Compu e s and Elec onics in Ag icul u e 221 (2024) 109007
A ailable online 10 May 2024
0168-1699/© 2024 The Au ho (s). Published by Else ie B.V. This is an open access a icle unde he CC BY-NC-ND license (h p://c ea i ecommons.o g/licenses/by-
nc-nd/4.0/).
A combined mul i- a ia e s a is ical analysis o es ablish dai y a m
ypologies in Can ab ia
Ib´
an V´
azquez-Gonz´
alez
a
,
*
, Elena Ga cía-Su´
a ez
b
, F ancisca Ruiz-Escude o
b
, Ana Isabel Ga cía-
A ias
a
a
Uni e si y o San iago de Compos ela, Escola Poli ´
ecnica Supe io de Enxe˜
na ía (Depa men o Applied Economics), Campus Uni e si a io s/n, 27002 Lugo, Spain
b
Cen e o Ag icul u al Resea ch and T aining (CIFA), Go e nmen o Can ab ia, 39600 Mu iedas-Can ab ia, Spain
ARTICLE INFO
Keywo ds:
Dai y a m ca ego isa ion
P incipal componen ac o analysis
Hie a chical clus e analysis
No he n Spain
Complex sample
ABSTRACT
In he las ew decades, dai y a ms ha e unde gone an in ense p ocess o s uc u al adjus men . Despi e his,
dai y a ming emains he mos impo an ag icul u al ac i i y in Can ab ia (no he n Spain), wi h many
di e en ypes o dai y a m exis ing. Howe e , he e a e ew s udies ha ha e cha ac e ised and es ablished
ypologies o unde s and his di e si y. This s udy aimed o de elop a me hod o cha ac e ising and ca ego ising
all he dai y a ms in Can ab ia ( a m popula ion) om a p oduc i e, economic and social poin o iew using
combined mul i- a ia e analysis echniques, including p incipal componen ac o analysis (PCFA) and hie a -
chical clus e analysis (HCA). Fo his pu pose, 86 su eys we e conduc ed on dai y ca le a ms in Can ab ia
om 2016 o 2017 using s a i ied andom sampling op imised wi h Neyman’s minimum a iance alloca ion.
The esul s, which ela e o all he dai y a ms in Can ab ia (SPSS complex sample module), ha e enabled us o
cha ac e ise and ca ego ise hese a ms. The sec o is mos ly made up o a ms wi h low-p oduc ion le els. Thei
main cha ac e is ics a e he impo ance o en ed land, he use o pas u e, good p oduc ion managemen , a
no able absence o young owne s, a s ong amily link and mode a e economic iabili y. The PCFA syn hesised
22 p oduc ion and socio-economic a iables in o 8 ac o s ha ep oduce 77.7 % o a iance, hal o hese ac o s
being economic in na u e. The HCA, which used a double decision c i e ion o de ine he op imum numbe o
clus e s, has classi ied he dai y a ms in Can ab ia in o ou p oduc ion ypologies, which di e in he le el o
ag icul u al ac i i y, p o i abili y and di e si ica ion. In addi ion, a i h g oup o a ms was iden i ied as singula
cases.
1. In oduc ion
The classi ica ion o a ms in o ypologies is a common p ac ice ha
has become widesp ead in ecen decades (Fe ei a-Golpe e al., 2021).
I enables a se ies o indi iduals o be placed in o homogeneous g oups
based on a se o a iables and c i e ia (Schwe ing e al., 2022). Se e al
classi ica ion c i e ia exis , and he mos common a e ela ed o he
p oduc ion sys em and o he socio-economic aspec s (S aï i and Lyoubi,
2003). The main ad an ages associa ed wi h he ca ego isa ion o
ag icul u al p oduc ion sys ems a e ha i e eals hei s uc u e,
unc ioning and s eng hs and enables he iden i ica ion o hose aspec s
o p oduc ion ha need o be imp o ed (Mąd y e al., 2013). Fu he -
mo e, he es ablishmen o ypologies minimises he di icul ies in un-
de s anding p oduc ion sys ems and p o ides knowledge ega ding hei
di e ences (Co ez-A iola e al., 2015; Hassall e al., 2023). I is also
use ul o de ining imp o emen s a egies in he amewo k o u u e
policy plans (G askempe e al., 2021), p omo ing sus ainable de el-
opmen (Cas el Genís e al., 2010) and ad ising on sus ainabili y
(Ande sen e al., 2007; B´
anku i e al., 2020).
Mos o he exis ing classi ica ions ha e employed mul i- a ia e
s a is ical analysis echniques, which enable a iables o di e en na-
u es o be used oge he ; Fe ei a-Golpe e al. (2021) used clus e
analysis wi h h ee p oduc i e and wo socio-economic a iables o
cha ac e ise he honey p oduc ion sec o in no hwes Spain. Di e en
echniques and me hods exis , which a y depending on he objec i es
pu sued and he na u e o he da a (Ande sen e al., 2007). Fo example,
Connell e al. (2007) used desc ip i e s a is ics o a echnical-
p oduc i e cha ac e isa ion o ca le sys ems in Anzoa egui S a e
(Venezuela). Sil a e al. (2007) adop ed equency dis ibu ion o de ec
c i ical nodes o managemen p ocesses in dual-pu pose ca le
* Co esponding au ho .
E-mail add ess: [email p o ec ed] (I. V´
azquez-Gonz´
alez).
Con en s lis s a ailable a ScienceDi ec
Compu e s and Elec onics in Ag icul u e
jou nal homepage: www.else ie .com/loca e/compag
h ps://doi.o g/10.1016/j.compag.2024.109007
Recei ed 19 Oc obe 2023; Recei ed in e ised o m 25 Ap il 2024; Accep ed 1 May 2024
Compu e s and Elec onics in Ag icul u e 221 (2024) 109007
2
p oduc ion sys ems in Zulia S a e (Venezuela). Siegmund-Schul ze and
Rischkowsky (2001) used and compa ed h ee di e en me hods (clus-
e analysis, logis ic eg ession and co espondence analysis) in he
iden i ica ion o he socio-economic cha ac e is ics o u ban sheep
keeping in wes A ica. In hese cases, he de e mina ion o he ypology
is he esul o a single analysis; howe e , he ypology can also be
de e mined in se e al s eps, o example, Pe o (1990) used a di ec ed
i e a i e me hod o agg ega e wi h he help o expe s he a ms in o
ypologies. O he s, howe e , es ablished ypologies o li es ock a m by
combining mul i- a ia e s a is ical analyses, i s a ac o analysis and
hen a clus e analysis. The la e has been used o he s udies con-
duc ed by Se ano-Ma ínez e al., (2004b), who iden i ied homoge-
neous g oups o ca le a ms in Le´
on (Spain). Co ez-A iola e al.
(2015), who cap u ed he di e si y o amily-based dai y a ms in
Michoacan (Mexico). Blanco-Penedo e al. (2019), who classi ied he
di e si y o o ganic dai y a ms in ou Eu opean coun ies. G¨
okdai e al.
(2020), who classi ied and cha ac e ised dai y goa a ms in I aly and
Tu key. Ruiz e al. (2020), who cha ac e ised ex ensi e li es ock
a ming sys ems (ca le, goa s and sheep) in a p o ec ed a ea o Spain
(Sie a Ne ada). Schwe ing e al. (2022), who s udied he ypologies and
mo i a ions o a m managemen in o ma ion sys ems in Ge many.
O he s udies, such as hose by Ma eus Sil ei a e al. (2022), who ep-
esen ed he di e si y o smallholde dai y p oduc ion sys ems in a
B azilian semi-a id egion, and Ri ei o-Vali˜
no e al. (2009), who ali-
da ed he dai y a m ypes in Galicia (Spain), wen u he and inco -
po a ed a hi d disc iminan - ype analysis.
Dai y a ming is a s a egic pa o he Spanish ag i- ood sec o owing
o i s economic and social signi icance, wi h an annual u no e o
a ound 13 billion eu os and c ea ion o mo e han 60,000 di ec jobs
(L´
opez Iglesias and Lainez And ´
es, 2022). In ecen decades, ca le a ms
ha e unde gone an in ense p ocess o s uc u al adjus men , cha ac-
e ised by a sha p educ ion in he numbe o a ms and employees,
alongside an inc ease in p oduc ion and he deg ee o in ensi ica ion
(Ri ei o-Vali˜
no e al., 2009; Gonz´
alez-Mejía e al., 2018).
Milk p oduc ion in Spain can be di ided in o wo di e en a eas
(Flo es-Cal e e e al., 2017). The no he n s ip, known as he Can a-
b ian Coas (Galicia, As u ias, Can ab ia and he Basque Coun y), is he
mos impo an in e ms o p oduc ion (accoun ing o 79 % o he dai y
a ms and 56 % o milk p oduc ion) (MAPA, 2023). Smalle p oduc ion
le els (on a e age 400 onnes pe yea ) and a die based on esh and
p ese ed o age p oduced on he a m, which includes, o a a ying
ex en , maize silage, and concen a e cha ac e ise he Can ab ian Coas
a ms (Flo es-Cal e e e al., 2017). Meanwhile, he es o he coun y
ea u es a ms wi h highe p oduc ion le els and i iga ed c opping
sys ems and, in some cases, a g ea e eliance on concen a es and o he
pu chased o age c ops.
Ou s udy ocused on Can ab ia, a small au onomous communi y
loca ed in he no h o Spain, whe e bo ines cons i u e he economic,
social and egional mains ay o he ag icul u al sec o (Calcedo, 2013;
V´
azquez-Gonz´
alez e al., 2023). Fu he mo e, i occupies a p ominen
posi ion in he dai y ca le sec o o he coun y (Ruiz-Escude o e al.,
2023). In 2022, i was he hi d mos impo an egion in Spain by
numbe o milk p oduce s (977 uni s, equi alen o 8.41 % o he o al)
and six h by numbe o cows (54,324 uni s, equi alen o 6.81 % o he
o al) and milk p oduc ion (408,130 onnes, equi alen o 5.57 % o he
o al) (MAPA, 2023).
As each ag icul u al p oduc ion sys em is di e en and aces speci ic
p oblems whose solu ions may be unique (Mąd y e al., 2013), i is
necessa y o cap u e he he e ogenei y and di e si y o a ms be o e
making any decisions (Ma eus Sil ei a e al., 2022). In Can ab ia, he e
is a wide a ie y o dai y a ms in ela ion o hei p oduc ion, socio-
economic and managemen cha ac e is ics (Ga cía-Su´
a ez, 2021).
Howe e , he e a e ew s udies ha ha e cha ac e ised hem, e en ewe
ha ha e es ablished ypologies and none ha ocused on all he a ms
in he e i o y; o ins ance, Blanco-Penedo e al. (2019) classi ied he
di e si y o o ganic dai y a ms in 14 Eu opean egions (one Can ab ia);
Calcedo (2013) analysed he e olu ion o he milk p oduc ion sec o in
Can ab ia du ing he quo a pe iod; Celo io e al. (2011) pe o med a
a m s uc u al cha ac e isa ion o he Pasiega local bo ine b eed in
Can ab ia; Dol a e al. (2018) iden i ied and cha ac e ised ou o age
managemen sys ems in 40 Can ab ia dai y a ms (g azing, ze o g azing,
conse ed o ages and maize silage); and Salcedo e al. (2022) calcu-
la ed he hyd ic oo p in o 53 dai y a ms in he no h o Spain ac-
co ding o six ood ypologies (o ganic, con en ional g azing, mange
g azing, g ass silage, maize silage, g ass and maize silage). The abo e
classi ica ions ha e he limi a ions o e e ing o an indi idual a m
(Dol a e al., 2018; Salcedo e al., 2022), a la ge geog aphic scope
(Blanco-Penedo e al., 2019; Salcedo e al., 2022) o a speci ic species
(Celo io e al., 2011) and do no use mul i- a ia e s a is ical analysis o
es ablish ypologies o he ypologies es ablished a e acco ding o eed/
o age managemen (Dol a e al., 2018; Salcedo e al., 2022).
Due o he lack o knowledge p esen ed abo e, his s udy aimed o:
i s , de elop a me hod ha cha ac e ises and ca ego ises all he dai y
a ms in a gi en e i o y om a p oduc i e, economic and social poin
o iew using combined mul i- a ia e analysis echniques; second, apply
his me hod o Can ab ia o comple e he a ailable knowledge abou i s
dai y sec o . Fo his pu pose, his s udy had h ee speci ic objec i es: o
ob ain ac o s ep esen a i e o a iables o di e en na u e ( ac o ial
cha ac e isa ion), o ob ain a ep esen a i e p oduc i e and socio-
economic ypology ( a m ypologies) and o cha ac e ise all dai y
a ms in Can ab ia and he esul ing ypologies.
We p esen esul s ha e e o all dai y a ms in Can ab ia, no jus
hose included in he su eys, using an unusual and e ec i e me hod o
ha pu pose (SPSS complex sample). In addi ion, we employed com-
bined mul i- a ia e s a is ical analysis o iden i y dai y a m ypologies
(non-exis en in his a ea), conside ing he decision on he op imal
numbe o g oups in he clus e analysis using a wo old c i e ion. The
ob ained esul s a e expec ed o expand he use o his me hodological
app oach, p o ide u he unde s anding o he di e si y o he exis ing
sys ems and help policymake s in o mula ing policies. I could also be
used o es ablish links wi h li es ock a m s a egies.
2. Ma e ials and me hods
The in o ma ion used was ob ained om a su ey o dai y ca le
a ms in Can ab ia conduc ed as ollows.
2.1. Su eys
A ep esen a i e sample (su eys) o dai y a ms in Can ab ia was
de e mined based on i e s a a o milk p oduc ion in he 2015–2016
season
1
. The milk p oduc ion s a a we e selec ed acco ding o he
popula ion size dis ibu ion (Table 1) and he usual s a a used in o he
Table 1
Sampling in o ma ion (popula ion size, numbe o su eys and ele a ion ac o ),
based on milk p oduc ion s a a ( onne) in he 2015–2016 season.
Milk p oduc ion s a a
2015–2016 ( onne)
Popula ion size
(N)
Numbe o
su eys (n)
Ele a ion ac o
(N/n)
<100 364 8 45.5
[100;250] 482 17 28.4
[250;500] 325 19 17.1
[500;1000] 164 15 10.9
≥1000 57 27 2.1
TOTAL 1392 86 16.2
Sou ce: own elabo a ion.
1
The 1392 a m da a used o sampling (mon hly milk p oduc ion) came
om he manda o y cow’s milk decla a ions, p o ided by he egional admin-
is a ion, wi h con inuous deli e ies in he pe iod om 1 Ap il 2015 o 31
Ma ch 2016.
I. V´
azquez-Gonz´
alez e al.
Compu e s and Elec onics in Ag icul u e 221 (2024) 109007
3
wo ks (Flo es-Cal e e e al., 2017; Dol a e al., 2018). The sampling
me hod employed was s a i ied andom sampling op imised using
Neyman’s minimum a iance alloca ion o a sampling e o o 5 % and
con idence le el o 95 % (V´
azquez-Gonz´
alez e al., 2023). The sample
size (n) was gi en by he Equa ion 1, which includes he ollowing pa-
ame e s: Nh he popula ion size, Sh he s anda d de ia ion, E he
maximum sampling e o (5 %), K he coe icien associa ed (1.96) wi h
con idence le el o 95 %, Y he popula ion alue o milk p oduc ion in
he 2015–2016 season (388,465 onnes), Sh
2
he a iance. The dis i-
bu ion by s a a (nh) was gi en by he Equa ion 2, ha uses any o he
abo e pa ame e s.
Sample size (n)
n=∑L
h=1(Nh x Sh)2
E2x Y2
K2+∑L
h=1Nh x Sh2(1)
Dis ibu ion by s a a (nh)
nh =n x Nh x Sh
∑L
h=1Nh x Sh (2)
n =sample size.
nh =sample size by s a um.
Nh =popula ion size in s a um h.
E =maximum sampling e o .
Y =popula ion alue o a iable i (Can ab ia milk p oduc ion in he
2015–2016 season).
K =coe icien associa ed wi h con idence le el.
Sh =s anda d de ia ion o a iable i (Can ab ia milk p oduc ion in
he 2015–2016 season) in s a um j.
Sh
2
= a iance o a iable i (Can ab ia milk p oduc ion in he
2015–2016 season) in s a um j.
L =las s a um.
The a ms in e iewed we e selec ed h ough a andom selec ion o
cases unc ion o he SPSS p og amme ( .21). This selec ion o a ms was
conduc ed sepa a ely o each o he i e milk p oduc ion s a a and in
wo s ages (ini ial and ese e a ms), andomly and wi hou
eplacemen .
The ques ionnai e, which was designed be ween Sep embe and
Oc obe 2016, consis ed o 623 i ems o in o ma ion s uc u ed in six
blocks: owne ship, p oduc ion, amily s uc u e, economic s uc u e
2
,
ecen e olu ion and p ospec s (Ga cía-Su´
a ez, 2021).
Be ween No embe 2016 and Feb ua y 2017, 86 su eys we e con-
duc ed on dai y ca le a ms in Can ab ia h ough di ec pe sonal in-
e iews las ing app oxima ely 1 h. The a me s we e p e iously
con ac ed ia elephone and e-con ac ed a e he ansc ip o missing
in o ma ion. I one ini ial a m e used o be in e iewed, i was eplaced
by a ese e belonging o he same s a a and in a nea by a ea. A ound
hal o he ini ial a ms we e eplaced by ese es as hey did no answe
he elephone o did no ag ee o he in e iew o easons o mis us ,
ea and lack o ime o in e es . The ques ionnai e was mainly answe ed
by he owne s i hey we e he s a egic decision-make s o he pe son
who made such decisions.
2.2. Da a analysis
A double mul i- a ia e analysis me hod was employed o classi y he
a ms, i s p incipal componen ac o analysis (PCFA) and second
hie a chical clus e analysis (HCA).
2.2.1. P incipal componen ac o analysis
PCFA is a s a is ical dimension educ ion me hod ha ep esen s a
wide ange o ela ionships be ween andom a iables h ough a subse
o dimensions called ac o s (Vilela-Fe ei a e al., 2021). The p incipal
componen me hod was employed o ob ain he ac o s, and eigen alues
g ea e han 1 we e selec ed as a c i e ion o he numbe o ac o s. To
acili a e he in e p e a ion o he ac o s, a o a ion was pe o med
using he Va imax me hod, whe eas he adequacy o he sample was
e alua ed using Ba le ’s sphe ici y es and he Kaise –Meye –Olkin
es (KMO). The s ages ca ied ou in he PCFA we e as ollows:
A-Selec ion o a iables
3
: a de ailed selec ion was made based on
he li e a u e consul ed and conside ing o he issues: a ailabili y,
quali y and ele ance (Kob ich e al., 2003; Ma eus Sil ei a e al., 2022)
(Table 2).
B-Co ela ion s udy
4
: highly co ela ed a iables we e elimina ed
(R
2
≥0,9) (Kob ich e al., 2003).
C-Ca ying ou successi e ac o analyses un il a alid one was
ound: he inal alid analysis mus be subs an ial (KMO >0.5 and
signi icance <0.05); u he mo e, all he a iables mus ha e high
commonali y (>0.5) and be in e p e ed in he o a ed componen ma ix
(co ela ion| |>0.5).
2.2.2. Hie a chical clus e analysis
The classi ica ion o he li es ock a ms in o ypologies has been
pe o med using mul i- a ia e HCA, he mos common ype o analysis,
wi h he ac o sco es o he alid PCFA as he a iables. Wa d’s me hod
was employed, and he dissimila i y measu es used we e he squa ed
Euclidean dis ance (Ca uso, 1997). A double decision c i e ion was
conside ed o de ine he op imal numbe o clus e s: a dend og am and
he calcula ion o he a es o a ia ion o he clus e ing coe icien s o
‘elbow ule’ (G askempe e al., 2021; Ma eus Sil ei a e al., 2022). The
o me , which is mo e commonly used, is g aphical and is based on he
in e p e a ion o he dend og am; he la e , which is analy ical, is based
on he calcula ion o he a es o change o clus e ing coe icien s be-
ween successi e s ages (Kob ich e al., 2003; P´
e ez and San in, 2007).
2.2.3. Desc ip i e s a is ics and pos -hoc es s
The esul s sec ion p o ides in o ma ion on he desc ip i e s a is ics
(mean alues, coe icien o a ia ion and con idence in e als a he 95
% le el) o a o al o 34 non-highly co ela ed p oduc ion and socio-
economic a iables. The alues e e o he o al popula ion, o which
he SPSS complex sample module has been applied. I is essen ial o use
his module o build a sampling plan (csa plan ile), wi h he speci ica-
ion o i e milk p oduc ion s a a and hei ele a ion ac o ,
5
using he
sampling assis an o complex samples. Signi ican di e ences in he
esul ing ypologies (Wald F- es ) and mul iple compa isons in he mean
alues o he g oups a e also de e mined ia pos -hoc es s
6
(5 % le el)
using he SPSS complex sample gene al linea model analysis (Zou e al.,
2020).
2
The economic da a collec ed in he su ey co espond o 2016 and he es
o he a iables o he ime a which he su ey was conduc ed. We a e awa e o
he di icul y and limi a ions in ol ed in collec ing eliable economic in o -
ma ion h ough a su ey, so he esul s ob ained a e an app oxima ion o he
economic eali y and should be in e p e ed wi h cau ion.
3
52 a iables we e selec ed (7 land base, 10 li es ock, 9 milk p oduc ion, 12
acili ies and machine y, 4 amily and wo k, 10 economic).
4
Bi a ia e Pea son’s o Spea man’s co ela ion was used, depending on he
dis ibu ion o he da a (Spea man’s i he dis ibu ion was no no mal). A o al
o 18 highly co ela ed a iables we e elimina ed.
5
Ra io o each s a um ha measu es he numbe o dai y a ms in Can a-
b ia (popula ion size) di ided by he numbe o su eys.
6
A pos hoc es was calcula ed o signi ican di e ences o ends in he
Wald F- es (p- alue <0.1).
I. V´
azquez-Gonz´
alez e al.
Compu e s and Elec onics in Ag icul u e 221 (2024) 109007
4
3. Resul s
3.1. Fac o cha ac e isa ion
The inal alid PCFA used 22 a iables (2 milk p oduc ion, 3 land
base, 5 li es ock, 4 acili ies and machine y, 2 amily and wo k, 6 eco-
nomic) and gene a ed 8 ac o s ep oducing 77.7 % o he o iginal
a iance. The analysis was sa is ac o y in s a is ical e ms as he mea-
su e o adequacy (KMO) was high (0.708), Ba le ’s Sphe ici y es was
signi ican (0.000) and he chi-squa ed alue was equal o 95.1.
Table 2
Lis o 52 a iables ini ially selec ed o he PCFA, g ouped acco ding o ca ego ies (milk p oduc ion, land base, li es ock, acili ies and machine y, amily and wo k,
economic).
CATEGORY Va iable Id. VARIABLES SIG. KS
1
Communali y Ex .
5
MILK PRODUCTION 1 Milk p oduc ion 15–16 pe a m
2
0.007
2 A e age milk p ice 15–16 (
€
/L)
3,4
0.720 0.700
3 Milk quo a pu chase since 1992 (% o quo a 15)
2
0.379
4 Milk p oduc ion 15–16 pe cow (L/cow)
2
0.728
5 Milk p oduc ion 15–16 pe UAA (L/ha UAA)
2
0.075
6 Milk p oduc ion 15–16 pe AWU (L/UTA)
2
0.423
7 Va ia ion in milk p oduc ion 16–17 (% o p oduc ion 15–16)
3
0.000
8 P oduc ion o se 14–15 (% o quo a 14–15)
3
0.030
9 Milk quo a pu chase since 1992 (kg)
3,4
0.000 0.787
LAND BASE 10 A e age plo size (ha)
3
0.000
11 UAA maize (% o UAA o al)
3,4
0.000 0.668
12 UAA g azing pas u es dai y he d (% o UAA o al)
3
0.000
13 UAA g een-cu pas u e mange dai y he d (% o UAA o al)
3
0.092
14 UAA o al (ha)
3,4
0.099 0.838
15 UAA used o pas u e (ha)
2
0.023
16 Ren ed UAA (% o UAA o al)
3,4
0.380 0.722
LIVESTOCK 17 Li es ock uni s (LU) (milk +o he ca le)
2
0.006
18 P oduc ion milk cows (% o o al cows)
3,4
0.188 0.771
19 Rea ing a e (% hei e s ≥12 mon hs o o al cows)
3
0.288
20 Bee cows (% o o al li es ock uni s)
3,4
0.000 0.699
21 S ocking a e (LU/UAA)
3,4
0.826 0.793
22 P oduc i i y pe annual wo k uni (LU/AWU)
3
0.580
23 To al dai y cows (p oduc ion +d y)
2
0.007
24 A e age longe i y be o e culling (lac a ions)
3,4
0.024 0.621
25 Cows moni o ed o milk yield (% o o al cows)
3,4
0.000 0.751
26 P imipa ous dai y cows (% o p oduc ion cows)
3
0.306
FACILITIES AND MACHINERY 27 N◦o headlocks/N◦o beds
3
0.000
28 N◦o headlocks/dai y cow
2
0.248
29 N◦o beds/dai y cow
2
0.251
30 Slu y pi capaci y pe cow (m
3
/cow)
2
0.022
31 Yea o cons uc ion mos ecen cowshed
3
0.010
32 Slu y pi capaci y (m
3
)
3,4
0.000 0.753
33 Mos powe ul ac o powe pe UAA (ho sepowe /ha)
3,4
0.087 0.812
34 To al su ace a ea o cowshed (m
2
)
2
0.000
35 Dis ance om cowshed o housing (m)
3,4
0.000 0.810
36 Maximum s o age ime slu y pi (mon hs)
2
0.001
37 Mos powe ul ac o powe (ho sepowe )
2
0.666
38 Cowshed su ace a ea pe cow (m
2
/cow)
3,4
0.075 0.857
FAMILY AND WORK 39 Numbe o amily membe s
3
0.000
40 Age o he owne (yea s)
3
0.422
41 To al AWU
3,4
0.000 0.819
42 AWU paid labou (% o o al AWU)
3,4
0.000 0.732
ECONOMIC 43 To al g oss p oduc (GP) (eu os)
2
0.003
44 Ne ma gin (eu os)
2
0.000
45 Ne ma gin pe AWU (eu os/AWU)
3,4
0.254 0.780
46 Ne ma gin pe 1000 L (eu os/1000 L)
3,4
0.191 0.868
47 Ne ma gin pe UAA (eu os/ha UAA)
2
0.155
48 To al cos (TC) (eu os)
2
0.003
49 Milk e enues (% o GP)
3,4
0.001 0.805
50 Pu chased ood (% o TC)
3,4
0.816 0.914
51 Gene al cos (% o TC)
3,4
0.080 0.771
52 Ex e nal ac o cos (% o TC)
3,4
0.030 0.824
1
S a is ical signi icance o he Kolmogo o –Smi no es (no mal dis ibu ion con as ). I he p- alue is equal o g ea e han 0.05 (no mal dis ibu ion).
2
18 a iables highly co ela ed (R2 ≥0.9), no analysed in PCFA.
3
34 a iables no highly co ela ed included in he PCFA.
4
22 a iables used in PCFA inal alid analysis.
5
Ex ac ion o communali ies o 22 a iables used in PCFA inal alid analysis.
Sou ce: own elabo a ion.
I. V´
azquez-Gonz´
alez e al.
Compu e s and Elec onics in Ag icul u e 221 (2024) 109007
5
Fu he mo e, all he a iables used exhibi ed high communali ies and
high co ela ion coe icien s in he o a ed componen ma ix (see Ta-
bles 2 and 3). The ac o s ha e been de ined acco ding o he na u e o
he highly co ela ed a iables (| |>0.5) in he o a ed componen
ma ix (Table 3).
Fac o 1, called ‘p oduc ion dimension and in ensi ica ion’, explains
28.7 % o he o al a iance. I is posi i ely co ela ed wi h he ollowing
a iables: milk p ice, slu y pi capaci y, o al annual wo k uni (AWU)
and pe cen age o paid labou , pe cen age o u ilised ag icul u al a ea
(UAA) unde maize cul i a ion, s ocking a e ( o al li es ock uni s [LU]/
UAA) and amoun o quo a pu chased since 1992. I is also nega i ely
co ela ed wi h a e age cow longe i y.
Fac o 2, called ‘economic p o i abili y’, accoun s o 11.4 % o he
o al a iance. The bes co ela ed a iables, all o hem being o an
economic na u e, a e he economic p o i abili y indica o s, such as he
ne ma gin (NM) pe AWU (NM/AWU) and he NM pe olume o milk
p oduced (NM/1000 L), bo h posi i ely co ela ed.
Fac o 3, called ‘ex ensi ica ion’, explains 9.2 % o he o al a iance.
The a iable posi i ely co ela ed wi h his ac o is he o al UAA and
he a iables nega i ely co ela ed wi h i a e he s ocking a e and he
powe in ho sepowe (hp) o he mos powe ul ac o on he a m (less
han 10 yea s old) exp essed as hec a es (hp/ha).
Fac o 4, labelled ‘high-p oduc ion-cos s uc u e’, explains 6.8 % o
he o al a iance. The a iables posi i ely co ela ed wi h his ac o a e
he pe cen age o pu chased ood o e o al cos s and he a iable
nega i ely co ela ed wi h i is he pe cen age o gene al cos s o e o al
cos s.
Fac o 5, de ined as ‘p oduc ion specialisa ion’, explains 6.3 % o he
o al a iance. The a iables posi i ely co ela ed wi h his ac o a e he
pe cen age o UAA en ed as well as he pe cen age o cows unde a milk
yield moni o ing egime and cows p oducing milk.
Fac o 6, de ined as ‘economic specialisa ion owa ds milk’, explains
5.5 % o he o al a iance. The a iable posi i ely co ela ed wi h i is
he pe cen age o e enues om milk sales compa ed wi h g oss p od-
uc
7
and he a iable nega i ely co ela ed wi h i is he pe cen age o
bee cows o e LU.
Fac o 7, de ined as ‘ex e nal ac o s’, explains 5 % o he o al
a iance. The mos impo an a iables, which a e posi i ely co ela ed
wi h i , a e he dis ance om he cowshed o he house and he pe -
cen age o ex e nal ac o cos s (land en , paid labou and in e es on
loans) o e o al cos s.
Table 3
In e p e a ion o he ac o s esul ing om inal alid dai y ca le PCFA (explained a iance, signi icance, iden i ica ion o a iables and co ela ion coe icien in he
o a ed componen s ma ix).
Fac o →
Eigen alue→
%Va iance→
(accumula ed)
Name → Meaning ac o Va iables
(Va iable Id.) → name
Co ela ion wi h
ac o
1
P oduc ion dimension and in ensi ica ion→ Highe milk p ices, dimension o he acili ies,
in es men s, labou needs, paid labou , maize cul i a ion and s ocking a e and lowe cow
longe i y
(2) →A e age milk p ize 15–16
(
€
/L)
0.525
(24) →A e age longe i y be o e
culling (lac a ions)
−0.570
(32) →Slu y pi capaci y (m
3
) 0.836
F1→6.32→ (41) →To al AWU 0.861
28.7 %→
(28.7 %) (42) →AWU paid labou (% o o al
AWU)
0.651
(11) →UAA maize (% o UAA o al) 0.736
(21) →S ocking a e (LU/UAA) 0.518
(9) →Milk quo a pu chase since
1992 (kg)
0.785
F2→2.52→ 11.4
%→
(40.2 %)
Economic p o i abili y→ Highe a m income pe annual wo k uni and pe olume o milk (45) →Ne ma gin pe AWU (eu os/
AWU)
0.800
(46) →Ne ma gin pe 1000L (eu os/
1000 L)
0.922
Ex ensi ica ion→ Highe u ilised ag icul u al a ea and lowe s ocking a e and powe pe
u ilised ag icul u al a ea
(14) →UAA o al (ha) 0.747
F3→2.02→ (21) →S ocking a e (LU/UAA) −0.641
9.2 %→ (49.3 %)
(33) →Mos powe ul ac o powe
pe UAA (ho sepowe /ha)
−0.820
F4→1.5→ 6.8 %→ High-p oduc ion-cos s uc u e→ Highe speci ic cos o pu chased ood and lowe gene al
cos
(50) →Pu chased ood (% o TC) 0.929
(56.2 %) (51) →Gene al cos (% o TC) −0.566
P oduc ion specialisa ion→ G ea e ele ance o en ed land and ca le managemen (milk
yield moni o ing egime, p oduc ion milk cows)
(16) →Ren ed UAA (% o UAA o al) 0.629
F5→1.38→ (25) →Cows moni o ed o milk
yield (% o o al cows)
0.653
6.3 %→ (62.5 %)
(18) →P oduc ion milk cows (% o
o al cows)
0.649
F6→1.22→ Economic specialisa ion owa ds milk→ Highe e enues om milk sales and lowe ele ance
o bee cows.
(49) →Milk e enues (% o GP) 0.708
5.5 %→ (68.0 %) (20) →Bee cows (% o o al
li es ock uni s)
−0.803
F7→1.09→
5.0 %→ (73.0
%)
Ex e nal ac o s→ G ea e dis ance om cowshed o house and he pe cen age o ex e nal
ac o cos s (land en , paid labou and in e es on loans)
(35) →Dis ance om cowshed o
housing (m)
0.861
(52) →Ex e nal ac o cos (% o TC) 0.650
F8→1.03→
4.7 %→ (77.7
%)
Animal wel a e→Mo e space o li es ock (38) →Cowshed su ace a ea pe
cow (m
2
/cow)
0.893
1
Only highly co ela ed a iables (| |>0.5) a e p esen ed.
Sou ce: own elabo a ion.
7
The ollowing i e e enues a e conside ed: sale o milk, subsidies, ca le,
o he li es ock and o he ag icul u al e enues (p ocessing o ag icul u al
p oduc s, e ilize s, c ops, insu ance, e c.).
I. V´
azquez-Gonz´
alez e al.
Compu e s and Elec onics in Ag icul u e 221 (2024) 109007
6
Finally, Fac o 8, de ined as ‘animal wel a e’, explains 4.7 % o he
o al a iance. The only single a iable ha is highly posi i ely co e-
la ed wi h i is cowshed a ea pe cow.
3.2. Fa m ypologies
Ou o he whole sample (86 su eys), we decided no o include
h ee a ms in he classi ica ion analysis as hey always o med a
di e en g oup in he di e en HCAs conduc ed due o he e y di e en
alues (ou lie s) o some o he 22 a iables used in he PCFA. This g oup
o h ee a ms, he ea e e e ed o as singula cases (SC), will be
conside ed in he cha ac e isa ion o he esul ing ypologies.
3.2.1. De e mina ion o he numbe o clus e s
The dend og am shows ha he op imum numbe o clus e s is 4,
d awing a e ical line a a dis ance be ween 14 and 20 poin s. This
in e al gi es he maximum dis ance o he ho izon al lines sepa a ing
he clus e s o all he clus e s wi h espec o he la e ( igh ) o ea lie
(le ) s ages (Fig. 1) (Kob ich e al., 2003).
The g aphical ep esen a ion o he analy ical solu ion, ela i e o
he clus e ing a es a he di e en s ages, sugges s, as does he dendo-
g am, ha he op imal numbe o clus e s is 4. I is no iceable how he
slope inc eases a a lowe numbe o clus e s, e.g. 3 (Fig. 2).
3.2.2. Cha ac e isa ion o dai y a ms in Can ab ia
In ela ion o he land base, he holdings ha e an a e age o 28.6 ha
o UAA, wi h ela i ely la ge plo s (2.7 ha), whe e en ing is he main
enancy egime (54.7 % o he UAA). The amoun o land dedica ed o
odde maize is low, accoun ing o less han 5 % o he UAA; con a ily,
pas u e unde g azing is he main use, wi h abou hal o he UAA
(Table 4).
The a e age he d size pe a m is 46 dai y cows, o which a high
pe cen age is in p oduc ion o being moni o ed o milk yield (82 % and
63 %, espec i ely). The a e age s ocking a e is 2.9 LU/ha, he a e age
longe i y be o e becoming cull cows is 4.2 lac a ions and he ea ing
a e is 41.1 %, which migh seem high o hese he d longe i y alues.
The di e si ica ion o p oduc ion in o o he li es ock ac i i ies, such as
bee cows, is no widesp ead, accoun ing o less han 5 % o he o al
LU.
In e ms o milk p oduc ion, he es ablishmen o milk quo as in
Spain in Ap il 1992 and hei subsequen aboli ion in Ma ch 2015 ha e
condi ioned he e olu ion o dai y a ms. Since 1992, a ms bough an
a e age quo a o 153,000 kg o milk. Fu he mo e, du ing he pe iod
immedia ely be o e (2014–2015 season) and a e (2016–2017 season)
he elimina ion o quo as, an opposi e p oduc i e beha iou was
de ec ed. In he i s pe iod, milk p oduc ion was almos 10 % lowe
han he a ailable quo a, whe eas in he second pe iod, i sligh ly
Fig. 1. Dendog am.
Fig. 2. G aphical ep esen a ion o he a ia ion a es o he clus e ing co-
e icien s (elbow ule).
Sou ce: own elabo a ion
I. V´
azquez-Gonz´
alez e al.
Compu e s and Elec onics in Ag icul u e 221 (2024) 109007
7
inc eased.
In ela ion o he acili ies and machine y a ailable, i should be
no ed ha on a e age, he mos ecen cowsheds a e amo ised, as hey
a e mo e han 30 yea s old. These cowsheds a e loca ed a an a e age
dis ance om he housing o 784 m and ha e o he cha ac e is ics such
as an a e age s all a ea pe cow o 12.7 m
2
and 1.03 headlocks pe bed.
The mechanisa ion index, measu ed as he a io o he powe (hp) o he
mos powe ul ac o on he a m (less han 10 yea s old) pe hec a e o
UAA, is 4.6 hp/ha.
Rega ding he cha ac e is ics o he owne , amily and wo k, almos
Table 4
P oduc ion and socio-economic cha ac e isa ion o he esul ing clus e s. Resul s ele a ed o he popula ion as a whole, aking as a e e ence he 34 non-highly
co ela ed a iables used in he PCFA.
GROUPS Desc. S . Complex Sample
4
GLM
Va iable
Id.
VARIABLES G1 G2 G3 G4 CS To al Coe .
Va .
LCI UCI F
Wald
Sig
3
.
N◦Su eys 19 42 15 7 3 86
Popula ion (n◦ a ms)
5
67 559 488 222 56 1392
MILK PRODUCTION
7 Va ia ion in milk p oduc ion 16–17 (% o
p oduc ion 15–16)
1.88a
2
0.70b 1.48ab 2.58a 0.00c 1.3 0.432 0.18 2.42 2.87 0.028
2 A e age milk p ice 15–16 (
€
/1000L)
1
313a 295b 270c 259c 264c 280 0.012 279 287 10,8 0.000
8 P oduc ion o se 14–15 (% o quo a 14–15) 15.7a 0.5ab −19.4bc −15.5b −31.6c −9.6 −0.304 −15 −3.8 4.42 0.030
9 Milk quo a pu chase since 1992 (1000 kg)
1
605a 213b 53.4c 34.1c 352ba 153 0.085 127 179 13.8 0.000
LAND BASE
14 UAA o al (ha)
1
55.0a 35.8b 20.7c 19.7c 30.6bc 28.6 0.092 23.4 33.9 9.21 0.000
16 Ren ed UAA (% o UAA o al)
1
68.8b 61.6b 49.4bc 37.5c 83.1a 54.7 0.086 45.3 64 5.84 0.000
11 UAA maize (% o UAA o al)
1
35.4a 3.3b 4.0b 0.0c 10.3ab 4.8 0.208 2.8 6.8 7.89 0.000
12 UAA g azing pas u es dai y he d (% o UAA
o al)
26.1b 34.3b 63.6a 59.4ab 56.2ab 49.1 0.092 40.1 58.1 3.04 0.022
13 UAA g een-cu pas u e o mange dai y he d
(% o UAA o al)
9.1 27.4 28.6 28.7 45.5 27.9 0.183 17.7 38.1 1.37 0.251
10 A e age plo size (ha) 3.0ab 3.2a 2.7ab 1.8b 1.1b 2.7 0.058 1.9 3.5 2.19 0.077
LIVESTOCK
18 P oduc ion milk cows (% o o al cows)
1
86.6a 85.4a 76.7b 84.4a 86.7a 82.3 0.012 80.3 84.3 4.34 0.003
26 P imipa ous dai y cows (% o p oduc ion cows) 36.9 31.4 31.1 39.8 39.6 33.2 0.073 28.4 38.1 1.0 0.410
25 Cows moni o ed o milk yield (% o o al
cows)
1
100a 100a 24.4b 33.3b 100a 62.9 0.099 50.5 75.2 26.7 0.000
19 Rea ing a e (% hei e s ≥12 mon hs o o al
cows)
36.1 35.2 38.8 61.9 42.7 41.1 0.079 34.6 47.5 1.96 0.107
24 A e age longe i y be o e culling (lac a ions)
1
3.1b 4.3ab 4.8a 2.9b 4.6a 4.2 0.048 3.8 4.6 12.54 0.000
21 S ocking a e (LU/UAA)
1
3.6a 2.8b 2.6bc 3.6a 2.1c 2.9 0.056 2.6 3.2 5.25 0.001
20 Bee cows (% o o al li es ock uni s)
1
0.0b 2.2ab 1.9ab 16.8ab 6.8ab 4.5 0.365 1.2 7.7 3.24 0.016
22 P oduc i i y pe annual wo k uni (LU/AWU) 56.8a 35.3b 28.8b 45.4ab 36.9b 35.7 0.069 30.8 40.6 8.15 0.000
FACILITIES AND MACHINERY
31 Yea o cons uc ion he mos ecen cowshed
(o e 1900)
99a 95a 74b 71b 90ab 84 0.002 77 91 3.99 0.005
35 Dis ance om cowshed o housing (1000 m)
1
0.97a 0.21b 0.092b 0.96ab 11.5a 0.784 0.459 0.07 1.49 2.43 0.055
32 Slu y pi capaci y (100 m
3
)
1
12.4a 3.69b 2.72bc 1.57c 4.71b 3.47 0.084 2.89 4.06 9.09 0.000
33 Mos powe ul ac o powe pe UAA (hp/ha)
1
2.9b 4.3ab 5.3a 4.2ab 5.1ab 4.6 0.093 3.75 5.46 3.18 0.018
38 Cowshed su ace a ea pe cow (m
2
/cow)
1
13.4a 14.1a 11.9a 7.5b 25.3a 12.7 0.086 10.5 14.9 7.6 0.000
27 N◦o headlocks/N◦o beds 1.00 1.06 1.03 1.00 0.98 1.03 0.024 0.98 1.08 1.21 0.314
FAMILY AND WORK
39 Numbe o amily membe s 3.7 3.6 3.5 3.0 3.5 3.5 0.053 3.1 3.83 0.39 0.816
40 Age o he owne (yea s) 49.8 52.6 51.8 51.0 46.6 51.7 0.032 48.4 55.0 1.49 0.213
41 To al AWU
1
3.6a 2.3b 1.8bc 1.6c 1.8bc 2.1 0.046 1.86 2.23 4.61 0.002
42 AWU paid labou (% o o al AWU)
1
37.9a 7.5b 1.0c 0.0 21.6ab 6.1 0.190 3.8 8.4 9.31 0.000
ECONOMIC
45 Ne ma gin pe AWU (1000 eu os/AWU)
1
38.9a 9.6b 2.9b 12.0b 11.9b 9.1 0.197 5.55 12.7 3.94 0.006
46 Ne ma gin pe 1000 L
1
124 51 17 115 62 53 0.424 8.3 98.5 1.7 0.158
49 Milk e enues (% o GP)
1
83.5a 81.9a 71.4b 58.2c 70.8bc 74.1 0.023 70.7 77.4 8.82 0.000
50 Pu chased ood (% o TC)
1
41.0b 50.0a 38.7b 47.1ab 34.1b 44.7 0.039 41.3 48.1 5.68 0.000
51 Gene al cos (% o TC)
1
20.7c 22.3bc 35.2a 28.9ab 24.7b 27.9 0.045 25.4 30.4 5.82 0.000
52 Ex e nal ac o cos (% o TC)
1
11.9b 6.2c 6.3c 8.2bc 19.9a 7.4 0.163 5.0 9.8 3.76 0.007
1
22 a iables used in PCFA inal alid analysis.
2
Subsc ip wi h a di e en le e ep esen s he s a is ical signi icance be ween he g oups (p- alue <0.05).
3
Signi icance le el in black (p- alue <0.05).
4
Desc ip i e s a is ics: coe icien o a ia ion (s anda d de ia ion/mean alue) and con idence in e als a he 95 % le el (lowe LCI and uppe UCI).
5
This a iable ep esen s he numbe o all dai y a ms in he e i o y (Can ab ia) ha belongs o each g oup.
Sou ce: own elabo a ion.
I. V´
azquez-Gonz´
alez e al.
Compu e s and Elec onics in Ag icul u e 221 (2024) 109007
8
all he dai y a ms in Can ab ia a e amily a ms (98.6 %), mainly wi h
he legal s a us o he owne as a na u al pe son (58.9 %), ollowed by
single- amily companies (21.7 %), whe e all he pa ne s a e amily
membe s li ing in he same household and mul i- amily companies (18
%) (Ga cía-Su´
a ez, 2021). The a e age numbe
8
o amily membe s is
3.5, wi h he a e age age o he owne being 51.7 yea s. The a e age
numbe o AWU is 2.1 AWU, o which 6.1 % is paid labou .
In ela ion o he economic in o ma ion o 2016, he a e age NM pe
AWU is 9,153
€
/AWU, and he NM pe uni o milk p oduced is 0.053
€
/L. In e ms o e enues, milk is by a he main income sou ce (74.1
%), and in e ms o expendi u e, pu chased eed is he main cos (44.7
%).
Finally, he p oduc ion s uc u e o he sec o is mos ly made up o
a ms wi h low p oduc ion, wi h 60.7 % o dai y a ms p oducing less
han 250 onnes (Table 5).
3.2.3. Cha ac e isa ion o a m ypology
O he 34 a iables analysed in Table 4, 9 do no subs an ially di e
in mean alues be ween ypologies (p- alue >0.05); in hese cases, he
mean alue o mos o he g oups is wi hin he con idence in e als (LCI
and UCI). The a iables, g ouped in ca ego ies, a e as ollows: land base
(pe cen age UAA de o ed o g een-cu pas u e o in-mange eeding,
a e age size o he plo s), li es ock (pe cen age o p imipa ous dai y
cows, ea ing a e), acili ies and machine y (dis ance om he cowshed
o he house, headlocks/bed a io), amily and wo k (numbe o amily
membe s, age o he owne ) and economic (MN/1000 L).
In he HCA, a o al o ou g oups ha e been ob ained, which a e
desc ibed below, oge he wi h he g oup o SC.
G oup 1–‘Fa ms wi h high p oduc ion and p o i abili y’. This
g oup is made up o 19 o he a ms su eyed, equi alen o 4.8 % o he
dai y a ms in Can ab ia; all o hem belong o he s a a wi h he highes
milk p oduc ion (500 onnes o mo e).
Thei land base is cha ac e ised by a highe UAA, a highe le el o
o age maize cul i a ion and en ed land. Thei li es ock is cha ac-
e ised by a la ge he d size, inc eased s ocking a e, highe milk p o-
duc i i y pe head, mo e con olled managemen (pe cen age o cows in
p oduc ion and unde milk yield moni o ing) and specialisa ion in milk
p oduc ion (absence o bee cows); howe e , he a e age longe i y
be o e culling is among he lowes . In e ms o p oduc ion, hese a ms,
which ha e he highes milk p oduc ion olume, ecei ed a highe p ice
pe li e o milk, ha e bough a la ge amoun o he milk quo a and, in
he las milking season (2014–2015) be o e he milk quo a was abol-
ished (Ma ch 2015), egis e ed a conside able excess o p oduc ion,
which di e s om he o e all si ua ion o all he a ms in Can ab ia.
In ela ion o he acili ies and machine y, hei cowsheds we e
cons uc ed mo e ecen ly and a e loca ed a a g ea e dis ance om he
house (signi ican end); hey also ha e a la ge su ace a ea pe cow
and a g ea e slu y pi capaci y. The powe a io o he mos powe ul
ac o pe hec a e o UAA has he lowes alues, which may be due o
he g ea e UAA a ailable. Rega ding labou , his g oup has highe la-
bou equi emen s and a g ea e pe cen age o paid wo k (six imes
highe han he a e age o he sec o ). On he economic side, G oup 1
has a g ea e economic p o i abili y pe wo ke (MN/AWU) as i com-
bines highe milk p oduc i i y pe AWU and uni a y NM (NM/1000 L).
Fu he mo e, a s ong economic specialisa ion exis s owa ds dai y, wi h
his g oup ha ing he highes pe cen age e enues om he sale o milk;
howe e , gene al cos s a e lowe due o he highe p opo ion o speci ic
cos s.
G oup 2–‘Fa ms wi h medium p oduc ion and p o i abili y’. This
is he la ges g oup, comp ising 42 o he a ms su eyed and ep e-
sen ing 40.2 % o he dai y a ms in Can ab ia. G oup 2 is made up o
a ms wi h lowe p oduc ion, all o hem wi h a p oduc ion o a leas
100 onnes; hose wi h p oduc ion be ween 250 and 500 onnes (46 %)
a e he la ges .
The land base is cha ac e ised by an UAA ha is subs an ially smalle
han ha o G oup 1, and en ing is he main land enu e egime, as in
G oup 1. Howe e , he impo ance o o age maize is less, and he a ea
unde g een-cu pas u e o g azing is g ea e . As in G oup 1, he e is a
high pe cen age o cows in p oduc ion, all o hem being moni o ed o
milk yield; howe e , he longe i y o he he d is highe , which is
possibly associa ed wi h a lowe s ocking a e and milk p oduc i i y.
The cha ac e is ics ela ed o milk p oduc ion (pu chased quo a, milk
p ice and p oduc ion a ia ion), annual labou equi emen s and he
numbe o paid wo ke s a e also lowe han in g oup 1. The cowsheds
a e ecen ly buil and well-sized buildings, as in G oup 1, bu he slu y
pi capaci y is much lowe .
G oup 2 has an NM/AWU ha is much lowe han ha o G oup 1
and simila o he a e age o he sec o . Simila o G oup 1, i spe-
cialises in milk p oduc ion, as can be seen in he high dependence on
milk e enues and he low pe cen age o gene al cos s. Howe e , he
pe cen age o ex e nal ac o s is he lowes o all g oups and indica es
lowe dependence on ex e nal esou ces (land, labou and capi al).
G oup 3–‘Fa ms wi h low p oduc ion and p o i abili y’. G oup 3
is made up o 15 o he a ms su eyed, ep esen ing 35.1 % o he dai y
a ms in Can ab ia. This g oup is composed almos en i ely o a ms
p oducing less han 250 onnes, o which he majo i y (56 %) p oduce
less han 100 onnes.
G oup 3 is cha ac e ised by a educed su ace, mainly dedica ed o
g azing. The he d is smalle in size and p oduc i i y (LU/AWU), as is he
pe cen age o cows in p oduc ion and unde milk yield moni o ing.
Fu he mo e, he longe i y o he cows is he highes , close o 5 lac a-
ions. The a e age p ice ecei ed o he milk sold, oge he wi h he
nega i e e olu ion o p oduc ion in he las quo a campaign and he
olume o quo a pu chased, has lowe alues di e en (p- alue <0.05)
om he wo p e ious g oups. The same is ue o he age o he cow-
sheds. The powe a io o he mos powe ul ac o pe hec a e o UAA
is he highes among all he g oups due o he smalle size o he a ms.
Mo eo e , he annual labou equi emen s a e lesse , and he e is
i ually no paid labou . In he economic side, G oup 3 has he lowes
p o i abili y pe wo ke , less han
€
3,000 (MN/AWU), which shows he
economic cons ain s hey a e expe iencing and he need o supplemen
hei incomes wi h o he e enues. The pe cen age o expendi u e on
Table 5
Pe cen age dis ibu ion o dai y ca le ypologies by milk p oduc ion s a a ( onne).
Milk p oduc ion s a a 2015–2016 ( onne)
GROUPS <100 [100;250] [250;500] [500;1000] ≥1000 To al
G1. Fa ms wi h high p oduc ion and p o i abili y 0.0 0.0 0.0 49.3 50.7 100
G2. Fa ms wi h medium p oduc ion and p o i abili y 0.0 30.4 45.9 19.5 4.2 100
G3. Fa ms wi h low p oduc ion and p o i abili y 55.9 34.8 7.0 2.2 0.0 100
G4. Di e si ied a ms wi h low p oduc ion 41.1 51.2 7.7 0.0 0.0 100
G5. Singula cases 0.0 50.3 30.3 19.4 0.0 100
To al 26.1 34.6 23.3 11.8 4.1 100
Sou ce: own elabo a ion.
8
In he case o mul i- amily companies and non- amily companies, he
numbe o membe s conside ed is he numbe o pa ne s.
I. V´
azquez-Gonz´
alez e al.
Compu e s and Elec onics in Ag icul u e 221 (2024) 109007
9
pu chased eed is lowe han he a e age, indica ing he g ea e use o
eed p oduced on he a m (g ea e au onomy), al hough he pe cen age
o gene al cos s is highe .
G oup 4–‘Di e si ied a ms wi h low p oduc ion’. This g oup is
made up o 7 o he a ms su eyed, ep esen ing 15.9 % o he dai y
a ms in Can ab ia. I is composed o a ms wi h low p oduc ion; how-
e e , i has mo e a ms wi h highe p oduc ion han G oup 3, since he
majo i y (51 %) o hese a ms p oduce be ween 100 and 250 onnes.
Al hough G oup 4 is simila o G oup 3, some di e ences exis . The
size, wi h 20 ha o UAA, is small, as is he case o G oup 3; howe e he
deg ee o en ing is lowe . In e ms o li es ock, despi e ha ing a simila
li es ock size, G oup 4 has a highe pe cen age o cows in p oduc ion
and a highe s ocking a e. Fu he mo e, he p esence o bee ca le and
a high a e o ea ing seem o indica e a di e si ica ion o ac i i y. The
p ice ecei ed o he milk and he amoun o quo a pu chased a e he
lowes , exhibi ing no signi ican di e ences om hose o G oup 3. This
also happens wi h he age o he he d, capaci y o he slu y pi , annual
labou equi emen o numbe o paid wo ke s. Howe e , he a ailable
su ace a ea pe cow is subs an ially lowe . Despi e i s small p oduc ion
size, G oup 4 has a much highe NM, bo h pe li e and pe wo ke , han
G oup 3, and e en highe han he a e age o dai y a ms in Can ab ia.
I has he lowes pe cen age o e enues om milk sales among all he
g oups (58.2 %), which is compensa ed o by he money ob ained om
he di e si ica ion o ag icul u al ac i i y (subsidies 22.3 %, cal es 11.2
%, ea ing hei e s 4.2 %, cull cows 4.1 %).
G oup 5–‘SC’. This g oup is made up o h ee o he a ms su eyed,
ep esen ing 4 % o he dai y a ms in Can ab ia. The SC ha e a la ge
p oduc ion s uc u e han hose o G oups 3 and 4; hey a e a ms
belonging o he in e media e-size s a a ( om 100 o 1000 onnes),
wi h hal o hem p oducing be ween 100 and 250 onnes.
G oup 5 p esen s some unique p oduc ion da a, such as he high
deg ee o en ing (83 % o he UAA), small plo s (1.1 ha), g ea e
longe i y be o e culling (4.6 lac a ions) and lowe s ocking a e (2.1 LU/
ha). In ela ion o milk p oduc ion, hey ha e pu chased a subs an ial
p opo ion o he quo a since 1992, bu he ecen e olu ion has been
ma ked by a much lowe p oduc ion han hei po en ial; u he mo e,
he a e age p ice ecei ed o hei milk is he lowes (264
€
/1000 L).
The mos ema kable aspec in ela ion o he acili ies is he g ea e
dis ance o he cowshed om he house (11 km) and he g ea e capaci y
o he cowshed (25 m
2
/cow). On he economic side, hei p o i abili y is
close o he a e age, bu hey ha e highe ex e nal ac o cos s (20 % o
he o al), which seems o be condi ioned by he highe pe cen age o
en ed land, paid labou and in es men s made.
4. Discussion
4.1. PCFA
Quan i a i e a iables we e used in he PCFA, which is he mos
common ype o analysis (Mąd y e al., 2013; V´
azquez-Gonz´
alez e al.,
2022). The same au ho s, who we e awa e ha he selec ion o a iables
depends on he p oduc ion sys em analysed and he objec i e pu sued in
he cha ac e isa ion, ound ha he mos used echnical a iables we e
su ace a ea, size and s ocking a e, paid labou , eed supply and p o-
duc i i y; in e ms o economic a iables, he mos used a iables we e
e enues, expendi u e and ma gins (income). Kaouche-Adjlane e al.
(2015), in es ablishing a ypology o dai y a ms in Alge ia, used a i-
ables ela ed o owne ship (age, educa ion), s uc u e (land base, li e-
s ock and equipmen ), managemen ( eeding, p oduc ion and
ep oduc ion) and economy. Maseda e al. (2004), in ca ego ising amily
dai y a ms in Galicia (no hwes Spain), used 94 a iables ela ed o he
loca ion o he a m, amily s uc u e, sou ces o income, p oduc ion,
cha ac e is ics o he cowshed, cha ac e is ics o he acili ies and ansi
ou es. In ou case, he p oduc ion and socio-economic a iables used
ha we e ela ed o milk p oduc ion, land base, li es ock, acili ies and
machine y, amily and wo k and economy la gely coincided wi h hose
used in he li e a u e.
The esul s ob ained om he PCFA ega ding he numbe o ac o s
(8) and he pe cen age o a iance (77.7 %) we e consis en wi h hose
ob ained in o he s udies cha ac e ising dai y a ms wi h a iance
anging om 50 % in England and Wales (Gonz´
alez-Mejía e al., 2018)
and o he Eu opean egions (Blanco-Penedo e al., 2019) o 84.2 % in
Michoacan (Mexico) (Co ez-A iola e al., 2015). The numbe o ac o s
may a y om 3 (Kaouche-Adjlane e al., 2015; Gonz´
alez-Mejía e al.,
2018) o 16, as epo ed by Maseda e al. (2004).
The eigh ac o s ob ained ha e a dec easing impo ance in he
pe cen age o a iance explained, some hing ha also occu s in he
wo ks consul ed and is a cha ac e is ic o he analysis. In e ms o he
na u e o he ac o s, he economic aspec is o g ea es impo ance as i
is p esen in ou o he eigh ac o s ob ained (Fac o s 2, 4, 6 and 7).
Thus, Fac o 2, called economic p o i abili y, is de ined by wo e y
impo an a iables in he economic iabili y o a a m (NM/AWU and
NM/1000 L). Bach e al. (2020) a gued ha in he economic analysis o a
dai y a m, i is necessa y o pay a en ion o milk p oduc ion and o he
uni ma gin, a iables ha we e conside ed in ou s udy. The Fac o 4,
e e ed o as he high-p oduc ion-cos s uc u e, is de ined by a highe
pe cen age o pu chased eed cos s; Salinas-Ma ínez e al. (2020)
a ibu ed highe eed cos s o la ge a ms owing o hei g ea e
dependence on concen a es. Fac o 6, called economic specialisa ion,
co esponds o a ms whose income mainly depends on dai y a ming.
The Eu opean Commission applies a simila concep when de ining he
concep o echnical economic specialisa ion, when a leas 66 % o he
g oss ma gin is di ec ly associa ed wi h his ac i i y (EC, 2012). P´
e ez-
M´
endez e al. (2020), who in es iga ed how heal h and ep oduc ion
a ec he echnical e iciency o dai y a ms in As u ias (no hwes
Spain), ha e de ined as specialised dai y a ms whe e milk accoun s o
mo e han 90 % o sales e enues. Finally, Fac o 7, e e ed o as
ex e nal ac o cos , is ela ed o hose a ms ha ha e ewe o hei own
esou ces (land, labou and capi al).
The numbe o ac o s ob ained, hei in e p e a ion and he posi ion
hey occupy exhibi simila i ies o hose o o he s udies. We conside ed
ha he esul ing ac o s can be g ouped in o wo ca ego ies, basic and
complemen a y, as epo ed by Se ano-Ma ínez e al., (2004a). The
basic ac o s a e ‘dimension and in ensi ied p oduc ion’ and ‘economic
p o i abili y’, which ep oduces a highe pe cen age o he a iance and
is ela ed o size, in ensi ied p oduc ion and economic p o i abili y. The
a iables ha de ine hese ac o s, such as size, p oduc ion, manage-
men , labou and economic ac o s, a e conside ed o be necessa y when
de ining p oduc ion sys ems o cha ac e ise and classi y a ms (Co ez-
A iola e al., 2015; Gonz´
alez-Mejía e al., 2018).
The e a e nume ous complemen a y ac o s ha ep oduce a smalle
pe cen age o he o iginal a iance (Se ano-Ma ínez e al., 2004b).
These ac o s a e de ined by speci ic a iables, o example, Fac o 3,
called ex ensi ica ion, is de ined by h ee a iables ha ake in o ac-
coun su ace a ea (UAA, s ocking a e and powe /a ea a io). Ma eus
Sil ei a e al. (2022) epo ed ha su ace a ea is a key a iable when
cha ac e ising ex ensi e p oduc ion sys ems. Fac o 5, e e ed o as
p oduc ion specialisa ion, co esponds o a ms wi h mo e con olled
managemen . P´
e ez-M´
endez e al. (2020) s a ed ha all he specialised
dai y a ms in he s udy we e moni o ing milk yield. Finally, Fac o 8,
called animal wel a e, is de ined by a single a iable, cowshed a ea pe
cow, which is conside ed in some s udies o be a measu e o animal
wel a e (Ga cía-P´
e ez e al., 2022).
4.2. Iden i ica ion and cha ac e isa ion o a m ypologies
In ou s udy, be o e he classi ica ion, we decided no o conside
h ee a ms (SC) in he HCA as hey had e y di e en alues (ou lie s)
in some o he 22 a iables used in he PCFA. The same decision has been
made in he p e ious li e a u e; hus, S aï i and Lyoubi (2003) and
Co ez-A iola e al. (2015) decided no o conside hese a ms in he
classi ica ion analysis (2 and 1 a ms espec i ely) and ea ed hem as a
I. V´
azquez-Gonz´
alez e al.
