El género Juniperus (Cupressaceae) en Mexico y Guatemala: análisis numérico y morfológico

THE GENUS JUNIPERUS (CUPRESSACEAE) IN MEXICO AND GUATEMALA: NUMERICAL AND MORPHOLOGICAL ANALYSIS

Thomas A. Zanoni ancl Robert P. Adams·:>

The genus ]uniperus ( Cupressaceae) consists of approximately 60 species distributed in the Northern Hemisphere, except for ]. procera Hoch. which extencls in to the Southern Hemisphere in Africa (Florin, 1963). The delimitation of the species of ] uniperus is difficult in all of the world; few taxonomic studies have ~ncompassed any large geographical or natural province. Gaussen (1968) lists 99 species in the most recent treatment of the iunipers of the world; however, this is an inflated number· since many of the species he recognizes had been reduced to synonymy by earlier authors.

]nniperus has been divided in to• three sections: Caryocedrus, Oxycedrus, and Sabina (Endlicher, 1841; Gaussen, 1968). Caryocedrus, a monorypic section, is restricted to the eastern l\Icditerranean basin; the other two sections are dis-tributed throughout the Northern Hemisphere. Section Oxycedrus is representecl i11 Norlh America by the circumboreal species ]. communis L. The other junipers (approxirnately 22 species) in No1th America belong to1 the section Sabina, which is characterizecl by opposite or ternate, scale-like aclulr leaves, peltate scales in male eones, and fcmale eones terminal on fertile twigs (peduncles). Engelrnann ( 1877) rnade thc {i rst stucly of the No1th American (mostly Unitecl Sta tes) j unipers.

The genus ]uniperus has receivecl considerable attention from Nortlt Ame -rican taxonomists in the past '.15 years. Fassett's (1944a, 1944b, 1945a, 1945b, lC)iJ.5c) ancl Hall's (1952a, 1952b, 1955, 1962, 1964, 1968) series for mo• rpho-logical studies concerning hybridization in severa! species of ]uniperus renewed intercst in this gcnus. More recent studies have used more sophisticatecl tech-niques for data collcction and analy5is. Va::ek (1966) ancl Vasek ancl Seora

(1967) examined thrce 11·esrern American species morphologically and chemi-cally. Van Haverbeke (1968), Flake, von H.udloff ancl Turner (1969, 1973), Adams ancl Turner (1970), Schurtz (1971), Adams (1969, 1970, 1972, 1973, * Department of Rotany allll Plant Pathology, Colorado State University, Fort Collins,

Colurado, 80523, U.S.A.

69

Boletín de la Sociedad Botánica de México35: 69-92, 1975 DOI: 10.17129/botsci.1154

BOLETJN DE LA SOCIEDAD BOTANICA DE MEXICO No. 35, 1975

1974a), and Powell and Adams ( 1973) imestigated evolutionary and systematic problems using sophisticated techniques in statistics ancl numerical taxonomy with several of these studies employing chemical data.

The junipers of North America extend southward into Mexico and Guate· mala, the southernmost localities being in the Guatemalan mountains. Standley (1920) presented the first evaluation of the Mexican junipers in this century ancl reco¡mized four species by studying herbarium specimcns. In 1943, Stanclley ancl Steyermark (1943) publishcd a previously undescribed species from Mexico and Guatemala. A total of nine species were known from Mexico, and one from Guatemala by 1944. Martínez (1944, 1946) critically examined the Mexican junipers and recognizecl 21 taxa, severa] of which hacl not been previously recognizccl.

Hall's (1954) review of ]uniperus ashei Buchholz reaffirmed the citation by Johnston (1943) that the species was in Mexico. Martínez (1946) hacl rejectecl Johnston's decision. Standley and Steyermark (1958) documentecl the presence of two species in Guatemala, which also occur in Mexico. ]. saltillensis was published as a new species in Mexico by Hall ( 1971).

Martínez (1946) reco?;nizecl ] 2 species, 6 varieties, and 3 forms (Fig. 1) of junipers from Mexico. Thcse taxa were basecl on fielcl observations, herbarium stuclies and literature review. It is doubtful that Martínez observed ali of the taxa in the fielcl, although he had placed his narne and collection numbers on many of the specimens he distributed (McVaugh, )972). The type specimens of the taxa established by Martínez are distinctly different from one another. Martínez' leve] of unclerslanding of the variation of the taxa in nature is not known.

Tlw ;:ection Sabina junipers o[ Mexico were clividecl by Martínez into five subsections of two or more taxa within each subscction. In a diagram (Fig. 1), the suhsections were interconnectecl by lines that appcar to inclicate rrlationship;:; howcver, tbe arrangement of the subsections in a pentagon appears to have no phylogenctic implications. Martínez (J <)4.6, 1053, ] 963) nevcr commentccl 011 tite interrelationships of the subsections.

The placement of the taxa into thc subsectic·ns implies relationships of the taxa within suhsections as indicated by Martíncz (Fig. ] ) . The ~ub~ection Monosperrnae consists of six laxa. ]uniperus comitarw, ]. monosperma var. 1no· nosperma, ]. monosperma var. gracilis, ]. c1")·throcarpa var. coahu.ilensis were apparc11tly thought to be closely relatecl to each othcr as inclicatecl by the lines clrawn by Martínez. f. californica and J. gamboa.ria are relatecl to thc other taxa of this subsection, but Martínez clid not inclicatc the a[rinities of thc~c two

JUNJPERUS IN MEXICO AND GUATEMALA

species. Subsection Deppeanae includes four varieties of

J.

deppeana and two forrns of f. patoniana. Martínez indicates reticulate relationships among these taxa.

The subsection Flaccidac consists of two varieties of f uniperus flaccida. J. blancoi and f. jaliscana are placed in subsection Jaliscanae. Subsection Monti-colae includes !. durangensis,

!.

standleyi, and three forms of f. monlicola. Heticulate relationships are indicrated by Martínez for this subsection.

The separation of the Mexican junipers into subsections was intended to reflect natural groupings. The use of reticulate patterns to show relationships of taxa within subsections suggests that the relationships were not readily ap-parent to Martínez.

The recen ti y described f uniperus saltillensis (Hall, 1971) was not considered by Martínez in his taxonomic treatment of the Mexican iunipers. Hall (1971) stated that this species is related to f. mo.nticola, !. standleyi, f. paliscana, and

J.

blancoi.

Although the Middle American junipers were recently studied by Martínez

( l 044, 1946), a reexamination of ,the morphology is needed for a comprehensive study of these junipers. A second phase of this study will include an examination of the terpenvids of the foliage. In addition, this study will provide information for the evaluation of the relationships of species in the genus f uniperus in North America.

MATERIALS AND METHODS

The specimens were selected to reprcsent the taxa known from Mexico and

Guatemala and to represent e~ch taxon from se\'cral localities, if possible (Fig.

2). Three to eight branches frorn female plants were selected from severa] lrccs at each locality in 1970 and 1972. Several herbarium specimens of the

'ame collector and collection number were used for !uniperns bumcoi and f. paloniana; few specimens of these two taxa were available.

Twcnty·four operational taxonomic units (Sokal and Sneatli, 1963) were

analyzecl. Ninetcen of the :24 operationa] taxo,nomic units (OTU's) represent the laxa known from Mexico ancl Guatemala. The five other OTU's represen! populations of uncertain affinities.

Female plants selectecl for analysis incluclecl the female eones (megastrobili or galbuli). Studies by Lemoine-Sebastian ( J 967) inclicatecl that the male eones (microstrobili) were not useful in cletecting specics differences in funiperus section Sabina.

BOLETIN DE LA SOCIEDAD BOTANICA DE MEXICO No. 35, 1975

For various reasons three species were not included in this study. f uniperus ashei is not included in this study; this species is known from a few localities and is uncommon in México. Adams and Tuner (1970) studied this species in detail and work is continuing (Adams, 1974.a). funiperus califomica and f uni-perus scopulorum were also r;ot included in this study. These species have their centers of distribution in the United States and will be studied at a later phase in The Junipers of North America Project. Distribution information concerning Juniperus calijomica and funipcrns scopulorum in Mexico have been recently reviewed by Zanoni and Adams (1973, 1974•).

Morphological characters. The selection of morphological characters was based on an examination of the literature and specimens. Earlier studies (Hall, 1952a, 1952b, 1955; Hall, McCormick and Fogg, 1962; Hall and Carr, 1964, 1968; van Haverbeke, 1968; Adams and Turner, 1970; Adams, 1972) included up to 29 morphological characters. Sixty-four characters (Tablei 1) were utilized in this study to provide a broader sampling of characters for analysis.

Peduncles, female CO!lles, seeds, terminal whip leaves, scale leaves, terminal branches and stem bark were examined and scored (Table 1). Ten measurements were taken from each plant or herbarium specimen whenever possible (except as noted in Table 1). Each linear measurement was taken using a reticel in a binocular dissecting microscope. Female cone length and width were measured with a millimeter ruler. Cone and seed colors were determined by comparison to the Munsell Color System charts. The !me, value, and chroma were recorded and analyzed as three separate characters. Cone pulp, cone bloom, seed and leaf shapes, gland shapes, leaf margins, and leaf and gland protrusions were scored by comparison to character states established by a preliminary survey of the specimens studied. Leaf tips were measured for the width at .66 mm from the tip of the whip leaves and at .33 mm from the tip on scale leaves to obtain an estimate of acuteness of the leaf tip. The angle of branching of the ultimate twig was measured with a protractor.

Stem bark exfoliation pattems were scored in the field when the specimens were collected. The bark descriptions of Martínez ( 1963) were used for the specimens distributed by Martínez.

Ali characters that were not present or not determinable from the specimens examined were scored with -1.0, indicating that no comparison was to be rnade for rhat charactcr on that tree in the data analysis.

Data analyses. Analysis o[ variance (ANOVA) was performecl on each of the 64 characters to detect (by use o( thc F ratio o[ tlic variance among OTU's/ variance within OTU's) which charactcrs exhibited statistically significant dif-ferences arnong OTU's. A modified Student-Newman-Keuls (SNK) multiple

ran-JUNIPERUS IN MEXICO AND GUATEMALA

ge test (Adams, 1970) for unequal OTU samples was used to analyze each character to detect which OTU means were significantly different at the .05 level. A weighted mean character difference (M. C. D.) similarity measure (Adams, 1970; Aclams, 1974b) was used to determine the similarities among the 24 OTU's. Three separate analyses were made with the data derived from the ANOV A and SNK multiple range tests. The first analysis consisted of 63 cha-racters and used yF - 1 as the weight of each character comparison; the

second used 45 characters and \/F - 1 character weighting; the third used 45 characters and F - 1 character weighting. The 45 characters were selected from the list of the 63 characters to eliminate many characters that were thought to be easily influenced by environmental differences or characters thought to be

subjectively scored during data collection.

The single linkage method of Sneath (1957) was used for clustering. RESULTS. Analysis of variance revealed that fifty-eight of the 64. characters had significant F ratios (P = .05). Two characters, branch flaccidness (FLC) and scale leaf tip clivergence. (LFD) had infinite F ratios because these charac-ters hacl no variation within OTU's. Fernale cone surface texture (FRS), hilum

scars per seecl ( HN O) , scale lea ves per 3 .33 mm ( RLO) , and scale leaf rupture (RRP) had non-significant F ratios. The SNK tests were run on the 62 characters (excluding FLC) and (LFD) to determine which OTU means were significantly different. Fifty-eight of these 62 characters ( excluding FLC and LFD) had significant SNK tests. The characters FRS, HNO, RLO, and RRP did not have

significant SNK tests.

The characlers of branch ílaccidness (FLC) and scale leaf tip divergence (LFD) were assigned F ratios equivalent lo the F ratio of cone pulp (RFP) for inclusion of these characters in the computation of similarity measures. The

F ratio of cone pulp was selected because the rneristic characters branch flac-cidness (FLC) and scale leaf tip divergence (LFD) exhibited variance within and among OTU's similar to that exhibited in cone pulp (FRP). Although no SNK tests were cornputed for FLC and LFD, these characters separate the flac-cidan complex from ali other OTU's examined in this study.

Before computation of the similarity measures, one character, hilum scars per seed (HNO) with a F ratio less than 1.0, was discarcled from the analysis. Cl ustering ( Fig. 3), using 63 characters and \/F - 1 character weights, illustrates severa! groupings of OTU's. The flaccidan complex,

f

uniperus flac-cida var. flaccida (FL) and /. flaccida var. poblana (FP), forms a quite distinct group. Tlie monlicolan complcx of !. nwnúcola f. monlicola (MM),!. nwnticola

BOLETIN DE LA SOCIEDAD BOTANICA DE MEXICO No. 35, 1975

deppeanan complex of J. deppeuna var. deppeana (DD), f. deppeana var.

ro-busta (DR), J. deppeana var. zacatecensis (DZ), the OTU DZA, and f. pato-niana (DP) is also fairly distinct. The one-seeded cornplex of J. gamboana (GA)

f. saltillensis (SL),

!.

comitana (CO),

!

.

erythrocarpa var. CO'ahuilensis (ERW

and EW),

!.

monosperma var. gracilis (MG), and the OTU's LLR, MLB and

MLT is apparent. The

!

.

monosperma var. monosperma from USA (MS), J. durangensis (DU), f. standleyi (ST), J.jaliscana (JA), and J. blancoi (BL)

cluster rather loosely and do not enter into these complexes.

Three of the OTU's rhat were of uncertain identities were quite similar to

other OTU's. DZA,

f

uniperus deppeana from El Alamo, Zacatecas, is most

simi-lar to

f.

deppeana var. robusta. This OTU (DZA) was collected near J. deppeana

var. zacatecensis which indicates that these taxa are sympatric in the state of

Zacatecas.

Se1·eral incongruities exist between the expected clustering pattern based

on the SNK tests interpretations and field observations, and the clustering

pat-terns using 63 characters with ·vF-=-· 1· character weights. f uniperus durangensis

(DU) and

!.

standleyi (ST) were expected to show greater similarity to the

montic0>lan complex of the

J.

monticola forms. f. monospemui var. monosperma

from USA (MS) was expected to be more similar to the one-seeded complex.

f.

errthrocarpa var. coahuilensis frorn western Mexico (EW) was expected to

be more similar to the OTU of the same taxon from eastern Mexico (ERW). The character list was examined and 18 characters were recognized as

pos-sibly being variable dueto envi ron mental influences or being subjectively scored.

Fifteen of these 18 charactcrs were removed because they were likely to be

easily influenced by the cnvironment. These characters include: seed length X

width (SEV), hilum scar length (HIL), whip leaf length (WLL), whip leaf

width (WLW), whip leaf length X width (WLV), number of whip leaves per

6.6 mm. (WLO), whip leaf gland length X width (WGV), length from whip

]Paf gland center to lcaf tip (WCT), scale leaf length (RLL), scale leaf width

!RLW), scale leaf length X width (RL V), number of scale Ieaves per 3.3 mm

(IU.0), scale leaf gland lcngth X wiclth (RGV), length from scale leaf gland

centcr to leaf tip (RCT), ancl diameter of the ultima te twig (DIA). The follorwing

three characters were removed frorn the1 _{elata analysis because they were believed} to he sub jectively seo red: cone surface texture (FRS), number of scale lea ves

per node (RLN), and stcrn bark exfoliation pattern (BRK). Although the

num-ber of scale leaves per noclc has frequently been used to charactcrize species of junipers, it appears that the selection of certain twip:s can bias the rcsults in

that binate ancl ternate lea[ arrangcments can orftcn he observecl on the same

JUN!PERUS IN MEXIl.O AND GUATEMALA

Figure 4, illustrates thc similarities oJ OTU's based on the remammg 45 cliaraclcrs with

ylf.----=--1

character weights. This recomputation o.f similarity • measures and the phenogram resulted in resolution of few of the incogruities observed in the similarity rneasures based on 63 characters and

vv=-1

cha-racter weights. Thc f uniperus monosperma var. monosperma from USA (MS) has a greater sirnilarity to lhe one-seeded OTU's. The positions of f. standleyi (ST) and

f.

durangensis (DU) are not similar to those expected from the SNK te~ts or from observations o.f specimens.

A hcavicr weighting \F - 1) was used as suggested by the recent work

of Adam ( 1975). Figure

5

illustrates the similarities of OTU's based on 45

characters with F - 1 character weights. The flaccidan, deppeanan, monticolan and monosperman complexes are still apparent as seen in Figures 3 and 4.

In

conlrast to Figures 3 and 4,

f

uniperus durangensis (DU) is more similar to the

monticolan cornplex but the clustering with

f.

monticola f. orizabensis (MO) was not cxpected.

!.

standleyi ~ST) is also more similar to the monticolan

com-plcx. J. monospenna var. monosperma from USA (MS) shows close affinities with the one-seeded cornplex. ]. erythrocarpa var. coahuilensis from eastern Mexico (ERW) is rnost similar to the plants of the same taxon in western Mexico (EW), although

EW

is most similar to LLR and

f.

monosperma var. gracilis C\IC) !

'.::ieveral groups are apparent in each of the three phenograms. The flaccidan

rnmplex of funiperus flaccida var. flaccida (FL) and f. flaccida var. poblana

¡ ¡:p) is quite clistinct from the orher OTU's. The deppeana cornplex of

f.

dep-¡11'1m11 ,·;:ir. deppeana (DD), J. deppeana var. robusta (DR), and

J.

deppeana var. :;acatecensis (DZ) remains a distinct group. The OTU DZA is most similar

lo f. deppeana var. robusta (DR). f. patoniana (DP) is most similar to

f.

dep-peana var. robusta ( DR) and clusters with the

f.

deppeana varieties.

Another distinct group is the monticolan complex, which is composed of f uniperus monticola f. monlicola (MM), f. monticola f. campacta (MC),

f.

monticola f. orizal;ensis (l\10).

f.

durangensis (DU) and

f.

standleyi (ST) have

affi11 itics lo the monticolan complex, but these affinities are not consistent in

lhc phenograms. Thc one-sceded complex o.f f. gamboana (CA), J. comitana ( CO), f. saltillensis (SL),

!.

crythrocarpa var. coahuilensis (ERW and EW), f. monosperma va r. gracilis (MG) form a group. The affinities of

!.

jaliscana (JAJ ancl J. blancoi (EL) are not clear from any of the data analyses. The r '>111¡;lr'scs of t~ixa rrcognized in these elata analyscs are similar to those of Mar-lír1rz ( Fig. 1) with various distinguishing characteristics shown in Table 2.

BOLETIN DE LA SOCIEDAD BOTANICA DE MEXICO No. 35, 1975

DISCUSSION AND CONCLUSIONS

This study of the Mexican and Guatemalan junipers revealed severa! new

relationships not appa rent in the studies by Martínez. !imiperus f laccida ( var.

flaccida and var. poblana) has lar ge, many-seeded eones, large scale leaves

with divergent leaf tips, and pendulant branches (see Table 2). The origin

and affinities of this specics are not known. f. jlaccida is most common in thc

hills and low mountains (800. 2 900 m.) of southern Mexico, where it apparently

evolved into the two varieties. The species is not found at the Isthmus of Tehu

:rn-tepec or in localities southeast of the isthmus. This arid, low-altitude (250 m)

region has been a barrier to the southern extension of the ranges of many

plants sin ce the late-Tertiary (Steyermark, 1950), as has been the case in tlie

junipers. The northernmost locality is in northeastern Sonora.

funipcrus deppeana, another spccies with largP, sevC'ral-seeded eones, bears

rnperficial resemblance to ]. flaccida. Martínez recognized four varieties o[

f.

deppeana. Three o[ these varieties;

f.

deppeana var. deppeana,

f.

deppeana var.

robusta, and ]. deppeana var. zacatecensis, wcre examined in this study. Thcse

varieties are distinct from one another. The similaritics of the OTU DZA and

f.

patoniana (DP) to the varieties of ]. deppeana are oí special interest. DZA

represents a population collected at El Alamo, Zacatecas. The idcntity of this

OTU was in question because thc population was located within the distribution

range off. deppeana v:H. zacatecensis, but thc morphoJogy appearecJ to be cJj Ífe

-rent from members of that variety. The similarity mcasures computccl for DZA

indicate that ir is most similar to f. deppeana var. robusta. This eviclence and ev

i-dence from the examination of other spccimens inclicatc that DZA is]. deppNuw

var. robusta and that f. deppeana var. robusta and

!.

deppeana var. :;acatPcensis

are sympatric in the state o[ Zacatecas, near the statc o[ Durango.

The OTU DP consistPcl of specimens identifiecl as funiperus patoniana by

l\fartínez. This OTU is mo~t similar to f. dcppewui. var. rnhusta anrl is similar

to the other varieties of ]. deppcana. Martínez (1946) indicatecl that it wa~ 11ot

conspecific with that taxon or with any othcr known taxon .. The clescriptions

(Martínez, 1946) of f. patonianci ancl

f.

deppeana var. robusta are similar.

except for the bark exfoliation patterns. f. patoniana has hark that is clivi<lccl

into long rectangular strips, that are sometimes interlacccl. f. dcppeana var.

robusta has bark that is divicled into small, rectanpilar plates that are alrno~t

¡:quare. A variant of f. patoniana,

!.

patoniana f. obscura, has bark with square

plates near the grouncl ancl laceratecl or rent bark a hove. Adams ( 1973) reportccl

a form off. deppearur.,

f.

deppea.na J. sperryi (ConC'll) Adarns, whic!t is

JUNIPERUS IN MEXICO AND GUATEMALA

Mountains of west Texas, except fer the furrowed bark ! He concluded that perhap;; only a difference of a few genes is responsible for this form. Thus it

seems reasonable that

f.

patoniana may represent only a few gene differences

from

f.

deppeana var. robusta. The other character states of

!.

patoniana are

within the range of character states for

!.

deppeana var. robusta. The taxon is

also within the geographic range of

!.

deppeana var. robusta. The data analyses in this study suggest that

!.

patoniana should be recognized as a subspecific

category of

f.

deppeana, i.e.,

f.

deppeana var. patoniana (Martínez) Zanoni

(Zanoni and Adams, 1976).

The varieties of

f

uniperus deppeana are generally widely distributed in Me-xico (Fig. 2).

f.

deppeana var. deppeana occurs in the Sierra Madre Oriental from the State of Mexico noo:thward to Coahuila.

!.

deppeana var. robusta i~ founcl in the Sierra Madre Occidental in the states of Zacatecas, Durango and Chihuahua. The range of

!.

deppeana var. zacatecensis is limited to the state of

Zacatecas.

f

uniperus durangensis,

f.

mónticola, and

f.

standleyi form a loose species

group, although the similarities of the..,«e species to

f

.

blancoi,

!.

jaliscana, and

]. saltillensis obscure these relationships in the phenograms. This group is cha·

racterizcd by small, many-seeded eones and ultimate branches that have a beaded

appcarance due to the shape and arrangement of scale leaves (Table 2). The similarities of these three species suggest common ancestry and divergence into

tlnee different geographic regions.

J.

durangensis is found in iso1ated populations in the pinc-oak forests at 2 100 to 2 700 m in the Sierra Madre Occidental. The most widespread species, f. nwnticola, is scattered on the higher volcanoes of lhe Neovolcanic Axis across southern Mexico and on peaks in the folded and faulted Sierra Madre Oriental at elevations generally over 3 000 m ( except at

El Chico, Hidalgo, where

f.

monticola f. monticola is growing at 2 4.SO m).

J.

monticola was divided in to three subspecific taxa (Martínez, 194.6) :

f.

monticola

f. compacta;

!.

monticola f. monticola; and

f.

monticola f. orizabensis. These

taxa were found to be distinct from one another in this study. Examination of

herbarium specimens indicated that

J.

monticola f. compacta and

f.

monticola f.

monticola intergrade and are coanmonly syrnpatric in the Neovolcanic Axis.

f.

11wnticoia f. compacta ancl /. monticola f. orizabensis are syrnpatric in the eastern

Ncovolcanic Axis and in the Sierra Madre Oriental.

f

uniperus standleyi is known from the Sierra de los Cuchumatanes, Gua.tema·

la ancl from Volcán Tacaná, Mexico, at elevations over 3 000 m. This species is located oo a land mfü:s that has been open to colonization of plants since the carly Tertiary (Steycrrnark, 1950).

f.

standleyi apparently has been geograp

hi-BOLETIN DE LA SOCIEDAD BOTANICA DE MEXICO No. 35, 1975

cally separated from

f.

monticola and ]. durangen:;is for a considerable period

of time.

The one-seeded junipers form the largest and most widespread group of junipers in North America. These taxa are very abundant in northern l\1exico and adjacent southwestern United States. Two taxa are known from Chiapas,

Mexico and Guatemala. The Mexican and Guatemalan junipers: Jwi'iperus co-mitana, f. erythrocarpa var. coalwilcnsis, ]. gamboana., ancl

J.

mo1wsperma var. gracilis are very similar to each other.

Analysis of the OTU's LLR ancl MLB inclicates that they are most similar to ]. monosperma var. gracilis. These OTU's probably represent variation of J. monosperma var. gracilis at the periphery oí its distribution. MLT is most similar

to ]. morwsperma var. gracilis, but the identity of this OTU is not clear from the phenograms. The OTU consisted of samples from very old, single-stemmccl trees at La Trinidad, Nuevo León. It is possible that MLT is J. morwsperma. var. gracilis; the differences being related to the ages of the trees. The junipers

sarn-plecl for J. morwsperma var. gracilis were shrubs and not as old as the tree~ at La Trinidad, Nuevo León.

Interestingly, ]uniperns monosperma var. gracilis is more similar to ]. err· throcarpa var. coahuilensis, J. camita.na, and ]. gambomw rhan it is to

J.

monos-perma. from the U.S.A. If this analysis is correct,

J.

monosperma var. gracilis

should probably be allied to a different species.

]uniperus erythrocarpa var. coahuilensis from western ::'dexico (EW) is Ycry simil:Jr to the plants from eastern Mexico (ERW) oí the same taxon, and to

]. monospenna var. gracilis. J. gamboana and ]. comitana are most similar to

each other and then to ]. monosperma var. gra.cilis. ]. saltillensis is most similar

to J. morwsperma var. gracilis.

Although the affinities of the one-seeded junipers are not clcar frorn this

study nor from Martíncz (1946), severa! obsen1ations may bt.: made concerning thesc taxa. Hall (1971) statecl that the J. sa.ltillensis was relaled to]. bl.ancoi, ]. durangensis,

f.

ja.liscana, and ]. monticol.a. Analysis of ]. saltillensis (mean

number of seecls per con e = 1.43) in this study indicates that it is more similar to the one-seeded junipers. The distribution and habitats of J. suüillensis are

also similar to the one-seccled iunipers, wilh which it is often found in Lhc same locations in eastern Mexico.

The subspecific epithet ]uniperus crythroca.rpa var. coahuilcnsis was applied to the Mexican plants of ]. erythrocarpa by Martínez (1946). The type spec i-men, other ~pecime11', and living material of ]. eryth.rocarpa from Texas has

JUNIPERUS IN MEXICO AND GUATEMALA

plants. Therefore, recognition of the Mexican plants as a different taxon is unwarrantecl.

Aclams (1972) incluclecl trees referable to

f

.

erythrocarpa in a stcluy of

f.

pinchoui populations frorn Texas. His results inclicate that these trecs (!. eryt

hro-carpa ) are very similar chemically to

f.

pinchotii and rather unsimilar to

J.

nonosperma, athlough these trees (/. erythrocarpa) were about equally similar

to

!.

monosperma and /. pinchotti when similarities were computecl with 17

morphological characters.

f

uniperus erythrocarpa was consiclerecl to be syno

ny-mous witlt

f.

pinchotii in the flora of Texas ( Correll ancl fohnson, 1970). The

!'la.tus of

!

.

erythrocarpa will be recvaluatecl la.ter when the terpenoicl elata have

been analyzed.

Differences among the one-~eeclecl ju ni pcrs were acknowledged by botanist~

by the rccognition of many species within this group. It appears that the

clif-ferences within this group have been OYerernphásized and that many of the

laxa shorulcl be reduced to subspecific categories. No changes are proposed in

this study, other thant the recognition o{

f

uniperus erythrocarpa var. coahuilensis

as

!.

erythrocarpa. Further work with thcse junipers ancl the one-seedecl junipers frorn the United Sta.tes is plannecl to properly evaluate the affinities of these taxa.

The affinities of the species

f

uniperus jaliscana and /. blancoi to the other

junipers are not clear in this study. Martínez ( 1946) indicated that these species

constituted a subsection in his taxonornic treatrnent; this stucly clid not subs

-tantiate his conclusion. These species exist in small, widely separated populations.

f.

jaliscana is known frorn El Salto, Durango ancl from two localitics in the

region of Río de Bavispe, Sonora.

Except for f uniperus scopulorwn (Zanoni ancl Adams, 1974b),

J.

blancoi is

tlic only Miclclle American juniper with smooth or entire leaf margins. Engelmann

(1877) ancl Hall (1952) were among tl1e first to recognize the importance of

leaf margins in the iclentification ancl taxonomy of the junipers. Formal taxo

-n0111ic recognition of subclivisions of the sabinoicl junipers on the basis of leaf

margi11 was given by Gaussen (1967). Scction Sabina (class Sabina in Gaussen,

]968) was divicled into the subsections (sections in Gaussen, 1967 ancl 1968):

Dt:ntir·ulatac (toorhecl leaf rnargins) and Integra.e (srnooth leaf margins). The

classificatio11 of the junipers of the world with subdivisions based on leaf margins

was prcsented by Gaussen ( 1967 ancl 1968). It is not surprising that /. blancoi

did not show close affinities to the other Mexican and Guatemalan iunipers, if

thr leaí n1ar¡!'in really indica.te two lines

oJ

evolution in the sabinoicl junipers.

f. blancoi would be even more dissimilar if this character wcre weightecl as

BOLETIN DE LA SOCIEDAD BOTANICA DE MEXICO No. 35, 1975

The authors are grateful to the following for assistance in this study: A. Gómez.Pompa, Universidad Nacional Autónoma de México; F. G. Hawksworth, U.S. Forest Service; R. Mac Vaugh, University of Michigan; Sra. Luz María Villarreal de Puga, Universidad de Guadalajara; Jerzy Rzedowski, Escuela

Na-cional de Ciencias BioJógicas; and J. Richards and R. Stavros, Colorado State University.

SUMMARY

The relationships of 18 truca of Mexican and Guatemalan f uniperus were

examined by use of 64 morphological characters in analysis of variance, multiple

range tests and numerical taxc1J1omic methods. Four species complexes were apparent. The flaccidan complex consisted of two varietis of J. /laccida. The

<leppeanan complex includes four varieties of J. deppearw. f unipcrus patoniana

Martínez is closcly associated with varieries ocf J. deppeana and is treated as a

variety of J. deppeana. Juniperus durangensis, J. standleyi, and the three forms

-0f

J. rrwnticola comprised the monticolan complex. The one-seeded complex

included

J.

comitana, J. erythrocarpa, J. gamboana, J. m(]lnosperma var. gracilis,

and J. sallillensis. ]uniperus erythrocarpa var. coahuiwnsis was not distinct from

]. erythrocarpa. Two species, f. búmcoi and J. jaliscana, did not appear to be

closely rclated to the other Middle American junipers.

RESUMEN

Las relaciones de los 18 taxa del género f uniperus en México y Guatemala

fueron examinadas usando el análisis de varianza, pruebas de intervalo múltiple y los métodos taxonómicos numéricos en los 64 caracteres morfológicos usados. Cuatro complejos de especies fueron aparentes. El complejo "flaccidan" que consta de dos variedades de J. deppeana.

!

uniperus patoniana Martínez la cual, ·está cercanamente asociada con las variedades de J. deppeana y se trata como

una variedad de J. deppeana. f uniperus durangensis, J. standleyi, y las tres formas de f. monticola comprenden el complejo "monticolan". El complejo de una semilla incluye a f. comitana, J. erythrocarpa, J. gamboana, J. monosperma

var. gracilis, y

!.

saltillensis. funiperus erythrocarpa var. coahuilensis no fue

diferente a

!.

erythrocarpa. Dos especies,

!

.

blancoi y f. jaliscana, no parecen estar relacionadas a los otros cedros de México y Guatemala.

JUN!PEIWS IN MEXICO ANO GUATE.MALA

BIBLIOGHAPHY

ADAMS, R. P. 1972. Chemosyslemalic and numerical sludies oí natural populations oí

]uni-perus pinc/wtii Sudw. Taxon 21: 407-427.

- -. 1973. Recrnlualion of the biological status of luniperus deppeana var. sperryi Cornell. Brittonia 25: 284-289.

- -. 1975. Population differentiation and relict disjunclions in ]uniperus ashei

(Cupres-saccae) based on terpenoid dala. ]. M olee. Evo/. (in press).

- -. 1975. Statistical character weighting and s·imilarity stability. Brittonia (in press).

AoAMS, R. P. and B. L. TURNEH. 1970. Chemosystematics and numerical studies oí natural

population oí ]uniperus ashei Buch. Taxon 19: 728-751. . .

CoRRELL, D. S. and !VI. C. ]OHNSTON. 1970. Manual of the vascular ¡ilants o/ Texas. Renner:

Texas Research Foundation.

E:-<DLICIIER, I. L. 1847. Synopsis coniferarum. Sangalli: Schcitlin and Zolikofer.

ENGELMANN, G. 1877. The American junipers of the !'ection Sabina. Trans. Acad. Sci. St. Louis 3: S83-S92.

FASSET'r, N. C. 1944·a. ]uniperus virginiana, ]. horizontalis, and ]. scop1iforum -I. The

specific charnclers. BulJ. Torrcy Bot. Club 71: 410-418.

- - . 1944b. ]unipcrus virginiana, ]. horizon:alis ancl ]. scopulorum II. Hybrid swarms.

Bull. Torrey Bot. Club í 1: 475-783.

- -. 1945a. ]nnipems virginiana, ]. horizontalis, and ]. scopulorum III. Possible hybridiz·

at-ion oí ]. horizontalis and J. scopulorwn. Bull. Torrey Bot. Club 72: 42-46.

- -. 1945b. ]uniperus virginiana, ]. horizontalis, and ]. scopulorum . IV. Hybrid swarms

of ]. virginiana and ]. horizontalis. Bu//. Torrey Bot. Club 72: 379-384.

- -. 1945c. ]uniperus virginiana, J. horizontalis and ]. scopu/orum . V. Taxonom~c treat

-111ent. Bulll. Torrey Bot. r:Lub 72: 480-482.

FLAKE, R. H., E. Vo:-< RuDLOFF, and B. L. TUR!'iER. 1969. Quantitative study oí clinal

variation in ]nniperus virginiana using terpenoid data. Proc. Natl. Ac.ad. U.S.A. 64:

487-494.

- -. 1973. Confirmalion of a clinal pattern of chemical differentiation in ]uniperus

virgilliana from terpenoid data obtained in successive years, p. 215-228, In V. C.

Runcckles and T. J. Mabry (ed.), Recent advances in Phytochemistry, Vol. 6. New

York: Academic Press.

F1.nu1-.:. R. 1963. The Jislribution o[ coniíer and taxad genera in time and space. Acta f-forti Berg. 20: 121-312.

GwssE-.:. H. 1%7. La classif.ication des gcnévricrs (Junjperus). Comp!. Rend. Hebd.

Seanccs Acad. Sci. 265: 954-957.

- -. 1968. Les gymnospermes actuelles et fossilcs. Les cupressacees. Trav. Lab. Forest.

Toulouse Tome !I, Sect. I, Vol. 1, partie JI 2, Fase. 10. 1-326.

HALL, M. T. 1952a. A hybrid swarm in ]uniperus L. Evolution 6: 347-357.

- -. 1952b. Variation and hybridizalion in ]uniperus. Ann. Missouri Bot. Carden 39: 1-64.

- -. 1954. Nomencaltural notes conccrning ]uniperus. Rhodora 56: 169-177.

-. l 955. Comparison of juniper pop11lation of an Ozark glade and old fields. Ann.

lfissonri Rut. Garrll'/1 -12: 171-194.

- -. 1971. A new spccics of ]unipems from i\J,.xico. F.ic/rlinna, Bot. 34: 45·53.

1111.1 .. M. T. and C. l CARR. 1964. Diíferential selcction in juniper population from the

Baum Vimeslone and Trinit1· sand of southern Oklahoma. Butler Univ. Bot. Stnd.

14: 21-40. .

- - -. 191iB. Variability in .funiperus in thc Palo Dnro Canyon of "·estern Texas. Sonth1u.

Naturalist 13: 75-98.

HALL, l\T. T., J. F. MccoR~HCK, and G. G. Focc. 1962. Hybridizalion betw!een ]unipems

ashci Buchholz and ]unipems pinchotii Suuworth in southwestern Texas. Butler Univ.

Bot. Stnd. 14: 9-28.

Jow,srn:-1, I. l\T. 194.3. Plants oí Coahuila. castcrn Chihuahua and adjoining Zacatecas

and Durango. ]. A rnold Arhor. 24: 306-339.

LE\J(lI'IE·SEIJ,\STI l'I,

e

1967. Apparcil rcproduclcur male des ]uniperus. Trnv. Lab.

BOLETIN DE LA SOCIEDAD BOTANICA DE MEXICO No. 35, 1975

MARTÍNEZ M. 1944. Nuevas especies de funiperus. Anal. Inst. Biot México 15: 7-15

- --194Ó. Los funiperus mexicanos Anal. Inst. Bio'l. México 17: 3-128.

- - . 1953. Las pináceas mexicanas México. Seer. Agric. Ganad. Subsecr. Recurs. Forest.

Caza.

- - . 1963. Las pináceas mexicanas. Tercera edición. México: Universidad Nacional Autónoma de México.

McVAUGJ-I, R. 1972. Botanical exploration in Nueva Galicia, Mexico. Contr. Univ_ Michigan

Herb. 9: 205-357.

PowELL, R. A. and R. P. ADAMS. 1973. Seasonal variation in the volatile terpeno-ids of

funiperus scopulorum (Cupressaceae). Amer. ]_ Bot. 60: 1041-1050.

ScHURTz, R. H. 1971. A taxonomic analysis of a triparental hybrid swarm in funiperus L.

Ph_ D. dissertation, Univ. Nebraska.

SNEATH, P. H. A. 1957. The application of computers to taxonomy. f. Gen. Microbio/_

17: 201-226.

SoKAL, R. R. and P. H. A. SNEATH. 1963. Principles of numerical taxonomy. San

Fran-cisco: W. H. Freeman and Company.

STANDLEY, P. C. 1920. Trees and shrubs of Mexico. Contr. U.S. Natl. H erb. 23: 1-170.

STANOLF.Y, P. C. and J. A. STEYERMARK. 1943. Studies of Central American plants - IIL

Field. Mus. Nat. Hist., Bot. Ser. 23: 1-28

- - . 1958. Flora of Guatemala. Fieldiana, Bot. 24: 1-478.

STEYERMARK, J_ A. 1950_ Flora of Guatemala. Ecology 31: 368-372.

VAN HAVERUEKE, D. F. 1968. A populational analysis of funiperus in the Missouri River

Basin. U ni v. N ebraska Stud. n.s. 38: 1-82.

VASEK, F. C. 1966. The distribution and taxonomy of three western junipers. Brittonia 18:

350-372.

VASF.K, F. C. and R. W. ScoRA. 1967. Analysis of the oils of western North American

junipers by gas-liquid chromatography. Amer. J. Bot. 54: 781-789.

ZANONI, T. A. and R. P. ADA:lfS. 1973. Distribution and synonymy of funiperus cali/ornica

Carr. (Cupressaceae) in Baja California, Mexico. 811//. Torrey Bot. Club 100:

364-367.

ZANONI, T. A. and R. P. AoAMS. 1975. Southern range extension of funipems scopulorum

Sarg. (Cupressaceae) into Mexico. SoLLthw. Naturalist 20 (in press).

- -. 1976. Distribution and synonymy of the Mcxican and Cuatemalan ]LLniperus

JUNIPERUS IN MEXICO AND GUATEMALA

TABLE 1

Murphological character& and thc states med. Superscrits (a) anrl (b) indicate 1hat the character was used in computation of similarity matrices hased on 63 and 45 morpho

-logical characters, respectively.

Characters

Peduncles PEDab

PEK•b

Canes

FRV0_b

FRRab

FRSa

FHU0_b

FVAª"

FCHª"

Seeds

SEV0

SERª"

SSHª"

SCHª"

States (if app].icahle)

LENGTH: average of up to 10 rneasurernents an<l not kss than 5 (in mm.).

PEDUNCLES CURVED: per cent curved, up to 10 rneasurements and

not ]ess than 5.

LENGTH X WIDTH: average of 10 measurements and not ]ess than 4 (in mm2_.).

LENGTH/WIDTH: average ratio of up to 10 measurements and not less than 4·.

SUR FA CE TEXTURE: l.

=

smooth; 2. sutures obvious; 4.

bumpy; 3. = scale

horny.

COLOR HUE: average !me of up to 10 eones and not less than 4.

COLOR VALUE: average value of up to 10 eones and not less than 4.

COLOR CHROMA: average ehroma of up to 10 eones and not less

than 4.

BLOOM: l. = very light; 2. = light 3. = medium; 4. = heavy;

5.

=

obscures cone color.

PU LP: l. ,oft fle,hy; 2.

=

fle~hy to fibcouo; 3. = soft fibrous;

4. = hard fibrous.

LENGTH X WIDTH: average of up to 10 rncasurernenls and not lcss

th,m 4 (in mm2_.l.

LENGTH/WIDTH: average ratio of up to 10 mcaourements and not

less 1han 4.

SHAPE: l.

=

widest "t basal half; 2. :.::. widcst at middlc; 3.

widesl al distal hall.

COLOR HUF.: a1·crap;c hue of up to 10 cecds and not les" than 4.

COLOR VALUE: average val u E of up to 10 seeds and not le>s' than

4 seeds.

COLOR Cl-IROMJ\: average chroma of up to 10 cceds and not less

BOLETIN DE LA SOCIEDAD BOTANICA DE lVIEXICO No. 35, 1975

Characters

SEG0b

SPfab

sszab

HNO

H/Sab

Whip Leaves

\\1LLª

WL\\I"

WLV•

WLRab

WLi\1ab

WLT0b

WLNab

WGVa

WCRª"

TABLE l. (Continued)

States (if appJ.icable)

GROOVES: average number of grooves per seed of up to 10 seeds

and not ]ess than 4.

SEEDS PER CONE: average of up to 10 eones and not lcss than 4.

RELATIVE SEED SIZE:

=

1 (% Iarge seeds)

+

2 (% medium

seeds)

+

3 ( % small seeds).

HILUM SCAR LENGTH: average of up to 10 mcasurcments and not

less than 4.

1-HLUM SCARS PER SEED: average of up to 10 measurcmenls and

not less than 4. (in mm).

HTLUM LENGTH/SEED LENCTH: average of up to 10 mcasurernents

and not Iess t han 4

LENGTH: average of up to 10 rncasurements and not Iess than 4

(in mm.).

WIDTH: average of up to 10 measurements and not lcss than 4

(in mm.).

LENCTH X \í'IDTI-l: average of up to 10 measurements and not

less than 4 (in mm2.).

LENCTI-I/WJDTI-I: avPrage of up to 10 111ea,uren1ents and not lcss

than 4.

SI-IAPE: l.

=

widc>st al ba,al half; 2.

=

widest at micldlc: 3.

=

widest al distal half.

MAH.GTN: l.

=

smooth; 2. cmarginate; 3.

=

light serration,

small teeth; 4 ..

=

heavy serralion, largc tecth.

TIP: width of lcaf tip. 66 mm. from tip, average o[ up to JO mearnr·

ements and not less than 4.

DOHS1_\L SlJRFACE: _1.

=

_rn11hn; 2.

=

_{smooth; 3.}

keeled; 4·.

=

strongly keeled.

LEAVES PER NODE: 2.

=

two; 3.

=

thrce.

LEAVES PEH. 6.6 mm.: average of up to 10 measurements and not

less than 4.

CLAND LENCTH X WIDTH: avt•rage of up lo 10 rneasurcmcnts and

not ]ess than 4 (in rnm".).

CL\NI) LENCTH/WJDTH: average of up to 10 measurcmrnts and

Characters

WGSab

WGPab

WRPªb

WCP

WGMab

Sea/e Leaves RLLª

RLWª

RLVª

RLRab

RLSab

RU-Jab

RL'fa0

RLDab

RLOa

RCVª

RCRªº

RCSno

RCPab

RRPab

!{('fa

RGMab

LFDao

JUNIPERUS IN lVIEXICO AND GUATEMALA

TABLE l. (Continued)

States (if app]foable)

GLAND SHAPE: l.

=

widest at basal half; 2.

=

widest at rnidd!e;

3.

=

widest at distal half.

GLAND PROTRUSION: l.

=

sunken; 2.

=

srnooth; 3.

=

protrudes.

GLAND RUTURE: l.

=

no; 2.

=

yes.

LENGTH FROM WHIP LEAF CLAND CENTER TO LEAF TIP:

average of up to 10 rneasurernents and not less than 4 (in mm.)

RATIO OF WLL/WCT: average of up to 10 rneasurements and not

less than 4 rneasurernents.

LENGTH: average of 10 measurements (in mm.).

WIDTH: average of 10 measurements (in mm.).

LENCTH X WIDTH: average of 10 measurements (in mm2.).

LENGTH/WIDTH: average of 10 measurements.

SHAPE: l.

=

widest at basal half; 2.

=

widest at rniddle; 3.

=

widest a t distal half.

J\fARCIN: l.

=

smooth; 2. = emarginate; 3.

=

light serration,

small teeth; 4.

=

heavy serration, lar ge teeth.

TIP: 11'idth of lea[ tip .33 mm. from tip, average of 10 measurements.

DORSAL SURFACE: l. = sunken; 2.

=

smooth; 3.

=

lightly

keeled; 4.

=

heavily kecJed.

LEAVES PER NODE: 2.

=

two; 3.

=

three.

LEA VES PER 3.3 mm.: average of 10 rneasuren1ents.

GLAND LENGTH X WIDTH: average of up to 10 measurements (in mm2 .).

GLAND LENCTH/WJDTH: average oí up to 10 measurements.

GL\:\I) SHAPE: l.

=

widcst at basal ltalf; 2. = widest at middle;

3.

=

widest a distal half.

GLAND PKOTil.USION: l. = sunken; 2.

=

smooth; 3. = protrudes.

GLAND RUPTURE: l. no; 2.

=

yes.

LENGTH FROJ\f SCALE LEAF CLAND CENTER TO LEAF TIP:

average of 10 measuremcnts (in mm.).

RATIO OF RLL/HCT: average of up to 10 measurements.

BOLETIN DE LA SOCIEDAD BOTANICA DE MEXICO No. 35, 1975

Characters

Bran.ches FLC•b

BTCab

BDG•b

BDSab

BRK•

TABLE l. (Continued)

Sta tes ( if applicable)

BRANCH FLACCIDNESS: l.

=

not flaccid; 2.

=

flaccid.

TERMINAL WHIP CURVATURE: l.

=

straight; 2. curved;

average of up to 5 measurements.

DEGREE OF BRANCHING ON TERMINAL WHIP BRANCH:

average of up to 5 measurements.

DI~TICHOUS BRANCHING ON LATERAL BRANCHES: l.

=

yes;

2.

=

no.

ANGLE OF BRANCHING OF ULTIMATE TWIG: average of up to

10 measurements to nearest 5 dcgrees).

DIAMETER OF ULTIMATE TWIG: average of 10 measurements (in

mm.).

STEM BARK EXFOLIATING PATTERN: l. = long strips; 2. =

fibrous strips; 3.

=

quadrangular plates; 4.

=

long plates;

5. = interlaced; 6. = papery-scaley.

TABLE 2.

Characterization of thc complexcs of llliddle American ]uniperus.

Character deppeanan jlaccidan one-seeded monticolan

Cone length X width (FRV) lar ge large small intermedia te

Cone pulp (FRP) fibrous fibrous soft, fibrous to fleshy soft, fibrous

Seeds per cone (SPF) severa! man y one severa]

Seed length X width (SEV) lar ge large intcrmediate small

Hilum scar length (HIL) long long intermediate short

Angle of branching of ultimate narrow narrow medium to wide wide

twig (BAN)

Cone color (FHU, FVA) dark, reddish brown dark, reddish brown variable variable

Peduncle length (PEO) intermedia te intermedia te short long

Peduncle curvature (PEK) straight straight straight curved

Leaves per node (WLN, RLN) 2 or 3 mostly 2 mostly 3 2

Terminal whip (BTCJ straight strnig or curved •traight (curved in curved

] . saltillensis)

Scale leaf width (RLW) wide variab]P, variabl,. variable

Scale leaf tip divergence (LFD) appressed divergent appressed appressed

BOLETIN DE LA SOOEDAD BOTANICA DE MEXICO No. 35, 1975

SlBSECCIOO /'12tí"TJC(X .. AE

a

B

SLBSECCICl'i t•ú\OSP(RMl..E

Figure l. Subsections and taxa of ]unipems in Mexico as recognized by Martínez

(adap-ted from Martínez, 1963). Subsection :rvionospermae: ]. comitana Martínez (COMIT.);

]. monosperma (Engelm.) Sarg. var. monosperma (MONO.) : ]. monosperma var. gracilis

Martínez (MONO. G.); ]. gamboana Martínez (GA:MB.); J. wl'ijomica Carriere (CALIF.);

.f. erythrocarpa var. coahuilensis Martínez (ERY. COAH). Subsection Deppeanae: ]. dep-¡1cana Steudel var. deppeana (DEP.); ]. deppeana var. pachyphlaea (Torrey) Martínez (DEP. PACHY.); ]. deppeana var. robusta Martínez (DEP. ROB.); J. deppeana var.

zacatecensis Martínez (DEP. ZAC.); J. patonwna l\1artínez f. patoniana (PAT.); ].

patoniana f. obscura Martínez (PAT. 0.). Subsection FJaccidae: J. /laccida Schlecht. var.

flaccida IFLAC.); ]. flaccida var. poblana Martínez (FLAC. POB.). Subsection Jali

s-canae: J. blancoi Martínez (BLANCO!); J. jaliscrma Marlínez (JALTS.). Subsection Mon· ticolae: J. standley Steyermark (STAND.); J. durangensis Martínez (DUR); J. monticola

Martínez f. monticola (MONT.); ]. monticola f. compacta ::Vlartínez 11\IONT. C.); ]. monticola f. ori:abensis Martínez {l\IONT. O.).

JUNIPERUS IN MEXICO AND GUATEMALA

¡/

MSs

OTU T;ixon BL ,r. :.Z:1;:Y:: co ,'. ;:;Jr1:.:a1:a

DD ~. i..:.:.,pc n:.~

var. dcpp.=:.:1:r. DP :a ~01:ic•:..;

DR .J. ::·eprea1:.:: var. z·u~:m;;r"! DZ ~. ..-p¡::.;m:.:r

\'ar. :;aC"a tei:oi:i

DZA d.:ppe.J.r ... <.

•1ar. ?

ER\.' ~·. ,-!'J ~;:r·~C'~rpc:

var ..... ·v:f::ii .. '.t!.1:;,;f,3

EiJ ,,r, er"j ci:roear;....:

var. aod!liilL~,::ii.t; FL J. ,·'[.J...:ai<=.'..:.

;~Zacddc:

FP J. flaccida va r. pob z,u;a

Source El Salto, Dgo.

La Trinitaria, Chis.

Apizaco, Pue.

El Chico, llgo ,' Llanos de P~o:ote, V. C.

ílgo.

Cd. Guerrero, Chih.

Huejotitan, Chih.

Hacienda Coyotes, Chih.

~:!na Dolores, Zac.

J !:nenez del Teul, Znc.

El A lamo, Zac.

Saltillo, Coah.

Chihu;thua '· Chih.

San Frnnc !reo del Oro, Chih.

L1s Nieves, Dgo.

El l.imon, Zac.

Puerto del Gato, ~lich.

Tchuacan, Pue,

~orelos, Owc ..

San MarCf)S •. N. L.

Ocotepec, Oax.

OTU Taxon

GA J. ganLoan-:: JA Lr. ,jc.liar:c:1:..-:

:·:G ,: . ·-:onosparnc:

var. !jl'C:ciZ~s

XS J •. "'IC>:ospe?"~ª var. mo,:oeocina MC J. monticoz.G

f. compacta ~C{ J. rionticola

f. monticoZa

MO J. M01lticola

f. 01•izabe1:sis

SL J. sa ZtiLLen::;is

ST J. standZcyi

LL·: J. ·~orioape'T"T!<1.

var, grac7.:i.e

m.n J. ,..,,,¡osvc1'Ma

var. g~cilic?

MLT J. morwspcrrra var. gi•aciZi.a?

Source

Co::iltan, Chis.

Cuale, Ja 1.

Tranc<l:>, ligo.

P;ümill.as, Tamps.

Dr. Arroyo, :-;. L.

Pto. de Pastores, '.'..:. L.

Cerro Potosi,.~!. L.

Ar izona ar.~ ~olorad.J 1 USA Nevado de Toluca, Mex.

El Chico, ligo.

Tan toco, Mex.

Pico de Orizaba, V. C.

Tokio, N. L.

Saltillo, Coah.

Encantada, Coah.

~cxico and Guateniti.la

Saltillo, Coah.

Saltillo, Coa!:.

La Tdnidad, .L L.

'°

o

OTU's

FP FL JA ST Bl MM f«) MC MS DU DZ 00 DP DR DZA · GA SI. CÓ ERW MLT EV. MG MLB LLR

.

1

l}J

~~

1

1

1

1

~~

T

T 1

63 CHAR.

w-_JF-1

Figure 3. Single linkage clustering of 24 OTU's using 63 morphological characters.

v

:;¡:;¡

o

¡-rrl

--3 ~ ü

t=l

I:""'

>

{f)

o

ti

-

t=l ü

"'

>

¡

ü

;:: to

,.

o

"'

--3

:::¡

>

-<

'2!

-

_n

>

ü t=l

is:

t=l

:><

-n

o

'2!

?

""

_<n

...

"'

__, <rt

l.D

-OTU'~

FP rL ST lll JA MO MM MC DIJ DZ DD DP CR DZA MS ERW Sl GA CO EW Mll MG MlB LLR

LJ

~

T

--

'-.,----1

1

T

-,--

1

~-.

45 CHA

R

.

W:_J

F- l

.9

'--..

ce

<

::¡;

t'"l ;::;:¡

ce

(/)

-z

:B

r.?I

~

3:

t'j

;:: i::S

~

8

)ll"<

-:

_>

....;

_z

t:i

0

e

>

....,

M

~

t-<

>

'°

G.?

OTU's

FP Fl DZ DO DP DR DZA JA BL ST MO DU MM MC MS EW Sl GA CO ERW MLT MG MLB LLR

45 CHAR.

w::F-

1

Figure 5. Single li11kagc clnstcring of 24 OTU's using 45 morpho1ogica1 charactcrs.

"'

!

_....

>

.8 ;ID

:::f

-<

1

1

.5

to

o

r

M

>-l

...

z

o

M

&:

Ul

o

o

M

o

>

o

to

o

>-l

>

z

...

(')

>

o

M

¡;¡::: M

::<

...

(')

o

z

?

<:;>

·"'

_... \O

__,