Genetic diversity and connectivity shape herbivore load within an oak population and its range limit

Gene ic di e si y and connec i i y shape he bi o e load wi hin an oak popula ion a i s ange limi TA ¨HTI POHJANMIES, 1,6, AYCO J. M. TACK, 2,3 PERTTI PULKKINEN, 4 SAKINA ELSHIBLI, 5 PEKKA VAKKARI, 5 AND TOMAS ROSLIN 1 1 Spa ial Foodweb Ecology G oup, Depa men o Ag icul u al Sciences, P.O. Box 27 (La oka anonkaa i 5), FI-00014 Uni e si y o Helsinki, Helsinki, Finland 2 Me apopula ion Resea ch G oup, Depa men o Biological and En i onmen al Sciences, P.O. Box 65 (Viikinkaa i 1), FI-00014 Uni e si y o Helsinki, Helsinki, Finland 3 Depa men o Ecology, En i onmen and Plan Sciences, S ockholm Uni e si y, Lilla F esca i, SE-106 91 S ockholm, Sweden 4 Haapas ensy ja¨ B eeding S a ion, Finnish Fo es Resea ch Ins i u e, Haapas ensy ja¨n ie 34, FI-12600 La¨ylia¨inen, Finland 5 Van aa Resea ch Cen e, Finnish Fo es Resea ch Ins i u e, PL 18, FI-01301 Van aa, Finland Ci a ion: Pohjanmies, T., A. J. M. Tack, P. Pulkkinen, S. Elshibli, P. Vakka i, and T. Roslin. 2015. Gene ic di e si y and connec i i y shape he bi o e load wi hin an oak popula ion a i s ange limi . Ecosphe e 6(6):101. h p://dx.doi.o g/10. 1890/ES14-00549.1 Abs ac . Hos gene ic di e si y and geno ypic iden i y ha e been epo ed o a ec he abundance, species ichness and species di e si y o associa ed he bi o e communi ies. Recen wo k, howe e , sugges s ha hese e ec s a e highly con ex -dependen and ha he magni ude and di ec ion o he e ec s may a y wi h e.g., spa ial ac o s and he amoun o gene ic a ia ion p esen in he hos popula ion. He e, we use obse a ional da a on a Finnish oak (Que cus obu ) popula ion o examine whe he low gene ic di e si y wi hin pe iphe al popula ions educes he impac o hos geno ype on associa ed he bi o e communi ies. We i s compa ed measu es o gene ic a ia ion wi hin Finnish oak popula ions wi h hose eco ded in mo e cen al pa s o he species’ ange, con i ming ha gene ic a ia ion wi hin he Finnish popula ions is compa a i ely low. Despi e his esul , we ound consis en imp in s o hos gene ic di e si y on he bi o e communi ies: he bi o e load, bu no he species ichness, inc eased wi h hos gene ic di e si y in bo h yea s and bo h spa ial scales examined. Spa ial connec i i y o hos s inc eased he bi o e di e si y as well as abundance. While he simila i y o he bi o e communi ies inc eased wi h he gene ic simila i y among hos s, he e ec o geog aphic dis ance was s onge . O e all, ou indings iden i y a majo ole o spa ial con ex in s uc u ing oak-associa ed he bi o e communi ies—bu we s ill ace de ec able imp in s o hos geno ype a mul iple spa ial scales e en in his pe iphe al, gene ically impo e ished oak popula ion. Key wo ds: communi y gene ics; ex ended pheno ype; gene ic a ia ion; Que cus obu . Recei ed 29 Decembe 2014; e ised 23 Feb ua y 2015; accep ed 4 Ma ch 2015; inal e sion ecei ed 10 Ap il 2015; published 25 June 2015. Co esponding Edi o : D. P. C. Pe e s. Copy igh : Ó2015 Pohjanmies e al. This is an open-access a icle dis ibu ed unde he e ms o he C ea i e Commons A ibu ion License, which pe mi s un es ic ed use, dis ibu ion, and ep oduc ion in any medium, p o ided he o iginal au ho and sou ce a e c edi ed. h p://c ea i ecommons.o g/licenses/by/3.0/ 6 P esen add ess: Depa men o Biological and En i onmen al Science, P.O. Box 35 (Su on ie 9 C), FI-40014 Uni e si y o Jy a¨skyla¨, Jy a¨skyla¨, Finland. E-mail: [email p o ec ed] INTRODUCTION The ex en o which he he bi o e communi ies o plan s ep esen an ex ended pheno ype o he hos geno ype has been he subjec o ecen deba e (Hughes e al. 2008, Bailey e al. 2009, Tack e al. 2012). Acco ding o he amewo k o communi y gene ics, gene ic a ia ion wi hin a www.esajou nals.o g 1June 2015 Volume 6(6) A icle 101 popula ion o one species may in luence he dynamics and di e si y o associa ed species—a no ion o which suppo has been ound in se e al empi ical s udies (He sch-G een e al. 2011). In pa icula , gene ic di e si y a he le el o plan indi iduals (i.e., he e ozygosi y; e.g., To a -Sa´nchez e al. 2013) o plan popula ions (e.g., C u singe e al. 2006, Johnson e al. 2006) has been p oposed o a ec he abundance and di e si y o associa ed a h opod communi ies, and he simila i y among plan indi iduals o a ec he simila i y o he associa ed communi- ies (e.g., Bange e al. 2006b). The e ec s o in aspeci ic gene ic a ia ion o hos plan s may occu a a ious le els and hus a ise om a ious mechanisms. Fo example, hos gene ic di e si y may inc ease he abundance o associ- a ed a h opods a he le el o indi idual plan s e.g., by he bi o e p e e ence (c . G ipenbe g e al. 2010, Kalske e al. 2014), o a he le el o plan popula ions e.g., by inc eased esou ce a ailabil- i y (C u singe e al. 2006). Recen wo k (Tack e al. 2012) s ess ha he ealized e ec s o hos plan geno ype should be compa ed wi h o he o ces ha may shape communi ies in hei na u al en i onmen . In pa icula , i has been p oposed ha spa ial e ec s— he ela i e posi ion o a hos plan wi h espec o congene ic indi iduals wi hin he landscape—may dwa gene ic e ec s (Tack e al. 2010). In addi ion o spa ial e ec s, he ela i e e ec s o hos geno ype may a y wi h he amoun o gene ic a ia ion p esen in he hos popula ion (Bange e al. 2006b). Howe e , s udies in communi y gene ics ha e consis en ly ocused on sys ems wi h ample gene ic a ia ion (e.g., hyb id sys ems). This s a us quo unde - mines he gene ali y o cu en heo y. Oaks (Que cus spp.) a e as a genus cha ac e - ized by high le els o gene ic a ia ion wi hin and among species and wi hin conspeci ic popula ions (K eme and Pe i 1993). Th oughou hei ange, oaks ha bo a di e se communi y o insec species. They may be conside ed ounda- ion species in he empe a e o es s o he no he n hemisphe e, whe e hei abundance is cu en ly in decline (Lindbladh and Fos e 2010). The peduncula e oak (Que cus obu ) g ows in no he n Eu ope a he no he n limi o i s na u al ange. As a likely consequence, he di e si y o chlo oplas DNA (cpDNA) in he Q. obu popula ions o no he n Eu ope has been ound o be lowe han in cen al Eu ope (Pe i e al. 2002). In addi ion o he ma ginal loca ion, low gene ic a ia ion in no h Eu opean Q. obu popula ions may de i e om small popula ion sizes, he absence o o he , in e e ile oak species, and he his o ic agmen a ion o habi- a s (Vakka i e al. 2006). He e, we use obse a ional da a om a na u al sys em o examine he e ec s o gene ic a ia ion in Finnish popula ions o Que cus obu a hei no he n ange edge on he associa ed commu- ni y o galling and lea -mining he bi o es. In pa icula , we ask: (1) Will he gene ic di e si y a he indi idual le el (i.e., he e ozygosi y) a ec he s uc u e o he he bi o e assemblage occu- pying he ees? (2) Will hos gene ic di e si y a ec he he bi o e assemblage a a neighbo - hood scale, i.e., clus e s o ees wi hin he landscape? (3) Will he gene ic simila i y among hos s inc ease he simila i y o hei he bi o e communi ies? (4) A e he esul s obse ed he e condi ional on hei se ing a he ange ma gin o he hos , i.e., a e he oaks o Finland gene ically mo e uni o m han a e oaks a he co e o hei Eu opean ange? METHODS S udy design Que cus obu occu s na u ally in he sou h- wes e n a chipelago and coas al zone o Finland (Fe is e al. 1998). To quan i y he le el o gene ic di e si y cha ac e izing Finnish Q. obu popula- ions, we measu ed he gene ic a ia ion wi hin h ee ee s ands in sou h-wes e n Finland: one in Inkoo, one in Salo, and one on he island o Wa kas (Fig. 1A). All h ee s ands a e sca e ed in s uc u e and loca ed in a mosaic o ag icul- u al a eas, wa e , and ba e ock. Samples o ees we e geno yped using 15 nuclea mic osa elli e loci ( o de ails, see Appendix A). Gene ic a ia ion wi hin each popula ion was measu ed as allelic ichness (A), calcula ed in A lequin 3.5.1.3 (Exco ie and Lische 2010). We examined he communi y-le el e ec s o hos geno ype wi hin he Q. obu popula ion on Wa kas , ocusing on wo spa ial scales: indi id- ual ees and ee neighbo hoods. One hund ed ees we e andomly selec ed (Fig. 1B) and su eyed o c. 20 galling and lea -mining www.esajou nals.o g 2June 2015 Volume 6(6) A icle 101 POHJANMIES ET AL. he bi o es in bo h 2006 and 2007. T ee neighbo - hoods we e de ined as including a minimum o h ee ees loca ed wi hin a maximum dis ance o 120 m o each o he , co esponding o he a e age dispe sal dis ance o one o he mine species, Tische ia ekebladella (Appendix A in Tack and Roslin 2011). Ele en such neighbo hoods we e iden i ied in he da a, wi h he numbe o ees in a neighbo hood anging om 3 o 12 (Fig. 1C). The neighbo hoods we e mu ually exclusi e, and oge he included a o al o 60 ou o he 100 ees su eyed. Fo u he de ails on he s udy design and he he bi o e su eys, see Appendix B. Fig. 1. Loca ions o he s udy popula ions, su eyed ees, and ee neighbo hoods. (A) Loca ions o he h ee oak popula ions in sou h-wes e n Finland wi hin which le els o gene ic a ia ion we e quan i ied. The loca ions a e ma ked on he map wi h black s a s. (B) A map o he island o Wa kas . All indi idual oak ees on he island a e ma ked on he map wi h black do s, and he 100 ees su eyed o he bi o es wi h whi e do s. The dashed squa e ma ks he a ea enla ged in (C). (C) A close-up o a pa o Wa kas . T ee neighbo hoods, as iden i ied o he analyses, a e ma ked on he map wi h di e en shades o g ey. Backg ound maps copy igh ed by he Na ional Land Su ey o Finland. www.esajou nals.o g 3June 2015 Volume 6(6) A icle 101 POHJANMIES ET AL. A he le el o indi idual ees, he bi o e communi ies we e desc ibed by o al species ichness, o al he bi o e abundance, and he Shannon-Wiene index o di e si y. Adop ing he app oach o To a -Sa´nchez e al. (2013), we desc ibed he gene ic di e si y o indi idual ees as le el o he e ozygosi y, which we measu ed as in e nal ela edness (IR; Apa icio e al. 2006). Values o IR can a y be ween 1 and 1, wi h nega i e IR alues indica ing highe he e ozy- gosi y and posi i e alues highe homozygosi y. Fo a mo e in ui i e in e p e a ion o he esul s, we used (IR) in he analysis, so ha highe alues indica e highe he e ozygosi y. Fo he neighbo hoods, o al species ichness and o al he bi o e abundance we e calcula ed by adding up he da a o e he ees in he neighbo hoods ( o al numbe o unique species encoun e ed and sum o he bi o e abundances). These alues we e a e ied o co ec o unequal neighbo hood sizes: all unique combina ions o h ee ees we e sampled o neighbo hoods o size .3, species ichness and he bi o e abun- dance we e calcula ed ac oss each such sample, and he esul an alues we e a e aged. These a e ages we e hen used as es ima es o neigh- bo hood-le el me ics. The gene ic di e si y wi hin neighbo hoods was desc ibed as allelic ichness (A), co ec ed o he a ia ion in he numbe s o ees in he neighbo hoods by he a e ac ion me hod o Hu lbe (1971). The me hod was used o calcula e es ima es o he expec ed numbe o alleles a each locus in a andom sample o 3 indi iduals om he neighbo hoods o size .3. Thus, he measu es o gene ic a ia ion, species ichness, and he bi- o e load we e all ea ed wi h e ec i ely simila co ec ions o unequal neighbo hood sizes. As p e ious wo k (Tack e al. 2010) has epo ed spa ial e ec s o be o majo impo ance in ou s udy sys em, we accoun ed o such imp in s in all o ou analyses. The connec i i y o indi idual ees was desc ibed wi h a me ic adop ed om Tack e al. (2010). In p inciple, connec i i y e lec s he expec ed immig a ion o he bi o es o a pa ch a maximum pa ch occu- pancy (Tack e al. 2010). A he neighbo hood le el, connec i i y was desc ibed as he connec- i i y alue a e aged ac oss he ees in he pa ch. S a is ical analyses To es o an e ec o hos geno ype on he bi o e communi y s uc u e, we applied h ee se s o s a is ical analyses: Fi s , we cons uc ed gene alized linea models o he he bi o e communi y desc ip o s (abundance, species ich- ness, and Shannon-Wiene di e si y a he indi idual and neighbo hood le el) as unc ions o hos connec i i y and gene ic di e si y. A log- link and a Poisson dis ibu ion we e assumed o each esponse a iable, excep o he Shannon- Wiene index, which was modeled by s anda d linea eg ession. Sepa a e models we e con- s uc ed o he indi idual (n¼100) and he neighbo hood (n¼11) le els, wi h all models i ed in R .2.15.3 (R Co e Team 2013). To check o spa ial au oco ela ion in he obse ed IR alues, we used a Man el es implemen ed in R using he egan package (Oksanen e al. 2013). No signi ican spa ial au oco ela ion was de ec - ed (P¼0.23). Second, o examine whe he species esponded simila ly o indi idual hos gene ic di e si y, we cons uc ed a gene alized linea mixed model (GLMM) o species-speci ic abundances in indi- idual ees (19 species 3100 ees 32 yea s) as a unc ion o species iden i y, ee connec i i y, IR, and hei in e ac ions as ixed e ec s, and ee iden i y as a andom e ec . A Poisson dis ibu- ion was assumed o he species abundances. The model was i ed using p oc Glimmix in SAS o Windows (SAS Ins i u e, Ca y, NC, USA), e sion 9.2. Su ey yea was included as an explana o y a iable in all o hese models, whe eas da a on h ee species (Cynips longi en is,Neu o e us numismalis, and Tische ia dodonea) no encoun- e ed in bo h yea s we e excluded. The numbe o lea es examined pe ee du ing he su eys was also included as a co a ia e in models i ed a he indi idual ee le el, and a species 3yea in e ac ion was added as a ixed e ec in he GLMM. Thi d, o examine whe he gene ically mo e simila hos s sha ed mo e simila he bi o e communi ies, we used Man el es s be ween communi y dissimila i y and gene ic dis ance, be ween communi y dissimila i y and geog aph- ic dis ance, and a pa ial Man el es be ween all h ee ma ices. The es s we e implemen ed using he egan package o R. These es s we e www.esajou nals.o g 4June 2015 Volume 6(6) A icle 101 POHJANMIES ET AL. pe o med o da a om yea 2006 and 2007, espec i ely. Fo indi idual ees, gene ic dis- ance was measu ed wi h Rousse ’sˆ a, compu ed wi h SPAGeDi .1.4 (Ha dy and Vekemans 2002). A he neighbo hood le el, gene ic dis ance was measu ed by pai wise F ST , calcula ed in A lequin wi h de aul se ings. Geog aphic dis ance be- ween wo neighbo hoods was de ined as he dis ance be ween hei cen oids. Fo cons uc - ing he communi y dissimila i y ma ices, da a on species abundances we e ans o med in o ela i e abundances by di iding each species- speci ic abundance wi h he o al he bi o e abundance obse ed on he ee/in he neighbo - hood. The communi y dissimila i y ma ices we e compu ed using he egan package o R wi h he B ay-Cu is index as he communi y dissimila i y measu e. Si es in which no species had been obse ed we e excluded om he da a, lea ing a he indi idual le el n¼92 ees in 2006 and n¼93 ees in 2007. The e we e no emp y si es a he neighbo hood le el. RESULTS As expec ed, gene ic di e si y among Finnish oak popula ions p o ed compa a i ely low. The a e age allelic ichness a ied be ween 9.37 in Wa kas and 13.44 in Inkoo. The numbe s o alleles pe locus in Finnish popula ions we e, on a e age, lowe han hose eco ded in mo e cen al popula ions in Eu ope (Appendix C: Table C1). Despi e low o e all a ia ion, gene ic di e si y le a s a is ically signi ican imp in on he bi o e abundance, and hese pa e ns we e simila a bo h indi idual and neighbo hood scales: highe indi idual he e ozygosi y and highe neighbo - hood-le el allelic ichness we e associa ed wi h highe he bi o e abundance (Table 1). A consis- en imp in o gene ic di e si y was de ec ed a he le el o indi idual he bi o e species: highe he e ozygosi y was ound o inc ease he species- speci ic abundances o he bi o es. In he GLMM, he species 3he e ozygosi y in e ac ion and he h ee-way species 3connec i i y 3he e ozygos- i y in e ac ion we e non-signi ican (F 18,3607 ¼ 0.61, P¼0.90, and F 19,3607 ¼0.44, P¼0.98, espec i ely), sugges ing ha mos species e- spond o he e ozygosi y in he same way. When hese non-signi ican in e ac ions we e d opped, he main e ec o he e ozygosi y p o ed s a is- ically signi ican (F 1,97 ¼5.17, P¼0.03). In con as o he impac o gene ic di e si y on he bi o e abundance, he e was no de ec able impac o gene ic di e si y on ei he species ichness o Shannon-Wiene di e si y (Table 1). Unlike gene ic di e si y, spa ial con ex had a Table 1. Gene alized linea models o he bi o e ichness, abundance and di e si y as unc ions o hos connec i i y, hos gene ic di e si y, su ey yea , and numbe o lea es examined. Each esponse was sepa a ely modeled a he le el o (A) indi idual ees and (B) neighbo hoods o mul iple ees. Response Va iable Es ima e SE Z/ P A) Indi idual le el (n¼100) Richness Gene ic di e si y 0.16 0.18 0.86 0.39 Connec i i y 0.01 0.002 7.65 ,0.0001 Yea 0.18 0.06 2.97 0.003 No. lea es examined 0.002 0.0002 9.33 ,0.0001 Abundance Gene ic di e si y 0.45 0.06 7.03 ,0.0001 Connec i i y 0.02 0.001 26.74 ,0.0001 Yea 0.37 0.02 18.24 ,0.0001 No. lea es examined 0.004 0.0001 62.61 ,0.0001 Shannon-Wiene Gene ic di e si y 0.07 0.19 0.37 0.71 Connec i i y 0.010 0.002 5.50 ,0.0001 Yea 0.16 0.07 2.46 0.02 No. lea es examined 0.001 0.0003 4.18 ,0.0001 B) Neighbo hood le el (n¼11) Richness Gene ic di e si y 0.31 0.44 0.69 0.49 Connec i i y 0.02 0.01 2.80 0.01 Yea 0.02 0.13 0.16 0.88 Abundance Gene ic di e si y 1.37 0.13 10.60 ,0.0001 Connec i i y 0.04 0.002 26.19 ,0.0001 Yea 0.30 0.03 9.40 ,0.0001 Z- alues o ichness and abundance modeled wi h Poisson eg ession; - alue o Shannon di e si y modeled wi h s anda d linea eg ession. www.esajou nals.o g 5June 2015 Volume 6(6) A icle 101 POHJANMIES ET AL. clea e ec on all o he desc ip o s o he bi o e communi y s uc u e a bo h he indi idual and he neighbo hood le el, wi h highe connec i i y inc easing he ichness, di e si y, and abundance o he bi o es (Table 1). Fo indi idual he bi o e species, ee connec i i y and he species 3 connec i i y in e ac ion also had signi ican e ec s (F 1,97 ¼7.67, P¼0.01, and F 18,3644 ¼4.46, P,0.0001, espec i ely), sugges ing ha hos connec i i y inc eases he bi o e abundance in gene al, bu ha di e en species espond o i in di e en ways. On a e age, gene ically mo e simila ees and ee neighbo hoods sha ed mo e simila he bi- o e communi ies (Table 2). Howe e , a simila e ec was obse ed o geog aphic dis ance: on a e age, he close o each o he wo ees/ neighbo hoods we e loca ed, he mo e simila we e hei he bi o e communi ies. This co ela- ion was s onge han ha be ween gene ic and communi y simila i y in all bu one case, he neighbo hood le el in 2006 (Table 2). Re lec ing his, he associa ion o gene ic and communi y simila i y weakened o non-signi ican when all h eedis ancema iceswe ecompa ed,i.e., when he geog aphic dis ances we e accoun ed o (Table 2). DISCUSSION We de ec ed a clea imp in o hos geno ype on he associa ed he bi o e communi ies bo h a he scale o indi idual ees and o ee neigh- bo hoods: he bi o e abundance was s ongly a ec ed by bo h hos gene ic di e si y and spa ial connec i i y. Impo an ly, he e ec s o hos geno ype eme ged despi e ou inding ha Finnish oak popula ions sus ain less gene ic a ia ion han do oak popula ions a he co e o hei ange, and despi e he sugges ion ha he ela i e impo ance o gene ic e sus en i on- men al e ec s may be smalle wi hin gene ically impo e ished popula ions (Bange e al. 2006b). Howe e , in con as o he majo i y o p e ious s udies (C u singe e al. 2006, Johnson e al. 2006, He sch-G een e al. 2011), we de ec ed no impac o hos gene ic di e si y on he di e si y o he he bi o e communi y. While compa isons o allelic ichness be ween samples o di e en size a e admi edly p ob- lema ic (Lebe g 2002), ou da a o e con incing e idence ha he gene ic di e si y o Finnish popula ions is lowe han ha cha ac e izing mo e cen al s ands. In i sel , he loss o gene ic di e si y om popula ions a he ange ma gin is consis en wi h heo y (Ecke e al. 2008) and ma ches wi h pa e ns obse ed in e ms o cpDNA a ia ion wi hin Q. obu popula ions (Pe i e al. 2002). While ma e nally inhe i ed cpDNA may show somewha di e en le els o a ia ion han nuclea mic osa elli es (Pe i e al. 2005), he le els o nuclea mic osa elli e di e - si y de ec ed in he p esen s udy sugges ha he his o ical coloniza ion p ocesses ha e a ec ed bo h (Fe is e al. 1998). In e ms o a e age allelic ichness, gene ic a ia ion wi hin he Q. obu popula ion on Wa kas was also he lowes measu ed among he h ee Finnish popula ions. Despi e all hese conside a ions, he gene ic a ia ion p esen wi hin he popula ion is appa - en ly su icien o cause de ec able e ec s on he bi o e load. In e ms o he imp in s o hos gene ic di e si y de ec ed, ou esul s we e s ikingly Table 2. Co ela ions be ween he bi o e communi y dissimila i y, geog aphic dis ance, and gene ic dis ance among (A) indi idual ees and (B) ee neighbo hoods measu ed using Man el and pa ial Man el es s. Communi y dissimila i y s 2006 2007 P P A) Indi idual le el Gene ic dis ance 0.10 0.02 0.14 0.002 Geog aphic dis ance 0.36 0.0001 0.39 0.0001 Gene ic dis ancejGeog aphic dis ance 0.03 0.28 0.05 0.15 B) Neighbo hood le el Gene ic dis ance 0.68 0.003 0.45 0.01 Geog aphic dis ance 0.68 0.01 0.60 0.002 Gene ic dis ancejGeog aphic dis ance 0.34 0.03 0.001 0.50 No e: Shown a e he Man el s a is ics compu ed ( ) and hei signi icance (P) based on 10,000 pe mu a ions o he communi y dissimila i y ma ix. www.esajou nals.o g 6June 2015 Volume 6(6) A icle 101 POHJANMIES ET AL. consis en o e species and spa ial scales: he he e ozygosi y o indi idual hos s as well as he allelic ichness wi hin ee neighbo hoods had a s a is ically signi ican , posi i e e ec on bo h he o e all abundance and he species-speci ic abun- dances o he bi o es. These esul s ag ee wi h p e ious indings ha hos geno ypic di e si y a he pa ch o popula ion le el inc eases a h opod abundance (C u singe e al. 2006, Johnson e al. 2006), and ha he he e ozygosi y and geno ype o indi idual hos s may a ec he bi o e abun- dance o signi ican deg ees (E ans e al. 2012, Kalske e al. 2014). E ec s o he e ozygosi y ha e p e iously been epo ed e.g., by To a -Sa´nchez e al. (2013), who ound o al a h opod biomass o dec ease wi h inc easing he e ozygosi y o he hos and in e p e ed his as a po en ial sign o weake de ense in mo e homozygous hos s. Ou esul s, con e sely, sugges ha highe he e ozy- gosi y may lead o highe suscep ibili y. We hypo hesize ha he galle and lea -mine species in ou s udy may espond posi i ely o he e o- zygosi y, which, assuming ha he e ozygosi y is associa ed wi h igo , ma ches wi h nume ous obse a ions es ing he plan igo hypo hesis (P ice 1991, Co nelissen e al. 2008). While he imp in s o hos gene ic di e si y on he bi o e abundance we e clea , he bi o e species ichness and di e si y we e no de ec ably a ec - ed by hos gene ic di e si y a ei he spa ial scale. This inding con as s wi h se e al p e ious epo s o he bi o e di e si y inc easing wi h popula ion-le el geno ypic di e si y. Such pa - e ns ha e been explained by addi i e and non- addi i e e ec s (C u singe e al. 2006, Johnson e al. 2006, Tack and Roslin 2011), i.e., by sampling among hos geno ypes wi h speci ic communi ies (addi i e e ec s) and syne gis ic e ec s ealized in mix u es o mul iple hos geno ypes (non-addi i e e ec s). I seems likely ha gene ic di e si y wi hin an indi idual hos does no p oduce he s uc u es, unc ions, o in e ac ions ha c ea e ei he addi i e o non-addi i e e ec s. None he- less, we also ailed o ind any e ec s o hos gene ic di e si y on species ichness a he neighbo hood le el—bu hen again, we mea- su ed neighbo hood-le el gene ic di e si y by a e age allelic ichness, which is no s ic ly analogous o geno ypic di e si y (Hughes e al. 2008). Thus, he cu en esul s do no su ice as e idence o a lack o hos gene ic di e si y e lec ing in o a lack o e ec on he bi o e di e si y. Ins ead, hey su ice o demons a e ha wha limi ed gene ic di e si y he e is, e lec s in o de ec able a ia ion in he bi o e abundance. The gene al associa ion o hos gene ic ela ed- ness and communi y simila i y, coined he ‘gene ic simila i y ule’, has been obse ed, o ins ance, in co onwood-based sys ems (Bange e al. 2006a, b) and in opical epiphy ic b omeliads (Zy ynska e al. 2012). In he p esen s udy, we ound gene ic simila i y among hos s o inc ease he simila i y o hei he bi o e communi ies, bu also geog aphic p oximi y among hos s inc eased communi y simila i y. Wi h he e ec s o geog aphic dis ances accoun ed o , he associa ion o communi y simila i y and gene ic ela edness gene ally dwin- dled in o s a is ical non-signi icance. This esul sugges s ha in ou s udy sys em, imp in s o spa ial con ex may o e ide hose o hos geno ype. Such an in e p e a ion is u he sup- po ed by p e ious indings om oak-based ood webs in Finland, whe e e ec s o hos geno ype ha e been p oposed o be seconda y o hose o landscape con igu a ion (Tack e al. 2010). No a- bly, he absence o a gene ic simila i y pa e n canno be di ec ly asc ibed o he lowe gene ic di e si y o Que cus obu a he ange ma gin: Gossne e al. (2015) ecen ly ound ha wi hin he co e a ea o he oak dis ibu ion (Ba a ia, sou he n Ge many) he deg ee o gene ic simila - i y among ma u e oaks was no e lec ed in he associa ed insec communi ies. Impo an ly, howe e , he e ec s o he spa ial con ex ailed o smudge all imp in s o hos geno ype. While hos connec i i y p o ed a signi ican ac o in luencing he bi o e abun- dance, ichness, and di e si y, he e ec s o hos geno ype on he bi o e abundance emained s a is ically signi ican e en when connec i i y was included in he models. Mo eo e , while ou esul s sugges ha di e en he bi o e species espond o hos connec i i y in di e en ways (as likely due o di e ences in dispe sal abili y), esponses o hos he e ozygosi y p o ed consis- en ac oss all species. Thus, bo h landscape con ex and hos geno ype seem o come wi h independen e ec s, and bo h may be impo an in de e mining he bi o e communi y s uc u e. These esul s ag ee closely wi h p e ious obse - a ions ha indi idual a ia ion in hos plan quali y is s ong enough o cause local adap a ion www.esajou nals.o g 7June 2015 Volume 6(6) A icle 101 POHJANMIES ET AL. by he bi o e popula ions o hei hos s, bu ha he s eng h o such adap a ion will a y wi h he landscape con ex (Tack and Roslin 2010). While ou pape p ima ily ocuses on he impac o gene ic di e si y and spa ial connec- i i y on he bi o e communi y pa e ns, we no e ha he esul s can also be in e p e ed om he pe spec i e o ee i ness. As low-le el he bi o y may s ongly a ec he i ness o oak ees (C awley 1985), and as ou esul s e eal a s ong link be ween gene ic di e si y, habi a con igu a ion, and he bi o e load, hey sugges ha ee i ness will a y p edic ably ac oss he landscape. Explo ing he impac o gene ic di e si y and he posi ion o a ee wi hin he landscape on ee i ness and he e olu ion o ee esis ance will hus o e an in e es ing a enue o u u e esea ch. In conclusion, ou s udy e eals how imp in s o hos geno ype and spa ial con ex may be ound e en in gene ically impo e ished hos popula ions a he ma gins o hei ange. By doing so, i poin s o in iguing in e ac ions be ween hos di e si y and he bi o e communi ies ac oss landscapes in any pa o he hos ’s ange. ACKNOWLEDGMENTS The s udy was suppo ed by he Academy o Finland (g an numbe 138346 o T. Roslin). We hank all he ieldwo ke s and all he landowne s o hei kind pe mission o sample oaks and insec s. LITERATURE CITED Apa icio, J. M., J. O ego, and P. J. Co de o. 2006. Wha should we weigh o es ima e he e ozygosi y, alleles o loci? Molecula Ecology 15:4659–4665. Bailey, J. K., J. A. Schwei ze , F. U ´beda, J. Ko iche a, C. J. LeRoy, M. D. Mad i ch, B. J. Rehill, R. K. Bange , D. G. Fische , G. J. Allan, and T. G. Whi ham. 2009. F om genes o ecosys ems: a syn hesis o he e ec s o plan gene ic ac o s ac oss le els o o ganiza ion. Philosophical T ans- ac ions o he Royal Socie y B 364:1607–1616. Bange , R. K., G. J. Allan, R. J. Tu ek, G. M. Wimp, N. Meneses, G. D. Ma insen, P. Keim, and T. G. Whi ham. 2006a. F om genes o geog aphy: a gene ic simila i y ule o a h opod communi y s uc u e a mul iple geog aphic scales. Molecula Ecology 15:4215–4228. Bange , R. K., R. J. Tu ek, B. Rehill, G. M. Wimp, J. A. Schwei ze , G. J. Allan, J. K. Bailey, G. D. Ma in- sen, P. Keim, R. L. Lind o h, and T. G. Whi ham. 2006b. A gene ic simila i y ule de e mines a h o- pod communi y s uc u e. Molecula Ecology 15:1379–1391. Co nelissen, T., G. Wilson Fe nandes, and J. Vascon- cellos-Ne o. 2008. Size does ma e : a ia ion in he bi o y be ween and wi hin plan s and he plan igo hypo hesis. Oikos 117:1121–1130. Co ell, J. E., R. C. Mun o, H. E. Tabbene , A. D. Milne , G. I. Fo es , and A. J. Lowe. 2003. Compa ison o ine-scale gene ic s uc u e using nuclea mic osa elli es wi hin wo B i ish oak- woods di e ing in popula ion his o y. Fo es Ecology and Managemen 176:287–303. C awley, M. J. 1985. Reduc ion o oak ecundi y by low-densi y he bi o e popula ions. Na u e 314:163–164. C u singe , G. M., M. D. Collins, J. A. Fo dyce, Z. Gompe , C. C. Nice, and N. J. Sande s. 2006. Plan geno ypic di e si y p edic s communi y s uc u e and go e ns an ecosys em p ocess. Science 313:966–968. Cu u,A.L.,O.Gailing,L.Leinemann,andR. Finkeldey. 2007. Gene ic Va ia ion and Di e en ia- ion wi hin a na u al communi y o i e oak species (Que cus spp.). Plan Biology 9:116–126. Degen, B., R. S ei , and B. Ziegenhagen. 1999. Compa a i e s udy o gene ic a ia ion and di e - en ia ion o wo peduncula e oak (Que cus obu ) s ands using mic osa elli e and allozyme loci. He edi y 83:597–603. Dow, B. D., M. V. Ashley, and H. F. Howe. 1995. Cha ac e iza ion o highly a iable (GA/CT)n mic osa elli es in he bu oak, Que cus mac oca pa. Theo e ical and Applied Gene ics 91:137–141. Ecke , C. G., K. E. Samis, and S. C. Lougheed. 2008. Gene ic a ia ion ac oss species’geog aphical anges: he cen al-ma ginal hypo hesis and be- yond. Molecula Ecology 17:1170–1188. E ans, L. M., J. S. Cla k, A. V. Whipple, and T. G. Whi ham. 2012. The ela i e in luences o hos plan geno ype and yea ly abio ic a iabili y in de e mining he bi o e abundance. Oecologia 168:483–489. Exco ie , L., and H. E. L. Lische . 2010. A lequin sui e e 3.5: a new se ies o p og ams o pe o m popula ion gene ics analyses unde Linux and Windows. Mo- lecula Ecology Resou ces 10:564–567. Fe is, C., R. A. King, R. Va¨ino¨la¨, and G. M. Hewi . 1998. Chlo oplas DNA ecognizes h ee e ugial sou ces o Eu opean oaks and sugges s indepen- den eas e n and wes e n immig a ions o Finland. He edi y 80:584–593. Gossne , M. M., M. B a¨ndle, R. B andl, J. Bail, J. Mu¨ lle , and L. Opgenoo h. 2015. Whe e is he ex ended pheno ype in he wild? The communi y composi- ion o a h opods on ma u e oak ees does no depend on he oak geno ype. PLoS ONE www.esajou nals.o g 8June 2015 Volume 6(6) A icle 101 POHJANMIES ET AL. 10:e0115733. G ipenbe g, S., P. J. Mayhew, M. Pa nell, and T. Roslin. 2010. A me a-analysis o p e e ence-pe o mance ela ionships in phy ophagous insec s. Ecology Le e s 13:383–393. G ipenbe g, S., and T. Roslin. 2005. Hos plan s as islands: esou ce quali y and spa ial se ing as de e minan s o insec dis ibu ion. Annales Zoo- logici Fennici 42:335–345. G ipenbe g, S., and T. Roslin. 2008. Nei he he de il no he deep blue sea: la al mo ali y ac o s ail o explain he abundance and dis ibu ion o Tische ia ekebladella. Ecological En omology 33:346–356. Ha dy, O. J., and X. Vekemans. 2002. SPAGeDi: a e sa ile compu e p og am o analyse spa ial gene ic s uc u e a he indi idual o popula ion le els. Molecula Ecology No es 2:618–620. He sch-G een, E. I., N. E. Tu ley, and M. T. J. Johnson. 2011. Communi y gene ics: Wha ha e we accom- plished and whe e should we be going? Philosoph- ical T ansac ions o he Royal Socie y B 366:1453– 1460. Hughes, A. R., B. D. Inouye, M. T. J. Johnson, N. Unde wood, and M. Vellend. 2008. Ecological consequences o gene ic di e si y. Ecology Le e s 11:609–623. Hu lbe , S. H. 1971. The nonconcep o species di e si y: a c i ique and al e na i e pa ame e s. Ecology 52:577–586. Johnson, M. T. J., M. J. Lajeunesse, and A. A. Ag awal. 2006. Addi i e and in e ac i e e ec s o plan geno ypic di e si y on a h opod communi ies and plan i ness. Ecology Le e s 9:24–34. Kalske, A., A. Muola, P. Mu ikainen, and R. Leimu. 2014. P e e ence o ou b ed hos plan s and posi i e e ec s o inb eeding on egg su i al in a specialis he bi o e. P oceedings o he Royal Socie y B 281:20141421. Kamp e , S., C. Lexe , J. Glo¨ssl, and H. S einkellne . 1998. Cha ac e iza ion o (GA)nmic osa elli e loci om Que cus obu . He edi as 129:183–186. K eme , A., and R. J. Pe i . 1993. Gene di e si y in na u al popula ions o oak species. Annals o Fo es Science 50:186–202. Lebe g, P. L. 2002. Es ima ing allelic ichness: E ec s o sample size and bo lenecks. Molecula Ecology 11:2445–2449. Lindbladh, M., and D. R. Fos e . 2010. Dynamics o long-li ed ounda ion species: he his o y o Que - cus in sou he n Scandina ia. Jou nal o Ecology 98:1330–1345. Oksanen, J., F. G. Blanche , R. Kind , P. Legend e, P. R. Minchin, R. B. O’Ha a, G. L. Simpson, P. Solymos, M. H. H. S e ens, and H. Wagne . 2013. egan: communi y ecology package. h p://CRAN. R-p ojec .o g/package¼ egan Pe i , R. J., J. Duminil, S. Fineschi, A. Hampe, D. Sal ini, and G. G. Vend amin. 2005. In i ed e iew: Compa a i e o ganiza ion o chlo oplas , mi o- chond ial and nuclea di e si y in plan popula- ions. Molecula Ecology 14:689–701. Pe i , R. J., e al. 2002. Chlo oplas DNA a ia ion in Eu opean whi e oaks: phylogeog aphy and pa - e ns o di e si y based on da a om o e 2600 popula ions. Fo es Ecology and Managemen 156:5–26. P ice, P. W. 1991. The plan igo hypo hesis and he bi o e a ack. Oikos 62:244–251. R Co e Team. 2013. R: A language and en i onmen o s a is ical compu ing. R Founda ion o S a is- ical Compu ing, Vienna, Aus ia. S einkellne , H., S. Fluch, E. Tu e schek, C. Lexe , R. S ei , A. K eme , K. Bu g, and J. Glo¨ssl. 1997. Iden i ica ion and cha ac e iza ion o (GA/CT)n- mic osa elli e loci om Que cus pe aea.Plan Molecula Biology 33:1093–1096. S ei , R., T. Labbe, R. Bacilie i, H. S einkellne , J. Glo¨ssl, and A. K eme . 1998. Wi hin-popula ion gene ic s uc u e in Que cus obu L. and Que cus pe aea (Ma .) Liebl. assessed wi h isozymes and mic osa elli es. Molecula Ecology 7:317–328. Tack, A. J. M., M. T. J. Johnson, and T. Roslin. 2012. Sizing up communi y gene ics: i ’s a ma e o scale. Oikos 121:481–488. Tack, A. J. M., O. O askainen, P. Pulkkinen, and T. Roslin. 2010. Spa ial loca ion domina es o e hos plan geno ype in s uc u ing an he bi o e com- muni y. Ecology 91:2660–2672. Tack, A. J. M., and T. Roslin. 2010. O e un by he neighbo s: landscape con ex a ec s s eng h and sign o local adap a ion. Ecology 91:2253–2260. Tack, A. J. M., and T. Roslin. 2011. The ela i e impo ance o hos -plan gene ic di e si y in s uc u ing he associa ed he bi o e communi y. Ecology 92:1594–1604. To a -Sa´nchez, E., L. Valencia-Cue as, E. Cas illo- Mendoza, P. Mussali-Galan e, R. V. Pe ´ ez-Ruiz, and A. Mendoza. 2013. Associa ion be ween indi idual gene ic di e si y o wo oak hos species and canopy a h opod communi y s uc u e. Eu o- pean Jou nal o Fo es Resea ch 132:165–179. Vakka i, P., A. Blom, M. Rusanen, J. Raisio, and H. Toi onen. 2006. Gene ic a iabili y o agmen ed s ands o peduncula e oak (Que cus obu )in Finland. Gene ica 127:231–241. Yu¨ ceda˘ g, C., and O. Gailing. 2013. Mo phological and gene ic a ia ion wi hin and among ou Que cus pe ea and Q. obu na u al popula ions. Tu kish Jou nal o Bo any 37:619–629. Zy ynska, S. E., M. S. Khud , E. Ha is, and R. F. P eziosi. 2012. Gene ic e ec s o ank- o ming b omeliads on he associa ed in e eb a e commu- ni y in a opical o es ecosys em. Oecologia 170:467–475. www.esajou nals.o g 9June 2015 Volume 6(6) A icle 101 POHJANMIES ET AL.