Evaluating mountain goat dairy systems for conversion to the organic model, using a multicriteria method

Animal (2012), 6:4, pp 693–703 &The Animal Conso ium 2011 doi:10.1017/S175173111100190X animal E alua ing moun ain goa dai y sys ems o con e sion o he o ganic model, using a mul ic i e ia me hod Y. Mena 1 - , J. Nahed 2 , F. A. Ruiz 3 ,J.B.Sa ´nchez-Mun˜oz 4 , J. L. Ruiz-Rojas 4 and J. M. Cas el 1 1 ETSIA, Uni e sidad de Se illa, Ca e e a de U e a km. 1, 41013 Se illa, Spain; 2 El Colegio de la F on e a Su , Ca e e a Paname icana y Pe i e ´ ico Su s/n, 29290 San C is obal de Las Casas, Chiapas, Mexico; 3 IFAPA Cen o ‘‘Camino de Pu chil’’, A ´ ea de Economı ´a y Sociologı ´a Ag a ias, Apdo. 2027, 18080 G anada, Spain; 4 Facul ad de Medicina Ve e ina ia y Zoo ecnia, Uni e sidad Au o ´noma de Chiapas, Mexico (Recei ed 2 Feb ua y 2011; Accep ed 3 Sep embe 2011; Fi s published online 13 Oc obe 2011) O ganic a ming conse es na u al esou ces, p omo es biodi e si y, gua an ees animal wel a e and ob ains heal hy p oduc s om aw ma e ials h ough na u al p ocesses. In o de o e alua e possibili ies o inc easing o ganic animal p oduc ion, his s udy p oposes a a m-scale mul ic i e ia me hod o assessing he con e sion o dai y goa sys ems o he o ganic model. In addi ion, a case s udy in he No he n Sie a o Se ille, sou he n Spain, is analysed. A consensus o expe opinions and a ield su ey a e used o alida e a lis o po en ial indica o s and issues o assessing he con e sion, which conside no only he Eu opean Communi y egula ions o o ganic li es ock a ming, bu also ag oecological p inciples. As a esul , he me hod includes 56 a iables in eg a ed in nine indica o s: Nu i ional managemen , Sus ainable pas u e managemen , Soil e ili y and con amina ion, Weed and pes con ol, Disease p e en ion, B eeds and ep oduc ion, Animal wel a e, Food sa e y and Ma ke ing and managemen . The nine indica o s a e inally in eg a ed in a global index named OLPI (O ganic Li es ock P oximi y Index). Applica ion o he me hod o a case s udy wi h 24 goa a ms e eals an OLPI alue o 46.5% o dai y goa a ms loca ed in moun ain a eas o sou he n Spain. The aspec s ha di e mos om he ag oecological model include soil managemen , animal nu i ion and p oduc ma ke ing. Resul s o he case s udy indica e ha he p oposed me hod is easy o implemen and is use ul o quan i ying he app oxima ion o con en ional a ms o an o ganic model. Keywo ds: dai y goa s, g azing sys ems, o ganic, ag oecology, indica o Implica ions In he Medi e anean Basin, o ganic p oduc ion could be a iable al e na i e o inc easing sus ainabili y o small uminan pas u e-based a ms. Indica o s a e equi ed o p o ide concise and aluable in o ma ion ega ding possi- bili ies o con e ing small uminan a ms o o ganics. The con e sion could in ol e a edesigned sys em, he e o e no only he obliga o y compliance issues should be conside ed bu also hose ela ed o ag oecological managemen . This a icle p oposes a concise me hod o imp o ing he adap a ion o dai y goa a ms o he o ganic model. In oduc ion Pas u e-based small uminan sys ems play an impo an socioeconomic and en i onmen al ole in he Medi e anean Basin; howe e , hey a e dec easing in numbe (De Rancou e al ., 2006). In gene al, ag icul u e wo ldwide has become g ea ly in ensi ied and mode nized o e he pas ew decades (LaSalle e al ., 2008). This has led o an inc ease in he use o e ilize s, syn he ic pes icides, an ibio ics, ho mones and ossil uels, and consequen ly led o an inc ease in en i on- men al p oblems (Pimen el e al ., 2005). Ag oecology, which de ines, classi ies and analyses ag a ian sys ems om ag o- nomic, ecological and socioeconomic iewpoin s (Lab ado and Sa andon, 2001), is p oposed o sol e hese en i on- men al p oblems. O ganic ag icul u e is by de ini ion he p ac ical applica ion o ag oecology. O ganic ag icul u e is based on ou p inciples (In e na ional Fede a ion o O ganic Ag icul u e Mo emen s (IFOAM), 2009): Heal h (o ganic ag icul u e should sus ain and enhance he heal h o soils, plan s, animals and humans as one and indi- isible); Ecology (o ganic ag icul u e should be based on li ing ecological sys ems and cycles, wo k wi h hem, emula e hem and help sus ain hem); Fai ness (o ganic ag icul u e should build on ela ionships ha ensu e ai ness wi h ega d o he common en i onmen and li e oppo uni ies); and Ca e (o ganic ag icul u e should be managed in a p ecau iona y - E-mail: [email protected] 693 and esponsible manne o p o ec he heal h and well-being o cu en and u u e gene a ions and he en i onmen ). O ganic li es ock is undamen ally based on g azing, in eg a ing he soil–plan –animal cycle, conse ing he en i onmen and biodi e si y and a ou ing animal wel a e. Fu he mo e, ood p oduced on o ganic a ms is ee om chemical subs ances and gene ically modi ied o ganisms, and has a ou able o ganolep ic p ope ies (Pimen el e al ., 2005). O ganic p oduc ion may also con ibu e o mi iga ing en i onmen al p oblems, and is hus accep ed by consume s and socie y in gene al (Pingali and Raney, 2005; Vaa s e al ., 2005). In addi ion, Ronchi and Na done (2003) s a e ha o ganic a ming p ac ices may help p omo e sus ainable land use and imp o e en i onmen conse a ion, animal wel a e and p oduc quali y, which is o pa icula in e es o he less a ou ed Medi e anean a eas. Small uminan a ms in he Medi e anean Basin mainly useindigenousb eeds(Cas el e al ., 2010), which a e g ea ly alued by o ganic p oduce s o hei success ul adap a ion o a high- o age die , good p oduc ion le el and longe i y. Fu he mo e, in his egion, many a ms g aze hei animals on na u al pas u es (Hadjigeo giou e al ., 2005), which a e also much app ecia ed by o ganic p oduce s o he ollowing easons: (i) lowe manpowe equi emen s (Ga cı ´a-Ma ı ´nez e al ., 2009); (ii) con ibu ion o be e milk quali y (Cuchillo e al ., 2009); (iii) be e image o he sys em as pe cei ed by he consume (Dubeu e al ., 2010); (i ) be e balance be ween ecosys ems, biodi e si y and landscape conse a ion (Be nue ´s e al ., 2011). In o de o inc ease o ganic husband y, i would be use ul o ollow a concise me hod o e alua e he deg ee o app oxima ion o con en ional a ms o he o ganic model o p oduc ion and iden i y aspec s o be imp o ed on each a m. This me hod should be b oad and mul idimensional (Munda, 2004; Bellon and Lamine, 2009), and should add ess he managemen o animals, soils and ege a ion, as well as en i onmen al, economic and social aspec s. They should be exp essed h ough indica o s so as o compa e di e en a ms in a egion o coun y and analyse he e olu ion o a g oup o a ms o e ime. P e ious s udies epo hose indica o s ha ha e been used o analyse a m sus ainabili y (Co ey e al ., 2004; Nahed e al ., 2006; Gala ´n e al . 2007; Peacock and She man, 2010); animal wel a e (Napoli ano e al ., 2009; Phy hian e al ., 2011), en i onmen al e ec s, animal wel a e and milk quali y o o ganic a ms (Mu ¨lle -Lindenlau e al ., 2010), and di e ences, in gene al, be ween o ganic and con en ional a ms (Nau a e al ., 2006; Rozzi e al ., 2007). O he s udies e alua e he echnical and economic pe o mance o o ganic a ming (Benoi and Laignel, 2009; Benoi e al ., 2009). Ne e heless, li e a u e on he con e sion o o ganic a ming is spa se (Bellon and Lamine, 2009), and ew speci ic me hods ha e been p oposed o e alua ing he possibili ies o his con e sion (Oli a es e al ., 2005). In his sense, p inciples-based indica o s should be conside ed o a oid he in oduc ion o a ming p ac ices ha unde mine he p inciples o o ganic a ming (Da nho e e al ., 2010). This a icle p oposes a a m-scale mul ic i e ia me hod o assessing he con e sion o dai y goa sys ems o he o ganic model ha conside s no only he EC egula ions o o ganic li es ock a ming bu also he ag oecological p inciples. In addi ion, applica ion possibili ies a e analysed in a case s udy in he No he n Sie a o Se ille (NSS), sou he n Spain. Me hodology Design o me hod Indica o design and selec ion o ele an issues o e alu- a ing he deg ee o app oxima ion o li es ock a ms o he o ganic model we e based on published a icles, and we e i s e alua ed by wo panels o expe s, ollowed by ield su eys. A pe son wi h solid expe ience in ag oecology, o ganic p oduc ion and/o small uminan g azing p oduc ion, was conside ed an expe . In 2007, eigh academic expe s in ag oecology and o ganic p oduc ion we e in i ed o pa icipa e in he i s panel o expe s o e alua e on- a m indica o s and issues, which had p e iously been iden i ied in a bibliog aphical e iew. The main c i e ion ollowed o he selec ion o issues was ha hey should be di ec ly ela ed wi h he Eu opean s anda ds on o ganic p oduc ion conce ning: (i) use o pe mi ed and banned subs ances and es ic ed d ugs ha help o p e en , con ol, cu e and e adica e diseases and a oid penaliza ion (G ay and Ho i, 2001; Ma a, 2001); (ii) use o ag oecological echnologies, which should no con- amina e, be dependen on capi al o deg ade he physical en i onmen , bu should pe mi he e icien use o local esou ces, he main enance o biological di e si y and he p oduc i e capaci y o he soil in he long e m (Guzma ´n and Alonso, 2001); (iii) implemen a ion o managemen and p omo ion mechanisms o o ganic li es ock a ming. Issues we e iden i ied as binomial a iables (yes 51, no 50). All we e conside ed as dummy a iables (1, 0) o homogenize he uni s o measu emen , and hus acili a e he calcula ion o each indica o . Dummy a iables could be c ea ed because he o ganic s anda ds a e based on well-de ined c i e ia o h esholds o he use o inpu s o pe mi ed and non-pe mi ed p ac ices, he e o e, mos a iables can be answe ed in his way. In he case o con- inuous a iables, such as a iable 2.3 o Table 2 conce ning S ocking a e , he o iginal da a we e s anda dized o dummy a iables assuming alue 1 when he esul was wi hin he pe mi ed o ecommended h eshold (in his case a s ocking a e below o equal o 13.3 goa o sheep/ha) and alue 0, in he opposi e case. The e o e, all a iables acqui ed an eigen alue wi h a binomial o Be noulli dis ibu ion (Za , 1984), which g ea ly simpli ied he me hod and made i easie o ob ain in o ma ion. Subsequen ly, each indica o akes i s alue om he a i hme ic mean o he alues o i s own a iable (G imm and Wozniak, 1990), mul iplied by 100: IjðkÞ¼½sum ½ iðj;kÞ=mj100 o i¼1 omðjÞ Mena, Nahed, Ruiz, Sa ´nchez-Mun˜o ´z, Ruiz-Rojas and Cas el 694 whe e j 51, 2, 3, y, 9 indica o s; i 51, 2, 3, y, mj a i- ables; mj ep esen ing he numbe o a iables i o he indica o j ; k 51, 2, 3, y, 24 a ms; i ( j , k )5 a iable i o indica o j and a m k ( alues anged om 0 o 1). In his way all he indica o s a e s anda dized o a common ela i e scale (%). Va iables we e no weigh ed o he calcula ion o he indica o s because hey eached an eigen alue wi h only wo possible and mu ually exclusi e esul s (0 o 1). The op imum o each indica o was 100%, which was ob ained i all o i s a iables elici ed a posi i e answe . The nex s age was o con i m he in e es o selec ed a iables and he lack o impo an issues, aking in o accoun he pa icula i ies o he o ganic small uminan p oduc ion in he Medi e anean Basin. Fo ha pu pose, a second panel o expe s was cons i u ed, in eg a ed by 12 expe s in o ganic and/o pas o al small uminan p oduc ion. In his case, echnicians, o ganic ce i ie s and a me s we e also included. An O ganic Li es ock P oximi y Index (OLPI) was d awn up wi h he collabo a ion o he second panel o expe s. This Index was based on he mul ic i e ia app oach o weigh ing and agg ega ing mul idimensional in o ma ion (Munda e al ., 1994; Falconi and Bu bano, 2004; Munda, 2004). The OLPI o each a m was he sum o i s weigh ed indica o alues: OLPI ðkÞ¼sum ½WCjIjðkÞ o j¼1 o9 whe e j 51, 2, 3, y, 9 indica o s; k 51, 2, 3, y, 24 a ms; WC j 5weigh ing coe icien assigned o each indica o (Table 1); Ij ( k )5 he alue o indica o j o a m k . The weigh ing coe icien o speci ic weigh assigned o each indica o (be ween 0 and 1) was de ined as a unc ion o : (i) i s impo ance acco ding o he p inciples o o ganic li es ock a ming and ag oecology and (ii) he di icul y in ul illing he equi emen s o he Eu opean s anda ds on o ganic p oduc ion (Table 1). In his sense, Nu i ional man- agemen and Ma ke ing and managemen we e he wo indica o s assigned he g ea es weigh s, bo h because o hei ag oecological impo ance and he di icul y o he dai y goa a ms in he Medi e anean Basin o each hem. In con as , Soil e ili y and con amina ion, Weed and pes con ol, B eeds and ep oduc ion and Animal wel a e ob ained a lowe sco e as, in gene al, he ini ial si ua ion o he sys ems o which he me hod is p oposed is closes o he o ganic model o p oduc ion. The global OLPI o all case-s udy a ms was he a e age o hei alues: Global OLPI ¼sum ½OLPIðkÞ=24 o k¼1 o24 whe e k 51, 2, 3, y, 24 a ms; OLPI( k ) is he OLPI o each a m. Applica ion o he me hod in a case s udy In he hi d phase, he me hod was es ed in he ield in a case s udy. Cha ac e is ics o he s udy a ea Fo y- i e pe cen o Spain’s o ganic li es ock a ms a e loca ed in Andalusia. Thi een pe cen o Andalusia’s o ganic li es ock a ms (p incipally mea -pu pose sheep and ca le) a e loca ed in he Andalusian p o ince o Se ille, whe e he e a e only 10 goa a ms, all o which aise he animals o mea . Fu he mo e, ew eed mills, slaugh e houses and cheese-making businesses exis in he p o ince o Se ille (3, 2 and 2, espec i ely; Minis y o he En i onmen and Ru al and Ma ine A ai s (MARM), 2009). The NSS is a adi ional goa - a ming egion in Andalusia, sou he n Spain. Moun ain dai y goa sys ems o his egion a e simila o hose o o he egions o he Medi e anean Basin. The NSS is loca ed be ween 388050and 378310No h la i ude, and 68300and 58230Wes longi ude. Al i ude anges om 236 o 968 m, and he a ea is cha ac e ized by sha p cli s and gen le slopes. The p edominan ecosys em is open woodland o dehesa (an ag osil opas o al sys em common o wes e n and sou hwes e n Spain), wi h a Medi e anean clima e, a e age annual ain all o 730 mm and an a e age annual empe a u e o 168C (S a is ical Ins i u e o Andalusia-Andalusian Go e nmen (SIA-AG), 2008). The a e age dai y goa he d size is 278 and he p edominan b eed is Flo ida, which yields an a e age o 346 kg o milk annually (Cas el e al ., 2011). Da a collec ion Da a we e aken om a sample o 24 con en ional goa a ms (20% o all he NSS goa a ms in he egion), selec ed wi h he assis ance o a local coope a i e called Co se illa . Th oughou he sp ing and summe o 2008, a ques- ionnai e ega ding all he a iables included in he me hod (Tables 2 o 6) was submi ed o a me s. I is no ewo hy ha in he cases whe e he a me ga e nega i e answe s o a iables 1.3 and 1.4 o Table 2, ha is, hey did no g ow c ops, a iables 2.1, 4.4, 4.5 and 5.6 o Tables 2 o 4 we e no aken in o accoun o he calcula ion o he co esponding indica o s (Sus ainable pas u e managemen and Weed and pes con ol). In o de o a oid subjec i i y in he answe s, guidelines we e d awn up indica ing which equi emen s had o be ul illed o each a iable o he answe o be a i ma i e. Al hough mos o he in o ma ion equi ed o espond o he a iables was elici ed h ough he in e iew wi h he a me , which ook place on he a m, in o he cases i was necessa y o eso o in o ma ion p o ided by b eeding associa ions, coope a i es, he milk indus y and li es ock heal h asso- cia ions ( a me s’ o ganiza ions o con ol animal heal h) o he egion. S a is ical analysis Goa a ms we e g ouped based on a K mean clus e analysis (Manly, 2004) using he weigh ed OLPI as a classi ica ion a iable. La e , he indica o s o he conglome a es we e analysed using a single ac o o one-way ANOVA. The SPSS s a is ical p og am (SPSS, 2006) was used. Con e sion o dai y goa a ms o o ganic 695 Resul s The i s panel o expe s sugges ed 31 a iables ela i e o P inciples o O ganic Ag icul u e (IFOAM, 2009), g ouped in o 10 indica o s (Table 1). The second panel o expe s p oposed 25 mo e a iables (Table 1) and g ouped indica o s Weed con ol and Pes con ol in a new a iable named Weed and pes con ol. The 56 a iables a e de ailed in Tables 2 o 6. Twen y- h ee o hem (which a e in i alics) ha e been di ec ly ex ac ed om Council Regula ion (EC) No. 834/2007 on o ganic p oduc ion and labelling o o ganic p oduc s and epealing Regula ion (ECC) No. 2092/91 (hence o h e e ed o as EC 834/2007) and om he Commission Regula ion (EC) No. 889/2008 o 5 Sep embe 2008 laying down de ailed ules o he implemen a ion o he Council Regula ion (EC) No. 834/2007 on o ganic p oduc ion and labelling o o ganic p oduc s wi h ega d o o ganic p oduc ion, labelling and con ol (hence o h e e ed o as EC 889/2008) . The emaining 33 a iables ha e been selec ed acco ding o hei impo ance o success ul ag oecological managemen , and acco ding o Bellon and Lamine (2009) and Da nho e e al .(2010)who poin ou he supe io i y o a edesigned sys em o an inpu subs i u ion sys em in he o ganic con e sion p ocess. Figu e 1 shows he mean alue o each indica o o he 24 NSS a ms: 1. Nu i ional managemen (23.6%); 2. Sus ainable pas u e managemen (42.6%); 3. Soil e ili y and con amina ion (20.0%); 4. Weed and pes con ol (70.4%); 5. Disease p e en ion (40.7%); 6. B eeds and ep oduc ion (62.5%); 7. Animal wel a e (68.9%); 8. Food sa e y (82.3%); 9. Ma ke ing and managemen (26.4%). The weigh ed mean OLPI alue was 46.5%. Table 1 Indica o s: p inciples linking, NV and WC used o calcula ing he ‘O ganic Li es ock P oximi y Index’ P inciple* Indica o NV** NV*** WC*** Heal h and ecology 1. Nu i ional managemen 3 6 0.16 Ecology and ca e 2. Sus ainable pas u e managemen 5 6 0.14 Heal h, ecology and ca e 3. Soil e ili y and con amina ion 1 5 0.06 Heal h and ca e 4. Weed con ol 1 Heal h and ca e 5. Pes con ol 1 6 0.08 Heal h and ca e 6. Disease p e en ion 4 9 0.12 Ecology and ca e 7. B eeds and ep oduc ion 2 3 0.06 Fai ness and heal h 8. Animal wel a e 5 11 0.08 Heal h 9. Food sa e y 4 4 0.14 Fai ness 10. Ma ke ing and managemen 5 6 0.16 To al 31 56 – NV 5numbe o a iables; WC 5weigh ing coe icien . *P inciples o o ganic ag icul u e (IFOAM, 2009). **Va iables included o he i s panel o expe s. ***Va iables and WC adop ed a e alida ion o second panel o expe s. Indica o s 4 and 5 we e me ged in o he indica o named ‘Weed and pes con ol’. The e o e, he p oposed me hod has only nine indica o s. Table 2 Va iables included in indica o s 1 o 2 (i ems o obliga o y compliance acco ding o EC 889/2008 a e in i alics) %* 1. Nu i ional managemen 1.1. Animals g aze daily o a leas 6 h. 58.3 1.2. A leas 50% o daily a ion ( o milked emales) and 60% ( o o he animals) is common o age and/o g ass. 33.3 1.3. The a me g ows c ops o ob ain ib e ( o pas u ing and/o odde ) o animal consump ion. 20.8 1.4. The a me cul i a es g ain o animal consump ion. 0.0 1.5. A leas 50% o eed consumed by he animals comes om he a m, en ed land o a nea by a m. 29.2 1.6. The a me does no use eed p ohibi ed by he ules (Commission Regula ion (EC) No. 889/2008). 0.0 2. Sus ainable pas u e managemen 2.1. The a me p ac ices c op o a ion. 40.0 2.2. The a me o ganizes animal g azing. 54.5 2.3. S ocking a e is less han o equal o 13.3 goa s o sheep pe ha. 86.4 2.4. S ocking a e is adequa e.** 36.4 2.5. The a me cul i a es leguminous c ops in isola ion o associa ed wi h g ains. 16.7 2.6. The a me imp o es na u al he baceous g asses. 40.9 *The alue eached o each a iable co esponds o he pe cen age o a ms wi h an a i ma i e esponse o ha a iable. **S ocking a e is adequa e i i is wi hin he limi s conside ed op imal o he ype o ecosys em and animal s udied (0.8 o 3.2 goa s/ha; Ruiz e al ., 2008) and signs o o e g azing a e no obse ed. Mena, Nahed, Ruiz, Sa ´nchez-Mun˜o ´z, Ruiz-Rojas and Cas el 696 As Figu e 1 shows, he indica o alues o he NSS a hes om he o ganic s anda ds a e: Nu i ional managemen , Soil e ili y and con amina ion and Ma ke ing and managemen . These a e ollowed by Sus ainable pas u e managemen and Disease p e en ion. Figu e 2 shows he deg ee o he e ogenei y o he sample wi h ega d o he OLPI. In global e ms, he deg ee o app oxima ion o he o ganic model o he a ms analysed does no each 50%, anging om a minimum alue o 31% o a maximum o 65% (Figu e 2), showing a wide he e o- genei y among a ms. Tables 2 o 6 show esul s o each a iable. They ep esen pe cen ages o a ms, which esponded a i ma i ely o he ques ion asked o each a iable. Table 7 shows mean pe cen ages o OLPI and o he nine indica o s included in OLPI, g ouped in o h ee conglome a es. ANOVA de ec ed signi ican di e ences in OLPI ( P ,0.001), as well as in indica o s o Nu i ional managemen ( P ,0.01) and Sus ainable pas u e managemen ( P ,0.001). The goa a ms wi h he lowes and highes pe cen ages o app ox- ima ion o he o ganic model belong o conglome a es C1 and C3, espec i ely. Discussion Rega ding he me hod p oposed In o de o de e mine he deg ee o app oxima ion o a a m o g oup o a ms o he o ganic p oduc ion model, i is necessa y o iden i y he mos impo an issues and sum- ma ize he in o ma ion ob ained in o a se o indica o s o es ablish he limi a ions, po en ial and oppo uni ies o he sys em. A mul ic i e ia index pe mi s a global quali ica ion o Table 3 Va iables included in indica o s 3 o 4 (i ems o obliga o y compliance acco ding o EC 889/2008 a e in i alics) %* 3. Soil e ili y and con amina ion 3.1. The a me uses only e ilize s allowed by Commission Regula ion (EC) No 889/2008. 95.8 3.2. The a me makes and applies compos , bu ies pos -ha es esidues and uses o he ypes o o ganic e ilize s. 0.0 3.3. The a me ca ies ou analysis o soil e ili y and con amina ion. 4.2 3.4. The e is no isk o soil o wa e con amina ion due o was e wa e . 29.1 3.5. The a me complies wi h na ional equi emen s o elimina ing manu e. 4.2 4. Weed and pes con ol 4.1. The a me only uses pes con ol p oduc s pe mi ed by Commission Regula ion (EC) No. 889/2008. 100.0 4.2. The a me only uses weed con ol p oduc s pe mi ed by Commission Regula ion (EC) No. 889/2008. 95.8 4.3. The a me does no use a m implemen s, which emo e a la ge quan i y o soil and p edispose o e osion. 33.3 4.4. The a me lea es land allow in o de o con ol weeds. 20.0 4.5. The a me ca ies ou c op o a ions in o de o con ol pes s and weeds. 40.0 4.6. The a me p ac ices in e c opping in o de o con ol pes s and weeds. 0.0 *The alue eached o each a iable co esponds o he pe cen age o a ms wi h an a i ma i e esponse o ha a iable. Table 4 Va iables included in indica o s 5 o 6 (i ems o obliga o y compliance acco ding o EC 889/2008 a e in i alics) %* 5. Disease p e en ion 5.1. The a me qua an ines animals, which a e sick o newly in oduced o he a m. 16.7 5.2. The a me ca ies ou na u al disease ea men (he balism o homeopa hy). 4.2 5.3. The a me ea s pa asi es only when necessa y and ne e mo e han wice pe yea . 100.0 5.4. The a me con ols wa e quali y. 91.7 5.5. Only p oduc s allowed by Commission Regula ion (EC) No. 889/2008 a e used o cleaning equipmen and acili ies. 33.3 5.6. Li es ock acili ies a e gene ally clean. 41.7 5.7. Hygienic-sani a y con ol o all aspec s o milking is adequa e.** 41.7 5.8. The a me does no use an ibio ics o o he con en ional e e ina y ea men s as p e en i e measu es. 0 5.9. The a me does no use accines as a p e en a i e measu e (obliga o y accines may be used). 16.7 6. B eeds and ep oduc ion 6.1. 75% o mo e o he animals a e au och honous and/o adap ed o he egion. 100.0 6.2. Animal ep oduc ion is na u al: no ho mones a e adminis e ed o synch onize hea , induce bi h, e c . 87.5 6.3. Bi hs a e dis ibu ed in o de o minimize dependence on pu chased eed.*** 66.7 *The alue eached o each a iable co esponds o he pe cen age o a ms wi h an a i ma i e esponse o ha a iable. **This includes e ision o milking equipmen a leas once e e y 6 mon hs. ***The seasons wi h he g ea es nu i ional needs should coincide wi h hose o g ea e g ass p oduc ion. Con e sion o dai y goa a ms o o ganic 697 he cha ac e is ics and unc ioning o he managemen sys em and a oids loss o in o ma ion. As epo ed by Munda (2004) a p io i , i is no possible o es ablish which is he bes mul ic i e ia me hod o a gi en empi ical p oblem, as his will depend on he objec i es and he con ex . The capaci y o each me hodology o build a mul ic i e ia index depends on how measu emen s and pe cep ions can be con e ed o a single scale o compa e elemen s o indica o s and o es ablish o de s o p io i y. The simples ways o d aw up a mul ic i e ia index wi h a speci ic scale o measu emen (such as he one used in his s udy) a e: (i) selec he a iables and indica o s; (ii) s anda dize he a iables o a oid scale and uni o measu emen e ec s (such as Z - alues o %); and (iii) weigh and agg ega e hem using mul ic i e ia echniques (Munda e al ., 1994). Table 6 Va iables included in indica o s 8 o 9 (i ems o obliga o y compliance acco ding o EC 889/2008 a e in i alics) %* 8. Food sa e y 8.1. The a m is ee o go e nmen ally con olled diseases (p incipally b ucellosis and ube culosis, al hough hese a y acco ding o species and zone). 100.0 8.2. Analyses o milk du ing he pas yea indica e less han 500 000 ge ms/ml o goa s and sheep. 86.4 8.3. Analyses o milk du ing he pas yea indica e less han 1 500 000 soma ic cells/ml o goa s and sheep. 59.1 8.4. Analyses o milk du ing he pas yea indica e an absence o bac e ial g ow h inhibi o s. 100.0 9. Ma ke ing and managemen 9.1. The a me has planned o con e o o ganic p oduc ion and has al eady aken some s eps in his di ec ion, ecei ing ad ice and/o aining by o ganic ce i ie s. 0.0 9.2. The a me has decided o con e o o ganic p oduc ion and al eady has an o ganic li es ock con e sion plan. 0.0 9.3. The a me adequa ely eco ds in o ma ion. ** 79.1 9.4. The a me has sold o sells his o he p oduc s o local indus ies o s o es. 83.3 9.5. The a me a ens his/he animals o inal sale weigh and ans o ms milk p oduced in o cheese o o he p oduc s. 0.0 9.6. The a me sells di ec ly o he inal consume . 0.0 *The alue achie ed o each a iable co esponds o he pe cen age o a ms wi h an a i ma i e esponse o ha a iable. **The a me has up- o-da e eco ds o e e ina y ea men s, eed managemen and pu chases and sales. 0 20 40 60 80 100 1. Nu i ional managemen 2. Sus ainable pas u e managemen 3. Soil e ili y and con amina ion 4. Weed and pes con ol 5. Disease p e en ion6. B eeds and ep oduc ion 7. Animal wel a e 8. Food sa e y 9. Ma ke ing and managemen Figu e 1 Mean indica o alues (pe cen age o app oxima ion o con en ional a ms o he o ganic model) o a ms in he No he n Sie a o Se ille. Table 5 Va iables included in indica o 7 (i ems o obliga o y compliance acco ding o EC 889/2008 a e in i alics) %* 7. Animal wel a e 7.1. The a me uses na u al lac a ion. 79.2 7.2. Lac a ion pe iod is a leas 45 days ( his applies only o eplacemen males and emales). 75.0 7.3. Co e ed a ea is a leas 1.35 m 2 pe adul animal. 33.3 7.4. Ou side space is a leas 2.5 m 2 pe adul animal. 79.2 7.5. Li es ock ha e pe manen access o open spaces, p e e ably o g asslands. 95.8 7.6. The a me does no sys ema ically ie up o isola e animals. 100.0 7.7. The a ea o housing o sp ing is su icien , p o ec ed om inclemen wea he and clean and well en ila ed. 75.0 7.8. Adul animals and newbo ns ha e su icien access o wa e , ood, en ila ion, ligh and adequa e empe a u e and humidi y. 83.3 7.9. The a me does no cu ho ns excep o he poin s, cas a e o ca y ou o he mu ila ion. 41.7 7.10. Animals gene ally a e in good heal h. 95.8 7.11. The a me has been ained in animal wel a e. 0.0 *The alue eached o each a iable co esponds o he pe cen age o a ms wi h an a i ma i e esponse o ha a iable. Mena, Nahed, Ruiz, Sa ´nchez-Mun˜o ´z, Ruiz-Rojas and Cas el 698 Indica o s Indica o s p oposed in his a icle ha e been designed o be included in a scheme o on- a m assessmen o con e sion o o ganics. Ne e heless, alues ob ained pe indica o may se e as a basis o s udies in which compa isons a e made be ween egions o o analyse he e olu ion o a g oup o a ms o e ime. Acco ding o Bellon and Lamine (2009), analysing con e sion and mo e gene ally ansi ion in ag icul u e as mul idimensional issues, in ol es bo h p oduc ion and social p ac ices, and en ails in e disciplina y app oaches. En i on- men , p oduc quali y (Bellon and Lamine, 2009) and animal wel a e (Napoli ano e al ., 2009) should be conside ed. The p oposed me hod includes indica o s ela i e o all issues, al hough a ming pe o mance and echnology a e he mos nume ous. Fo en i onmen al impac assessmen , se e al ypes o indica o s can be dis inguished. Bocks alle e al . (2009) poin ou h ee g oups: he i s consis s o simple indica o s based on one a iable o a simple combina ion o a iables ob ained by su ey o om da abases and a e no di ec ly measu ed; he second g oup includes indica o s based on calcula ion and in eg a e mo e han one ype o ac o , o example, a m p ac ices and soil condi ions; and he hi d g oup includes indica o s based on one and se e al mea- su emen s, o example, biodi e si y index. In con as , Napoli ano e al . (2009) conside ha he indica o s o be included in a scheme o on- a m assessmen o animal 0 10 20 30 40 50 60 70 123456789101112131415161718192021222324 Olpi alue (%) Figu e 2 No he n Sie a o Se ille a ms classi ied by inc easing o de o app oxima ion o O ganic Li es ock P oximi y Index (OLPI). Fa ms 1 o 8 belong o conglome a e 1; a ms 9 o 21 belong o conglome a e 2 and a ms 22 o 24 belong o conglome a e 3. Table 7 Mean alue and s anda d e o o nine indica o s in he OLPI o goa a ms in he No he n Sie a o Se ille (sou heas Andalusia, Spain), g ouped by conglome a es Conglome a e C1 C2 C3 Indica o * n 58 n 513 n 53 F ; P - alue 1. 6.25 c (63.04) 26.92 b (66.41) 55.55 a (65.5) 8.285; 0.002 2. 19.16 b (65.37) 51.02 a (65.36) 68.88 a (65.87) 12.273; 0.0001 3. 20.00 (60.0) 21.53 (61.53) 13.33 (66.66) nd 4. 68.75 (64.91) 67.94 (65.48) 83.33 (616.66) nd 5. 41.66 (65.45) 38.46 (63.47) 48.14 (67.40) nd 6. 62.50 (64.16) 61.53 (63.47) 66.66 (60.00) nd 7. 68.18 (65.15) 65.03 (64.09) 87.87 (63.03) nd 8. 75.00 (68.18) 88.46 (65.38) 75.00 (614.43) nd 9. 22.91 (64.38) 26.92 (62.34) 33.32 (60.01) nd OLPI 38.75 c (61.43) 48.38 b (60.69) 58.64 a (63.15) 40.729; 0.0001 OLPI 5O ganic Li es ock P oximi y Index. a, b, c Mean alues wi h di e en le e s in he same ow a e signi ican ly di e en ; nd 5non-signi ican di e ence. *1. Nu i ional managemen ; 2. Sus ainable pas u e managemen ; 3. Soil e ili y and con amina ion; 4. Weed and pes con ol; 5. Disease p e en ion; 6. B eeds and ep oduc ion; 7. Animal wel a e; 8. Food sa e y; 9. Ma ke ing and managemen . The indica o ’s alue anges om 1 o 100. The g ea e he alue, he close he app oxima ion o he o ganic model. Con e sion o dai y goa a ms o o ganic 699 wel a e should be alid (meaning ul wi h espec o animal wel a e), eliable ( e lec ing he endency o gi e he same esul s on epea ed measu emen s) and easible (conce ning ime and inancial equi emen s). When selec ing he indica o s p oposed in his a icle, p io i y was gi en o he h ee c i e ia highligh ed by Napoli ano e al . (2009): alidi y, eliabili y and easibili y. Consequen ly, he indica o s p oposed belong o he i s and second ype desc ibed by Bocks alle e al . (2009). In his sense, he indica o s Soil e ili y and con amina ion, Weed and pes con ol, Disease p e en ion, B eeds and ep oduc- ion and Ma ke ing and managemen belong o he i s g oup, and he e o e in o ma ion is ob ained by su ey. In con as , he indica o s Nu i ional managemen , Sus ain- able pas u e managemen , Animal wel a e and Food sa e y, included issues ha should be measu ed di ec ly, al hough in some cases he a me al eady had he da a. Issues O he 56 a iables, 33 a iables p oposed a e no explici ly included in EC 889/2008, bu a e alid om an ag oecological poin o iew. Issues conside ed a e impo an o a oid he con en ionaliza ion o o ganic a ming (de ined by Da nho e e al . (2010), as he in oduc ion o a ming p ac ices ha unde mine he p inciples o o ganic a ming) and o p omo e ue o ganic a ming, which may in ol e a edesign o he a m sys em ins ead o simple inpu subs i u ion (Bellon and Lamine, 2009). The combined app oach using a scien i ic li e a u e e iew, expe opinion and ield s udies, yielded a lis o ele an and easible issues o assessing he con e sion o small uminan sys ems o he o ganic model. A i s app oach o his me hod was p esen ed du ing he 9 h In e na ional Con e ence on Goa s held in Que e a o, Mexico (Mena e al ., 2009a). The lis o a iables p oposed by he i s panel o expe s was simple and sho e han he lis inally included in his a icle. The second panel o expe s ( echnical expe s om he a me s associa ions, coope a i es and ce i ie s ha had g ea e knowledge o he goa p oduc ion sys ems o he Medi e anean Basin) added new a iables in o de o p o- ide mo e p ecise in o ma ion speci ic o he Medi e anean Basin goa sys ems. They also sugges ed me ging he indi- ca o s Weed con ol and Pes con ol in o a single indica o called Weed and pes con ol, so as o a oid epe i ions in he ques ions and answe s on ag icul u al p ac ices. I he e a e di e en li es ock species on a a m, ce ain conside a ions will be necessa y. I he a me manages bo h sheep and goa s oge he , hey can be conside ed as a single species, bu i hey a e managed e y di e en ly, hen he me hod should be applied sepa a ely. I animals o he han sheep o goa s a e a med, hen he me hod should be adjus ed sligh ly and each species should be deal wi h sepa a ely. Ne e heless, al hough he species a e ea ed sepa a ely, Soil e ili y and con amina ion and Weed and pes con ol can be ob ained oge he . A e conduc ing he ield su ey, i was obse ed ha he local ex ension agen s, who pe iodically isi he a ms ound i easy o ob ain in o ma ion on mos o he issues included. Fu he mo e, his in o ma ion would gi e a p ecise idea o he issues o be imp o ed o con e sion o o ganic a ming. OLPI Fo in eg a ed analysis o all he in o ma ion ob ained, he au ho s o his s udy g ouped he 56 a iables in o nine indica o s and hen calcula ed a mul ic i e ia index called OLPI, which is a weigh ed mean o he indica o s. Expe opinion has been conside ed o design he OLPI, especially on he selec ion o he a iables, he de ini ion o he indica o s and on hei ele an impo ance, as he main goal o he me hod was o de ec he main de iciencies o a a m o g oup o a ms when con e ing o o ganics. Weigh s o indica o s o calcula ing he OLPI we e based on he impo ance con e ed o hem by he expe s and we e ans o med o a pe cen age scale, as in Rozzi e al . (2007) o build he o ganic o al me i index o o ganic dai y a me s. I he me hod is used o compa e a ms o di e en egions, OLPI should no be conside ed, as he weigh ing coe icien s mus be adjus ed in acco dance wi h speci ic local c i e ia (Gallopı ´n, 1997). In he con ex o mul ic i e ia decision making (whe e a decision-making p oblem consis s o choosing he bes al e - na i e acco ding o ce ain c i e ia, o amoun o knowledge) mos applica ions equi e c i e ia o be weigh ed acco ding o impo ance, hus implying an ex ension o usual non-weigh ed ope a o s (G abisch, 1995). An in e es ing class is he O de ed Weigh ed A e aging (OWA) ope a o s, in oduced by Yage (1991), which is a weigh ed sum wi h o de ed a gumen s. In OWA he weigh mus be anked in he o de in which he ac o s a e agg ega ed, based on uzzy measu es (Boongoen and Shen, 2009). Fuzzy measu es ha e been used as a new agg ega ion ool in he ield o subjec i e mul ic i e ia e alua ion (G abisch, 1995). The heo y o uzzy logic p o ides a ma he- ma ical means o cap u e he unce ain ies associa ed wi h human cogni i e p ocesses, such as hinking and easoning. Fu he mo e, he heo y o mul ic i e ia decision making based on uzzy measu es (i.e. Choque in eg al) has p o ed o be a be e app oxima ion o eali y, as i admi s ha sys ems ha e di e en nuances and ha decisions a e made acco ding o a wide ange o possibili ies and c i e ia ha a e all equally alid, al hough some imes hey a e con a- dic o y o con lic ing. The me i o he Choque in eg al is o b ing a powe ul ool o model in e ac ion and dependency be ween a iables (G abisch, 1995). Despi e hei immense alue, uzzy in eg als a e di icul o apply o eal si ua ions (G abisch, 1995). The main ad an age o he OLPI me hod o e uzzy logic in mul ic i e ia analysis is ha i is easy o calcula e. In la e s udies, once he me hod p oposed has been applied o mo e a ms, we will ha e a mo e p ecise idea o he ela ionship be ween di e en c i e ia and condi ioning ac o s, and he e o e uzzy logic me hodologies, such as he Choque in eg al, can be used o decision making. Mena, Nahed, Ruiz, Sa ´nchez-Mun˜o ´z, Ruiz-Rojas and Cas el 700 The case s udy Fo 31 o he 56 a iables (Tables 2 o 6), he le el o compliance is less han 50%. This indica es ha despi e he ac ha he a ms aise animals o au oc honous b eeds and mos g aze hei animals, many aspec s mus be imp o ed o a m managemen o be conside ed ag oecologically adequa e. These esul s ag ee wi h he indings o Ronchi and Na done (2003), who epo ed ha on he basis o he cu en p oduc ion sys ems in he Medi e anean a eas, only a limi ed numbe o a ms ha e he op imal equi emen s o a ansi ion om con en ional o o ganic a ms, and o sus ainable economic iabili y. In o de o illus a e he po en ial o his me hod, a iables a hes om he ag oecological op imum, and consequen ly om he o ganic s anda d, we e analysed. In addi ion, sug- ges ions a e made o imp o ing he a ms o b ing hem close o he o ganic model. Dai y goa sys ems in he s udied a ea ha e de iciencies in he o ganic equi emen s o nu i ional managemen , as sel - su iciency is low and hey ha e a sho p o ision o common odde . This ag ees wi h he obse a ions o Mena e al . (2009a) and Na done e al . (2004), who poin ou ha lack o eed au onomy is one o he main di icul ies o o e come in he ansi ion om con en ional goa a ms o o ganic. As Ruiz e al . (2009) epo , goa a me s who base animal eeding on local g asses in moun ainous a eas do no manage g azing app op ia ely. Dai y goa s g aze h oughou he whole yea , e en hough in summe and win e na u al g asses a e sca ce in he Medi e anean ecosys em. Fu he mo e, a me s belie e ha goa s ob ain enough ib e when g azing, and he e o e do no p o ide su icien oughage in he s able. This leads o a ib e-poo die when g ass is in sho supply (Ruiz e al ., 2009). The e o e, i is necessa y o inc ease p o ision o common odde , especially du ing he summe and ea ly au umn, in o de o comply wi h he EC 889/2008 minimum ib e equi emen s ( a iable 1.2 o Table 2). Acco ding o he p inciples o o ganic li es ock a ming, o ganic odde s should mainly be p oduced on he a m i sel , o by ano he local a me . I is also necessa y o g ow g ain o animal eed ( a iable 1.4. o Table 2), and use only eed pe mi ed by he EC egula ion o o ganic li es ock a ming ( a iable 1.6. o Table 2; EC 889/2008). Leguminous c ops in monocul u es o associa ed wi h g asses, and imp o emen s ( e iliza ion, eplan ing, e c.) o na u al g asses ( a iables 2.5 and 2.6 o Table 2) will help o inc ease eed sel -su iciency. Howe e , as Mena e al . (2009a) ha e epo ed, in he Medi e anean Basin, s eep slopes and poo soil quali y educe he possi- bili y o g owing animal eed on he a m. Acco ding o Ruiz e al . (2009), he mean quan i y o plan ed g asses is 0.09 ha pe goa pe a m. Ano he impo an aspec ela ed o he use o g asses is he li es ock ca ying capaci y. Al hough mos o he a ms ha e less han 13.3 goa o sheep/ha ( a iable 2.3 o Table 2), which is equi alen o 2 Li es ock Uni s/ha – he maximum s ocking a e allowed by he EC 889/2008 – only one- hi d ha e an adequa e s ocking a e ( a iable 2.4 o Table 2), conside ing he cha ac e is ics o he Medi e anean eco- sys em whe e 0.8 o 3.2 goa s/ha a e ecommended (Ruiz e al ., 2008). The e o e, li es ock ca ying capaci y and ep oduc ion planning should be adjus ed. In his sense, he seasons wi h he g ea es milk p oduc ion (when he animals ha e g ea e nu i ional needs) should coincide wi h hose o g ea e g ass p oduc ion, namely in sp ing. This would imp o e eed sel -su iciency on he o ganic a m. Success ul ag oecological a m managemen equi es adequa e soil managemen and soil quali y. Al hough he e a e ew e e ences o his in EC 889/2008, he me hod includes wo ela ed indica o s: Soil e ili y and con amina ion and Weed and pes con ol. Acco ding o he esul s o his s udy, only 4.2% o a me s should s op using chemical e ilize s i hey wish o change o o ganics. P ac ices, which imp o e hese wo indica o s acco ding o he speci ica ions o EC 889/2008 a e: imp o emen in manu e managemen (e.g. compos ing), pe iodic analyses o soil e ili y, allow pe iods, c op o a ion and in e c opping (which all help o con ol pes s wi hou eso ing o pes icides). As o he indica o Disease p e en ion, a me s should imp o e h ee a iables: na u al disease ea men ( a iable 5.2 o Table 4), which no a me does; co ec use o an i- bio ics ( a iable 5.8 o Table 4), some a me s use an ibio ics wi hou e e ina y guidance; and co ec use o accines ( a iable 5.9 o Table 4). In humid egions, pa asi ic disease con ol is di icul in o ganic li es ock (Lu e al ., 2010); howe e , pa asi ic disease con ol and p e en ion is no a p oblem in he s udied a ea, as he clima e o Andalusia (low ain all and high empe a u es) does no a ou pa asi e de elopmen . The indica o Animal wel a e has an accep able le el o compliance. Only one a iable, co e ed space o he animals ( a iable 7.3 o Table 5), was no p o ided by he a ms, in ac only 33% ha e he minimum co e ed space equi ed by EC 889/2008, which is 1.5 m 2 o adul goa s. Howe e , his is no a p oblem in Andalusia whe e he clima e is gen le, and he e o e a minimum co e ed space is no obliga o y. Finally, Ma ke ing and managemen is one o he h ee indica o s wi h he lowes le el o compliance. The low alues o he a iables included in his indica o (Table 6) a e because o he lack o speci ic ma ke s o o ganic p oduc s, which a e o en sold in Spain as con en ional p oduc s (Ga cı ´a e al ., 2008). As Mena e al . (2009b) epo , p omo ing di ec ma ke ing channels is also necessa y o a success ul con e sion o o ganics. Table 7 shows h ee g oups (clus e s) o a ms, classi ied acco ding o he OLPI alue. Signi ican di e ences a e only ound among clus e s o OLPI and o wo indica o s: Nu i ional managemen and Sus ainable pas u e manage- men . Fo all indica o s excep Soil e ili y and con amina ion, g ea e alues a e ob ained o g oup C3. This is because, in gene al, a me s do no e ilize he soil in any o he g oups. This is consis en wi h Na done e al . (2004), who poin ou ha many a ms loca ed in a id, sub-humid/humid opical and empe a e en i onmen s a e based mainly on ex ensi e use o non- e ilized na u al pas u es. Con e sion o dai y goa a ms o o ganic 701