- Top of page
- Materials and Methods
- Conflict of Interest
- Appendix A1
Climate change, land-use change, and harvesting are drivers of species extinction of global importance (Foley et al. 2005; Beever and Belant 2012). Therefore, understanding the effects of these factors is essential for the design of conservation, sustainable management, and restoration plans. In many cases, these factors act simultaneously on the species, but their combined effects need to be investigated. The little evidence available reveals that a common species response to the combined effects of such factors is unlikely. For instance, habitat degradation and climate change are likely to decrease habitat specialists and benefit some mobile and generalist species of British butterflies (Warren et al. 2001). For exploited species, harvesting effects adds a further complication, which combined effects with other potential drivers of species extinction are yet to be explored.
Epiphytes comprise 10% of all vascular plants, and some epiphyte species are among the most vulnerable organisms to climate change, harvesting, and disturbance (Benzing 1998; Ozanne et al. 2003). Climate change may affect epiphytes directly and indirectly. As some of these plants obtain nutrients and water from the atmosphere, changes in climate, in particular in temperature and humidity, affect their water and nutrient supplies (Obregona et al. 2011). Climate change affects epiphytes indirectly by affecting the host plants on which they live (Hsu et al. 2012). Global warming, for instance, is expected to promote uphill displacements and range contractions of populations inhabiting mountain ranges (Wilson and Gutiérrez 2012). Such changes may benefit the epiphytes at certain altitudes and harm at others. The hosts are likely to be affected as well, although not necessarily with the same effects observed in the epiphytes.
Land-use change usually reduces the abundance of host trees and, consequently, is expected to decrease the abundance and diversity of epiphytes. Disturbances associated with land-use change frequently result in higher solar radiation, temperature and wind velocity, and lower humidity compared with the original conditions, resulting in a gradual substitution of mesic epiphytes by more xerophytic epiphytes (Cascante-Marín et al. 2006). Epiphyte harvesting is very common in certain geographical regions (Flores-Palacios and Valencia Díaz 2007), but its consequences are poorly known. Harvesting may endanger some populations (Borba et al. 2007). However, harvesting effects may depend on human culture and taxonomic groups and may allow a higher degree of biodiversity conservation, compared with other land-use alternatives, such as the conversion of forest to pastures (Ruiz-Pérez 1997).
Assessing the mechanisms underlying the impacts of human activities on epiphytes such as harvesting requires the inclusion of multiple ecological levels ranging from genes to ecosystems (Ticktin 2004). Population genetic studies are useful to detect the genetic diversity of populations. Theory predicts that the capability of populations to respond to environmental changes depends on their genetic diversity. This appears to be the case for climate change in aquatic plant species (Ehlers and Reusch 2008), and for fragmentation in tropical tree species (del Castillo et al. 2011). Human activities have yielded conflicting results in epiphyte genetic diversity. Harvesting has been associated with decreases in genetic diversity (Cruse-Sanders and Hamrick 2004), but some epiphytes maintain considerable levels of genetic diversity in fragmented habitats (Avila Diaz and Oyama 2007). In addition to estimating genetic diversity, population genetic analyses allow exploring the effects of fragmentation on gene flow (Murren 2003), and detecting past bottlenecks (Alcántara et al. 2006), processes that may affect the population's abundance and persistence. Demographic studies are useful for assessing the effects of anthropogenic factors on populations. Harvesting, for instance, is one source of mortality, which affects the population vital rates and may increase the extinction risks in epiphytes (Mondragón 2009; Mondragón and Ticktin 2011). Finally, the consideration of the epiphyte ecosystems and its dynamics may help to interpret the observed population characteristics of epiphytes in ecosystems affected by anthropogenic activities. Forest dynamics, in particular, is expected to play a crucial role in epiphyte abundance as shown in simulation models (e.g., Hsu et al. 2012). Vegetation dynamics usually alter the distribution of epiphytes' microhabitats (Zartman 2003).
Despite their importance, comprehensive studies on epiphytes that include population genetics, demography, and the dynamics of the specific portions of the landscape at which the studied population dwells are lacking. As part of a long-term project aiming at fomenting the conservation of native forests and their species in southern Mexico, we conducted a multidisciplinary study on Catopsis compacta in Santa Catarina Ixtepeji, Oaxaca. Deforestation is significant, and the current climatic conditions appear to be displaced to higher altitudes. Furthermore, this epiphyte is subjected to extraction. Because of harvesting, habitat fragmentation, and climate change, we expect a declining and subdivided population of Catopsis in Santa Catarina with low genetic diversity, inbreeding, and declining population size. Our objectives were as follows: (1) to assess the status of the population through demographic and population genetic analyses; (2) to identify the patterns of distribution of C. compacta within the forest fragments; and (3) to assess the patterns of deforestation in the vegetation that host C. compacta using satellite images and GIS, and use this information to explain the demographic and genetic results obtained in (1).
- Top of page
- Materials and Methods
- Conflict of Interest
- Appendix A1
Species inhabiting human-altered environments are commonly affected by a combination of potential threatening factors of anthropogenic origin, such as harvesting, deforestation, and climate change, which may jeopardize the persistence of their populations. Harvesting is predicted to modify the structure and dynamics of the populations, alter population subdivision, decrease genetic diversity, and induce selective genetic changes (Allendorf et al. 2008). When the entire individual is removed, as is the case of C. compacta, harvesting imposes an additional source of mortality, because the harvested individual no longer contributes offspring and genes to the population. Habitat fragmentation, a widespread consequence of perturbation of natural habitats, is commonly expected to cause habitat losses and increases in the degree of isolation of the remaining populations (Lindenmayer and Fischer 2006). Habitat losses or at least habitat displacements, due to changes of the precipitation and temperature regimens are also an expected consequence of climate change (Wilson and Gutiérrez 2012). Habitat losses, mortality, population, isolation, and reductions in population genetic diversity, in turn, are likely to increase the probabilities of extinction by genetic and demographic stochasticty (Caswell 2001; Allendorf and Luikart 2007). As such, the net impact of climate change, fragmentation and harvesting can be expected to be even harsher than the separate impacts of each of these factors, because all of them can stress individually the populations. The present study, however, shows that C. compacta can maintain a growing population and displays a high genetic variation, even with harvesting and high levels of fragmentation and provides evidence that antagonistic effects of fragmentation with harvesting can explain, at least in part, these results.
In small or subdivided populations, such as those affected by harvesting and habitat fragmentation, genetic drift is expected to reduce population genetic diversity (Allendorf and Luikart 2007). Indeed, according to a recent meta-analysis, habitat fragmentation decreases population genetic diversity in most studied species (Aguilar et al. 2008), and harvesting may do so in some cases (e.g., Cruse-Sanders and Hamrick 2004). Ideally, assessing the effects on potential stressors of the populations requires comparisons of several populations before and after the effects of such stressors take place. In lieu of such kind of comparison, testing for population subdivision, and looking for evidences of past bottlenecks can help to reveal the expected effects of such stressors on the genetic attributes of the populations. The reported genetic variation of other species with similar breeding systems, dispersal mechanisms, or taxonomic affinity can provide a surrogate reference for comparative purposes with the genetic variation observed in the tested population. For this purpose, the use of allozymes as genetic markers results convenient given the extensive number of species studied and published meta-analysis, such as that of Hamrick and Godt (1996) with 247 plant species analyzed (but see Lowe et al. 2004, for some drawbacks of this kind of meta-analysis).
In spite of at least 50 years of harvesting and a continuous fragmentation of the habitat for at least 27 years, the population of C. compacta does not show evidences of population subdivision or past bottlenecks. The significant Fis detected in our multilocus analysis is most likely due to mating among relatives particularly between individuals sharing the same host trees as pollinators tend to fly short distances (personal observation). On the other hand, the nonsignificant Fst value suggests that the current trends of fragmentation have not caused population subdivisions between the oak forest and the chaparral fragments. The estimators of genetic diversity found in C. compacta were the highest reported in Bromeliaceae (Table 2) and are among the highest reported in plant species from other families. For instance, our values of genetic diversity (0.405 vs. 0.157, outcrossing-wind-dispersed seeds; or 0.165, outcrossing-monocots) and percentage of polymorphic loci (81.8 vs. 62.4%, outcrossing-wind-dispersed seeds; or 52.5% outcrossing-monocots) were higher than the average reported in species with similar breeding system and seed dispersal mechanism, or taxonomic group using isozymes (cf., Hamrick and Godt 1996). In summary, our data do not show evidences that genetic erosion, population subdivision between adjacent vegetation fragments, or past bottlenecks have affected the population of C. compacta.
The high genetic variation in C. compacta requires further investigations, but at least five factors may contribute to this result. Firstly, dioecy prevents selfing, the most extreme source of inbreeding. Secondly, among wind-dispersed propagules, plumed seeds, such as those C. compacta, are reportedly highly effective for long-distance dispersal (Cousens et al. 2008). Thirdly, the epiphytic condition may favor long-distance dispersal, by positioning the plants on the host trees, from which height dispersal can reach longer distances than ground positions. Fourthly, as explained below, the expansion of the perimeter of the fragments, during vegetation fragmentation, as shown by the GIS analyses, may have favored long-distance seed dispersal. Gene flow reduces inbreeding, decreases the losses of genetic variation by drift, and homogenizes subpopulations as supported by the nonsignificant Fst value. Finally, the expansion of the C. compacta population, as suggested by the demographic analyses, may contribute to decrease the probabilities of losses of genetic variation by drift, and to generate genetic variation by mutations.
Fragmentation is not a synonymous of habitat loss, but a change in landscape structure (Fahrig 2003). One of the most obvious changes in structure during fragmentation is the formation of borders in the vegetation fragments, which are interfaces between the fragments and the habitat that surrounds them. In Santa Catarina, the total area of the forest decreased 76% on average in 27 years. However, our plant distribution analysis shows that most C. compacta individuals grow on host plants located in a fringe of vegetation at the border of the fragments. The fragment perimeter increased at a relatively high rate in both the oak forest (0.047 years−1, td = 14.8 years) and the oak-chaparral fragments (0.034 years−1, td = 20.2 years), displaying similar values to the estimated intrinsic rate of increase in the population (0.046 years−1, td = 15.1 years). Logging is a factor that may have contributed to increase open spaces in the vegetation. Wood is commonly extracted for diverse purposes including firewood and lumber. Climate change is another factor. A drier and warmer climate trend is likely affecting the study area. Under such conditions, oak forests are expected to be replaced by oak-chaparrals with low height and low canopy cover (Fig. 2). The greater abundance of open spaces in the chaparral may have increased the probabilities of C. compacta colonization (see Fig. 3). Other species of epiphytes appear to be benefited by open spaces such as those with water-absorbing trichomes, the atmospheric-type epiphytes, which usually replace more mesic species under such conditions (e.g., Flores-Palacios and García-Franco 2004; Winkler et al. 2005; Cascante-Marín et al. 2006; Werner 2011).
In C. compacta, the edges of the vegetation may result more convenient than the forest interior for both physiological and genetic reasons. Firstly, C. compacta has thick leaves with a thick cuticle and water-absorbing trichomes, which are well-recognized adaptations to xeric environments in Bromeliaceae (Medina 1974). Trichomes, however, partially block the incidence of light to the chloroplasts and hinder gas-exchange reducing, therefore, photosynthetic rates, particularly in shady environments. The oak forests near Santa Catarina have a relatively dense foliage (73% cover; Zacarías-Eslava and del Castillo 2010), which may reduce the probabilities that C. compacta individuals in the core of fragments display a positive carbon balance, particularly during the rainy season, when the host trees are fully covered by leaves. By contrast, the plants on hosts at the edge of the fragments are expected to receive more lateral light on the open side of the edges, increasing the probabilities that xeromorphic epiphytes, such as C. compacta, display a positive carbon balance. This conclusion is also supported by the fact that our PVA analyses show that quasi extinction probabilities are zero even when the frequency of rainy years decrease to 2%. Secondly, open spaces favor a laminar flow of the wind, compared to spaces with complex vegetation, which promotes turbulence (Whitehead 1983). Thus, for species with wind-dispersed seeds, the edge of the fragments likely increase gene flow and decrease population subdivision caused by habitat fragmentation by favoring long-distance dispersal. Gene flow appears to be favored by fragmentation in other plant species (see Young and Clarke 2000, for some examples).
The increases in the perimeter of the vegetation patches and the openness of the vegetation, likely enhanced by climate change and logging, appear to favor the population expansion of C. compacta in Santa Catarina, by increasing the habitat at which new individuals can become established and by favoring long-distance seed dispersal. This result and the fact that all possible hosts have not been colonized indicate that habitat size is not constraining the population growth of Catopsis, under the prevailing conditions. By contrast, harvesting increases the mortality of the adults, one of the stages that most contributes to population growth, according to our elasticity analysis. Harvesting was also the factor with the highest influence in quasi-extinction probabilities according to our PVA. Climate change and deforestation appear to increase the habitat and the opportunities for higher dispersal of Catopsis, but harvesting partially counteracts such effects by reducing the number of reproductive individuals. At present, however, the levels of harvesting do not appear to harm significantly the population, which appears to keep growing, maintain a stable size structure and a high genetic diversity.
The conditions on which our analyses based are likely to change, and the prevalence of this species in the future in Santa Catarina is not warranted. If the levels of harvesting duplicates, the risks of extinction increases sharply, as indicated by our PVA. In general, the total edge of the fragments tends to increase fast during the early stages of fragmentation. As fragmentation proceeds, the rate of increase of the perimeter is expected to decline, and, eventually, the total edge decreases first slow and then fast had fragmentation continue (Fahrig 2003). Indeed, in the less fragmented oak forest, the rate of increase in the perimeter was higher than that in the more open oak-chaparral. If the estimated rates of deforestation continue, the perimeter of the vegetation patches are likely to decline, together with the extinction risks of C. compacta owing to the ecological and genetic stochasticity. Eventually, the distribution of C. compacta is expected to be positively correlated with vegetation as predicted in simulations (Hsu et al. 2012).
Traditionally, the creation of reserves has been identified as the ultimate solution for biodiversity conservation. For species such as C. compacta, however, undisturbed reserves are not an option. These species requires periodic disturbances that create forest openings. We have shown that C. compacta can take advantage of moderate levels of disturbance and a warmer and drier climate both of which may generate open spaces in the vegetation on whose borders this species can prosper. For edge-inhabiting species, the creation of artificial edges appears to be a good prescription for conservation and management (Lindenmayer and Fischer 2006). Moderate plant extraction and moderate logging may create artificial edges in Santa Catarina bringing about benefits to the local people. However, a further complication arises, because C. compacta and their hosts have contrasting requirements for survival and reproduction, as suggested by the fact that C. compacta population has a positive and significant stochastic λ-value but the forest is decreasing and regeneration is poor (Zacarías-Eslava and del Castillo 2010). Conservation and management practices should take a careful consideration of the requirements of both the host and the epiphytes together with the dynamics of the habitat, and the demographic and genetic context of the species to be preserved. Certainly, a close monitoring at population and community level is needed in combination with adaptive management techniques (Millar et al. 2007). These practices should be particularly relevant for species inhabiting transient habitats such as those affected by climate change and disturbance for a sound management and conservation planning.
Catopsis compacta is an example of a species in which different anthropogenic activities, such as harvesting and deforestation, probably induced by logging and climate change, can inflict antagonistic effects on the population under certain conditions, and display contrasting responses to such factors relative to its hosts. The conservation and sustainable management of this kind of system require a clever administration of disturbance rather than preventing it through reserves in which disturbance is suppressed. A full understanding of the consequences of potential threatening factors on species persistence or extinction should only be accomplished by inspecting the action and interaction of all factors that might affect the population and with the consideration of such effects on the species on which the focus species depends. For these goals, our results illustrate the advantages of multidisciplinary approaches, each of which provide relevant and essential information.