Mexico: Los Tuxtlas Tropical Forest Restoration

Reforested plots in Los Tuxtlos; Credit: Anika Hazra
Reforested plots in Los Tuxtlos


The Los Tuxtlas project was started with the hypothesis that planting seedlings of interior forest species after land abandonment could sharply accelerate the process of re-vegetation of complex communities. Pioneer stands or monocultural plantations may be enriched with seedlings of late-successional animal-dispersed trees, or initial plantings could be mixes of late-successional and pioneer species. This project sets criteria for selecting species for enrichment and in some cases for overstoreys. The underlying idea for the restoration work at Los Tuxtlas Biological Station in the state of Veracruz, Mexico was that in the process of conversion of rainforest to pastoral or agricultural use and then back to rainforest has shown a difficulty in the recovery of equivalent levels of biodiversity found in the original forest. In order to evaluate the restoration of deep-forest species, researchers looked at both natural regeneration and experimental plantings in pasture that had been abandoned. By evaluating the variability in leaf traits, which is required during the development from a seedling to a tree, it was found that it is already measurable in juvenile trees distributed over a range of environments. Variability in leaf traits, and therefore ecological performance, was expected to correlate with tree stature because of the light gradient that occurs from the bottom to the top of the forest. Tall species may experience low light levels as seedlings or juveniles inhabiting the forest floor, but higher light levels as they gain height. Therefore, they show leaves with traits related to high-light environments while short-statured trees show leaves with traits of low-light environments. In addition, tall species experience higher heterogeneity in light levels within their crowns than shorter species that spend most of their lives under lower light levels. For tall tree species, labile leaf characteristics provide the functional means of adjusting easily to the changing environmental conditions of early-successional environments. The project was most interested in whether: (1) the growth and survival of late-successional species growing in pastures, edges and secondary forests is related to intraspecific variation of SLM (specific leaf mass) and leaf density, (2) growth and survival and leaf traits of late-successional trees are related to the maximum mature height of species, and (3) if those late-successional species that grow faster survive better in early-successional environments. It was found through the course of the project that the planting of late-successional species in early-successional environments helps enrich low-diversity forests colonized by pioneers that arrive by unassisted dispersal. To select late-successional species to enrich early-successional environments, we relate plant traits to ecological performance. Maximum mature height and variation in leaf traits are associated with survival or growth rates of late successional species planted in early-successional environments.

Quick Facts

Project Location:
Unnamed Road, Oaxaca, Mexico, 17.41442177435896, -95.18307502812502

Geographic Region:
Latin America

Country or Territory:

Tropical Forest

Tropical Forest - Seasonal Broadleaf

Area being restored:
25 hectares

Organization Type:
University / Academic Institution


Project Stage:

Start Date:

End Date:

Primary Causes of Degradation

Agriculture & Livestock, Deforestation, Fragmentation

Degradation Description

The area near Los Tuxtlas has been inhabited since pre-Columbian times. Its location in southern Veracruz state places it within a mix of cultures ranging from Olmec to Mayan to Creole. Subsistence agriculture has been practiced continually since the days of the Olmec, but with the arrival of the Spanish came cattle and the necessity of clearing land for pasture. The area around Los Tuxtlas is today a mix of small agricultural fields, pasture, and remnant forest. In 1968, the project site was deforested and subsequently used for agriculture for four years and then as pasture for fifteen years. It has undergone natural regeneration since abandonment in 1987.

Reference Ecosystem Description

The preserve near Los Tuxtlas is the northernmost extension of lowland evergreen rainforest in the neotropics. These forests were once dominant in the lowlands along the Gulf of Mexico and stretch south through Central America and into South America. Los Tuxtlas is a part of the Mexican volcanic belt, a mountain range formed in the Oligocene period with the rock foldings and the intense volcanic activity present from its origins. In general the flora of Los Tuxtlas is associated with neotropical floras of the Caribbean Region and specifically, the Province of the Coast of the Gulf of Mexico. An important component of its flora originated in Central and South America and its distribution seems to have played a preponderant role in the processes of speciation of the Mexican flora. Although around 75 % of the species of plants are found in Central America, Los Tuxtlas is one of the five areas with the highest endemism of trees in Mexico and it has been indicated that nearly 10 % of the trees of the superior canopy, endemic to the warm-humid zones of Mexico. Historically, these rainforests are some of the most biologically diverse ecosystems on the planet. Neotropical rainforest is composed of four distinct layers: the emergent layer, the canopy, the understory and the forest floor. Each layer has its own flora and fauna, which in some instances can mean upwards of 10,000 species or more in a single acre.

Project Goals

1. Planting of 45 species of pioneer and late-successional trees, palms and lianas to evaluate their survival and growth rates in abandoned pastures.

2. Identify vegetative traits (e.g. leaf traits, maximal tree height) that predict success in survival and growth of 23 species of trees under the stressful conditions of open pasture or recently abandoned cropland. Also, to use those traits as selection criteria for species to enrich or dominate forest restoration.

3. Planting of 12 late-successional tree species in secondary forest and pastures to evaluate simulated arrested succession and identify those traits that predict survival and growth rates in persistent adverse conditions such as pastures with no input of pioneer species (arrested succession).

4. Evaluate survival and growth rates of late-successional tree species under initial arrested succession and subsequent natural succession.


The project does not have a monitoring plan.


Large areas of tropical rainforest are being lost annually to deforestation. These lands are being used for a variety of agricultural activities. At the same time, thousands of acres of land are being abandoned annually, allowing for forest succession. The project focused on the consequences of abandonment, including how on top of losing the initial biodiversity of the rainforest there was an additional “˜time tax’ on the delay in colonization of deep forest species. The project focused on how intentional plantings could increase the colonization and hence diversity of those areas undergoing succession. This land belongs now to a forest reserve. The study site is close to a 640 ha reserve of the northernmost rain forest in the Neotropics, which belongs to UNAM (Universidad Nacional Autonomia de Mexico). The researchers were interested in how to integrate low-input methods of forest restoration to enhance levels of biodiversity as succession progressed.

Description of Project Activities:
The project is part of a permanent project examining the restoration of neotropical rainforests using intentional planting of late-successional tree species. From 23 late-successional tree species evaluated, those with high intraspecific variability in leaf mass per unit leaf area (specific leaf mass; SLM) showed higher survival and growth rates across different microhabitats of early successional environments than species with low variability in SLM. The predictive power was higher for mid-canopy species (>11-25 m). For twelve late-successional tree species evaluated, those species with low mean SLM measured in sun (pastures) or shade (secondary forest) leaves showed higher growth rates in height when growing in abandoned pastures with long-lasting high light conditions for 17 months. Maximum tree height was related to variation in SLM but it was not correlated to survival and growth rates in early successional ecosystems and neither in pastures. Leaf traits offer easily measurable variables that may provide criteria for selection of species for planting mixed species stands. This alleviates the need to individually screen large numbers of late-successional species for performance in different scenarios of restoration (i.e. early successional ecosystems, pastures). The general procedure was to germinate seeds collected from the Los Tuxtlas Biological Station (LTBS) in a shaded nursery, and transplant seedlings to three microhabitats in the field. In August - December 1996, seeds of 45 species of pioneer and late-successional trees, palms and lianas were collected from the LTBS and germinated under shade in a nursery (100 individuals per species, on average; total of N = 5000 individuals). Seeds were collected from more than three maternal individuals to ensure genetic diversity. During January - March 1997, researchers marked a grid of 5000 2 m x 2 m squares covering the field, including each of the above-designated microhabitats. Seedlings of the 45 species approximately 4 months old were planted in mixed stands, one seedling in the middle of each square. The microhabitat that each experimental individual inhabited was described with respect to crown illumination index in 2001. This index ranges from 1 to 5, with 1 representing the dark understory of the secondary forest (no direct light reaching individuals) and 5 representing sunny pasture (individuals totally exposed to light). To characterize the physical environment in the five different levels of crown illumination index, researchers measured photon-flux density (PFD), gravimetric soil moisture, and bulk density in a sub-sample of individuals during the dry season. PFD was measured with a Light Meter (Li-cor LI-189 quantum sensor). Gravimetric soil moisture and bulk density were measured following Chapman (1976), bulk density was used as an estimate of soil compaction. Individuals of all species were exposed to significantly different photon-flux densities (Krus- kall - Wallis(4,495) = 115.41, P < 0.0001) and gravimetric soil moisture (F (4,115) = 4.02, P < 0.005) at the different levels of crown illumination index. Individuals experienced similar bulk densities at all levels of the index (F (4,115) = 0.04, P > 0.5; Table 1). Twenty-four tree species from 15 families were used to represent the variety of fruit types and maximum mature heights of late-successional trees in the area. Maximum mature height was considered to be the tallest published height for adults of a given species at LTBS. To calculate seedling and juvenile growth rates, heights and diameters at the stem base were measured in April 1998, October 1998, April 1999 and April 2001. Grasses and herbs were removed within a radius of 25 cm around individuals at the time of measurement, but neighboring plants above each experimental plant were left untouched. Increments in height (heightt2 - height t1/time) and in basal stem diameter were calculated for the three periods between censuses. The survival percentage for each species was calculated as the number of individuals surviving from 1997 to 2001, divided by the number of individuals present in 1997. Seven species that experienced some losses of seedlings during a small accidental fire in 1998 were omitted from the analysis of survival, leaving 17 species for that analysis. To determine intraspecific variation in leaf traits, three leaves from one individual per species in each microhabitat were measured for a total of five individuals per species (N = 115 individuals). For those species with compound leaves, one leaflet at the same position was sampled from three leaves. Leaves were sampled from different branches just behind the new leaves to select for similar leaf ages. For those species with few leaves, individuals selected were those with enough leaves to be sampled; when all individuals of a species had many leaves, individuals were chosen at random. Leaves were measured in a leaf area meter (Ci-202, CID, Inc. Camas, WA, USA; leaf size, LS) and weighed to the nearest 0.1 g (fresh weight, FW) within 2 h of collection (Balance Pocket Pro 250-B). Leaves were oven dried to constant mass at 100 8C and weighed to the nearest 0.001 g (Balance Mettler PM 1200; dry weight, DW). With these data, leaf mass per unit area [SLM = (DW/LS)] and leaf density [leaf density = FW/DW] were calculated. Leaf density, as measured in this study, reflects the amount of cytoplasm versus hemicellulose and cellulose in plant tissues. Intraspecific variation for each leaf trait is reported as its coefficient of variation (hereafter CV) = (standard deviation of given leaf trait) x (100)/ (mean of leaf trait across all microhabitats), corrected for sample size.

Ecological Outcomes Achieved

Eliminate existing threats to the ecosystem:
Planting late-successional species in early-successional environments helps enrich low-diversity forests colonized by pioneers that arrive by unassisted dispersal. To select late-successional species to enrich early-successional environments, researchers were able to relate plant traits to ecological performance. Maximum mature height and variation in leaf traits are associated with survival or growth rates of late successional species planted in early-successional environments. Mid-canopy species with higher intraspecific variation in SLM show higher growth rates and survival, as expected. Also, taller species show higher growth increments in diameter, but not in height. Finally, those species with higher growth rates experience higher survival. All relationships between measured growth rates are strongest for mid-canopy species (R2 greater than or equal to 0.6 - 0.8), and may only consistently exist with them. Among canopy species, growth rate is the best predictor of survival (R2 > 0.45). It should be noted that survival of canopy and mid-canopy species is generally high and similar (canopy 73.25 plus or minus 19.36%, N = 6; mid-canopy 66.96 plus or minus 23.52%, N = 8). It may be that this sample of six canopy species is insufficient to show clear differences with so little variation in key variables, or that other factors are at play with canopy as compared with mid-canopy species. The successful contribution is an easy-to-measure plant trait that predicts growth and survival of mid-canopy trees; a strong positive relationship exists between variation in SLM and growth and survival of this segment of the forest community. There are good reasons why variation in leaf traits should be a good indicator of growth and survival under varied conditions. Leaf traits vary within a species when individuals grow under different environmental conditions. Plants exposed to high-light conditions have smaller leaves, with higher SLM and leaf density than plants in low-light environments. The ability of species to develop leaves adapted to high-or low-light conditions explains, in part, differential growth rates under different light levels. Within-individual variation in leaf traits has previously been related to success of individuals exposed to heterogeneous environments, for example, in vines, modular plants (ramets of a single genet) and canopy trees. A higher intraspecific variation in leaf traits may be interpreted as a greater capacity to adjust to current environmental conditions of an early-successional environment, thereby predicting higher growth and survival of plants in these heterogeneous habitats. Early-successional environments (pastures and edge conditions) have higher light levels, and lower availability of water, than the understory of the mature forest. Leaf density is expected to increase under water stress to enhance water-holding capacity of leaves. However, in this study even when there is a correlation between CV of leaf density and CV of SLM (Pearson correlation = 0.70, P < 0.001), leaf density is less variable than SLM, this variation is only weakly related to survival, and it is not at all related to growth rates. Similar lack of response to situational shortages of water was found in seedlings of Inga edulis and Ocotea withei growing in pastures in Costa Rica with a dry season as long as the Los Tuxtlas season (ca. 3 months). Perhaps low intraspecific variation in the trait and low variation in water availability reduce the effectiveness of this character as a predictor of ecological performance. Both studies do show the tolerance of late-successional species to shortages of water, a characteristic that allows successful planting in deforested areas. A gradient in light, temperature and relative humidity is found from the forest floor to the upper canopy. Higher light, temperature and relative humidity are found at the top of the canopy, where the highest temperatures remain for longer times and the greatest fluctuations in relative humidity occur. It follows that the leaves of taller trees are exposed to higher light levels and higher fluctuation in environmental conditions than those of short species. Besides the environment, leaf traits are affected by height because the cost of foliar construction as tree size increases due to reduced water flow at the top of the canopy. Reduced water flow results in higher water stress for canopy leaves and limits maximum mature height. Therefore, tall trees show leaves with characteristics related to high-light levels and water stress (i.e., high specific leaf mass). The interplay of adaptation to high-light and water stress may give canopy trees a different adaptive trajectory than mid-canopy species. The conditions to which plants must respond change with ontogeny. These changes differ for canopy species of tall stature that eventually experience exposed canopy, as compared with mid- canopy species that experience full sun only intermittently, and shorter species that may never be exposed to the sun. Canopy and mid-canopy species experience low-light levels early in their lives, but higher light levels as they grow to the canopy. They develop different leaves during ontogeny for this predictable increase in light levels (developmental changes in leaf traits) and concomitant higher water stress as height increases. As adults, these species develop sun leaves at the top of the canopy and shade leaves at the bottom of the same crown. Therefore, higher variability of leaf traits of canopy and mid-canopy species is anticipated and seems to arise in early ontogenetic stages. The ability to develop various types of leaves likely helps these species accommodate suppression and release while growing to reach different levels in the canopy. Mean leaf traits may change during ontogeny, but the trees have the capacity to develop sun or shade leaves as required at any ontogenetic stages. In mid-canopy, species variability in leaf traits may translate accurately in growth rates because they never experience full sun conditions, therefore, leaves within their canopy will always show different traits depending on environmental conditions. Researchers could only speculate about the absence of response in canopy species. The possibility exists that a developmentally programmed response makes them less sensitive to minor changes in environmental light. Performance of canopy species may be fixed since they have a predictable future habitat: sun conditions in the canopy. A positive relationship between growth rate and tree height appears to be general. The relationship holds in more than 50 tree species in Malaysian rainforest, and 44 species in Costa Rica. This has implications for restoration that involves community-wide management of succession. With this project, relationships between tree height and growth rates differed when species are partitioned by their maximum mature height. The strength of the relationship found in this and other studies (up to R2 = 0.46 for mid-canopy species) may not be enough to predict growth rates by tree height only, but it will be a useful tool to start selection of species before further testing is done. Short species, that are expected to show low growth rates in most environments, may be introduced at later times to further enrich the forest. Maximal growth of late-successional species primarily occurs in partial to full vertical illumination, suggesting that the absence of late-successional species in open areas is due to the lack of propagules and/or germination capacity instead of inability to grow in open conditions. Therefore, if planted for management or reforestation, late-successional species are not only able to cope with high-light levels and low water supply, they grow faster in early-successional conditions than under their more normal shaded seedling habitats (e.g., Pouteria sapote and Diospyros digyna).

Factors limiting recovery of the ecosystem:
Currently people and cattle are able to enter the study plantation, as no fence is present. Measures were not taken to erect a fence. As natural succession took place, the vegetation became denser and people and cattle did not cross the area anymore. An open area inside the restored parcel remains covered with the exotic grass Cynodon plectostachyus, which grows up to 1.5 m tall in the wet season. This area experienced one accidental fire in 1998, one year after the plantation was established. Fire is no longer a threat.

Socio-Economic & Community Outcomes Achieved

Economic vitality and local livelihoods:
Late-successional tropical tree species represent >80% of rain forest tree species in undisturbed habitats, providing much of the structural diversity of the communities as well as food resources for fruit-and seed-eating animals. For places where natural regeneration is possible, but the species that are present in the old growth forest require many years to arrive by themselves (i.e. are dispersal-limited), the planting of late-successional species is a viable means for restoring diversity much more rapidly than natural regeneration. Here researchers show that tree species with high intraspecific variability in SLM survive and grow better across different microhabitats of early-successional environments, with much the strongest relationships for mid-canopy species. The relationship for canopy species did not hold for this project and remains to be confirmed. Based on the project's results, successful enrichments are likely to include mid-canopy species with high growth rates predicted by high variation of specific leaf mass (SLM), and canopy species with high growth rates predicted by maximum mature height. Use of variability in leaf traits and other indices (e.g., maximum mature height) that are related to performance may alleviate the need to individually screen large numbers of late-successional species for high growth rates and survival in restoration projects. Use of such easily assessed measures would free time and resources for evaluation of other criteria, such as economic value or dispersal attributes that influence animal populations, for combinations of species that will generate desired restored forests. The range of variation in these traits may differ from one site to another, but in any particular forest mosaic the rankings of variation are likely to be consistent. Some of these traits may be measured in herbarium specimens. Mean values of SLM will be higher than those where leaf area is measured in fresh leaves. However, by measuring a large number of specimens it may be possible to capture the variation of SLM of the species evaluated. Secondary selections of species should be made taking into account those with the highest dispersal limitation (large seeds, low dispersal ability). Enrichment of early-successional environments that includes as many species as possible will maximize diversity and complexity of regenerating forests.

Long-Term Management

Los Tuxtlas is now a part of a forest reserve that belongs to UNAM (Universidad Nacional Autonomia de Mexico). The site is under a permanent monitoring regime as a consequence.

Sources and Amounts of Funding

These activities where supported by initial funding from the UNAM (Mexico) for 7 years obtained by M. Ricker. Further funding was obtained by Martínez-Garza from the Lincoln Park Zoo Neotropic Fund (Chicago) for 2 years and from UNAM thereafter.

Other Resources

Martínez-Garza, C., V. Peña, M. Ricker, A. Campos and H. F. Howe (2005). Restoring tropical biodiversity: Leaf traits predict growth and survival of late-successional trees in early-successional environments. Forest and Ecology Management 217: 365-379

Martínez-Garza, C. and H. F. Howe (2005). Developmental strategy or immediate responses in leaf traits of tropical tree species? International Journal of Plant Sciences 166: 41-48

Martínez-Garza, C. and H. F. Howe (2003). Restoring tropical diversity: beating the time tax on species loss. Journal of Applied Ecology 40: 423-429.

Cristina Martínez-Garza
Centro de Educación Ambiental e Investigación
Sierra de Huautla (CEAMISH),
Universidad Autónoma del Estado de Morelos (UAEM),
Av. Universidad 1001, Col Chamilpa,
Cuernavaca, Morelos 62209.

Primary Contact

Organizational Contact