Most studies on frugivorous bat assemblages in secondary forests have concentrated on differences among successional stages, and have disregarded the effect of forest management. Secondary forest management practices alter the vegetation structure and fruit availability, important factors associated with differences in frugivorous bat assemblage structure, and fruit consumption and can therefore modify forest succession. Our objective was to elucidate factors (forest structural variables and fruit availability) determining bat diversity, abundance, composition and species-specific abundance of bats in (i) secondary forests managed by Lacandon farmers dominated by Ochroma pyramidale, in (ii) secondary forests without management, and in (iii) mature rain forests in Chiapas, Southern Mexico. Frugivorous bat species diversity (Shannon H’) was similar between forest types. However, bat abundance was highest in rain forest and O. pyramidale forests. Bat species composition was different among forest types with more Carollia sowelli and Sturnira lilium captures in O. pyramidale forests. Overall, bat fruit consumption was dominated by early-successional shrubs, highest late-successional fruit consumption was found in rain forests and more bats consumed early-successional shrub fruits in O. pyramidale forests. Ochroma pyramidale forests presented a higher canopy openness, tree height, lower tree density and diversity of fruit than secondary forests. Tree density and canopy openness were negatively correlated with bat species diversity and bat abundance, but bat abundance increased with fruit abundance and tree height. Hence, secondary forest management alters forests’ structural characteristics and resource availability, and shapes the frugivorous bat community structure, and thereby the fruit consumption by bats.
Citation: Vleut I, Levy-Tacher SI, de Boer WF, Galindo-González J, Vazquez L-B (2013) Tropical Secondary Forest Management Influences Frugivorous Bat Composition, Abundance and Fruit Consumption in Chiapas, Mexico. PLoS ONE 8(10): e77584. https://doi.org/10.1371/journal.pone.0077584
Editor: Brock Fenton, University of Western Ontario, Canada
Received: May 29, 2013; Accepted: September 3, 2013; Published: October 11, 2013
Copyright: © 2013 Vleut et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Funding: This research was funded by Etnobiología para la Conservación A.C. and by a doctoral scholarship awarded to the first author by CONACyT-Mexico (Reg. 239503). Idea Wild and Fondo Institucional de Fomento Regional para el Desarrollo Científico, Tecnológico y de Innovación (FORDECYT) (Agreement 116306) provided mist nets. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Competing interests: The authors have declared that no competing interests exist.
Due to the rapid conversion of rain forest into agricultural fields and secondary forest, with a deforestation rate of 0.5%/year, secondary forests will occupy a high percentage of the total forested area in the world in the next decades [1,2]. However, little evidence is available on the impact of secondary forest management strategies on frugivorous bats. Bats play a key role in forest regeneration processes, they are important contributors of propagules, facilitate the reproductive success of plants, and play an important role in the economic value of forests [3,4], as 70-98% of flowering woody plants from rain forest are dispersed by bats and other vertebrates [5–7].
The Mexican Maya Lacandon, developed effective management strategies for the fallow periods [8–10], which includes the sowing of selected tree species during this period, such as Swietenia macrophylla or Ceiba pentandra for their timber qualities, or Ochroma pyramidale, a pioneer tree known for its rapid growth and capacity to restore soil fertility . Ochroma pyramidale is a fast growing light demanding early-successional species [9,12–14], able to reach up to 6 m in a year, that can improve soil fertility through the acceleration of soil organic matter accumulation [15,16]. Changes in structure and composition due to specific traits of this dominant tree could trigger important cascading effects, as compositional and structural differences in tropical forests have been associated with differences in the bat species composition or species-specific abundance [17–21]. Ochroma pyramidale secondary forests presented higher canopy openness and tree height, lower tree density and higher shrub density in the understory in comparison with control secondary forests . A higher canopy openness could negatively affect bat assemblage, due to increasing predation risk , but could increase the abundance of Carollia and Sturnira bats, preferring fruits from light demanding early-successional shrub species (e.g. Piperaceae and Solanaceae) [7,20,21,23,24].
To test this, we first compared bat species diversity, bat abundance, composition and the abundance of the most dominant bat species among three forest types: managed secondary forest (with O. pyramidale), control secondary forests (without management) and rain forests. Secondary, we evaluated environmental variables, divided into structural variables (canopy openness, tree density, height and diversity) and fruit availability variables (diversity and abundance) among forest types, and correlated these variables with the frugivorous bat diversity, bat abundance and species-specific responses.
We expected to find that: (i) a different bat species composition among forest types with a higher abundance of bat species associated with open canopies, such as Carollia and Sturnira, in managed secondary forest with O. pyramidale compared to control secondary forests and rain forests (ii) the species composition of consumed fruits differs among forest types and more bats consume fruits of early-successional shrub species in the more open O. pyramidale forests in comparison with control secondary forests and rain forests (iii) bat diversity and abundance increases with decreasing tree density and increasing tree height.
Materials and Methods
The bats were captured with a Scientific Collector’s permit (Colector Científico de Flora y Fauna) issued by the environmental authority in Mexico (SEMARNAT permit number 04787/10). The guidelines of investigating flora and fauna in the Mexican territory conform to the policies of the Division of Graduate Studies and the Ethics Committee of El Colegio de la Frontera Sur. They approved the research study on “Factors that determine the presence of bats in secondary forests in the Maya Lacandon community Lacanhá, Chiapas, Mexico” to the first author on July, 2009. The bats were captured using mist nets and released after identifying and measuring bat individuals. None of the captured individuals were sacrificed.
We studied 12 sites in the Maya Lacandon community in Chiapas, Mexico (16° 46’ 08” N, 91° 08’ 12” W); eight sites were secondary forest patches of 0.5-1.0 ha extent and the other four were sites in the matrix of rain forest (1.0 ha). The community of Lacanhá is located on the margin of Montes Azules Biosphere Reserve, Chiapas, Mexico with an elevation of 355-370 m above sea level  and a dominating surface of rain forest (Maldonado Unpublished data). For a more detailed view of the study area and the locations of the study sites see Vleut et al. . All study sites were located on private lands and owners Manuel Castellano and Eva Chankin gave consent to conduct the study on their lands. The secondary forest patches were cultivated using slash-and-burn, and abandoned 10-15 years ago. The eight secondary forest patches were divided over two treatments: four patches of secondary forest were dominated by the pioneer tree O. pyramidale (Malvaceae), hereafter called O. pyramidale forests, product of management by Lacandon farmers before the end of the cultivation period; and four patches of secondary forests without any preference for tree species (hereafter called secondary forests), so we have three treatments (O. pyramidale forests, secondary forest and rain forest) with four replicates each. The proportion of surrounding rain forest around secondary forest patches was similar among sites to reduce the effect of confounding variables on bat diversity .
Bats were captured each month using 3 (12.0 x 2.4 m, 36 mm mesh) mist nets per site, set at ground level, starting 0.5 h before sunset until 4h after, one night per site, from August 2010-July 2011. This study focused on phyllostomid bats only, and the techniques we used are widely accepted as effective sampling methods for these bats [18,26]. The total sampling effort was 74,650 m2·h . Trails, roads and rivers were avoided while capturing bats, for their potential bias towards bat species . Vegetation structure can differ between the edges in comparison with the interior of the forests and possibly affecting affect bat assemblage. The influence of the proximity towards the boundary is significantly reduced beyond 15-25 m into the forest and we therefore placed mist nets at least 25 m from the border of sites . Bats were not captured during and 2 days prior and after a full moon. In the case of heavy rain we waited until the rain passed, shook excess of the mist nets and resumed counting the 4 and a half hour of bat sampling following Medellín et al. . Captured bats were identified using field guides [30,31]. Before releasing at the capture site, the forearm of each individual was marked with a colored marker, to avoid recounting during the same night.
Although frugivorous bat are considered to specialize in the consumption of fruits, they supplement their diet with other nitrogen or protein rich sources such as insects, foliage or nectar [32–34]. However, during this manuscript we focus on the consumption of fruits by bats by either determining the fruit species from the collected fecal samples or by the fruits carried by the bats. Carried fruits can include fruits with larger sized seeds which are not swallowed by the bat, but are dispersed when the pulp is consumed.
Fruit consumption data from fecal samples was gathered by collecting seeds from plastic sheets placed below the mist nets or from cloth bag in which the bats were individually held for a period of 30 min . In the case of overlapping bat individuals of different species in the mist nets we excluded the fecal samples on the plastic sheets from the database, to avoid assigning the fecal samples to the wrong species. The collected fecal samples were stored, dried and kept in small paper bags, and seeds were then separated from fruit pulp, insects and other fecal remains. Carried fruits and seeds from fecal samples were identified to species level when possible. Fruit consumption was quantified by the number of different fruit species carried or defecated per bat per site. We categorized fruit consumption into life form and successional stage; life form and successional stage comprised two classes, respectively shrubs or trees, and early-successional stage (light demanding plant species) or late-successional species (shade tolerant plant species), based on tolerance to shade and seed size [36,37].
Forest structure and fruit availability
Canopy trees were measured in six randomly selected quadrats of 10 × 10 m per site in which dbh (diameter at breast height) and height of all trees ≥5 cm dbh (diameter at breast height) were measured and identified. The mean tree density, height and diversity (Shannon H’) was estimated per site as well as the percentage of canopy openness at 15 random points using a hemispherical crown densiometer (Forestry Suppliers, Inc, Jackson, MS, USA).
Fruit production was monitored at monthly intervals, prior to each bat capturing, for all bat food plant species within the study sites, identifying and counting individual trees and shrubs that carried fruit. Following Wulms , food plants were identified along ten line-transects of approximately 80 m long, and a maximum detection distance between line-transects of 3 m. It was impossible to count the number of ripe fruits for certain tree species, especially for larger trees such as Ficus spp, so instead of counting fruits we recorded the number of plants with at least one mature fruit for all plant species. We quantified fruit availability as the number of shrub, trees and vines with fruits as potential food source for frugivorous bats per site. Fruit not encountered in fecal samples or carried by bats were not considered as part of the diet of frugivorous bats and eliminated from the analysis.
We estimated the species diversity (Shannon H’; ) of frugivorous bats, fruit consumption and fruit availability using EstimateS . The effect of forest type bat abundance was tested with separate General Linear Models (GLM), with forest type as fixed factor. We constructed residual plots to check for normality and log-transformed bat abundance to meet GLM requirements. Tukey post-hoc tests were performed on fixed factors if significant in the GLMs.
We used a one-way ANOSIM  on Bray-Curtis similarity with Bonferroni corrected p-values to test whether frugivorous bat species composition and the composition of fruit species consumption differed among the forest types.
To test for differences between the diversity of frugivorous bats, fruit consumption and fruit availability we used the two-tailed t-test proposed by Hutcheson . To compare the species abundance of the five most abundant frugivorous bat species among forest types we tested the species abundance for normality using Shapiro-Wilk normality tests and log-transformed C. sowelli abundance and Box-Cox (Box and Cox 1964 ) transformed C. perspicillata abundance species abundance to meet GLM requirements. We tested for the effect of forest type on the abundance of both Carollia species with separate General Linear Models (GLM), including forest type as fixed factor, followed by Tukey post-hoc tests. Unlike Carollia abundance, both Artibeus species and S. lilium abundance presented a Poisson distribution and we therefore used a Generalized Linear Model (GZLM), followed by Bonferroni corrected Mann-Whitney U post-hoc tests, to test for the effect of forest type on the abundance of both species’ abundances .
We tested for the effect of forest type on fruit consumption per life form and successional stage with a Kruskal-Wallis test, due to non-normal data, followed by a Bonferroni corrected Mann-Whitney U post-hoc tests for non-parametric data. We were not able to normalize the fruit consumption per life form and successional stage, and therefore we tested for the effect of forest type on fruit consumption per life form and successional stage with a Kruskal-Wallis test, followed by a Bonferroni corrected Mann-Whitney U post-hoc tests for non-parametric data.
Furthermore, we tested for differences in environmental variables, including structural variables (tree density, tree height, diversity of canopy trees and canopy openness) and fruit availability (fruit diversity and abundance) among forest types with an ANOVA, followed by Tukey post-hoc tests for normally distributed data, and a Kruskal-Wallis test, followed by a Bonferroni corrected Mann-Whitney U post-hoc tests for non-parametric data. All residuals were tested for normality using Shapiro-Wilk tests. All tests were performed in SPSS v17.
We used the informative-theoretic approach , to model the bat diversity and abundance as dependent variables with environmental variables as independent variables, based on the second order Akaike’s Information Criterion corrected for small sample size (AICc; i.e. n /K < ~40). The AICc values calculated from GLM’s has no meaning on its own, but comparing AICc estimations among different models is used as an approximation for the most meaningful model. Two measures associated with the AICc to compare the model fits were used; the delta AICc (Δi) and Akaike weights (wi). The models returning an Δi < 2 are suggested to be the most meaningful models. Akaike weights (wi) give a measure of the relative strength of evidence, and indicate the weight of evidence in favour of being the best model. We used model averaging when not one of the models proved overwhelmingly supported by the data, by calculating the precision (SE) of the model-averaged estimate, termed as the unconditional SE, unconditional 95% confidence intervals (CI) and the relative importance of each independent variable . Analyses on the informative-theoretic approach and model averaging were conducted in R v.3.0.0 .
Finally, we carried out a non-metric multidimensional scaling (NMDS) ordination, with Bray-Curtis as a measure of compositional dissimilarity , including abundance per bat species in relation to the differences in environmental variables per site, with 999 permutations, using the MetaMDS function in the R [44,46]. We used the stress value to measure the “goodness of fit” or the mismatch between distance measures and the distance in ordination space, with values smaller than 20 indicates a good ordination . The final solution for the NMDS ordination after 999 permutations, was achieved within two runs of the data, with a stress value of 5.24 over a scale of 0 to 100.
Frugivorous bat composition, diversity and abundance
A total of 1645 frugivorous bats, belonging to 18 species (Table 1), were captured and only two individuals were recaptured, with similar diversity (H’) between O. pyramidale forests and secondary forests (Hutcheson t-test t = -0.71, P > 0.05), O. pyramidale and rain forests (Hutcheson t-test t = -1.19, P > 0.05), and between secondary forests and rain forests (Hutcheson t-test t = -0.42, P > 0.05); Figure 1a). Bat abundance was affected by forest type, and was highest in O. pyramidale forests and in rain forests (F2,141 = 8.75, R2 = 0.110, P < 0.001; Figure 1b). The Anosim revealed a significant bat species compositional difference among forest types (global R = 0.1946, P < 0.001) and a pairwise comparison indicated significant differences between O. pyramidale forest and secondary forest (R = 0.1067, P < 0.01), O. pyramidale forest and rain forests (R = 0.2876, P < 0.01), and finally between secondary forest and rain forests (R = 0.1823, P < 0.01).
|Subfamily||Species||O. pyramidale secondary forest||Secondary forest||Rain forest||Total||% of total|
|Total number of species||12||12||18|
Bars representing; (a) frugivorous bat diversity (Shannon H’), (b) log frugivorous bat abundance and (c) bat fruit consumption diversity (Shannon H’). Bars with equal letters are not significantly different; based on a two-tailed Hutcheson t-test for the frugivorous bat diversity and bat fruit consumption diversity and an ANOVA and Tuckey post-hoc test for the comparison of bat abundance.
Comparison between most dominant bat species among the forest types revealed highest abundance of both C. sowelli (F2,141 = 9.40, P < 0.001) and C. perspicillata (F2,141 = 3.000, P = 0.053) in O. pyramidale forests, but a similar abundance of C. perspcillata in comparison with secondary forests (Table 1). Both A. jamaicensis (χ2,141 = 148.32, P < 0.001) and A. lituratus (χ2,141 = 111.204, P < 0.001) had higher abundances in rain forests, but their abundances in O. pyramidale forests and secondary forests was similar. Sturnira lilium abundance was highest in O. pyramidale and similar between secondary forests and rain forests (χ2,141 = 14.29, P = 0.001).
Fruit consumption by bats
Bats consumed fruits of a total of 27 plant morphological species (Table 2), of which 16 were identified to species level, three to genus level and two only to family level. We were unable to identify 6 seed species, which represented 6.5 % of the total fruit species consumed. The diversity of consumed fruit species (H’) was similar between the two secondary forest types (Hutcheson t-test t = -0.057, P > 0.05), O. pyramidale forest and rain forest (Hutcheson t-test t = 0.61, P > 0.05) as well as between secondary forest and rain forest (Hutcheson t-test t = -0. 54, P > 0.05; Figure 1c). The number of frugivorous bats that consumed fruits from early-successional trees (χ22,141 = 3.32, P = 0.191) and late-successional shrubs fruits (χ22,141 = 0.51, P = 0.773) was similar among forest type (Table 3). More bats that consumed fruits from early-successional shrubs were captured in O. pyramidale forests (χ22,141 = 25.11, P < 0.001), and more bats consuming late-successional trees larger were captured in rain forests (χ22,141 = 37.40, P < 0.001). The species composition of consumed fruits proved different among forest types (global R = 0.114, P < 0.001). Pairwise tests showed no difference in composition of consumed fruits species between O. pyramidale forests and secondary forests (R = 0.0224, P = 0.156), but did reveal a difference between O. pyramidale forests and rain forests (R = 0.1706, P < 0.001), as well as between secondary forests and rain forests (R = 0.1428, P < 0.001).
|Species||Successional stage||Life form||O. pyramidale secondary forests||Diverse secondary forests||Rain forests|
|Calophyllum brasiliense var.j*||Late||Tree||0||0.0||0||0.0||2||0.6|
|Ficus spp 1||Late||Tree||3||0.9||0||0.0||4||1.2|
|Ficus spp 2||Late||Tree||0||0.0||0||0.0||13||3.8|
|Family Solanacea 1||Shrub||0||0.0||0||0.0||2||0.6|
|Family Solanacea 2||Shrub||0||0.0||0||0.0||1||0.3|
|O. pyramidale forests||Secondary forests||Rain forests|
Forest structure and fruit availability
The forest structure differed among forest types (Table 4), with lower tree density (χ22,69 = 36.30, P < 0.001) in rain forests and higher tree height (χ22,69 = 22.83, P < 0.001) in O. pyramidale forests compared to secondary forests. The diversity of trees was similar between the forest types; O. pyramidale forests and secondary forests (Hutcheson t-test t = -1.24, P > 0.05), O. pyramidale forests and rain forests (Hutcheson t-test t = -1.38, P > 0.05), and between secondary forests and rain forests (Hutcheson t-test t = -0.51, P > 0.05). Canopy openness was highest in O. pyramidale forests (χ22,178 = 135.33, P < 0.001).
|Environmental variable||Ochroma pyramidale forests||Secondary forests||Rain forests|
|Tree density (N/ha)||1450||±||3.39||b||1988||±||6.85||c||875||±||307||a|
|Tree height (m)||12.4||±||1.8||b||9.8||±||1.7||a||12.8||±||2.9||ab|
|Tree diversity (H')^||1.53||±||0.31||a||1.7||±||0.23||a||1.87||±||0.39||a|
|Canopy openness (%)||13.5||±||2.6||c||10.2||±||1.8||b||5.3||±||1.2||a|
|Fruit diversity (H')^||1.1||±||0.23||a||1.5||±||0.2||a||0.8||±||0.14||a|
|Fruit abundance (N/ha)*||76.9||±||47.8||b||102.0||±||57.8||b||18.3||±||10.2||a|
|Piper auritum (N/ha)||1.3||±||1.3||a||4.3||±||1.9||b|
|P. aduncum (N/ha)||4.8||±||5.8||a||2.6||±||2.6||a|
|P. hispidum (N/ha)||3.8||±||3.3||a||3.3||±||3||a|
|P. aeruginosibaccum (N/ha)||0.54||±||0.9||a||2.7||±||4.2||a||1.4||±||2.3||a|
|P. aequale (N/ha)||0.77||±||1.4||a||1.4||±||2.4||a|
|Cecropia obtusifolia (N/ha)||1.5||1.5||a||2.9||±||2.8||b|
Fruits in O. pyramidale forests and secondary forests were originated from only four shrub species (Table 4): Piper auritum, P. aduncum, P. hispidum, P. aeruginosibaccum, and one tree species Cecropia obtusifolia. Piper aequale was only found in secondary forests. Ficus americana, F. maxima and Calophyllum brasiliense were only reported in the rain forest, where also the late-successional shrub species P. aeruginosibaccum and P. aequale were found.
Fruit availability of P. auritum (Z1,94 = -6.64, P < 0.001) and C. obtusifolia (Z1,94 = -0.57, P = 0.034) was higher in secondary forests than in O. pyramidale forests, but for the other Piper species, fruit availability was similar between secondary forest types.
The diversity of available fruit species was similar between all three forests types (Hutcheson t-test t < 1.15, P > 0.05). Whereas the fruit abundance (χ22,36 = 14.35, P = 0.001) was highest in both O. pyramidale forests and secondary forest, and lowest in rain forests.
Predictors of bat diversity and abundance
The best model explaining the differences in bat species diversity included a negative relation with tree density and canopy openness, (Table 5). Bat abundance was best explained by a positive relation with fruit abundance and tree height, and a negative relation with tree density and canopy openness.
|Best model||Number of parameters||AICc||Delta AICc (Δi)||Akaike weight (wi)||Coeff||SE||95% CI||Relative importance|
|(Log) Bat abundance||4||61.73||0||0.30|
Frugivorous bat species abundance ordination
The ordination clustered sites according to forest types, with a higher degree of similarity between O. pyramidale and secondary forest sites (Figure 2). Four of the six environmental variables (canopy openness, tree density, fruit abundance and diversity) proved significantly in the ordination (Table 6). The canopy openness seemed an important factor in separating O. pyramidale forests from secondary forest sites, with several characteristics bat species that tolerate higher canopy openness, such as Carollia sowelli, C. perspicillata, Sturnira lilium and Uroderma billobatum (Figure 2). Tree density, fruit abundance and fruit diversity were highest in secondary forests, and seemed negatively correlated to Phylloderma stenops and Glossophaga commissarisi species abundances, which were more common in rain forest.
Sites: 1-4 O. pyramidale managed forests, 5-8 secondary forests and 9-12 rain forests. Directions of arrows indicate increasing variable value. Italic letters represent frugivorous bat species names, with 1st letter for genus and 2nd letter for species (exception: Centurio senex is C. sx; see Table 1).
Bat composition, diversity, abundance and species-specific abundance
The management of secondary forests with the use of fast growing pioneer tree species O. pyramidale increases the abundance of frugivorous bats, equaling to that in rain forests. The bat species composition differed significantly among the forest types, with relatively more Carollia sowelli and S. lilium in O. pyramidale forests. These relatively small frugivorous bat species are probably attracted by the open canopy as seen from Figure 2, since they prefer early- to mid-successional secondary forests  for its abundance of fruit plants, such as the light demanding Piper spp. and Solanum spp. These plant species generally occur in a clumped distribution, but yield a constant year-round fruit production and therefore allow for food specialization of these bat species [50–53]. Even though the availability of Piper auritum, which is preferred by Carollia sowelli and generally forms an important part of its diet [54,55], was more abundant in secondary forests, this did not seem to cause a higher abundance of Carollia sowelli than in O. pyramidale forests. This suggests that other factors, such as structural vegetation characteristics like canopy openness might also be important for these bat species when searching for food. Larger frugivorous bats Artibeus lituratus and A. jamaicensis bats were abundantly present in rain forests, which provided them temporary but massive fruit availability of asynchronous, patchy distributed late-successional tree species [50,53]. Besides a preference for Ficus spp. , both large Artibeus bat species are also know for consuming Cecropia fruits . The secondary forest patches with Cecropia fruit availability could function as an alternative source of food for both large frugivorous bat species in periods of low Ficus fruit abundance, but even though the fruit availability of Cecropia was higher in secondary forests in comparison with O. pyramidale forest, this did not result in a higher abundance of either Artibeus species.
Fruit availability and consumption
Managed O. pyramidale forests showed higher density of shrubs , but this did not result in higher fruit diversity or fruit abundance in comparison with other forest types. The composition of consumed fruits was significantly different between rain forests and both O. pyramidale forests and secondary forests, probably because various species of late-successional fruits that formed part of bat consumption (e.g. Ficus spp.) were only available in rain forests. This is confirmed by the fact that more bats consumed late-successional tree fruits in rain forests. Overall consumption consisted of mostly early-successional shrubs, which supports the fact that bats are known for their preference in early-successional shrub and tree fruits, and therefore have been indicated as facilitating successional development of early-successional areas [7,56]. As expected, secondary forests management with O. pyramidale increased the consumption of early-successional shrub fruits. Even though the diversity and abundance of fruits available in each of the secondary forest types was similar, the positive association between canopy openness and abundance of light demanding early-successional shrub species could be sufficient to attract more individuals of certain bat species such as Carollia and Sturnira lilium, with a preference for their fruits . The consumption of late-successional fruits by bats in both secondary forest types was low, probably due to the absence of late-successional tree species and low abundance of large frugivorous bats, such as A. lituratus and A. jamaicensis that feed on these fruits.
Structural variables and fruit availability underlying differences in frugivorous bats
Both secondary and O. pyramidale forests, although with 10-15 years of vegetation succession, provided fruits from mostly early-successional and mid-to late-successional shrubs, with highest fruit diversity in secondary forests. However, only the abundance of fruits was positively correlated to the bat abundance, instead of fruit diversity, supporting the findings of an earlier study on increasing bat abundance with increasing fruit availability in agroecosystems . Additionally, in areas of similar (temporal) fruit abundance, bat abundance and diversity is probably determined more by vegetation structural differences, such as tree density, height and canopy openness. In our study these variables were related with the bat assemblage structure, as also supported by other publications [18,21,57].
Tree density, previously reported to affect bat assemblages [58–60], was an important factor in the best models, negatively affecting both frugivorous bat diversity and abundance respectively. Higher densities of trees, as found in secondary forests , usually restricts the movement of bats, reducing their flight efficiency  and interferes with clutter echoes . Tree height has been previously identified as an important characteristic related with bat roosts [63,64]. Tree height was highest in rain forest, offering suitable bat roosting sites .
Besides a lower tree density, O. pyramidale forests presented higher canopy openness than secondary forests, which was important in the second strongest model, negatively related to bat abundance. This is also found by previous studies [21,23,24,50,66], where bat movement of certain species, such as Vampyressa thyone, Vampyrodes caracciolii and Phyllostomus discolor was reduced due to lack of cover from predators  or because their preferred fruit species are generally not found under open canopies. As shown from the NMDS, the open canopy could positively affect the arrival of certain bat species such as C. sowelli, C. perspicillata and abundantly present in secondary forest types and tolerant towards areas of higher canopy openness, where light demanding fruit plants are more likely to occur. The relative open canopy in secondary forest dominated by O. pyramidale permits light to reach the understory, which is beneficial for the colonization of early-successional plant species and consequently provide food and shelter for generalist dispersers , as well as improving conditions favorable for the arrival and growth of mid-to late-successional plant species. However, without the reduction of canopy closure in O. pyramidale forests and the continuous dispersion of mostly early-successional shrub species by S. lilium and Carollia spp., the growth of late-successional plant species could be suppressed, delaying the successional development of these 10-15 year old forests .
We found that secondary forest management promoting trees of O. pyramidale affects the forests’ structural characteristics and fruit availability, and thereby alter the abundance of frugivorous bats, overall frugivorous species composition and affecting the movement of specific bat species such as Carollia sowelli and S. lilium towards more open O. pyramidale managed secondary forests. Hence, the application of forest management strategies can trigger cascading effects and consequently direct or change the speed of forest succession by affecting the arrival of important seed dispersers such as bats.
We are grateful to M. Castellano and E. Chankin for allowing us to work on their lands. Furthermore, we would like to thank J. Chankin, A. Chankin, C. Peñaloza Guerrero, R. van Toor, V. Hommersen, M. Wulms and A. Sánchez for their help in the fieldwork.
Conceived and designed the experiments: IV SILT JGG LBV. Performed the experiments: IV. Analyzed the data: IV WFdB. Wrote the manuscript: IV SILT JGG LBV WFdB.
- 1. Asner GP, Rudel TK, Aide TM, Dedfries RS, Emerson R (2009) A contemporary assessment of change in humid tropical forests. Conserv Biol 23: 1386-1395. doi:https://doi.org/10.1111/j.1523-1739.2009.01333.x. PubMed: 20078639.
- 2. FAO (2010) Global forest resources assessment 2010, Main report, FAO Forestry Paper 163. Rome.
- 3. Muscarella R, Fleming TH (2007) The role of frugivorous bats in tropical forest succession. Biol Rev 82: 573-590. doi:https://doi.org/10.1111/j.1469-185X.2007.00026.x. PubMed: 17944618.
- 4. Kunz TH, Braun de Torrez E, Bauer D, Lobova T, Fleming TH (2011) Ecosystem services provided by bats. Ann N Y Acad Sci 1223: 1-38. doi:https://doi.org/10.1111/j.1749-6632.2011.06004.x. PubMed: 21449963.
- 5. Howe HF, Smallwood J (1982) Ecology of seed dispersal. Annu Rev Ecol Syst13: 201-228. doi:https://doi.org/10.1146/annurev.es.13.110182.001221.
- 6. Janson CH (1983) Adaptations of fruit morphology to dispersal agents in Neotropical forest. Science 219: 187-189. doi:https://doi.org/10.1126/science.219.4581.187. PubMed: 17841688.
- 7. Galindo-González J, Guevara S, Sosa VJ (2000) Bat-and bird-generated seed rains at isolated trees in pastures in tropical rain forest. Conserv Biol 14: 1693-1703. doi:https://doi.org/10.1046/j.1523-1739.2000.99072.x.
- 8. Nations J, Nigh R (1980) The evolutionary potential of Lacandon Maya sustained yield tropical forest agriculture. J Anthropol Res 36: 1-30.
- 9. Levy-Tacher SI (2000) Sucesión causada por roza-tumba-quema en las selvas de Lacanhá, Chiapas. PhD dissertation, Colegio de Posgraduados; Montecillo Texcoco, Estado de México, México. 165 p.
- 10. McGee RJ (2002) Watching Lacandon Maya. Boston: Allyn and Bacon.
- 11. Vleut I, Levy-Tacher SI, de Boer WF, Galindo-González J, Ramírez-Marcial N (2013) Can a fast-growing early-successional tree (Ochroma pyramidale, Malvaceae) accelerate forest succession? J Trop Ecol 29: 173-180. doi:https://doi.org/10.1017/S0266467413000126.
- 12. Dalling JW, Lovelock CE, Hubbell SP (1999) Growth responses of seedlings of two neotropical pioneer species to simulated forest gap environments. J Trop Ecol 15: 827-839. doi:https://doi.org/10.1017/S0266467499001200.
- 13. Pearson TRH, Burslem DFRP, Goeriz RE, Dalling JW (2003) Regeneration niche partitioning in neotropical pioneers: eﬀects of gap size, seasonal drought and herbivory on growth and survival. Oecologia 137: 456-465. doi:https://doi.org/10.1007/s00442-003-1361-x. PubMed: 12920642.
- 14. Levy-Tacher SI, Golicher JD (2004) How predictive is traditional ecological knowledge? the case of the Lacandon Maya Fallow Enrichment System. Interciencia 29: 496-503.
- 15. Diemont SA, Martin JF, Levy-Tacher SI, Nigh RB, Lopez PR, Golicher JD (2006) Lacandon Maya forest management: restoration of soil fertility using native tree species. Ecol Eng 28: 205-212. doi:https://doi.org/10.1016/j.ecoleng.2005.10.012.
- 16. Douterlungne D, Levy-Tacher SI, Golicher JD, Román-Dañobeytia F (2010) Applying indigenous knowledge to the restoration of degraded tropical rain forest dominated by bracken. Res Ecol 18: 322-329.
- 17. Medellín RA, Equihua M, Amin MA (2000) Bat Diversity and Abundance as Indicators of Disturbance in Neotropical Rain forests. Conserv Biol 14: 1666-1675. doi:https://doi.org/10.1046/j.1523-1739.2000.99068.x.
- 18. Kalko EKV, Handley CO Jr (2001) Neotropical bats in the canopy: diversity, community structure and implications for conservation strategies. Plant Ecol 153: 319-333. doi:https://doi.org/10.1023/A:1017590007861.
- 19. Willig MR, Presley SJ, Bloch CP, Hice CL, Yanoviak SP, Diaz MM, Chauca LA, Pacheco V, Weaver SC (2007) Phyllostomid bats of lowland Amazonia: effects of habitat alteration on abundance. Biotropica 39: 737-746. doi:https://doi.org/10.1111/j.1744-7429.2007.00322.x.
- 20. Castro-Luna AA, Galindo-González J (2012) Enriching agroecosystems with fruit-producing tree species favors the abundance and richness of frugivorous and nectarivorous bats in Veracruz, Mexico. Mamm Biol 77: 32-40.
- 21. Vleut I, Levy-Tacher SI, Galindo-González J, de Boer WF, Ramírez-Marcial N (2012) Tropical rain forest matrix quality affects bat assemblage structure in secondary forest patches. J Mamm 93: 1469-1479. doi:https://doi.org/10.1644/12-MAMM-A-005.1.
- 22. Russo D, Cistrone L, Jones G (2007) Emergence time in forest bats: the influence of canopy closure. Acta Oecol 31: 119-126. doi:https://doi.org/10.1016/j.actao.2006.11.001.
- 23. Gorresen PM, Willig MR (2004) Landscape-scale responses of bats to habitat fragmentation in Atlantic Rain forest of Paraguay. J Mamm 85: 688-697. doi:https://doi.org/10.1644/BWG-125.
- 24. Ford WM, Menzel MA, Rodrigue JL, Menzel JM, Johnson JB (2005) Relating bat species presence to simple habitat measures in a central Appalachian forest. Biol Conserv 126: 528-539. doi:https://doi.org/10.1016/j.biocon.2005.07.003.
- 25. Pennington TD, Sarukhán J (2005) Árboles tropicales de México. Tercera edición. Universidad Nacional Autónoma de México y Fondo de Cultura Económica, México, D.F.
- 26. MacSwiney G, Clarke FM, Racey PA (2008) What you see is not what you get: the role of ultrasonic detectors in increasing inventory completeness in Neotropical bat assemblages. J Appl Ecol 45: 1364-1371. doi:https://doi.org/10.1111/j.1365-2664.2008.01531.x.
- 27. Straube FC, Bianconi GV (2002) Sobre a grandeza e a unidade utilizada para estimar esforço de captura com utilização de redes-deneblina. Chiroptera Neotrop 8: 150-152.
- 28. Palmeirim JM, Etheridge K (1985) The influence of man-made trails on foraging by tropical frugivorous bats. Biotropica 17: 82-83. doi:https://doi.org/10.2307/2388385.
- 29. Williams-Linera G (1990) Vegetation structure and environmental conditions of forest edges in Panama. J Ecol: 356-373.
- 30. Reid FA (1997) A field guide to the mammals of Central America and southeast Mexico. New York: Oxford University Press.
- 31. Medellín RA, Arita HT, Sánchez O (2008) Identificación de los murciélagos de México: C. de campoS. Edición. México: Instituto de Ecológica, UNAM.
- 32. Kunz TH, Ingalls KA (1994) Folivory in bats: an adaptation derived from frugivory. Funct Ecol 8: 665-668.
- 33. Herrera GL, Gutierrez E, Hobson KA, Altube B, Díaz WG, Sánchez-Cordero V (2002) Sources of assimilated protein in five species of New World frugivorous bats. Oecologia, 133: 280-287. doi:https://doi.org/10.1007/s00442-002-1036-z.
- 34. Fleming TH (1986) Opportunism versus specialization: the evolution of feeding strategies in frugivorous bats. In: A. EstradaTH Fleming. Frugivores and seed dispersal. Netherlands: Springer Verlag. pp. 105-118.
- 35. Galindo-González J, Vázquez-Domínguez G, Saldaña-Vázquez RA, Hernández-Montero JR (2009) A more efficient technique to collect seeds dispersed by bats. J Trop Ecol 25: 205-209. doi:https://doi.org/10.1017/S0266467409005859.
- 36. Swaine MD, Whitmore TC (1988) On the definition of ecological species groups in tropical rain forests. Vegetatio 75: 81-86. doi:https://doi.org/10.1007/BF00044629.
- 37. Whitmore TC (1989) Canopy gaps and the two mayor groups of forest trees. Ecology 70: 536-538. doi:https://doi.org/10.2307/1940195.
- 38. Wulms M (2009) Bats as indicator species of forest recovery in areas of secondary forest along a gradient of time since abandonment in Lacanhá Chansayab, Chiapas, Mexico. Bachelor dissertation. , University of Applied Sciences, Van Hall-Larenstein, Velp, the Netherlands.
- 39. Krebs CJ (1989) Ecological methodology. New York: Harper and Row.
- 40. Colwell RK (2009) EstimateS: statistical estimation of species richness and shared species from samples, version 7.0.0. Available http://viceroy.eeb.uconn.edu/EstimateS.
- 41. Clarke KR (1993) Non-parametric multivariate analyses of changes in community structure. Aust J Ecol 18: 117-143. doi:https://doi.org/10.1111/j.1442-9993.1993.tb00438.x.
- 42. Hutcheson K (1970) A test for comparing diversities based on the Shannon formula. J Theor Biol 29: 151-154. doi:https://doi.org/10.1016/0022-5193(70)90124-4. PubMed: 5493290.
- 43. Box GE, Cox DR (1964) An analysis of transformations. J R Stat Soc B Stat Methodol) 26: 211-252.
- 44. Zar JH (1999) Biostatistical analysis. India: Pearson Education.
- 45. Burnham KP, Anderson DR (2002) Model selection and multi-model inference: a practical information-theoretic approach. Springer Verlag.
- 46. TEAM, R. Development Core (2005) R: A language and environment for statistical computing. ISBN 3-900051-07-0. Vienna, Austria: R Foundation for Statistical Computing. p. 2013. Available: http://www. R-project. org, 2005 .
- 47. Magurran AE (2004) Measuring Biological Diversity. Oxford, UK: Blackwells Publishing House.
- 48. McCune B, Grace JB, Urban DL (2002) Analysis of ecological communities. MjM Software Design, Gleneden Beach.
- 49. Castro-Luna AA, Sosa VJ, Castillos-Campos G (2007) Bat diversity and abundance associated with the degree of secondary succession in a tropical forest mosaic in south-eastern Mexico. Anim Conserv 10: 219-228. doi:https://doi.org/10.1111/j.1469-1795.2007.00097.x.
- 50. Gentry AH (1974) Coevolutionary patterns in Central American Bignoniaceae. Ann Mo Bot Gard 69: 728-759.
- 51. Fleming TH, Heithaus ER (1981) Frugivorous Bats, Seed Shadows, and the Structure of Tropical Forests. Biotropica 13: 45-53. doi:https://doi.org/10.2307/2388069.
- 52. Marinho-Filho JS (1991) The Coexistence of Two Frugivorous Bat Species and the Phenology of Their Food Plants in Brazil. J Trop Ecol 7: 59-67. doi:https://doi.org/10.1017/S0266467400005083.
- 53. Dumont ER (2003) Bats and fruit: an ecomophological approach. In: TH KunzMB Fenton. Bat Ecology. Chicago, IL.: University of Chicago Press. pp. 398-429.
- 54. da Silva AG, Gaona O, Medellín RA (2008) Diet and trophic structure in a community of fruit-eating bats in Lacandon Forest, Mexico. J Mamm 89: 43-49. doi:https://doi.org/10.1644/06-MAMM-A-300.1.
- 55. Fleming TH (1991) The relationship between body size, diet, and habitat use in frugivorous bats, genus Carollia (Phyllostomidae). J Mamm 72: 493-501. doi:https://doi.org/10.2307/1382132.
- 56. Medellín RA, Gaona O (1999) Seed dispersal by bats and birds in forest and disturbed habitats of Chiapas, Mexico. Biotropica 31: 478-485. doi:https://doi.org/10.1111/j.1744-7429.1999.tb00390.x.
- 57. Schulze MD, Seavy NE, Whitacre DF (2000) A comparison of the Phyllostomid bat assemblages in undisturbed Neotropical forest and in forest fragments of a Slash-and-Burn Farming Mosaic in Petén, Guatemala. Biotropica 32: 174-184. doi:https://doi.org/10.1111/j.1744-7429.2000.tb00459.x.
- 58. Estrada A, Coates-Estrada R (2001) Bat species richness in live fences and in corridors of residual rain forest vegetation at Los Tuxtlas, Mexico. Ecography 24: 94-102. doi:https://doi.org/10.1034/j.1600-0587.2001.240111.x.
- 59. Lumsden LF, Bennett AF (2005) Scattered trees in rural landscapes: foraging habitat for insectivorous bats in southeastern Australia. Biol Conserv 122: 205-222. doi:https://doi.org/10.1016/j.biocon.2004.07.006.
- 60. Fischer J, Stott J, Law BS (2010) The disproportionate value of scattered trees. Biol Conserv 143: 1564-1567. doi:https://doi.org/10.1016/j.biocon.2010.03.030.
- 61. Jones G, Rayner JMV (1991) Flight performance, foraging tactics and echolocation in the trawling insectivorous bat Myotis adversus (Chiroptera: Vespertilionidae). J Zool Lond 225: 393-412. doi:https://doi.org/10.1111/j.1469-7998.1991.tb03824.x.
- 62. Schnitzler HU, Kalko EK (2001) Echolocation by Insect-Eating Bats: We define four distinct functional groups of bats and find differences in signal structure that correlate with the typical echolocation tasks faced by each group. BioScience 51: 557-569. doi:https://doi.org/10.1641/0006-3568(2001)051[0557:EBIEB]2.0.CO;2.
- 63. Crampton LH, Barclay RMR (1998) Selection of roosting and foraging habitat by bats in different-aged aspen mixedwood stands. Conserv Biol 12: 1347-1358. doi:https://doi.org/10.1046/j.1523-1739.1998.97209.x.
- 64. Lacki MJ, Baker MD (2003) A prospective power analysis and review of habitat characteristics used in studies of tree-roosting bats. Acta Chiropt 5: 199-208. doi:https://doi.org/10.3161/001.005.0211.
- 65. Kunz TH, Lumsden LF (2003) Ecology of cavity and foliage roosting bats. In: TH KunzB. Fenton. Bat ecology. Chicago, Illinois: University of Chicago Press. pp. 3-89.
- 66. Crome FHJ, Richards GC (1988) Bats and gaps: microchiropteran community structure in a Queensland rain forest. Ecology 69: 1960-1969. doi:https://doi.org/10.2307/1941173.
- 67. Fleming TH (1986) Opportunism versus specialization: the evolution of feeding strategies in frugivorous bats. In: A. EstradaTH Fleming. Frugivores and seed dispersal. Dordrecht, The Netherlands: Dr. W. Junk Publishers. pp. 1105-1118.
- 68. Connell JH, Slatyer RO (1977) Mechanisms of succession in natural communities and their role in community stability and organization. Am Nat 111: 1119-1144. doi:https://doi.org/10.1086/283241.
- 69. Greig N (1993) Regeneration mode in neotropical Piper: habitat and species comparisons. Ecology 74: 2125-2135. doi:https://doi.org/10.2307/1940857.
- 70. Guevara S, Laborde J, Liesenfeld D, Barrera O (1997) Potreros y ganadería. In: E. González-SorianoR. DirzoR. Vogt. Historia Natural de Los Tuxtlas. Instituto de Biología, Instituto de Ecología, Universidad Nacional Autónoma de México, D.F. pp. 43-58.
- 71. Levy-Tacher SI, Aguirre-Rivera JR (2005) Successional pathways derived from different vegetation use patterns by Lacandon Mayan Indians. J Sustain Agric 26: 49-82. doi:https://doi.org/10.1300/J064v26n01_06.