Can Overharvesting of a Non-Timber-Forest-Product Change
Forest Ecology and Management 324 (2014) 117–125
Contents lists available at ScienceDirect
Forest Ecology and Management
journal homepage: www.elsevier.com/locate/foreco
Can overharvesting of a non-timber-forest-product change the regeneration dynamics of a tropical rainforest? The case study of Euterpe edulis ⇑ Ana Elena Muler a, Débora C. Rother b, , Pedro S. Brancalion c, Rafaela P. Naves a, Ricardo R. Rodrigues a, Marco A. Pizo d a Departamento de Ciências Biológicas, ESALQ – Universidade de São Paulo (USP), Av. Pádua Dias, 11, Caixa Postal 9, Piracicaba, CEP 13418-900 São Paulo, Brazil b Programa de pós-graduação em Biologia Vegetal, Campus Rio Claro – Universidade Estadual Paulista (UNESP), Av. 24A, 1515, Caixa Postal 199, CEP 13506-900 Rio Claro, São Paulo, Brazil c Departamento de Ciências Florestais, ESALQ – Universidade de São Paulo (USP), Av. Pádua Dias, 11, Caixa Postal 9, Piracicaba, CEP 13418-900 São Paulo, Brazil d Departamento de Zoologia, UNESP – Universidade Estadual Paulista ‘‘ Júlio de Mesquita Filho’’ (UNESP), Av. 24-A, 1515, Bela Vista, CEP 13506-900 Rio Claro, São Paulo, Brazil article info abstract
Article history: The exploitation of non-timber forest products is often considered a low-impact activity in tropical for- Available online 7 October 2013 ests. However, assessments of the impacts of such activity are mostly focused on the harvested species and not on the plant community, thus limiting our understanding for establishing forest management Keywords: recommendations. We investigated the consequences of Euterpe edulis palm heart harvesting on the seed Seed rain rain in the Brazilian Atlantic rainforest. We compared the density of E. edulis individuals, as well as the Seed dispersal density of E. edulis seeds, and the density, richness and functional composition of seed rain of the whole Keystone species plant community, before and after palm heart harvesting in a 10 ha permanent plot. This assessment was Harvesting carried out in preserved (typical old-growth Atlantic rainforest) and in disturbed (more open habitat Forest succession dominated by the native bamboo Guadua tagoara) forest patches. Palm harvesting reduced the E. edulis population from 202.16 to 25.67 ind/ha and its seed rain density from 0.362 to 0.3 seeds/m2 and from 2.395 to 0.15 seeds/m2 in preserved and disturbed forest patches, respectively. Seed density of light- dependent climbers, pioneer trees, bamboo and animal-dispersed seeds increased after palm harvesting, especially in the disturbed forest patches, where palm harvesting was more intense and may have chan- ged the light regime of the understory. On the other hand, species richness of the plant community declined by half. We observed a remarkable decline in the number of animal-dispersed species, especially for those with large seeds, suggesting that the activity of seed dispersers, including many species attracted by E. edulis fruits, was reduced. Therefore, harvesting of E. edulis palm heart may change the regeneration dynamics of the Atlantic rainforest, both due to shifts in forest structure, mediated by the removal of individuals from the forest canopy, and in community functioning, mediated by the interfer- ence on the activity of seed dispersers. Ó 2013 Elsevier B.V. All rights reserved.
1. Introduction management on the regeneration dynamics of ecological commu- nities and ecosystems (Ticktin, 2004;Ticktin and Shackleton, What is the impact of non-timber forest product (hereafter 2011). The scarcity of studies on these issues for tropical forests NTFP) harvesting on species population dynamics? How does this across the world is an important limitation for their sustainable impact vary with the life history of plants harvested? Are current management, although much of the livelihood of traditional com- rates of exploitation sustainable? These are key questions ad- munities living in such ecosystems is based on NTFPs (Belcher dressed by Peres (2010) about NTFP exploitation. In addition to et al., 2005; Kar and Jacobson, 2012; Zenteno et al., 2013). In spite understanding impacts at the population level, another important of the growing knowledge about the ecological consequences of question without clear answers is related to the impacts of NTFP NTFP exploitation (Ticktin, 2004), overharvesting of NTFP is still a relevant concern for forest management and biological conserva- tion (Ticktin et al., 2002; Peres, 2010; Fernandez et al., 2012). ⇑ Corresponding author. Tel.: +55 19 34294431x233. Overharvesting occurs when the harvest rate of any given nat- E-mail addresses: [email protected] (A.E. Muler), [email protected] ural population exceeds its natural replacement rate (Peres, (D.C. Rother), [email protected] (P.S. Brancalion), [email protected] (R.P. Naves), [email protected] (R.R. Rodrigues), [email protected] (M.A. Pizo). 2010). This process has become one of the most important global
0378-1127/$ - see front matter Ó 2013 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.foreco.2013.09.001 118 A.E. Muler et al. / Forest Ecology and Management 324 (2014) 117–125 threats to the maintenance of biodiversity, both for terrestrial and investigated the extraction of leaves and fruits/seeds, so that aquatic ecosystems (Peres, 2010). Unfortunately, the increasing the ecological consequences of palm heart extraction are less global demand for land and food to support the 9 billion people known (Ticktin, 2004). The case study of the exploitation of the estimated to live on Earth by 2050 (Godfray et al., 2010;Smith endangered palm Euterpe edulis in Brazil, locally known as ‘‘pal- et al., 2010; Phalan et al., 2011) does not provide a favorable pros- miteiro’’, provides a valuable template for assessing the possible pect to reduce the pressure on commercially important native spe- consequences of palm heart overharvesting for the dynamics of cies in tropical rainforests. Especially in developing countries, tropical rainforests. This endemic palm species provides the most where most of the population growth, rural poverty, deforestation important NTFP exploited in the Brazilian Atlantic Forest hotspot, and remaining tropical rainforest cover are concentrated, the chal- the palm heart – the apical meristem and developing undifferen- lenge for sustainably exploiting native species within their natural tiated leaves of the palm’s stem (Reis et al., 2000). Given that the habitat is even greater (Kareiva et al., 2007). This holds true not apical meristem is extracted to obtain the palm heart and that only because of the socioeconomic challenge to reduce poverty palmiteiro is a single-stemmed palm that does not re-sprout after and promote economic development without environmental deg- harvesting, palm heart extraction leads to the death of palmiteiro radation, but also due to the inherent ecological complexity of individuals (Reis et al., 2000). According to Peres (2010), ‘‘exploi- tropical rainforests. Whereas many species are more resilient to tation of NTFPs often involves partial or entire removal of indi- exploitation and remain abundant under relatively intense har- viduals from the population, but the extraction method and vesting regimes, others can be driven to local extinction by even whether vital parts are removed usually determine the mortality the lightest exploitation levels (Peres, 2010). Furthermore, if the level in the exploited population’’. Based on this definition, we overexploited species plays an important role for ecosystem func- consider palm heart harvesting as NTFP exploitation, even though tioning, like keystone species for seed dispersers, the reduction in the entire plant is cut. As a result of the predatory exploitation of abundance or extinction of this species can negatively affect the palm heart in the last well-conserved remnants of Atlantic Rain- structure and composition of the ecological community through forest, including protected areas, this species is currently under cascading effects caused by dispersal limitation (Terborgh, 1986; extinction risk (Dransfiled et al., 1988; Galleti and Chivers, Scott Mills et al., 1993). In spite of these risks for biodiversity per- 1955; Galetti and Fernandez, 1998). A recent less-impacting op- sistence, community management of tropical forests has played a tion for the economic use of palmiteiro has been the exploitation relevant role for controlling deforestation in the tropics (Porter- of fruits for producing a southeastern equivalent of the Amazo- bolland et al., 2011). Thus, the exploitation of NTFPs should be bet- nian ‘‘açaí’’ (Euterpe oleracea), the pulp of palm fruits (Brancalion ter investigated in order to obtain the benefits derived from con- et al., 2012), which is currently used for many healthy, nutri- servation-through-use, without the risks of overharvesting and tional perspectives due to its high content of energy, minerals its potential consequences for ecosystem degradation (Fernandez and antioxidants. et al., 2012). Palmiteiro has two important characteristics that can amplify Although there are many case studies in the literature the ecological consequences of its overharvesting for palm heart. reporting overharvesting of timber species in tropical forests, Firstly, it is the most abundant species in the Atlantic rainforest, such as Brazilwood (Caesalpinia echinata – Dean, 1996), East representing nearly 20% of individuals with diameter at breast African blackwood (Dalbergia melanoxylon – Ball, 2004) and height (DBH) > 4.8 cm in conserved areas, with more than 200 indi- broad-leaved mahogany (Swietenia macrophylla –(Grogan viduals per hectare (Brancalion et al., 2012); consequently, the re- et al., 2008), little is known about the consequences of overhar- moval of adults due to palm heart harvesting may result in vesting NTFPs (see Peres et al., 2003 for an example). Tradi- important changes in forest structure. Secondly, palmiteiro plays tional NTFP extractive practices are often understood as a a keystone role for vertebrate frugivores (Fadini et al., 2009), since low-impact economic activity in tropical forests compared to it produces large amount of fruits every year and fruiting occurs in logging and shifting agriculture (Peters et al., 1989). Indeed, a season of food scarcity for frugivores in the Atlantic rainforest NTFP harvesting does not generally cause such evident modifi- (Galetti and Aleixo, 1998; Castro et al., 2007). Consequently, the cations in forest structure like timber exploitation, and thus density reduction of palmiteiro adults may be harmful to the has not been perceived as a degrading activity. For instance, regeneration dynamics of the Atlantic rainforest, which often has the sustainability of exploitation of natural resources in Brazil more than 80% of tree species dispersed by vertebrate frugivores was higher for NTFPs compared to timber and hunting, although (Almeida-Neto et al., 2008), as a result of the depletion of seed rain. two of the nine case studies of NTFP exploitation were consid- Although the negative consequences of palm harvesting may take ered unsustainable (Fernandez et al., 2012). However, the few some years to be expressed in the plant community regeneration studies available on NTFP overharvesting have focused on the as a result of the time required for the establishment and growth impacts on the species itself (see examples in (Mitja and Les- of woody species, the assessment of seed rain may provide a faster cure, 2000; Peres and Lake, 2003; Fernandez et al., 2012). From overview of the outcomes of palm harvesting for the regeneration the 68 studies reviewed by Ticktin (2004) in which NTFP har- dynamics of the Atlantic rainforest. Based on this case study, we vesting was quantitatively assessed, 58 (82.8%) investigated sought to investigate the ecological consequences of palmiteiro the ecological implications of NTFP exploitation at the level of palm heart harvesting in seed rain dynamics by studying the den- individuals and/or populations, while only 9 (12.8%) and 3 sity of E. edulis seeds, and the density, richness and functional com- (4.3%) focused on the level of communities and ecosystems, position of seed rain of the whole plant community in a Brazilian respectively. Thus, it is necessary to increase our knowledge Atlantic rainforest. about the consequences of NTFP exploitation for community functioning in order to provide more reliable systems of tropical forest sustainable management. Palms, in particular, have shown a highlighted importance as a 2. Materials and methods source of NTFPs in tropical rainforests. In the same review men- tioned above, nearly 41% of the studies focused on the impacts of 2.1. Study site NTFP harvesting on palm species, followed by tree (26%), herb (26%), vines and lianas (3%), and bryophytes species (3%) (Ticktin, This work was carried out in a permanent plot of 10.24 ha, im- 2004). However, most of these works on palm species have planted in 2003 and located in the Carlos Botelho State Park A.E. Muler et al. / Forest Ecology and Management 324 (2014) 117–125 119
(37.793 ha 24°000 24°150S and 47°450 48°100W), São Paulo (Veblen, 1982; Rother et al., 2009, 2013; Lima et al., 2012), and state, southeastern Brazil. the regeneration of palmiteiro is compromised (Fantini and Guries, The permanent plot was composed of a grid of 256 subplots 2007). (20 m 20 m), in which all individuals with DBH >4.8 cm were sampled. Palmiteiro was the most abundant species in the implementation of the permanent plot (240 individuals per hect- are – 21.5% of all individuals sampled over this diameter class), 2.2. Data collection where 204 other species were sampled. Overall, palmiteiro pro- duces 3000 fruits and 4.5 kg of fruits per individual in the stud- We found a conserved population of palmiteiro when the perma- ied site (Leite et al., 2012). Carlos Botelho State Park is covered nent plot was implemented in 2003, with large individuals and a by different physiognomies of Atlantic rainforest (dense ombro- high abundance of plants. When Rother et al. (2009) evaluated seed phylous forest) (Veloso and Góes-Filho, 1982), which are deter- rain in this permanent plot between 2004 and 2005, the population mined by the altitudinal gradient established between 30 m of palmiteiro was still conserved, and no signs of illegal harvesting and 1003 m above sea level (Domingues and Silva, 1988). The lo- were observed by that time. However, after 2005, most large individ- cal climate is classified as humid subtropical without dry season uals of palmiteiro of the permanent plot were illegally cut for palm (CFA – Köppen system), with an average annual temperature of heart extraction. Illegal harvesting of palmiteiro is unfortunately a 22 °C and rainfall of 1584 mm, concentrated in summer. A less common problem in protected areas in the Atlantic Forest (Galetti rainy season is observed from April to September, when rainfall and Fernandez, 1998; Reis et al., 2000), which was intensified by is below 50 mm and palmiteiro seeds are dispersed (Castro et al., the population reduction of the species in most private landholdings 2007). This protected area has the highest levels of species rich- due to harvesting. Based on this historical context, this study was ness of São Paulo state and has more than 60 endangered spe- conducted in two periods: the first from 2004 to 2005 (prior to pal- cies, mainly in the Myrtaceae, Lauraceae, and Gesneriaceae miteiro harvesting – (Rother et al., 2009) and the other in 2008 (after families (Lima et al., 2011). palmiteiro harvesting). This study was carried out in two forest patch types which are Seed rain was first evaluated from June 2004 to May 2005 with scattered in the above described permanent plot: one considered 1 1 m seed traps made of wood and nylon screen and positioned preserved and the other considered disturbed, characterized by 10 cm above the ground (Rother et al., 2009). Eighty seed traps the dominance of the bamboo Guadua tagoara in forest patches. were established in the permanent plot (40 in each forest patch) We studied these two forest patch types because palmiteiro har- and were allocated with a minimum distance of 20 m away from vesting creates gaps and has favored bamboo proliferation in the each other. In the second period of evaluation, seed rain was eval- forest (Fantini and Guries, 2007). Thereby, we were interested in uated from January to December 2008 in the same permanent plot, understanding if palmiteiro palm harvesting in more degraded but in different zones. In 2008, most of the bamboo patches evalu- sites, where this species may share a higher proportion of canopy ated by Rother et al. (2009) were in a post-reproductive phase, cover due to the lower density of trees, would have stronger con- with adults dead and many young seedlings regenerating on the sequences in the forest regeneration dynamics. The ‘‘preserved’’ forest floor. As a result, our seed traps were established in other forest patches were covered by typical old-growth Atlantic rainfor- sites within the permanent plot where vegetative stands of bam- est, with mean canopy height of 20 m and a highly stratified verti- boo were found. A total of 61 50 50 cm seed traps, made with cal structure, with tree branches densely covered with epiphytes the same materials and installed at the same height as the seed and the understory full of ferns, palms, shrubs and plenty of spe- traps of the previous period, were set in the permanent plot (31 cies of Araceae and Maranthaceae (Lima et al., 2011). The ‘‘dis- in disturbed forest patches and 30 in preserved forest patches). turbed’’ sites were characterized by having dense patches of the In both periods (pre- and post-harvesting), the content of each seed endemic bamboo Guadua tagoara, which occupied 3 ha of the trap was evaluated every month, with the identification and count- 10.24 ha permanent plot by 2005 (Rother et al., 2009). Each bam- ing of seeds. The identified species were classified according to boo stem is 10–15 m tall, 5–10 cm in diameter and has many their seed dispersal syndromes (abiotic-dispersed: mediated by branches, with thorns, deciduous leaves and broad-triangular wind, gravity and explosion; and animal-dispersed: mediated by sheaths, which create a dense layer of litter over the forest floor animals – according to Pijl van der (1982), successional group (pio- (Londoño and Clarck, 2002). Frutification is massive and occurs neer and non-pioneer - according to (Swaine and Whitmore, 1988) every two years, with large production of small seeds (Veldman and seed size (Large seeds: seeds with a larger diameter or/and and Putz, 2011). This bamboo species also propagates vegetatively, length when compared to palmiteiro seed size; and Small seeds: which allows a fast colonization of forest gaps (Burman and Filgue- seeds with a lower diameter or/and length when compared to pal- iras, 1993). Once the forest is colonized by this species, the struc- miteiro seed size. We considered the value of 1.1 cm as the mean ture and dynamics of the forest patches are highly modified diameter and 1.2 cm as the mean length of palmiteiro globular seeds (Galetti et al., 2011).
Fig. 1. Comparison of Euterpe edulis individuals and seed density before (2003 and 2004) and after (2008) it is illegal cut for palm heart extraction in preserved and disturbed forest patches at the Atlantic rainforest of the Carlos Botelho State Park, São Paulo state, southeastern Brazil. Bars followed by different letters, for each patch type, differ by Tukey test (P < 0.05); vertical lines in each bar represent the standard error. 120 A.E. Muler et al. / Forest Ecology and Management 324 (2014) 117–125
The abundance of palmiteiro individuals was assessed within 2.3. Data analysis circular plots (radius of 10 m) established around each seed collec- tor described above in the two periods of evaluation, were all pal- The density of E. edulis adults (DBH P 4.8 cm) was submitted to miteiro individuals with DBH P 4.8 cm were sampled. ANOVA, followed by Tukey tests for significant results, to compare the density among the periods: implementation of the permanent plot (2003); pre-harvesting (2004–2005); and post-harvesting
Fig. 2. Dominance-diversity curve of community seed species in preserved (A) and disturbed (B) forest patches in the pre- and post-harvesting periods. Species code: Alchornea glandulosa (Alc gla), Anthurium scandens (Ant sca), Bauhinia sp. (Bau. sp.), Cabralea canjerana (Cab can), Cecropia glaziovii (Cec gla), Cissampelos andromorpha (Cis and), Costus spiralis (Cos spi), Coussapoua microcarpa (Cou mic), Cryptocarya moschata (Cry. mos), Euterpe edulis (Eut edu), Geonoma sp. (Geo sp.), Guadua tagoara (Gua tag), Hyeronima alchorneoides (Hye alc), Maytenus sp. (May sp), Monstera adansonii (Mon ada), Philodendron sp. (Phi sp.), Phytolacca dioica (Phy dio), Psychotria pubigera (Psy pub), Rollinea sericea (Rol ser), Rudgea jasminoides (Rud jas), Sloanea guianensis (Slo gui), Solanum pseudoquina (Sol pse), Chrysophyllum viride (Chr vir), Tetrastylidium grandifolium (Tet gra), Virola bicuhyba (Vir bic), sp21, sp71, sp117, sp119, sp131,sp140, sp153, sp155, sp34, sp36, sp53, sp60, sp70, sp71, sp76, sp78, sp79, sp80, sp87, sp90. A.E. Muler et al. / Forest Ecology and Management 324 (2014) 117–125 121
(2008), considering the two forest patch types (Preserved and posed by few abundant species and a large proportion of rare spe- Disturbed). cies (Fig. 2). However, the dominance profile of seed rain was An ANOVA with repeated measures for each forest patch, fol- markedly different in the post-harvesting period, especially in dis- lowed by Tukey tests for significant results, was used to compare turbed forest patches (Fig. 2). In these forest patches, the main the density of seed rain between pre- and post-harvesting periods. changes observed were: (1) the dominance increase of C. glaziovii; In this model, the period was considered as treatment and mea- (2) the massive seed dispersal of the wind-dispersed, pioneer clim- surements over the months were considered repeated measures. ber Mikania sp., which was not sampled in the pre-harvesting per- We have also used this statistical method to compare, between iod but was the most abundant species after palm harvesting pre- and post-harvesting periods, the seed density (seeds/m2)of (average of 1377 seeds/m2 – this species was not included in the (1) E. edulis, (2) plant community, (3) successional groups (pioneer dominance-diversity curve shown in Fig. 3 because it was consid- and non-pioneer species), (4) seed dispersal syndrome (animal- ered an outlier); (3) the increased dispersal of the bamboo G. tag- dispersed and abiotic-dispersed species, and (5) seed size of ani- oara, which did not appeared as a dominant species in the pre- mal-dispersed species (smaller and larger seeds). The specie Mika- harvesting period, but was one of the most abundant species nia sp. was excluded from the analysis because it was considered post-harvesting; and (4) the dominance reduction of palmiteiro an outlier e. All analyses were carried out with software R (R Devel- in the seed rain, dropping from the 2nd to the 9th position in the opment Core Team, 2011). rank. Although less affected by palmiteiro harvesting, the seed rain We compared the total species richness and species richness dominance profile of preserved forest patches showed two impor- for each functional group between pre- and post-harvesting peri- tant differences (Fig. 2): (1) the climber Mikania sp. appeared as the ods, for each forest patch, using rarefaction. Matrices were pre- third most abundant species (average of 94 seeds/m2) and (2) pal- pared containing data on species abundance, for both periods miteiro dropped from the 6th most abundant species in the seed and in each forest patch, based on the content sampled in each rain to the 13th position after palm harvesting. seed trap. Given that the number of seeds sampled differed In the disturbed forest patches, seed density was seven-fold among forest patch types and study periods, we considered the higher in the post-harvesting period (F = 6.54; df =2; p = 0.010), smallest sample as the baseline for rarefaction. We used the first- whereas species richness obtained by rarefaction declined from order Jackknife estimator (Burnham and Overton, 1979) to pro- 153 to 64 species (Fig. 3). In contrast, in preserved forest vide the number of species expected. A total of 100 randomiza- patches the density of seeds did not differ between study peri- tions were performed for the order of samples (Gotelli and ods (F = 3.53; df =2; p = 0.060), but species richness obtained by Colwell, 2001) based on the average of the accumulated richness rarefaction was slightly higher in the post-harvesting period associated with a standard deviation with 95% confidence (Mor- (Fig. 3). aes et al., 2007). We used the program Estimate S for this analysis The main changes in the functional profile of seed rain are (Colwell, 2005). shown in Fig. 4. A higher seed density in the disturbed sites after palm harvesting was mainly driven by the increase of pioneer 3. Results (F = 6.66; df =2; p = 0.001) and animal-dispersed (F = 6.48; df =2; p = 0.011) seeds, although species richness was reduced between The density of palmiteiro individuals differed between the stud- periods, especially for animal-dispersed species (Fig. 4). Although ied periods (implementation of the permanent plot, pre- and post- seed density of abiotic-dispersed species was also higher after har- harvesting), both in the preserved (F = 20.43; df =2;p = 0.001) and vesting (F = 7.99; df =2;p = 0.004), its contribution to the seed rain in the disturbed forest patches (F = 24.4; df =2;p = 0.001). As ex- was too reduced compared to animal-dispersed species to have pected, palm heart harvesting reduced palmiteiro density in both any effect on the overall seed density of the plant community forest patch types, but this reduction was more intense in dis- (Fig. 4). On the other hand, the seed density of non-pioneer species turbed forest patches (Fig. 1). Although palmiteiro seed density did not differ between periods in this forest patch (F = 1.99; df =2; did not differ between study periods in preserved forest patches p = 0.16) and no relevant differences in species richness were ob- (F = 0.178; df =2;p = 0.67), it was 16-fold lower in disturbed forest served (Fig. 4). In contrast, in preserved forest patches seed density patches after harvesting (F = 10.87; df =2;p = 0.001) (Fig. 1). of pioneer species did not differ between periods (F = 2.39; df =2; The profile of seed rain was markedly different in preserved and p = 0.12), but non-pioneer species were more abundant (F = 2.39; disturbed forest patches and between pre- and post-harvesting df =2;p = 0.12) and represented by fewer species (Fig. 4). Similarly, (Supplementary material S1). In both forest patch types and peri- animal-dispersed species were also more abundant (F = 4.13; ods, the animal-dispersed, pioneer species Cecropia glaziovii and df =2;p = 0.04) and represented by fewer species (Fig. 4). Hieronyma alchorneoides dominated seed rain, which was com-
Fig. 3. Comparison of seed density and richness, of plant community before (2003) and after (2008) illegal cut of Euterpe edulis individuals for palm heart extraction in preserved and disturbed forest patches at the Atlantic rainforest of the Carlos Botelho State Park, São Paulo state, southeastern Brazil. Species richness was compared between periods using rarefaction. Bars followed by different letters, for seed density evaluation in the same patch type, differ by Tukey test (P < 0.05); vertical lines in each bar represent the standard error. 122 A.E. Muler et al. / Forest Ecology and Management 324 (2014) 117–125
Fig. 4. Comparison of seed density of successional and seed dispersal functional groups before (2003) and after (2008) illegal cut of Euterpe edulis individuals for palm heart extraction in preserved and disturbed forest patches found at the Atlantic rainforest of the Carlos Botelho State Park, São Paulo state, southeastern Brazil. Species richness was compared between periods using rarefaction. Bars followed by different letters, for seed density evaluation in the same patch type, differ by Tukey test (P < 0.05); vertical lines in each bar represent the standard error.
Fig. 5. Comparison of seed density of animal-dispersed, large-seeded species before (2003) and after (2008) illegal cut of Euterpe edulis individuals for palm heart extraction in preserved and disturbed forest patches found at the Atlantic rainforest of the Carlos Botelho State Park, São Paulo state, southeastern Brazil. Seed size was evaluated only for animal-dispersed species; we considered as large-seeded species those with seed diameter equal or higher than E. edulis seeds. Species richness was compared between periods using rarefaction. Bars followed by different letters, for seed density evaluation in the same patch type, differ by Tukey test (P < 0.05); vertical lines in each bar represent the standard error. A.E. Muler et al. / Forest Ecology and Management 324 (2014) 117–125 123
Contrary to our expectations, seed density of animal-dispersed, Although the proliferation of pioneers in disturbed forests is a large-seeded species did not differ between periods in both pre- natural and expected response of ecological succession to changes served (F = 3.72; df =2; p = 0.05) and disturbed (F = 1.40; df =2; in forest structure, and apparently does not represent a threat to p = 0.24) forest patches. However, we observed a remarkable de- forest conservation, Atlantic Forest remnants are particularly vul- cline in the number of animal-dispersed species, especially for nerable to such successional setbacks resulting from human-med- those with large seeds in disturbed forest patches after harvesting iated disturbances (Hill and Curran, 2003; Tabarelli et al., 2004), (Fig. 5). such as illegal cutting of palmiteiro. The intense fragmentation and reduced forest cover of the Atlantic Forest (Ribeiro et al., 2009), combined with historical degradation of forest remnants, 4. Discussion has intensified edge effects and reduced the recruitment of tree species with higher dispersal limitations and sensitivity to habitat Palm heart harvesting resulted in marked shifts in the regener- quality (Tabarelli et al., 2012). The exclusion of a large set of late ation dynamics of the Atlantic rainforest, especially in the dis- successional species, along with the proliferation of a small set of turbed forest patches, as shown by the changes in seed rain generalist, r-strategist species has resulted in a massive process density, richness and composition of functional groups. The num- of biotic homogenization of Atlantic Forest remnants (Lôbo et al., ber of gaps may have favored the higher seed production and pro- 2011). Thus, palmiteiro illegal harvesting may be one more factor liferation of pioneer species such as Mikania sp., the tree C. glaziovii contributing to the loss of biological diversity in the biome. The and the bamboo G. tagoara in disturbed sites, and the overall dom- proliferation of pioneer climbers, like Mikania sp., may bring addi- inance of animal-dispersed, pioneer species in the seed rain sug- tional limitations to the reestablishment of biodiversity through gest that palm harvesting has driven the disturbed forest patches the competitive exclusion of tree and shrub species (Laurance, to an earlier successional stage. However, preserved sites seem to 2002; Schnitzer and Bongers, 2011). Mikania sp. in particular is be less affected by palm harvesting, as indicated by few changes well recognized as an aggressive pioneer climber with high com- in the seed rain profile. The arrival of palmiteiro seeds was also af- peting potential (Zhang et al., 2004; Lian et al., 2006). The massive fected after harvesting, with a 15-fold reduction in the number of colonization of forest canopy by this species may impose an ar- seeds deposited per area compared to the pre-harvesting period in rested succession on forest patches submitted to palmiteiro har- the disturbed forest patches, although palmiteiro deposition in vesting, as observed in other forest remnants dominated by preserved forest patches remained similar. climbers (Schnitzer and Bongers, 2011). Concomitantly, the succes- The extraction of palm stems may result in different conse- sional setback of areas submitted to palmiteiro harvesting may fa- quences for the regeneration dynamics of the studied rainforest vor the colonization of forest gaps by the shade-intolerant bamboo and palmiteiro population. One of them is the change in forest G. tagoara (Rother et al., 2009). Other studies carried out at this structure mediated by disturbances in the canopy, which can be same permanent plot have demonstrated that habitats dominated basically determined by the intensity of harvesting and the integ- by this bamboo species are indeed subjected to drastic shifts in for- rity of forest structure previous to it. The harvesting of a lower est regeneration dynamics (Rother et al., 2009; Lima et al., 2012). number of individuals would evidently reduce the impact on can- One such shift is the reduction in the arrival of seeds of certain spe- opy cover, since palmiteiro has a very high natural abundance in cies (Rother et al., 2009). Thus, palmiteiro harvesting may indi- the Atlantic Forest. Given that palm heart extraction was more in- rectly change forest regeneration dynamics by stimulating tense in disturbed forest patches, where 90% of the individuals bamboo regeneration and the proliferation of Mikania. with DBH > 4.8 cm were cut, it is expected that this forest patch The second consequence of palmiteiro harvesting for forest type was more affected by harvesting. Regarding the importance regeneration dynamics is a change in the functional profile of seed of the forest structure integrity, we believe that palmiteiro harvest- dispersers that visited the sites. Palmiteiro is known for producing ing in more preserved forest patches may bring lower impacts in many fruits that are heavily consumed by large assemblages of canopy cover than in disturbed sites. Since in old-growth Atlantic medium- to large-bodied vertebrate seed dispersers (Galetti and rainforests large trees dominate the canopy and palmiteiro occur- Fernandez, 1998), which may be attracted or not to a given patch rence is restricted to the understory (Fantini and Guries, 2007), its of forest depending on the availability of palmiteiro fruits. For in- extraction would not create large openings in forest canopy, stance, Castro et al. (2007; 2012) described the migration of avian although small gaps may be produced. Thus, the massive harvest- seed dispersers in an altitudinal gradient of Atlantic rainforest to ing in disturbed sites may create large gaps in the forest, thus stim- accompany the fruiting phenology of palmiteiro, which varies ulating the proliferation of pioneer species of different life forms, according to the altitude. In this context, we initially hypothesized such as tress, climbers and bamboos (Swaine and Whitmore, that the massive harvesting of palmiteiro in forest patches would 1988). The combination of these two factors – higher intensity of reduce seed rain density and richness of animal-dispersed, large- palm harvesting and lower initial canopy cover – probably deter- seeded species attracted by palmiteiro fruits. Contrary to our mined the different patterns of seed rain in disturbed and pre- expectations, seed density of this group of species did not differ be- served forest patches. tween periods in both forest patch types. However, the number of In disturbed forest patches, the potential reduction of canopy animal-dispersed, large-seeded species collected was three times cover resulting from palmiteiro harvesting may have stimulated lower after harvesting in the disturbed forest patches, where palm the seed bank of pioneer species, such as Mikania sp. and C. glazi- harvesting was more intense. Rother et al. (2013) evaluated the ovii, which germinated, developed and reached maturity, thus pro- assemblages of birds in preserved and disturbed forest patches ducing the large amounts of seeds (Garwood, 1989; Tabarelli et al., after harvesting of palmiteiro and did not find relevant differences 2004; Oliveira-filho et al., 2004) that were sampled after harvest- in the richness and abundance of frugivorous birds between forest ing. The five-year period (2003 to 2008) between seed rain assess- patch types (18 species and 136 individuals in disturbed plots and ments would be certainly enough for a complete reproductive 15 species and 167 individuals in preserved plots). Thus, the differ- cycle of such pioneer species. The functional profile of the seed rain ences in forest structure may not have affected the assemblage of after harvesting, in which pioneer, animal-dispersed species dom- seed dispersers between forest patch types and study periods. On inated seed rain in the disturbed forest patches, reinforces the the other hand, the reduction of 90% of palmiteiro density in dis- hypothesis that early-successional species were favored after the turbed plots may have had strong consequences for the attraction harvesting of palmiteiro. 124 A.E. Muler et al. / Forest Ecology and Management 324 (2014) 117–125 of large-bodied vertebrate frugivores that disperse large-seeded Burman, A.G., Filgueiras, T., 1993. A review of the woody bamboo genera of Brazil species. (Gramineae: Bambusoideae: Bambuseae). Thaiszia 3, 53–88. Burnham, K., Overton, W.S., 1979. Robust Estimation of Population Size When The role played by palmiteiro for avian frugivorous in the Atlan- Capture Probabilities Vary Among Animals Author(s): Burnham, K.P, Overton, tic rainforest is the theme of a controversial and not yet defined is- W.S. Ecological Society of America Stable URL: Vary among animalS1. Ecology sue. Since palmiteiro fruits are consumed by at least 30 bird and 13 letters vol. 60, pp. 927–936
Mitja, D., Lescure, J.P., 2000. Madeira para perfume: qual será o destino do pau- Schnitzer, S.A., Bongers, F., 2011. Increasing liana abundance and biomass in rosa? Floresta em Jogo: o Extrativismo na Amazônia Central. Editora UNESP, tropical forests: emerging patterns and putative mechanisms. Ecology Letters Imprensa Oficial do Estado. 14, 397–406. Moraes, R.A., Sawaya, R.J., Barrela, W., 2007. Composição e diversidade de anfíbios Scott Mills, L., Soulé, M.E., Doak, F.D., 1993. The keystone-species concept in ecology anuros em dois ambientes de Mata Atlântica no Parque Estadual Carlos Botelho, and conservation. Bioscience 43, 219–224. São Paulo, sudeste do Brasil. Biota Neotropica, Campinas 7, 27–36. Smith, P., Gregoy, P.J., Van vuuren, D., Obersteiner, M., Havlínk, P., Rousevell, M., Oliveira-filho, A.T., Carvalho, D.A., Vilela, E.A., Curi, N., Fontes, M.A.L., 2004. Diversity Woods, J., Stehfest, E., Bellarby, J., 2010. Competition for land. Philosophical and structure of the tree community of a fragment of tropical secondary forest Transactions of the Royal Society B. Biological Sciences 365, 2941–2957. of the Brazilian Atlantic Forest domain 15 and 40 years after logging. Revista Swaine, M.D., Whitmore, T.C., 1988. On the definition of ecological species groups in Brasileira de Botânica 27, 685–701. tropical rain forests. Vegetatio 75, 81–86. Peres, C.A, Baider, C., Zuidema, P.A, Wadt, L.H.O., Kainer, K.A, Gomes-Silva, D.A.P., Tabarelli, M., Peres, C.A., Melo, F.P.L., 2012. The ‘‘few winners and many losers’’ Salomão, R.P., Simões, L.L., Franciosi, E.R.N., Cornejo Valverde, F., Gribel, R., paradigm revisited: Emerging prospects for tropical forest biodiversity. Shepard, G.H., Kanashiro, M., Coventry, P., Yu, D.W., Watkinson, A.R., Freckleton, Biological Conservation 155, 136–140. R.P., 2003. Demographic threats to the sustainability of Brazil nut exploitation, Tabarelli, M., Silva, J.M.C., Gascon, C., 2004. Forest fragmentation, synergisms and 302. Science, (New York, N.Y.), pp. 2112-4. the impoverishment of neotropical forests. Biodiversity and Conservation 13, Peres, C.A., Lake, I.R., 2003. Extent of nontimber resource extraction in tropical 1419–1425. forests: accessibility to game vertebrates by hunters in the amazon basin. Team, R.D.C., 2011. A language and environment for statistical computing. R Conservation Biology 17, 521–535. Foundation for Statistical Computing [WWW Document]. URL http://www.R- Peres, C.A., 2010. Overexploitation. Oxford University Press. project.org/. (accessed 3.15.2013). Peters, C.M., Gentry, A.H., Mendelsohn, R., 1989. Valuation of an Amazonian Terborgh, J., 1986. Community aspects of frugivory in tropical forests. Princeton rainforest. Nature 339, 655–656. University. Phalan, B., Balmford, A., Green, R.E., Scharlemann, J.P.W., 2011. Minimising the harm Ticktin, T., 2004. The ecological implications of harvesting non-timber forest to biodiversity of producing more food globally. Food Policy 36, S62–S71. products. Journal of Applied Ecology 41, 11–21. Pijl van der, L., 1982. Principles of dispersal in higher plants. Springer-Velag, New Ticktin, T., Nantel, P., Ramirez, F., Johns, T., 2002. Effects of variation on harvest York, Berlin Heidelberg. limits for nontimber forest species in Mexico. Conservation Biology 16, 691– Porter-bolland, L., Ellis, E.A., Guariguata, M.R., Ruiz-mallén, I., Negrete- 705. yankelevich, S., Reyes-garcía, V., 2011. Forest ecology and management Ticktin, T., Shackleton, C., 2011. Non-Timber Forest Products in the Global Context. community managed forests and forest protected areas : an assessment of Verlag, Berlin, pp. 149–169. their conservation effectiveness across the tropics. Forest Ecology and Veblen, T.T., 1982. Growth patterns of Chusquea bamboos in the understory of Management 268, 6–17. Chilean Nothofagus forests and their influences in forest dynamics. Bulletin of Reis, M.S., Fantini, A.C., Nodari, R.O., Reis, A., Guerra, M.P., Mantovani, A., 2000. the Torrey Botanical Club 109, 474–487. Management and conservation of natural populations in atlantic rain Veldman, J.W., Putz, F.E., 2011. Grass-dominated vegetation, not species-diverse forest: the case study of palm heart (Euterpe edulis Martius). Biotropica natural savanna, replaces degraded tropical forests on the southern edge of the 32, 894. Amazon Basin. Biological Conservation 144, 1419–1429. Ribeiro, M.C., Metzger, J.P., Martensen, A.C., Ponzoni, F.J., Hirota, M.M., 2009. The Veloso, H.P., Góes-Filho, L., 1982. Fitogeografia brasileira: classificação fisionômico Brazilian Atlantic Forest: How much is left, and how is the remaining forest ecológica da vegetação Neotropical. In: Boletim Técnico do Projeto distributed? Implications for conservation. Biological Conservation 142, 1141– RADAMBRASIL. 1153. Zenteno, M., Zuidema, P.A., de Jong, W., Boot, R.G.A., 2013. Livelihood strategies and Rother, D.C., Jordano, P., Rodrigues, R.R., Pizo, M.A., 2013. Demographic forest dependence: New insights from Bolivian forest communities. Forest bottlenecks in tropical plant regeneration: A comparative analysis of causal Policy and Economics 26, 12–21. influences. Perspectives in Plant Ecology, Evolution and Systematics 15, 86– Zhang, L.Y., Ye, W.H., Cao, H.L., Feng, H.L., 2004. Mikania micrantha HBK in China – 96. an overview, 42–49. Rother, D.C., Rodrigues, R.R., Pizo, M.A., 2009. Effects of bamboo stands on seed rain and seed limitation in a rainforest. Forest Ecology and Management 257, 885– 892.