Absence of geographical structure of morphological variation in Juniperus oxycedrus L. subsp. oxycedrus in the Balkan Peninsula

Absence of geographical structure of morphological variation in Juniperus oxycedrus L. subsp. oxycedrus in the Balkan Peninsula
of 15
All materials on our website are shared by users. If you have any questions about copyright issues, please report us to resolve them. We are always happy to assist you.
Related Documents
  ORIGINAL PAPER Absence of geographical structure of morphologicalvariation in  Juniperus oxycedrus  L. subsp.  oxycedrus in the Balkan Peninsula Robert Brus  • Dalibor Ballian  • Peter Zhelev  • Marija Pandz ˇa  • Martin Bobinac  • Jane Acevski  • Yannis Raftoyannis  • Kristjan Jarni Received: 26 March 2010/Revised: 20 October 2010/Accepted: 25 November 2010/Published online: 21 December 2010   Springer-Verlag 2010 Abstract  We examined leaf and mature seed conevariation of   Juniperus oxycedrus  L. subsp.  oxycedrus  in12 natural populations across the species range in theBalkan Peninsula. We measured 10 morphological traitsfrom a minimum of 100 leaves in each of 190 individuals,and two morphological traits from 30–50 seed cones ineach of 94 females. High phenotypic variation was found,but no geographical structure or cline across populationswas detected for any of the studied traits. Mean values of comparable leaf and cone morphological traits did notdiffer considerably from values reported elsewhere.Gender dimorphism in leaf morphology was detected, butit was not distributed uniformly throughout the studiedarea. An ANOVA model with both nested and crossedeffects revealed that the largest proportion of the totalvariation was, as expected, contained within populations,partly as among-tree variation (18–47%, depending on thetrait) and partly as within-tree variation (33–77%), whichwas remarkably high. Gender dimorphism explained only0–3% of the total variation. Differences among popula-tions (2–23%) were significant for all studied traits exceptone; however, PCA showed no clear geographical dif-ferentiation of the studied populations. This lack of phylogeographical structure may be the consequence of repeatedly occurring colonisation-retreat scenarios andsuggests the existence of several small refugial popula-tions scattered over a large part of the Balkan Peninsulain the Pleistocene. Further research including palaeobo-tanical and molecular genetic studies will be needed tobetter understand the forces that shaped current variationpatterns of   J. oxycedrus  L. subsp.  oxycedrus  in the BalkanPeninsula. Keywords  Phenotypic variation    Plant variation    Plantmorphology    Biometry    Sexual dimorphism   Geographical differentiation    Pleistocene refugia Communicated by C. Ammer.R. Brus ( & )    K. JarniBiotechnical Faculty, Department of Forestry and RenewableForest Resources, University of Ljubljana, Vecˇna pot 83,1000 Ljubljana, Sloveniae-mail: robert.brus@bf.uni-lj.siD. BallianFaculty of Forestry, University of Sarajevo, Zagrebacˇka 20,71000 Sarajevo, Bosnia and HerzegovinaP. ZhelevDepartment of Dendrology, University of Forestry,10 Kliment Ochridsky Bulvd., 1756 Sofia, BulgariaM. PandzˇaPrimary School ‘Murterski sˇkoji’, Put sˇkole 8,22243 Murter, CroatiaM. BobinacFaculty of Forestry, University of Belgrade, Kneza Visˇeslava 1,11030 Belgrade, SerbiaJ. AcevskiFaculty of Forestry, University of Skopje ‘Cyril and Methodius’,Bul. Aleksandar Makedonski bb, 1000, Skopje,Republic of MacedoniaY. RaftoyannisDepartment of Forestry and Environmental Management,Technological and Educational Institute of Lamia,36100 Karpenisi, Greece  1 3 Eur J Forest Res (2011) 130:657–670DOI 10.1007/s10342-010-0457-1  Introduction  Juniperus oxycedrus  L. subsp.  oxycedrus  is considered atypical subspecies of Prickly Juniper (  J. oxycedrus  L.) andis distributed from Morocco, Algeria and Tunisia in northAfrica into Portugal, Spain, France, Italy, the BalkanPeninsula, Turkey and eastward to southern Caucasus andnorthern Iran (Franco J do 2002; Farjon 2005; Roloff and Ba¨rtels 2008). Throughout its distribution range, it is acommon element in Mediterranean sclerophyllous shrub-land, such as maquis and garrigue, and in dry woodlandwith termophyllous tree species, as well as in montane andmesic forest. It occurs on dry, stony slopes in thin soil onall kinds of parent rock but is rare on sand dunes; in Eur-ope, it can grow up to 2,300 m a.s.l. (Christensen 1997;Franco J do 2002; Farjon 2005). It is an evergreen, dioe- cious erect shrub or small tree up to 8(12) m tall, withspreading or ascending branches, leaves (6)8–15(25) mmlong and 1–1.5(2) mm wide and orange- to reddish brownseed cones that are (6)8–10(13) mm thick and normallycontain three seeds (Franco J do 2002; Schulz et al. 2003; Farjon 2005; Roloff and Ba¨rtels 2008).Beside  J. oxycedrus  L. subsp.  oxycedrus , another threesubspecies are usually recognised within the species:  J. oxycedrus  subsp.  macrocarpa  (Sibth. & Sm.) Ball,  J. oxycedrus  subsp.  badia  (H. Gay) Debeaux and  J. oxycedrus subsp.  transtagana  Franco (Franco J do 1963; Farjon 2001; Franco J do 2002; Farjon 2005). In addition, Adams et al. (2005) argue the existence of two cryptic, genetically dis-tinct but morphologically very similar species within  J. oxycedrus  L. subsp.  oxycedrus.  These species are believedto be largely allopatric;  J. oxycedrus  subsp.  oxycedrus  isonly distributed in the areas west of Italy, while therecently recognised new species,  J. deltoides  R. P. Adams,based on published data on DNA RAPDs, nrDNAsequence, morphology and terpenoids (Adams 2004), isthought to be only present from Italy eastward throughTurkey into the Caucasus Mts. and Iran. According to thisview, all  J. oxycedrus  L. subsp.  oxycedrus  populations inthe Balkan Peninsula would in fact be treated as  J. delto-ides  R. P. Adams.Much of the recent research within  J. oxycedrus  L.  sensulato  has so far been focused on  J. oxycedrus  L. subsp. macrocarpa  (Lewandowski et al. 1996; Cantos et al. 1998; Garcı´a-Berlanga and Serrano 2003; Klimko et al. 2004; Mun˜oz-Reinoso 2004; Redondo and Saavadera 2004; Sezik  et al. 2005; Juan et al. 2006; Massei et al. 2006), which is a rare and endangered species in several parts of its range.Within  J.oxycedrus L.subsp. oxycedrus ,besidetaxonomicalstudies(Adams2003,2004;Adamsetal.2005),researchhas dealtwithnewforms(Yaltiriketal.2007;AvciandZielin´ski2008), reproduction (Cantos et al. 1998; Ortiz et al. 1998; Arista et al. 2001), demographic dynamics (Biondi 1990; Baldoni et al. 2004), phytosociology (Cano et al. 2007), somatic embryogenesis (Gomez and Segura 1996) orchemicalcompositionoftheplant(MilosandRadonic2000;Koukosetal.2002;Valentinietal.2003;Loizzoetal.2007). Only a few studies addressed the morphologic variation of   J. oxycedrus  L. subsp.  oxycedrus . Lebreton et al. (1991)found the seed cones to be discriminant between subsp. oxycedrus  and subsp.  macrocarpa,  while the mean numberof seeds per cone was not. More recently, Klimko et al.(2007) performed a biometric examination, based on mor-phological characters of the leaves, cones and seeds, of 13  J.oxycedrus L.subsp. oxycedrus populationsfromEast-andWest-Mediterranean regions and reported significant dif-ferences between both groups of populations.In the Balkan Peninsula in particular,  J. oxycedrus L. subsp.  oxycedrus  has been poorly studied. Most of theearlier-mentioned research on this taxon was performed onthe material from either West- or East-Mediterranean(including Greek) populations with the exception of thestudy of Klimko et al. (2007), which included one popu-lation from Bosnia and Herzegovina and another fromCroatia. When defining a new  J. deltoides  species, Adams(2004) only cites one representative specimen from theBalkan Peninsula outside Greece; this is a herbariumspecimen collected by W. B. Turill in 1922 near Porecˇ,Istria, Croatia. Alexandrov et al. (1993) described a newlocality of   J. oxycedrus  in Bulgaria, Milos and Radonic(2000) analysed its volatile compounds in Croatia, andKoukos et al. (2002) examined the chemical compositionof the berry oil in Greece.In flowering plants, sexual dimorphism is often observednot only in sexual organs but also in traits that are notdirectly related to reproduction. Species can be sexuallydimorphic in resource acquisition and allocation, in inter-actions with other community members, and in size, colouror longevity of vegetative structures (Dawson and Geber1999; Delph 1999). Differences in leaf size between gen- ders have been reported for several species with femalesoften having larger and/or longer leaves than males (Wallaceand Rundel 1979; Bond and Midgley 1988; Mitchell 1998; Delph et al. 2002; Iszkuło et al. 2009). For  J. communis, sexual dimorphism was identified in lifetime performance(Ward 2007), in water use in drier sites (Hill et al. 1996), and males were less grazed than females (McGowan et al.2004). Males were taller than females in  J. virginiana (Vasiliauskas and Aarssen 1992), while females of   J. thu-rifera  appeared to be taller than males but had lower radialgrowth (Gauquelin et al. 2002). Considering these differ-ences, sexual dimorphism might be expected in  Juniperusoxycedrus  L. as well. So far, no sex-related leaf dimorphismin any  Juniperus  species has been studied or reported.Until now, phenotypic variation of   J. oxycedrus  L. subsp. oxycedrus  in the Balkan Peninsula has only been studied 658 Eur J Forest Res (2011) 130:657–670  1 3  briefly in several studies; however, there has been no inte-gral research that systematically analysed phenotypic vari-ation across different factors (i.e. population, individual andgender) and geographical differentiation of this species inthe entire region. Also, a hypothesis on the existence of   J. deltoides  R. P. Adams in the eastern Mediterranean hasnever been tested on a large set of samples, and sex-relatedleaf dimorphism in any  Juniperus  species and  J. oxycedrus subsp.  oxycedrus  in particular has not been studied so far.To answer these questions, we performed a detailed analysisof leaf and seed cone morphology of   Juniperus oxycedrus L. subsp.  oxycedrus  in the entire Balkan Peninsula. Methods Study areaIn the NW part of the Balkan Peninsula,  J. oxycedrus L. subsp.  oxycedrus  is mainly distributed along the Adriaticcoast and in the adjacent areas. In Slovenia, it is a very rarespecies only found in isolated locations on steep slopes inlow altitudes in the Mediterranean region (Brus 2008).Further to the south, it is one of the most common speciesof various termophyllous plant communities of the Cro-atian regions Istria, Kvarner and Dalmatia, and in thecoastal part of Montenegro and Albania; however, it canalso be found more inland either continuously in sub-mediterranean forests or in isolated locations in more thecontinental part of Croatia, Bosnia and Herzegovina, andMontenegro (Vidakovic´ and Franjic´ 2004; Sˇilic´ 2005). InGreece, it is widespread on the mainland and some of thelarger islands except the Ionian Islands and Cyclades and isgenerally found on hillslopes at 200–1,300 m, sometimesascending to sub-alpine levels (Christensen 1997). In thecentral part of the Balkan Peninsula with a mildly conti-nental climate, it is relatively common in Macedonia and inisolated locations in central, western and southwesternSerbia (Omanovic´ 1938; Jovanovic´ et al. 1997; Jovanovic´2000; Dinic´ et al. 2006). It is frequent and abundant on thedry sunny slopes of the mountains in Southern Bulgaria,usually at lower altitudes (Alexandrov et al. 1993).SamplingTwelve populations of   J. oxycedrus  L. subsp.  oxycedrus were sampled in 2007 across the entire species’ range inthe Balkan Peninsula (Table 1; Fig. 1). In each population, 14–17 adult individuals (at least 7 males and 7 females)were selected. The sample size was based on the number of individuals that assures the determination of statisticallysignificant differences among populations and betweengenders within those and is comparable with the number of individuals included in similar studies (e.g., Klimko et al.2007). The sex of each individual was determined by thepresence of either seed cones or male strobili or their res-idue on the branches at the time of sampling. The distancebetween selected individuals was at least 25 m. From eachindividual, a normally developed branch, approx. 35 cmlong, was taken from the outer, sunny portion of the south,south-east or south-west facing part of the crown at aheight of 1–2 m above the ground. In the laboratory, one-year-old leaves were collected from yearly increments of atleast 10 long shoots of the same branch, which were chosenrandomly. From the central part of each shoot, approxi-mately 10 leaves were collected randomly. Altogether, aminimum of 100 leaves were sampled from each individ-ual. From each sampled female tree, 30–50 ripe seed coneswere collected from the same part of the crown.For the morphological analysis, the length and diameter(average of two measurements at an angle of 90  ) of freshcones were measured to the nearest 0.1 mm. The freshleaves were scanned and measured with the software‘‘WinFOLIA Pro 2005’’ of Regent Instruments Inc. Thescanning accuracy was set to 0.03 mm. For each leaf, 10morphological traits were measured or calculated (Table 2;Fig. 2). In total, 190 individuals, 19,908 leaves and 4,421cones were analysed.Statistical analysisStandard formulae of descriptive and multivariate statistics(Sokal and Rohlf  1989; McGarigal et al. 2000) were used to obtain an objective picture of the morphological parame-ters. Descriptive statistics, such as arithmetic means, stan-dard deviations and coefficients of variation were used todescribe the main features of morphological variation. Wefound no noticeable deviation from normality for anycharacter studied. Analysis of variance was used to deter-mine the differences among populations, between differentgenders and among trees in a single population. For thispurpose, a nested analysis of variance was designed wheretrees were nested within populations and gender. The nesteddesign was carried out with the model: Y   ¼  P  þ  G  þ  P    G  þ  T ð PG Þ þ  e The model tests the main effects of population (P) andgender (G), the interaction between them (P 9  G), and thenested effects of trees within a population and gender T(PG)on all the measured traits of leaves and cones. The model hadtwofixed(crossed)factors(PandG)andonerandom(nested)effect (T). The contribution of a hierarchical level to totalvariance was presented as a share of total variance. Principalcomponent analysis (PCA) was used to determine thedifferences among the populations .  The aim of the analysiswas to combine the original variables into independent Eur J Forest Res (2011) 130:657–670 659  1 3  synthetic variables thatexplained the greatest partof the totalvariation observed among the populations. In the method, avarimax rotation was used. The units that were discriminatedwere the populations. All computations were performed withSPSS Statistics 17.0 and Statistica for Windows software. Results Most of the studied leaf and cone morphological traitsvaried greatly across the geographical range representedin the study (Table 3). In general, the differences amongpopulations were of similar magnitude. However, popu-lation GRE differed distinctly from all other populationsin several traits, such as L, A, W80, W/L and coef. Thedistance from the lamina’s base to the point of maximalwidth of the leaf (dW) was by far the most variable trait,with coefficients of variation ranging from 36.00 to46.88% for particular populations and 42.64% on aver-age for all samples together. The geographically inter-mediate and large population BIH 2 was the mostvariable for the largest number of studied traits (L, A,W/L, coef), while the least variable populations wereMNE and SRB, the latter being situated at one geo-graphical extreme and isolated. The least variable traitswere cone height (H), leaf width (W) and cone diameter(D). Coefficients of variation among populations forparticular morphological traits were generally of a sim-ilar order of magnitude. However, two populations, SLOand CRO 1, showed particularly large deviation fromthis trend in seed cone diameter (D), with very highcoefficients of variation. In general, neither studiedmorphological traits nor coefficients of variation showedany clear patterns of change across the geographicalgradient (Table 3).Gender dimorphism was established for all analysed leaf morphological traits. A comparison of the leaves of maleand female sub-populations revealed significant differencesin particular traits in most populations (Table 4). The dif-ferences were generally very small. In some morphologicaltraits, however, the difference between genders was larger,and as shown in Table 5, statistically significant within thewhole sample. In the population CRO 2, dW (position of the widest point of the leaf) of females was 0.224, while Table 1  Main characteristics of the studied populationsPop. No. Acronym Pop. name Country Sample size Coordinates Altitude (m)M F Tot.1 SLO Kosˇtabona, near Koper Slovenia 8 8 16 45   29 0 03 00 N 25013   43 0 42 00 E2 CRO 1 Jablanac Croatia 7 7 14 44   43 0 05 00 N 22014   55 0 07 00 E3 CRO 2 Tisno Croatia 10 7 17 43   48 0 21 00 N 1015   39 0 56 00 E4 BIH 1 Kravica, near Ljubusˇki, Herzegovina Bosnia and Herzegovina 8 8 16 43   12 0 23 00 N 19017   34 0 03 00 E5 BIH 2 Hutovo, near Ljubinje, Herzegovina Bosnia and Herzegovina 8 8 16 42   57 0 24 00 N 32517   48 0 37 00 E6 MNE Vladimir, near Ulcinj Montenegro 8 8 16 41   57 0 49 00 N 13519   16 0 44 00 E7 SRB Gocˇ, near Kraljevo Serbia 8 8 16 43   33 0 25 00 N 73020   40 0 13 00 E8 MAC 1 Jabolci, near Skopje Macedonia 8 8 16 41   54 0 23 00 N 62021   19 0 50 00 E9 MAC 2 Trpejca, Lake Ohrid, near Mt. Galicˇica Macedonia 8 8 16 40   56 0 59 00 N 1,00020   46 0 51 00 E10 GRE Karpenisi Greece 8 8 16 38   55 0 32 00 N 1,28021   47 0 39 00 E11 BUL 1 Goleshevo Bulgaria 7 8 15 41   26 0 34 00 N 70023   35 0 01 00 E12 BUL 2 Yasenovo, near Kazanlak, Stara Planina Bulgaria 8 8 16 42   41 0 30 00 N 55025   14 0 34 00 E660 Eur J Forest Res (2011) 130:657–670  1 3  dW of males was only 0.167. Likewise, in most otherpopulations the widest point of the leaf was positionedcloser to the leaf base in males than in females, but in onepopulation (MAC 2) the situation was reversed. In othermorphological traits, even more populations expresseddeviation from the general pattern: in leaf length (L) andleaf area (A) these were BIH 1 and BUL 2, while in othertraits the number was even higher. However, in none of thetraits were the male–female differences distributed in thesame way uniformly throughout all populations. Namely,in some populations a higher value of a particular trait wasfound in the male sub-population, while in other popula-tions it was found in the female sub-population. Forexample, in 9 populations the leaves of females were Fig. 1  Geographical distribution of the studied populations. The  dark   shaded area represents the distribution range of   Juniperus oxycedrus L. subsp.  oxycedrus  (after Jalas and Suominen 1973) Table 2  Measured orcalculated morphological traits Abbr. Morphological traitL Lamina lengthW Lamina maximal widthA Leaf areaW10 Width of lamina on 10% of lamina’s length from lamina’s base upW25 Width of lamina on 25% of lamina’s length from lamina’s base upW80 Width of lamina on 80% of lamina’s length from lamina’s base upW90 Width of lamina on 90% of lamina’s length from lamina’s base upW/L Lamina width/length ratiocoef 4pi*A/P 2 (P  =  leaf perimeter)dW Distance from the lamina’s base to the point of maximal widthH Height of seed coneD Diameter of seed coneEur J Forest Res (2011) 130:657–670 661  1 3
Similar documents
View more...
Related Search
We Need Your Support
Thank you for visiting our website and your interest in our free products and services. We are nonprofit website to share and download documents. To the running of this website, we need your help to support us.

Thanks to everyone for your continued support.

No, Thanks