ARTICLE IN PRESS

Dendrochronologia 22 (2005) 187–193

1125-7865/$ - se

doi:10.1016/j.de

�Correspondfax: +54261 42

E-mail addr

www.elsevier.de/dendro

ORIGINAL ARTICLE

Anatomy of growth rings at the Yucatan Peninsula

Fidel A. Roiga,�, Juan J. Jimenez Osorniob, Jose Villanueva Diazc, Brian Luckmand,Holm Tiessene, Andrea Medinaf, Elke J. Noellemeyerf

aLaboratorio de Dendrocronologıa, IANIGLA-CONICET, Mendoza, ArgentinabUniversidad Autonoma de Yucatan, Merida, Yucatan, MexicocCENID-RASPA-INIFAP, Durango, MexicodDepartment of Geography, University of Western Ontario, London, Ontario, CanadaeInstitut fur Tropen- Pflanzenbau, Universitat Gottingen, GermanyfFacultad de Agronomıa, Universidad Nacional de La Pampa, Santa Rosa, Argentina

Received 7 December 2004; accepted 28 April 2005

Abstract

The dendrochronological characteristics of 52 tree species from the semi-tropical forests of the Yucatan Peninsulawere opportunistically explored in a salvage dendrochronological study. The existence of clear growth rings in thesetrees is a key prerequisite for further studies and a convincing demonstration of the dendrochronological potential oftropical tree species will allow the development of future research programs concerning the ecology of the species andinferences about past environmental changes detected from tree rings. Many aspects of the conservation andmanagement of Yucatec forests should be urgently addressed to aid in the development of improved strategies beyondthe scope of more traditional agricultural uses. Development of tree-ring analyses from selected local species can be ofsubstantial assistance in these initiatives.r 2005 Elsevier GmbH. All rights reserved.

Keywords: Growth rings; Tropical dry forests; Yucatan; Mexico

Introduction

Tropical dry regions are some of the most threatenedecosystems in the world. Economic and demographicchanges, climatic instability and inappropriate agricul-tural practices are the principal drivers of land usechanges and the ongoing degradation of land quality,often leading to desertification. Radical procedures forthe management of these dry forests are essential toimprove forest production and preserve or recover

e front matter r 2005 Elsevier GmbH. All rights reserved.

ndro.2005.05.007

ing author. Tel.: +54261 4287029;

85940.

ess: froig@lab.cricyt.edu.ar (F.A. Roig).

endangered areas. Tree-ring analysis can contributesignificantly to the measurement of forest sustainabilityand ecosystem resilience in these semi arid regions.The usefulness of tree-rings derives from their annual

resolution, precise dating, and sensitivity to climaticvariables. A variety of environmental stressors (includ-ing climate) affect the physiology of trees and thesevariations are recorded in the tree rings. Thus,dendrochronology has been extensively employed inenvironmental studies of temperate and high altitudeforests, but there have been few applications to datefrom the tropics (Roig, 2000). One of the main reasons isthat many tropical trees do not produce anatomicallydistinctive annual growth rings that can be used for the

ARTICLE IN PRESSF.A. Roig et al. / Dendrochronologia 22 (2005) 187–193188

development of reliable tree-ring chronologies (Stahleet al., 1999). However, many tropical forests experiencea seasonal climate that result in the formation ofdistinct, annual rings in their woods. Therefore, severalresearch contributions on tropical dendrochronologyhave been published in the last years worldwide (e.g.,Bormann and Berlyn, 1981; Pant, 1983; Jacoby, 1989;Bhattacharyya et al., 1992; Buckley et al., 1995; Ecksteinet al., 1995; Pumijumnong et al., 1995; Stahle et al.,1999; Roig, 2000; Worbes, 2002; Schongart et al., 2004).In Yucatan, the dry forests have been subjected to a

variable degree of anthropogenic modification, espe-cially by shifting cultivation, sisal (Agave fourcroydes

Lem.) plantation and cattle farming activities. Lowdeciduous forests that originally covered an area of1.8� 106 ha have been reduced to ca. 4� 105 ha, and, forexample, 2� 105 ha have been deforested simply for theproduction of sisal. The pressures on these forests wereparticularly strong during the 19th and the first half ofthe 20th after which commercial sisal cultivationdeclined (Gonzalez-Iturbe et al., 2002). Now, theabandoned cultivated fields slowly recover. Contraryto these aggressive agricultural practices, the Mayaculture traditionally promoted the use and protection ofnative species in the home gardens (Solares). Withinthese agricultural units the Mayas have employed treesfor different purposes, such as for fodder, medicine orjust as protection for their houses and cattle.Little information exists on the dendrochronology of

trees from the deciduous low-stature tropical dry forestsof northern Yucatan and cultivated trees from theSolares. The absence of tree-growth information sig-nificantly hampered studies of the productivity anddevelopment of natural, recolonizing and cultivatedforests in this region. In the fall of 2002, hurricaneIsidore devastated a large part of this region felling largeareas of forest and woodland providing abundantdeadfall. This paper reports initial results of a salvagedendrochronological survey of cross-sections of some ofthis material that sought to identify suitable species tobe used in tree-ring analysis, particularly through therecognition of anatomical properties defining the growthrings. This would provide valuable information to beused in programs concerning tree-growth estimation andstudies based on interactions between tree growth andvarying environmental controls. These studies couldprovide future guidelines to improve the management ofnatural forests and Solares ecosystems.

The study area and wood collections

Sampling was carried out in northwestern Yucatanduring April 2003 (Fig. 1A). Most of this territory is akarst flat plain developed on Pleistocene limestones, on

which precipitation infiltrates into a well developedunderground drainage system. Soils are thin, calcareousand with a high percentage of rocks in the surface(Aguilera, 1958). The climate is semi-arid warm tropical(Aw0 according to Koppen’s classification; Garcıa,1973), with a seasonal rainfall in May–October(Fig. 1C–D). This rain fall pattern is modulated bytrade winds and by the northward displacement of theIntertropical Convergence Zone (Mosino and Garcıa,1974). As a result of a south-to-north precipitationgradient there is a dry area in the north of the YucatanPeninsula, with marked drought conditions duringNovember to April (rainfall amounts to only 30% ofthe annual precipitation).The precipitation gradient and the rainfall season-

ality, strongly influences the physiology of plantscausing leaf fall and reduced growth. In the north ofthe Peninsula, deciduous trees are dominant whiletoward the south there is a preponderance of evergreentrees (Flores and Espejel, 1994; see Fig. 1B). Vegetationtypes at the north are classified as low-stature and spinydeciduous forest and median-stature sub-deciduousforest, with differences in species composition, treeheight and rainfall stress. Leguminosae constitute thedominant tree species in all of these vegetation types(Flores, 2001).Wood samples were collected at Chicxulub site (CIC

code), at the north of the study region (Fig. 1A).Toward the south, collections were made at Hocaba(HOC) and Sahcaba (SAH) sites and at the JardınBotanico from the Centro de Investigacion Cientıfica deYucatan, Merida (HEC). Hocaba and Sahcaba corre-spond to zones with permanent agricultural populationand the species collected are those traditionally con-served and cultivated in the Solares systems. Woodsample preparation included cutting, polishing, andstaining. Ring anatomy definition follows the IAWAnomenclature (IAWA, 1989). Samples were examinedunder a Wild M8 stereomicroscope. All samples arestored at the Wood Collection of the Laboratorio deDendrocronologıa, Mendoza.

Results

Table 1 contains a list of the collected species studied,sorted by botanical family. 52 species from 22 familieswere analyzed in this study. Other 19 species werecollected (RES site as in Fig. 1A) but remain to betaxonomically identified and therefore were not includedin this analysis. About 35% of the taxa we analyzedshow distinct growth ring boundaries and ring bound-aries can be identified with some difficulty in a further52%. Almost 88.5% of woods display porosity of thediffuse-porous type, supporting the general trend for

ARTICLE IN PRESS

Fig. 1. (A) Location of the wood sample sites in the Yucatan Penınsula, Mexico. (B) Vegetation types of Yucatan. (C) Climate zones

of Yucatan according to Koppen’s nomenclature. (D) Climate diagram for Merida state.

F.A. Roig et al. / Dendrochronologia 22 (2005) 187–193 189

tropical (Brazilian) woods (Metcalfe and Chalk, 1983;Alves and Angyalossy-Alfonso, 2000). This featurerestrain the use of the vessel diameter differences asdescriptor of the growth ring boundaries from most ofour set of tropical samples. However, the marginalparenchyma is the most common character associatedwith anatomical growth limits, since 61% of woods withdistinct growth rings show this character (this increasesto 72% if we include the woods with difficult growthrings). Distinct growth ring boundaries marked by otherstructural changes are less frequent, e.g., differences invessel diameter between early and late-wood (11%) ordifferences in radially flattened and thick-walled fibres inlate- vs. earlywood (22%), or combination of bothcharacters (17%). The absence of growth rings seems tobe more related to Moraceae, Nyctaginaceae andPalmae families, although it is well known that the last

two groups form unusual woods or no ring structuresper se (monocot) respectively.Wedging rings were found in 25% of trees but

particularly abundant in Capparis incana Kunth, Mal-

pighia glabra L., Gymnopodium floribundum Rolfe,Citrus sp., Thouinia paucidentata Radlk. and Vitex

gaumeri Greenm. The discontinuous rings are ofcommon occurrence in other tropical floras (Worbes,2002) and certainly constitute a real problem indendrochronological dating procedures.

Conclusions and recommendations

The climatic and vegetation characteristics of Yuca-tan provide a potential for the development of

ARTICLE IN PRESS

Table 1. Growth ring characteristics of Yucatan tree species. Key to classifications

Wood characters

Family GR P LWF MP RB WR

Species

Anacardiaceae

Spondias purpurea L. B 3 1 — FW —

Annonaceae

Annona squamosa L. B 3 — 1 MP —

Annona sp. B 3 — 1 MP —

Boraginaceae

Cordia dodecandra A. DC. A 3 — 1 MP —

Ehretia tinifolia L. A/B 2 2 — VD/FW 1

Burseraceae

Bursera simaruba (L.) Sarg. B 3 1 — FW —

Capparidaceae

Capparis incana Kunth B 3 — 1/2 MP 2

Lauraceae

Persea americana Mill. A 3 1 — VD/FW —

Leguminosae

Acacia dolichostachya Blake A/B 3 — 1 MP —

Acacia gaumeri Blake A 3 — 1 MP —

Acacia riparia Kunth B

Albizia tomentosa (Micheli) Standl. A 3 — 1 MP —

Apoplanesia paniculata Presl B 3 — 1 MP —

Caesalpinia gaumeri Greenm. A 2/3 — 1 MP/VD —

Caesalpinia yucatanensis Greenm. A/B 3 — 1 MP —

Chloroleucon mangense (Jacq.) Britton & Rose B 3 — 1 MP —

Enterolobium cyclocarpum (Jacq.) Griseb. B 3 1 2 MP/FW —

Erythina standleyana Krukoff C 3 — — — —

Havardia albicans (Kunth) Britton & Rose B 2/3 — 1 MP —

Leucaena leucocephala (Lam.) de Wit. A 3 — 1 MP —

Lonchocarpus xuul Lundell B 3 — 1 MP 1

Lysiloma latisiliquum (L.) Benth. B 3 1 1 MP —

Mimosa bahamensis Benth. B 3 — 2 MP —

Piscidia piscipula (L.) Sarg. A 2/3 — 1 MP —

Pithecellobium dulce (Roxb.) Benth. B 3 — 1 MP 1

Senna racemosa (Mill.) Irwin & Barneby B 3 — 1 MP —

Tamarindus indica L. A 3 — 1 MP —

Malpighiaceae

Malpighia glabra L. B 3 — 1 MP 2

Meliaceae

Cedrela odorata L. A 1/2 1 1 MP/VD —

Trichilia hirta L. A 3 — 1 MP 1

Moraceae

Brosimum alicastrum Sw. B 3 — 2 MP —

Chlorophora tinctoria (L.) Gaudich. ex Benth. B 3 — 2 MP 1

Ficus benjamina L. C 3 — — — —

Ficus cotinifolia Kunth C 3 — — — —

Myrtaceae

Psidium guajaba L B 3 1 — FW 1

Nyctaginaceae

Neea sp. C 3 — — — —

Pisonia aculeata L. C 3 — — — —

F.A. Roig et al. / Dendrochronologia 22 (2005) 187–193190

ARTICLE IN PRESS

Table 1. (continued )

Wood characters

Palmae

Cocos nucifera L. C 3 — — — —

Polygonaceae

Gymnopodium floribundum Rolfe A 3 3 — FW 2

Rhamnaceae

Colubrina elliptica (Sw.) Brizicky & Stern A 3 1 1 MP/FW —

Karwinskia humboldtiana (Schultes) Zucc. A 3 — — VD —

Rubiaceae

Guettarda elliptica Sw. A 3 1 — FW/VD —

Randia obcordata Watson B 3 1 — FW —

Rutaceae

Citrus sp. B 3 — 1 MP 2

Sapindaceae

Melicoccus bijugatus Jacq. B 3 — 1/2 MP 1

Talisia olivaeformis (Kunth) Radlk. A 3 — 1 MP —

Thouinia paucidentata Radlk. A 3 1 — FW 2

Sapindus saponaria L. B 2/3 — 1 MP —

Sapotaceae

Manilkara zapota (L.) Royen B/C 3 — 2 MP —

Sterculiaceae

Guazuma ulmifolia Lam. A 3 1 — FW —

Theophrastaceae

Jacquinia flammea Millsp. ex Mez B 3 — — FW —

Verbenaceae

Vitex gaumeri Greenm. A 3 1 — FW 2

GR ¼ Growth ring: A: distinct; B: distinct with difficulty; C: indistinct or absent.

P ¼ Porosity: 1: ring-porous; 2: semi-ring-porous; 3: diffuse-porous.

LWF ¼ Latewood fibre wall thickness: 1: thin- to thick-walled; 2: very thick walled.

MP ¼Marginal parenchyma: 1: continuous; 2: discontinuous.

RB ¼ Ring boundary structure (as observed in the latewood portion): FW: change in fibre wall thickness or fibre radial size; VD: changes in vessel

diameter; MP: presence of marginal parenchyma.

WR ¼Wedging rings: 1: rare; 2: abundant.

F.A. Roig et al. / Dendrochronologia 22 (2005) 187–193 191

dendrochronological studies. The long dry seasonoccurring between November and April induces growthseasonality in plants. Therefore, over almost all of thePeninsula and particularly at the northern side (withmore extreme aridity conditions) the vegetation totallyor partially sheds its leaves during the summer months(Flores and Espejel, 1994). This leaf loss must necessa-rily influence the activity of the growth tissue (cambium)inasmuch as, without carbohydrate production and/orthe possibility of efficient gradients to transport water,the vascular cambium would enter a stage of relative restcoinciding with the dry period. As a result many woodsin this study region will show anatomical growthmarkers.The fact that 35% of the taxa in Yucatan show clearly

identifiable boundaries (mostly because of the presenceof marginal parenchyma bands) implies a very high

incidence of distinct growth rings for a tropical zone.Alvim (1964) and Chowdhury (1961) reported a 25–45%of species in the tropics of the Amazon and India withdistinguishable growth rings. Even though the numberof samples we analyzed in the Yucatan is relativelylimited, the percentage of plants with distinct rings iscomparable to or substantially higher than thoserecorded for some humid tropics (Fig. 2).The impressive tree species diversity of the Yucatan

flora poses interesting possibilities for dendrochronolo-gical studies. Our observations are based on only 52species (22 families), but this figure represents only asmall portion of the whole number of species in Yucatanthat could be analyzed. For example, we sampled 19leguminous trees (42% of which had visible rings), butthe Yucatan flora contains 103 tree and 30 woody shrubspecies that belong to this family (Flores, 2001).

ARTICLE IN PRESS

Fig. 2. Anatomical morphology of growth rings from some northern Yucatan Penınsula trees. (A) Cedrela odorata, (B) Trichilia

hirta, (C) Karwinskia humboldtiana, (D) Guettarda elliptica Sw. Arrows indicate the growth ring boundaries. All pictures � 40.

F.A. Roig et al. / Dendrochronologia 22 (2005) 187–193192

Therefore our observations are quite limited and muchfurther work is necessary to establish the full dendro-chronological potential in these environments.The data presented above suggest two major strate-

gies for future work. One would be to focus on thosespecies forming distinct, and probably annual, growthrings that could be used for dendrochronological studiessensu stricto. In this sense, recommended species forstudy include Cordia dodecandra A.DC., Gymnopodium

floribundum, Thouinia paucidentata and Talisia olivae-

formis (Kunth) Radlk., several leguminous species(Acacia, Caesalpinia, Leucaena, Piscidia, etc.), meliaceas(Trichilia hirta L., Cedrela odorata L.) and rhamnaceas(Karwinskia humboldtiana (J.A. Schultes) Zucc., Colu-

brina elliptica (Sw) Brizicky & W.L. Stern). Such speciescould provide data on tree ages and forest productivityof stands with additional possibilities for reconstructingclimate records in some cases. For this purpose it isunquestionable that further dendrochronological studiesare needed in order to explore in detail into the nature ofthe ring formation. The annuity of the growth rings isthe necessary prerequisite for the success of the cross-dating approach and the development of climatereconstructions. A second field of study would bedirected to the use of those species with ring boundariesdiscerned with some difficulty, in order to documentapproximate tree ages and biomass volume sufficient toestimate forest production at a broad scale.

Acknowledgements

We wish to express our gratitude to Jose Castillo-Caamal from the Universidad Autonoma de Yucatanand Jose Antonio Gonzalez-Iturbe from Centro deInvestigacion Cientıfica de Yucatan, Merida, for valu-able field guidance and to Rafael Bottero fromIANIGLA for graphics. Funds were provided by theInter-American Institute for Global Change Researchand the US National Science Foundation to support the

sampling program by members of the IAI collaborativeResearch Networks CRN001 and CRN003 followingthe 3rd CRN PI/IAI–IGBP meeting (‘‘Building GlobalChange Networks in the Americas’’) held in Mendoza,27–28 January 2003.

References

Aguilera HN. Suelos de la Penınsula de Yucatan. In: Los

recursos naturales del sureste y su aprovechamiento, vol.

II–V, Inst. Nacional Recursos Naturales Renovables, 1958.

p. 177–212.

Alves ES, Angyalossy-Alfonso V. Ecological trends in the

wood anatomy of some Brazilian species. 1. Growth rings

and vessels. IAWA Journal 2000;21:3–30.

Alvim PT. Tree growth periodicity in tropical climates. In:

Zimmermann MH, editor. The formation of wood in forest

trees. New York: Academic Press; 1964. p. 479–96.

Bhattacharyya A, Yadad RR, Borgaonkar HP, Pant GB.

Growth ring analysis of Indian tropical trees: dendrocli-

matic potential. Current Science 1992;62:736–41.

Bormann FH, Berlyn G, editors. Age and growth rate of

tropical trees: new directions for research. New Haven:

Yale University Press; 1981 137pp.

Buckley BM, Barbetti M, Watanasak M, D’Arrigo RD,

Boonchirdchoo S, Sarutanon S. Dendrochronological

investigations in Thailand. IAWA Journal

1995;16:393–409.

Chowdhury KA. Growth rings in tropical trees and taxonomy.

10th Pacific Science Congress. Hawai: Honolulu; 1961.

Eckstein D, Sass U, Baas P, editors. Growth periodicity in

tropical trees. IAWA Journal 1995;16:323–442.

Flores JS. Leguminosae. Florıstica, Etnobotanica y Ecologıa.

Etnoflora Yucatanense No. 18, Universidad Autonoma de

Yucatan, Merida, Yucatan, 2001. 320pp.

Flores JS, Espejel IC. Tipos de vegetacion de la Penınsula de

Yucatan. Etnoflora Yucatanense 1994;3 135pp.

Garcıa E. Modificaciones al sistema de clasificacion climatica

de Koppen. Apuntes de climatologıa. Talleres Larios.

Mexico, DF 1973.

Gonzalez-Iturbe JA, Olmsted I, Tun-Dzul F. Tropical dry

forest recovery after long term Henequen (sisal, Agave

ARTICLE IN PRESSF.A. Roig et al. / Dendrochronologia 22 (2005) 187–193 193

fourcroydes Lem.) plantation in northern Yucatan, Mex-

ico. Forest Ecology and Management 2002;167:67–82.

IAWA. IAWA list of microscopic features for hardwood

identification. IAWA Bulletin 1989;10:219–332.

Jacoby GC. 1989. Overview of tree-ring analysis in tropical

regions. IAWA Bulletin 1995;10:99–108.

Metcalfe CR, Chalk L. Anatomy of the Dicotyledons. 2nd ed.,

vol. 2. Oxford: Clarendon Press; 1983. 297pp.

Mosino PA, Garcıa E. The climate of Mexico. In: Bryson RA,

Hare FK, editors. Climates of North America. Amsterdam,

London, New York: Elsevier; 1974. p. 345–409.

Pant GB. Climatological signals from the annual growth rings

of selected tree species of India. Mausam 1983;34:251–6.

Pumijumnong N, Eckstein D, Sass U. Tree-ring research on

Tectona grandis in northern Thailand. IAWA Journal

1995;16:385–92.

Roig FA. Dendrocronologıa en los bosques del Neotropico:

revision y prospeccion futura. In: Roig FA, editor.

Dendrochronology in Latin America. Mendoza: EDIUNC

Pub.; 2000. p. 307–55.

Schongart J, Junk WJ, Piedade MTF, Ayres JM, Huttermann

A, Worbes M. Teleconnection between tree growth

in the Amazonian floodplains and the El-Nino-Southern

Oscillation effect. Global Change Biology 2004;1:

683–92.

Stahle DW, Mushove PT, Roig FA, Cleveland MK. Manage-

ment implications of annual growth rings in Pterocarpus

angolensis from Zimbabwe. Forest Ecology and Manage-

ment 1999;4760:1–13.

Worbes M. One hundred years of tree-ring research in the

tropics–A brief history and an outlook to future challenges.

Dendrochronologia 2002;20:217–31.