Niche characterization of shrub functional groups along an atlantic-mediterranean gradient

A icle Niche Cha ac e iza ion o Sh ub Func ional G oups along an A lan ic-Medi e anean G adien Ad iana E. Ol ho 1, Ca olina Ma ínez-Ruiz 1,2,* and Josu G. Alday 3,4,*   Ci a ion: Ol ho , A.E.; Ma ínez-Ruiz, C.; Alday, J.G. Niche Cha ac e iza ion o Sh ub Func ional G oups along an A lan ic-Medi e anean G adien . Fo es s 2021,12, 982. h ps:// doi.o g/10.3390/ 12080982 Academic Edi o : Panayo is Dimopoulos Recei ed: 16 May 2021 Accep ed: 22 July 2021 Published: 24 July 2021 Publishe ’s No e: MDPI s ays neu al wi h ega d o ju isdic ional claims in published maps and ins i u ional a il- ia ions. Copy igh : © 2021 by he au ho s. Licensee MDPI, Basel, Swi ze land. This a icle is an open access a icle dis ibu ed unde he e ms and condi ions o he C ea i e Commons A ibu ion (CC BY) license (h ps:// c ea i ecommons.o g/licenses/by/ 4.0/). 1Á ea de Ecología, Depa amen o Ciencias Ag o o es ales, E.T.S. de Ingenie ías Ag a ias, Uni e sidad de Valladolid, A da. de Mad id 50, 34071 Palencia, Spain; aol ho [email p o ec ed] 2Sus ainable Fo es Managemen Resea ch Ins i u e, Uni e si y o Valladolid-INIA, E.T.S.II.AA., A da. Mad id 50, 34071 Palencia, Spain 3Join Resea ch Uni CTFC–AGROTECNIO–CERCA, 25198 Lleida, Spain 4Depa men o C op and Fo es Sciences, Uni e si a de Lleida, A da Alcalde Ro i a Rou e 191, 25198 Lleida, Spain *Co espondence: [email p o ec ed] (C.M.-R.); [email p o ec ed] (J.G.A.) Abs ac : The iden i ica ion o he ac o s con olling he unde s o y species dis ibu ion and abun- dance is essen ial o unde s and he ecology and dynamics o na u al o es s and hei managemen esponse. We assess he ela ionships be ween en i onmen al g adien s and sh ub unc ional g oups dis ibu ion pa e ns and niche cha ac e is ics in a ansi ional a ea be ween he Eu osibe ian and Medi e anean biogeog aphic egions in No he n Spain. He e, 772 plo s om he 3 d Spanish Na ional Fo es In en o y we e used. Sh ub unc ional g oups espond o he same complex en i- onmen al g adien s as ees, i.e., he no h-sou h clima ic g adien and a slope g adien . Unimodal esponse cu es o sh ub unc ional g oups and amilies domina e along bo h g adien s, p o iding e idence o success ul unc ional u no e . Simila o ee species, he niche loca ion o unc ionally ela ed sh ubs is close. Func ional g oups occupying en i onmen s wi h sha p con as o ansi ional en i onmen s ha e he b oades niches, whe eas hose specialized unc ional g oups occupying local- ized habi a s showed he na owes niches. The knowledge o sh ub species dis ibu ions and niche cha ac e is ics along complex en i onmen al g adien s will imp o e ou abili y o discuss po en ial conse a ion managemen goals o h ea s due o land-use changes and u u e clima e change. Keywo ds: coenoclines; o es in en o y da a; Huisman-Ol -F esco (HOF) modeling; niche wid h; No he n Spain; sh ub unc ional g oup dis ibu ion 1. In oduc ion Sh ub species li ing in he o es unde s o y a e impo an elemen s o ecosys em s uc u e and unc ion, p o iding habi a and o age o wildli e [ 1 – 4 ] and con ibu ing signi ican ly o plan di e si y [ 5 – 7 ]. Being pa icula ly impo an in a eas whe e sh ub unde s o y is signi ican ly mo e di e se han o es o e s o y, such as he Palencia p o ince (No he n Spain) [ 8 , 9 ]. I is widely ecognized ha sh ubs acili a e he g ow h o p o ec ed species unde hei canopy [ 10 ], besides he ec ui men o ee seedlings mos ly om Que - cus L. (oaks) [ 11 – 14 ] and Pinus L. (pines) species [ 15 – 17 ] in Medi e anean en i onmen s. These acili a i e e ec s a e undamen al o egene a e o es ecosys ems in Medi e anean egions [ 18 , 19 ]. Addi ionally, sh ubs play a undamen al ole as ecosys em enginee s o inc ease he e ogenei y in mic o-en i onmen al condi ions a o ing la e-successional species es ablishmen [ 12 , 20 – 22 ]. Con e sely, i has been shown ha ce ain sh ubs a o exo ic species in asion (e.g., in he No he n Cali o nian coas al dunes E icame ia e icoides (Less.) Jeps. enhances he coloniza ion o he in asi e B omus diand us Ro h [ 19 ]), and e en some sh ub species become in asi e in he lack o app op ia e managemen p ac ices (e.g., Cy isus scopa ius (L.) Link in F ance and Aus alia [ 23 , 24 ]). Thus, he desc ip ion o en i onmen al ac o s shaping he dis ibu ion and abundance o unde s o y sh ub Fo es s 2021,12, 982. h ps://doi.o g/10.3390/ 12080982 h ps://www.mdpi.com/jou nal/ o es s Fo es s 2021,12, 982 2 o 16 species has undamen al implica ions o o es di e si y conse a ion [ 25 , 26 ] and o es managemen [27,28], pa icula ly in he cu en scena io o clima e change [29–31]. Despi e he ac ha sh ubs a e key componen s o o es ecosys ems [ 32 , 33 ], he e ha e been ew e o s o model sh ubs abundance o dis ibu ion [ 5 , 34 ], and e y ew models o simula ion o o es dynamics inco po a e he esponse o unde s o y ege a- ion laye s [ 26 ]. Mos likely, he unde s o y esponses a e mo e complex han o e s o y esponses. Fo ins ance, mic oclima ic ex emes a e amelio a ed by mode a e o e s o y co e , which would o he wise be s ess ul o g ound-laye plan s, in con as , high o e - s o y co e supp esses unde s o y species i espec i e o en i onmen al cons ain s [ 5 ]. Thus, i migh be assumed ha unde s o y ege a ion has a unimodal esponse o o es bio ic a iables, such as o e s o y co e [ 5 ], along wi h en i onmen al g adien s [ 35 , 36 ]. Ne e heless, as epo ed o opical o es s [ 37 , 38 ], unde s o y sh ub species migh no espond o he same complex en i onmen al g adien s as ee species. A he same ime, sh ub communi y esponses a e o en highly complex [ 39 ]. The e o e, o educe complexi y, sh ub species a e usually g ouped acco ding o he ai s hey sha ed in o di e en unc- ional g oups o plan unc ional ypes [ 40 ], such as Raunkiæ ’s li e- o ms (chame ophy e, phane ophy e), dispe sal mode (anemocho y, zoocho y . . . ) o egene a ion me hod (ge - mina o , esp ou e ). In any case, a deepe unde s anding o sh ub unc ional esponses o complex en i onmen al g adien s is needed. In his sense, we assessed he ela ionships be ween sh ub unc ional g oup dis ibu- ion pa e ns and niche cha ac e is ics and wo main en i onmen al g adien s (coenoclines) p e iously iden i ied [ 41 ] in a ansi ional a ea be ween he Eu osibe ian and Medi e - anean biogeog aphic egions in No he n Spain. We cha ac e ized he ealized niche o sh ub unc ional g oups along bo h coenoclines, conside ing ha he species ealized niche is concei ed as a mul idimensional esponse su ace and is composed by he undamen al equi emen s o he species and i s unc ion, i.e., in e ac ions wi h o he species h ough compe i i e exclusion and acili a ion [ 42 ]. He e, conside ing ha unc ional g oups a e impo an componen s o ecosys em unc ion [ 43 ], we used a sh ub- unc ional-g oups app oach. Ou speci ic objec i es we e: (1) o assess whe he unde s o y sh ub unc ional g oups esponded o he same egional complex en i onmen al g adien s as ee species (i.e., al i ude and empe a u e, and s eepness); (2) o de e mine whe he sh ub unc ional g oups exhibi ed a unimodal esponse o hose en i onmen al g adien s; (3) o quan i y he op ima and niche wid hs o sh ub unc ional g oups ac oss he main g adien s. The desc ip ion o sh ub unc ional g oups dis ibu ion and niche ea u es along complex en i- onmen al g adien s will help o imp o e ou knowledge o discuss po en ial conse a ion managemen h ea s and goals because o land uses and clima e changes [15]. 2. Ma e ials and Me hods 2.1. S udy A ea and Da a Sou ce The en i onmen al g adien unde s udy uns om he no h o he sou h o he Palencia p o ince (No he n Spain; 43 ◦ 04 0 N and 41 ◦ 46 0 N la i ude, and 3 ◦ 53 0 W and 5 ◦ 02 0 W longi ude; Figu e 1A). The g adien is 140 km in leng h and comp ises a high a ie y o en i onmen al condi ions and landscapes [ 9 , 44 ], as a esul o he con luence o he A lan ic and Medi e anean biogeog aphic egions and wo geomo phological uni s ( he Can ab ian Range and he Cas ilian pla eau), which ha bo g ea ege a ion di e si y [ 8 ]. The clima e changes om Alpine o he Medi e anean om no h o sou h [ 45 ], which is ela ed o opog aphy (Figu e 1B), mainly de e mined by he p esence o he Can ab ian Range in he no h (al i ude up o 2540 m), and he Cas ilian pla eau (mean al i ude ca. 800 m) in he Cen e and sou he n pa s o he g adien . The empe a u e inc eases ma kedly om no h o sou h ollowed by a no o ious inc ease in p ecipi a ion (highe xe ici y) ha leads o a ise in he con inen al na u e o he clima ic condi ions [ 46 ]. Thus, he clima ic and en i onmen al g adien h oughou he Palencia p o ince is o pa icula in e es when desc ibing and cha ac e izing he o es ege a ion composi ional changes om spa ial pe spec i es along b oad he e ogeneous en i onmen al g adien s. Fo es s 2021,12, 982 3 o 16 Fo es s 2021, 12, x FOR PEER REVIEW 3 o 16 800 m) in he Cen e and sou he n pa s o he g adien . The empe a u e inc eases ma k- edly om no h o sou h ollowed by a no o ious inc ease in p ecipi a ion (highe xe ici y) ha leads o a ise in he con inen al na u e o he clima ic condi ions [46]. Thus, he cli- ma ic and en i onmen al g adien h oughou he Palencia p o ince is o pa icula in e - es when desc ibing and cha ac e izing he o es ege a ion composi ional changes om spa ial pe spec i es along b oad he e ogeneous en i onmen al g adien s. Figu e 1. (A) Loca ion o 772 ield plo s om 3SNFI conside ed o s udy in he Palencia p o ince (Spain). Plo s a e ep e- sen ed o e he Digi al Ele a ion Model (DEM 200 m) using QGIS 3.16.6. (B) Topog aphic g adien o he Palencia p o ince om he no h (moun ains) o he sou h (limes one moo s), ob ained om DEM 25 m. Bo h DEM om MDT2000 © Ins i- u o Geog á ico Nacional de España). Sh ub species da a om 772 pe manen ield plo s (Figu e 1A) om he 3 d Spanish Na ional Fo es In en o y (3SNFI; 1997–2007) we e used. The SNFI measu e ci cula plo s o a iable adius (5, 10, 15, and 25 m) sys ema ically dis ibu ed and loca ed on he in e - sec ion nodes o a 1 × 1 km 2 Uni e sal T ans e se Me ca o g id; only plo s loca ed inside o es a eas a e measu ed (see [41,46] o u he de ails on plo s selec ion, cha ac e is ics, and dis ibu ion). T ees wi h a diame e a b eas heigh o 75, 125, 225, and 425 mm a e measu ed, espec i ely, in each one o he concen ic ci cum e ences. The co e (%) o all Figu e 1. ( A ) Loca ion o 772 ield plo s om 3SNFI conside ed o s udy in he Palencia p o ince (Spain). Plo s a e ep esen ed o e he Digi al Ele a ion Model (DEM 200 m) using QGIS 3.16.6. ( B ) Topog aphic g adien o he Palencia p o ince om he no h (moun ains) o he sou h (limes one moo s), ob ained om DEM 25 m. Bo h DEM om MDT2000 © Ins i u o Geog á ico Nacional de España). Sh ub species da a om 772 pe manen ield plo s (Figu e 1A) om he 3 d Spanish Na ional Fo es In en o y (3SNFI; 1997–2007) we e used. The SNFI measu e ci cula plo s o a iable adius (5, 10, 15, and 25 m) sys ema ically dis ibu ed and loca ed on he in e - sec ion nodes o a 1 × 1 km 2 Uni e sal T ans e se Me ca o g id; only plo s loca ed inside o es a eas a e measu ed (see [ 41 , 46 ] o u he de ails on plo s selec ion, cha ac e is ics, and dis ibu ion). T ees wi h a diame e a b eas heigh o 75, 125, 225, and 425 mm a e measu ed, espec i ely, in each one o he concen ic ci cum e ences. The co e (%) o all sh ub species p esen in a ixed plo adius o 10 m [ 46 ] was in en o ied. A o al o 86 sh ub species we e egis e ed in he 772 selec ed plo s. Howe e , only 47 o hem we e used in subsequen analyses o ha ing su icien co e and equency in he plo s, a co ec axonomical iden i ica ion in he ield [ 41 ], and a ailable in o ma ion o hei unc ional ai s in da abases o Sou he n Eu opean Flo a. These 47 sh ub species we e classi ied in habi a - ela ed unc ional g oups (see Table A1) such as amily, geog aphical dis ibu- Fo es s 2021,12, 982 4 o 16 ion (A lan ic, Medi e anean, A lan ic-Medi e anean, endemic o he Ibe ian Peninsula), bio ype (pe ennial, deciduous), Raunkiæ ´ s li e- o m (chame ophy es, phane ophy es), dispe sal mode (anemocho y, zoocho y, au hoco y, ba oco y, o he ), and egene a ion me hod (ge mina o , esp ou e ). 2.2. Coenoclines Cha ac e iza ion In p e ious s udies, an indi ec o dina ion echnique (de ended co espondence analysis, DCA) was applied on he co e (%) ma ix o all woody species ( ees and sh ubs) p esen in each o he 772 selec ed plo s [ 41 , 46 ]. DCA showed ha changes in he o es ege a ion composi ional changes along he Palencia p o ince we e mainly de e mined by he no h-sou h clima ic di e ences (DCA1, p ima y coenocline), al hough a seconda y coenocline also showed a u no e o species ela ed o he slope g adien (DCA2, s eepness). In pa icula , Fagus syl a ica L. and Que cus pe aea (Ma .) Liebl. domina ed deciduous moun ain o es s ha a e eplaced by Que cus py enaica Willd. o es s and hose, in u n, by Que cus aginea Lam. and Que cus ilex subsp. ballo a (Des .) Samp. Fo es s, as a idi y inc eases owa ds he sou h. Pinus syl es is L. and Pinus nig a J.F. A nold domina ed coni e ous woodlands in moun ain a eas and de i al moo s (middle pa o he p o ince; Figu e 1B), whe eas he sou he n limes one moo s (Figu e 1B) a e domina ed by plan a ions o Pinus halepensis Mill., Pinus pinea L. and Cup esus sempe i ens L. In e es ingly, along he slope g adien he e is also a u no e o ee species, pa ing na u al Junipe us spp., Que cus pe aea,Que cus py enaica, and Fagus syl a ica o es s loca ed on s eeply sloping si es, and pine plan a ions (Pinus nig a,Pinus pinas e Ai on) ha domina e in la a eas. The e o e, he sea ch o pa e ns ha ela e ee species ends wi h unde s o y sh ub unc ional g oup ends along hese wo main coenoclines will gi e us aluable in o ma ion o managemen and conse a ion. 2.3. Da a Analyses Fi s , he main wo coenoclines we e p e iously iden i ied (DCA1, DCA2) [ 41 , 46 ]. The e o e, he e we modeled he abundance (co e pe cen age) o sh ub unc ional g oups along hese coenoclines using HOF models (HOF = Huisman-Ol -F esco [ 47 ]) h ough he ‘eHOF’ package ( e sion 1.7 [ 48 ]) in he R-language en i onmen ( e sion 4.0.3; R De elop- men Co e Team, Vienna, Aus ia, h p://www. -p ojec .o g, accessed on 1 Feb ua y 2020). HOF models allow desc ibing he species esponses, which may esul om bo h en i on- men al condi ions and in a- and in e speci ic in e ac ions [ 49 ]. HOFs a e a hie a chical se o esponse models, anked by hei inc easing complexi y (Model I, no species end; Model II, inc easing o dec easing end; Model III, inc easing o dec easing end below maximum a ainable esponse; Model IV, symme ical esponse cu e; Model V, skewed esponse cu e). The selec ion o he mos app op ia e model o each o he unc ional g oups analyzed was done using he Akaike In o ma ion C i e ium (AIC) [ 50 ]; smalle AIC alues indica e mo e pa simonious models. Finally, o hose unc ional g oups wi h unimodal esponses, he loca ion o niche op ima ( µ ) and niche wid hs (2 ) we e de i ed om he HOF models [49]. 3. Resul s The mos equen unde s o y sh ub species we e E icaceae:Calluna ulga is (L.) Hull (in 13.6% o plo s), E ica spp. (12.8%) and E ica aus alis L. (10.1%). O he sh ub species such as Rosa spp., Cy isus scopa ius (L.) Link, Thymus spp., A c os aphylos u a-u si (L.) Sp eng., E ica cine ea L.,and E ica a bo ea L. we e p esen in ca. 3–7% o plo s. Fabaceae,Rosaceae, and E icaceae accoun ed o 49% o sh ub gene a and 44% o sh ub species. 3.1. Dis ibu ion Pa e ns o Main Sh ub Families along Coenoclines Fabaceae and Cap i oliaceae we e he unique amilies wi h inde e mina e esponse cu es (HOF model I) since bo h had low and cons an co e (<1%) along inc easing a idi y DCA1 coenocline. Thymelaeaceae and Oleaceae (HOF model II wi h a dec easing end) had Fo es s 2021,12, 982 5 o 16 he g ea es co e in he no he n moun ains wi h a subsequen educ ion owa ds he sou h (DCA1 le -end; Figu e 2A). Only As e aceae (HOF model II wi h an inc easing end) showed an inc easing end as a idi y inc eases owa ds he sou h (DCA1 igh -end). The emainde axonomic g oups showed unimodal esponse cu es (HOF models IV o V; Figu e 2A and Table 1) wi h op ima a di e en poin s along he g adien , o example, Aqui oliaceae and Be ulaceae wi h op ima in he no he n moun ains ( DCA1 < 2 ) had he na owes niche wid h ( 2 = 0.5 ), E icaceae had in e media e niche wid h ( 2 = 2.3 ) wi h i s op imum in he ansi ional a ea be ween he A lan ic and Medi e anean condi ions ( DCA1 ≈2 ), Rosaceae and Cis aceae had he b oades niche wid hs (2 = 6.1 and 4.1, espec- i ely) wi h op ima in he g adien middle (DCA1 ≈ 4), and Labia ae had in e media e niche wid h (2 = 2.6) wi h i s op imum a he igh end o DCA1, i.e., owa ds he sou h. Fo es s 2021, 12, x FOR PEER REVIEW 5 o 16 Sp eng., E ica cine ea L., and E ica a bo ea L. we e p esen in ca. 3–7% o plo s. Fabaceae, Rosaceae, and E icaceae accoun ed o 49% o sh ub gene a and 44% o sh ub species. 3.1. Dis ibu ion Pa e ns o Main Sh ub Families along Coenoclines Fabaceae and Cap i oliaceae we e he unique amilies wi h inde e mina e esponse cu es (HOF model I) since bo h had low and cons an co e (<1%) along inc easing a id- i y DCA1 coenocline. Thymelaeaceae and Oleaceae (HOF model II wi h a dec easing end) had he g ea es co e in he no he n moun ains wi h a subsequen educ ion owa ds he sou h (DCA1 le -end; Figu e 2A). Only As e aceae (HOF model II wi h an inc easing end) showed an inc easing end as a idi y inc eases owa ds he sou h (DCA1 igh - end). The emainde axonomic g oups showed unimodal esponse cu es (HOF models IV o V; Figu e 2A and Table 1) wi h op ima a di e en poin s along he g adien , o example, Aqui oliaceae and Be ulaceae wi h op ima in he no he n moun ains (DCA1 < 2) had he na owes niche wid h (2 = 0.5), E icaceae had in e media e niche wid h (2 = 2.3) wi h i s op imum in he ansi ional a ea be ween he A lan ic and Medi e anean condi- ions (DCA1 ≈ 2), Rosaceae and Cis aceae had he b oades niche wid hs (2 = 6.1 and 4.1, espec i ely) wi h op ima in he g adien middle (DCA1 ≈ 4), and Labia ae had in e medi- a e niche wid h (2 = 2.6) wi h i s op imum a he igh end o DCA1, i.e., owa ds he sou h. Figu e 2. HOF-de i ed esponse cu es o main axonomic g oups among sh ubs in Palencia (No he n Spain), ela i e o he i s (A) and second (B) main coenoclines (DCA1 and DCA2, espec i ely). Fabaceae (Faba) and Cap i oliaceae (Cap i ) wi h HOF model I a e no shown in Figu e 2A. Codes a e p esen ed in Table 1, wi h he addi ional: Thymelaeaceae = Thymel; Oleaceae = Olea. Table 1. Loca ion, along DCA1 and DCA2, o op imum (μ), op alue (p edic ed maximum p oba- bili y), and niche wid hs based on 2 ole ances o axonomic g oups wi h unimodal esponses (HOF model IV; only HOF model V o Be ulaceae along bo h coenoclines and Aqui oliaceae along DCA1). Taxonomic G oup Code μ Top Value 2 -niche DCA 1 Rosaceae Rosa 3.923 0.006 6.08 Cis aceae Cis a 3.786 0.010 4.10 Labia ae Labia 6.387 0.070 2.55 E icaceae E ica 2.224 0.079 2.25 Be ulaceae Be ul 0.882 0.040 0.49 Aqui oliaceae Aqui o 0.358 0.051 0.48 DCA2 Aqui o Be ul Cis a As e a E ica Labia Olea Rosa Thymel 0 1 2 3 4 5 012345678 Family co e (%) DCA1 (A) Aqui o Be ul Cap i Cis a As e a E ica Labia Faba Olea Rosa Thymel 0 1 2 3 4 5 012345 Family co e (%) DCA 2 (B) Figu e 2. HOF-de i ed esponse cu es o main axonomic g oups among sh ubs in Palencia (No he n Spain), ela i e o he i s ( A ) and second ( B ) main coenoclines (DCA1 and DCA2, espec i ely). Fabaceae (Faba) and Cap i oliaceae (Cap i ) wi h HOF model I a e no shown in Figu e 2A. Codes a e p esen ed in Table 1, wi h he addi ional: Thymelaeaceae = Thymel; Oleaceae = Olea. Table 1. Loca ion, along DCA1 and DCA2, o op imum ( µ ), op alue (p edic ed maximum p obabil- i y), and niche wid hs based on 2 ole ances o axonomic g oups wi h unimodal esponses (HOF model IV; only HOF model V o Be ulaceae along bo h coenoclines and Aqui oliaceae along DCA1). Taxonomic G oup Code µTop Value 2 -Niche DCA1 Rosaceae Rosa 3.923 0.006 6.08 Cis aceae Cis a 3.786 0.010 4.10 Labia ae Labia 6.387 0.070 2.55 E icaceae E ica 2.224 0.079 2.25 Be ulaceae Be ul 0.882 0.040 0.49 Aqui oliaceae Aqui o 0.358 0.051 0.48 DCA2 Rosaceae Rosa 4.517 0.014 3.42 E icaceae E ica 1.454 0.056 2.71 Labia ae Labia 1.734 0.023 1.23 As e aceae As e a 1.742 0.006 0.95 Aqui oliaceae Aqui o 3.469 0.030 0.94 Be ulaceae Be ul 3.574 0.087 0.13 Along DCA2 coenocline o inc easing s eepness, any axonomic g oup showed HOF model I, only Cis aceae (HOF model II wi h a dec easing end) dec eased i s co e owa ds he s eepes si es (DCA2 igh -end), whe eas Fabaceae,Oleaceae,Thymelaeaceae, and Cap i oli- Fo es s 2021,12, 982 6 o 16 aceae (HOF model II wi h an inc easing end) inc eased hei co e owa ds he s eepes si es (DCA2 le -end). The o he axonomic g oups showed unimodal esponse cu es (HOF models IV o V; Figu e 2B and Table 1) wi h op ima a di e en poin s along he g adien : Be ulaceae,Aqui oliaceae, and Rosaceae owa ds he s eepes si es (DCA2 > 3.5), whe eas E icaceae,As e aceae, and Labia ae owa ds he less sloping si es (DCA2 < 2); while Rosaceae and E icacea had he b oades niche wid hs (2 = 3.4 and 2.7, espec i ely). 3.2. Sh ub Func ional G oup Responses along Coenoclines Conside ing he geog aphical dis ibu ion, only Medi e anean sh ubs species did no shows a unimodal esponse cu e along DCA1 coenocline o inc easing a idi y. They showed an inc easing end below maximum a ainable esponse (HOF model III; Figu e 3A ) owa ds he sou h (DCA1 igh -end). The sh ubs classi ied as A lan ic, A lan ic- Medi e - anean, and endemic o Ibe ian Peninsula showed skewed unimodal esponse cu es along DCA1 coenocline (model V; Figu e 3A; Table 2). A lan ic sh ubs wi h op imum in he no he n moun ains (DCA1 < 2) had he b oades niche (2 = 4.4), whe eas A lan ic- Medi e anean, and endemic o Ibe ian Peninsula sh ubs wi h na owe niches (2 = 1.7 and 2 = 1.1, espec i ely) had he op ima in he ansi ional a ea be ween he A lan ic and Medi e anean condi ions (DCA1 ≈2). When bio ype o Raunkiæ ´ s li e- o m was conside ed, bo h deciduous and pe ennial o phane ophy es and chame ophy es sh ubs showed skewed unimodal esponse cu es along DCA1 coenocline (model V; Figu e 3B,C; Table 2). Phane ophy es and deciduous sh ubs showed close niche wid hs (2 = 3.0 and 2.6, espec i ely) wi h he op ima in he no he n moun ains (DCA1 < 2), whe eas pe ennial and chamaephy es wi h he same b oades niche wid h (2 = 4.0) had op ima in he ansi ional a ea be ween he A lan ic and Medi e anean condi ions (DCA1 ≈2). In he case o sh ub dispe sal modes, only au ocho y and ba ocho y did no display unimodal esponse cu es along DCA1 coenocline (Figu e 3D): au ocho y showed a HOF model II wi h a dec easing end owa ds he sou h (DCA1 igh -end), while ba ocho y showed HOF model III wi h an inc easing end below maximum a ainable esponse owa ds he sou h (DCA1 igh -end). Anemocho y and zoocho y had he op ima in he ansi ional a ea be ween he A lan ic and Medi e anean condi ions (DCA1 ≈ 2), whe eas ‘o he ’ had he op imum in he igh end o DCA1 coenocline, i.e., owa ds he sou h; he h ee dispe sal modes had simila and in e media e niche wid hs (2 = 2.0–3.3; Table 2). Bo h ca ego ies o egene a ion me hod displayed skewed unimodal esponse cu es along DCA1 coenocline (model V; Figu e 3E; Table 2): esp ou e s wi h he op imum in he no he n moun ains (DCA1 < 2) and ge mina o s wi h he op imum in he middle pa o he g adien (DCA1 ≈4); bo h wi h na ow niche wid hs (2 < 1.5). Along DCA2 coenocline o inc easing s eepness, and conside ing sh ub geog aphical dis ibu ion, only A lan ic sh ubs did no show unimodal esponse cu es bu a HOF model III wi h an inc easing end (Figu e 3F) owa ds he s eepes si es (DCA2 igh -end). Sh ubs classi ied as he Medi e anean, A lan ic-Medi e anean, and endemic o he Ibe ian Peninsula showed skewed unimodal esponse cu es (model V; Table 2) wi h op ima in he s eepes si es (DCA2 < 2); he A lan ic-Medi e anean sh ubs had he b oades niche wid h (2 = 3.9). When bio ype o Raunkiæ ´ s li e- o m we e conside ed (Figu e 3G,H), bo h deciduous and pe ennial, as well as chame ophy es showed skewed unimodal esponse cu es along DCA2 coenocline (model V; Table 2) wi h close niche wid hs (2 = 1.5–1.8). Chame ophy es and pe ennial sh ubs had he op ima in he s eepes si es (DCA2 < 2), whe eas deciduous sh ubs had he op imum in he middle pa o he slope g adien (DCA2 = 4.6). Phane o- phy es showed a HOF model II wi h an inc easing end owa ds he s eepes si es (DCA2 igh -end). Fo es s 2021,12, 982 7 o 16 Fo es s 2021, 12, x FOR PEER REVIEW 7 o 16 Figu e 3. HOF-de i ed esponse cu es o sh ub unc ional g oups in Palencia (No he n Spain), ela i e o he i s (A–E) and second (F–J) coenoclines (DCA1 and DCA2, espec i ely); (A,F): geo- g aphical dis ibu ion, (B,G): bio ype, (C,H): Raunkiæ ´s li e- o m, (D,I): dispe sal mode, and (E,J): egene a ion me hod. 0 5 10 15 20 25 30 012345678 Co e (%) (A) A l-Med A lan ic Medi e anean Endemic o IP 0 5 10 15 20 25 30 35 40 45 012345678 Co e (%) (B) Pe ennial Deciduous 0 5 10 15 20 25 30 35 40 012345678 Co e (%) (C) Chamephy es Phane ophy es O he 0 5 10 15 20 25 30 35 012345678 Co e (%) (D) Au hoco y Ba ocho y Zoocho y Anemocho y 0 5 10 15 20 25 30 35 40 012345678 Co e (%) DCA1 (E) Resp ou e s Ge mina o s 0 5 10 15 20 25 30 35 40 45 50 012345 Co e (%) (F) ATL-MED Medi e anean A lan ic Endemic IP 0 5 10 15 20 25 30 35 40 45 50 55 012345 Co e (%) (G) Pe ennial Deciduous 0 5 10 15 20 25 30 35 40 45 50 55 012345 Co e (%) (H) Chamephy es Phane ophy es 0 5 10 15 20 25 30 35 40 45 50 012345 Co e (%) (I) Anemoco y Au hoco y Zoocho y Ba ocho y O he 0 5 10 15 20 012345 Co e (%) (J) Resp ou e s Ge mina o s DCA 2 Figu e 3. HOF-de i ed esponse cu es o sh ub unc ional g oups in Palencia (No he n Spain), ela i e o he i s ( A – E ) and second ( F – J ) coenoclines (DCA1 and DCA2, espec i ely); ( A , F ): geog aphical dis ibu ion, ( B , G ): bio ype, ( C , H ): Raunkiæ ´ s li e- o m, ( D , I ): dispe sal mode, and (E,J): egene a ion me hod. Fo es s 2021,12, 982 8 o 16 Table 2. Loca ion o op imum ( µ ), op alue (p edic ed maximum p obabili y), and niche wid hs based on 2 ole ances o sh ub unc ional g oups wi h unimodal esponses (HOF model V; only HOF model IV o ‘au hoco y’ along DCA1 and ‘A lan ic-Medi e anean’ along DCA2). Da a o A lan ic sh ubs along DCA2 a e no a ailable since hei op ima all ou side he sampled pa o he g adien . Fuc ional G oup µTop Value 2 -Niche DCA1 Geog aphical dis ibu ion A lan ic 1.69 0.18 4.39 A lan ic-Medi e anean 2.68 0.11 1.69 Endemic o Ibe ian Peninsula 2.08 0.04 1.12 Bio ype Pe ennial 1.95 0.28 4.01 Deciduous 1.07 0.14 3.05 Raunkiæ ’s li e- o m Chamaephy es 2.49 0.23 4.02 Phane ophy es 1.26 0.21 2.62 Dispe sal mode Anemocho y 1.95 0.13 2.06 Zoocho y 2.64 0.20 3.30 O he 6.75 0.08 2.02 Regene a ion me hod Ge mina o s 0.85 0.08 1.04 Resp ou e s 1.63 0.22 1.15 DCA2 Geog aphical dis ibu ion Medi e anean 1.05 0.14 1.39 A lan ic-Medi e anean 1.90 0.07 3.9 Endemic o Ibe ian Peninsula 1.27 0.03 1.29 Bio ype Pe ennial 1.07 0.30 1.78 Deciduous 4.64 0.33 1.52 Raunkiæ ’s li e- o m Chamaephy es 1.25 0.25 1.72 Dispe sal mode Zoocho y 4.57 0.28 2.53 Au hoco y 0.97 0.14 0.99 Ba ocho y 1.82 0.01 0.74 O he 1.88 0.04 0.40 Regene a ion me hod Ge mina o s 3.62 0.10 1.18 Resp ou e s 1.06 0.13 1.39 In he case o sh ub dispe sal modes (Figu e 3I), anemocho y showed a HOF model III wi h an inc easing end owa ds he s eepes si es (DCA2 igh -end). The emainde g oups showed unimodal esponse cu es (model V; Table 2) along DCA2 coenocline. Au hocho y, ba ocho y, and ‘o he ’ wi h e y na ow niche wid hs (2 < 1) had op ima in he s eepes si es (DCA2 le -end), whe eas zoocho y wi h he b oades niche wid h (2 = 4.6) had he op imum owa ds he less slopping si es (DCA2 igh -end). Bo h ca ego ies o egene a ion me hod showed skewed unimodal esponse cu es along DCA2 coenocline (model V; Figu e 3J; Table 2). Resp ou e s wi h he op imum in he s eepes si es (DCA2 le -end), and ge mina o s wi h he op imum owa ds he less slopping si es (DCA2 igh -end); bo h wi h na ow niche wid hs (2 < 1.5). Fo es s 2021,12, 982 9 o 16 The heigh o he esponse (h), i.e., p obabili y o occu ence, was e y low along bo h coenoclines, pa icula ly o axonomic g oups (Table 1). Only Labia ae and E icaceae along DCA1, and Be ulaceae along DCA2 showed p obabili ies o occu ence abo e 7%. P obabili ies below 20% we e p edic ed o mos o he o he sh ub unc ional g oups along bo h coenoclines (Table 2). 4. Discussion O e all, ou indings sugges ha unde s o y sh ub main amilies and unc ional g oups espond o he same egional complex en i onmen al g adien s as ee species (i.e., al i ude empe a u e, and s eepness) since mos g oups showed unimodal ends o bo h g adien s (coenoclines). Howe e , he e we e some amilies (Figu e 2) and unc ional g oups (Figu e 3) ha showed inc easing and dec easing ends along he coenoclines, sugges ing ha he op imal niches o hese g oups a e ou o he leng h o hose g adien s. In any case, he desc ip ion o niche wid h along g adien s o mos impo an sh ub amilies and unc ional g oups is undamen al o wo ace s: i) o educe he complexi y o he successional p ocess enabling us o unde s and i be e and being able o use science- based managemen o biodi e si y and ecosys em se ices conse a ion, and ii) o de elop sus ainable o es p ac ices. Iden i ying such pa e ns in he ield is c ucial o ad ance ecological o es knowledge aking in o conside a ion no only he ee componen bu also sh ub di e si y and unc ion [51]. 4.1. Dis ibu ion Pa e ns o Main Sh ub Families along Coenoclines The dis ibu ion pa e n o he main sh ub amilies along he inc easing a idi y coen- ocline (DCA1) e lec ed he ecognized en i onmen al g adien om A lan ic empe a e a eas o he Medi e anean con inen al clima e o he Cas ilian pla eau [ 44 ]. The sh ub amilies dominance changes h ough he i s conenocline om he unde s o y sh ubby eg- e a ion belonging o Aqui oliaceae,Thymelaceae, and Be ulaceae amilies in he mo e A lan ic, cold and mois o es s o he no he n moun ains and alleys (DCA1 le -end; Figu e 2A ) o E icaceae,Rosaceae, and Cis aceae amilies in he less mois o es s o p e-moun ainous ange (DCA1 cen e), and he mo e heliophilous sh ubby unde s o y wi h Labia ae and As e aceae amilies cha ac e is ic o he mo e xe ic Medi e anean o es s on limes one and gypsum soils in he sou h (DCA1 igh -end). Howe e , he e we e wo amilies, Fabaceae and Cap i oliaceae, ha showed cons an co e alues along he main a idi y coenocline. I is well known ha Fabaceae g ow in e y di e en a idi y equi emen s, which a o s hei abili y o inhabi di e en ecosys ems [ 8 ]. Simila ly, Cap i oliaceae is a amily wi h gen- e a like Sambucus,Vibu num, and Lonice a widely dis ibu ed in he no he n hemisphe e adap ed o wide niche equi emen s [8]. The deciduous b oadlea o es s o A lan ic no he n a eas domina ed by Fagus syl- a ica and Que cus pe aea a e associa ed wi h an unde s o y o Aqui oliaceae,E icaceae, and Thymelaceae species such as Daphne lau eola L., E ica a bo ea L., Vaccinium my illus L. Howe e , in moun ain a eas he o es s o he mois es alleys a e composed o Rosaceae and Be ulaceae species mixed wi h So bus aucupa ia L. and Ilex aqui olium L. unde Be ula pubescens Eh h. As he a idi y g adien inc eased (i.e., mo ing sou hwa ds), mo e scle o- phyllous ee species appea ed like Que cus py enaica, which has an unde s o y domina ed by E icaceae (e.g., Calluna ulga is (L.) Hull, Daboecia can ab ica (Huds.) K.Koch, E ica cine ea L., E ica scopa ia L.) ha cons i u e he subs i u ion sh ubby unde s o y o mo e ansi ional- humid woodlands. In u n, coni e woodlands (Pinus syl es is,Pinus uncina a Mill. ex Mi b.) in he cold and ainy moun ains o he no h, sh ub-unde s o y is domina ed by E icaceae amily wi h E ica a bo ea a bo ea as he main species. Finally, mo e xe ophy ic and he mophilic sh ub amilies eplaced his unde s o y when mo ing sou hwa ds, such as Labia ae and As e aceae (e.g., La andula spp. and Helich ysum s oechas (L.) Moench) domina - ing unde Que cus ilex subsp. ballo a,Que cus aginea, and Junipe us hu i e a o es s, whe eas in he limes one moo s o he sou h, unde Pinus halepensis,Pinus pinea, and Cup essuss sempe i ens plan a ions, heliophyllous and summe d ough sh ub amilies domina e, Fo es s 2021,12, 982 16 o 16 63. Ga cía-Du o, J.; Muñoz, A.; Ál a ez, R.; Pesquei a, X.M.; Ri as, M.; Reyes, O.; Casal, M. Papel de los g adien es ambien ales en la egene ación ege al as incendio en ecosis emas de ma o al de Galicia. In Ac as del 5 ◦ Cong eso Fo es al Español. Mon es y Sociedad: Sabe quéhace ; Sociedad Española de Ciencias Fo es ales: Pon e ed a, Spain, 2009. 64. Cal o, L.; Tá ega, R.; Luis, E. Regene a ion pa e ns in a Calluna ulga is hea hland in he Can ab ian moun ains (NW Spain): E ec s o bu ning, cu ing and ploughing. Ac a Oecol. 2002,23, 81–90. [C ossRe ] 65. Ga cía, D.; Ma ínez, D. Species ichness ma e s o he quali y o ecosys em se ices: A es using seed dispe sal by ugi o ous bi ds. P oc. Royal Soc. B. 2012,279, 3106–3113. [C ossRe ]