Monitoring High Nature Value Grassland in Transylvania, Rom ania Laura SUTCLI FFE1 and Kryst yna LARKHAM2 1 Dept Veget at ion Analysis, Univ. Göt t ingen, Unt ere Karspüle 2, 37073 Göt t ingen, Germ any sut cliffe.laura@gm ail.com 2 I m perial College London, Sout h Kensingt on, London SW7 2AZ, UK KEYW ORD S: HNV farm ing, Rom ania, grassland biodiversit y, agri- environm ent schem es, plant indicat or species, m onit oring. ABSTRACT Sem i- nat ural grassland has in t he last 50 years becom e increasingly rare in m uch of Europe, however, low levels of int ensificat ion and sm all- scale farm ing have led t o t he preservat ion of subst ant ial areas of t his valuable habit at in Rom ania. Nevert heless, here t oo land m anagem ent pract ices have in recent years begun t o change, and t here is increasing t hreat t o Rom anian grassland biodiversit y from abandonm ent and int ensificat ion. One m eans of count ering t his t hreat is t o offer financial support for ‘biodiversit y- friendly’ farm ing t hrough agri- environm ent schem e paym ent s, such as t he schem e init iat ed in Rom ania for High Nat ure Value grassland in 2007. One im port ant aspect of agri- environm ent schem es is t he m onit oring of it s efficacy in m aint aining or im proving t he ecological qualit y of t he far m land. Therefore, t his invest igat ion uses m ult ivar iat e analysis of veget at ion survey dat a t o ident ify a list of plant indicat or species t hat can be used t o m onit or t he qualit y of lowland grassland under agrienvironm ent schem e, based on t he st udy region of Sout hern Transylvania. I N TROD UCTI ON Low- int ensit y agricult ural habit at s are one of t he m ost valuable sources of biodiversit y in Europe ( Bignal & McCracken, 1996) . This is part icularly apparent in sem i- nat ural grasslands, i.e. nat urally occurring ( not plant ed) veget at ion t hat is regularly grazed or cut , result ing in a st at e t hat m im ics nat ural habit at s ( Beaufoy, 2008) , which can cont ain som e of t he highest plant densit ies in Europe, if not t he world. Nevert heless, sem i- nat ural grassland is under t hreat due t o it s low econom ic viabilit y in m odern agricult ure. A dram at ic decline in t he surface area of sem i- nat ural grassland has been seen in t he last cent ury, part icular ly in nort hern and west ern European count ries such as t he UK, where som e 97% of sem i- nat ural grassland has been lost since 1930 ( Fuller, 1987) . The m aj or drivers of t his loss have been int ensificat ion t o increase yields of m ore fert ile ar eas, and associat ed abandonm ent of m ore m arginal land. I n recognit ion of bot h t he im port ance of areas of high agricult ural biodiversit y and t he t hreat t hey face, t he concept of High Nat ure Value ( HNV) farm land was creat ed t o describe “ t hose areas in Europe where agricult ure is a m aj or ( usually t he dom inant ) land use and where t hat agricult ure support s or is associat ed wit h eit her a high species and habit at diversit y, or t he presence of species of European conservat ion concern, or bot h.” ( Andersen et al., 2003) . Eligible farm land is ident ified on t he basis of 1) a high proport ion of sem i- nat ural veget at ion; 2) t he dom inance of low- int ensit y agricult ure or a m osaic of sem i- nat ural and cult ivat ed land and sm all- scale feat ures; 3) t he presence rare species or a high proport ion of European or world populat ions. The conservat ion of HNV farm land is now an EU obj ect ive and m easures such as agrienvironm ent schem es ( AES) t o provide support for t hese areas have been built int o t he 20072013 rural developm ent plans of t he Mem ber St at es ( EEA, 2004) . As part of it s accession t o t he EU in 2007, Rom ania now provides agri- environm ent paym ent s for areas ident ified as HNV ( 34% of t he ut ilisable agricult ural area: Paracchini et al., 2008: see Figure 1) . Figure 1: Current dist ribut ion of HNV grassland ( shaded grey) in Rom ania according t o CORI NE Land Cover dat a ( adapt ed from : NRDP 2007) . St udy area out lined w it h a black box. To qualify for t he paym ent s, landowners m ust follow cert ain m anagem ent prescript ions designed t o prevent t he effect s of agricult ural int ensificat ion. The basic HNV package 1 forbids, for exam ple, any ploughing, rolling, seeding and t he use of chem ical fert ilisers and pest icides during t he 5- year cont ract per iod, as well as delaying m owing unt il aft er t he 1st of July and rest rict ing st ocking rat es t o below 1 Livest ock Unit ha- 1 ( NRDP, 2007) . The effect iveness of t hese m easures for all areas under t he schem e is, however, not guarant eed. Monit oring of t he quant it y and qualit y of HNV grassland is t herefore necessary t o det erm ine whet her t he qualit y of grassland is being m aint ained, or if adj ust m ent t o t he m easures is needed. I n Rom ania, t his m onit oring is int ended t o be carried out t hrough regular surveys of t he relat ive num bers of vascular plant indicat or species present in sam ple areas of HNV grassland ( NRDP, 2007) . The prelim inary list of 22 plant indicat or species given in t he 2007– 2013 Nat ional Rural Developm ent Plan of Rom ania is incom plet e ( NRDP, 2007) . Therefore in order t o cont ribut e t he im provem ent of t his list , t his st udy used a m ult ivar iat e analyt ical approach t o ident ify appropriat e plant indicat or species based on t he st udy area of t he lowland grasslands of Sout hern Transylvania. I t was assum ed t hat t he m onit oring procedure will t ake t he form of a line t ransect or sim ilar, w hich w ill provide a rapid, r epresent at ive overview of t he parcel. Thus, highly visible and rapidly ident ifiable vascular plant t axa wit h a preference for low- int ensit y HNV grassland were sought . M ATERI AL AN D M ETH OD S Bet ween May and July 2009, t he vascular flora of lowland m esic ( Molinio- Arrhenat herat ea R. Tx. 1937, and Fest uco- Brom et ea Br.- Bl. et Tüxen ex Soó 1947) grassland around t he villages of Saschiz, Bunest i and Viscri in Sout hern Transylvania was surveyed. Three m anagem ent t ypes ( m ixed past ure, sheep past ure and hay m eadow) and t hree int ensit ies ( ‘int ensive’, wit h overgrazing or previous use of input s; ‘ext ensive’, w it h no input s and low st ocking rat es; and ‘abandoned’ or very undergrazed) were dist inguished, and t hree repeat s were sam pled for each com binat ion, giving 27 separat e grassland parcels. For each parcel, five 100 m 2 plot s were sam pled wit h five random ly placed 0.25m 2 relevés in each, recording each species and it s abundance/ coverage ( Braun- Blanquet ) . I n addit ion, a t ransect was walked in each parcel, recording all species wit hin 2 m eit her side. Const rained m ult ivariat e analysis ( Canonical Correspondence Analysis: Fig. 2) was used t o ident ify t hose species which had t he st rongest relat ionship wit h ext ensive grassland ( i.e. t hose sit uat ed closest t o t he ‘ext ensive’ variable on t he ordinat ion plot ) . These species wer e screened for suit abilit y ( ease and speed of recognit ion in t he field) t o creat e a final list of indicat or species. The relat ive occurrences of t hese species at each int ensit y level were t est ed wit h t he species list s recorded from t he t ransect ( t he ant icipat ed m onit oring m et hodology) using an independent t wo- t ailed t - t est . RESULTS AN D D I SCUSSI ON Figure 2 shows t he dist ribut ion of species relat ive t o t he use int ensit y. Those select ed for inclusion in t he final list are displayed in Table 1. Figure 2: i) Sim ple ordinat ion plot of a CCA of all relevé dat a ( 135 sam ples) using focus on int erspecies dist ances and biplot scaling ( excluding t he lower 3% of species weight and fit ranges, nam es rem oved for clar it y) . Species select ed for inclusion in t he prelim inary list are displayed as black circles, all ot her species displayed as em pt y t riangles. The variables 'I nt ensive', 'Ext ensive' and 'Abandoned' are displayed as filled t riangles. ii) Sim ple ordinat ion plot of a CCA of all t ransect dat a ( 27 sam ples) t reat ed as above, and excluding t he lower 6% of species w eight and fit ranges. 1: List of 28 indicat or species for lowland m esophilic grassland Fa m ily Spe cie s Fa m ily Spe cie s Apiaceae Pim pinella saxifraga Fabaceae Trifolium alpest re Ast eraceae Cent aurea j acea Cent aurea nigrescens Trifolium m ont anum Gent ianaceae Cent aurium ryt hraea Cent aurea pseudophrygia Lam iaceae Bet onica officinalis Leucant hem um vulgare Clinopodium vulgare Cam panulaceae Cam panula pat ula Prunella laciniat a Caryophyllaceae Diant hus arm eria Salvia prat ensis Diant hus cart husianorum St ellaria gram inea Thym us pulegioides Poaceae Ant hoxant hum odorat um Colchicaceae Colchicum aut um nale Dipsacaceae Knaut ia arvensis Pr im ulaceae Prim ula veris Euphorbiaceae Euphorbia cyparissias Rosaceae Filipendula vulgaris Fabaceae Genist a t inct oria Ononis repens Briza m edia Fragaria viridis Rubiaceae Galium verum When t est ing against t he t ransect dat a, t here was a significant ly higher num ber of indicat or species in ‘ext ensive’ t han ‘int ensive’ ( P < 0.005, t = 5.15, t wo t ailed, df = 7) or ‘abandoned’ t ransect s ( P < 0.05, t = 2.52, t wo t ailed, df = 7) , as shown in t he boxplot in Figure 3. Figure 3. Boxplot of t he num ber of indicat or species recorded in t ransect s on different use int ensit ies ( N= 9 for each int ensit y) . Different let t ers indicat e significant differences ( t - t est , 7 d.f., p< 0.05) . Regarding t he applicabilit y of t he result s, out of t he 25 grassland habit at t ypes list ed for t he Rom anian HNV grassland area in Sârbu et al. ( 2004) , t his invest igat ion includes only ‘Hill m ount ain m esophilous m eadows’ ( Cynosurion crist at i) , ‘Hill m ount ain m esophilous m anured m eadows’ ( Arrhenat herion) and ‘Hill and plat eau xero- m esophilous grasslands’ ( Cirsiobrachypodion pinnat i) . Alt hough t hese are by far t he m ost com m on grassland t ypes, in order t o use indicat or species at t he nat ional scale m odificat ions of t he list for ot her grassland t ypes based on sim ilar st udies will be necessary. I n t heir analysis of plant indicat or species for ext ensive grassland at a nat ional level, Opper m ann et al. ( 2009) dist inguished six geographical zones in Germ any, w it h separat e list s for each zone. A sim ilar approach could be considered using t he five biogeographical zones of Rom ania. A furt her out com e of t his st udy is t he confirm at ion t hat a st at ist ical approach can be used t o successfully ident ify plant indicat or species. An alt ernat ive approach t o ident ifying plant indicat or species is t hat of collect ing biot ope and species dat a for t he region, from which species are select ed or rej ect ed based on crit eria such as dist ribut ion, frequency and ecological preferences, as w ell as professional experience. This m et hod was successfully used in Germ any t o ident ify plant indicat or species for grassland biodiversit y ( Mat zdorf et al., 2008) . However, t his relies heavily on com prehensive, accurat e and up- t o- dat e bot anical and habit at inform at ion of t he kind not yet fully available for Rom ania. I f chosen w ell and recorded appropriat ely, regular m onit oring of t he relat ive num ber of indicat or species in select ed HNV grassland parcels t hroughout t he count ry should det ect changes in habit at qualit y. The success of t he HNV schem e can t hus be evaluat ed and, in t he case of a decline in qualit y, changes m ade t o t he m easures before pot ent ially irreversible losses of biodiversit y are suffered. Such a feedback syst em is vit al, as t he HNV m anagem ent prescript ions as t hey st and are unlikely t o be sufficient t o conserve t he exist ing level of biodiversit y, in t he face of t he m odernisat ion of Rom anian agricult ure. Alt hough som e t radit ional m et hods ( e.g. t he use of anim al drawn m achinery) are rewarded by a prem ium ( ‘package 2’) , such labour int ensive pract ices are likely t o disappear as m echanisat ion becom es m ore w idespread in t he count ry. As Kleij n & Sut herland ( 2003) observe, even if an AES is financially beneficial, if t here is lit t le support , feedback or evaluat ion ( as is current ly t he case in Rom ania) , t hen agri- environm ent prescript ions w ill be viewed as m erely an inconvenience, t o be carried out wit h t he least possible effort . The success of AES t hus lies as m uch in t he at t it udes of t he part icipat ing farm ers, as in t he m easures t hem selves ( Wilson & Hart , 2001) , and m ore effort needs t o be m ade across Europe t o educat e landowners and increase involvem ent in, and awareness of, achieving biodiversit y goals. Whilst t hey can be crit icised for not going far enough, m anagem ent prescript ions can also lead t o t he danger of inflexibilit y. Agri- environm ent schem es m ust involve som e addit ional act ivit y or change in pract ice on t he part of t he farm er. This conflict s w it h one of t he cent ral charact erist ics of t radit ional farm ing landscapes, i.e. t he flexible responsiveness and adapt at ion of pract ices t o t he prevailing regional environm ent al condit ions. Changes t o t his m anagem ent pose a considerable risk of dist urbing t he nat ural rhyt hm s of agricult ural wildlife, which are int im at ely bound up in t he annual farm ing cycle. One exam ple from t he Rom anian HNV farm ing guidelines is t hat m owing m ay not occur before t he 1st of July, in order t o prot ect ground nest ing birds – around t wo t o t hree weeks lat er t han it is t radit ionally carried out in t he st udy area ( E Ghilea pers. com m ., 2009) . This can im pact t he com posit ion of bot h fauna and flora, as t he t im ing of t he cut affect s which species can cont ribut e r ipe seed t o t he seed bank and which anim als are dist urbed by m owing. Whilst such t arget ed conservat ion m anagem ent , when carried out sensit ively, can benefit m any ot her species ( t he so- called ‘um brella’ effect : Andelm an & Fagan, 2000) , experience has shown t hat when it differs from t he hist orical m anagem ent , it m ay lead t o a decline in populat ions of ot her species and unbalance ecosyst em s. A dram at ic exam ple of t his is given by Konvicka et al. ( 2008) , who found t hat changed m owing pat t erns in Czech m eadows t o conform t o agri- environm ent regulat ions, direct ly caused t he ext inct ion of a large populat ion of t he endangered but t erfly species, Colias m yrm idone. AES should t hus ideally enable and encourage farm ers t o cont inue wit h t he m ost ecologically fr iendly pract ices. This is claim ed t o be t he advant age of ‘result s- based’ schem es. Her e, in cont rast t o t he usual ‘m anagem ent - based’ schem es, paym ent s are direct ly linked t o biodiversit y, leaving t he farm ers free t o m anage t heir fields as t hey w ish. This approach has been used since 1994 for a regional AES in t he Germ an st at e of Brandenburg ( Kaiser et al., 2009) and has in recent years also been int roduced t o t he st at es of Baden- Würt t em berg ( Opperm ann et al., 2002) and Lower Saxony ( Keienburg et al., 2006) . Under t hese schem es, fields are assessed using a st andardised t ransect m et hodology, recording t he num ber of plant indicat or species for ext ensive m anagem ent . Fields cont aining over a cert ain num ber of indicat or species t hen qualify for paym ent s. By linking paym ent s t o out com es, not only does it guarant ee effect iveness and efficiency of invest m ent , but also increases t he engagem ent of t he landowners, giving t hem a great er degree of aut onom y and sense of involvem ent wit h t he conservat ion process ( Mat zdorf et al., 2008) . The result s- based schem e is essent ially a sm all- scale, t arget ed ( or ‘deep and narrow’) approach t o conservat ion, requiring subst ant ial infrast ruct ure and support . I n cont rast , t he st rengt h of t he HNV schem e is t hat it encom passes large expanses of t he m ost ecologically valuable agr icult ural land. This w ill com bat t he fragm ent at ion effect s t hat sm aller- scale AES cannot prevent , allow ing t he nat ural m igrat ion of species and t he support of large, resilient populat ions. However , as a ‘broad and shallow’ schem e, resources are t hinly spread and result s are difficult t o m easure. The current lack of det ailed baseline ecological inform at ion in Rom ania ( and in m any ot her EU count ries) also m eans t hat any evaluat ion of success will necessar ily r ely heavily on expert j udgem ent , rat her t han hard figures. Thus, it is all t he m ore crit ical t hat a m onit oring schem e wit h clear goals is put in place quickly and effect ively, so t hat fut ure changes in biodiversit y can be m easured and declines in t his nat ural resource prevent ed. CON CLUSI ON S The analysis yielded 28 vascular plant indicat or species ( shown in Table 1) . The significant differences in t he num ber of species present in t he different m anagem ent int ensit ies ( Fig. 3) suggest t hat t hese species w ill decline in response t o int ensificat ion or abandonm ent . Regular nat ional m onit oring of t he num ber of indicat or species in sam ple areas under HNV AES agreem ent ( relat ive t o cont rol areas) w ill provide a m easure of success of t he Rom anian AES, inform ing any im provem ent s t hat m ay need t o be m ade. This st udy covers t he com m on lowland m esophilic grassland t ype, however, furt her st udies of t he 26 or so ot her grassland t ypes in Rom ania will be necessary t o provide specific list s for different areas. ACKN OW LED GEM EN TS The aut hors would like t o t hank t he NGO Fundat ia ADEPT ( www .fundat ia- adept .org) for t heir generous support t hroughout t he fieldwork for t his st udy, as well as Prof. Erwin Bergm eier. REFEREN CES Andelm an, S. J., Fagan, W. F., 2000 – Um brellas and flagships: efficient conservat ion surrogat es or expensive m ist akes? Proceedings of t he Nat ional Academ y of Sciences of t he Unit ed St at es of Am erica 97: 5954–5959. Andersen, E., Baldock, D., Bennet t , H., Beaufoy, G., Bignal, E., Brouwer, F., Elbersen, B., Eiden, G., Godeschalk, F., Jones, G., McCracken, D.I ., Nieuwenhuizen, W., van Eupen, M., Hennekens, S. & Zervas, G., 2003 – Developing a High Nat ure Value indicat or. Report for t he European Environm ent Agency, Copenhagen, 76 pp. Beaufoy, G., 2008 – HNV farm ing – explaining t he concept and int erpret ing EU and nat ional policy com m it m ent s. European Forum on Nat ure Conservat ion and Past oralism . 15 pp. Bignal E., McCracken D., 1996 – Low int ensit y far m ing syst em s in t he conservat ion of t he count ryside. Journal of Applied Ecology 33: 413–424. EEA - European Environm ent Agency, 2004 – High Nat ure Value farm land - Charact erist ics, t rends and policy challenges. European Environm ent Agency, Luxem bourg: Office for Official Publicat ions of t he European Com m unit ies. 32 pp. Fuller, R.M., 1987 – The changing ext ent and conservat ion int erest of lowland grasslands in England and Wales ‐ a r eview of grassland surveys 1930‐84. Biological Conservat ion 40: 281‐300. Kaiser, T., Rohner, M.- S., Reut t er, M., Mat zdorf, B., Schaepe, A., Hoffm ann, E.. 2009 – Die Ent wicklung einer Kennart enm et hode zur Förderung von art enreichem Grünland in Brandenburg. Nat urschut z und Landschaft spflege in Brandenburg 18: 44–50 ( in Germ an) . Keienburg, T., Most , A., Pr üt er, J. ( Eds.) , 2006 – Ent w icklung und Erprobung von Met hoden für die ergebnisorient iert e Honorierung ökologischer Leist ungen im Grünland Nordwest deut schlands. NNA- Bericht e 19, 257 pp. ( in Germ an) . Kleij n, D., Sut herland, W. J., 2003 – How effect ive are European agri- environm ent schem es in conserving and prom ot ing biodiversit y? Journal of Applied Ecology 40: 947–969. Konvicka, M., Benes, J., Cizek, O., Kopecek, F., Konvicka, O., Vit az, L., 2008 – How t oo m uch care kills species: Grassland reserves, agri- environm ent al schem es and ext inct ion of Colias m yrm idone but t erfly from it s form er st ronghold. Journal of I nsect Conservat ion 12: 519–525. Mat zdorf, B., Kaiser , T., Rohner, M.- S., 2008 – Developing biodiversit y indicat or t o design efficient agri- environm ent al schem es for ext ensively used grassland. Ecological I ndicat ors 8: 256–269. NRDP - Governm ent of Rom ania, Minist ry of Agricult ure and Rural Developm ent , 2007 – Nat ional Rural Developm ent Program m e 2007–2013. 812 pp. Opperm ann, R., Br iem le, G., 2002 – Blum enwiesen in der landw irt schaft lichen Förderung. Erst e Erfahrungen m it der ergebnisorient iert en Förderung im baden- würt t em bergischen Agrar- Um welt program m MEKA I I . Nat urschut z und Landschaft splanung 34: 203–209 ( in Germ an) . Opperm ann, R., Kr ism ann, A., Sonnberger, M., Weiß, B., 2009 – Bundesweit es Biodiversit ät sm onit oring zur Grünlandveget at ion. Met hodik und erst e Erfahrungen. Nat ur Landsch. 84: 62- 70 ( in Germ an) Paracchini, M.L., Pet ersen, J.E., Hoogeveen, Y., Bam ps, C., Burfield, I ., van Sw aay, C., 2008 – High Nat ure Value Farm land in Europe - An est im at e of t he dist ribut ion pat t erns on t he basis of land cover and biodiversit y dat a. Report EUR 23480 EN. European Com m ission Joint Research Cent re, I nst it ut e for Environm ent and Sârbu A, Coldea G, Negrean G, Crist ea V, Hanganu J, Veen P., 2004 – Grasslands of Rom ania. Final report on Nat ional Grassland I nvent ory 2000–2003. Bucharest : Universit y of Bucharest & Royal Dut ch Societ y for Nat ure Conservat ion. 71 pp. Wilson, G. A., Hart , K., 2001 – Farm er part icipat ion in agri- environm ent al schem es: t owards conservat ion- orient ed t hinking? Sociologia Ruralis 41: 254–274. Mount ain hay m eadows: hot spot s of biodiversit y and t radit ional cult ure, Ed. Barbara Knowles, Societ y of Biology, London. © Pogány- havas Associat ion 2011.