Forest Ecology and Management 334 (2014) 331–343

Contents lists available at ScienceDirect

Forest Ecology and Management

journal homepage: www.elsevier .com/locate / foreco

Recovery and early succession after experimental disturbancein a seasonally dry tropical forest in Mexico

http://dx.doi.org/10.1016/j.foreco.2014.09.0180378-1127/� 2014 Elsevier B.V. All rights reserved.

⇑ Corresponding author.E-mail address: mavaldez@ecosur.mx (M. Valdez-Hernández).

Mirna Valdez-Hernández a,⇑, Odilón Sánchez b, Gerald A. Islebe a, Laura K. Snook c,Patricia Negreros-Castillo d

a El Colegio de la Frontera Sur, Av. del Centenario Km. 5.5, Chetumal, Quintana Roo C.P. 77014, Mexicob Centro de Investigaciones Tropicales, Universidad Veracruzana, Xalapa, Mexicoc Bioversity International, Via dei Tre Denari 472/a, Maccarese (Rome), Italyd Instituto de Investigaciones Forestales, Universidad Veracruzana, Xalapa, Mexico

a r t i c l e i n f o a b s t r a c t

Article history:Received 31 May 2014Received in revised form 1 September 2014Accepted 4 September 2014

Keywords:Semi-deciduous forestNatural regenerationSilvicultureSecondary forestYucatan peninsula

We studied succession over five years in a seasonally dry tropical forest in Quintana Roo, Mexico, follow-ing three different types of experimental disturbance (slashing and complete felling; slashing, felling andburning; and machine-clearing), each one implemented in 1996 on two 0.5 ha treatment plots. Beforeexperimental disturbances, the floristic composition, dominance and diversity of the forest vegetationhad been determined. In 1997, after treatments were applied, a second survey characterized early sec-ondary vegetation at one year. A third survey was conducted in 2001. The 1996 vegetation compositionrevealed no significant differences among the six treatment plots. In 1997, floristic composition on the sixtreatment plots showed differences in dominance and diversity: the post-treatment vegetation on theslash/fell treatment was clearly distinct from that on the other two treatments. In 2001, differencesamong the plots had decreased considerably. Comparisons among seral stages revealed that one-year-old secondary vegetation differed from the pre-disturbance original vegetation, while five-year-old veg-etation was similar to the original in its diversity, floristic composition and dominance. Felling alonefavors species with a high resprouting capacity. The frequency of species with resprouting capacitywas lower on slash/fell/burn treatments and lowest on machine-cleared plots. Results indicate that theeffect of disturbance tends to decline over time and that complete clearing of small areas is effectiveas a silvicultural treatment to favor regeneration of valuable timber species and sustain diversity.

� 2014 Elsevier B.V. All rights reserved.

1. Introduction

Seasonally dry tropical forests (SDTF) comprise 42% of all trop-ical vegetation types in the world (Murphy and Lugo, 1995) andsuffer severely from deforestation. In Mesoamerica, the originalarea covered by SDTF was 550,000 km2 and only 2% are well con-served (Miles et al., 2006). The ever-increasing fragmentation oftropical forests urges the study of the regeneration process in for-est ecosystems under timber management. Regeneration processstudies can help us to predict which species will colonize disturbedareas, as well as the future species composition in these areas(Westman, 1990).

The ability of species to establish immediately after a distur-bance is determined in large part by their presence on the site(as stumps, roots, seeds or seedlings) as well as their capacity to

arrive there. The ability of species to maintain themselves as partof a new stand also depends on individual structural characteristicsand competitive capacities (Yih et al., 1991; Williams-Linera et al.,1997, 2011; Gaudet and Kaedy, 1988; Vandermeer et al., 2000;Dupuy et al., 2012; López-Martínez et al., 2013). Floristic composi-tion is influenced by the nature of the disturbance (magnitude, fre-quency and intensity) and the condition of a community when thedisturbance occurs (Kellman, 1970). Natural regeneration has beenstudied more frequently in moist tropical forests than in SDTF(Vieira and Scariot, 2006). In SDTF, species present adaptations toseasonal drought and the regeneration process is relatively rapid,facilitated by the high proportion of small seeds dispersed by wind(Gentry, 1995). Many species are able to resprout after a distur-bance event (Ewel, 1980; Murphy and Lugo, 1986; Kennard,2002; Negreros-Castillo and Hall, 2000; McLaren and McDonald,2003; Vieira et al., 2006). A relatively simple structure and lowdiversity (Murphy and Lugo, 1986; Kennard, 2002) give SDTF ahigh recovery capacity (Vieira et al., 2006).

332 M. Valdez-Hernández et al. / Forest Ecology and Management 334 (2014) 331–343

Two techniques have commonly been used to study secondarysuccession: (1) chronosequences and (2) vegetation dynamicsstudies. Chronosequence studies evaluate changes based on a com-parison of various plots of different seral ages, while vegetationdynamics studies document the development of the vegetation inpermanent plots, through remeasurements (Chazdon et al.,2007). Most studies in tropical vegetation have been carried outon chronosequences (Westman, 1990; Williams-Linera, 1990;Randall and Pickett, 1992; Snook, 1993; Denslow, 1995;Williams-Linera et al., 2011; Dupuy et al., 2012; López-Martínezet al., 2013). The main disadvantage of these studies is that theyuse spatially discrete locations to represent changes taking placeover time (Pickett, 1989), so data show the net cumulative effectof both spatial and temporal differences. Dynamic vegetation stud-ies examine the gradual shifts over time and can reveal more aboutecological processes that produce cumulative changes (Chazdonet al., 2007). Vegetation dynamics studies have been carried outin Mexico (Miranda et al., 1960; Sarukhán, 1964; Sánchez andIslebe, 1999; Valdez-Hernández, 1999) as well as in Central Amer-ica, notably to document recovery of forests in Nicaragua after hur-ricane Joan (Vandermeer et al., 1990, 2000; Yih et al., 1991).

In this paper, we analyze the species composition of seasonallydry forest on plots before and after three experimental distur-bances: (1) slash and burn; (2) slash and fell; and (3) mechanizedclearing. These were applied as silvicultural treatments to favor theregeneration of valuable timber species in managed forests. Weevaluated the composition, structure and diversity of regeneratingvegetation at one and five years after the treatments.

2. Methods

2.1. Study area

The Yucatan peninsula includes one of the largest continuoustracts of SDTF, making it a priority area for conservation (Mileset al., 2006). Seventy-four percent of Quintana Roo is covered byforest (Díaz-Gallegos et al., 2008). Most of the forest area is undermanagement by Mayan communities (Jhones et al., 2000; Brayet al., 2004). Although Quintana Roo presents the lowest deforesta-tion rate of the Mesoamerican Biological Corridor (0.6% year�1), atleast 23% of its territory is covered by secondary vegetation (Díaz-Gallegos et al., 2008). Historically, the forests of the region haveexperienced human impacts like the traditional slash and burnmethod of clearing agricultural fields and, more recently, timberextraction (Snook, 1998; Bray et al., 2004). Large scale agriculturalclearing and livestock production have affected smaller areas inYucatan (Jhones et al., 2000). Natural disturbances, notably hurri-canes, are also frequent. In the last century at least 100 hurricaneshave been recorded as affecting Quintana Roo (Boose et al., 2003).Forest fires are also common. In recent years Quintana Roo wasamong the ten states with the largest area affected by fire(CONAFOR, 2009, 2010, 2011).

This research was conducted in X-pichil (19�410 N, 88�220 W), aMayan community located in the central region of the Mexicanstate of Quintana Roo (Fig. 1). The community controls an area of31,314 ha, of which nearly 30,000 ha have been defined as produc-tion forest. These forests are included within the Calakmul-SianKaán portion of the Mesoamerican Biological Corridor (Díaz-Gallegos et al., 2008). Floristic richness at the local scale is between80 and 120 tree species per hectare (Cortés-Castelán and Islebe,2005; Sánchez et al., 2007).

The dominant soils are lithosols associated with rendzinas andluvisols (CEEM, 1987). The climate is sub-humid warm with asummer precipitation regime (García, 1988). The mean annualtemperature is 27.3 �C, and the mean annual precipitation is

1035.6 mm. The study area has a dry-warm season between Febru-ary and May, and a rainfall season between June and October(INEGI, 1990). Based on the precipitation and temperature regime,the vegetation can be classified as Seasonal Dry Tropical Forests(Murphy and Lugo, 1995).

Local communities produce corn and associated crops on fieldscleared using slash and burn methods, felling the understory with‘‘machetes’’ and large trees with axes, leaving the vegetation to dry.Once dry, the vegetation is burned before the rainy season. Thearea is usually used for agriculture for one to three years andallowed to fallow for 7–15 years, depending on land availability.Local agriculture is subsistence-oriented and agricultural fields or‘‘milpas’’ average about one hectare in size (Faust, 2001). Forestextraction began in the seventeenth century and focused on theextraction of large individuals of mahogany (Swietenia macrophyllaKing) (Snook, 1998). Since 1950 hardwood species have also beenexploited in X-pichil, originally to make railroad ties, notably ‘che-chen’ (Metopium brownei (Jacq.) Urb.), ‘jabin’ (Piscidia piscipula (L.)Sarg.), ‘chakte-kok’ (Simira salvadorensis (Standl.) Steyerm.), ‘cha-kte-viga’ (Coulteria platyloba (S. Watson) N. Zamora), ‘yaxnik’ (Vitexgaumeri Greenm.), ‘boop’ (Coccoloba spicata Lundell), ‘catalox’(Swartzia cubensis (Britton & P. Wilson) Standl.), ‘tzalam’ (Lysilomalatisiliquum (L.) Benth.), ‘pucte’ (Bucida buceras L.), and ‘kaniste’(Pouteria campechiana (Kunth) Baehni) (Murphy, 1990). The studysite is located in the production forest of the X-pichil ejido. Thevegetation prior to the disturbances consisted of more than 100tree species (Murphy, 1990; Valdez-Hernández, 1999; Sorensen,2006). According to local residents, the trees were at least 50 yearsold. A detailed description of the study area can be found in Snookand Negreros-Castillo (2004).

2.2. Sampling design and treatments

2.2.1. Pre-disturbance samplingEarly in 1996, six treatment plots of 5000 m2 each (50 m N–

S � 100 m E–W) were laid out in well conserved forest vegetation.Distances between plots were less than 500 m, and soil and envi-ronmental conditions were similar. In each treatment plot, twosample plots of 500 m2 (50 � 10 m) were established. In thesesample plots, before the clearing treatments were carried out, flo-ristic composition was sampled in the following subplots: arboreallayer (dbh P3 cm) in two sample plots of 500 m2; seedling layer(10 cm 6 height < 1 m) in five randomly selected subplots of1 m2. For each individual, species, life-form, and dbh wererecorded. In addition, the number of individuals of each speciesper plot was estimated. Density was estimated considering genetsas one individual if the multiple stems were rooted in one subplot.Once all the vegetation data was recorded, two treatment plotsreceived each of the three experimental disturbances describedbelow, applied as silvicultural treatments to regenerate valuabletimber species (see Snook and Negreros-Castillo, 2004):

� Slash and Burn (B): slashing of understory vegetation and man-ual felling of trees. After drying for several weeks, the debriswas burned, the treatment used locally to open agriculturalclearings.� Slash and Fell (F): this treatment included slashing and felling,

but debris was left on site.� Machine clearing (M): tree farmers or Caterpillar tractors were

used to knock over trees and underbrush, uprooting them andpushing them to the side of the plot, leaving the soil free ofvegetation.

2.2.2. Post-disturbance samplingIn 1997 and 2001, post-disturbance regeneration was sampled

on two permanent sample plots of 100 m2 (10 � 10 m) laid out

Fig. 1. Location of the study area in X-pichil ejido, Quintana Roo, Mexico.

M. Valdez-Hernández et al. / Forest Ecology and Management 334 (2014) 331–343 333

in the central part of each 500 m2 sample plot established in 1996.This gave a total of twelve post-disturbance sample plots, two100 m2 sample plots on each treatment plot, for a total of four sam-ple plots in each type of disturbance. In both 1997 and 2001 indi-viduals were sampled using the following size categories:seedlings (10 cm 6 height < 1 m) were sampled in five subplots of1 m2 distributed randomly in each plot. Individuals P3 cm indbh were sampled across the full 100 m2 plot. Within the sampleplots, all individuals were identified to species and life form anddbh was measured for stems >1.3 m height. For sprouts the densitywas estimated considering all sprouts from one stem or root as oneindividual regardless of the number of sprouts. The basal area (BA)was determined for each stem >3 cm in dbh. The total basal areafor each individual was derived from the sum of the BA of all thestems from a single base.

2.3. Data analysis

Change in dominance of individuals <1 m height was evaluatedusing the distribution index (DI) derived from the density and fre-quency of each species for each sample year (Sarukhán, 1964)according to the formula:

DI = density ⁄ frequency, whereFrequency = (plots where present/total plots) ⁄ 100Density = number of individuals/area sampled

Dominance of species for individuals >3 cm dbh was calculatedbased on the importance value (IV), an index derived from density,frequency and BA of each species (Kent and Coker, 1996), accordingto the formula:

IV = relative density + relative basal area + relative frequencyWhere:Relative density = (number of individuals of a species /totalnumber of individuals) ⁄ 100Relative BA = (BA of a species/BA of all species) ⁄ 100Relative frequency = (frequency of a species/ frequency of allspecies) ⁄ 100

Diversity was calculated using the Shannon–Wiener index. Thisindex combines two aspects of diversity: species number andevenness (Magurran, 1996).

For data analysis, a generalized linear model (GLM) was applied,using repeated measures and an a posteriori Tukey test, based on anested block design and using years, disturbance and plot as blocks.Sample plots were handled as repetitions using the same code forboth sample plots of each treatment. A nested design was usedbecause the response of treatments depends on the initial condi-tions of the plots. Variables analyzed were diversity, BA, densityand percent of individuals derived from sprouts. All statistical anal-yses were performed with the Statistica program version 12.

Finally, the density per species was used to run detrended cor-respondence analysis (DCA) (Hill and Gauch, 1980), using PC-ORD5.1 (McCune and Mefford, 2006). Ordination was applied for eachyear (1996, 1997, 2001), and one ordination included all plots inorder to evaluate similarity between vegetation characteristics onareas affected by different types of disturbance.

3. Results

3.1. Floristic composition and life forms

In total, among all three sampling periods and plots, 179 specieswere registered, of which 135 were identified to species using her-barium specimens and 44 were identified to family or genus orcould not be identified and were considered as morphospecies.The list of species, families and life forms identified during theyears 1996, 1997 and 2001 is presented in Appendix A. Beforethe application of treatments we found the lowest species rich-ness: 81 species from 67 genera and 44 families. Families withthe highest number of species in the pre-disturbance forest(1996) were: Fabaceae, Sapindaceae, Polygonaceae, Rubiaceae,Euphorbiaceae, and Bignoniaceae (Appendix A).

The highest species richness was registered one year after thetreatments, in 1997 (Appendix A): 158 species from 113 generaand 55 families. Families with the highest number of species inall treatments were Fabaceae, Rubiaceae and Bignoniaceae. Themost important difference among treatments was the presence of

334 M. Valdez-Hernández et al. / Forest Ecology and Management 334 (2014) 331–343

a higher number of species of the family Asteraceae and Solanaceaeafter the machine clearing.

In 2001, five years after the disturbance, we registered 109 spe-cies from 76 genera and 44 families (Appendix A). The followingfamilies were the most abundant in each treatment. Slash and fell:Fabaceae, Bignoniaceae, Rubiaceae, Sapindaceae, Sapotaceae; slashand burn: Fabaceae, Bignoniaceae, Myrtaceae, Polygonaceae;machine clearing: Fabaceae, Rubiaceae, Euphorbiaceae, Sapinda-ceae, Polygonaceae, Myrtaceae.

In 1996, before the experimental disturbances, the dominantlife forms were trees, lianas and herbs (Table 1). Of these, threearboreal species (Ceiba aesculifolia, Guapira costaricana, Ficus sp)and one herbaceous species (Neea psychotrioides) were only pres-ent in the pre-disturbance vegetation. In 1997, after the experi-mental disturbances, the percentages of trees, lianas and herbsdecreased, particularly in machine-made clearings (Table 1). Ninespecies of trees and 25 species of lianas were found only in theone-year-old vegetation. In 2001, all three treatments had similarproportions of life forms and the percentages were similar to thosebefore the disturbances (Table 1).

3.2. Density, diversity and basal area

A total of 7043 individuals were surveyed in all sampled years.Comparisons of density (total number of individuals on sampledplots, counting all species and life forms) among years revealedthat the one-year-old post-disturbance vegetation of 1997 hadthe lowest density (8.9 individuals m�2; F = 28.02, P < 0.001,df = 12). Comparison of the density among treatments and yearsrevealed significant differences (Fig. 2a; F = 1.92, P 6 0.05). Thelowest value was documented in 1997 in the burned treatments(5.5 individuals m�2; P 6 0.05); while in 1996 and 2001 the density(Fig. 2a) showed no significant differences among treatments(P P 0.05). The percentage of individuals derived from sprouts inthe vegetation in 1997 was higher than in 2001 (14.8% and 7.7%,respectively; Fig. 2b; F = 28.02, P < 0.001), so the regeneration fromsprouts was more important at one year. The proportion of stemsderived from sprouting was highest on felled plots in 1997(24.2%; Fig. 2b; P P 0.05). In 2001 the percentage of individualsderived from sprouts (Fig. 2b) showed no significant differencesamong treatments (11.7% on felled clearings, 6.9% on burned clear-ings and 4.5% on machine clearings; Fig. 2b; P P 0.05).

The comparison of diversity values (Shannon–Wiener index)among years revealed that the diversity was highest in 1997(3.3; Fig. 2c; F = 28.03, P < 0.001). Significant differences amongyears and treatments were found (Fig. 2c; F = 1.99, P 6 0.05). Diver-sity in 1997 in the machine clearings was low (3.1; P = 0.05); in1996 and 2001 the diversity (Fig. 2c) showed no significant differ-ences among treatments (P = 0.14).

Basal area (BA) varied by treatment and year (Fig. 2d), being low-est in 1997 (0.5 m2 ha�1; F = 28.02, P < 0.001). That year themachine clearings had the lowest BA (0.1 m2 ha�1; P 6 0.05). In2001, a considerable increase in BA was registered for alltreatments, but differences among types of disturbance were not

Table 1Percentage of individuals of each life form recorded in the pre-disturbance (1996) and po

Live form Pre-disturbance (%) Burned (%)

1996 1997 2001

Tree 70 54 70Liana 18 25 15Herb 5 10 5Shrub 5 8 8Others 2 3 2

statistically significant (Fig. 2d; P P 0.05). By 2001 the BA was sim-ilar to that documented prior to the disturbance, in 1996 (2001:10.8 m2 ha�1, 1996: 12.7 m2 ha�1; P P 0.05; Fig. 2d), largely dueto the number of sprouts.

Sprouting was common in several species: the species with thehighest resprouting response in the slash and burn treatment wereCroton sp., Hampea trilobata, Mosannona depressa, Piper sp., Thoui-nia paucidentata and Psidium sartorianum. In the felling treatment,many species sprouted from the cut stump of the trees, notablyGymnanthes lucida, Piper sp., Pouteria reticulata, Trichilia minutiflora,Casearia nitida, Nectandra coriacea, Manilkara zapota and Alseisyucatanensis. Species resprouting on machine clearings, probablyfrom roots, included N. coriacea, P. sartorianum, Coccoloba diversifo-lia and Caesalpinia gaumeri.

3.3. Dominance in the seedling layer (<1 m height)

Table 2 shows all species with high dominance values in thesize class <1 m height in each seral stage and treatment, basedon the Distribution Index (DI). In 1996, there were 12 dominantspecies among all six plots. Most dominant species were tree (9species), followed by lianas (2 species), and herbs (1 species). Thetree P. reticulata was calculated to have the highest DI.

In 1997, one year after clearing, the dominance valuesdecreased considerably and 19 dominant species were registeredamong the six treatment plots. Dominant species included 12trees, five lianas, two herbs, and one shrub (Table 2). N. coriaceashowed the highest DI. The post-disturbance vegetation analysisof one-year-old plots showed that the species with the highestdominance values on the felled treatments and those cleared bymachine were Panicum trichoides (herb) and T. minutiflora (tree).On burned plots N. coriacea and Petrea volubilis (liana) weredominant.

The dominance values in 2001 were higher than those regis-tered for previous years. We registered 15 dominant speciesamong the six treatment plots (Table 2). Dominance values werehighest for eight species of trees, followed by five species of liana.In addition, one herb species and one shrub species were docu-mented. The species with the highest DI was the tree P. reticulata,on plots cleared using all three treatments. On felled plots, Stizo-phyllum riparium (liana) was also important, while on burned plotsN. coriacea had high dominance values and on machine-clearedplots, M. zapota also had a high dominance value.

3.4. Dominance among individuals >1 m height

Species dominance for individuals >3 cm in dbh are presentedin Table 3. Before the treatment (1996), the three species withthe highest Importance Value (IV) were Sabal yapa, an understorypalm, and two canopy species, P. reticulata and M. zapota.

In the secondary vegetation one year after clearing (1997), thedominant species were Carica papaya, L. latisiliquum and P. piscipula(Table 3). In the burned and machine-cleared treatments, the dom-

st-disturbance vegetation (1997, 2001).

Felled (%) Machine-cleared (%)

1997 2001 1997 2001

58 74 46 6922 13 30 15

9 4 16 48 7 7 93 2 1 3

Fig. 2. (a) Average density of individuals for each treatment in sampled years. (b) Average percentage of individuals from sprouts for each treatment in sampled years. (c)Average diversity (Shannon–Wiener index) for each treatment in sampled years. (d) Average total basal area per ha of individuals P3 cm dbh for each treatment in sampledyears. PD = pre-disturbance, error bars are SE based on comparisons among four sample plots per treatment.

M. Valdez-Hernández et al. / Forest Ecology and Management 334 (2014) 331–343 335

inant species were L. latilisiquum and P. piscipula. In the felled treat-ment, the dominant species were C. papaya and P. piscipula.

The dominant species five years after clearing (2001) wereL. latisiliquum and P. campechiana (Table 3). The species with thehighest dominance in burned and machine-cleared treatmentswas L. latisiliquum. In the felled treatment, P. campechiana wasthe dominant species. In this seral stage the floristic compositionhad recovered. Eighty percent of the dominant species in the pre-disturbance forest were dominant five years after disturbance. Sev-eral tree species that were dominant prior to the experimental dis-turbances showed an increase in IV after five years: P. reticulata, M.zapota, T. minutiflora, S. salvadorensis, P. sartorianum, Exothea pan-iculata and Astronium graveolens. Conyza sp and C. papaya andshort-lived species such as Solanum hazenii and Neurolaena lobata,registered in 1997, were not recorded in 2001 (Table 3).

3.5. Indirect ordination analysis

The DCA calculated for each seral stage and between the seralstages is presented in Fig. 3(a–c). The analysis of the 1996 data,did not reveal well defined groups, indicating that sample plotswere not distinctive, so the graph is not presented.

For 1997 data, the eigenvalue of axis 1 was 0.367 and the valuefor axis 2 was 0.126. A clear separation could be observed whichdefined the two groups (Fig. 3a). Plots cleared using the fellingtreatment are on the right side of the x-axis; plots cleared usingburning are in the upper left part of the figure, with the exceptionof sample plot B4S in the lower left part, with the lowest percent-age of resprouting (7.2%, Fig. 3b), and the highest number of uniquespecies within that treatment block. Machine-cleared plots are dis-persed in the grouping. Machine clearing resulted in the lowestpercentage of resprouting (6.1%, Fig. 3b), so their regenerationwas less reflective of the predisturbance species composition andvaried more among plots. The x-axis indicates the disturbancegradient. Based on floristic composition and density, the felling

treatment caused the least change in species composition, mainlydue to the resprouting capacity of arboreal species including P. pis-cipula and G. lucida (Table 3).

In 2001 (Fig. 3b), the eigenvalues were high (axis 1, 0.611 andaxis 2, 0.333). A clear separation could be observed of nine plots.Independent of the treatment, they are grouped in the lower cen-tral part of the graph. Only three parcels are not clustered: twoplots on machine cleared treatments (M3 N, M5 N) and one plotcleared using burning (B4 N).This seems to show that the differen-tiating effect of disturbance type decreases with time.

The ordination for all sampled years (Fig. 3c) had high eigen-values on the axes 1 and 2, 0.523 and 0.345, respectively. Threegroups were identified according to their age: the 1996 vegetationis grouped in the left central part of the DCA plot; the 1997 vege-tation is located on the right central part; the third group, formedby the 2001 vegetation, is located in the lower central part. Thevegetation of 2001 showed recovery in density and floristic compo-sition. This recovery is seen in all plots, but mainly in the felledplots, with 24% species composition recovery since 1997, mainlydue to resprouting. Resprouting in the machine cleared plots wasmuch lower (6%), and originated from roots. DCA shows an age gra-dient along the first axis and reveals that the impact of disturbancedecreases over time.

4. Discussion

In the study area, local inhabitants consider that mature vegeta-tion 40 years old or more can be recognized by the presence of aclosed canopy made up of trees with large diameters, and anunderstory with juveniles of tree species and some palm species,without pioneer species. These characteristics were found in thepredisturbance vegetation, which was dominated by M. zapota, P.reticulata and S. yapa. P. reticulata has been reported as the mostimportant species in terms of basal area in forests of QuintanaRoo (Dickinson et al., 2001; Cairns et al., 2003). In the early seral

Table 2Average distribution index for dominant species < 1 m height on 10 m2 sample plots established on each pair of treatment plots prior to disturbance (1996), and on 10 m2 of plotsestablished on the same three pairs of treatment plots after Slash and Burn, complete Felling and Machine clearing one year (1997) and five years (2001) after treatments. Valuesare presented for each pair of treatment plots.

Species Distribution Index

Pre-disturbance Burned Pre-disturbance Felled Pre-disturbance Machine-cleared

1996 1997 2001 1997 2001 1997 2001

TreesPouteria reticulata 6.13 5.87 1.6 9.07Piper sp. 4.44 1.03 1.11 12.6 1.11Nectandra coriacea 2.2 1.55 2.4 0.86 5.8 1.03Gymnanthes lucida 3.33 2.13 0.85 4.19Manilkara zapota 1.78 1.63 2.13 0.59Trichilia minutiflora 1.33 1.56 2.44Calyptranthes pallens 0.27 0.79 1.18 0.4 2.17Hippocratea sp 1.78Piscidia piscipula 0.87 0.85Guettarda combsii 0.86 0.37Eugenia capuli 0.93Laetia thamnia 0.8Lonchocarpus rugosus 0.2 0.57Caesaria nitida 0.74Psidium sartorianum 0.36 0.3Diospyrus salicifolia 0.32Coccoloba diversifolia 0.21

PalmsSabal yapa 0.27

LianasBignoniaceae1 2.49 1.25 6.87 2.18 0.52 1.33 1.24 4.65Stizophyllum riparium 1.6 2.76 3.2 0.55 1.21 2.2 0.33 1.03Hiraea obovata 0.45 1.63 1.77 0.71 2.66Cissus biformifolia 0.45 0.45Petrea volubilis 0.69Bignoniaceae2 0.43Bignoniaceae3 0.38

HerbsLasiacis ruscifolia 12.5 0.67 17.3 20.1Poaceae 8.8 1.71 6.6Panicum trichoides 1.18 2.18Justicia breviflora 0.42

336 M. Valdez-Hernández et al. / Forest Ecology and Management 334 (2014) 331–343

stage, the families Fabaceae, Poaceae and Asteraceae dominated.Several genera found in the secondary vegetation of 1997 and2001, such as Bauhinia, Croton, Cecropia, Eupatorium, Hamelia,Lasiacis, Panicum, Piper, Solanum, Spondias and Trema, have beenidentified as pioneers in different seral stages in the Neotropics(Whitmore, 1975; Halle et al., 1978; Chavelas and Contreras,1990; White and Hood, 2004; Williams-Linera et al., 2011). How-ever, tree species from mature vegetation were also registered inthese early stages, including M. zapota, A. graveolens, Dendropanaxarboreus, Bursera simaruba, P. reticulata and M. brownei (Dickinsonet al., 2001; Cairns et al., 2003), which made up 75% of the numberof species found in the three samples. These six species had respro-uted, which allowed their rapid recovery after clearing. Other stud-ies in SDTF have shown that resprouting is a common and effectivemechanism of regeneration (Ewel, 1980; Murphy and Lugo, 1986;Negreros-Castillo and Hall, 2000; Kennard, 2002; McLaren andMcDonald, 2003; Vieira et al., 2006; Sampaio et al., 2007; López-Martínez et al., 2013). Another study, in gaps caused by similar dis-turbances, found that 81% of the species had resprouted within14 months after cutting stems of dbh >2 cm (McLaren andMcDonald, 2003). Kennard (2002) reported that in post-fire treat-ments 60% of all individuals of 2.5 m height had resprouted18 months after intense fire, while low intensity fires resulted inresprouting among 90% of the individuals.

L. latisiliquum was one of the species with high IV after distur-bance, mainly after burning and machine clearing (Table 3). In thisspecies no resprouting was observed; it seems new individualsoriginated from seed. In dry forests in Mexico where slash and

burn activities are applied, fire has been found to reduce seedbanks up to 93% (Miller, 1999). The regeneration of trees derivefrom seeds fallen from nearby trees around the experimental clear-ings or dispersed into the clearings by animals. In another study ofdisturbed sites in Yucatan, however, L. latisiliquum was the onlytree of which seeds were detected in the soil, along with abundantseeds of herbs (Rico-Gray and García-Franco, 1992), so its seedsmay have remained in the seed bank undamaged by fire.

One year after clearing, sprouting seemed to vary among clear-ing treatments. On the burned treatment Croton sp., H. trilobata, M.depressa, Piper sp., T. paucidentata and P. sartorianum were the mostabundant resprouters. The most abundant sprouters on machinecleared treatments were N. coriacea, P. sartorianum, C. diversifoliaand Caesalpinia gaumeri. Sprouting on these two treatments camemostly from roots. On the felling treatment, where trees werecut off at stump height and no additional disturbance wasimposed, a number of canopy species resprouted, notably M.zapota, A. yucatanensis, P. reticulata and T. minutiflora as well as G.lucida, Piper sp., C. nitida and N. coriacea.

In 1996, the most abundant life form was trees, but after theclearing treatments life forms were distributed among trees, lianasand herbs (Table 1). The most likely reason that herbaceous speciesand lianas have more possibilities of rapid establishment(Schnitzer et al., 2000) is that in addition to their vegetative regen-eration capacity, most of these species have small seeds (Vieiraet al., 2006) and rapid growth (Schnitzer et al., 2000). Seeds withlow water content are less susceptible to dehydration in open areas(Holl, 1999). The high number of lianas in 1997 supports the obser-

Table 3Dominant species >3 cm dbh based on calculations of pre-disturbance Importance Value (1996), using averages between two 500 m2 pre-treatment plots; and amongindividuals >1 m height at one year (1997) and five years (2001) after clearing, on the same sample plots, derived from average values among a total of four 100 m2 sample plotsestablished on the two replicate treatment plots subjected to each type of disturbance.

Species Importance value

Pre-disturbance Felled Pre-disturbance Burned Pre-disturbance Machine-cleared

1996 1997 2001 1997 2001 1997 2001

TreesLysiloma latisiliquum 5.61 2.91 3.34 1.23 53.8 47.4 1.23 58.9 49.9Piscidia piscipula 9.24 13.5 18.1 12.49 34.5 27.2 17.13 31.3 26.8Pouteria reticulata 37.21 4.53 6.32 19.77 1.12 2.72 34.26 1.12 3.3Alseis yucatanensis 17.14 10.1 10.3 26.21 0.84 0.49 21.91 0.82 0.49Manilkara zapota 17.72 5.29 6.27 34.64 1.84 1.71 31.01 1.54 3.73Cecropia peltata 0.83 10.42 6.6 14.7 11.8 24.3Gymnanthes lucida 19.94 15.4 17.2 2.88 1.12 3.91 1.85 1.62 2.87Caesalpinia gaumeri 11.11 6.24 5.97 19.07 4.53 5.09 11.5 3.78 9.45Bursera simaruba 12.03 8.04 10.2 9.81 2.26 3.42 10.97 5.9 4.32Coccoloba spicata 11.79 1.68 8.4 12.82 8.85 5.12 11.12 2.97 6.36Nectandra coriacea 1.23 10.6 11.8 1.23 5.4 9.6 6.12 3.54 9.03Pouteria campechiana 9.2 10.7 20.3 9.19 3 3.29 8.75 1.4 13.5Hampea trilobata 8.08 9.33 6.01 15.6 4.24 6.25Vitex gaumeri 7.24 6.41 6.99 1.85 5.18 4.54 3.92 7.1 22Trichilia minutiflora 17.47 4.46 6.39 2.47 2.66 3.92 16.66 2.24 6.39Guettarda combsii 3.76 2.44 6.3 2.86 4.91 13.8 3.06 2.9 11.2Trema micrantha 3.25 6.39 9.58 6.37 7.38 3.25Simira salvadorensis 3.09 2.56 11.7 4.99 0.56 2.79 11.53 0.56 10.38Laetia thamnia 10.29 0.96 2.99 11.01 0.56 0.73 9.13 0.56 2.99Luehea speciosa 1.23 1.95 2.39 7.32 7.84 8.78 1.23 3.78 2.39Psidium sartorianum 4.17 0.84 1.71 6.54 5.65 5.47 7.37 0.84 10.7Exothea paniculata 6.55 3.67 13.9 4.19 1.42 1.66 9.53 1.12 7.27Croton sp. 7.15 5.24 5.9 3.07 2.23 5.24Solanum hazenii 1.96 5.49 11.6Exothea diphyla 0.62 2.18 2.56 0.62 0.83 0.98 11.47 0.84 2.56Neurolaena lobata 1.81 2.69 9.25Astronium graveolens 5.5 2.49 8.85 2.47 0.83 2.19 7.54 0.84 11.52Eupatorium odoratum 3.63 1.68 5.11 4.13 8.47 1.68Lonchocarpus rugosus 0.62 2.02 3.08 0.62 2.85 5.88 2.08 3.27 8.3Thevetia gaumeri 1.90 2.08 2.31 7.53 1.82 2.08Piper sp. 3.23 6.91 2.02 6.46 1.68 6.91

HerbsCarica papaya 66.5 37.2 24.9Conyza sp. 1.68 8.85 2.97

PalmsSabal yapa 38.41 0.69 3.1 76.04 0.86 6.13 28.71 0.56 3.1

M. Valdez-Hernández et al. / Forest Ecology and Management 334 (2014) 331–343 337

vation that disturbances maintain liana diversity, as their require-ments include high light levels (Schnitzer and Bongers, 2002).

The basal area values of the 1996 vegetation were low (8.7–14.7 m2 ha�1; these values show the range of variation in all plots,Fig. 2d). However, they are within the range of data recorded forother tropical forests from the region. Basal area values between20.7 and 28.4 m2 ha�1 are reported from Yucatan Peninsula(White and Hood, 2004). In medium statured forest of northernQuintana Roo 32.6 m2 ha�1 (Sánchez, 2000) and 22.4 m2 ha�1 (LaTorre-Cuadros and Islebe, 2003) are reported. For the transitionbetween medium and low forest 13.9 m2 ha�1 has been docu-mented (La Torre-Cuadros and Islebe, 2003). In 1997, the basal areawas close to zero in the felling and burning treatments and in themachine clearing plots the dbh had not reached 3 cm (Fig. 2d). In2001 we found similar basal area values to the pre-disturbancevegetation (2001: 10.8 m2 ha�1, 1996: 12.7 m2 ha�1) which canbe explained by the high number of sprouts, producing a hugeincrease in the number of stems.

In the first year of the recovery (1997), the type of disturbanceinfluenced the dominant species and diversity. At this early stage,the felling and the burning treatments showed the fastest recoveryin species composition, influenced by remnant individuals. Otherstudies have shown that, in SDTFs, resprouting after slash and slashand burn can accelerate recovery (Miller and Kauffman, 1998;Kammesheidt, 1999; Kennard et al., 2002) which is important in

the development of secondary successions. We consider this tobe the main reason why the felling and burning treatments pre-sented a more predictable vegetation recovery, while in themachine clearing treatment, the only remaining vegetative struc-tures were roots and rhizomes. Here, in the absence of stumps,many more individuals were able to colonize as seeds.

Although plots were close together and presented similar pre-cipitation and edaphic conditions, microenvironmental factorsexist in every type of treatment which interact with biotic condi-tions to facilitate or delay the vegetation development. Amongthe microenvironmental differences were the availability of waterand nutrients. The water availability could have been reduced inthe machine clearings as the soil had no protective layer of leaves,resulting in a higher rate of evaporation and the soil may have beensomewhat compacted, which could inhibit rapid germination ofseeds. Where vegetation had been slashed and burned, ash mayhave increased the water holding capacity of the soil. Furthermore,plots subjected to the burning treatment had more nutrients avail-able. It has been documented that after slash and burn, phosphorusavailability can double and there is a significant increase in theamount of nitrogen available in the first 5 cm of soil (Giardinaet al., 2000). It has been found that species richness and the num-ber of juveniles in dry forest is related to phosphate levels in soils(Ceccon et al., 2003), and access to water also affects the diversityof arboreal species, mainly juveniles (Williams-Linera and Lorea,

Fig. 3. Detrended Correspondence Analysis (DCA) of permanent plots; (a) ordina-tion 1997, (b) ordination 2001, (c) ordination among years of sampling (1996, 1997,2001). Treatment (B) slash and burn, (F) slash and fell, (M) machine clearing;treatment plots numbers (1–6); sample plots (N) north, (S) south; years of sampling(96) 1996, (97) 1997, (01) 2001.

338 M. Valdez-Hernández et al. / Forest Ecology and Management 334 (2014) 331–343

2009). The supply of nutrients depends on the quality of humusand decomposition rate (Bradshaw, 1997; Hasselquist et al.,2010) and water supply, as absorption of nutrients in plants is only

possible in a soluble state (Gibson and Nobel, 1986). The combina-tion of high evaporation with low nutrient supply in the machine-cleared plots could create a more stressful microenvironment. Thiscould have slowed the development of vegetation in the earlyyears. At the other extreme, on plots that had been slashed andfelled, abundant debris shaded the soil and contributed with leafand other organic materials to the soil. Experiments in Brazil haveshown that the use of organic matter as a soil cover can reduceevaporation by 40% compared to exposed soils. These differencesdiminish as leaf area and transpiration increase (Gomes et al.,2011).

Despite the differences among disturbances, the regrowingstands after five years had a similar floristic composition to thepre-disturbance vegetation (Appendix A). Even some species listedas threatened in the IUCN Red List for Mexico, notably A. graveolens,Simarouba amara Aubl. and Zamia loddigesii (SEMARNAT, 2002),reappeared. Recovery in SDTF has been proposed to be faster thanin moist tropical forests because they have a less complex structure(Ewel, 1980; Murphy and Lugo, 1986; Kennard, 2002).

Forests of the Yucatan Peninsula have been under human influ-ence for millennia (Carrillo-Bastos et al., 2010). Natural distur-bances like hurricanes and fires are frequent, but agriculture andforest management are also major drivers of forest composition.Our results reveal that making small clearings in the forest, likethose used in slash and burn agriculture, even if made using appar-ently destructive practices like uprooting the vegetation bymachine or burning it, do not threaten the diversity of the season-ally dry tropical forests in the region. Such clearings can be used assilvicultural treatments to foster natural regeneration of commer-cially important species while maintaining biodiversity. Fosteringregeneration of valuable species is important because a lack ofcommercial species may render forests more susceptible to con-version to nonforest land uses that are more lucrative (Putz andFredericksen, 2004). Currently the major threats to Quintana Roo’sforests are government development programs for the Yucatanpeninsula, which focus on large scale agricultural production andthe establishment of grasslands for animal husbandry. These pro-duction systems are intended to replace forests (Sohn et al.,1999), unlike slash and burn agriculture, a land use system whosesuccess is based on the continuous regeneration of the forest. Slashand burn agriculture could be combined with forest managementto ensure the regeneration of commercial species (M. zapota, L. lat-isiliquum, P. piscipula, D. arboreus, B. simaruba, M. brownei, amongothers), the conservation of biodiversity and the maintenance offorest cover.

Acknowledgements

CONACYT is acknowledged for financial support to the firstauthor. The establishment of the experimental treatments waspossible through support from the USDA Forest Service, underthe terms of agreement No. USDA-95-CA-118 with the WorldWildlife Fund, US. Financial support was also provided by the Bio-diversity Support Program, a consortium of World Wildlife Fund,The Nature Conservancy and the World Resources Institute, withfunding from the US Agency for International Development. Theopinions expressed in this article are those of the authors and donot necessarily reflect the views of the US Department of Agricul-ture, Biodiversity Support Program and US Agency for InternationalDevelopment. Thanks to the interest of Robert Petterson, Caterpil-lar Inc. contributed the time of bulldozers to create the machineclearings. Data collection was done with the support of NaDeneSorensen, Nicolas Can y Juan Can. Finally we thank the ejidoX-Pichil for their authorization and support for the establishmentof permanent sampling plots of 1996–2001 and two anonymousreviewers whose comments improved this manuscript.

M. Valdez-Hernández et al. / Forest Ecology and Management 334 (2014) 331–343 339

Appendix A

Complete list of species registered in the pre-disturbance(1996) and post-disturbance vegetation (1997, 2001). Family andlife form of each species is listed. Year of observation of each spe-cies indicated by an asterisk (⁄).

Family

Scientific name Lifeform

1996

1997 2001

Trees (T) and Palms (P)

Anarcadiaceae Astronium

graveolens Jacq.

T ⁄ ⁄ ⁄

Anarcadiaceae

Metopiumbrownei (Jacq.)Urb.

T

⁄ ⁄ ⁄

Annonaceae

Annona glabra L. T ⁄ Annonaceae Mosannona

depressa (Baill.)Chatrou

T

⁄ ⁄ ⁄

Apocynaceae

Thevetia gaumeriHemsl.

T

⁄ ⁄

Araliaceae

Dendropanaxarboreus (L.)Decne. & Planch.

T

⁄ ⁄ ⁄

Asteraceae

Porophyllumpunctatum (Mill.)S.F. Blake

T

⁄

Burseraceae

Bursera simaruba(L.) Sarg.

T

⁄ ⁄ ⁄

Burseraceae

Protium copal(Schltdl. &Cham.) Engl.

T

⁄ ⁄

Bixaceae

Cochlospermumvitifolium (Willd.)Spreng.

T

⁄

Ebenaceae

Diospyrossalicifolia Humb.& Bonpl. ex Willd.

T

⁄ ⁄ ⁄

Euphorbiaceae

Astrocasia sp T ⁄ ⁄ Euphorbiaceae Croton sp T ⁄ ⁄ ⁄ Euphorbiaceae Gymnanthes

lucida Sw.

T ⁄ ⁄ ⁄

Fabaceae

Acacia centralis(Britton & Rose)Lundell

T

⁄ ⁄

Fabaceae

Bauhiniadivaricata L.

T

⁄ ⁄ ⁄

Fabaceae

Caesalpiniagaumeri Greenm.

T

⁄ ⁄ ⁄

Fabaceae

Chloroleuconmangense (Jacq.)Britton & Rose

T

⁄

Fabaceae

Coulteriaplatyloba (S.Watson) N.Zamora

T

⁄ ⁄

Fabaceae

Lonchocarpusrugosus Benth.

T

⁄ ⁄ ⁄

Fabaceae

Lysilomalatisiliquum (L.)Benth.

T

⁄ ⁄ ⁄

Fabaceae

Lysiloma auritum(Schltdl.) Benth.

T

⁄

Fabaceae

Platymiscium T ⁄ ⁄

Appendix A (continued)

Family

Scientific name Lifeform

1996

1997 2001

yucatanumStandl.

Fabaceae

Piscidia piscipula(L.) Sarg.

T

⁄ ⁄ ⁄

Fabaceae

Swartzia cubensis(Britton & P.Wilson) Standl.

T

⁄ ⁄ ⁄

Fabaceae

Zygia recordiiBritton & Rose

T

⁄ ⁄

Celastraceae

Hippocrateafloribunda Benth.

T

⁄ ⁄ ⁄

Celastraceae

Hippocratea sp T ⁄ ⁄ ⁄ Lamiaceae Callicarpa

acuminata Kunth

T ⁄

Lamiaceae

Vitex gaumeriGreenm.

T

⁄ ⁄ ⁄

Lauraceae

Licariacampechiana(Standl.)Kosterm.

T

⁄ ⁄ ⁄

Lauraceae

Nectandracoriacea (Sw.)Griseb.

T

⁄ ⁄ ⁄

Malvaceae

Ceiba aesculifolia(Kunth) Britten &Baker f.

T

⁄

Malvaceae

Hampea trilobataStandl.

T

⁄ ⁄ ⁄

Malvaceae

Luehea speciosaWilld.

T

⁄ ⁄ ⁄

Malvaceae

Pseudobombaxellipticum(Kunth) Dugand

T

⁄ ⁄ ⁄

Meliaceae

Trichilia glabra L. T ⁄ ⁄ ⁄ Meliaceae Trichilia hirta L. T ⁄ ⁄ ⁄ Meliaceae Trichilia

minutifloraStandl.

T

⁄ ⁄ ⁄

Meliaceae

Trichilia moschataSw.

T

⁄

Moraceae

Brosimumalicastrum Sw.

T

⁄ ⁄ ⁄

Moraceae

Ficus sp T ⁄ Myrtaceae Calyptranthes

pallens Griseb.

T ⁄ ⁄ ⁄

Myrtaceae

Eugenia buxifoliaLam.

T

⁄ ⁄

Myrtaceae

Eugenia capuli(Schltdl. &Cham.) Hook. &Arn.

T

⁄ ⁄ ⁄

Myrtaceae

Psidiumsartorianum (O.Berg) Nied.

T

⁄ ⁄ ⁄

Nyctaginaceae

Guapiracostaricana(Standl.)Woodson

T

⁄

Picramniaceae

Alvaradoaamorphoides

T

⁄

(continued on next page)

340 M. Valdez-Hernández et al. / Forest Ecology and Management 334 (2014) 331–343

Appendix A (continued)

Family

Scientific name Lifeform

1996

1997 2001

Liebm.

Piperaceae Piper sp T ⁄ ⁄ ⁄ Polygonaceae Coccoloba

acapulcensisStandl.

T

⁄ ⁄ ⁄

Polygonaceae

Coccolobadiversifolia Jacq.

T

⁄ ⁄ ⁄

Polygonaceae

Coccoloba spicataLundell

T

⁄ ⁄ ⁄

Polygonaceae

Gymnopodiumfloribundum Rolfe

T

⁄ ⁄ ⁄

Polygonaceae

Neomillspaughiaemarginata (H.Gross) S.F. Blake

T

⁄ ⁄

Primulaceae

ArdisiaescallonioidesSchltdl. & Cham.

T

⁄ ⁄ ⁄

Rhamnaceae

Krugiodendronferreum (Vahl)Urb.

T

⁄ ⁄ ⁄

Resedaceae

Forchhammeriatrifoliata Radlk.

T

⁄ ⁄ ⁄

Rubiaceae

AlseisyucatanensisStandl.

T

⁄ ⁄ ⁄

Rubiaceae

Guettarda combsiiUrb.

T

⁄ ⁄ ⁄

Rubiaceae

Psychotria sp1 T ⁄ Rubiaceae Psychotria sp2 T ⁄ Rubiaceae Randia sp T ⁄ Rubiaceae Simira

salvadorensis(Standl.)Steyerm.

T

⁄ ⁄ ⁄

Rubiaceae

Rubiaceae1 T ⁄ Rutaceae Casimiroa

tetrameria Millsp.

T ⁄ ⁄

Salicaceae

Caesaria nitida(L.) Jacq.

T

⁄ ⁄ ⁄

Salicaceae

Laetia thamnia L. T ⁄ ⁄ ⁄ Salicaceae Zuelania guidonia

(Sw.) Britton &Millsp.

T

⁄ ⁄ ⁄

Sapindaceae

Allophyluscominia (L.) Sw.

T

⁄ ⁄

Sapindaceae

Exothea diphylla(Standl.) Lundell

T

⁄ ⁄ ⁄

Sapindaceae

Exotheapaniculata (Juss.)Radlk.

T

⁄ ⁄ ⁄

Sapindaceae

Talisia oliviformis(Kunth) Radlk.

T

⁄ ⁄ ⁄

Sapindaceae

ThouiniapaucidentataRadlk.

T

⁄ ⁄ ⁄

Sapotaceae

ChrysophyllummexicanumBrandegee exStandl.

T

⁄ ⁄

Appendix A (continued)

Family

Scientific name Lifeform

1996

1997 2001

Sapotaceae

Manilkara zapota(L.) P. Royen

T

⁄ ⁄ ⁄

Sapotaceae

Pouteriacampechiana(Kunth) Baehni

T

⁄ ⁄ ⁄

Sapotaceae

Pouteria reticulata(Engl.) Eyma

T

⁄ ⁄ ⁄

Sapotaceae

Sideroxylonsalicifolium (L.)Lam.

T

⁄

Simaroubaceae

Simarouba amaraAubl.

T

⁄ ⁄

Urticaceae

Cecropia peltata L. T ⁄ ⁄ Sp1 T ⁄ Sp2 T ⁄

Aracaceae

Chamaedorea sp P ⁄ Aracaceae Desmoncus

orthacanthosMart.

P

⁄ ⁄

Aracaceae

Sabal yapa C.Wright ex Becc.

P

⁄ ⁄ ⁄

Herbs (H) and Shrubs (S)

Acanthaceae Aphelandra scabra

(Vahl) Sm.

H ⁄

Acanthaceae

Justicia breviflora(Nees) Rusby

H

⁄

Amaryllidaceae

Habranthustubispathus(L’Hér.) Traub

H

⁄

Asteraceae

Ageratum sp H ⁄ ⁄ Asteraceae Conyza sp H ⁄ Asteraceae Tridax

procumbens L.

H ⁄

Boraginaceae

Cordia sp H ⁄ Euphorbiaceae Euphorbia

cyathophoraMurray

H

⁄

Euphorbiaceae

Euphorbia dentataL.

H

⁄

Malvaceae

Sida acuta Burm.f.

H

⁄

Nyctaginacea

NeeapsychotrioidesDonn. Sm.

H

⁄

Phytolaccaceae

Phytolaccarivinoides Kunth& C.D. Bouché.

H

⁄

Poaceae

Lasiacis ligulataHitchc. & Chase

H

⁄

Poaceae

Lasiacis ruscifolia(Kunth) Hitchc.

H

⁄ ⁄

Poaceae

Lasiacis sp H ⁄ ⁄ Poaceae Panicum

trichoides Sw.

H ⁄

Poaceae

Poaceae1 H ⁄ ⁄ Rubiaceae Hamelia sp1 H ⁄ ⁄ ⁄ Rubiaceae Hamelia sp2 H ⁄ Solanaceae Physalis sp1 H ⁄ Solanaceae Physalis sp2 H ⁄

M. Valdez-Hernández et al. / Forest Ecology and Management 334 (2014) 331–343 341

Appendix A (continued)

Family

Scientific name Lifeform

1996

1997 2001

Violaceae

HybanthusyucatanensisMillsp.

H

⁄ ⁄

Zamiaceae

Zamia loddigesiiMiq.

H

⁄ ⁄

Sp8

H ⁄ Apocynaceae Plumeria rubra L. S ⁄ ⁄ ⁄ Asteraceae Eupatorium

odoratum L.

S ⁄ ⁄

Asteraceae

Neurolaena lobata(L.) Cass.

S

⁄

Asteraceae

Verbesinagigantea Jacq.

S

⁄ ⁄

Cannabaceae

Trema micrantha(L.) Blume

S

⁄ ⁄

Caricaceaea

Carica papaya L. S ⁄ Euphorbiaceae Cnidoscolus

aconitifolius(Mill.) I.M. Johnst.

S

⁄ ⁄ ⁄

Malvaceae

Malvaviscusarboreus Cav.

S

⁄ ⁄ ⁄

Rubiaceae

Randia aculeata L. S ⁄ Rubiaceae Randia longiloba

Hemsl.

S ⁄

Rubiaceae

Randia sp1 S ⁄ Rubiaceae Randia sp2 S ⁄ Solanaceae Solanum asperum

Rich.

S ⁄ ⁄

Solanaceae

Solanum hazeniiBritton

S

⁄ ⁄

Solanaceae

SolanumrudepannumDunal

S

⁄

Lianas (L) and epiphytes (E)

Acanthaceae Mendoncia

lindavii Rusby

L ⁄

Apocynaceae

Echites tuxtlensisStandl.

L

⁄

Apocynaceae

Gonolobusstenanthus(Standl.)Woodson

L

⁄

Apocynaceae

Mandevillasubsagittata (Ruiz& Pav.) Woodson

L

⁄

Apocynaceae

Sarcostemmabilobum Hook. &Arn.

L

⁄

Araceae

SyngoniumpodophyllumSchott

L

⁄ ⁄

Bignoniaceae

Cydistaaequinoctialis (L.)Miers

L

⁄

Bignoniaceae

Dolichandraunguis-cati (L.)L.G. Lohmann

L

⁄ ⁄

Bignoniaceae

Mansoaverrucifera(Schltdl.) A.H.

L

⁄ ⁄

Appendix A (continued)

Family

Scientific name Lifeform

1996

1997 2001

Gentry

Bignoniaceae Stizophyllum

riparium (Kunth)Sandwith

L

⁄ ⁄ ⁄

Bignoniaceae

Bignonaceae 1 L ⁄ ⁄ ⁄ Bignoniaceae Bignonaceae 2 L ⁄ ⁄ ⁄ Bignoniaceae Bignonaceae 3 L ⁄ ⁄ ⁄ Bignoniaceae Bignonaceae 4 L ⁄ Celastraceae Crossopetalum

gaumeri (Loes.)Lundell

L

⁄ ⁄ ⁄

Connaraceae

Rourea glabraKunth

L

⁄ ⁄

Convolvulaceae

Ipomoea tricolorCav.

L

⁄

Convolvulaceae

Jacquemontiapentanthos (Jacq.)G. Don

L

⁄

Cucurbitaceae

Cucurbita sp L ⁄ Cucurbitaceae Melothria pendula

L.

L ⁄

Dioscoreaceae

DioscoreaspiculifloraHemsl.

L

⁄

Euphorbiaceae

Dalechampiascandens L.

L

⁄

Euphorbiaceae

Plukenetia sp L ⁄ ⁄ Euphorbiaceae Tragia

yucatanensisMillsp.

L

⁄ ⁄

Fabaceae

Bauhinia herrerae(Britton & Rose)Standl. &Steyerm.

L

⁄ ⁄

Fabaceae

Centrosema sp L ⁄ Fabaceae Dalbergia glabra

(Mill.) Standl.

L ⁄

Fabaceae

Desmodiumincanum (Sw.)DC.

L

⁄

Malpighiaceae

Bunchosiaglandulosa (Cav.)DC.

L

⁄ ⁄

Malpighiaceae

Hiraea obovataHuber

L

⁄ ⁄ ⁄

Malpighiaceae

Stigmaphyllon sp L ⁄ Menispermaceae Cissampelos

owariensis P.Beauv. ex DC.

L

⁄

Passifloraceae

Passiflora sp L ⁄ Passifloraceae Passifloraceae1 L ⁄ Rhamnaceae Sageretia sp L ⁄ Rubiaceae Morinda royoc L. L ⁄ Rubiaceae Margaritopsis

microdon (DC.)C.M. Taylor

L

⁄ ⁄

Rubiaceae

Notopleura sp L ⁄ Sapindaceae Paullinia clavigera

Schltdl.

L ⁄ ⁄ ⁄

Sapindaceae

Serjania L ⁄ (continued on next page)

342 M. Valdez-Hernández et al. / Forest Ecology and Management 334 (2014) 331–343

Appendix A (continued)

Family

Scientific name Lifeform

1996

1997 2001

goniocarpa Radlk.

Smilacaceae Smilax mollis

Humb. & Bonpl.ex Willd.

L

⁄

Smilacaceae

Smilax spinosaMill.

L

⁄ ⁄ ⁄

Solanaceae

Solanum sp L ⁄ Verbenaceae Petrea volubilis L. L ⁄ ⁄ ⁄ Vitaceae Cissus biformifolia

Standl.

L ⁄

Vitaceae

Cissus verticillata(L.) Nicolson &C.E. Jarvis

L

⁄

Sp3

L ⁄ Sp4 L ⁄ Sp5 L ⁄ ⁄ Sp6 L ⁄ ⁄ Sp7 L ⁄ Sp9 L ⁄

Araceae

SyngoniumangustatumSchott

E

⁄

Bromeliaceae

Bromeliaceae1 E ⁄

References

Boose, E.R., Foster, D.R., Plotkin, A.B., Hall, B., 2003. Geographical and historicalvariation in hurricanes across the Yucatan peninsula. In: Gómez-Pompa, A.,Allen, M.F., Fedick, S.L., Jiménez-Osorio, J.J. (Eds.), The Lowland Maya Area:Three Millennia at the Human Wildland Interface. Haworth Press, New York, pp.495–516.

Bradshaw, A., 1997. The importance of soil ecology in restoration science. In:Urbanska, K.M., Weeb, N.R., Edwards, P.J. (Eds.), Restoration Ecology andSustainable Development. Cambridge University Press, Cambridge, UK, pp. 33–64.

Bray, D.B., Ellis, E., Armijo-Canto, N., Beck, C., 2004. The institutional drivers ofsustainable landscapes: a case study of the ‘‘Mayan zone’’ in Quintana RooMexico. Land Use Policy 21, 333–346.

Cairns, M.A., Olmsted, I., Granados, J., Argaez, J., 2003. Composition andaboveground tree biomass of a dry semi-evergreen forest on Mexico’sYucatan Peninsula. For. Ecol. Manage. 86, 125–132.

Carrillo-Bastos, A., Islebe, G.A., Torrescano-Valle, N., González, N.E., 2010. Holocenevegetation and climate history of central Quintana Roo, Yucatan PeninsulaMexico. Rev. Palaeob. Palynol. 160, 189–196.

Ceccon, E., Huante, P., Campo, J., 2003. Effects of nitrogen and phosphorusfertilization on the survival and recruitment of seedlings of dominant treespecies in two abandoned tropical dry forests in Yucatan Mexico. For. Ecol.Manage. 182, 387–402.

CEEM, 1987. Monografía municipal de Felipe Carrillo Puerto. Gobierno del Estado deQuintana Roo, México.

Chavelas, P., Contreras, G., 1990. Caracterización de la estructura de acahuales dediferentes edades. Taller internacional sobre investigación en silvicultura ymanejo de selvas, SARH/INIFAP, México.

Chazdon, R.L., Letcher, S.G., van Breugel, M., Martínez-Ramos, M., Bongers, F.,Finegan, B., 2007. Rates of change in tree communities of secondary neotropicalforests following major disturbances. Philos. Trans. R. Soc. B. 362, 273–289.

CONAFOR, 2009. Reporte semanal de resultados de incendios forestales 2009.CONAFOR-SEMARNAT, México.

CONAFOR, 2010. Reporte semanal de resultados de incendios forestales 2010.CONAFOR-SEMARNAT, México.

CONAFOR, 2011. Reporte semanal de resultados de incendios forestales 2011.CONAFOR-SEMARNAT, México.

Cortés-Castelán, J.C., Islebe, G.A., 2005. Influencia de factores ambientales en ladistribución de especies arbóreas en las selvas del sureste de México. Rev. Biol.Trop. 53 (1–2), 115–133.

Denslow, S.L., 1995. Disturbance and diversity in tropical rain forests: the densityeffect. Ecol. Appl. 5 (4), 962–968.

Díaz-Gallegos, J.R., Mas, J.F., Velazquez-Montes, A., 2008. Monitoreo de los patronesde deforestación en el corredor biológico Mesoamericano Mexico. Interciencia33 (12), 882–890.

Dickinson, M.B., Hermann, S.M., Whigham, D.F., 2001. Low rates of backgroundcanopy-gap disturbance in a seasonally dry forest in the Yucatan Peninsula witha history of fires and hurricanes. J. Trop. Ecol. 17, 895–902.

Dupuy, J.M., Hernández-Stefanoni, J.L., Hernández-Juárez, R.A., Tetetla-Rangel, E.,López-Martínez, J.O., Leyequién-Abarca, E., Tun-Dzul, F.J., May-Pat, F., 2012.Patterns and correlates of tropical dry forest structure and composition in ahighly replicated chronosequence in Yucatan Mexico. Biotropica 44 (2), 151–162.

Ewel, J., 1980. Tropical succession: manifold routes to maturity. Biotropica 12, 2–7.Faust, B.B., 2001. Maya environmental successes and failures in the Yucatan

Peninsula. Environ. Sci. Policy 4, 153–169.García, E., 1988. Modificación al sistema de clasificación de Köppen (para adaptarlo

a las condiciones de la República Mexicana), fourth ed. Instituto de Geografía,UNAM, México.

Gaudet, C.L., Kaedy, P.A., 1988. A comparative approach to predicting competitiveability from plant traits. Nature 334, 242–243.

Gentry, A.H., 1995. Diversity and floristic composition of neotropical dry forests. In:Bullock, S.H., Mooney, H.A., Medina, E. (Eds.), Seasonally Dry Tropical Forests.Cambridge University Press, Cambridge.

Giardina, C.P., Sanford, R.L., Døckersmith, I.C., Jaramillo, V.J., 2000. The effects ofslash burning on ecosystem nutrients during the land preparation phase ofshifting cultivation. Plant Soil. 220, 247–260.

Gibson, A., Nobel, P.S., 1986. The Cactus Primer. Harvard University, USA.Gomes, J., Pacheco, J., Carlesso, R., Trois, C., Knies, A.E., 2011. Perdidas de agua por

evaporación en maíz con siembra convencional y directa para diferentes nivelesde cobertura muerta resultados experimentales. Rev. Cienc. Tec. Agrop. 20 (2),60–64.

Halle, F., Oldeman, R.A., Tomlinson, P.B., 1978. Tropical Trees and Forests: AnArchitectural Analysis. Springer-Verlag, New York.

Hasselquist, N.J., Santiago, L.S., Allen, M.F., 2010. Belowground nitrogen dynamics inrelation to hurricane damage along a tropical dry forest chronosequence.Biogeochemistry 98, 89–100.

Hill, M.O., Gauch, H.G., 1980. Detrended correspondence analysis: an improvedordination technique. Vegetatio 42, 47–58.

Holl, K.D., 1999. Factors limiting tropical rain forest regeneration in abandonedpasture: seed rain, seed germination, microclimate, and soil. Biotropica 31,229–242.

INEGI, 1990. Censo agropecuario y forestal por ejidos. INEGI, México.Jhones, C.C., Gama, L., Giddings, L., Rico-Gray, V., Velázquez, A., 2000.

Caracterización de los paisajes terrestres actuales de la península de Yucatán.Inv. Geog. 42, 28–39.

Kammesheidt, L., 1999. Forest recovery by root suckers and above-ground sproutsafter slash-and-burn agriculture, fire and logging in Paraguay and Venezuela. J.Trop. Ecol. 15, 143–157.

Kellman, M.C., 1970. On the nature of secondary plant succession. In: Archibold,O.W. (Ed.), Ecology of World Vegetation. Chapman & Hall, London (1955).

Kennard, D.K., 2002. Secondary forests succession in a tropical dry forest: patternsof development across a 50-year chronosequence in lowland Bolivia. J. Trop.Ecol. 18, 53–66.

Kennard, D.F., Gould, F., Putz, F.E., Fredericksen, T.S., Morales, F., 2002. Effect ofdisturbance intensity on regeneration mechanisms in a tropical dry forest. For.Ecol. Manage. 162, 197–208.

Kent, M., Coker, P., 1996. Vegetation Description and Analysis: A Practical Approach.John Wiley & Sons, New York.

La Torre-Cuadros, M.A., Islebe, G., 2003. Traditional ecological knowledge and use ofvegetation in southeastern Mexico: a case study from Solferino, Quintana Roo.Biodiv. Consv. 12, 2455–2476.

López-Martínez, J.O., Sanaphre-Villanueva, L., Dupuy, J.M., Hernández-Stefanoni,J.L., Meave, J.A., Gallardo-Cruz, J.A., 2013. b-Diversity of functional groups ofwoody plants in a tropical dry forest in Yucatan. PloS One 8 (9), 1–9.

Magurran, A., 1996. Ecological Diversity and Its Measurement. Chapman & Hall,London.

McCune B., Mefford M.J., 2006. PC-ORD. Multivariate analysis of ecological data.Version 5.10. MjM Software, Gleneden Beach, Oregon.

McLaren, K.P., McDonald, M.A., 2003. Coppice regrowth in a disturbed tropical drylimestone forest in Jamaica. For. Ecol. Manage. 180, 99–111.

Miles, L., Newton, A.C., DeFries, R., Ravilious, C., May, I., Blyth, S., Kapos, V., Gordon,J., 2006. A global overview of the conservation status of tropical dry forest. J.Biogeo. 33, 496–510.

Miller, P.M., 1999. Effects of deforestation on seed bank in a tropical deciduousforests of western Mexico. J. Trop. Ecol. 15, 179–188.

Miller, P.M., Kauffman, J.B., 1998. Seedling and sprout response to slash-and-burnagriculture in a tropical deciduous forest. Biotropica 30, 538–546.

Miranda F., Gómez-Pompa A., Hernández E., 1960. Un método para la investigaciónecológica de regiones tropicales húmedas de México. Comisión de estudiossobre la ecología de las dioscóreas (vol 3). SAG – INIF, México.

Murphy J., 1990. Indigenous forest use and development in the ‘‘Maya Zone’’ ofQuintana Roo, Mexico. Major paper/Master in Environmental Studies, GraduateProgram in Environmental Studies, York University, Toronto, Canada, 181 pp.

Murphy, P.G., Lugo, A.E., 1986. Ecology of tropical dry forest. Ann. Rev. Ecol. Syst. 17,67–88.

Murphy, P.G., Lugo, A.E., 1995. Dry forests of Central America and Caribbean islands.In: Bullock, S.H., Mooney, H.A., Medina, E. (Eds.), Seasonally Dry TropicalForests. Cambridge University Press, Cambridge.

Negreros-Castillo, P., Hall, R., 2000. Sprouting capability of seventeen tropical treespecies in Quintana Roo Mexico. For. Ecol. Manage. 126, 399–403.

M. Valdez-Hernández et al. / Forest Ecology and Management 334 (2014) 331–343 343

Pickett, S.T.A., 1989. Space for time substitution as an alternative to long-termstudies. In: Likens, G.E. (Ed.), Long-Term Studies in Ecology. Springer, New York.

Putz, F.E., Fredericksen, T.S., 2004. Silvicultural intensification for tropical forestconservation: a response to Sist and Brown. Biodiv. Conserv. 13 (12), 2387–2390.

Randall, W.M., Pickett, S.T.A., 1992. Dynamics of associations between plants in tenold fields during 31 years of succession. J. Ecol. 80, 291–302.

Rico-Gray, V., García-Franco, J., 1992. Vegetation and soil seed bank on successionalstages in tropical lowland deciduous forest. J. Veg. Sci. 3, 617–624.

Sampaio, A.B., Holl, K.D., Scariot, A., 2007. Does restoration enhance regeneration ofseasonal deciduous forests in pastures in Central Brazil. Restor. Ecol. 15 (3),462–471.

Sánchez, O., 2000. Análisis estructural de la selva del jardín botánico. In: Sánchez,O., Islebe, G.A. (Eds.), El jardín botánico Dr. Alfredo Barrera Marín fundamento yestudios particulares, ECOSUR-CONABIO, México.

Sánchez, O., Islebe, G.A., 1999. Hurricane Gilbert and structural changes in tropicalforest in south-eastern México. Global Ecol. Biogeogr. Lett. 8, 29–38.

Sánchez, O., Islebe, G.A., Valdez-Hernández, M., 2007. Flora arbórea ycaracterización de gremios ecológicos en distintos estados sucesionales de laSelva Mediana de Quintana Roo. For. Veracruzana 9 (2), 17–26.

Sarukhán, J., 1964. Estudio sucesional de un área talada en Tuxtepec Oaxaca. Bol.Esp. Inst. Nac. Invest. For.. INIF, México.

Schnitzer, S.A., Bongers, F., 2002. The ecology of lianas and their role in forests.Trends Ecol. Evol. 17 (5), 223–230.

Schnitzer, S.A., Dalling, J.W., Carson, W.P., 2000. The impact of lianas on treeregeneration in tropical forest canopy gaps: evidence of an alternative pathwayof gap-phase regeneration. J. Ecol. 88, 655–666.

SEMARNAT, 2002. Norma Oficial Mexicana NOM-059-ECOL-2001, Protecciónambiental–Especies nativas de México de flora y fauna silvestres-Categoríasde riesgo y especificaciones para su inclusión, exclusión o cambio-Lista deespecies en riesgo. Diario Oficial de la Federación, México.

Snook, L. 1993. Stand dynamics of Mahogany (Swietenia macrophylla King) andassociated species after fire and hurricane in the tropical forests of the YucatanPeninsula, Mexico. Doctoral Dissertation, Yale University, New Haven, CT, USA.

Snook, L.K., 1998. Sustaining harvests of Mahogany (Swietenia macrophylla King) inMexico’s Yucatan forests: past, present and future. In: Primack, R.B., Bray, D.,Galletti, H.A., Ponciano, I. (Eds.), Timber, Tourists and Temples: Conservationand Development in the Maya Forest of Belize. Island press, Washington, DC,Guatemala and Mexico, pp. 61–80.

Snook, L.K., Negreros-Castillo, P., 2004. Regenerating mahogany (Swieteniamacrophylla King) on clearings in Mexico’s Maya Forest: the effects of clearing

treatment and cleaning on seedling survival and growth. For. Ecol. Manage. 189,143–169.

Sohn, Y., Moran, E.F., Gurri, F., 1999. Deforestation in North-Central Yucatan (1985–1995): mapping secondary succession of forest and agricultural land use inSotuta using the cosine of the angle concept. Photo Eng. Rem. Sens. 65, 947–958.

Sorensen N., 2006. Regeneration and growth of several canopy tree species in theMaya forest of Quintana Roo, Mexico: The role of competition and microhabitatconditions. Ph.D. Dissertation. Oregon State University, Corvallis, OR.

Valdez-Hernández M., 1999. La vegetación después de tres tipos de perturbación;en la comunidad de X-pichil, Quintana Roo. Dissertation. Instituto Tecnológicode Chetumal, Chetumal, México.

Vandermeer, J., Zamora, N., Yih, K., Boucher, Y.D., 1990. Regeneración inicial en unaselva tropical de la costa caribeña de Nicaragua después del huracán Juana. Rev.Biol. Trop. 38 (2B), 347–359.

Vandermeer, J., Granzow-De La Cerda, I., Boucher, D., Perfecto, I., Ruiz, J., 2000.Hurricane disturbance and tropical tree species diversity. Science 209, 788–791.

Vieira, D.L.M., Scariot, A., 2006. Principles of natural regeneration of tropical dryforests for restoration. Restor. Ecol. 14 (1), 11–20.

Vieira, D.L.M., Scariot, A., Sampaio, A.B., Holl, K.D., 2006. Tropical dry forestregeneration from root suckers in Central Brazil. J. Trop. Ecol. 22, 353–357.

Westman, W., 1990. Structural and floristic attributes of recolonizing species inlarge rain forest gaps, North Queensland. Biotropica 22 (3), 226–234.

White, D.A., Hood, C.S., 2004. Vegetation patterns and environmental gradients intropical dry forests of the northern Yucatan Peninsula. J. Veg. Sci. 15, 151–160.

Whitmore, T.C., 1975. Tropical Rain Forests of the Far East. Oxford Clarendon Press,London.

Williams-Linera, G., 1990. Origin and early development of forest edge vegetation inPanama. Biotropica 22 (3), 235–241.

Williams-Linera, G., Lorea, F., 2009. Tree species diversity driven by environmentaland anthropogenic factors in tropical dry forest fragments of central VeracruzMexico. Biodiv. Conserv. 18, 3269–3293.

Williams-Linera, G., Domínguez-Gastelú, R., García-Zurita, M.E., 1997.Microenvironment and floristics of different edges in a fragmented tropicalrainforest. Conserv. Biol. 12 (5), 1091–1102.

Williams-Linera, G., Alvarez-Aquino, C., Hernández-Ascención, E., Toledo, M., 2011.Early successional sites and the recovery of vegetation structure and treespecies of the tropical dry forest in Veracruz Mexico. New For. 42, 131–148.

Yih, K., Boucher, D., Vandermeer, J., Zamora, N., 1991. Recovery of the rain forest ofsoutheastern Nicaragua after destruction by hurricane Joan. Biotropica 23 (2),106–113.