Three ways to deliver a net positive impact with biodiversity offsets

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/ Th ee ways o deli e a ne posi i e impac wi h biodi e si y o se s © 2020 The Au ho s. Conse a ion Biology published by Wiley Pe iodicals LLC on behal o Socie y o Conse a ion Biology Published e sion Moilanen, A e; Ko iaho, Janne S. Moilanen, A., & Ko iaho, J. S. (2021). Th ee ways o deli e a ne posi i e impac wi h biodi e si y o se s. Conse a ion Biology, 35(5), 197-205. h ps://doi.o g/10.1111/cobi.13533 2021 Essay Th ee ways o deli e a ne posi i e impac wi h biodi e si y o se s A e Moilanen 1,2 ∗and Janne S. Ko iaho 3,4 1Finnish Na u al His o y Museum, Uni e si y o Helsinki, P.O. Box 17, Helsinki, FI-00014, Finland 2Depa men o Geosciences and Geog aphy, Uni e si y o Helsinki, Helsinki, FI-00014, Finland 3Depa men o Biological and En i onmen al Science, Uni e si y o Jy äskylä, P.O. Box 35, Jy äskylä, FI-40014, Finland 4School o Resou ce Wisdom, Uni e si y o Jy äskylä, Jy äskylä, FI-40014, Finland Abs ac : Biodi e si y o se ing is he p ac ice o using conse a ion ac ions, such as habi a es o a ion, man- agemen , o p o ec ion, o compensa e o ecological losses caused by de elopmen ac i i y, including cons uc- ion p ojec s. The ypical goal o o se ing is no ne loss (NNL), which means ha all ecological losses a e compensa ed o by commensu a e o se gains. We ocused on a concep ual and me hodological explo a ion o ne posi i e impac (NPI), an ambi ious goal ha implies commi men beyond NNL and ha has ecen ly ecei ed inc easing a en ion om big business and en i onmen al nongo e nmen al o ganiza ions. We iden i ied 3 main ways NPI could be deli e ed: use o an addi ional NPI mul iplie ; use o slowly de eloping pe manen o se s o deli e addi ional gains a e NNL has i s been eached du ing a sho e o se e alua ion ime in e al; and he combina ion o pe manen o se s wi h pa ially empo a y losses. An impo an and no el a ian o he las mechanism is he use o an al e na e mi iga ion hie a chy so ha gains om he adi ional hi d s ep o he mi iga ion hie a chy (i.e., onsi e ehabili a ion) a e no longe be coun ed owa d educed o se equi emen s. The ou come om hese 3 ac o s is ha o he same ecological damage, la ge o se s will be equi ed han p e iously, he eby imp o ing o se success. As a co olla y, we show ha o se s a e NNL only a 1 epheme al poin in ime, be o e which hey a e ne nega i e and a e which hey become ei he NPI o ne nega i e impac , depending on whe he pe manen o se s a e combined wi h pa ially empo a y losses o i empo a y o se gains a e combined wi h pa ially pe manen losses. To achie e NPI, o se s mus be made pe manen , and hey mus achie e NNL du ing an ag eed-upon o se e alua ion pe iod. An addi ional NPI-mul iplie and use o he modi ied mi iga ion hie a chy will deli e addi ional NPI gains. Achie ing NPI is ully condi ional on p io achie emen o NNL, and NNL o se s ha e been equen ly obse ed o ail due o inadequa e policy equi emen s, poo planning, o incomple e implemen a ion. Ne e heless, achie ing NPI becomes s aigh o wa d i NNL can be c edibly eached i s . Keywo ds: al e na e mi iga ion hie a chy, biodi e si y o se ing, ecological compensa ion, ne gain, ne nega- i e impac , pe manence T es Mane as de P opo ciona un Impac o Posi i o Ne o con Compensaciones po Biodi e sidad Resumen: La compensación po biodi e sidad es una p ác ica que consis e en usa las acciones de conse - ación, como la es au ación, manejo o p o ección del hábi a , pa a compensa las pé didas ecológicas causadas po las ac i idades de desa ollo, incluidos los p oyec os de cons ucción. La me a ípica de la compensación es la nula pé dida ne a (NNL), lo que implica que odas las pé didas ecológicas es án compensadas po las ganancias p opo cionales. Nos en ocamos en una explo ación concep ual y me odológica del impac o posi i o ne o (NPI), una me a ambiciosa que implica un comp omiso más allá de la NNL y que ecien emen e ha ecibido una mayo a ención po pa e de los g andes negocios y las o ganizaciones no gube namen ales ambien ales. Iden i icamos ∗Add ess co espondence o Moilanen A., email [email p o ec ed]i A icle impac s a emen : Achie ing a ne posi i e impac o biodi e si y o se ing is s aigh o wa d i no ne loss can be c edibly eached i s . Pape submi ed May 31, 2019; e ised manusc ip accep ed Ap il 30, 2020. This is an open access a icle unde he e ms o he C ea i e Commons A ibu ion License, which pe mi s use, dis ibu ion and ep oduc ion in any medium, p o ided he o iginal wo k is p ope ly ci ed. 1 Conse a ion Biology,Volume0,No.0,1–9 © 2020 The Au ho s. Conse a ion Biology published by Wiley Pe iodicals LLC on behal o Socie y o Conse a ion Biology DOI: 10.1111/cobi.13533 2Biodi e si y O se s es mane as p incipales median e las cuales se pod ía p opo ciona el NPI: el uso de un mul iplicado adicional de NPI; el uso de compensaciones pe manen es de len o desa ollo pa a en ega ganancias adicionales después de que p ime o se haya log ado el NNL du an e un in e alo de iempo más co o pa a la e aluación de las compensaciones; y la combinación de las compensaciones pe manen es con las pé didas pa cialmen e empo- ales. Una a ian e impo an e y no edosa del úl imo mecanismo es el uso de una je a quía al e na de mi igación de al mane a que las ganancias p o enien es del adicional e ce paso de la je a quía de mi igación (es deci , la ehabili ación in si u) ya no se con abilizan pa a los eque imien os educidos de las compensaciones. El esul ado de es os es ac o es consis e en que pa a el mismo daño ecológico se eque i án compensaciones mayo es a las necesa ias p e iamen e, aumen ando así el éxi o de las compensaciones. Como co ola io, demos amos que las compensaciones sólo alcanzan el NNl du an e un pun o e íme o en el iempo, an es del cual ienen un saldo ne o nega i o y después del cual se ans o man en un impac o ne o posi i o o un impac o ne o nega i o dependiendo de si las compensaciones pe manen es se combinan con pé didas pa cialmen e empo ales o de si las ganancias empo ales de las compensaciones se combinan con pé didas pa cialmen e empo ales. Pa a alcanza el NPI, las compensaciones deben ol e se pe manen es y deben llega al NNL du an e un pe iodo aco dado de e aluación de compensaciones. El uso de un mul iplicado adicional de NPI y de una je a quía al e ada de mi igación p opo - ciona á ganancias adicionales al NPI. La ob ención del NPI es comple amen e dependien e de la ob ención p e ia del NNL; se ha obse ado con ecuencia que las compensaciones po NNL allan debido a los eque imien os inadecuados de las polí icas, la pob e planeación o la implemen ación incomple a. Sin emba go, llega al NPI se uel e una a ea sencilla si p ime o se puede alcanza el NNL de mane a e osímil. Palab as Cla e: compensación ecológica, compensación po biodi e sidad, ganancia ne a, impac o nega i o ne o, je a quía al e na de mi igación, pe manencia In oduc ion Biodi e si y is almos uni e sally in a s a e o decline, and he a e o ecosys em deg ada ion, which unde lies much o biodi e si y loss, is accele a ing a he han s abilizing (e.g., Mace e al. 2018; IPBES 2019; IPCC 2019). In as- uc u e de elopmen is one o he majo causes o he ecosys em deg ada ion and biodi e si y loss (Maxwell e al. 2016), and i is p ojec ed ha he nex ew decades will see an unp eceden ed u he expansion o i (e.g., zu E mgassen e al. 2019b). Ecological losses om in as uc u e de elopmen as well as om o he an h opogenic land uses could po en- ially be cu bed by a la ge-scale adop ion o biodi e si y o se ing (Rainey e al. 2015; IUCN 2016a; Willemen e al. 2020). Biodi e si y o se ing (he ea e o se ing) is he p ac ice o balancing ecological losses wi h eco- logical gains gene a ed h ough ac ions, such as ecolog- ical es o a ion, es ablishmen o new p o ec ed a eas, o some o m o habi a managemen (Gibbons & Lin- denmaye 2007; McKenney & Kiesecke 2010; Wende e al. 2018). Habi a es o a ion is used o educe pas ecological deg ada ion and he eby p o ides o se gains. P o ec ed a ea es ablishmen is a common o m o so- called a oided (a e ed) loss o se s, which aim o p o- ide gains ia educ ion o p essu es ha would lead o u u e habi a deg ada ion o popula ion declines unless add essed (Gibbons & Lindenmaye 2007; Ma on e al. 2015). Simila o p o ec ion, habi a managemen ope - a es by p e en ing o slowing he habi a de e io a ion ha would occu wi hou conse a ion ac ion. Bo h p o- ec ion and es o a ion come wi h complica ions, such as pa ial gains, ime delays in he deli e y o gains, leak- age, and ques ions o addi ionali y. Moilanen and Ko iaho (2018) p o ide a uni ied ea men o hese unde lying ac o s. A leas on pape , biodi e si y o se ing seems o be an inc easingly adop ed policy ool. The e a e a leas 37 coun ies wi h manda o y biodi e si y o se ing policies o a leas some in as uc u e sec o s o habi a ypes and a u he 64 coun ies ha ecommend o enable ol- un a y o se s (zu E mgassen e al. 2019b). S ill, land use o biodi e si y o se s is limi ed compa ed wi h he scale o human impac s on he wo ld; in 2018, he e we e globally only jus o e 150,000 km2o a iable- ype o - se s comple ed o in he p ocess o being implemen ed (Bull & S ange 2018). Despi e limi ed implemen a ion so a , he ela i ely common global policy indica es ha biodi e si y o se ing can ha e an impo an ole o play in minimizing losses o e en deli e ing ne posi- i e biodi e si y impac s o compensa e o in as uc- u e de elopmen and land use. Nego ia ions a e ongoing o include combined ecosys em es o a ion and no ne loss (NNL) o ne posi i e impac (NPI) o ecosys ems in he global pos -2020 global biodi e si y amewo k (CBD 2020). O se ing is p esen ly seen as he ou h s age o he so-called mi iga ion hie a chy in which ecological loss is i s a oided al oge he , hen minimized by app op i- a e local p ojec design, co ec ed by onsi e ehabili a- ion ( es o a ion), and inally o se o si e (e.g., IUCN 2016b; OECD 2016; A lidge e al. 2018). The commonly s a ed goal o o se ing is NNL (i.e., ecological losses in- cu ed by de elopmen a e ully balanced by commen- su a e gains) (McKenney & Kiesecke 2010; Ma on e al. 2018; Maseyk e al. 2020). Despi e o se p og ams being Conse a ion Biology Volume 0, No. 0, 2020 Moilanen & Ko iaho 3 inc easingly adop ed all a ound he wo ld, he e is wide a ia ion in wha an o se is aken o mean ope a ionally (Bull & S ange 2018). Al hough NNL is hough o be di icul o achie e, and many NNL policies ha e indeed been shown o ail hei in ended aims (Cal e e al. 2015; Lindenmaye e al. 2017; Bezombes e al. 2019; zu E mgassen e al. 2019a), ecen ly an appeal was made s a ing ha a de- si able ou come o economic de elopmen should be he e en mo e ambi ious goal o NPI o biodi e si y. Ne posi i e impac is gaining ac ion om businesses and conse a ion NGOs alike (Gibbons & Lindenmaye 2007; Rainey e al. 2015; IUCN 2016a; Bull & B ownlie 2017). Al hough an o se p og am ha aims a NNL only com- pensa es o he ecological losses, NPI aims o deli e mo e, o imp o e he s a e o he en i onmen ela i e o i s p ede elopmen s a e. F om a co po a e pe spec i e, he social license o op- e a e can bene i g ea ly om well-managed and epo ed ne posi i e en i onmen al impac s (Riche e al. 2015; IUCN 2016a; OECD 2016). Despi e he po en ial, and some ini ial en husiasm o he implemen a ion o NPI o se p og ams (Rainey e al. 2015), only 38 companies ha e made commi men s o NPI policies o en i onmen since 2001, and ou o hese jus 26 s ill had ac i e com- mi men s in 2016 (de Sil a e al. 2019). Mo eo e , an analysis o he sus ainabili y epo s o he op 100 global companies e ealed ha al hough 49% o hese big busi- nesses men ion biodi e si y, hey in gene al gi e i lim- i ed ea men , and no company has epo ed quan i a- i e ou comes o hei ac i i ies o biodi e si y (Addison e al. 2019; Smi h e al. 2019). A conclusion ha bio- di e si y impac s a e unde epo ed, and in mos cases con ined o gene ic o ague s a emen s, has also been made ecen ly (e.g., de Sil a e al. 2019; Skouloudis e al. 2019; Weissge be e al. 2019). One eason o he lack o up ake by hese companies may be ha a clea ope a- ional ecipe is missing and hus he e is signi ican con- usion as o how––i a all––NPI can be achie ed (Walke e al. 2009; Rainey e al. 2015; Bull & B ownlie 2017). We de ised 3 ways o do so. Th ee Ways o Achie e NPI wi h Biodi e si y O se s Ne posi i e impac can be achie ed h ough he appli- ca ion o an NPI mul iplie on op o he equi emen s o NNL h ough long- e m gains ha g ow om slowly de eloping pe manen o se s and h ough pe manen o se s combined wi h pa ially empo a y losses. A e- duc ion in long- e m losses can a ise in 2 ways, om ces- sa ion o indi ec impac s and om he use o he gains om he hi d s ep o he mi iga ion hie a chy (onsi e es o a ion) owa d NPI a he han o educed NNL e- qui emen s. Be o e examining hese ou es o NPI, we ei e a e ha i NNL canno be achie ed o he p ojec , nei he can NPI. In his concep ual ea men , we as- sumed ha o se ing e o s equi ed o achie e NNL ha e i s been success ully p ede e mined and accoun o all ele an conside a ions, such as pa ial es o a ion and a oided loss gains, ime delays, leakage, simpli ied biodi e si y measu emen , unce ain y, e c. (Moilanen & Ko iaho 2018). (The common wisdom is o cou se, ha o se s–– o wha e e eason––usually ail o achie e NNL.) No ably, he le el o lexibili y in space, ime, and biodi e si y allowed in he de e mina ion o NNL (Bull e al. 2015; Moilanen & Ko iaho 2018) has no bea ing on he achie emen o NPI. The same mechanisms apply i espec i e o he deg ee o lexibili y allowed. The i s and simples mechanism o achie ing NPI is o apply an a ea mul iplie addi ional o he NNL o se - ing e o , which hen p oduces he gi en deg ee o NPI (Moilanen e al. 2009; Bull & B ownlie 2017; Moilanen & Ko iaho 2018). When he o se ing e o o each NNL has been co ec ly de e mined, hen applying an addi- ional mul iplie o , o example, 1.3 would p oduce 30% NPI ela i e o he o iginal esidual loss caused by he p ojec . Applica ion o a la ge enough mul iplie is he mos common eason o success o an NNL policy (zu E mgassen e al. 2019a). Fundamen ally, i he equi e- men s o NNL a e known, doing mo e will p o ide mo e han NNL, which implies NPI. The second mechanism o deli e NPI is he use o slowly ma u ing pe manen o se s o deli e long- e m NPI. The desi abili y o pe manen o se s has o cou se been poin ed ou many imes (e.g., McKenney & Kiesecke 2010; an Oos e zee e al. 2012; Moilanen & Ko iaho 2018), bu less so in he con ex o NPI. Ne posi i e impac om his mechanism builds on a ypi- cal (hypo he ical), mono onically inc easing esponse o habi a condi ion o es o a ion measu es (Figs. 1a & 1b). (A oided loss gains ypically de elop slowly o e yea s, jus like es o a ion gains, so he logic o he igu e also applies o a oided loss o se s.) As backg ound, unless all o se s a e a p io i pu chased om a habi a bank, an o se e alua ion ime ame has o be se , o e which NNL is achie ed, ha app op ia ely accoun s o he slow de elopmen o es o a ion and a oided loss gains (Moila- nen & Ko iaho 2018). Unless se explici ly, such a ime- ame will be adop ed implici ly (e.g., ia how much gain is assumed om habi a es o a ion). Because he e alua- ion pe iod needs o be easonable and no so long ha socie al c edibili y is los (e.g., 20–30 yea s) bu es o a- ion bene i s ma u e mo e slowly, habi a condi ion will in some cases con inue o slowly imp o e a e he end o he e alua ion pe iod. Fo example, o es s ake many decades o e en cen u ies o ully ma u e. This means ha he o se will gene a e u he gains he eby ansi- ioning o long- e m NPI condi ional on losses emaining s able o dec easing o e ime (Fig. 1b). The deg ee o NPI achie ed ia his mechanism can be es ima ed by di iding long- e m a e age gains by sho - e m a e age Conse a ion Biology Volume 0, No. 0, 2020 4Biodi e si y O se s Figu e 1. Illus a ion o es o a ion o se gains and he e alua ion ime ame: (a) habi a condi ion s a s o eco e and gains (g ay a ea) accumula e a e es o a ion has been implemen ed (pe manence is c i ical o he long- e m pe o mance o he o se ); (b) a e age o se gains (dashed lines) can be calcula ed o e di e en e alua ion pe iods (Te), wi h gains ansi ioning om NNL o NPI o e ime; (c) pe manen o se s can deli e NPI when coupled wi h pa ially empo a y loses; and (d) local ehabili a ion can also gene a e NPI ia he educ ion o a e age long- e m losses, assuming ha i is no allowed o educe o se equi emen s. gains, ma ked in Fig. 1 wi h gain(2Te) and gain(Te), e- spec i ely. The hi d mechanism o deli e NPI is combining pe - manen o se s wi h (pa ially) empo a y losses (Figs. 1c & 1d). Fo example, losses caused by indi ec dis u - bance, such as noise, ligh , dus , human p esence, e c., may decline quickly a e he end o a empo a y p ojec , such as esou ce ex ac ion (Fig. 1c), allowing o NPI o de elop, again ully condi ional on he pe manence o he o se s and p o ided ha all losses ha e i s been o se o NNL inside he ag eed-upon o se e alua ion pe iod. O cou se, losses a e e ec i ely pe manen o many ypes o p ojec s, such as oads, dams, o expanded u ban a eas, la gely p ecluding NPI ia he second mech- anism in hese cases. The deg ee o NPI achie ed ia his mechanism can be es ima ed by di iding sho - e m a e - age losses by long- e m a e age losses, ma ked in Fig. 1 by L(Te)andL(2Te), espec i ely. An impo an a ian o he hi d mechanism o achie e NPI is educ ion o long- e m losses ia a majo ans o ma ion o he hi d s ep o he mi iga ion hie - a chy. We p opose ha gains om local ehabili a ion should no be deduc ed om he o e all esidual loss caused by he new de elopmen . In o he wo ds, hey should no be coun ed owa d educing he esidual im- pac coun ed om he di ec oo p in o he p ojec . Local ehabili a ion should ne e heless be implemen ed ully. As a consequence, all losses a e i s o se o NNL, and any la e gains om local ehabili a ion con ibu e owa d NPI ins ead (Fig. 1d). E ec i ely, we ecas local ehabili a ion as a p ice o be paid owa d NPI. Wha we p opose is a eo ganiza ion o he mi iga ion hie - a chy whe ein i s hi d s age becomes he las s ep and p o ides NPI a he han educing equi emen s o NNL (Fig. 2). This change in how he mi iga ion hie a chy is pe cei ed also educes he signi icance o unce ain ies inhe en in local ehabili a ion due o ime delays o di - icul ies wi h he es o a ion o a signi ican ly damaged habi a (Schoukens & Clique 2016). Addi ional ac ion o NNL i sel is shi ed o wa d in ime, allowing mo e immedia e moni o ing and e i ica ion o o se gains. Ne e heless, local ehabili a ion is likely o eco e only a ac ion o losses, leading o a smallish deg ee o NPI only (Fig. 1d). As some hing o pay a en ion o, he e- qui emen s o he quali y o local ehabili a ion should no be signi ican ly educed i i is implemen ed o NPI ins ead o NNL. The message o he second and hi d mechanisms is ha pe manen gains a e c i ical o o se ing o achie e NPI. I o se s a e pe manen , i is likely ha gains will con inue o acc ue h ough ime, e en i slowly, allowing he ansi ion om NNL o NPI (Fig. 1a). Howe e , com- bining pe manen o semipe manen losses wi h empo- a y gains seems o ensu e ailu e o he o se ing, lead- ing o ne nega i e impac (NNI) o ne loss (Fig. 1a). Fo example, conside a 20-yea empo a y o se o a Conse a ion Biology Volume 0, No. 0, 2020 Moilanen & Ko iaho 5 Figu e 2. Re ised mi iga ion hie a chy o impac a oidance, minimiza ion, and o se ing in which gains om local ehabili a ion a e no deduc ed om he o e all loss caused by he p ojec and o se s p o ide no ne loss (NNL) wi hou eliance on local ehabili a ion, which la e con ibu es o ne posi i e impac (NPI) (black backg ound, loss o ecological unc ion). oad o u ban a ea ha could exis e ec i ely o e e . An NNI om empo a y o se s can e en app oach comple e loss o ecological alue i he o se is e alua ed o e a e y long pe iod, sugges ing ha all pas , p esen , and u u e o se p ojec s ha use empo a y o se s o o - se pa ially o ully pe manen losses a e p ime candi- da es o majo ailu e o NNL. In e es ingly, NNL i sel occu s only a an epheme al poin in ime: be o e his poin o se s a e ne nega i e, a his poin hey ha e im- p o ed jus enough in ecological quali y o each NNL, and a e his poin NPI ollows wi h u he ma u a ion o gains, p o ided ha he o se is pe manen . O e all, he dis inc ion be ween pe manen and empo a y is su - icien by i sel o make he di e ence be ween po en ial long- e m NPI and almos ce ain NNI, as illus a ed in Fig. 1a. O cou se, pe manence alone canno gua an ee NPI. The a ea and quali y o he o se needs o be o h- e wise adequa e o achie e NNL inside he s a ed o se e alua ion pe iod. Rele an o selec ion o o se ac ion, he equi emen o pe manence is a se ious complica- ion o o se s implemen ed ia habi a managemen o main enance because i is di icul o gua an ee in pe - pe ui y unding and pe manence o on- he-g ound op- e a ions ha need epe i ion yea a e yea and decade a e decade (No on & Wa bu on 2015). Finally, Table 1 summa izes he deg ee and iming o NPI p oduced by he me hods desc ibed abo e. Only gains om an NPI mul iplie (MNPI)a egene a edsimul- aneously wi h he NNL o se ing. The o he 2 o ms o NPI can only be gene a ed a e he o se e alua ion pe iod ends, which could coincide wi h he end o he p ojec . A eal-wo ld o se case will p obably in ol e many ypes o impac s and di e en o se ac ions imple- men ed in di e en en i onmen s. I so, he e alua ion o p ojec -le el NPI equi es agg ega ion o e all impac s and hei o se s. Requi emen ha NNL be Achie ed Fi s The majo p e equisi e o achie ing NPI is ha NNL be co ec ly speci ied i s , meaning ha all o he chal- lenges o achie ing NNL need o be sol ed be o e mo ing o NPI. We emphasize ha , in gene al, ailu e o o se - ing has been mo e he ule han he excep ion, and way bigge o se s (la ge mul iplie s) a e needed o achie e NNL han he e is poli ical will o p esen ly. Failu e o o se s can be a ibu ed o insu icien policy equi e- men s (i.e., lack o adequa e speci ica ion and en o ce- men ) (e.g., Cu an e al. 2013; Spash 2015; Guille & Conse a ion Biology Volume 0, No. 0, 2020 6Biodi e si y O se s Table 1. Timing and deg ee o ne posi i e impac (NPI) p oduced by he 3 sou ces o NPI.∗. Timing and deg ee o NPI, ela i e o NNL a he end o he o se e alua ion ime ame (Te) Th ee NPI componen s by Teby 2 ×Te NPI mul iplie MNPI MNPI Inc eased long- e m gains om slowly de eloping pe manen o se ac ions (Fig. 1b) 0RP Reduced long- e m losses combined wi h pe manen gains, including educed indi ec losses om dis u bance and local ehabili a ion om he al e na e mi iga ion hie a chy (Figs. 1c, d and 2). 0RL Wi h all componen s used MNPI MNPI ×RP×RL ∗See Fig. 1 o g aphical explana ion o RP=gains(2Te)/gains(Te)andRL=L(Te)/L(2Te). Semal 2018), poo design (e.g., Walke e al. 2009; Gib- bons e al. 2018; Moilanen & Ko iaho 2018), incomple e implemen a ion o he o se s (e.g., Gibbons e al. 2018; Weissge be e al. 2019), o e en aud (Moilanen & Ko i- aho 2018). Se e al e iews ha e been conduc ed on he ecological ou comes o o se p ojec s (e.g., Bi keland & Knigh -Lenihan 2016; Theis e al. 2020; zu E mgassen e al. 2019a). In hei e iew o compliance wi h o se - ing policy, Theis e al. (2020) conclude ha al hough many p ojec s a e o icially labeled a success due o hei high compliance wi h legisla i e equi emen s, he e is only a weak posi i e ela ionship be ween compliance and success in achie ing good ecosys em unc ion. This sugges s ha he policy equi emen s do no suppo c edible ecological equi emen s o NNL (Moilanen & Ko iaho 2018). The ope a ional decisions ha de ine he ecological success o a biodi e si y o se ing p ojec ha e been dis- cussed in he scien i ic and g ay li e a u e (e.g., OECD 2016; Ga dene e al. 2013; G imm & Köppel 2019). Moilanen and Ko iaho (2018) g ouped impo an o se design ac o s a ound objec i es, cha ac e is ics o o - se ac ions, and 3 well-known undamen al axes o he ecological eali y: space, ime, and biodi e si y. The ac- o s ela ed o objec i es a e he deg ee o adhe ence o he mi iga ion hie a chy; he de ini ion o NNL, including e ec o unce ain y; and he desi ed deg ee o compen- sa ion ela i e o NNL. Fo he spa ial axis, decisions o be made include he ex en o he o se implemen a ion a ea (how a om damage can o se s be implemen ed) and he spa ial e e ence ame o biodi e si y alua ion (e.g., local, na ional, and con inen al), which in luences, o example,wha couldbeconside ed adingup.Fo he empo al axis, he decisions a e whe he he o se is empo a y o pe manen and wha he o se e alua ion pe iod and s eng h o ime discoun ing o delayed gains will be. Fo biodi e si y, decisions include how biodi e - si y will be measu ed and whe he o how ading up is allowed. Fo ac ions, addi ionali y o ac ions, habi a es o a ion esponse unc ions, esponse o a oided loss ac ion, baseline o a oided loss compa ison, and leakage associa ed wi h a oided loss all need o be add essed. Gains and losses should be balanced accoun ing o a leas hese 15 design ac o s and quan i ies. Igno ing any o hem du ing planning may lead o some unin ended ou come o ailu e in o se ing. Fo ins ance, igno ing o unde es ima ing ime delays in es o a ion gains, leak- age, o unce ain ies can s aigh o wa dly lead o o se s speci ica ions ha a e in eali y inadequa e o achie ing NNL. One eason o se s ail ecologically seems o be ha o se s do no apply o all ypes o losses. Ano he is an imbalance be ween he e alua ion o biodi e si y losses and biodi e si y gains such ha he loca ion, na u e, and ex en o impac s a e well documen ed, whe eas ex- pec ed gains a e p edic ed in a compa a i ely ague and unce ain manne (Weissge be e al. 2019; zu E mgassen e al. 2019a). This imbalance is in pa una oidable be- cause losses conce n obse able biodi e si y and can be measu ed on he g ound, whe eas he gains will only oc- cu in he u u e and hus can e e only be es ima ed o p ojec ed wi h unce ain y. Ne e heless, we belie e ha he science o es o a ion ecology is ma u e enough o p o ide he means o p ojec he gains wi h eason- able and su icien accu acy, a leas in many ecosys ems (IPBES 2018). Thus, he imbalance in ope a ionaliza ion is a ailu e ha can be add essed by inc eased emphasis on capaci y building o allow compe en es ima ion o gains in planning. I policy de elope s, consul an s, plan- ne s, e c., we e knowledgeable abou he ope a ional e- qui emen s o success ul o se s, he ha sh c i ique o biodi e si y o se ing pe se migh be answe ed. In es - ing in p ope knowledge ans e would likely build ca- paci y and esul in mo e success ul o se s. The Al e na e Mi iga ion Hie a chy As a co e componen o his wo k, we p opose an al e - na i e o he adi ional mi iga ion hie a chy (e.g., IUCN 2016b; A lidge e . al. 2018) ha is used o posi ion o - se ing in he b oad scheme o en i onmen al impac Conse a ion Biology Volume 0, No. 0, 2020 Moilanen & Ko iaho 7 a oidance a he beginning o p e y much e e y pub- lica ion abou o se s (Fig. 2). Fi s , al hough we do no a all disag ee wi h he 2 i s s eps o he mi iga ion hi- e a chy, we a e skep ical abou hei unc ionali y. S ep one, a oidance, can ce ainly be implemen ed by mo - ing a p ojec such as a oad in o an ecologically less ha m ul loca ion, and land-use planning ools o doing his a e eadily a ailable (e.g., Ka eksela e al. 2013). Ne e heless, we belie e ha economic ac i i y is gen- e ally encou aged and ha cancella ion o p ojec s is un- common. I has indeed been ound ha he e is o en a ailu e o a oid impac s (Phalan e al. 2017) and ha en i onmen al impac assessmen s a ely ind signi ican impac s, which hen allows p ojec s o go ahead despi e some en i onmen al impac s (Mu ay e al. 2018). S ep 2 is minimiza ion o impac s, o example, ia planning o echnical solu ions. P ojec impac s seem o be educed o he le el o compliance wi h en i onmen al egula ion (Theis e al. 2020), bu i is ques ionable whe he busi- nesses a e willing o in es subs an ial ex a money in o educing impac s e en u he . Pu ing hese ini ial mis- gi ings aside, ou main a gumen he e is abou he hi d le el o he mi iga ion hie a chy, local ehabili a ion. We p opose ha i should become he las s ep in he hie a - chy and ha i s ole be ecas in o a ime-delayed sou ce o (ideally manda o y) NPI a he han allowing i o be coun ed owa d he educ ion o he o iginal esidual im- pac s o he p ojec ha need o be o se (Fig. 2). The e is a logical ine i abili y ha he ans o ma ion o he mi iga ion hie a chy would imp o e o se pe o - mance, subjec o app op ia e egula ion and checks o ensu e p ope implemen a ion. Fi s , because local e- habili a ion would no longe coun owa d he educ- ion o p ojec ed losses, mo e o se ac ion would be equi ed o p oduce he gains needed o NNL. Second, mo e conse a ion ac ion would be b ough o wa d in ime. Inc eased o se s can be implemen ed immedia ely o e en pu chased om a habi a bank, which could be decades ea lie han ehabili a ion du ing o a e he end o he p ojec . Thi d, unce ain y would be educed be- cause he e would be less conce n abou he easibili y, eliable implemen a ion, o e all success, and e i ica ion and du abili y o es o a ion e o in a signi ican ly dam- aged impac a ea a in he u u e du ing o a e he end o he p ojec . Mo eo e , he e is he isk o any p ojec ex ending om wha was expec ed o iginally, he eby delaying u he he deli e y o gains om local eha- bili a ion. Hence, he e a e mul iple easons o gi ing up he p ac ice o allowing local ehabili a ion o educe he size o he o se needed. In o he wo ds, we p o- pose adop ion o he al e na i e mi iga ion hie a chy in Fig. 2 ha mo es owa d NPI. We also p opose ha NPI be compulso y, as he p ice paid o he p i ilege o he o se s being allowed in he i s place, implying ull com- mi men o local ehabili a ion on op o ini ial NNL and NPI o se s ha should be pe manen . F om NNL o NPI Al hough NNL has been discussed ex ensi ely, he e is no ha much li e a u e ocusing on he ansi ion om NNL o NPI. In one o he ea lies ea men s, Bull and B ownlie (2017) a gue ha achie ing NPI is undamen- ally di e en om achie ing NNL and ha mo ing om one o he o he p esen s signi ican challenges ha will be less i ial han o en hough . They suppo his po- si ion wi h 4 lines o a gumen s, which we do no ind in luence ou easoning: i s , NNL and NPI ha e dis inc unde lying conse a ion philosophies; second, he e is unce ain y in achie ing NNL; hi d, app op ia e ames o e e ence may depend on he case; and ou h, s ake- holde expec a ions o NNL and NPI may di e . We con- side ed each o hese a gumen s. The i s a gumen e ol es exclusi ely a ound he in- kind and ou -o -kind o se s and he deg ee o lexibili y allowed (Bull & B ownlie 2017). O e all, he dis inc ion be ween in-kind and ou -o -kind biodi e si y is an illusion c ea ed by he accu acy, o lack he eo , wi h which bio- di e si y is measu ed. Taken o he ex emes, e e y indi- idual and ecological communi y is unique and hus i e- placeable in he ully in-kind sense (Moilanen & Ko iaho 2018). Ou a gumen is ha doing mo e han equi ed o NNL will p o ide NPI i espec i e o he deg ee o lexibili y allowed in he NNL o se o in he addi ional ac ions done o NPI. I would be pe ec ly easonable o allow ele a ed lexibili y o addi ional ac ions ha a e made wi h he in en ion o deli e ing NPI. The second a gumen es s on he claim ha mee ing he NNL objec i es al eady equi es o e compensa ing o losses, which makes i di icul o speci y how la ge he gains should be o mee he NPI objec i es (Bull & B ownlie 2017). Acco ding o he p esen easoning, i he le el o compensa ion needed o NNL is known, he ansi ion o NPI is s aigh o wa d (e.g., by apply- ing he NPI mul iplie ). We ei e a e ha he majo p e- equisi e o achie ing NPI is ha he gains equi ed o achie e NNL a e i s co ec ly speci ied, which na u ally accoun s o ele an unce ain ies. I his is no he case, he p oblem is no wi h he ansi ion om NNL o NPI bu a he wi h he speci ica ion o NNL i sel . The hi d a gumen is based on he ac ha he choice o e e ence ame (ecological baseline; coun e ac ual) can signi ican ly in luence he gains expec ed (Bull & B ownlie 2017). Consequen ly, case-speci ic es ima es o gain and he le el o compensa ion needed o NNL will be a ec ed. Al hough his is a ac , in an o se - ing p ojec , he gains needed o NNL and NPI bo h depend on he same esidual loss caused by he de el- opmen . The choice o e e ence ame in no way a - ec s he easibili y o any o he 3 NPI mechanisms we desc ibed. The ou h a gumen is abou public o s akeholde pe cep ion o he c edibili y o NPI objec i es (Bull & Conse a ion Biology Volume 0, No. 0, 2020 8Biodi e si y O se s B ownlie 2017). We do no see public pe cep ion as a undamen al ecological o logical p inciple ha de ines wha can be achie ed wi h o se ing. Ce ainly, s ake- holde suppo and b oad accep ance can gene a e own- e ship ha acili a es good design, implemen a ion, and moni o ing o o se s, bu i does no in luence, o exam- ple, he abili y o habi a es o a ion o deli e ecological gains needed o NNL o NPI. Based on he e iews ci ed abo e, i is ne e heless clea ha ue p oblems exis wi h inadequa e policy equi emen s, design, and implemen a ion o o se s. The e o e, a majo challenge o biodi e si y o se ing, be he objec i e NNL o NPI, seems o be knowledge ans e and capaci y building so ha unde s anding in he scien i ic communi y is communica ed o socie y and ope a ionalized in he implemen a ion o he o - se s. Also, businesses would be equi ed o accep he compa a i ely la ge o se a ea equi emen s and cos s ha a ise when accoun ing o pa ial addi ionali y, pa - ial and delayed gains o es o a ion and p o ec ion, ime discoun ing, leakage, unce ain y, addi ional NPI equi e- men s, and possible ex a mul iplie s ha compensa e o lexibili y in space o biodi e si y (Moilanen & Ko- iaho 2018). The equi emen o uly pe manen o - se s, as p esen ed he e, would also cause p oblems in con ex s whe e legisla i e o adminis a i e p ac ices do no suppo pe manen land alloca ion o conse a ion o he applica ion o o he e ec i ely pe manen mech- anisms. Tha said, legisla ion is a human cons uc and can be changed o allow pe manen o se s i he e is he public desi e o do so. O e all, he p esen a gumen emphasizes how NNL and NPI a e una oidably dependen on he pe iod o e which he balance o losses and gains is e alua ed (Fig. 1), meaning ha ques ions o ime (i.e., o se e al- ua ion pe iod), de elopmen o gains, pe manence, and ime discoun ing become o p ima y impo ance o o - se success ( o u he discussion o hese in he con ex o achie ing NNL, see Moilanen and Ko iaho [2018]). Fo simplici y, we p esen ed ma e ial wi hou he com- plica ion o ime discoun ing, which can, howe e , be implemen ed in calcula ions ha apply s anda d ime- discoun ing echniques (Lai ila e al. 2014; Moilanen & Ko iaho 2018). Ou a gumen and Fig. 1 a e p esen ed in e ms o agg ega e gains in ela ion o agg ega e losses. In on- he-g ound planning, calcula ion o agg ega e loss needs an in e media e s ep o accoun o he size and quali y o a ea a ec ed and ac ional ecological losses he ein. Likewise, es ima ion o gains needs o accoun o he size o o se a ea and ac ional gains he ein, plus all he o he ac o s summa ized by Moilanen and Ko iaho (2018). As a inal cau ion, he e a e some en- i onmen s, such as slowly ma u ing la e successional ecosys ems, o which c edible o se ing may be impos- sible i any hing close o in-kind is equi ed (Pilg im e al. 2013). This means ha he abili y o o se ing o deli e NNL and NPI in a c edible manne has o be examined case by case. Once he equi emen s o NNL ha e been speci ied and once i has been e i ied ha adequa e implemen a- ion op ions exis (which will no always be he case), he ansi ion o NPI becomes simple enough. O se ing is only one pa o he solu ion o global en i onmen- al p oblems. Ne e heless, wi hin i s scope, i should be made as good a ool as possible. Acknowledgmen s A.M. was suppo ed by he S a egic Resea ch Council p ojec IBC-Ca bon, g an 312559. We hank I. R. Pe e - son o cons uc i e commen s on he manusc ip . Li e a u e Ci ed Addison PF, Bull JW, Milne -Gulland EJ. 2019. Using conse a ion sci- ence o ad ance co po a e biodi e si y accoun abili y. Conse a- ion Biology 33:307–318. A lidge WN, e al. 2018. A global mi iga ion hie a chy o na u e con- se a ion. BioScience 68:336–347. Bezombes L, Ke bi iou C, Spiegelbe ge T. 2019. Do biodi e si y o - se s achie e no ne loss? An e alua ion o o se s in a F ench depa - men . Biological Conse a ion 231:24–29. Bi keland J, Knigh -Lenihan S. 2016. Biodi e si y o se ing and ne pos- i i e design. Jou nal o U ban Design 21:50–66. Bull JW, B ownlie S. 2017. The ansi ion om no ne loss o a ne gain o biodi e si y is a om i ial. O yx 51:53–59. Bull JW, Ha dy MJ, Moilanen A, Go don A. 2015. Ca ego ies o lexi- bili y in biodi e si y o se ing, and hei implica ions o conse a- ion. Biological Conse a ion 192:522–532. Bull JW, S ange N. 2018. The global ex en o biodi e si y o se imple- men a ion unde no ne loss policies. Na u e Sus ainabili y 1:790– 798. Bull JW, e al. 2020. Ne posi i e ou comes o na u e. Na u e Ecology and E olu ion 4:4–7. Cal e C, Napoléone C, Salles JM. 2015. The biodi e si y o se ing dilemma: be ween economic a ionales and ecological dynamics. Sus ainabili y 7:7357–7378. CBD (Con en ion on Biological Di e si y). 2020. Ze o d a o he pos - 2020 global biodi e si y amewo k. CBD/WG2020/2/3. CBD, Mon- eal. Cu an M, Hellweg S, Beck J. 2013. Is he e any empi ical suppo o biodi e si y o se policy? Ecological Applica ions 24:617–632. de Sil a GC, Regan EC, Polla d EHB, Addison PFE. 2019. The e olu- ion o co po a e no ne loss and ne posi i e impac biodi e si y commi men s: unde s anding appe i e and add essing challenges. Business S a egy and he En i onmen 28:1481–1495. Ga dne TA, e al. 2013. Biodi e si y o se s and he challenge o achie - ing no ne loss. Conse a ion Biology 27:1254–1264. Gibbons P, Lindenmaye DB. 2007. O se s o land clea ing: no ne loss o he ail wagging he dog? Ecological Managemen and Res o a- ion 8:26–31. Gibbons P, Macin osh A, Cons able AL, Hayashi K. 2018. Ou comes om 10 yea s o biodi e si y o se ing. Global Change Biology 24:e643-54. G imm M, Köppel J. 2019. Biodi e si y o se p og am design and im- plemen a ion. Sus ainabili y 11:6903. Guille F, Semal L. 2018. Policy laws o biodi e si y o se ing as a conse a ion s a egy. Biological Conse a ion 221:86–90. Conse a ion Biology Volume 0, No. 0, 2020