land
A icle
Land Use Changes in Ibe ian Peninsula 1990–2012
Da id Fe nández-Noguei a * ID and Edua do Co belle-Rico * ID
Land labo a o y (Labo aTe), Depa men o Ag icul u al and Fo es Enginee ing,
Uni e sidade de San iago de Compos ela, 15782 San iago de Compos ela, Spain
*Co espondence: [email p o ec ed] (D.F.-N.); [email p o ec ed] (E.C.-R.);
Tel.: +34-680-546-041 (D.F.-N.); +34-982-823-324 (E.C.-R.)
Recei ed: 8 June 2018; Accep ed: 19 Augus 2018; Published: 20 Augus 2018
Abs ac :
This wo k aims o p o ide a comp ehensi e, wall- o-wall analysis o land use/co e
changes in he con inen al a eas o Po ugal and Spain be ween 1990 and 2012. This o e all objec i e
is de eloped in o wo main esea ch ques ions: (1) Whe he di e ences be ween he ex en and
p e alence o changes exis be ween bo h coun ies and (2) which a e he ho spo s o change (a eas
whe e a gi en land use/co e ansi ion domina es he landscape) in each coun y. We used Co ine
Land Co e in h ee di e en poin s in ime (1990, 2000, 2012) o explo e eigh cha ac e is ic land
co e ansi ions and ca ied ou a clus e analysis a LAU2 le el (municipali ies in Spain, pa ishes
in Po ugal) ha allowed o iden i y he a eas in which each ansi ion was dominan . The main
indings include he decline o ag icul u al a ea and he inc ease o u banized and a i icial co e s in
bo h coun ies, bu di e en ends ollowed by o es co e , wi h an inc ease in Spain and a dec ease
in Po ugal. A he same ime, he spa ial analysis p o ided an o e iew o he main g adien s o
change ela ed o ensions be ween ag icul u al in ensi ica ion–ex ensi ica ion, on he one hand, and
de o es a ion–a o es a ion, on he o he .
Keywo ds:
land use/co e change; Co ine Land Co e ; Ibe ian Peninsula; clus e analysis;
in ensi ica ion; a o es a ion; ex ensi ica ion; u baniza ion; a o es a ion; de o es a ion
1. In oduc ion
In he u n o he wen ie h cen u y, majo land use/co e changes ook place in Eu ope. Among
hei known d i e s we e socie al changes a ec ing moun ain communi ies, globaliza ion and majo
policy/ins i u ional changes, such as he collapse o o me socialis egimes and he e o ms o he
Eu opean Common Ag icul u al Policy [
1
–
3
]. These ans o ma ions ha e a ec ed mos o he mo e
common (in e ms o a ea) land uses/co e s, bu abandonmen o a mland is p obably he change ha
a ec ed a la ge a ea in he Eu opean landscapes [
4
]. Fo he mos pa , land use/co e changes a e
he mos isible ace o a cons an end o adjus men ollowed by land sys ems ha also encompasses
changes in land use in ensi y. Al hough he i s could be said o be mo e d as ic and e lec in mo e
impo an changes in he le els o goods and se ices p o ided by he landscape, he la e a e a guably
mo e silen bu also ele an . Ne e heless, published e idence sugges s ha land use in ensi y has
become a he s able in Eu ope [
5
] o ha , a leas , inpu and ou pu in ensi y in he a ming sec o a e
inc easingly decoupled [
6
]. Occupa ion o o me ag icul u al land by u baniza ion is also a ma e o
conce n [7].
Wi h ega d o he implica ions de i ed om he adjus men o ag icul u al land uses, hese
a e usually coupled o en i onmen al and landscape- ype isks [
8
,
9
]. In ensi ica ion o ag icul u al
p ac ices is known o be linked o he use o pes icides and e ilize s, locally inc easing he le els
o human app op ia ion o ne p ima y p oduc ion (HANPP) [
10
] and a o ing he deg ada ion o
soils ha a e inc easingly echnological [
2
,
6
]. In he a eas o sou he n Eu ope his is o en associa ed
Land 2018,7, 99; doi:10.3390/land7030099 www.mdpi.com/jou nal/land
Land 2018,7, 99 2 o 12
o o e exploi a ion o wa e esou ces and he ad ance o dese i ica ion [
11
]. On he con a y, he
dec ease in p oduc i e capaci y and he ma ginaliza ion o emo e a eas [
3
,
12
] has caused a gene al
inc ease in biomass [
13
] which o en couples o inc eased isk o wild i es [
14
,
15
]. Ag icul u al
ex ensi ica ion and e en ual abandonmen o p e iously cul i a ed a mland is also closely ela ed o
he loss o adi ional ag icul u al sys ems, s uc u es and p ac ices in he coun yside, mos ly a ec ing
ag a ian landscapes and he economic sus ainabili y o ce ain egions [4,16].
Ve y much ela ed o wha happened o ag icul u al a eas, o es co e has inc eased s eadily in
mos Eu opean Membe S a es [
17
–
19
]. Ne e heless, de o es a ion can be s ill locally impo an as
a esul o o es i es in sou he n coun ies such as, o example, Spain, F ance, Po ugal, C oa ia o
G eece [
15
,
20
,
21
]. Human p essu e has inc eased o es agili y [
22
], bu in spi e o ac i e logging and
ex eme clima ic e en s, o es biomass is on he ise in hese coun ies, also con ibu ing o he isk o
wild i es [23].
The ela i e impo ance o hese and o he land use/co e changes a ies among di e en
coun ies in he Eu opean Union. Fo example, u baniza ion p ocesses show he highes a ios in and
a ound he so-called “Eu opean pen agon” (Holland, Belgium, Ge many and F ance), wi h he Ibe ian
Peninsula, I aly and I eland in a second le el and he lowes a es o u baniza ion in Eas e n and
Sou h-Eas e n coun ies [17,24]. On he o he hand, he Bal ic coun ies, I eland, Hunga y, Spain and
Po ugal p esen high a es o ag icul u al in ensi ica ion and ex ensi ica ion, showing a much s onge
ension han o he pa s o he con inen [
2
,
6
]. A eas mos a ec ed by ma ginaliza ion o ag icul u e
a e o en linked o moun ain a eas, inland a eas o Eas Eu ope and he Medi e anean a eas [
3
,
12
,
25
].
Encompassing wo biogeog aphic egions, la ge moun ainous a eas, se e al majo me opoli an
a eas and a la ge a ie y o ag icul u al sys ems, he Ibe ian Peninsula is an excellen s udy a ea
o land use/co e change esea ch. In he las decades, a e y ele an expansion o g eenhouse
ag icul u e [
26
] has coexis ed wi h he ex ensi ica ion and e en ual abandonmen o ex ensi e a ming
sys ems [
27
–
31
]. In ensi e a o es a ion p ojec s ha e inc eased he a ea o o es plan a ions o wood
and pulp p oduc ion [
15
,
32
,
33
], bu his was simul aneous wi h spon aneous ege a ion enc oachmen
whe e ag icul u al on ie e ea ed. U baniza ion ook place a a ele an pace [
31
], bu e y o en
concen a ed in a s ip o land e y close o he coas [30,34–36].
The aim o his wo k is o p o ide a comp ehensi e, wall- o-wall analysis o land use/co e
changes in he con inen al a eas o Po ugal and Spain (583.118 km
2
) be ween 1990 and 2012. This
o e all objec i e is ela ed o wo main esea ch ques ions: (1) Whe he di e ences be ween he ex en
and impo ance o changes exis be ween bo h coun ies; and (2) which a e he ho spo s o change
in each coun y. Fo he pu poses o his wo k, we de ined ho spo s as a eas whe e a gi en land
use/co e ansi ion domina es he landscape and co e s a la ge p opo ion o o al a ea.
2. Ma e ials and Me hods
The main da a sou ces o his s udy a e he 1990, 2000 and 2012 edi ions o he Co ine Land Co e
(CLC) p ojec downloaded om he Spanish Na ional Geog aphic Ins i u e (Ins i u o Geog á ico Nacional)
and om he Po uguese Te i o ial Gene al Di ec o a e (Di eçao Ge al do Te i ó io). The ca og aphic
scale o he o iginal maps is 1:100.000 and he a ea o he minimum mapping uni is 25 hec a es.
All he subsequen analyses we e ca ied ou in as e o ma using a spa ial esolu ion o
25 m ×25 m,
GRASS GIS 7 [
37
] and he R p ojec o s a is ical compu ing [
38
]. We eclassi ied he o iginal le el
3 CLC legend (44 di e en ca ego ies) in o a simpli ied sys em wi h wo le els (L1 and L2) and 12
and 6 ca ego ies, espec i ely (Table 1). The eclassi ica ion is in ended o ocus on co e s ha a e
managed o p oduc i e pu poses. Acco dingly, di e en subca ego ies (L1) we e kep o di e en
ypes o ag icul u al and o es a eas, as changes be ween hese o en indica e changes in he in ensi y
wi h which hey a e used. We used he esul ing maps o p oduce simple ansi ion ma ixes in
o de o p oduce a i s assessmen , in e ms o o al a ea occupied, o ne changes and ansi ions
among ca ego ies.
Land 2018,7, 99 3 o 12
Table 1.
Reclassi ica ion o he o iginal Co ine Land Co e (CLC) le el 3 legend in o simpli ied L1 and
L2 land use/co e classes.
CLC Le el 3
Reclassi ica ion
L1 L2
212,213 I iga ed land
Ag icul u al a eas
211 Rain ed c ops
231 Pas u es
221,222,223,241,242 Pe manen c ops
243,244 Dehesa/Mon ado 1
321,322,323,324,333,334 Sh ubland/Na u al a eas Sh ubland/Na u al a eas
331,332,335 Unp oduc i e a eas Unp oduc i e a eas
311 Ha dwood o es
Fo es a eas
312 Coni e ous o es
313 Mixed o es
411,412,421,422,423,511,512,521,522,523 We lands and wa e bodies We lands and wa e bodies
111,112,121,122,123,124,131,132,133,141,142 A i icial/U ban a eas A i icial/U ban a eas
1
Dehesa (ES) and Mon ado (PT) a e a o m o adi ional sil opas o al sys em in which ex ensi e li es ock g azing is
combined wi h spa se some imes de ined as an “o cha d meadow”.
A e he i s assessmen , we classi ied all he obse ed changes in o a se o 8 land use/co e
ansi ions (LCT, Figu e 1), ollowing a simila app oach o Fe anec e al. (2010) [
15
]: A o es a ion
(LCT1), Con e sion o ag icul u e (LCT2), Fa mland abandonmen (LCT3), Ag icul u al ex ensi ica ion
(LCT4), De o es a ion (LCT5), Change o o es composi ion (LCT6), Ag icul u al in ensi ica ion
(LCT7), U baniza ion (LCT8). A simila app oach has also been used o he Spanish case by
Ma ínez-Fe nández e al. (2015) [
39
]. These LCT a e in ended o cap u e he main g adien s o
change ha a e p esen in he a ea o s udy and, acco dingly, some o hem ep esen opposi e ends
(e.g., LCT1 and LTC5, LCT4 and LCT7), while o he s may ake place sequen ially in ime (e.g., LCT3
o en ep esen s he las s ages o LCT4).
Land 2018, 7, 99 3 o 12
ex ensi ica ion (LCT4), De o es a ion (LCT5), Change o o es composi ion (LCT6), Ag icul u al
in ensi ica ion (LCT7), U baniza ion (LCT8). A simila app oach has also been used o he Spanish
case by Ma ínez-Fe nández e al. (2015) [39]. These LCT a e in ended o cap u e he main g adien s
o change ha a e p esen in he a ea o s udy and, acco dingly, some o hem ep esen opposi e
ends (e.g., LCT1 and LTC5, LCT4 and LCT7), while o he s may ake place sequen ially in ime (e.g.,
LCT3 o en ep esen s he las s ages o LCT4).
Table 1. Reclassi ica ion o he o iginal Co ine Land Co e (CLC) le el 3 legend in o simpli ied L1
and L2 land use/co e classes.
CLC Le el 3
Reclassi ica ion
L1
L2
212,213
I iga ed land
Ag icul u al a eas
211
Rain ed c ops
231
Pas u es
221,222,223,241,242
Pe manen c ops
243,244
Dehesa/Mon ado 1
321,322,323,324,333,334
Sh ubland/Na u al a eas
Sh ubland/Na u al a eas
331,332,335
Unp oduc i e a eas
Unp oduc i e a eas
311
Ha dwood o es
Fo es a eas
312
Coni e ous o es
313
Mixed o es
411,412,421,422,423,511,512,521,522,523
We lands and wa e bodies
We lands and wa e bodies
111,112,121,122,123,124,131,132,133,141,142
A i icial/U ban a eas
A i icial/U ban a eas
1 Dehesa (ES) and Mon ado (PT) a e a o m o adi ional sil opas o al sys em in which ex ensi e
li es ock g azing is combined wi h spa se some imes de ined as an “o cha d meadow”.
Figu e 1. A schema ic ep esen a ion o he 8 Land Co e T ansi ions (LCT) de ined in his wo k,
ollowing he app oach sugges ed by Fe anec e al. (2010) [15].
In o de o iden i y he egions wi hin he s udy a ea whe e each LCT (o combina ion o se e al
LCT) is mo e impo an in e ms o he a ea i occupies, we calcula ed he p opo ion o a ea ha each
o he eigh LCT ep esen ed in each LAU-2 uni along he whole pe iod 1990–2012. We used his
in o ma ion o pe o m a clus e analysis on he eigh LCT a iables, so we could di ide he 12,099
LAU-2 uni s in he s udy a ea in o a numbe o homogenous g oups o municipali ies/pa ishes wi h
simila pa e ns o land co e ansi ion. We used Wa d’s minimum a iance me hod (a hie a chical
clus e ing me hod) on he Euclidean dis ance among obse a ions. Each esul ing clus e was hen
Figu e 1.
A schema ic ep esen a ion o he 8 Land Co e T ansi ions (LCT) de ined in his wo k,
ollowing he app oach sugges ed by Fe anec e al. (2010) [15].
In o de o iden i y he egions wi hin he s udy a ea whe e each LCT (o combina ion o se e al
LCT) is mo e impo an in e ms o he a ea i occupies, we calcula ed he p opo ion o a ea ha
each o he eigh LCT ep esen ed in each LAU-2 uni along he whole pe iod 1990–2012. We used
Land 2018,7, 99 4 o 12
his in o ma ion o pe o m a clus e analysis on he eigh LCT a iables, so we could di ide he
12,099 LAU-2 uni s in he s udy a ea in o a numbe o homogenous g oups o municipali ies/pa ishes
wi h simila pa e ns o land co e ansi ion. We used Wa d’s minimum a iance me hod
(a hie a chical clus e ing me hod) on he Euclidean dis ance among obse a ions. Each esul ing
clus e was hen desc ibed ega ding he land co e ansi ion/s ha accoun o he la ges sha e in
e ms o a ea. We assume a clus e o municipali ies/pa ishes in which a gi en LCT is dominan as a
ho spo o ha gi en ansi ion.
3. Resul s
3.1. Ne Changes
The occupa ion da a o he en i e peninsula e i o y (Table 2; de ailed esul s o each coun y
in Tables A1 and A2) sugges s ha a ound 65.5% (381.924 km
2
) o i s o al a ea has no unde gone
any changes a L1 le el o e he 1990–2012 pe iod. I we look a each S a e sepa a ely, he Spanish
e i o y seems o ha e been subjec ed o sligh ly mo e changes (36.6% o i s a ea unde wen some
kind o change a L1 le el) han Po ugal (22.4%).
Table 2.
Ne esul s: a ea occupied by L1 and L2 classes in he Ibe ian Peninsula in 1990, 2000 and 2012.
Land Use/Co e
1990 2000 2012
km2%km2%km2%
I iga ed c ops 22,989 3.94 26,096 4.48 28,239 4.84
Rain ed c ops 113,174 19.41 108,786 18.66 107,709 18.47
Pas u es 6846 1.17 6648 1.14 9386 1.61
Pe manen c ops 87,709 15.04 88,644 15.20 77,902 13.36
Dehesas 62,118 10.65 62,232 10.67 52,319 8.97
(Ag icul u al a eas) 292,836 50.22 292,405 50.15 275,556 47.26
Ha dwood o es 48,204 8.27 49,056 8.41 60,983 10.46
Coni e ous o es 46,797 8.03 45,499 7.80 49,516 8.49
Mixed o es 20,040 3.44 20,369 3.49 18,803 3.22
(Fo es a ea) 115,041 19.73 114,924 19.71 129,301 22.17
Sh ubland 159,682 27.38 157,523 27.01 155,021 26.58
Unp oduc i e 2366 0.41 2344 0.40 2362 0.41
We lands and wa e bodies 4932 0.85 5367 0.92 5713 0.98
A i icial Su ace 8262 1.42 10,555 1.81 15172 2.60
O e all, ag icul u al a eas occupy he la ges sha e in he Ibe ian Peninsula, co e ing almos
50% o o al a ea o he whole pe iod. Al hough hey showed a sligh decline ( om 50.22% in 1990
o 47.26% in 2012), his concen a ed on some o he L1 classes o Dehesa/Mon ado, ain ed and
pe manen c ops, while i iga ed c ops and pas u es expe ienced a sligh inc ease o e he same
pe iod ha was no enough o compensa e o he losses o o he ag icul u al ca ego ies. The loss o
ag icul u al land appea ed o be sligh ly mo e p onounced in Spanish e i o y han in Po ugal.
Following he opposi e end, o es a ea expanded o e he same pe iod and wen om 19.73%
o o al a ea in 1990 o 22.17% in 2012. This expansion seemed o be concen a ed in he 2000–2012,
as he pe iod 1990–2000 showed la ge s abili y o o es a ea. Ha dwood o es s showed he la ges
inc ease, ollowed by coni e ous o es s (which unde wen a sligh dec ease du ing he i s en yea s
ha was mo e han compensa ed by he a ea gained du ing 2000–2012). The ends in Spain and
Po ugal, ne e heless, appea ed o be clea ly di e en . Fo es a ea in con inen al Spain inc eased om
18.4% in 1990 o 22.1% in 2012, while in con inen al Po ugal wen om 27.0 o 22.6%.
Sh ublands appea ed as he second L2 ca ego y in e ms o o al a ea in he Ibe ian Peninsula
(26.6%), and emained la gely s able along he whole pe iod. Again, hey seem o ha e ollowed
opposi e ends in bo h coun ies, om 28.6% in 1990 o 26.8% in 2012 in Spain and om 20.3% o 25.5%
in Po ugal. Buil (impe ious) a eas, while occupying much less a ea in o al, showed a ela i ely
impo an inc ease in bo h coun ies.
Land 2018,7, 99 5 o 12
3.2. Land Use/Co e T ansi ions
In his sec ion we ocus on he ela i e impo ance, in e ms o a ea, o he land co e ansi ions
(LCT) as de ined in he me hodology. Resul s sugges ha he changes ha ook place in 1990–2000
we e much mo e limi ed han hose occu ed in 2000–2012 (Table 3). Acco dingly, we will b ie ly
commen he esul s based on he ela i e impo ance o each LCT, o he whole Ibe ian Peninsula,
in 2000–2012.
Table 3.
Es ima ion o a ea a ec ed by land use/co e ansi ions (LCT) in he Ibe ian Peninsula,
con inen al Spain and con inen al Po ugal, in 1990–2000 and 2000–2012.
LCT
Spain Po ugal Peninsula
km2% km2% km2%
1990–2000
LCT1 (A o es a ion) 3810 0.77 2776 3.10 6586 1.13
LCT2 (Con . o ag ic.) 2688 0.54 425 0.48 3113 0.53
LCT3 (Abandonmen ) 914 0.19 482 0.54 1396 0.24
LCT4 (Ex ensi ica ion) 2310 0.47 253 0.28 2563 0.44
LCT5 (De o es a ion) 3263 0.66 2706 3.03 5969 1.02
LCT6 (Ch. o o es comp.) 149 0.03 120 0.13 269 0.05
LCT7 (In ensi ica ion) 3666 0.74 831 0.93 4497 0.77
LCT8 (U baniza ion) 1645 0.33 718 0.80 2363 0.41
2000–2012
LCT1 (A o es a ion) 39,630 8.03 2280 2.55 41,910 7.19
LCT2 (Con . o ag ic.) 22,585 4.57 1495 1.67 24,080 4.13
LCT3 (Abandonmen ) 23,954 4.85 1572 1.76 25,526 4.38
LCT4 (Ex ensi ica ion) 20,028 4.06 2512 2.81 22,540 3.87
LCT5 (De o es a ion) 16,237 3.29 5128 5.73 21,365 3.66
LCT6 (Ch. o o es comp.) 17,223 3.49 1432 1.60 18,655 3.20
LCT7 (In ensi ica ion) 29,041 5.88 1145 1.28 30,186 5.18
LCT8 (U baniza ion) 5500 1.11 606 0.68 6106 1.05
The mos impo an ansi ion in he Ibe ian Peninsula, in bo h pe iods, was LCT1 (a o es a ion).
Resul s indica e ha his p ocess was mo e in ense in Po ugal in 1990–2000, bu ha since hen
changed and a ec ed a la ge sha e o o al a ea in Spain in 2000–2012. Ag icul u al in ensi ica ion
(LCT7) appea ed as he second la ges p ocess in e ms o he sha e o a ea a ec ed (a o al o 5.18% o
he Peninsula in 2000–2012). Again, esul s sugges ha his p ocess a ec ed mo e a ea in 2000–2012
han in 1990–2000, and appea s o ha e been much mo e impo an in Spain han in Po ugal.
On a hi d le el (in 2000–2012) a e LCT2 (con e sion o ag icul u al use) and LCT3 (abandonmen
o ag icul u al use). Bo h appea ed o occupy sligh ly mo e han 4% o o al a ea in he Ibe ian Peninsula
in 2000–2012, and be conside ably mo e impo an in Spain han in Po ugal. These ansi ions a e
closely ollowed by LCT4 (ex ensi ica ion o ag icul u al use), LCT5 (de o es a ion) and LCT6 (change
o o es composi ion), which occupied a ound 3–4% o he Ibe ian Peninsula in 2000–2012. O hese
h ee, only LCT5 (de o es a ion) occupied a la ges sha e in Po ugal.
The las p ocess in e ms o he p opo ion o a ea occupied is LCT8 (u baniza ion/inc ease o
impe ious su ace). This seemed o ha e been mo e impo an in Po ugal han in Spain in 1990–2000,
bu a ec ed a la ge sha e o a ea in Spain in 2000–2012.
3.3. Clus e Analysis
The esul s o he clus e analysis allowed o iden i y se en dis inc g oups o municipali ies/
pa ishes, in e nally homogeneous in e ms o he ela i e impo ance o land use/co e ansi ions
(LCT) along he whole pe iod o s udy (Figu es 2and 3). Along he ollowing lines we will sho ly
desc ibe he main p ocesses in each g oup (de ailed a e ages o a ea occupied by each LCT in each
g oup o municipali ies/pa ishes a e shown in Table A3).
Land 2018,7, 99 6 o 12
Land 2018, 7, 99 6 o 12
lines we will sho ly desc ibe he main p ocesses in each g oup (de ailed a e ages o a ea occupied
by each LCT in each g oup o municipali ies/pa ishes a e shown in Table A3).
Figu e 2. Rada cha o a ea a ec ed by each ansi ion in each clus e .
Figu e 3. G oups o municipali ies/pa ishes (LAU2) esul ing om he clus e analysis.
Adminis a i e di ision a NUTS-2 le el is included as e e ence. Each clus e name appea s n he
legend along wi h he main land co e ansi ion (in e ms o a ea) in each case.
Clus e 1 combines municipali ies in which mo e han 45% o a ea was ans o med. Fo he
mos pa , his is explained by he expansion o o es s (LCT1), co e ing an a e age o 13.7% o
municipal/pa ish a ea, and in e nal changes in o es s (LCT6, 17.46% o municipal/pa ish a ea).
Figu e 2. Rada cha o a ea a ec ed by each ansi ion in each clus e .
Land 2018, 7, 99 6 o 12
lines we will sho ly desc ibe he main p ocesses in each g oup (de ailed a e ages o a ea occupied
by each LCT in each g oup o municipali ies/pa ishes a e shown in Table A3).
Figu e 2. Rada cha o a ea a ec ed by each ansi ion in each clus e .
Figu e 3. G oups o municipali ies/pa ishes (LAU2) esul ing om he clus e analysis.
Adminis a i e di ision a NUTS-2 le el is included as e e ence. Each clus e name appea s n he
legend along wi h he main land co e ansi ion (in e ms o a ea) in each case.
Clus e 1 combines municipali ies in which mo e han 45% o a ea was ans o med. Fo he
mos pa , his is explained by he expansion o o es s (LCT1), co e ing an a e age o 13.7% o
municipal/pa ish a ea, and in e nal changes in o es s (LCT6, 17.46% o municipal/pa ish a ea).
Figu e 3.
G oups o municipali ies/pa ishes (LAU2) esul ing om he clus e analysis. Adminis a i e
di ision a NUTS-2 le el is included as e e ence. Each clus e name appea s n he legend along wi h
he main land co e ansi ion (in e ms o a ea) in each case.
Clus e 1 combines municipali ies in which mo e han 45% o a ea was ans o med. Fo he
mos pa , his is explained by he expansion o o es s (LCT1), co e ing an a e age o 13.7% o
municipal/pa ish a ea, and in e nal changes in o es s (LCT6, 17.46% o municipal/pa ish a ea).
Land 2018,7, 99 7 o 12
Municipali ies in clus e 1 a e mainly loca ed in he Spanish egions o Galicia, Euskadi (Basque
Coun y), A agon and Ca alunya.
Clus e 2 is o med by municipali ies ha expe ienced a la ge inc ease o u ban a eas (an a e age
o 12.90% o municipal/pa ish a ea). While some o hem a e sca e ed along he whole Ibe ian
Peninsula, some clea concen a ions can be iden i ied a ound he main ci ies (e.g., Mad id, Lisbon,
Ba celona, Za agoza, Valencia, Po o).
Clus e 3 is o med by municipali ies wi h he highes deg ee o s abili y along he wen y- wo
yea s o s udy, as only an a e age o 11% o municipal/pa ish a ea was modi ied. The e is a sligh
endency owa ds a o es a ion, hough, as his showed sligh ly highe alues han de o es a ion.
Clus e 4 is o med by municipali ies wi h he highes a es o de o es a ion (16.9% on a e age).
Many o hese a eas a e loca ed in Po ugal, whe e hey occupy a la ge p opo ion o he o al a ea
along in e io a eas No h o Tejo (Tajo) i e and coas al a eas o he Sou h. Ne e heless, clea
ho spo s can be also app ecia ed in Spain, pa icula ly in he No hwes o Cas illa y León egion o in
he p o ince o Huel a in Andalucía.
Clus e 5 is o med by municipali ies ha expe ienced high a es o ag icul u al ex ensi ica ion
(15.17% o municipal a ea, on a e age). Ag icul u al in ensi ica ion and o he land use/co e
ansi ions a e also p esen in hese a eas, albei in a lowe scale. Many o he municipali ies/pa ishes
in his clus e appea in he Spanish egions o Andalucia and Ex emadu a and he Po uguese egion
o Alen ejo, al hough some a e also p esen in o he egions.
Clus e 6 ga he s municipali ies in which a o es a ion played a la ge ole (11.09% o municipal
a ea on a e age), bu which also show an in e es ing ension be ween con e sion o na u al and
semi-na u al a eas o ag icul u e (LCT2, ep esen ing on a e age 7.2% o municipal/pa ish a ea) and
a mland abandonmen (LCT3, 6.78%). The la ges a ea o med by hese municipali ies is loca ed in
he Spanish egion o Ex emadu a and he wes e n po ion o Cas illa la Mancha.
Finally, land use/co e ansi ions in clus e 7 appea clea ly domina ed by ag icul u al
in ensi ica ion (LCT7, 21.01%), e en hough a o es a ion, ag icul u al ex ensi ica ion and a mland
abandonmen a e also p esen in he ange o 4–5% o municipal/pa ish a ea. Along wi h clus e 1,
clus e 7 includes municipali ies wi h he highes amoun o changes (a ec ing a ound 43% o he
municipal/pa ish a ea). A eas in his clus e appea in se e al Spanish egions, mos ly in he Eas e n
po ion o he coun y.
4. Discussion
I is likely ha he pe iod o ime co e ed by his s udy has been one o e y subs an ial changes.
Some au ho s sugges ha he ins i u ional change ep esen ed by he access o Po ugal and Spain o
he Eu opean Economic Communi y in 1986 g ea ly accele a ed he e ec s o social and echnological
ans o ma ions o he second hal o he wen ie h cen u y [40].
The esul s ob ained a e in line wi h he indings o wo ks al eady published on land use/co e
changes in Spain and Po ugal. These ha e poin ed ou he exis ence o na u al egene a ion o o es
a eas and a o es a ion in se e al a eas o Spain, such as Galicia [
40
], Py enees [
20
], Cen al and Ibe ian
Sys em [
41
,
42
], as well as in Po ugal [
32
]. Many o he ci ed s udies coincide wi h a eas co esponding
o clus e s 1 and 6, which we iden i ied in his wo k as ho spo s o o es expansion. Al hough p obably
ini ia ed ea lie , o es expansion may ha e inc eased a e he access o bo h coun ies o he Eu opean
Economic Communi y in 1986, as a consequence o policy and ma ke changes ha accele a ed he
ans o ma ion o he ag icul u al sec o , including a o es a ion measu es unde Regula ion (EC)
2080/92 [
15
,
43
]. On he o he hand, he ne educ ion o o es a ea in Po ugal is also consis en wi h
published li e a u e [
19
,
21
], is la gely concen a ed in a eas included in clus e 4 and may be ela ed o
he ex en and ecu ence o o es i es [44].
Ano he p ocess commonly egis e ed by Eu opean li e a u e is ag icul u al ex ensi ica ion and
e en ual abandonmen o ag icul u al ac i i y [
3
,
25
]. In Spain, i has been documen ed in many a eas
included in clus e s 5 and 6. Fo example, in he Py enees and Eb o Dep ession [
45
,
46
], Andalucía [
47
],
Land 2018,7, 99 8 o 12
and Mu cia [
26
]. On he con a y, p oduc i e in ensi ica ion would concen a e in a eas included in
clus e 7. This dis ibu ion is consis en wi h he expansion o i iga ed g eenhouse c ops in whe e he
highes a es appea in coas al a eas [26,48] and he main i e basins [49].
Ano he one o he mos esea ched p ocesses is u baniza ion and he inc ease in impe ious
su ace. Howe e , o en imes he sp awl occu s di usely [
30
,
34
,
50
] and i is he e o e di icul
o quan i y using da a sou ces as CLC. This p ocess is well ep esen ed in clus e 2, including
municipali ies nea he main me opoli an a eas and Medi e anean coas al a eas in bo h coun ies.
As a as o es s composi ion is conce ned, a sho coming o he o iginal le el- h ee CLC da a is
he inabili y o ell apa di e en ypes o ha dwood o es s. The o iginal ca ego y 311 (“B oad-lea ed
o es s”) includes bo h as g owing species, mos ypically included in genus Eucalyp us, and slow
g owing na i e species. Conside ing ha he o igin o new s ands is usually di e en in bo h g oups
( e y o en human made plan a ions in he o me , spon aneous enc oachmen in he la e ) and ha
hei en i onmen al alue is also qui e di e en , i would ha e been nice o be able o classi y hem in
sepa a e ca ego ies. A ela ed di icul y is associa ed wi h he size o he minimum mapped a ea in
he o iginal da a (25 ha), ha is o en much la ge han he a e age plo size in he highly agmen ed
landscapes o he no he n and no hwes e n pa s o he Ibe ian Peninsula [
27
,
51
]. Pa icula ly when
conside ing he o iginal le el- h ee CLC class 313 (“Mixed o es s”), his means ha many pa ches
a e no necessa ily “mixed o es s” in s ic sense, bu a spa ial mix u e o small homogeneous o es
s ands [
13
]. The mosaic s uc u e o many landscapes in he Medi e anean basin makes i di icul o
assess changes using ela i ely coa se da ase s such as CLC.
Finally, esul s sugges a e y small quan i y o changes be ween 1990 and 2000 in Spain, compa ed
o wha is ound in Po ugal du ing he same pe iod [
52
], and specially compa ed o wha is ound
o bo h coun ies in 2000–2012. We we e conce ned ha his may be due o quali y p oblems o he
o iginal da a o he yea 2000 (mos likely) o he yea 1990 which would esul in an inabili y o
adequa ely cap u e changes be ween bo h yea s, pa icula ly in Spanish e i o y. Ou e o s o ca y
ou a clus e analysis using he land use/co e ansi ions o bo h ime pe iods p o ided unsa is ac o y
esul s, and esul s imp o ed g ea ly when we op ed o use ansi ions o e 1990–2012 as he inpu da a
o he clus e analysis. Howe e , ano he explana ion could be ela ed o changes in he me hodology
o CLC 2012 in compa ison wi h he one used in p e ious edi ions, as he Spanish e sion o CLC12
was p oduced by gene aliza ion o mo e de ailed, al eady exis ing, land use/co e maps, ins ead o
manual pho oin ep e a ion [53].
5. Conclusions
The gene al image p o ided by ou esul s indica es ha peninsula e i o y has been subjec ed o
in ense p ocesses o change as a esul o he cul u al, social and ins i u ional changes o ecen decades.
The decline o ag icul u al a ea and he inc ease o u banized and a i icial co e s appea as common
ends in he con inen al e i o y o bo h Po ugal and Spain. The ends ollowed by o es co e ,
ne e heless, appea as opposi e—an inc ease in Spain bu a dec ease in Po ugal. A he same ime,
he spa ial analysis allowed o p o ide an o e iew o he main g adien s o change ela ed o ensions
be ween ag icul u al in ensi ica ion–ex ensi ica ion, on he one hand, and de o es a ion–a o es a ion,
on he o he . The esul o he spa ial clus e analysis also allowed o iden i y he ho spo s o he
a o emen ioned p ocesses.
Au ho Con ibu ions:
Concep ualiza ion, E.C.-R. and D.F.-N.; Fo mal Analysis, D.F.-N.; Valida ion, E.C.-R.;
W i ing-O iginal D a P epa a ion, D.F.-N.; W i ing-Re iew & Edi ing, E.C.-R. and D.F.-N.; Visualiza ion, E.C.-R.
and D.F.-N.
Funding: This esea ch ecei ed no ex e nal unding.
Con lic s o In e es : The au ho s decla e no con lic o in e es .
Land 2018,7, 99 9 o 12
Appendix A
Table A1. Resul s: a ea o L1 and L2 classes in Spain in 1990, 2000 and 2012.
Land Use/Co e
1990 2000 2012
km2%km2%km2%
I iga ed c ops 21,054 4.26 23,501 4.76 25,577 5.18
Rain ed c ops 102,220 20.70 98,559 19.96 99,154 20.08
Pas u es 6299 1.28 6225 1.26 8584 1.74
Pe manen c ops 71,446 14.47 72,574 14.70 62,162 12.59
Dehesas 48,347 9.79 48,887 9.90 38,316 7.76
(Ag icul u al a eas) 249,366 50.51 249,745 50.58 233,793 47.36
Ha dwood o es 37,600 7.62 37,794 7.65 50,663 10.26
Coni e ous o es 38,900 7.88 38,384 7.77 44,691 9.05
Mixed o es 14,403 2.92 14,894 3.02 13,738 2.78
(Fo es a ea) 90,903 18.41 91,072 18.45 109,093 22.10
Sh ubland 141,531 28.67 139,043 28.16 132,534 26.85
Unp oduc i e 2015 0.41 1993 0.40 2263 0.46
We lands and wa e bodies 3768 0.76 4149 0.84 4228 0.85
A i icial Su ace 6139 1.24 7721 1.56 11,811 2.39
Table A2. Resul s: a ea o L1 and L2 classes in Po ugal in 1990, 2000 and 2012.
Land Use/Co e
1990 2000 2012
km2%km2%km2%
I iga ed c ops 1935 2.16 2596 2.90 2662 2.98
Rain ed c ops 10,954 12.25 10 228 11.44 8555 9.56
Pas u es 547 0.61 423 0.47 803 0.90
Pe manen c ops 16,264 18.19 16,070 17.97 15,741 17.60
Dehesas 13,772 15.40 13,346 14.93 14,004 15.67
(Ag icul u al a eas) 43,471 48.62 42,663 47.71 41,765 46.71
Ha dwood o es 10,604 11.86 11,263 12.60 10,322 11.54
Coni e ous o es 7897 8.83 7116 7.96 4825 5.40
Mixed o es 5637 6.30 5475 6.12 5065 5.66
(Fo es a ea) 24,139 27.00 23,854 26.68 20,211 22.60
Sh ubland 18,160 20.31 18,488 20.68 22,488 25.15
Unp oduc i e 351 0.39 351 0.39 100 0.11
We lands and wa e bodies 1173 1.31 1218 1.36 1485 1.67
A i icial Su ace 2123 2.37 2834 3.17 3361 3.76
Table A3.
A e age pe cen ages o a ea a ec ed by LCT o he municipali ies in each g oup, 1990–2012.
Clus e LCT1 LCT2 LCT3 LCT4 LCT5 LCT6 LCT7 LCT8
1 13.71 2.75 2.09 1.81 2.88 17.46 3.39 1.71
2 1.96 2.24 1.38 2.70 2.60 1.01 1.39 12.90
3 2.22 1.12 1.62 1.76 1.43 0.73 1.75 0.50
4 4.83 1.95 2.47 2.14 16.89 3.79 1.54 0.68
5 3.14 3.84 3.60 15.17 0.96 0.64 4.94 1.36
6 11.09 7.21 6.78 2.26 2.50 1.21 2.21 0.82
7 5.93 3.59 5.19 4.03 0.96 0.64 21.01 2.06
