Saproxylic Insect Ecology and the Sustainable Management of Forests Author(S): Simon J

Saproxylic Insect Ecology and the Sustainable Management of Forests Author(s): Simon J. Grove Source: Annual Review of Ecology and Systematics, Vol. 33 (2002), pp. 1-23 Published by: Annual Reviews Stable URL: http://www.jstor.org/stable/3069254 . Accessed: 11/03/2011 07:27

Your use of the JSTOR archive indicates your acceptance of JSTOR's Terms and Conditions of Use, available at . http://www.jstor.org/page/info/about/policies/terms.jsp. JSTOR's Terms and Conditions of Use provides, in part, that unless you have obtained prior permission, you may not download an entire issue of a journal or multiple copies of articles, and you may use content in the JSTOR archive only for your personal, non-commercial use.

Please contact the publisher regarding any further use of this work. Publisher contact information may be obtained at . http://www.jstor.org/action/showPublisher?publisherCode=annrevs. .

Each copy of any part of a JSTOR transmission must contain the same copyright notice that appears on the screen or printed page of such transmission.

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].

Annual Reviews is collaborating with JSTOR to digitize, preserve and extend access to Annual Review of Ecology and Systematics.

http://www.jstor.org Annu. Rev. Ecol. Syst. 2002. 33:1-23 doi: 10.1146/annurev.ecolsys.33.010802.150507 Copyright? 2002 by Annual Reviews. All rights reserved First publishedonline as a Review in Advance on August 6, 2002

SAPROXYLICINSECT ECOLOGY AND THE SUSTAINABLEMANAGEMENT OF FORESTS

Simon J. Grove Division of ForestResearch and Development,Forestry Tasmania, GPO Box 207, Hobart, Tasmania7001, Australia;email: [email protected]

Key Words dead wood, insect conservation, logging, beetles, forestry * Abstract Saproxylic insects comprise a diverse, species-rich and dominant func- tional group that share a dependence on dead wood and the old trees that generate it (maturetimber habitat). Recent research has highlighted their sensitivity to forest management, with managed or secondary forests generally supporting fewer individuals, fewer species, and different assemblages compared to old-growth or primary forests. This sensitivity is a product of their association with a habitat that tends to dimin- ish in managed forests. Many species also have low powers of dispersal relative to human-induced fragmentation, making breaks in habitat continuity particularlyharm- ful. In western Europe, many species are now regionally extinct. Information is largely lacking elsewhere, but similar ecological and management principles should apply. Measures taken to protect the habitat of hollow-dependent vertebrates may ensure the survival of some saproxylic insects, but unless their needs are expressly consid- ered, there remains the risk that many others may be lost as forest areas shrink and management of remaining areas intensifies.

INTRODUCTION

Scope of Review In this paper I review the literature on saproxylic insect ecology and conservation with particular reference to their relationships with forest management. I dwell particularly on the western European experience and also consider my own research findings from the tropical rainforests of northeastern Queensland, Australia. In so doing I highlight some of the common ecological traits shared by many saproxylic insects (especially the beetles, the subject of my own research) and relate these to the forest management practices that have so often led to their demise. Assum- ing that saproxylic insects are likely to respond similarly to these practices wherever they occur, I offer some thoughts on the extent to which they have a secure long- term future in managed forests worldwide, together with some suggestions for how their needs could better be addressed.

0066-4162/02/1215-0001$14.00 1 2 GROVE

What are SaproxylicInsects and Why do They Have Their Own Word? Saproxylic insects are defined as those that are "dependent,during some part of theirlife cycle, uponthe deador dying wood of moribundor deadtrees (standing or fallen), or upon wood-inhabitingfungi, or upon the presenceof othersaproxylics" (Speight 1989). The saproxylichabit includes representatives from all majorinsect orders(especially beetles andflies), andaccounts for a largeproportion of the insect faunain any naturalforest (Table 1). For example,56% of all forest beetle species in a region of Germanywere consideredsaproxylic (Kohler 2000). Beetles alone may representalmost 40% of all arthropodspecies (Grove& Stork2000), andthere areat least twice as many species of saproxylicbeetles as thereare terrestrial vertebrates(Parker 1982)-probably manymore. Saproxylic insects, then,are a species- richand functionally important component of forestecosystems. Other invertebrate groups,fungi, andother microorganisms could equallybe describedas saproxylic. The word saproxylic was coined in France (Dajoz 1966) and has been widely adoptedin Europe since the publicationsof Harding& Rose (1986) and Speight (1989). Publicationsfrom Canada(Hammond 1997) and Australia(Michaels & Bomemissza 1999, Grove & Stork 1999, Yee et al. 2001, Grove 2002a) have recentlyfollowed suit. Frequentlyused equivalentterms include deadwood(Elton 1966) and wood-living (SG Nilsson 1997). Saproxylic is a particularlyuseful word, because includedin the definition(besides wood-feeders)are bark-feeders, feederson wood-decomposingfungi, associatedpredators, parasitoids, detritivores feeding on theirwaste products,and other commensals. Hence, it refersto an entire functionalgroup that has obligateassociations with an arrayof dead-woodhabitats. Maturetimber habitat (Grove 2001a) is perhapsthe only termin frequentuse that can be used to sum up all the key habitatfeatures on which saproxylic insects depend. Examples include standing and fallen dead wood of various diameters and in various states and stages of decay, wood-rottingand otherdependent fungi (hyphaeand sporocarps),fissures and crevices in bark,water- or humus-filledrot- holes and other tree cavities, sap-runs,and the tunnels and frass of wood-borers. In fire-adaptedforests, charredwood is another example (Wikars 2001), as is waterloggedwood for aquaticsaproxylic insects (Braccia & Batzer2001).

"Deadwoodology":The Basics In the context of forest management,specific concern for saproxylic insects is acute in Europebut remainsat a much lower level elsewhere.Because the study of maturetimber habitat and saproxylicinsects may thereforebe unfamiliarto many readers,I offer the following six guiding principles:

THE RICHNESSOF LIFEIN THE "ARBOREALMEGALOPOLIS" Larger-diameter, overmature, senescent, moribund,decadent, or veteran trees (Read 1991) form the centerpiece of the mature timber habitat concept. They may be commercially overmaturebut are in their ecological prime of life and have been likened to an SAPROXYLICINSECTS AND FORESTRY 3

TABLE 1 Examples demonstratingthe richness of the saproxylicinsect fauna Country Reference Observations

Australia Hammond 21%of all beetle speciesin canopysamples from SE et al. (1996) Queenslandwere xylophagous Australia Yee et al. (2001) 104 beetlespecies hand-collected from Eucalyptus logs in a single forestregion in Tasmania Australia Grove(2002a) 339 saproxylicbeetle species recorded from flight intercepttraps in rainforestin NE Queensland FrenchGuyana Tavakilian 500 speciesof cerambycidbeetle were reared from dead et al. (1997) wood fromleguminous forest trees Indonesia Hammond(1990) 20%of the 3488 beetle speciescollected in north Sulawesiwere xylophagous or xylomycophagous UnitedKingdom Elton(1966) 456 speciesof invertebrateswere recorded from dead wood habitatsin a singlewoodland Finland Hanski& 287 speciesof saproxylicbeetles were recorded in a Hammond(1995) singleforest Finland Siitonen(2001) 20-25% of all forest-dwellingspecies (not just insects) were saproxylic,including about 800 beetle species Finland Martikainen 42%of 553 beetle species sampledin old spruceforest et al. (2000) were saproxylic Finland Martikainen(2001) 42%of 780 beetle speciescollected from dead aspen treeswere saproxylic Sweden Palm(1959) 342 speciesof beetleassociated with deadaspen in southernand central Sweden Sweden Ehnstrom(2001) 405 saproxylicbeetle speciesassociated with deadbirch, 389 with deadScotch pine and 354 with deadNorway spruce Norway 0kland Nearly700 obligatesaproxylic beetle speciesnationally et al. (1996) and200 facultativelysaproxylic species Germany Derksen(1941) 217 speciesof saproxylicinsects in deadbeech in a singleforest Germany K6hler(2000) 56%of all forest-dwellingbeetle speciesin north Rhinelandconsidered saproxylic Germany Blab et al. (1994) 25%of Germanbeetle speciesconsidered saproxylic Canada Hammond(1997) 257 speciesof saproxylicbeetle recorded from dead aspensin one foresttype UnitedStates Deyrup(1976) Morethan 300 saproxylicinsect species recorded from Douglas-firin Washington UnitedStates Howden& 188 speciesof beetlesrecorded from dead pines in a Vogt(1951) singleMaryland forest UnitedStates Savely(1939) 162 speciesof beetlesrecorded from pine andoak logs in a singleNorth Carolina forest 4 GROVE

arborealmegalopolis (Speight 1989) or termedmegatrees (Nilsson & Baranowski 1994). They probablyassume greatest relative importancein forests dominated by broadleaves, whether temperate (Speight 1989) or tropical (Grove 2002b) (Figure 1). In boreal and other coniferous or sclerophyllous forests, the standing dead trees (snags) and dead wood on the forest floor derivedfrom megatrees may be more important(Berg et al. 1994). However, despite the importanceof megatrees, we know relatively little about what lives in them, compared with what lives in snags or in dead wood on the forest floor. Some saproxylic insects are found in all three types, but many have distinct preferences(Palm 1959). In Swedish broadleavedforests megatrees supportmore substrate-specificspecies than either snags or dead wood on the forest floor and include most of the rarest and most threatenedsaproxylic insects (Nilsson & Baranowski1997, Jonsell et al. 1998).

THERE'SDIVERSITY IN DECAY A key source of saproxylic insect diversity is the rangeof decay stages andtypes thatoccur within dead wood, each providinghabi- tats for differentassemblages. As wood decomposesit is colonized by a succession of saproxylicinsect species (Blackman& Stage 1924, Kletecka 1996, Hammond et al. 2001), as are the fruiting-bodiesof wood-rottingascomycetes and basid- iomycetes (Jonsell et al. 1999). Insects that colonize living or freshly dead wood often show narrowhost specificity (Hamilton 1978); the same applies for species on wood-decaying fungi (Kaila et al. 1994). Host specificity is rare at the tree species level but common at higher plant taxonomic levels (e.g., Tavakilianet al. 1997). In Germanforests Kohler (2000) found that 13% of the saproxylicbeetle species were tree genus specific. As decompositionproceeds host specificitydrops off, though there remains a big difference between the fauna of hardwoodsand that of softwoods (Savely 1939, As 1993). Within either, decay type is important (Harmonet al. 1986, Jonsell et al. 1998). For instance, there is a big difference in the fauna depending on whether the wood was decayed by white-rot or brown-rotfungi (Araya 1993, Wood et al. 1996, Yee et al. 2001). Furtherdiffer- ences are attributableto wood moisturecontent and exposureto sun (Martikainen 2001).

ALL DEAD WOOD IS GOOD, BUT BIGGERIS BETTER Size is importantfor saproxylic insects. Some are able to make use of their chosen substratewhatever its dimensions(Palm 1959), but most are more particular(Elton 1966), which results in different assemblages in substratesof different sizes. Most studies suggest a positive relationshipbetween tree or dead-wood diameterand species richness, incidence, or abundance(Table 2), with larger-diametermaterial being especially importantfor rareand threatenedspecies (e.g., Warren& Key 1991). Variousex- planationsexist for this phenomenon.First, larger-diametertrees and dead wood are highly heterogeneous habitats, allowing many specialist species to occupy them at the same time (Kolstrom& Lumatjarvi2000). Second, larger-diameter SAPROXYLICINSECTS AND FORESTRY 5

BA45

15-

BA65

a 10- CL BA SNAG40 SNAG DW40 L0

E DW Z 5 CWD40

CWD

-5 - Figure 1 Numbersof saproxylicbeetle species showing significantcorrelations (p < 0.05; Pearsontwo-tailed) in abundancewith maturetimber habitat attributes acrossthe author's81 studyplots in theDaintree lowlands of northeasternQueensland, Australia.Based on all species representedby five or more individuals(n = 118). Correlationsabove the horizontalaxis (blackbars) are positive; correlations below the horizontalaxis (whitebars) are negative.Note that some species may contributeto morethan one column.BA, basalarea; BA45, basalarea of trees >45 cm dbh(diameter at 1.5 m); BA65, basalarea of trees >65 cm dbh;CWD, volume of coarsewoody debrison the forestfloor; CWD40, volume of CWD >40 cm dia (mid-pointdiameter); SNAG,volume of snags;SNAG40, volume of snags >40 cm dia;DW, total volume of deadwood; DW40, total volume of deadwood >40 cm dia. 6 GROVE

TABLE 2 Examples demonstratingrelationships between saproxylicinsects and mature timberhabitat diameter

Country Reference Relationship8 Observation

Australia Grove (2002b) + Larger-diameterliving trees a betterpredictor of saproxylicbeetle species richness and assemblagecomposition than total basal area in lowland tropicalrainforest Australia Yee et al. + Saproxylicbeetle fauna of rottinglogs variedby (2001) rot type, with some types primarilyfound in larger-diameterlogs Finland Kolstrom& + Model of forest-standstructure linked to database Lumatjarvi of saproxylicbeetle habitatrequirements (2000) Predictedthat retainingaspen trees in managed conifer standswould result in a progressive increasein saproxylicspecies as tree diameter increased.Thinning aspen along with conifers would not allow this increase Finland Vaisanen + Numberof species of subcorticalbeetles in trees et al. (1993) in a managedforest positively correlatedwith diameter Finland Siitonen & + Incidence of threatenedsaproxylic beetle Pytho Saaristo kolwensisin sprucelogs positively correlated (2000) with diameter Finland Martikainen + Numberof saproxylicbeetle species per plot et al. (2000) positively correlatedwith abundanceof larger- diameterliving and dead trees Sweden Jonsell + Models based on known habitatassociations of et al. (1998) threatenedsaproxylic invertebrates predicted that most would occur in dead wood in the largest diameterclass, including 178 not found in smaller-diameterclasses. Only 94 species would occur in the smallest, with only 13 of these not found in largerclasses Sweden Ranius & + Species richness of saproxylicbeetles in hollow Jansson oaks positively relatedto tree diameter,especially (2000) amongstthreatened species and those associated with sporocarpsof wood-rottingfungi Sweden T Nilsson + Probabilityof occurrenceof saproxylicbeetle (1997) Bolitophagusreticulatus increased with tree diameter Norway Thunes + Sporocarpsof wood-rottingfungus Fomitopsis et al. (2000) pinicola harboredmore saproxylicbeetle species on largertrees (Continued) SAPROXYLICINSECTS AND FORESTRY 7

TABLE 2 (Continued) Country Reference Relationship8Observation Germany Kleinevoss + Highersaproxylic beetle species richness on et al. (1996) larger-diameterdead wood United Elton(1966) -/+ Assemblagecomposition of smaller-diameter Kingdom deadwood intermediate between that of larger-diameterdead wood and leaf litter United Edmonds& + Higherincidence of thecerambycid beetle States Eglitis(1989) Monochamusscutellatus on larger-diameter Douglas-firlogs United Torgersen& + Camponotus(carpenter) ants prefer larger States Bull (1995) logs to smallerones United Hespenheide -/+ Somespecies of buprestidbeetle have particular States (1976) branch-diameterpreferences, whereas others makeuse of a widerange of diameters Japan Araya (1993) - The lucanidbeetle Prismognathusacuticollis occurredmore predictably in smaller-diameter logs a+, positive relationship;-, negativerelationship

pieces take longer to decompose, and a more stable microclimateis maintained within, favoring many species (Palm 1959). Third, larger-diameterpieces sup- port more species of fungi, including species specific to larger-diameterwood (Kruys& Jonsson 1999, Nordtn & Paltto2001), on which many saproxylicinsects depend.

DEAD WOOD IS DYNAMIC AND SO ARE ITS INHABITANTSDead-wood abundance partly depends on its rates of input and output (wood growth and decay) and whetherthese are in equilibrium.Equilibrium is most likely in those old-growth forests in which the normal disturbanceregime is one of small-scale gap replacement. If disturbancedynamics are more sporadic (e.g., occasional wild- fires, windstorms,or earthquakes),then dead wood abundancemay vary greatly dependingon the time since the last disturbanceevent (Siitonen 2001, Spies et al. 1988). A region's saproxylicinsect faunawill presumablybe pre-adaptedto its normal disturbanceregime (McPeek & Holt 1992). Species whose habitatis naturallylong- lived and/or abundantare usually poor dispersers(e.g., many specialists of tree hollows) (Nilsson & Baranowski 1997; Ranius & Hedin 2001) compared with those whose habitatis scarce and/or ephemeral(e.g., many scolytine beetles on wind- or fire-damagedtrees) (Nilssen 1984). Alterationof disturbancedynamics of a forest throughmanagement may have knock-oneffects on saproxylicinsects 8 GROVE

(Schiegg 2000). The most likely alterationis a reductionin matureforest stands, potentiallythreatening many species (Ranius2002) while turninga few othersinto pests (Heli6vaara& Vaiisanen1984, Safranyik& Linton 1999). Alternatively,fire suppressioncan lead to a reductionin early successional stages and an increasein standdensity (Johnsonet al. 2001), threateningmany fire-adapted, thermophilic or shade-intolerantsaproxylic insect species, which can comprise a large proportion of the regional fauna (0kland et al. 1996, Martikainen2001). Ecologists are familiarwith metapopulationdynamics operating over the scale of kilometersamong habitatpatches measured in hectares(Hanski 1999, Thomas 2000), but for some saproxylic insect species, metapopulationdynamics may operate among habitat patches measured in cubic meters (Table 3). Individual trees, logs, snags, or fungal sporocarpscan host several generationsof a particular species, but eventually the entire populationwill become extinct once the resource decomposes, while new substratepatches appearfor colonization elsewhere. If forest managementalters patch abundanceand distribution,then the populationdynamics of species dependenton them may break down, leading to landscape-levelextinction. For example, a common harvestingpractice in forests pre-adaptedto wildfiresis clearfelling.Although often justifiedon the groundsof mimickingnatural disturbance (Attiwill 1994), the resultingdead wood dynamics may be very different(Harmon et al. 1986, Price et al. 1998, Grove et al. 2002). Not only has much of the wood been harvested,but the remainderis fragmented and positioned entirely on the ground, where it may decay so rapidly that little remainswithin a few decades. Saproxylic insects may then face a critical habitat continuity gap, during which metapopulationdynamics might well break down, even thoughdead wood levels might subsequentlyincrease later in the silvicultural cycle.

CONNECTIONSIN SPACEAND TIME: CONTINUITY IS THE KEY One of the key managementissues for saproxylicinsects is the maintenanceof ecological continuity in space (connectivity) and in time. In much of western Europeecological continuity(or lack of it) is thoughtto be a major limiting factor for many of the less vagile saproxylic insect species (Speight 1989). Following millennia of intensive use (Thirgood 1989), there are probablyfew remainingpatches of forest that have experienced sufficient ecological continuity to supportanything close to the naturalcomplement of saproxylic insect species (see "SaproxylicInsects and the UnsustainableManagement of Forests,"below). This has led to Euro- pean conservationistspaying particularattention to old-forest species, which are confined to primaryforest, despite there being many mature secondary forests in the region (Speight 1989, Martikainenet al. 1999, Siitonen & Saaristo2000). In the United Kingdom a suite of such species can be used to discriminatebe- tween maturewoodland dating from the eighteenthcentury or earlierand the other 78% of the forest estate, which consists of more recently establishedplantations or regrowth(Fowles 1997). However, where forests are less fragmentedand/or SAPROXYLICINSECTS AND FORESTRY 9

TABLE 3 Examplesdemonstrating relationships between saproxylicinsects and availabilityof maturetimber habitat

Country Reference Observation

Australia Grove (2002b) Basal area of larger-diametertrees the best predictorof saproxylicbeetle abundanceand assemblagecomposition; coarse woody debrisvolume a betterpredictor of species richness in lowland tropicalrainforest Finland Martikainen Total volume of dead wood providedthe best predictorof et al. (2000) beetle species richness Finland Siitonen Specialist beetle species were more abundantin a forest (1994a) containingabundant dead wood than in one in which dead wood was scarce Finland Sippola & Positive relationshipbetween dead-woodvolume and Kallio (1995) beetle species richness at the stand(1 ha) scale Finland Komonen Numberof saproxylictrophic levels based on the bracket et al. (2000) fungus Fomitopsisrosea decreasedfrom three in intact old-growthforest to one in fragmentedold-growth forest Finland/ Siitonen On the Russian side of the borderdead aspen trees were Russia (1994b) common, but on the Finnish side they were rare.On the Russian side they supporteda much richerbeetle fauna than on the Finnish side, including2 species extinct in Finland and 18 species regardedas nationallyrare in Finland, comparedwith 5 rare species on the Finnish side Sweden Jonsell Fungusbeetles showed differingabilities to colonize et al. (1999) sporocarps,dependent on distance and the natureof the interveningland. Most of the species studiedcould persist in the managedforest landscapeif suitablebreeding substrate were createdor maintainedat the landscape(1 km2)scale. To some extent largerhabitat patches compensatedfor isolation Sweden Nilsson & Beetle species specializing on tree-hollowsin living trees Baranowski (a rarebut long-lived resource)less vagile than species living (1997) in dead wood on the forest floor (a commonerbut less persistentresource). Logged forest supportedall species living in dead wood on the forest floor but not severaltree- hollow species, even though the trees there were now large enough to supportsuitable hollows Norway Thunes The majorfactor influencingthe numberof species of beetles et al. (2000) and the numberof threatenedspecies on dead sporocarpsof the bracketfungus Fomitopsispinicola was the amountof dead wood in and aroundthe sampling site. Some of the species colonizing dead sporocarpswere also able to live in dead wood and only occurredon sporocarpswhen dead wood was abundant

(Continued) 10 GROVE

TABLE 3 (Continued) Country Reference Observation Norway Rukke Incidenceof beetlespecies associated with the bracket (2000) fungusFomesfomentarius related to degreeof isolation of habitattrees, as well as to habitatsize at treeand landscapescale Norway Sverdrup- Probabilityof occurrenceof thetenebrionid fungus-beetle Thygeson& Bolitophagusreticulatus increased with tree diameter Midtgaard andwith the number of deadsporocarps on thetree and (1998) decreasedas thedistance to surroundinginhabited treesincreased Norway 0kland Positiverelationship between dead-wood volume and beetle et al. (1996) speciesrichness at thelandscape (32 ha)but not the stand (1 ha) scale.At all scalesdiversity of deadtree parts, number of large-diameterdead trees and number of polyporefungi speciesall correlatedwith beetle species richness and with abundanceof manyspecies; several species absent below a certaindensity of deadwood Switzerland Schiegg Differentspecies of beetleand fly foundin forestswith (2000) differingdegrees of connectivityof deadwood; plots with higherdead-wood connectivity associated with higher speciesrichness Canada Kehler& Probabilityof occurrenceof thetenebrionid fungus-beetle Bondrup- Bolitotheruscornutus decreased with distance at the scale Nielsen of sporocarps,logs, andforest blocks in an agricultural (1999) matrix.At anyscale, distance proved the most consistent indicatorof isolation USA Chandler Greaterabundance of manybeetle species in old-growth (1991) comparedwith regrowth forest attributed to thegreater prevalenceof suitabledead-wood habitat in old growth

where occasional catastrophicdisturbance is the norm, the present availabilityof the right dead-woodhabitat may have greatersignificance for saproxylic species than does its continuityover time. This is held to be the case in the boreal forests of Fennoscandia(Kouki et al. 2001) and might also be trueelsewhere.

DEADWOOD IS NOT WASTEWOOD For classically trained foresters and for many members of the public, wood allowed to rot or burn is wood that has gone to waste ("zero waste tolerance")(Lofroth 1998). Yet in naturewaste recycling is a criticalecological process. Decomposingdead wood is the meansby which a large proportionof the nutrientsand energy accumulatedby the living tree is returned to the soil (Laiho & Prescott 1999). Dead wood also acts as a medium-termstore SAPROXYLICINSECTS AND FORESTRY 11

of carbon(Mackensen & Bauhus 1999). As decompositionprogresses dead wood becomes incorporatedinto soil, helping to maintainlevels of organic matterand carbon(Tate et al. 1993). Decompositionis broughtabout primarily by the activity of fungi andother microorganisms but is often mediatedby saproxylicinvertebrates (Swift 1977, Edmonds& Eglitis 1989, Schowalteret al. 1992, Hanula1996). Quite apartfrom their own rightsto existence, saproxylicinsects are a principalsource of food for otherforest-dwelling organisms such as woodpeckers(Hanula & Franzreb 1998, Torgersen& Bull 1995).

CONSERVATIONAND MANAGEMENTISSUES

Death by a Thousand Cuts:How Humans and Saproxylic Insects Compete for Trees Over the past few thousandyears humanshave increasinglyended up competing with saproxylic insects for timber (Hagan & Grove 1999). A small minority of saproxylicspecies such as some barkbeetles (Peltonen 1999, Weslien& Schroeder 1999) can compete successfully with humansand are thereforetermed pests. The vast majorityare more likely to end up on the losing side, resulting in managed forests supportingaltered saproxylic insect species assemblages, often coupled with lower numbersand fewer species overall (Table4). By managingforests so that we can remove wood before it is lost to decay, we progressively eliminate maturetimber habitat. Eliminationof maturetimber habitat may be most rapidand thoroughthrough excessive forest hygiene (Schmitt 1992), salvage logging (Maser 1996), or fuelwood or biomass harvesting(Tritton et al. 1987, Wall 1999, Grove et al. 2002). In other situationsits loss may be more insidious and may startwith a temporary increase in visible dead wood throughwasteful logging practices(Harmon 2001) or fire suppressionand pest outbreaksbrought on by management(Edmonds & Marra1999). However,with successive silviculturalcycles an incrementalloss is likely: As existing old trees die, the dead wood they produce decomposes, and the younger,more healthy trees of the managedforest do not live long enough to generatesufficient replacement habitat. Mature timber habitat may thusbe reduced from abundanceto rarity.Goodburn & Lorimer(1998) reportedthat levels of dead wood on the forest floor of selectively managed hardwoodforests in Wisconsin andMichigan were only about60% those of old-growthforests, with differencesin larger-diametermaterial being even more pronounced.Fennoscandian production forests provide a more extreme example, with coarse woody debris levels falling by 90-98% over the past century (Siitonen 2001). In other parts of Europe it is not even possible to say how much dead wood an old-growthforest would have had, because there are none left. Writingfrom a Britishperspective, Elton (1966) statedthat "if fallen timberand slightly decayed trees areremoved [from a natural forest]the whole system is gravelyimpoverished of perhapsmore thana fifth of its fauna."If anything,this is likely an underestimate,as the studiesin Table 1 suggest. 12 GROVE

TABLE 4 Examples of relationshipsbetween saproxylicinsects and forest managementa

Country Reference Nb Sc Cd Observations

Australia Grove + + + Analyses of data concerning339 2002a saproxylicbeetle species in old-growth, selectively logged and regrowthtropical rainforest.Differences subtle but consistent Australia Michaels & + + NR Lucanidbeetle species richness and Bomemissza abundancelower in regrowthforest than 1999 in recent clear-fells owing to continued presence of old-growthlogs in clear- felled areas Finland Martikainen + + + 78% of 232 saproxylicbeetle species et al. 2000 more abundantin old-growththan in managedmature forest; almost no overlapin assemblagecomposition Finland Martikainen + + NR Scolytine beetles in old-growth, et al. 1999 overmature,and managedmature forest Finland Vaiisanen NR NR + Among saproxylicbeetles living et al. 1993 subcorticallyin dead trees, proportion of scolytines much higher in managed forest;proportion of rare species higher in old growth Poland Gutowski + 0 + Cerambycidbeetles in old-growthversus 1986 managedforest. More species present only in old-growththan only in managedforest Germany Schmitt NR + + Recordedmore species of saproxylic 1992 beetle in logs and more threatened species in unmanagedthan in adjacent managedforest. Managedforest assemblagesmore dominated by scolytines Canada Spence + + + Several saproxylicbeetle species found et al. 1996 in old-growththat were rareor absent in maturemanaged forest United Chandler + + NR More species and individualsof leiodid States & Peck beetles in old-growthversus "regrowth" 1992 (formerclear-cut) forest United Chandler + NR NR More individualsof saproxylicbeetles States 1991 in old-growthversus regrowthforest -[ a Relationshipsare expressedwith respectto "old-growthness";+, positive relationship;0, no relationship;NR, not reported b N, abundanceof individuals c S, species richness d C, assemblagecomposition SAPROXYLICINSECTS AND FORESTRY 13

Referringto NorthAmerican forests, Huston(1996) stated,"No othermanageable propertyof the forest environmenthas a greaterimpact on biodiversitythan coarse woody debris. Even [timber]harvesting... probablyhas a greatereffect on total forest biodiversitythrough the alterationand removalof coarse woody debristhan throughits effect as a disturbancethat 'resets' forest succession." The situation may be equally critical wherevernative forests are heavily ex- ploited for fuelwood [e.g., China (Nalepa et al. 2001), South Africa (DuPlessis 1995), Australia (Driscoll et al. 2000)]. It may also be comparableto parts of the eastern United States, where the extensive mature forests are the result of "old-field"succession and are less than 200 years old. Within my study area in northeasternQueensland, the volume of dead wood on the forest floor in selectively logged tropical rainforestaveraged about 80% that of nearby old growth, with much greater deficiencies among larger-diametermaterial (Grove 2001b). Differenceswere more markedamong living trees, and in these forests it appeared to be the loss of larger-diametertrees from logged forest that was the main cause of the differencesin the saproxylicbeetle fauna (Grove 2002b) (Figure2). This is likely also the case in forests managedby selection silvicultureelsewhere.

SaproxylicInsects and the UnsustainableManagement of Forests:A 5000-YearEuropean Experiment The historyof forest use andabuse in westernEurope is not uniquebut is well documented. Forests had scarcely reached their maximumpost-glacial extent when farmersstarted clearing them. Over the following millenniaforest cover was dras- tically reduced,and the structureand compositionof remainingfragments greatly altered(Thirgood 1989). By 1000 A.D., therewas probablyno truly naturalforest left in Europeoutside Fennoscandia(Greig 1982). It was not just wolves, bears, and lynx that retreated.As forests were cleared and the remainingfragments were managedfor firewood,poles, and timber,mature timber habitat and saproxylic insects also began to vanish (Speight 1989). Theirprehistoric extinction in the United Kingdomis well documented.Buckland & Dinnin (1993) list 17 saproxylicbeetle species known in the United Kingdom only from subfossils, mostly from peat deposits datingfrom about2900 B.C., and speculate that the list will grow with furtherexcavations. None of these species is yet globally extinct, but most now survive only in tiny refugia elsewhere in Europe.Many belong to genera containingspecies that are currentlycommon in other parts of the world (e.g., Rhysodes, Prostomis, Pycnomerus,Dromaeolus, Platycerus, Cerambyx,and Eremotes) and provide a clear reminderof what any region standsto lose if the processes that operatedin westernEurope are repeated elsewhere. The 5000-year-longtrend in forest loss in westernEurope has now largelybeen reversed,with the expansionof secondarywoodlands and plantations. However, the extinction of saproxylic insects continues apace. For instance, Hammond(1974) reportedthe loss from the United Kingdom of a further20 beetle species over 14 GROVE

Reduced Reduced saproxylic saproxylic insect insectspecies abundance richness

Reduced Reduced dead wood Logging basalarea, volumes,especially and especially larger-diameter clearanc of larger- size-classes diameter trees

Altered Altered saproxylic saproxylic insectassemblage insectguild composition composition

Figure 2 Likely impactsof logging and forest clearanceon maturetimber habitat, and consequencesfor the saproxylicinsect fauna.Based on the main findingsof the author's researchon saproxylicbeetles in northeasternQueensland. The thicknessof each arrowis approximatelyproportional to the strengthof the relationshipas gaugedby correlations, analysesof variance,or multivariateanalyses as discussedin Grove(2002b). Relationships concerningguild composition are derived from the author'sunpublished data.

the previous two centuries, and a fifth of the remainingUK cerambycidbeetle fauna is now consideredthreatened (Twinn & Harding1999). Todayextinction is mostly drivenby increasinglyintensive forest management(Hanski & Hammond 1995, Siitonen 2001) coupled with the delayed effects of past fragmentation.In Fennoscandia,fire suppressionand no-burn silviculture is one of the mainthreats to saproxylicspecies (Martikainen2001, Johnsonet al. 2001). Saproxylicinsects now comprise a disproportionatelylarge percentageof nationallyrare and threatened species in Europe (Geiser 1983, Shirt 1987). For instance, in Sweden, of the 739 threatenedforest invertebratespecies, old living trees are considereda key habitat for 33%, logs 28%, and snags 35% (Berg et al. 1994). The removal of mature timberhabitat is consideredthe main threatfor 65% of the United Kingdom's 150 threatenedwoodland insect species (Hambler& Speight 1995). Many saproxylic species now survive in Europeonly as relictualpopulations, "hangingon by the tips of theirtarsi" (Buckland & Dinnin 1993) in small patches SAPROXYLICINSECTS AND FORESTRY 15

of forest or pasturewoodland-or even single trees-which for historicalreasons did not experiencethe same loss of maturetimber habitat. In the absenceof positive management,the ultimateextinction of some such species (trulythe "livingdead") is almost inevitablethrough stochastic events (Thomas2000).

Silviculture,Morticulture, and the Living Dead: Catering for SaproxylicInsects in ManagedForests The forestryprofession is increasinglyaware of its responsibilitiestowards dead wood management["morticulture" (Harmon 2001)], especially in North America (Franklin 1989) and northernEurope (Siitonen 2001). Done well, morticulture should cater for saproxylic insects. Whereas it is unrealisticto expect even sus- tainably managed commercial forests to retain a full complement of saproxylic insect species at the stand level, they ought to be able to at the landscape level. Some brief suggestions for how to do so follow. At the landscapelevel the bottom line is that managementshould retain sufficient elements of naturalforest dynamics to maintainthe ecological processes in which maturetimber habitatis involved-especially the chance for trees to live to ecological maturityand decay without intervention.Sufficient trees or patches should be retainedto allow for some landscape-leveltemporal and spatialhabitat continuity(Martikainen 2001, Grove et al. 2002). Managementshould also allow for the occasional occurrenceof pest and disease outbreaks,windthrow, and fire if these arenormal disturbance events. At the standlevel thereare many measures that can be taken(e.g., Grove2001a, Ehnstr6m2001). For all forest types these should include using reduced-impactlogging techniques,avoiding incidentaldamage to logs and snags, and avoidingharvesting of woody debris.Additional measures for selectively logged forests should include retaininghabitat trees and legacy trees and avoiding silviculturalrefinement. Additional measures for forests subjected to clearfelling should include extending rotationslong enough for maturetimber habitatto accumulatebetween felling cycles (at least over part of the production forest estate), a preferencefor small logging coupes and/oraggregated retention, and separatingconsecutive logging coupes to ensure that there is always some matureforest in the productionforest matrix. In regions where species are regionallyextinct througha history of forest frag- mentationand intensive use (as in western Europe),reintroductions are the only option. If they still surviveas "living dead"in refugia,then morticulturalmanage- ment of refugia and the surroundingmatrix may yet save them from extinction. Managingrefugia is perhapsbest developed in the United Kingdom, where most of the remainingkey sites are culturallandscapes (such as royal huntingparks), now owned by nature conservationorganizations (Alexander 1995), allowing a level of intensive care that would be uneconomicalfor commercialforestry. An example is the reinstatementof pollarding(Read 1991). Originallycarried out to providea supplyof pole-wood above the reachof browsinganimals, it incidentally helps prolong tree life, providingthe necessary maturetimber habitat and habitat 16 GROVE

continuity for a wide range of saproxylic insects and other organisms.In North America other suggestions for enhancing mature timber habitat in production forests have included the use of explosives (Bull & Partridge1986) to produce snags from living trees, attractingbark beetles to living trees with pheromonesto produce snags for "wildlife"(Ross & Niwa 1997), and the felling of "cull"trees to supplementcoarse woody debris (Franklin1989). Killing trees with herbicides has also been promoted,but the resultantsnags provide much poorer conditions for saproxylicinsects thando those resultingfrom natural tree death(Aulen 1991).

ManagementIndicators: The Way Forwardfor SustainingSaproxylic Insects? In Europeenough is now known about saproxylicinsects for theirpotential use in identifyingkey sites for natureconservation or for monitoringthe sustainabilityof forest management.Many taxonomic groups have been promotedfor this purpose, especially beetles (Harding& Rose 1986, Harding& Alexander 1994, Nilsson et al. 1995). However, their use still remains the domain of entomologists and academics, ratherthan forestry practitioners.No such level of knowledge-or indeed interest-appears to exist in any other part of the world, so their chances of being widely adoptedas indicatorsare slim. As an alternative,mature timber habitatcould well provide useful structural surrogatesfor saproxylicinsects (e.g., Thuneset al. 2000), and incorporatingsuch surrogatesinto guidelines, criteria,and indicatorsis perhapsour best chance of addressingthe needs of saproxylicinsects worldwide.One advantageis thathabitat is more readily measurablethan are the saproxylicinsects themselves (Hodge & Peterken 1998). Measuringthe basal area or standingvolume of larger-diameter living trees is a simple procedureand merely an extension of a standardforestry activity. Snags and dead wood are slightly more difficult to measure accurately owing to their naturalpatchiness (Siitonen 1994a, Sippola et al. 1998, Grove 2001b). Currently,most initiativesassume that snags and dead wood on the forest floor are the key features.This may be so in forests where occasional large-scale disturbancesare the norm,but as this review has demonstrated,it is not the whole picture,especially in forests where smaller-scaledisturbances operate. Europeanguidelines for certificationand sustainablemanagement increasingly promotethe retentionand maintenanceof dead wood and old trees (UKFC 1998, PEFC 1999). The level of managementconcern for coarse woody debris, snags, and dependentvertebrates in NorthAmerica (Fenger 1996, Hagan& Grove 1999, Harmon2001) and Australia(Gibbons & Lindenmayer2001, Woldendorpet al. 2001) provides some hope that the needs of some saproxylic insects will be addressed by default. In the tropics nearly all available sustainableforest management guidelines are too general to refer specifically to the maturetimber habitat, with one main exception (CIFOR 1999). Much work remains to be done worldwide to develop more appropriateguidelines and to get them implemented(Grove 2001a). SAPROXYLICINSECTS AND FORESTRY 17

CONCLUSION

FromEuropean studies it is clear that saproxylicinsects are peculiarlysusceptible to forest management,in no small part owing to their dependence on a suite of habitatsthat is naturallyabundant in unmanagedforest but often rare in long- managedforests. Many species may havelow powersof dispersal,with populations governed by metapopulationdynamics. These characteristicsmake evolutionary sense, but do not favor survivalin managedand fragmentedforest settings, where spatial and temporalbreaks in habitatcontinuity can lead to populationdeclines and extinctions. I hope this reviewhas demonstratedthat saproxylic insects deservemuch greater considerationamongst forest managers and their ecologist advisers. Sustaining saproxylic insects may not be any more challenging than sustainingother forest biota, but unless theirneeds are specificallyaddressed there remains a seriousrisk that we will lose a large proportionof the forest fauna from our managedforests before we even know it. One only has to look at westernEurope to see what might happenif we fail to act.

ACKNOWLEDGMENTS

I thank the CooperativeResearch Centre for Tropical Rainforest Ecology and Management, James Cook University, and the Australian Tropical Research Foundationfor funding my doctoralresearch in northeasternQueensland, which formedthe basis of this review.I also thankChristine Nalepa, Donna Scheungrab, Tim Schowalter,Roger Sandquist,and ChristineNiwa for following up on my requests for advice on the status of research on saproxylic insects in North America. I thank Petri Martikainen,Anne Sverdrup-Thygeson,Jim Hammond, and Marie Yee for their helpful comments on a previous version of this manuscript.

The Annual Reviewof Ecology and Systematicsis online at http://ecolsys.annualreviews.org

LITERATURE CITED

Alexander KNA. 1995. Historic parks and ous forest fragments in boreal forest. Eco- pasture-woodlands:the National Trust re- graphy 16:219-28 source and its conservation.Biol. J. Linn. Attiwill PM. 1994. Ecological disturbanceand Soc. London 56 (Suppl.):155-75 the conservation management of eucalypt Araya TE. 1993. Relationshipbetween decay forests in Australia. For. Ecol. Manage. types of dead wood and occurrence of lu- 63:301-46 canid beetles (Coleoptera:Lucanidae). Appl. Aulen G. 1991. Increasing insect abundance Entomol.Zool. 28:27-33 by killing deciduoustrees-a method of im- As S. 1993. Are habitat islands islands? proving the food situation for endangered Woodliving beetles (Coleoptera)in decidu- woodpeckers.Holarct. Ecol. 14:68-80 18 GROVE

Berg A, EhnstromB, Gustafsson L, Halling- fallen beech wood]. Z. Morphol. Okol. Tier. back T, Jonsell M, Weslien J. 1993. Threat- 37:683-734. In German ened plant, animal and fungus species in Deyrup MA. 1976. The insect community Swedish forests: distributionand habitatas- of dead and dying Douglas-fir: Diptera, sociations. Conserv.Biol. 8:718-31 Coleoptera, and Neuroptera. PhD thesis. Blab J, Nowak E, TrautmannW, Sukopp H. Univ. Seattle 1994. Rote Liste der gefiihrdetenTiere und DriscollD, MikovitsG, FreudenbergerD. 2000. Pflanzenin derBundesrepublikDeutschland. Impact and Use of Firewood in Australia. Greven,Ger: Kilda Canberra,ACT: CSIRO Sustainable Ecosys- BlackmanMW,Stage HH. 1924. On the succes- tems sion of insects living in the barkand wood of DuPlessis MA. 1995. The effects of fuelwood dying, deadand decaying hickory. Tech. Bull. removalon the diversityof some cavity-using NY State Coll. For. 17:3-268 birds and mammals in South Africa. Biol. Braccia A, Batzer DP. 2001. Invertebratesas- Conserv.74:77-82 sociated with woody debris in a southeast- EdmondsRL, Eglitis A. 1989. The role of the ernUS forestedfloodplain wetland. Wetlands Douglas-fir beetle and wood borers in the 21:18-31 decompositionof and nutrientrelease from Buckland PC, Dinnin MH. 1993. Holocene Douglas-fir logs. Can. J. For. Res. 19:853- woodlands, the fossil insect evidence. In 59 Dead WoodMatters: The Ecology and Con- EdmondsRL, MarraJL. 1999. Decomposition servation of Saproxylic Invertebrates in of woody material:nutrient dynamics, inver- Britain, ed. KJ Kirby,CM Drake, pp. 6-20. tebrate/fungirelationships and management Peterborough,UK: English Nat. in Northwest forests. In Proc. Pac. North- Bull EL, Partridge AD. 1986. Methods of west For. Rangeland Soil OrganismSymp., killing trees for use by cavity nesters. Wildl. ed. RT Meurisse,WG Ypsilantis,C Seybold, Soc. Bull. 14:142-46 pp. 68-79. Corvallis, OR: USDA For. Ser. Chandler DS. 1991. Comparison of some Gen. Tech. Rep. PNW-GTR461 slime-mold and fungus feeding beetles EhnstromB. 2001. Leaving dead wood for in- (Coleoptera)in an old-growthand 40-year- sects in boreal forests-suggestions for the old forest in New Hampshire.Coleopt. Bull. future.Scand. J. For.Res. Suppl. 3:91-98 45:239-56 Elton C. 1966. Dying and dead wood. In The ChandlerDS, Peck SB. 1992. Diversityand sea- Patternof AnimalCommunities, ed. C Elton, sonalityof leiodid beetles (Coleoptera:Leio- pp. 279-305. London:Methuen didae) in an old-growth and a 40-year-old Fenger M. 1996. Implementing biodiversity forest in New Hampshire.Environ. Entomol. conservationthrough the British Columbia 21:1283-91 Forest Practices Code. For. Ecol. Manage. CIFOR. 1999. C&I ToolboxSeries No. 2: The 85:67-77 CIFORCriteria andIndicators Generic Tem- Fowles AP. 1997. The Saproxylic Quality In- plate. Bogor, Indonesia:Cent. Int. For. Res. dex: an evaluation of dead wood habitats 53 pp. based on rarity scores, with examples from Dajoz R. 1966. Ecologie et biologie des Wales. Coleopterist6:61-66 col6opteresxylophages de la hetraie. [Ecol- FranklinJ. 1989. Towarda new forestry.Am. ogy and biology of xylophagous beetles of For. 95:37-44 the beechwood] Vie Milieu 17:525-636 In Geiser R. 1983. Rote Liste ausgewihlter Frenchwith English summ. Familien xylobionter Kafer (Coleoptera)in Derksen W. 1941. Die Succession der ptery- Osterreich[Red list of threatenedxylobiont goten Insekten im abgestorbenBuchenholz beetle families (Coleoptera)in Austria] (In [The succession of pterygote insects in German). In Rote Listen GefahrdeterTiere SAPROXYLICINSECTS AND FORESTRY 19

[RedListsof ThreatenedSpecies], ed. J Gepp, chetosum association in Bialowieza Forest pp. 131-37. Wien, Austria: Osterr. Bun- (NE Poland).J. Appl. Entomol. 102:380-91 desminist. Gesundh.Umweltschutz HaganJM, Grove SL. 1999. Coarsewoody de- Gibbons P, LindenmayerD. 2001. Tree Hol- bris:humans and naturecompeting for trees. lows and WildlifeConservation in Australia. J. For.97:6-11 Canberra:CSIRO. 240 pp. Hambler C, Speight MR. 1995. Biodiversity GoodbumJM, LorimerCG. 1998. Cavity trees conservation in Britain-science replacing and coarse woody debris in old growth tradition.Br. Wildl.6:137-47 and managed northernhardwood forests in Hamilton WD. 1978. Evolution and diversity Wisconsin and Michigan. Can. J. For. Res. underbark. In DiversityofInsect Faunas, ed. 28:427-38 LA Mound, N Waloff, pp. 154-75. London: Greig J. 1982. Past and present limewoods in R. Entomol. Soc. London Europe.In ArchaeologicalAspects of Wood- Hammond HEJ. 1997. Arthropod biodiver- land Ecology, ed. M Bell, S Limbrey,pp. 89- sity from Populus coarse woody mate- 99. Oxford:Br. Archaeol.Rep. rial in north-centralAlberta-a review of Grove SJ. 2001a. Developing appropriate taxa and collection methods. Can. Entomol. mechanisms for sustaining mature timber 129:1009-33 habitatin managednatural forest stands.Int. Hammond HEJ, Langor DW, Spence JRS. For.Rev. 3:272-84 2001. Earlycolonization of Populuswood by Grove SJ. 2001b. Extent and composition of saproxylicbeetles. Can.J. For.Res. 31:1175- dead wood in Australian lowland tropical 83 rainforestwith differentmanagement histo- Hammond PM. 1974. Changes in the British ries. For.Ecol. Manage. 154:35-53 coleopterous fauna. In The Changing Flora Grove SJ. 2002a. The influence of forest man- and Fauna of Britain, ed. DL Hawksworth, agement history on the integrity of the sap- pp. 323-69. London:Academic roxylic beetle fauna in an Australian low- HammondPM. 1990. Insect abundanceand di- land tropical rainforest.Biol. Conserv. 104: versity in the Dumoga-Bone National Park, 149-71 N. Sulawesi, with special reference to the Grove SJ. 2002b. Tree basal area and dead beetle fauna of lowland rainforest in the wood as surrogateindicators of saproxylic Torautregion. In Insects and the RainForests insect faunalintegrity: a case study from the of South East Asia (Wallacea), ed. WJ Australianlowland tropics. Ecol. Indicators Knight,JD Holloway,pp. 197-254. London: 1:171-88 R. Entomol. Soc. London GroveSJ, Meggs J, GoodwinA. 2002. A review HammondPM, Kitching RL, Stork NE. 1996. of biodiversityconservation issues relating The composition and richness of the tree- to coarse woody debris managementin the crown Coleoptera assemblage in an Aus- wet eucalypt productionforests of Tasma- tralian subtropicalforest. Ecotropica 2:99- nia. Hobart,Aust.: For.Tasmania Tech. Rep. 108 22:1-72 Hanski I. 1999. MetapopulationEcology. Ox- Grove SJ, Stork NE. 1999. The conservation ford: Oxford Univ. Press. 313 pp. of saproxylicinsects in tropicalforests: a re- Hanski I, HammondPM. 1995. Biodiversityin searchagenda. J. Insect Conserv.3:67-74 boreal forests. TrendsEcol. Evol. 10:5-6 Grove SJ, StorkNE. 2000. An inordinatefond- Hanula JL. 1996. Relationship of wood- ness for beetles. Invertebr.Taxon. 14:733-39 feeding insects and coarse woodydebris. See Gutowski JM. 1986. Species composition and McMinn & Crossley 1996, pp. 55-81 structure of the communities of longhorn HanulaJL, FranzrebK. 1998. Source, distribu- beetles (Col., Cerambycidae)in virgin and tion and abundanceof macroarthropodson managed stands of Tilio-Carpinetumsta- the barkof longleafpine: potential prey of the 20 GROVE

red-cockadedwoodpecker. For. Ecol. Man- on dead birch trunks decayed by different age. 102:89-102 polypore species. Ann. Zool. Fenn. 31:97- Harding PT, Alexander KNA. 1994. The use 107 of saproxylic invertebratesin the selection KehlerD, Bondrup-NielsenS. 1999. Effects of and conservationof relic forest in pasture- isolation on the occurrenceof a fungivorous woodland. Br. J. Entomol. Nat. Hist. 7:21- forestbeetle, Bolitotherus cornutus, at differ- 26 ent spatialscales in fragmentedand continu- HardingPT, Rose F. 1986. Pasture-Woodlands ous forests. Oikos 84:35-43 in Lowland Britain. Huntingdon,UK: Inst. KleinevossK, ToppW, Bohac J. 1996. Buchen- Terr.Ecol. 89 pp. Totholz im Wirtschaftswaldals Lebensraum Harmon ME. 2001. Moving towards a new fir xylobionte Insekten [Dead beech wood paradigm for woody detritus management. in the commercialforest as habitatfor xylo- Ecol. Bull. 49:269-78 biont insects]. Z. Okol. Nat.schutz 5:85-95. HarmonME, FranklinJF, Swanson FJ, Sollins In German P, GregorySV, et al. 1986. Ecology of coarse KleteckaZ. 1996. The xylophagousbeetles (In- woody debris in temperateecosystems. Adv. secta,Coleoptera) community and its succes- Ecol. Res. 15:133-302 sion on Scotch elm (Ulmusglabra) branches. Heliovaara K, Vaisanen R. 1984. Effects of Biologia (Bratislava)51:143-52 modem forestry on northwesternEuropean Kohler F. 2000. Totholzkaferin Naturwaldzel- forest invertebrates:a synthesis. Acta For. len des nordlichenRheinlandes. Vergleich- Fenn. 189:5-30 ende Studienzur Totholzkdferfauna Deutsch- HespenheideHA. 1976. Patternsin the use of lands und deutschen Naturwaldforschung single plant hosts by wood-boring beetles. [Saproxylic Beetles in Nature Forests of Oikos 27:61-64 the NorthernRhineland. Comparative Stud- Hodge SJ, Peterken GF. 1998. Deadwood in ies on the Saproxylic Beetles of Germany British forests: priorities and a strategy. and Contributionsto German Nature For- Forestry71:99-112 est Research]. Recklinghausen:Landesamt Howden HF, Vogt GB. 1951. Insect communi- AgrarordnungNRW. 351 pp. In German ties of standingdead pine (Pinus virginiana Kolstrom M, LumatjarviJ. 2000. Saproxylic Mill.). Ann. Entomol.Soc. Am. 44:581-95 beetles on aspen in commercial forests: a HustonMA. 1996. Modelingand management simulationapproach to species richness.For. implicationsof coarse woody debris impacts Ecol. Manage. 126:113-20 on biodiversity. See McMinn & Crossley, KomonenA, PenttilaR, LindgrenM, HanskiI. 1996, pp. 55-81 2000. Forest fragmentationtruncates a food Johnson EA, Miyanishi K, Bridge SRJ. 2001. chain based on an old-growthforest bracket Wildfireregime in the boreal forest and the fungus. Oikos 90:119-26 idea of suppressionand fuel buildup. Con- Kouki J, Lofman S, MartikainenP, Rouvi- serv.Biol. 15:1554-57 nen S, Uotila A. 2001. Forest fragmenta- Jonsell M, NordlanderG, Jonsson M. 1999. tion in Fennoscandia:linking habitat require- Colonization patterns of insects breeding ments of wood-associatedthreatened species in wood-decaying fungi. J. Insect Conserv. to landscape and habitatchanges. Scand. J. 3:145-61 For.Res. Suppl. 3:27-37 Jonsell M, Weslien J, EhnstromB. 1998. Sub- Kruys N, Jonsson BG. 1999. Fine woody de- strate requirementsof red-listed saproxylic bris is importantfor species richness on logs invertebratesin Sweden.Biodivers. Conserv. in managedboreal spruceforests of northern 7:749-64 Sweden. Can. J. For.Res. 29:1295-99 Kaila L, MartikainenP, Punttila P, Yakovlev LaihoR, PrescottCE. 1999. The contributionof E. 1994. Saproxylic beetles (Coleoptera) coarse woody debristo carbon,nitrogen, and SAPROXYLICINSECTS AND FORESTRY 21

phosphoruscycles in three Rocky Mountain Scolytidae and Curculionidae)in northern coniferousforests. Can. J. For.Res. 29:1592- Finland.Ann. Ent. Fenn. 50:37-42 1603 Nilsson SG. 1997. Grynocharis oblonga L. LofrothE. 1998. The dead wood cycle. In Con- (Coleoptera: Trogossitidae): a specialized servation Biology Principles for Forested wood beetle with a relict distribution.Ento- Landscapes, ed. J Voller, S Harrison, pp. mol. Tidskr.118:1-9. In Swedish with En- 185-214. Vancouver,Can.: UBC Press glish summ. Mackensen J, Bauhus J. 1999. The Decay Nilsson SG, ArupU, BaranowskiR, EkmanS. of Coarse WoodyDebris. Canberra,Aust.: 1995. Tree-dependentlichens and beetles as Aust. GreenhouseOff. 41 pp. indicatorsin conservationforests. Conserv. MartikainenP. 2001. Conservationof threat- Biol. 9:1208-15 ened saproxylic beetles: significance of re- Nilsson SG, BaranowskiR. 1994. Indikatorer tained aspenPopulus tremulaon clearcutar- pa jattetradskontinuitet-svenskafirekom- eas. Ecol. Bull. 49:205-18 ster av knapparesom ar beroende av grova, MartikainenP, Siitonen J, Kaila L, Punttila levande trad [Indicatorsof megatree conti- P, Rauh J. 1999. Bark beetles (Coleoptera, nuity-Swedish distributionof click beetles Scolytidae) and associated beetle species (Coleoptera,Elateridae) dependent on hol- in mature managed and old-growth boreal low trees]. Entomol. Tidskr.115:81-97. In forests in southernFinland. For. Ecol. Man- Swedish with English summ. age. 116:233-45 Nilsson SG, BaranowskiR. 1997. Habitatpre- MartikainenP, Siitonen J, PunttilaP, Kaila L, dictabilityand the occurrenceof wood bee- RauhJ. 2000. Species richnessof Coleoptera tles in old-growthbeech forests. Ecography in mature managed and old-growth boreal 20:491-98 forests in southernFinland. Biol. Conserv. Nilsson T. 1997. Spatial population dynamics 94:199-209 of the black tinderfungus beetle, Bolitopha- Maser C. 1996. Salvage logging-the loss of gus reticulatus(Coleoptera: Tenebrionidae). ecological reason and moral restraint.Int. J. PhD thesis. Univ. Uppsala, Sweden. 44 pp. Ecofor. 12:176-78 Norden B, Paltto H. 2001. Wood-decay fungi McMinn J, Crossley DA, eds. 1996. Biodiver- in hazel wood: species richnesscorrelated to sity and Coarse WoodyDebris in Southern standage anddead wood features.Biol. Con- Forests. Athens GA: USDA For. Ser. Gen. serv. 101:1-8 Tech. Rep. SE-94 0kland B, Bakke A, Hagvar S, Kvamme T. McPeek MA, Holt RD. 1992. The evolutionof 1996. What factors influence the diversity dispersalin spatiallyand temporallyvarying of saproxylic beetles? A multiscaled study environments.Am. Nat. 140:010-27 froma spruceforest in southernNorway. Bio- Michaels K, Bornemissza G. 1999. Impact divers. Conserv.5:75-100 of clearfell harvesting on lucanid beetles Palm T. 1959. Die Holz- und Rindenkafer (Coleoptera:Lucanidae) in wet and dry scle- dersiid- und mittelschwedischen Laubbaume rophyll forests in Tasmania.J. Insect Con- [The wood- and bark-beetlesof south and serv. 3:85-95 mid-Swedishbroadleaves]. Opusc. Entomol. Nalepa CA, Li LI, Wen-HuaLU, Lazell J. 2001. 16 (Suppl.):1-374. In German Rediscovery of the wood-eating cockroach ParkerSP. 1982. Synopsisand Classificationof Cryptocercusprimarius (Dictyoptera:Cryp- Living Organisms.New York:McGraw-Hill. tocercidae)in China, with notes on ecology 1232 pp. anddistribution. Acta Zootaxon. Sin. 26:184- PEFC. 1999. Pan Eur. For. Certif. Website: 90 http://www.pefc.org/ Nilssen AC. 1984. Long-rangeaerial dispersal Peltonen M. 1999. Windthrows and dead- of barkbeetles andbark weevils (Coleoptera, standing trees as bark-beetle breeding 22 GROVE

materialat forest-clearcutedge. Scand.J. For. Insects. Peterborough,UK: Nat. Conserv. Res. 14:505-11 Counc. PriceK, PojarJ, RoburnA, BrewerL, PoirierN. Siitonen J. 1994a. Decaying wood and saprox- 1998. Windthrownor clearcut-what's the ylic Coleopterain two old spruce forests: a difference?Northwest Sci. 72:30-32 comparisonbased on two samplingmethods. RaniusT. 2002. Influenceof standsize andqual- Ann. Zool. Fenn. 31:89-95 ity of tree hollows on saproxylic beetles in Siitonen J. 1994b. Occurrence of rare and Sweden. Biol. Conserv. 103:85-91 threatenedinsects living on decaying Pop- Ranius T, Hedin J. 2001. The dispersalrate of ulus tremula:a comparisonbetween Finnish a beetle, Osmodermaeremita, living in tree and Russian Karelia. Scand. J. For. Res. 9: hollows. Oecologia 126:363-70 185-91 RaniusT, JanssonN. 2000. The influenceof for- Siitonen J. 2001. Forest management,coarse est regrowth,original canopy cover and tree woody debris and saproxylic organisms: size on saproxylicbeetles associatedwith old Fennoscandianboreal forests as an example. oaks. Biol. Conserv.95:85-94 Ecol. Bull. 49:11-42 Read HJ. 1991. Pollardand VeteranTree Man- Siitonen J, Saaristo L. 2000. Habitat require- agement.London: Corp. London. 60 pp. ments and conservationof Pytho kolwensis, Ross DW, Niwa CG. 1997. Using aggre- a beetle species of old-growthboreal forest. gation and antiaggregationpheromones of Biol. Conserv.94:211-20 the Douglas-fir beetle to produce snags for Sippola A, Kallio R. 1995. Species composi- wildlife habitat. West.J. Appl. For. 12:52- tion of beetles (Coleoptera)in differenthabi- 54 tats within old-growthand managed forests Rukke BA. 2000. Effects of habitat fragmen- in Finnish Lapland.In Northern Wilderness tation:increased isolation and reducedhabi- Areas:Ecology, Sustainability, Management, tat size reduces the incidence of dead wood ed. A Sippola,P Alaraudanjoki,B Forbes,V fungi beetles in a fragmented forest land- Hallikainen,pp. 59-77. Rovaniemi,Finland: scape. Ecography23:492-502 Arctic Cent. SafranyikL, Linton DA. 1999. Spruce beetle Sippola A, Siitonen J, Kallio R. 1998. Amount (Coleoptera:Scolytidae) survivalin stumps and quality of coarse woody debris in natu- and windfall. Can. Entomol. 131:107-13 ral and managedconiferous forests near the Savely HE. 1939. Ecological relations of cer- timberlinein FinnishLapland. Scand. J. For. tain animalsin dead pine and oak logs. Ecol. Res. 13:204-14 Monogr.9:321-85 Speight MCD. 1989. SaproxylicInvertebrates Schiegg K. 2000. Effects of dead wood volume and their Conservation. Strasbourg, Fr: and connectivityon saproxylicinsect species Counc. Eur.79 pp. diversity.Ecoscience 7:290-98 SpenceJR, Langor DW, Niemela J, CarcamoH, Schmitt M. 1992. Buchen-Totholzals Lebens- CurrieCR. 1996. Northernforestry and cara- raumfiir xylobionte Kifer [Deadbeech wood bids: the case for concern about old-growth as habitat for xylobiont beetles]. Waldhy- species. Ann. Zool. Fenn. 33:173-84 giene 19:97-191. In German Spies TA, Franklin JF, Thomas TB. 1988. Schowalter TD, Caldwell BA, CarpenterSE, Coarse woody debris in Douglas-fir forests GriffithsRP, Harmon ME, et al. 1992. De- of westernOregon and Washington. Ecology composition of fallen trees: effects of ini- 69:1689-702 tial conditions and heterotrophcolonization Sverdrup-ThygesonA, Midtgaard F. 1998. rates. In TropicalEcosystems: Ecology and Fungus-infectedtrees as islandsin borealfor- Management,ed. KP Singh, JS Singh, pp. est: spatial distributionof the fungivorous 373-83. New Delhi, India:Wiley Eastern beetle Bolitophagusreticulatus (Coleoptera, Shirt DB. 1987. British Red Data Books: 2. Tenebrionidae).Ecoscience 5:486-93 SAPROXYLICINSECTS AND FORESTRY 23

Swift MJ. 1977. The ecology of wood decom- Vaisinen R, Bistrom 0, HeliovaaraK. 1993. position. Sci. Prog. 64:175-99 Sub-cortical Coleoptera in dead pines and Tate KR, Ross DJ, O'Brien BJ, Kelliher FM. spruces: Is primeval species composition 1993. Carbonstorage and turnover,and res- maintained in managed forests? Biodivers. piratoryactivity, in the litter and soil of an Conserv.2:95-113 old-growthsouthern beech (Nothofagus)for- Wall J. 1999. Fuelwood in Australia: im- est. Soil Biol. Biochem. 25:1601-12 pacts and opportunities.In TemperateEu- Tavakilian G, Berkov A, Meurer-GrimesB, calypt Woodlands in Australia: Biology, Mori S. 1997. Neotropical tree species Conservation, Management and Restora- and their faunas of xylophagous longi- tion, ed. RJ Hobbs, CJ Yates, pp. 372-81. corns (Coleoptera:Cerambycidae) in French Chipping Norton, Aust.: Surrey Beatty & Guyana.Bot. Rev. 63:303-55 Sons ThirgoodJV. 1989. Man's impacton the forests WarrenMS, Key RS. 1991. Woodlands:past, of Europe. J. WorldFor. Resour. Manage. presentand potential for insects. In The Con- 4:127-67 servation of Insects and their Habitats, ed. Thomas CD. 2000. Dispersal and extinction in NM Collins, JA Thomas,pp. 155-203. Lon- fragmentedlandscapes. Proc. R. Soc. London don: Academic Ser.B 267:139-45 WeslienJ, SchroederLM. 1999. Populationlev- Thunes KH, MidtgaardF, Gjerde I. 2000. Di- els of barkbeetles and associated insects in versity of Coleopteraof the bracketfungus managedand unmanagedspruce stands. For. Fomitopsispinicola in a Norwegian spruce Ecol. Manage. 115:267-75 forest. Biodivers. Conserv.9:833-52 WikarsL-O. 2001. The wood-decayingfungus TorgersenTR, Bull EL. 1995. Down logs as Daldinia loculata (Xylariaceae) as an indi- habitatfor forest-dwellingants-the primary cator of fire-dependentinsects. Ecol. Bull. prey of pileatedwoodpeckers in northeastern 49:263-68 Oregon.Northwest Sci. 69:294-303 WoldendorpG, Keenan RJ, Ryan MF. 2001. TrittonLM, MartinCW, HornbeckJW, Pierce Coarse WoodyDebris in Australian Forest RS. 1987. Biomass and nutrient removals Ecosystems.Canberra, Aust.: AFFA Bur. Ru- from commercial thinning and whole-tree ral Sci. clear-cuttingof centralhardwoods. Environ. Wood GA, Hasenpusch J, Storey RI. 1996. Manage. 11:659-66 The life historyof Phalacrognathusmuelleri TwinnPFG, HardingPT. 1999. ProvisionalAt- (Macleay) (Coleoptera, Lucanidae). Aust. las of the LonghornBeetles (Coleoptera,Ce- Entomol.Mag. 23:37-48 rambycidae)of Britain. Abbots Ripton, UK: Yee M, Yuan Z, Mohammed C. 2001. Not Nat. Environ.Res. Counc. 96 pp. just waste wood: decaying logs as key habi- UK For. Comm. 1998. The UK Forestry tats in Tasmania'swet sclerophyllEucalyp- Standard:The UK Government'sApproach tus obliqua productionforests: the ecology to Sustainable Forestry. Edinburgh: For. of large and small logs compared.Tasforests. Comm. 13:119-28