Does catchment geodiversity foster stream biodiversity?

This is a sel -a chi ed e sion o an o iginal a icle. This e sion may di e om he o iginal in pagina ion and ypog aphic de ails. Au ho (s): Ti le: Yea : Ve sion: Copy igh : Righ s: Righ s u l: Please ci e he o iginal e sion: CC BY 4.0 h ps://c ea i ecommons.o g/licenses/by/4.0/ Does ca chmen geodi e si y os e s eam biodi e si y? © The Au ho s, 2019 Published e sion Kä nä, Olli-Ma i; Heino, Jani; Laamanen, Tiina; Jy känkallio-Mikkola, Jenny; Pajunen, Vi pi; Soininen, Janne; Tolonen, Kimmo T.; Tukiainen, Helena; Hjo , Jan Kä nä, O.-M., Heino, J., Laamanen, T., Jy känkallio-Mikkola, J., Pajunen, V., Soininen, J., Tolonen, K. T., Tukiainen, H., & Hjo , J. (2019). Does ca chmen geodi e si y os e s eam biodi e si y?. Landscape Ecology, 34(10), 2469-2485. h ps://doi.o g/10.1007/s10980-019-00901-z 2019 RESEARCH ARTICLE Does ca chmen geodi e si y os e s eam biodi e si y? Olli-Ma i Ka ¨ na ¨.Jani Heino .Tiina Laamanen .Jenny Jy ka ¨nkallio-Mikkola . Vi pi Pajunen .Janne Soininen .Kimmo T. Tolonen .Helena Tukiainen . Jan Hjo Recei ed: 14 Decembe 2018 / Accep ed: 5 Sep embe 2019 / Published online: 14 Sep embe 2019 ÓThe Au ho (s) 2019 Abs ac Con ex One app oach o main ain he esilience o bio ic communi ies is o p o ec he a iabili y o abio ic cha ac e is ics o Ea h’s su ace, i.e. geodi- e si y. In e es ial en i onmen s, he ela ionship be ween geodi e si y and biodi e si y is well ecog- nized. In s eams, he abio ic p ope ies o ups eam ca chmen s in luence s eam communi ies, bu he ela ionships be ween ca chmen geodi e si y and aqua ic biodi e si y ha e no been p e iously es ed. Objec i es The aim was o compa e he e ec s o local en i onmen al and ca chmen a iables on s eam biodi e si y. We speci ically explo ed he use ulness o ca chmen geodi e si y in explaining he species ichness on s eam mac oin e eb a e, dia om and bac e ial communi ies. Me hods We used 3 geodi e si y a iables, 2 land use a iables and 4 local habi a a iables o examine species ichness a ia ion ac oss 88 s eam si es in wes e n Finland. We used boos ed eg ession ees o explo e he e ec s o geodi e si y and o he a iables on biodi e si y. Resul s We de ec ed a clea e ec o ca chmen geodi e si y on species ichness, al hough he adi- ional local habi a and land use a iables we e he s onges p edic o s. Especially soil- ype ichness appea ed as an impo an ac o o species ichness. While a iables ela ed o s eam size we e he mos impo an o mac oin e eb a e ichness and pa ly o bac e ial ichness, he impo ance o wa e chem- is y and land use o dia om ichness was no able. Conclusions In addi ion o adi ional en i onmen al a iables, geodi e si y may a ec species ichness a ia ion in s eams, o example h ough changes in wa e chemis y. Geodi e si y in o ma ion could be used as a p oxy o p edic ing s eam species ichness and o e s a supplemen a y ool o conse a ion e o s. Keywo ds F eshwa e s En i onmen al he e ogenei y Ca chmen ea u es  Mac oin e eb a es Dia oms Bac e ia Species ichness Elec onic supplemen a y ma e ial The online e sion o his a icle (h ps://doi.o g/10.1007/s10980-019-00901-z) con- ains supplemen a y ma e ial, which is a ailable o au ho ized use s. O.-M. Ka ¨ na ¨(&)H. Tukiainen J. Hjo Geog aphy Resea ch Uni , Uni e si y o Oulu, P. O. Box 8000, 90014 Oulu, Finland e-mail: [email p o ec ed] J. Heino T. Laamanen Finnish En i onmen Ins i u e, F eshwa e Cen e, Paa o Ha aksen ie, 90570 Oulu, Finland J. Jy ka ¨nkallio-Mikkola V. Pajunen J. Soininen Depa men o Geosciences and Geog aphy, Uni e si y o Helsinki, P. O. Box 33, 00014 Helsinki, Finland K. T. Tolonen Depa men o Biological and En i onmen al Science, Uni e si y o Jy a ¨skyla ¨, P.O. Box 35, 40014 Jy a ¨skyla ¨, Finland 123 Landscape Ecol (2019) 34:2469–2485 h ps://doi.o g/10.1007/s10980-019-00901-z(0123456789().,- olV)(0123456789().,- olV) In oduc ion Global change can be seen as deg ada ion o na u al ecosys ems, which in u n is he mos impo an eason unde lying biodi e si y change (Fulle e al. 2007). Clima e and land use changes modi y bio ic communi- ies in all kinds o en i onmen s (Sala e al. 2000;Vilmi e al. 2017), including eshwa e ecosys ems (Donohue e al. 2009; Heino e al. 2009). Recen ly, i has been es ima ed ha dec ease o biodi e si y due o an h o- pogenic s esso s is clea ly highe in eshwa e ecosys- ems han in e es ial ecosys ems (Abell 2002;Wiens 2016). The undesi able end o biodi e si y loss is associa ed wi h key s esso s, such as pollu ion, in asi e species, dams, and modi ica ion o in-s eam habi a s (Vo ¨ o ¨sma y e al. 2010). The e ec s o hese s esso s a e highly ala ming because eshwa e ecosys ems co e only a small ac ion o he Ea h’s su ace a ea (0.8%), bu ha bo a conside able (6%) p opo ion o all known species on Ea h (Dudgeon e al. 2006). S eam ecosys ems a e dependen on ca chmen - scale p ope ies and p ocesses (Hynes 1970; Allan and Cas illo 2007). Thus, one could assume ha a use ul solu ion o main ain biodi e si y is o es ablish p o ec ed a eas by conside ing he in e -connec ed ea u es o he ca chmen and s eam en i onmen s (Wa d e al. 2002; Moilanen e al. 2008; Soininen e al. 2015). T adi ionally, p o ec ed a eas ha e been designed o he p o ec ion and main enance o biodi- e si y in land and ma ine ecosys ems (IUCN 1994), al hough hey a e seldom designed o conse ing eshwa e biodi e si y only. In some cases, i e s a e used as he bo de s o p o ec ed a eas (Abell e al. 2007). The wide use o he ca chmen -based conse - a ion p obably su e s om a lack o posi i e empi ical examples, he unique posi ion o eshwa e s (e.g. s eam co ido s) in he landscape (Abell e al. 2007), and he absence o comp ehensi e and s an- da dized knowledge abou he spa ial dis ibu ion o he mos i al a eas o biodi e si y (Ca izo e al. 2017). Howe e , i is gene ally unde s ood ha inclu- si e conse a ion o eshwa e ecosys ems equi es a whole-ca chmen app oach (Dudgeon e al. 2006). Conse ing na u e’s s age (CNS) is known as a s a egy o p o ec biodi e si y by main aining geodi- e si y (Beie e al. 2015). In CNS, he ocus is on he abio ic ‘s age’ upon which ecological p ocesses ake place ins ead o using popula ions, species o com- muni ies as he uni o conse a ion planning (Beie and B os 2010). Geodi e si y e e s o he a iabili y o abio ic cha ac e is ics o Ea h’s su ace and subsu ace, including ma e ials such as soils, p o- cesses like e osion, and land o ms such as alleys, which a e ela i ely s able o e long ime pe iods (Ande son and Fe ee 2010; G ay 2013). In unning wa e en i onmen s, geodi e si y could be conside ed as a d i e o species dis ibu ions and ecological p ocesses a di e en scales (Lawle e al. 2015). In addi ion, he ole o geodi e si y has been acknowl- edged o i s posi i e associa ion wi h biodi e si y in e es ial ecosys ems (Ande son and Fe ee 2010; S ein e al. 2014; Bailey e al. 2017; Tukiainen e al. 2017a), ma ine ecosys ems (Kaskela e al. 2017) and ecen ly in s eams a a local-scale (Ka ¨ na ¨e al. 2018). Running wa e s a e hie a chically s uc u ed ecosys ems whe e he dis ibu ions o species on a ce ain loca ion depend on il e ing p ocesses based on clima e, geology, dispe sal p ocesses, channel mo - phology and physical–chemical p ope ies o local habi a s (Po 1997; Fig. 1). In s eams, habi a - and each-scale il e s comp ise nu ien s, ligh , pH, s eam wid h and cu en eloci y (F issel e al. 1986). In addi ion, abio ic d i e s ope a ing a ca chmen scales (e.g. land-use, soil- ype and geology) a e also consid- e ed impo an o species dis ibu ions. Fo example, geology, geomo phological ea u es and land use a he ca chmen -scale cons ain local habi a condi ions (F issel e al. 1986; Richa ds e al. 1996; Da ies e al. 2000; Pajunen e al. 2017). The signi icance o abio ic a iables in s uc u ing species dis ibu ion a ies wi h he scale and o ganism g oup. Fo ins ance, wa e chemical p ope ies shape he s uc u e o mic oalgal communi ies (e.g. dia oms) (Soininen 2007; Jy ka ¨n- kallio-Mikkola e al. 2016) and bac e ial communi ies (Heino e al. 2014; Wang e al. 2017). Local mac oin e eb a e di e si y o en esponds o local s eam ea u es like s eam wid h and cu en eloci y, subs a e p ope ies and wa e chemis y (Malmq is and Ma ¨ki 1994; Heino e al. 2013) and se e al o ca chmen p ope ies, such as geology and land use (Richa ds e al. 1996; Sandin and Johnson 2004). In his s udy, we explo ed he possibili y o explain species ichness a ia ion in s eams using geodi e - si y in o ma ion. In addi ion, we had land use da a and adi ionally used local-scale en i onmen al a iables as explana o y a iables in ou s a is ical models. We compiled geodi e si y in o ma ion o a numbe o bo eal ca chmen s in wes e n Finland a 1-km 2 scale 123 2470 Landscape Ecol (2019) 34:2469–2485 and examined biodi e si y a ia ion along geodi e - si y and land use g adien s using dia oms, bac e ia and mac oin e eb a e as ocal o ganismal g oups. A local-scale, we expec ed o ind a ela ionship be ween mac oin e eb a e ichness and s eam si e a iables especially ela ed o s eam size and wa e chemis y (Heino e al. 2003). Fo dia om and bac e ial ichness, we expec ed o ind a clea ela ionship o wa e chemis y a iables (Soininen 2007; Jy ka ¨nkallio- Mikkola e al. 2016). O land use a iables, we expec ed o ind e ec s o ag icul u al and a i icial a eas on dia om, bac e ial (Leland and Po e 2000; Lea and Lewis 2009), and mac oin e eb a e ichness (Lena and C aw o d 1994; Sponselle e al. 2001). Mic obes may be e y sensi i e o en i onmen al changes due o hei small size (Zeglin 2015), which could lead o changes in species ichness in ca ch- men s in luenced by ag icul u e wi h associa ed inc ease in nu ien le els (Allan and Cas illo 2007). We also expec ed o ind an indi ec e ec o geodi e si y on species ichness, because geological ea u es, besides o he ca chmen p ope ies, con ol many each-scale cha ac e is ics, such as ege a ion in he ipa ian co ido , low a iabili y and wa e chemis y (Leland and Po e 2000; Soininen 2015). Fo example, subsu ace p ope ies (e.g. soil ype) a ec he p ecipi a ion– uno p ocesses in a wa e - shed (e.g. wa e in il a ion capabili ies), esul ing in changes in wa e chemis y and consequen a ia ion in species ichness in headwa e s eams (Fig. 1). Ma e ials and me hods Cha ac e is ics o he s udy a ea We sampled al oge he 88 bo eal s eams om 21 main i e basins in he coas al a eas o wes e n Fig. 1 Rep esen a ion o he en i onmen al ea u es om egional-scale and ca chmen -scale o local-scale habi a con- di ions. En i onmen al ea u es and p ocesses measu ed a di e en scales can be conside ed as il e s impo an o s eam biodi e si y pa e ns (F issel e al. 1986; Po 1997). Geodi e - si y a ca chmen -scale (soil- ype ichness in his example) can a ec he wa e in il a ion p ocesses and hus wa e chemis y impo an o s eam o ganisms (Leland and Po e 2000). The isualized clima e a iable is mean annual empe a u e ac oss Finland. Pho o c edi J. Jy ka ¨nkallio-Mikkola 123 Landscape Ecol (2019) 34:2469–2485 2471 Finland, which all d ain o he Bal ic Sea (Fig. 2). The loca ion o su eyed s eam si es (1 si e pe s eam) s e ched app oxima ely 520 km in no h–sou h di ec- ion and mo e han 300 km in eas –wes di ec ion. The landscapes in he sou he n pa s o he s udy a ea comp ise i e alleys wi h sligh ly undula ing opog aphy, la ge coas al plains and some lakes. The sou he n pa s o he s udy a ea a e cha ac e ized by ill-co e ed bed ock hills and s uc u ally con olled alleys (Fogelbe g and Seppa ¨la ¨1986). No he n a eas a e mainly la e ain, wi h ew lakes and ine- sedimen and ill deposi s. Topog aphy a ies mo e in he no he nmos pa s o he s udy a ea. Al i ude anges om sea le el up o 200 m a.s.l. in he eas e n pa s. Geomo phology o he s udy a ea is cha ac e - ized by glacial and glacio lu ial elie . Fo example, eske s cause a ia ion in o he wise qui e la coas al landscapes. The bed ock o he s udy a ea is p ima ily Fig. 2 Loca ions o he 88 sampling si es in 21 majo i e basins in Finland 123 2472 Landscape Ecol (2019) 34:2469–2485 composed o c ys alline ocks, and he soils in he a ea a e mos ly g ound mo aine (A o e al. 1990). In i e alleys and coas al egions, he e a e also sand o ma ions (e.g. dunes) and glacio lu ial deposi s wi h so ed ma e ials. Biogeog aphically, he s udy a ea anges om hemibo eal (whe e mixed o es a e dominan ) o middle bo eal ege a ion zones (whe e bo h coni e - ous and mixed o es s occu commonly) (Ah i e al. 1968). We lands wi h di e en pea deposi s a e ela i ely a e in he sou h bu a e inc easingly common no hwa ds (Ha ¨me -Ah i e al. 1988). Mean annual ai empe a u e ypically a ies om o e 5 °C in he sou hwes o 2 °C in he no h (Pi inen e al. 2012). Mean annual p ecipi a ion in he s udy a ea ypically anges om 500 mm in he no hwes coas o o e 700 mm in he sou he nmos a ea (Pi inen e al. 2012). The land use o he s udy a ea a ies subs an ially. The sou he nmos s eams a e gene ally si ua ed in human-domina ed landscapes (e.g. ag icul u al a eas), whe eas he no he n s eams a e ypically loca ed in o es -domina ed landscapes. In addi ion, local en i- onmen al condi ions o s eams a y om nea - p is ine o es ed headwa e s eams o mo e modi ied s eams in he ca chmen s o in ensi e ag icul u e (Jy ka ¨nkallio-Mikkola e al. 2017; Heino e al. 2018). Biological sampling To con ol o seasonal a ia ion in biological com- muni ies, s eam mac oin e eb a e, dia om and bac- e ial samples we e collec ed wi hin 2 weeks in Sep embe 2014. Sep embe is a sui able mon h o sampling o ganisms dwelling in bo eal s eams because di e si y is high and na u al dis u bances (e.g. snowmel -caused loods) a e ypically less equen han in he sp ing pe iod (Heino e al. 2013). In addi ion, one- ime snapsho sampling, i done wi hin a na ow ime window, is sui able o unco e ing spa ial pa e ns and species ichness– en i onmen ela ionship. Howe e , i emains silen on empo al a ia ion, which was beyond he ocus o his s udy. Fo mac oin e eb a es, a 2-min kick sample (ne mesh size 0.5 mm) co e ing mos mic ohabi a s in a i le sec ion o app oxima ely 100 m 2 was aken. Such samples con ain usually mo e han 70% o species a a si e in a gi en season (Myk a ¨e al. 2006). Mac oin e eb a es and associa ed ma e ial we e immedia ely p ese ed in e hanol in he ield, and samples we e aken o he labo a o y o u he p ocessing and iden i ica ion. Mac oin e eb a es we e iden i ied o he lowes possible axonomic le el, i.e. species, species g oup o genus. Dia om and bac e ial samples we e aken simul a- neously wi h he mac oin e eb a e sampling. A each si e, 10 andomly chosen cobble-sized s ones we e collec ed om ca. 20 cm dep h om di e en pa s o he i le si e om an a ea co e ing a ound 20 m o he s eam si e leng h. Dia oms we e collec ed om s ones by b ushing h ough a ubbe empla e (5 95cmin size) and he composi e sample was immedia ely p ese ed in a cool and da k box. In he labo a o y, dia om us ules we e cleaned om o ganic ma e ial using we combus ion wi h hyd ogen pe oxide (30%, H 2 O 2 ) and moun ed in Naph ax. A leas 500 us ules pe sample we e coun ed and iden i ied o species le el wi h a Nikon Op ipho 2 phase con as ligh mic oscope. Bac e ial samples we e wiped o om cobble sized s ones using s e ile pieces o oam plas ics and we e ozen s aigh away in he ield un il labo a o y analyses. Supplemen a y de ails o ield sampling can be ound om Vilmi e al. (2016) and Jy ka ¨nkal- lio-Mikkola e al. (2017). In he labo a o y samples we e i s eeze-d ied and hen he DNA was ex ac ed using a Powe Soil DNA Isola ion Ki (MoBio, Ca lsbad, USA). PCRs we e pe o med by Ve i i The mal Cycle (Li e Technologies) The 16S DNA egion o bac e ia was ampli ied wi h p ime s 519F 50-CAGCMGCCGCGGTAATWC-30and 926 P1 50-CCTCTCTATGGGCAGTCGGT- GATCCGT CAATTCCTTTRAGTTT-30. Unique, nine base ba code p ime s we e used o each sample. This ba coding sys em allows us o iden i y each sample in pos -sequencing analyses. The amplicons we e sequenced using ion o en semiconduc o sequencing, whe e he o al numbe o aw sequences was 2,708,611. The sequence lib a y was spli by samples and quali y il e ed based on he quali y sco es o each sequence. Sequences wi h quali y sco es below 25, sho e han 200 bp o longe han 1000 bp we e emo ed. A e quali y con ol, a o al o 549,548 sequences we e e ained, wi h an a e age sequence leng h o 3414 bp. The sequences we e clus e ed as ope a ional axonomic uni s (OTUs) using he Usea ch61 algo i hm (Edga 2010) wi h 97% 123 Landscape Ecol (2019) 34:2469–2485 2473 sequence simila i ies. OTUs (97% simila i y) we e de e mined using he Quan i a i e Insigh s In o Mic obial Ecology (QIIME) pipeline e sion 1.8.0 (Capo aso e al. 2010). The OTU da ase was a e ied o he lowes numbe o sequences de ec ed (1052), because sequence numbe s a ied among samples. The labo a o y and bioin o ma ics me hods a e p e- sen ed in mo e de ailed in Heino e al. (2015) and Jy ka ¨nkallio-Mikkola e al. (2017). In all s a is ical analysis, we used o e all species ichness (i.e. numbe o species o OTUs) o each o ganism g oup as a measu e o biodi e si y. We acknowledge ha by using such a simple measu e o biodi e si y we could lose in o ma ion ega ding a ia ion in species ai s and phylogene ic ela edness (e.g. Heino and Tolonen 2017). Howe e , as he numbe o species emains he mos commonly u ilized measu e o biodi e si y in gene al (e.g. Magu an 2004) and among eshwa e s udies in pa icula (e.g. Feld e al. 2009), we decided o concen a e on species ichness as he esponse a iable. En i onmen al a iables In each s eam si e, cu en eloci y and wa e dep h we e measu ed a 30 loca ions, and s eam wid h was measu ed om 10 c oss-s eam ansec s co e ing he s udy si e. Mean alues o he physical a iables we e used in s a is ical analyses. Wa e samples o de e - mining o al phospho us, o al ni ogen and wa e colo we e collec ed. Elec ic conduc i i y and pH we e measu ed wi h YSI-P o essional Plus wa e quali y me e (YSI Inco po a ed, Yellow Sp ings, USA). All en i onmen al a iables we e collec ed simul aneously wi h biological sampling. Ca chmen ea u es and geodi e si y Fo each s udy si e, ups eam ca chmen a ea was de e mined using a digi al ele a ion model (g id esolu ion 10 910 m, Na ional Land Su ey o Finland 2013) in A cGIS 10.5 so wa e. Ca chmen size, land use ype and geodi e si y we e u he calcula ed o each ca chmen . Land use was ob ained om CORINE Land Co e da a (20 920 m, Finnish En i onmen Ins i u e 2013). Geodi e si y a iables consis ed o geomo pholog- ical, soil- and ock- ype ichness a esolu ion o 1-km 2 g id cells (Table 1). Geomo phological, o land o m, da a we e quan i ied using land o m obse a ions, GIS-based en i onmen al a iables and gene alized addi i e modelling (Hjo and Luo o 2012). Land o m obse a ions we e ob ained by an expe geomo phol- ogis who sys ema ically examined geomo phological maps (1:50,000) and ae ial pho og aphs (*30 cm esolu ion). A e ha , he geomo phological dis i- bu ion modelling app oach was used o p edic he numbe o land o ms in each 1-km 2 g id cell co e ing he whole Finland. The e, he land o m obse a ions and GIS-based en i onmen al a iables we e used in gene alized addi i e modelling o gene a e geomo - phological ichness alues o he g id cells (Hjo and Luo o 2012,2013; Tukiainen e al. 2017a). Soil and ock- ype ichness we e calcula ed by summing he numbe o di e en soil and ock ypes in a 1-km 2 g id cell. The calcula ions we e based on digi al soil and bed ock maps, espec i ely (Geological Su ey o Finland, GSF 2010a,b). Rock- ypes we e classi ied by an expe in o 16 gene ically and geochemically dis inc classes (Tukiainen e al. 2017a). All geodi- e si y measu es we e i s calcula ed o he whole Finland and a e wa ds educed o ma ch wi h ca ch- men bounda ies in A cGIS 10.5 en i onmen . A mo e de ailed desc ip ion o he geodi e si y da a can be ound om Tukiainen e al. (2017a). Selec ion o inal p edic o a iables We selec ed al oge he nine p edic o a iables o he inal analyses, o which wo we e land use, h ee we e geodi e si y and ou we e local en i onmen al a i- ables (Table 2). Ha el e al.’s (1996) sh inkage ule ecommends including no mo e han n/10 p edic o s in he inal model, which suppo s he s ic selec ion o nine a iables o he inal analyses. The selec ion o p edic o s was based on he heo e ical and empi ical backg ound o impo an a iables o s eam o gan- isms. In addi ion, p elimina y examina ions o Pea son co ela ions be ween species ichness and en i on- men al a iables we e used o de ec he p elimina y ela ionship be ween he a iables and bio ic ichness, and o p e en mul icollinea i y. To limi he numbe o physico-chemical a iables, we selec ed he ones ha ha e been ound o in luence bio ic communi ies in bo eal s eams: pH, wid h, dep h and eloci y (Malmq is and Ma ¨ki 1994; Myk a ¨e al. 2007; Lea e al. 2009; Heino e al. 2012,2014). Fu he mo e, we 123 2474 Landscape Ecol (2019) 34:2469–2485 excluded o al phospho us, o al ni ogen and wa e colo om u he analysis because hey had low Pea son co ela ions wi h bio ic ichness in ou p elimina y analyses. In addi ion, conduc i i y was emo ed because o high co ela ion wi h land use a iables (see Appendix S1). We used wo land use classes, a i icial and ag icul u al a eas in o de o desc ibe an h opogenic e ec s in he ca chmen a ea abo e ou s eam sampling poin (Tonkin e al. 2016; Jy ka ¨nkallio-Mikkola e al. 2017; Pajunen e al. 2017). S a is ical me hods Fi s , all p edic o a iables we e es ed o no mali y and ans o med when necessa y. Fo en i onmen al a iables (wid h and dep h) and geodi e si y a i- ables, loga i hmic ans o ma ions we e used. Fo land Table 1 In o ma ion on geomo phological ea u es, ock ypes and soil ypes based on which geodi e si y a iables we e calcula ed No e ha geomo phological ichness included land o ms om a ious geomo phological p ocess g oups (e.g. see Hjo e al. 2012, Supplemen al Ma e ial 1) Geodi e si y a iable Fea u es o p ocesses Geomo phological ichness Aeolian Biogenic C yogenic Flu ial Glacigenic Glacio lu ial Li o al Ma ine Mass-was ing polygene ic bed ock Slope Wea he ing Rock- ype ichness Ul ama ic in usi e o olcanic ocks Ma ic in usi e o olcanic ocks In e media e, in usi e olcanic ocks G ani ic o esic ocks Peli ic sedimen a y ocks Conglome a es A kosic sedimen a y ocks Black schis s Qua z- ich sedimen a y ocks Sedimen a y ca bona e ocks o ca bona i es Gneisses and migma i es I on o e High-g ade me amo phic ocks Me asoma ic ocks Impac mel ocks Sulphide o e Soil- ype ichness Rock (ba e ock o hin soil co e , 1m) Till (glacigenic deposi s) S ony a eas and block ields Sand and g a el Sil Clay Gy ja (lake and sea sedimen s, [6% o ganic ma e ial) Pea 123 Landscape Ecol (2019) 34:2469–2485 2475 use a iables, a csine-squa e oo - ans o ma ion was used. Second, Pea son co ela ion was used o exam- ine cong uence be ween he bio ic ichness and he p edic o a iables (see also Appendix S1–S2). We used boos ed eg ession ees (BRTs) o explain he a ia ion o species ichness and o measu e he ela i e in luence o he di e en p edic o s on species ichness using he package ‘gbm’ ( e sion 2.1.1.) in R 3.1.2 (R Co e Team 2017). BRTs is a non-pa ame ic, machine lea ning me hod ha can be unde s ood as a p og essi e ype o eg ession modelling (Eli h e al. 2008). Machine lea ning me hods ha e many bene i s o e adi ional s a is ical models, such as hei obus ness o missing alues and mul icollinea i y in he da a (Do mann e al. 2013) and, especially in he case o ou s udy, hei abili y o handle nonlinea ela ionships and a iable in e ac ions (Eli h e al. 2008). In ecen yea s, simila me hods o decision ees ha e been used in many ields, including ecology (Thuille e al. 2003; Mouche e al. 2015; Jy ka ¨nkal- lio-Mikkola e al. 2017), land-use change (Mu ¨lle e al. 2013) and geodi e si y–biodi e si y explo a ions in e es ial en i onmen (Bailey e al. 2017; Tukiainen e al. 2017a). BRTs consis o eg ession ees, which explain he de iance o a dependen a iable by i ing simple models on pa i ions o he whole da a space. Pa i- ions a e esul s om spli ing up he da a space in o assemblages ha a e as simila as possible in e ms o esponse and ha minimize p edic ion e o s. A e - wa ds, BRTs combine simple decision ees (i.e. boos ing) by adding ees in a o wa d and s age-wise ashion o minimize he loss unc ion o he model (Eli h e al. 2008). Using he ‘gbm.s ep’ unc ion allowed us o calib a e models wi h h ee pa ame e s o speci y. Fi s , ee complexi y ( c) means he model complexi y in e ms o allowed in e ac ions be ween independen a iables. Second, bag ac ion (b ) sepa a es he inpu da a o calib a ion and e alua ion da a. Thi d, lea ning a e (l ), also known as he sh inkage pa ame e , can be speci ied. A e es ing, we se c o h ee, b o 0.75, l o 0.001 and used a Gaussian e o dis ibu ion o all esponse a iables, excep o bac e ial ichness o which he Poisson e o dis ibu ion was used. The e iciency o he models was e alua ed using he pe cen o explained de iance [(null de iance - esidual de iance)/null de iance]. To unde s and he e ec s o indi idual a iables on species ichness, he ela i e in luence (sum up o 100%) o e e y p edic o a iable was acqui ed om ‘gbm.s ep’. We also c ea ed pa ial dependency plo s o explo e he ela ionship be ween species ichness and he p edic o a iables (Mu ¨lle e al. 2013; Mouche e al. 2015). Final models we e alida ed using 10- old c oss- alida ion (CV). This me hod subsamples he da a 10 imes acco ding o de ined b (0.75). This means ha Table 2 De ails o he en i onmen al a iables used in boos ed eg ession ee analyses o examining he ela ionship be ween he en i onmen and species ichness o s eam mac oin e eb a es, dia oms and bac e ia in wes e n Finland (n = 88 s eams) En i onmen al a iable Uni Mean (min o max) Sou ce Soil- ype ichness Numbe o soil ypes 2.6 (1–6) GSF Rock- ype ichness Numbe o ock ypes 1.6 (1–6) GSF Geomo phological ichness Numbe o geomo phological ea u e ypes 4.7 (1–9) GAM Ag icul u al a eas % 16.8 (0–60.9) CORINE A i icial a eas % 2.2 (0–9.5) CORINE Wa e pH 7.2 (5.8–8.1) Field Wid h m 3.7 (0.72–15.5) Field Dep h m 0.18 (0.1–0.4) Field Veloci y m/s 0.24 (0.04–0.5) Field No e ha mean alues o he geodi e si y (soil-, ock- and geomo phological- ypes) in he able a e based on he mean alues o all 88 ca chmen s a 1 km-scale. Minimum and maximum alues o geodi e si y a iables e e o he 1 km-scale minimum and maximum alues om all he s udied ca chmen s GAM gene alized addi i e model, GSF Geological Su ey o Finland, CORINE coo dina ion o in o ma ion on he en i onmen 123 2476 Landscape Ecol (2019) 34:2469–2485 Allan JD, Cas illo MM (2007) S eam ecology. S uc u e and unc ion o unning wa e s. Sp inge , Do d ech Ande son MG, Fe ee CE (2010) Conse ing he s age: clima e change and he geophysical unde pinnings o species di e si y. PLoS ONE 5:e11554 A o K, Lai aka i I, Lukka inen H, Luukkonen E, Simonen A, Tal i ie J, Vo ma A, Vuo ela P (1990) Geology. In: Ala- lammi P (ed) A las o Finland, Folio 123. Na ional Boa d o Su ey and Geog aphical Socie y o Finland, Helsinki Bailey JJ, Do een SB, Hjo J, La e s CP, Field R (2017) Modelling na i e and alien ascula plan species ichness: a which scales is geodi e si y mos ele an ? Glob Ecol Biogeog 26:763–776 Beie P, B os B (2010) Use o land ace s o plan o clima e change: conse ing he a enas, no he ac o s. Conse Biol 24:701–710 Beie P, Hun e ML, Ande son M (2015) Special sec ion: con- se ing na u e’s s age. Conse Biol 29:613–617 Beisel J-N, Usseglio-Pola e a P, Thomas S, Mo e eau J-C (2000) The spa ial he e ogenei y o a i e bo om: a key ac o de e mining mac oin e eb a e communi ies. Hyd obiologia 422(423):163–171 Beseme K, Singe G, Ho ¨dl I, Ba in TJ (2009) Bac e ial com- muni y composi ion o s eam bio ilms in spa ially a i- able- low en i onmen s. Appl En i on Mic obiol 75:7189–7195 Capo aso JG, Kuczynski J, S ombaugh J, Bi inge K, Bushman FD, Cos ello EK, Fie e N, Conzalez Pen ˜a A, Good ich JK, Go don JI, Hu ley GA, Kelley ST, Knigh s D, Koenig JE, Ley RE, Lozupone CA, McDonald D, Muegge BD, Pi ung M, Reede J, Se insky JR, Tu nbaugh PJ, Wal e s WA, Widmann J, Ya sunenko T, Zane eld J, Knigh R (2010) QIIME allows analysis o high- h oughpu communi y sequencing da a. Na Me hods 7:335–336 Ca izo SF, Lengyel S, Kapusi F, Szabolcs M, Kaspe idus HD, Scholz M, Ma ko ic D, F eyho J, Cid N, Ca doso AC, Da wall W (2017) C i ical ca chmen s o eshwa e bio- di e si y conse a ion in Eu ope: iden i ica ion, p io i i- za ion and gap analysis. J Appl Ecol 54:1209–1218 Da ies NM, No is RH, Thoms MC (2000) P edic ion and assessmen o local s eam habi a ea u es using la ge- scale ca chmen cha ac e is ics. F eshw Biol 45:343–369 Donohue I, Jackson AL, Pusch MT, I ine K (2009) Nu ien en ichmen homogenizes lake ben hic assemblages a local and egional scales. Ecology 90:3470–3477 Do mann CF, Eli h J, Bache S, Buchmann C, Ca l G, Ca e ´G, Ca cı ´a Ma que ´z JR, G ube B, La ou cade B, Lei a ˜o PJ, Mu ¨nkmu ¨lle T, McClean C, Osbo ne PE, Reineking B, Sch o ¨de B, Skidmo e A, Zu ell D, Lau enbach S (2013) Collinea i y: a e iew o me hods o deal wi h i and a simula ion s udy e alua ing hei pe o mance. Ecog aphy 36:27–46 Dudgeon D, A hing on AH, Gessne MO, Kawaba a ZI, Knowle DJ, Le eque C, Naiman RJ, P ieu -Richa d AH, So o D, S iassny MLJ, Sulli an CA (2006) F eshwa e biodi e si y: impo ance, h ea s, s a us and conse a ion challenges. Biol Re 81:163–182 Edga RC (2010) Sea ch and clus e ing o de s o magni ude as e han BLAST. Bioin o ma ics 26:2460–2461 Eli h J, Lea hwick JR, Has ie T (2008) A wo king guide o boos ed eg ession ees. J Anim Ecol 77:802–813 Feld CK, da Sil a PM, Sousa JP, De Bello F, Bug e R, G andin U, He ing D, La o el S, Moun o d O, Pa do I, Pa ¨ el M, Ro ¨mbke J, Sandin L, Jones KB, Ha ison P (2009) Indi- ca o s o biodi e si y and ecosys em se ices: a syn hesis ac oss ecosys ems and spa ial scales. Oikos 118:1862–1871 Finnish En i onmen Ins i u e (2013) CORINE Land Co e 20 m. h ps://a aa. da a. i/web/pai uli/la auspal elu. Accessed Oc 2015 Fogelbe g P, Seppa ¨la ¨M (1986) Geomo phology. In: Alalammi P (ed) A las o Finland, Folio 122. Na ional Boa d o Su ey and Geog aphical Socie y o Finland, Helsinki F issel CA, Liss WJ, Wa en CE, Hu ley MD (1986) A hie a - chical amewo k o s eam habi a classi ica ion: iewing s eams in a wa e shed con ex . En i on Manag 10:199–214 Fulle RA, I ine KN, De ine-W igh P, Wa en PH, Gas on KJ (2007) Psychological bene i s o g eenspace inc ease wi h biodi e si y. Biol Le 22:390–394 Geological Su ey o Finland, GSF (2010a) Supe icial deposi s o Finland 1:200 000. GSF, Espoo Geological Su ey o Finland, GSF (2010b) Bed ock o Finland 1:200 000. GSF, Espoo Gi adoux P (2018) Spa ial analysis and da a mining o ield ecologis s. R package e sion 1.6.9 G ay M (2013) Geodi e si y: aluing and conse ing abio ic na u e, 2nd edn. Wiley-Blackwell, Chiches e , p 508 Ha ¨me -Ah i L, Ruuhija ¨ i R, Suominen J (1988) Vege a ion and lo a. In: Alalammi P (ed) A las o Finland, Folio 141. Na ional Boa d o Su ey and Geog aphical Socie y o Finland, Helsinki Ha el FE, Lee KL, Ma k DB (1996) Mul i a iable p ognos ic models: issues in de eloping models, e alua ing assump- ions and adequacy, and measu ing and educing e o s. S a Med 15:361–387 Heino J, G o ¨n oos M, Ilmonen J, Ka hu T, Ni a M, Paasi i a L (2013) En i onmen al he e ogenei y and be a di e si y o s eam mac oin e eb a e communi ies a in e media e spa ial scales. F eshw Sci 32:142–154 Heino J, G o ¨n oos M, Soininen J, Vi anen R, Muo ka T (2012) Con ex dependency and me acommuni y s uc u ing in bo eal headwa e s eams. Oikos 121:537–544 Heino J, Melo AS, Jy ka ¨nkallio-Mikkola J, Pe sch DK, Sai o VS, Tolonen KT, Bini LM, Landei o VL, Sil a TSF, Pajunen V, Soininen J, Siquei a T (2018) Sub opical s eams ha bou highe genus ichness and lowe abun- dance o insec s compa ed o bo eal s eams, bu scale ma e s. J Biogeog 45:1983–1993 Heino J, Muo ka T, Paa ola R (2003) De e minan s o mac oin e eb a e di e si y in headwa e s eams: egional and local in luences. J Anim Ecol 72:425–434 Heino J, Nokela T, Soininen J, Tolkkinen M, Vi anen L, Vi - anen R (2015) Elemen s o me acommuni y s uc u e and communi y–en i onmen ela ionships in s eam o gan- isms. F eshw Biol 60:973–988 Heino J, Tolkkinen M, Pi ila ¨AM, Aisala H, Myk a ¨H (2014) Mic obial di e si y and communi y–en i onmen ela- ionships in bo eal s eams. J Biogeog 41:2234–2244 Heino J, Tolonen KT (2017) Un angling he assembly o li o al mac oin e eb a e communi ies h ough measu es o 123 Landscape Ecol (2019) 34:2469–2485 2483 unc ional and phylogene ic alpha di e si y. F eshw Biol 62:1168–1179 Heino J, Vi kkala R, Toi onen H (2009) Clima e change and eshwa e biodi e si y: de ec ed pa e ns, u u e ends and adap a ions in no he n egions. Biol Re 84:39–54 Hild ew AG, Gille PS (1994) Pa chiness, species in e ac ions, and dis u bance in he s eam ben hos. In: Gille PS, Hil- d ew AG, Ra aelli DG (eds) Aqua ic ecology: scale, pa - e n and p ocess. Blackwell Science, Ox o d, pp 21–62 Hjo J, Go don JE, G ay M, Hun e ML (2015) Why geodi- e si y ma e s in aluing na u e’s s age. Conse Biol 29:630–639 Hjo J, Heikkinen R, Luo o M (2012) Inclusion o explici measu es o geodi e si y imp o e biodi e si y models in a bo eal landscape. Biodi e s Conse 21:3487–3506 Hjo J, Luo o M (2012) Can geodi e si y be p edic ed om space? Geomo phology 153–154:74–80 Hjo J, Luo o M (2013) S a is ical me hods o geomo phic dis ibu ion modeling. In: Sh ode J J (ed) T ea ise on geomo phology. Academic, San Diego, pp 59–73 Hynes HBN (1970) The ecology o unning wa e s. Li e pool Uni e si y P ess IUCN (1994) Guidelines o p o ec ed a ea managemen ca e- go ies. IUCN, Gland Jy ka ¨nkallio-Mikkola J, Heino J, Soininen J (2016) Be a di e si y o s eam dia oms a wo hie a chical spa ial scales: implica ions o biomoni o ing. F eshw Biol 61:239–250 Jy ka ¨nkallio-Mikkola J, Meie S, Heino J, Laamanen T, Pajunen V, Tolonen KT, Tolkkinen M, Soininen J (2017) Disen- angling mul i-scale en i onmen al e ec s on s eam mic obial communi ies. J Biogeog 44:1512–1523 Ka ¨ na ¨O-M, Heino J, G o ¨n oos M, Hjo J (2018) The added alue o geodi e si y indices in explaining a ia ion o s eam mac oin e eb a e di e si y. Ecol Indic 94:420–429 Kaskela AA, Rousi H, Ronkainen M, O lo a M, Babin A, Gogobe idze G, Kos amo K, Ko ilainen AT, Nee in I, Ryabchuk D, Se gee A, Zhaimoda V (2017) Linkages be ween ben hic assemblages and physical en i onmen al ac o s: he ole o geodi e si y in Eas e n Gul o Finland ecosys ems. Con Shel Res 142:1–13 Lawle JJ, Acke ly DD, Albano CM, Ande son MG, Dob owski SZ, Gill JL, Helle NE, P essey RL, Sande son EW, Weiss SB (2015) The heo y behind, and he challenges o , con- se ing na u e’s s age in a ime o apid change. Conse Biol 29:618–629 Lea G, Lewis GD (2009) Impac o ca chmen land use on bac e ial communi ies wi hin s eam bio ilms. Ecol Indic 9:848–855 Lea G, Niyogi D, Ha ding J, Dong Y, Lewis G (2009) Bio ilm bac e ial communi y s uc u e in s eams a ec ed by acid mine d ainage. Appl En i on Mic obiol 75:3455–3460 Legend e P, Legend e L (2012) Nume ical ecology, 3 d edn. Else ie , Ams e dam Leland HV, Po e SD (2000) Dis ibu ion o ben hic algae in he uppe Illinois Ri e Basin ela ion o geology and land use. F eshw Biol 44:279–301 Lena DR, C aw o d JK (1994) E ec s o land use on wa e quali y and aqua ic bio a o h ee No h Ca olina Piedmon s eams. Hyd obiologia 294:185–199 Likens GE (2013) Biogeochemis y o a o es ed ecosys em. Sp inge , New Yo k, p 208 Magu an AE (2004) Measu ing biological di e si y, 2nd edn. Blackwell Science L d., Ox o d Malmq is B, Ma ¨ki M (1994) Ben hic mac oin e eb a e assemblages in no h Swedish s eams: en i onmen al ela ionship. Ecog aphy 17:9–16 Moilanen A, Lea hwick J, Eli h J (2008) A me hod o spa ial eshwa e conse a ion p io i iza ion. F eshw Biol 53:577–592 Mouche M, Le e s C, Zupan L, Kuemme le T, Plu za C, E b K, La o el S, Thuille W, Habe l H (2015) Tes ing he e ec i eness o en i onmen al a iables o explain eu o- pean e es ial e eb a e species ichness ac oss biogeo- g aphical scales. PLoS ONE. h ps://doi.o g/10.1371/ jou nal.pone.0131924 Mu ¨lle D, Lei a ˜o PJ, Siko T (2013) Compa ing he de e mi- nan s o c opland abandonmen in Albania and Romania using boos ed eg ession ees. Ag ic Sys 117:66–77 Myk a ¨H, Heino J, Muo ka T (2007) Scale- ela ed pa e ns in he spa ial and en i onmen al componen s o s eam mac oin e eb a e assemblage a ia ion. Glob Ecol Bio- geog 16:149–159 Myk a ¨H, Ruokonen T, Muo ka T (2006) The e ec o sample du a ion on he e iciency o kick sampling in wo s eams wi h con as ing subs a um he e ogenei y. Ve h In Ve Theo Angew Limnol 29:1351–1355 Na ional Land Su ey o Finland (2013) Ele a ion model 10 m. h ps://a aa. da a. i/web/pai uli/la auspal elu. Accessed Oc 2015 Nichols WF, Killingbeck KT, Augus PV (1998) The in luence o geomo phological he e ogenei y on biodi e si y II. A landscape pe spec i e. Conse Biol 12:371–379 Pajunen V, Luo o M, Soininen J (2017) Un a elling di ec and indi ec e ec s o hie a chical ac o s d i ing mic obial s eam communi ies. J Biogeog 44:2376–2385 Pa ks KE, Mulligan M (2010) On he ela ionship be ween a esou ce based measu e o geodi e si y and b oad scale biodi e si y pa e ns. Biodi e s Conse 19:2751–2766 Passy SI (2010) A dis inc la i udinal g adien o dia om di e - si y is linked o esou ce supply. Ecology 91:36–41 Pausas JC, Ca e as J, Fe e ´A, Fon X (2003) Coa se-scale plan species ichness in ela ion o en i onmen al he e ogenei y. J Veg Sci 14:661–668 Piha H, Luo o M, Piha M, Me ila ¨J (2007) Anu an abundance and pe sis ence in ag icul u al landscapes du ing a clima ic ex eme. Glob Change Biol 13:300–311 Pi inen P, Simola H, Aal o J, Kauko an a J, Ka lsson P, Ruuhela R (2012) Clima ological s a is ic o Finland 1981–2010. Finnish Me eo ological Ins i u e, Helsinki Po NL (1997) Landscape il e s and species ai s: owa ds mechanis ic unde s anding and p edic ion in s eam ecol- ogy. J N Am Ben hol Soc 16:391–409 Po apo a MG, Cha les DF (2002) Ben hic dia oms in USA i e s: dis ibu ions along spa ial and en i onmen al g a- dien s. J Biogeog 29(2):167–187 Richa ds C, Johnson LB, Hos GE (1996) Landscape-scale in luences on s eam habi a s and bio a. Can J Fish Aqua Sci 53:295–311 123 2484 Landscape Ecol (2019) 34:2469–2485 Ridgeway G (2015) gbm: gene alized boos ed eg ession models. R package e sion 2.1.1. h p://CRAN.R-p ojec . o g/packages=gbm. Accessed No 2017 Sala OE, Chapin FS, A mes o JJ, Be low E, Bloom ield J, Di zo R, Hube -Sanwald E, Huenneke LF, Jackson RB, Kinzig A, Leemans R, Lodge DM, Mooney HA, Oes e held M, Po NL, Sykes MT, Walke BH, Walke M, Wall DH (2000) Global biodi e si y scena ios o he yea 2100. Science 287:1770–1774 Sandin L, Johnson RK (2004) Local, landscape and egional ac o s s uc u ing ben hic mac oin e eb a e assemblages in Swedish s eams. Landscape Ecol 19:501–514 Soininen J (2007) En i onmen al and spa ial con ol o esh- wa e dia oms—a e iew. Dia om Res 22:473–490 Soininen J (2015) A e ca chmen p ope ies use ul p oxies o eshwa e biodi e si y? In: Daniels JA (ed) Ad ances in en i onmen al esea ch, ol 38. No a Science Publishe s, New Yo k, pp 29–40 Soininen J, Ba els P, Heino J, Luo o M, Hilleb and H (2015) Towa d mo e in eg a ed ecosys em esea ch in aqua ic and e es ial en i onmen s. Bioscience 65:174–182 Spi ale D, Nascimbene J (2012) Spa ial s uc u e, ock ype, and local en i onmen al condi ions d i e moss and lichen dis- ibu ion on calca eous boulde s. Ecol Res 27:633–638 Sponselle RA, Ben ield EF, Vale HM (2001) Rela ionships be ween land use, spa ial scale and s eam mac oin e e- b a e communi ies. F eshw Biol 46:1409–1424 S ein A, Ge s ne K, K e H (2014) En i onmen al he e o- genei y as a uni e sal d i e o species ichness ac oss axa, biomes and spa ial scales. Ecol Le 17:866–880 R Co e Team (2017) 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. h p://www.R-p ojec .o g/. Accessed No 2017 Thuille W, Vay eda J, Pino J, Saba e S, La o el S, G acia C (2003) La ge-scale en i onmen al co ela es o o es ee dis ibu ions in Ca alonia (NE Spain). Glob Ecol Biogeog 12:313–325 Toi anen M, Hjo J, Heino J, Tukiainen H, A o ii a J, Alahuh a J (2019) Is ca chmen geodi e si y a use ul su oga e o aqua ic plan species ichness? J Biogeog 46:1711–1722 Tonkin JD, Heino J, Sunde mann A, Haase P, Ja ¨hnig SC (2016) Con ex dependency in biodi e si y pa e ns o cen al Ge man s eam me acommuni ies. F eshw Biol 61:607–620 Tukiainen H, Alahuh a J, Field R, Ala-Hulkko T, Lampinen R, Hjo J (2017a) Spa ial ela ionship be ween biodi e si y and geodi e si y ac oss a g adien o land-use in ensi y in high-la i ude landscapes. Landscape Ecol 32:1049–1063 Tukiainen H, Bailey JJ, Field R, Kangas K, Hjo J (2017b) Combining geodi e si y wi h clima e and opog aphy o accoun o h ea ened species ichness. Conse Biol 31:364–375 Va anka S, Luo o M (2012) En i onmen al de e minan s o wa e quali y in bo eal i e s based on pa i ioning me h- ods. Ri e Res Appl 28:1034–1046 Vilmi A, Alahuh a J, Hjo J, Ka ¨ na ¨O-M, Leinonen K, Rocha MP, Tolonen KE, Tolonen KT, Heino J (2017) Geog aphy o global change and species ichness in he No h. En i on Re 25:184–192 Vilmi A, Ka jalainen SM, Nokela T, Tolonen K, Heino J (2016) Un a elling he d i e s o aqua ic communi ies using dis- pa a e o ganismal g oups and di e en axonomic le els. Ecol Indic 60:108–118 Vo ¨ o ¨sma y CJ, McIn y e P, Gessne MO, Dudgeon D, P use- ich A, G een P, Glidden S, Bunn SE, Sulli an CA, Lie - mann CR (2010) Global h ea s o human wa e secu i y and i e biodi e si y. Na u e 467:555–561 Wang J, Meie S, Soininen J, Casamayo EO, Pan F, Tang X, Yang X, Zhang Y, Wu Q, Zhou J, Shen J (2017) Regional and global ele a ional pa e ns o mic obial species ich- ness and e enness. Ecog aphy 40:393–402 Wa d JV, Tockne K, A sco DB, Cla e C (2002) Ri e ine landscape di e si y. F eshw Biol 47:517–539 Wei T (2017) Visualiza ion o a co ela ion ma ix. R package e sion 0.84 Wiens JJ (2016) Clima e- ela ed local ex inc ions a e al eady widesp ead among plan and animal species. PLoS Biol 14(12):e2001104 Zeglin LH (2015) S eam mic obial di e si y in esponse o en i onmen al changes: e iew and syn hesis o exis ing esea ch. F on Mic obiol 6:454 Publishe ’s No e Sp inge Na u e emains neu al wi h ega d o ju isdic ional claims in published maps and ins i u ional a ilia ions. 123 Landscape Ecol (2019) 34:2469–2485 2485