects of a bio- all of which may increase their overall fragility.Temperate- logicalcontrol introduction on threenontarget native species zone forests, such as those we have highlighted here, have of Satumiidmoths. Conserv Biol 14: 1789-1806. few tree species relative to the species-rich tropical forests CatovskyS, HolbrookNM, andBazzaz FA. 2002.Coupling whole- that garer much attention from ecologists and conserva- treetranspiration and canopy photosynthesis in coniferousand tion biologists. When there are only one or two foundation broad-leavedtree species. Can J ForRes 32: 295-309. DavisonEM and TayFCS. 1987. The effect of waterloggingon species in a forest, there is little functional in redundancy infectionof Eucalyptusmarginata seedlings by Phytophthoracin- many important respects, and their loss is likely to lead to namomi.New Phytol 105: 585-94. rapid, possibly irreversible,shifts in biological diversity and Dayton PK. 1972. Towardan understandingof community system-wide changes in structure and function (Ebenman resilienceand the potentialeffects of enrichmentsto the ben- and Jonsson in press). Regrettably, the lack of detailed thos at McMurdoSound, Antarctica.In: ParkerBC (Ed). of the natural of most in most Proceedingsof the colloquiumon conservationproblems in knowledge history species Antarctica.Lawrence, KS: Allen Press. and the abandonment of forests, courses and curricula in DaytonPK. 2003. The importanceof the naturalsciences to con- natural history (Dayton 2003), will leave us unawareof the servation.Am Nat 162: 1-13. collapse of the intricate webs of interactions and processes DudgeonS and PetraitisPS. 2005. Firstyear demography of the foundation and its that are lost when foundation species disappear. species,Ascophylum nodosum, community Oikos109: 405-15. These species provide fundamental structureto a system, implications. Ebenman B and Jonsson T. Using community viability analysisto and thus are by definition irreplaceable.For example, with- identify fragile systems and keystone species. TrendsEcol Evol www.frontiersinecology.org ©C The Ecological Society of America www.frontiersinecology.org

This content downloaded from 158.135.136.72 on Fri, 19 Sep 2014 10:31:20 AM All use subject to JSTOR Terms and Conditions Loss of foundation species AM Ellison et at.al.

doi:10.1016/j.tree.2005.06.011.In press. SF,and Keane RE (Eds). Whitebarkpine communities,ecology EldrigeN. 1998. Life in the balance:humanity and the biodiversity and restoration.Washington, DC: IslandPress. I crisis. Princeton, NJ: Princeton University Press. KizlinskiML, Orwig DA, Cobb RC, et al. 2002. Direct and indirect Ellison AM and FarnsworthEJ. 2001. Mangrovecommunities. In: ecosystem consequences of an invasive pest on forests domi- BertnessMD, Gaines SD, and Hay ME (Eds). Marine commu- nated by easternhemlock. J Biogeogr29: 1489-503. nity ecology. Sunderland,MA: SinauerAssociates. MacDonald GI and Hoff RJ. 2001. Blister rust: an introduced Ellison AM, Chen J, Diaz D, et al. 2005. Changes in ant commu- plague. In: Tomback DF, Aro SF, and Keane RE (Eds). nity structure and composition associated with hemlock Whitebark pine communities, ecology and restoration. decline in New England. In: Onken B and Reardon R (Eds). Washington,DC: IslandPress. Proceedingsof the 3rd symposiumon hemlock woolly adelgid MacDonaldGM, CwynarLC, and Whitlock C. 1989. Late quater- in the eastern United States. Morgantown,WV: USDA Forest nary dynamics of pines: northern North America. In: Service. RichardsonDM (Ed). Ecology and biogeographyof Pinus.New Fames PE. 1990. SNOTEL and snow course data describing the York,NY: CambridgeUniversity Press. hydrologyof whitebarkpine ecosystems. In: Schmidt WC and Mattson DJ, Kendall KC, and Reinhart DP. 2001. Whitebarkpine, McDonald KJ (Eds). Proceedingsof a symposiumon whitebark grizzlybears, and red squirrels.In: Tomback DF, Aro SF, and pine ecosystems:ecology and managementof a high mountain Keane RE (Eds). Whitebark pine communities, ecology and resource. Ogden, UT: USDA Forest Service Intermountain restoration.Washington, DC: IslandPress. ResearchStation. McMartin B. 1992. Hides, hemlocks, and Adirondack history. Foster DR, Kittredge D, Donahue B, et al. 2005. Wildlands and Utica, NY: North Country Books. woodlands:a vision for the forestsof Massachusetts.Petersham, McWilliams WH and Schmidt TL. 2000. Composition, structure, MA: HarvardForest. and sustainability of hemlock ecosystems in eastern North FrostR 1923. Dust of snow. In: New Hampshire:a poem with notes America. In: McManusKA, Shields KS, and Souto DR (Eds). and grace notes. New York,NY: Henry Holt and Co. Proceedings:symposium on sustainable management of hem- Griffin JM, Lovett GM, Arthur MA, et al. 2003. The distribution lock ecosystems in eastern North America, June 22-24, 1999. and severity of beech bark disease in the Catskill Mountains, Durham,NH: USDA ForestService. NY. CanJ For Res 33: 1754-60. OrwigDA, FosterDR, and MauselDL. 2002. Landscapepatterns of Grime JP. 1984. Dominant and subordinatecomponents of plant hemlock decline in New England due to the introducedhem- communities: implications for succession, stability and diver- lock woolly adelgid.J Biogeogr29: 1475-87. sity. In: Gray AJ and Crawley MJ (Eds). Colonization, succes- Paillet F 2002. Chestnut: history and ecology of a transformed sion and stability.Oxford, UK: Blackwell. species.J Biogeogr29: 1517-30. Halpern CB, McKenzie D, Evans SA, et al. 2005. Initial responses Paine RT. 1966. Food web complexity and species diversity.Am Nat for forest understoriesto varying levels and patterns of green- 100: 65-75. tree retention. EcolAppl 15: 175-95. ParkerGG, Harmon MA, and LefskyJQ. 2004. Three-dimensional HamburgSP and Cogbill CV. 1988. Historicaldecline of red spruce structure of an old-growth Pseudotsugatsuga canopy and its populationsand climatic warming.Nature 331: 428-31. implications for radiation balance, microclimate, and gas Hansen EM, Goheen DJ, Jules ES, et al. 2000. Managing Port- exchange. Ecosystems7: 440-53. Orford cedar and the introduced pathogen Phytophthoralater- Reid WV, Mooney HA, CropperA, et al. 2005. Millennium ecosys- alis. PlantDis 84: 4-10. tem assessmentsynthesis report.Washington, DC: Millennium Hanski I. 1982. Dynamics of regional distribution:the core and Assessmentand WorldResources Institute. satellite species hypothesis. Oikos38: 210-21. Sharitz R and Mitsch WJ. 1993. Southern floodplain forests. In: HoldenreiderO, PautassoM, WeisbergPJ, et al. 2004. Tree diseases Echtemacht AC (Ed). Biodiversityof the southeasternUnited and landscapeprocesses: the challenge of landscapepathology. States: lowland terrestrialcommunities. New York, NY: John TrendsEcol Evol 19: 446-52. Wiley and Sons. Holling CS. 1992. Cross-scalemorphology, geometry, and dynam- Smock LA and MacGregorCM. 1988. Impact of the American ics of ecosystems.Ecol Monogr 62: 447-502. chestnut blight on aquatic shredding macroinvertebrates.J N Hollingsworth RG and Hain FP. 1991. Balsam woolly adelgid Am BentholSoc 7: 212-21. (Homoptera,Adelgidae) and spruce-firdecline in the southern Snyder CD, YoungJA, Smith D, et al. 2002. Influence of eastern Appalachians: assessing pest relevance in a damaged ecosys- hemlock (Tsuga canadensis)forests on aquatic invertebrate tem. FloridaEntomol 74: 179-87. assemblagesin headwater streams. Can J Fish Aquat Sci 59: Howarth FG. 1991. Environmental impacts of classical biological 262-75. control. Ann Rev Entomol36: 485-509. Sun G, McNulty SG, ShepardJP, et al. 2001. Effectsof timberman- Huston MA.1994. Biological diversity:the coexistence of species agement on the hydrology of wetland forests in the southern on changing landscapes. Cambridge, UK: Cambridge United States. ForestEcol Manag 143: 227-36. University Press. Tingley MW, Orwig DA, Field R, et al. 2002. Avian response to Jacobs DF and Severeid LR. 2004. Dominance of interplanted removal of a forest dominant:consequences of hemlock woolly American chestnut (Castanea dentata) in southwestern adelgidinfestations. J Biogeogr29: 1505-16. Wisconsin, USA. ForestEcol Manag 191:111-20. VandermastDB, Van Lear DH, and Clinton BD. 2002. American Jenkins,J, Aber JD, and Canham CD. 1999. Hemlock woolly adel- chestnut as an allelopath in the southern Appalachians.Forest gid impacts on community structure and N cycling rates in EcolManag 165: 173-81. eastern hemlock forests.Can J For Res 29: 630-45. Wallace JB, WebsterJR, EggertSL, et al. 2001. Largewoody debris Johnson DM, Bj0mstad ON, and Liebhold AM. 2004. Landscape in a headwaterstream: long-term legacies of forest disturbance. geometry and travelling waves in the larch budmoth. Ecol Lett Int RevHydrobiol 86: 501-13. 7: 967-74. Wills RT. 1992. The ecological impactof Phytophthoracinnamomi in Jones CG, LawtonJH, and Shachak M. 1994. Organismsas ecosys- the Stirling Range National Park, Western Australia. Aust J tem engineers. Oikos69: 373-86. Ecol 17: 145-59. Kendall KC and Keane RE. 2001. Whitebark pine decline: infec- Wills RT and KeigheryGJ. 1994. Ecologicalimpact of plant disease tion, mortality,and population trends. In: TombackDF, Arno on plant communities.J Roy Soc WAust 77: 127-31.

This content downloaded from 158.135.136.72 on Fri, 19 Sep 2014 10:31:20 AM All use subject to JSTOR Terms and Conditions