Agroforest Syst (2010) 80:303–313 DOI 10.1007/s10457-010-9321-z

Growth of Euterpe edulis Mart. (Arecaceae) under forest and agroforestry in southern Brazil Rodrigo Favreto • Ricardo Silva Pereira Mello Luı´s Rios de Moura Baptista

•

Received: 30 October 2009 / Accepted: 1 June 2010 / Published online: 20 June 2010 Ó Springer Science+Business Media B.V. 2010

Abstract The palm Euterpe edulis has high ecological and economic importance in Brazil. Currently, this species is being cultivated and managed for spontaneous regeneration in banana plantations. However, there are no data comparing its plantation growth performance to its native forest growth. We evaluated growth and mortality (M) of individuals of E. edulis planted in secondary dense ombrophilous forest and in banana plantations, as well as their relationships with site variables (canopy opening, soil nutrient availability, density of existing E. edulis, and herbivory).

Electronic supplementary material The online version of this article (doi:10.1007/s10457-010-9321-z) contains supplementary material, which is available to authorized users. R. Favreto (&) State Foundation for Agricultural Research – FEPAGRO, RS484 km5, CEP 95530-000 Maquine´, RS, Brazil e-mail: [email protected]

Twelve banana plantation sites and 12 sites in secondary dense ombrophilous forests were selected. At each site, 25 young individuals of E. edulis were planted in 2003. Annually until 2008, morphometric, herbivory, and M of the individuals were evaluated. In 2008, canopy and soil variables were measured at each plot. E. edulis growth was five times higher in banana plots compared to forest plots; current annual increment on height reached 38.9 cm in banana plots, compared to 7.3 cm in forest plots. M was relatively low and similar at both sites, presenting an intraspecific density-dependence pattern. Significant correlations were found between morphometric variables, M, and herbivory of E. edulis and canopy and soil variables. Euterpe edulis presented plasticity that allows for its establishment in banana plantations, indicating high potential for management in agroforestry consortia. Such management may be a useful conservation strategy for this and other shade-tolerant species.

R. S. P. Mello Department of Ecology, Federal University of Rio Grande do Sul, Av. Bento Gonc¸alves, 9500, CEP 91540-000 Porto Alegre, RS, Brazil

Keywords Domestication  Density-dependence  Gap effect  Juc¸ara palm

R. S. P. Mello Ac¸a˜o Nascente Maquine´-ANAMA, Rua do Come´rcio 507, CEP 95532-000 Maquine´, RS, Brazil

Introduction

L. R. de Moura Baptista Post-Graduate Program in Botany, Federal University of Rio Grande do Sul – UFRGS, Av. Bento Gonc¸alves 9500, CEP 91501-970 Porto Alegre, RS, Brazil

The ‘‘Juc¸ara’’ palm or ‘‘palmiteiro’’—Euterpe edulis Mart. (Arecaceae)—has great ecological importance in the food chain of the Brazilian Atlantic Rainforest, producing large quantities of flowers and fruits (Reis

123

304

and Kageyama 2000). It is characterized by an ample geographical distribution and abundance in the subcanopy of the dense Brazilian Atlantic Coastal Rainforest, as well as in tropical forests of the Parana´ River Basin (Henderson 2000). The species has significant economic and social importance due to the extraction industry of heart of palm (also called ‘‘palmito’’ in Brazil) and fruit harvesting, not to mention other products derived from its leaves and stem. Unfortunately, natural regeneration and conservation of E. edulis are seriously impaired by deforestation and by intense, illegal extraction of heart of palm, causing habitat reduction and population fragmentation which have depleted the remaining native populations. Cultivation and natural regeneration by adopting intercropping or agroforestry systems (Bovi et al. 1987) have been identified as alternatives to alleviate this depletion, making agricultural areas ‘‘habitats’’ for conservation of E. edulis and other species. Farmers are now planting and allowing for the natural establishment of Juc¸ara palm in banana plantations (Vivan 2002). Banana plantations, with an area exceeding 500,000 hectares in Brazil (IBGE 2006) and largely distributed in areas previously covered by Atlantic Rainforest, could play an important role in the conservation of E. edulis, if their potential for establishment of this species is confirmed. Many banana plantations are situated near or adjacent to forests, thus promoting mutual environmental interaction that includes seed dispersal of forest species such as E. edulis. Since it is a rainforest sub-canopy species, E. edulis shows high rates of seedling mortality (M) when exposed to direct sunlight (Ruschel et al. 1997), thus requiring shade during its initial phase of development (Conte et al. 2000). Forest conditions generally allow for incidence of up to 100 lmol m-2 s-1 (0.5–4% of total incidence) of photosynthetically active radiation at soil surface (Chazdon and Fetcher 1984). Growth of E. edulis increases with exposure to photosynthetic active radiation up to 360 lmol m-2 s-1 (about 20% of total); from 360 to 1260 lmol m-2 s-1 (20–70% of total, respectively) growth is equivalent; and over 1260 lmol m-2 s-1, seedling development can be impaired (Nakazono et al. 2001). Thus, although E. edulis needs shade in the early stages, its growth can be limited by excessive shade (Paulilo 2000).

123

Agroforest Syst (2010) 80:303–313

Low photosynthetically active radiation inside forests is an important factor that limits the establishment, growth, and regeneration of plant species (Denslow 1987). Moreover, forest dynamics create spatial heterogeneity of environments, from gaps to dense-stratified foliage layers, producing various micro-habitats and limiting the occurrence of many species (Givnish 1999), including palms (Kahn and Castro 1985). Many species whose seedlings survive under forest canopies benefit from, and even depend on, canopy openings for regeneration (Augspurger 1984; Brokaw 1985). According to Lee et al. (1996), light is the most important physical factor in controlling development of tree seedlings in tropical forests. Growth acceleration and increase of survival occur when a canopy gap is formed, a behavior commonly known as the gap effect (Denslow 1987). Euterpe edulis has recalcitrant seeds that germinate and form seedling banks within forests (Reis and Kageyama 2000; Reis et al. 2000). Recruitment of seedlings to the reproductive phase appears to be associated with conditions of more light in gaps or along riverbanks (Sanchez et al. 1999). Soil fertility is undoubtedly an important factor influencing plant growth (Malavolta 2006). Young individuals of E. edulis respond to different nutrient levels in greenhouses (Venturi and Paulilo 1998), depending on luminosity conditions (Illenseer and Paulilo 2002). However, there are no existing studies on this topic for E. edulis in field conditions. Besides light and nutrient levels, temperature, humidity, and other factors associated with the gap effect can influence the growth of E. edulis. Domesticated landscapes where natural regeneration of E. edulis has been allowed may present certain edafoclimatic characteristics similar to forests in early stages of succession or to gaps, with edge effects such as increased light incidence and temperature variation, among other biotic and abiotic factors. Consequently, the behavior of forest species that establish themselves in managed areas tends to be similar to behavior of individuals in gaps or forest edges. Depending on the species, such behavior leads to success or failure in establishment and reproduction. Although studies show that land use is an important factor in environmental change, including interference in plant phenology (Laurance et al. 2003), research is scarce on evaluating the effect of human action on possible changes in these processes,

Agroforest Syst (2010) 80:303–313

since species respond differently to such action. Moreover, the majority of previous studies of E. edulis were conducted under controlled greenhouse conditions; therefore, field studies are necessary to obtain a better understanding of its reactions to diverse natural growth factors (Paulilo 2000). Thus, this study is a comparative study of E. edulis growth in banana plantations and secondary dense ombrophilous forest, with two primary goals: (1) to assess the growth and M of young individuals of E. edulis; (2) to identify relationships between site variables (canopy opening, soil nutrient availability, density of existing E. edulis; and herbivory) and responses of young individuals of E. edulis (growth and M).

Materials and methods The study was conducted in the municipality of Maquine´-RS (between 51°210 –50°050 W and 29°200 – 29°500 S), situated in the north coastal region of Rio Grande do Sul state, Brazil (Fortes 1959), whose original vegetation cover is dense ombrophilous forest (Atlantic Forest). The climate is classified as being Cfa (subtropical), according to Ko¨eppen’s classification, characterized by annual mean temperature of 20°C, relative humidity of 80%, and annual rainfall of 1700 mm well-distributed throughout the year. The mean temperature of the warmest month (January) is 24.5°C and the coldest month (July) is 15.5°C; the winter is mild, with few frosts of low intensity (Model and Sander 1999). The experimental design consisted of planting E. edulis in 24 plots of 100 m2 (10 9 10 m): 12 plots in areas of secondary dense ombrophilous forest, and 12 plots in banana plantations. The secondary dense ombrophilous forest areas were in advanced stage of regeneration, with canopy height of about 12 m, possessing predominance of species of the families Myrtaceae, Rubiaceae, and Fabaceae (Sevegnani and Baptista 1996). The selected banana plantations (Musa sp. AAB, subgroup Prata) for this study were characteristic of southern Brazil (Moreira 1987), cultivated at a density of about 1,333 plants per hectare, limed and fertilized regularly, with yields of around 10 tons ha-1. All plots were installed in areas with altitudes varying between 30 and 130 m, and slope varying between zero and 25% at different solar

305

orientations. The main soil types were Chernozem and Leptosoil (IUSS Working Group WRB 2006) of the Cirı´aco-Charrua soil map unit (Brasil, Ministe´rio da Agricultura 1973). The plots were installed in 2003 by planting 25 young individuals of E. edulis with the same origin (seeds from matrices close to the sites) and similar size (11.9 ± 2 cm tall) in each plot, spaced every 2 meters, thus totaling 600 planted individuals. Morphometric variables, herbivory, and M of individuals were recorded annually in June from 2004 to 2008. Morphometric variables evaluated for each individual were stem base diameter (SD), height (H), and quantity of live leaves (QL). The latter was defined as those which were fully expanded and presented green color on at least 1/3 of their surface area. For statistical analyses, plot averages were made from individual data of SD, H, and QL. Current annual increment (CAI) of SD and H in each annual period were calculated using the equation CAI = Ym - Ym - 1, where Y is the value of the variable and m the year (Iman˜a Encinas et al. 2005). M was recorded as the percentage of dead individuals of E. edulis per plot. The incidence of herbivory (HE) was also evaluated as the percentage of individuals of E. edulis per plot with clear evidence of HE (at least 10% of foliar damage by insects). In 2008, canopy openness (CO), leaf area index (LAI), selected soil properties, and density of preexisting E. edulis (Eu) were measured at each plot. In order to calculate the percentage of CO, three hemispheric canopy photos were obtained for each plot using a Nikon Coolpix 995 digital camera with a ‘‘fish-eye’’ lens facing skyward, at 1.5-m height above ground, leveled and aligned with magnetic north. Image processing was performed using the software Gap Light Analyzer 2.0 (Frazer et al. 1999). The LAI was obtained from the same images, using the portion of images of 60° from zenith—LAI 4 Ring (Stenberg et al. 1994). The density of preexisting E. edulis (Eu) was evaluated only in forest plots, where E. edulis was naturally present, by counting individuals [with diameter at breast height (DBH)—higher than 5 cm], and then converting the value to hectares. Soil samples were obtained from 0 to 20 cm layers in six subsamples distributed in a ‘‘W’’ configuration per plot. Soil analyses were conducted at the Agricultural Chemistry Lab of FEPAGRO—State

123

306

Agroforest Syst (2010) 80:303–313

Foundation for Agricultural Research, as described by Tedesco et al. (1995). The selected soil properties for evaluation were phosphorus (P), potassium (K), and sodium (Na), by Mehlich-1 extraction; organic matter (OM), by Na2Cr2O7 ? H2SO4 combustion; calcium (Ca), magnesium (Mg), and manganese (Mn), by KCl extraction; zinc (Zn) and copper (Cu), by HCl extraction; boron (B), by hot water extraction; iron (Fe), by C2H8N2O4 extraction; hidrogenionic potential (pH); hydrogen ? aluminum (H ? Al), by SMP estimation; cation exchange capacity (CEC); and percent of base saturation (V). Boxplots (McGill et al. 1978) were created to summarize the data of all the variables in 2008. The comparison of means between forests and banana plantation plots was made by analysis of variance with randomization testing (Pillar and Orlo´ci 1996). In order to verify exploratory interpretable patterns and identify relationships among plot variables and individuals of E. edulis, ordination analysis was conducted (PCOA) from a dissimilarity matrix (Euclidean distance) between plots, as described by the variables mentioned above (Podani 1994). For this analysis, data were transformed (centralization and normalization), allowing comparison of variables with different scales. In the scattergram, only variables with a minimum correlation of |0.6| with at least one of the axes were plotted. Mean of M in time (2003–2008) was illustrated by regression analysis, and the best models were selected by significance of Ftest, high R2, biological logic, and significance of the regression parameters (Steel et al. 1997); the data were also subjected to residuals analysis (independence, homogeneity of variance, and normality of distribution), and no transformation was necessary. Spearman correlation coefficients were calculated and evaluated in terms of significance between the morphometric variables of E. edulis and certain variables representative of biotic (M and HE), soil (V), and canopy (CO) conditions. The randomization test and ordination were conducted using the program Multiv 2.3.20 (Pillar 2006), and other analyses using SPSS 10.0.1 (SPSS Inc. 1999).

was five times higher compared with individuals in forest plots. Individuals in banana plots also presented twice as many QL in relation to forest plots (Fig. 1; online resource). The CAI of H of E. edulis observed in forest plots was 7.3 cm year-1, remaining stable; for banana

Results

Fig. 1 Boxplots of stem base diameter (SD), height (H), and quantity of live leaves (QL) of individuals of Euterpe edulis in 2008, planted in 2003 in secondary dense ombrophilous forest and banana plantations in southern Brazil. P probability, randomization test

Five years after initial planting, average size of individuals of E. edulis (SD and H) in banana plots

123

7

SD (cm)

6 5 4 3 2 1

P<0.001

0

Height (cm)

300

200

100

P<0.001

0

Quantity of leaves

9

6

3

0

P<0.001

Forest plots

Banana plots

– median interquartile range amplitude o outlyers

Agroforest Syst (2010) 80:303–313

307

30 25 20

y = -0.5952x2 + 8.5048x + 2.2 R2 = 0.99**

15 10

Forest plots

5

Banana plots 0 2004

2005

2006

2007

2008

Fig. 3 Cumulative mortality (%) of young individuals of Euterpe edulis planted in 2003 in secondary dense ombrophilous forest and banana plantations in southern Brazil. * F2,2 = 295.9, P = 0.003. ** F2,2 = 97.44, P = 0.010

total data variation, summarizing and corroborating to Fig. 4. Banana plots, located predominantly in the lower right quadrant, were primarily associated with high SD, CO, HE, soil pH, and V. The inverse was found for forest plots in other quadrants of the scattergram, being mainly associated with high levels of OM, Mn, and H ? Al. For both forest and banana plots, there was a significant positive correlation between SD and H, as would be expected. For forest plots, significant negative correlations of H and SD with M were observed, and positive correlations occurred between HE, M, and Eu. For banana plots, a significant negative correlation was detected between H and V, whereas CO and V showed a significant positive correlation (Table 1).

60

100

Herbivory (%)

Mortality (%)

Fig. 2 Boxplots of mortality and incidence of herbivory of individuals of Euterpe edulis in 2008, planted in 2003 in secondary dense ombrophilous forest and banana plantations in southern Brazil. P probability, randomization test

y = -0.5x2 + 6.7667x + 9.4667 R2 = 0.99*

35

Mortality (%)

plots, CAI was 38.9 cm year-1 on average, being 17.6 cm in year one, and 63.4 cm in year five. The CAI of SD in forest plots was 0.07 cm year-1, whereas in banana plots the mean was 0.89 cm year-1, being 0.33 cm in year one, and 1.04 cm in year five. During the five-year study period, there were no individuals with exposed stipe in forest plots. In 2007, banana plots presented 15 individuals of E. edulis (6.9% of the remaining 218) with exposed stipe, one of them with DBH of 6.0 cm. In 2008, 63 individuals (30% of the remaining 210) showed exposed stipe, and 14 of them (6.7% of remaining) presented stipes greater than 1.3 m in height, with DBH minimum and medium of 4.7 and 5.7 cm, respectively. Mortality showed high amplitude without significant difference between forest and banana plots, and HE was twice as high for banana plots compared with forest plots (Fig. 2). For both forest and banana areas, M was high in early years tending to stabilize at the end of the evaluation period (Fig. 3). Average values of canopy variables and soil properties are shown in Fig. 4. Banana plots presented significantly higher CO compared to forest plots, indicating greater access to light under banana canopies. Banana plots showed higher soil P, Zn, Fe, V, and pH compared to forest plots, whereas Mn content was higher for forest plots. Ordination analysis (Fig. 5) revealed a separation between forest and banana plots, associated with different variables correlated with axes I and II that explain 67.9% of

40

20

P=0.781

0

Forest plots

75

50

25 0

P=0.001

Forest plots

Banana plots

Banana plots

– median interquartile range amplitude

123

4

25

Leaf area index

20 15 10 5

P=0.003

0

– median 3

interquartile range amplitude

2

Zinc (mg/L)

30 20

Iron (%)

.6 40

20

10

P=0.009

CEC (me/100mL)

pH

V (%)

60

5 40

P<0.001

10

P=0.370

0

8 6 4 2

P=0.477

0

90 60 30

P=0.617

0 10 8 6 4 2

P=0.093

0

Forest plots

Banana plots

Discussion The high growth of E. edulis in banana plots can be attributed to the greater availability of soil nutrients associated with higher luminosity in comparison with forest plots. Fertilization and liming are causes of the

30 20 10 P=0.454

2

1 P=0.258

0 3

300 200 100

P=0.627

0 120 100 80 60 40 20 0

P=0.001

Forest plots

Banana plots

2

1 P=0.229

0

Organic matter (%)

120

40

3

10

Boron (mg/L)

Potassium (mg/L)

150

Manganese (me/100mL)

Sodium (mg/L)

12

400

180

P=0.013

0

Copper (mg/L)

20

P=0.011

20

Magnesium (me/100mL)

Calcium (me/100mL)

30

.3

50

80

4

.4

.1 0.0

P<0.001

0 100

6

.5

.2

10

7

H+Al (me/100mL)

outlyers

.7

30

0

123

o

P=0.092

1

50

Phosphorus (mg/L)

Fig. 4 Boxplots of canopy openness, leaf area index, and selected soil properties in plots under secondary dense ombrophilous forest and banana plantations in 2008 with individuals of Euterpe edulis in southern Brazil. P probability, randomization test

Agroforest Syst (2010) 80:303–313

Canopy oppeness (%)

308

9 8 7 6 5 4 3

P=0.078

Forest plots

Banana plots

great availability of soil nutrients at banana plantations. For E. edulis, high levels of available radiation lead to the optimum use of nutritional conditions (Illenseer and Paulilo 2002). Individuals of E. edulis benefit from canopy openings, since they can adjust their growth and nutrient use efficiency when light

Agroforest Syst (2010) 80:303–313

309

II = 21,6 %

OM º º

º

º H+Al Mn º º º º

NaCEC K Ca Mg B

º º

apparently is not present at banana plantations. Higher growth of E. edulis in banana plots can also be attributed to the gap effect (Denslow 1987), mainly represented by increased radiation, and other associated factors such as temperature and soil humidity. Thus, banana plantations may present microclimatic conditions similar to forest gaps and edges. The high growth of E. edulis in banana plots points to the feasibility of establishing E. edulis at these sites. The shade provided by banana plantations in southern Brazil may indeed be satisfactory, since the young palm E. edulis does not tolerate direct sunlight incidence (Ruschel et al. 1997). These results indicate a plasticity of E. edulis in the acclimatization of individuals in the managed environment of banana plantations, which have different biotic and abiotic conditions compared to their natural forest habitat. This plasticity allows its regeneration in forest gaps and edges, including margins of water bodies. Euterpe edulis, in relation to its successional group, has been classified as a shade-tolerant— climax species (Schorn and Galva˜o 2006). Marcos and Silva Matos (2003) found large numbers of young individuals of E. edulis in impacted forest areas, contradicting the literature that considers E. edulis as a climax species (Macedo et al. 1993; Mantovani 1993; Sa´-Rocha et al. 2002; among others). The high growth of E. edulis in banana plots

º

º I = 46,3 %

V pH

CO

HE

SD

o Forest plots Banana plots

Fig. 5 Scattergram generated through ordination analysis (PCOA) from Euclidean distance between plots of secondary dense ombrophilous forest and banana plantations of southern Brazil with individuals of Euterpe edulis planted in 2003, and correlated variables ([|0.6|) with at least one of the axes. Percentages indicate the representation of each axis in total data variation

and nutrient availability are varied (Illenseer and Paulilo 2002). Within forests, low luminosity is a limiting factor on E. edulis growth (Paulilo 2000), but such a limitation Table 1 Spearman correlation coefficients between stem base diameter (SD), height (H), herbivory (HE), and mortality (M) of planted individuals of Euterpe edulis, and plot variables in H

secondary dense ombrophilous forest and banana plantations in southern Brazil

HE

M

V

CO

Eu

–0.37

–0.72

***

0.26

0.10

–0.26

–0.32

–0.80

***

0.20

0.18

–0.41

*

0.14

0.11

0.48

0.19

0.05

0.76

0.32

0.22

Forests SD

0.91

****

H HE

0.54

M V CO

***

0.38

Bananas SD H HE M

0.67

**

0.38

–0.37

–0.11

–0.17

–0.23

–0.73

***

0.28

0.53 –0.06

*

V

0.25 –0.47 0.72 0.13

***

0.75

***

N = 12, V percent of base saturation of soil, CO canopy openness, Eu density of pre-existing E. edulis *, **, ***, **** Significant at P \ 0.1, 0.05, 0.01, and 0.001, respectively

123

310

in this study is in agreement with the results found by Marcos and Silva Matos (2003). Brokaw (1985) argues that many species are only partially shade tolerant because they depend on increased luminosity levels to reach maturity; this seems to be the case of E. edulis. Perhaps for these and other reasons, Sanchez et al. (1999) classified E. edulis as early secondary. In sum, there is ongoing debate over the classification of species because current classification systems are inadequate to describe the continuum that exists between shade-tolerant and shade-intolerant species. The results concerning forest plots corroborate previous studies indicating that E. edulis is a species that responds to density-dependent effects on M of young individuals (Freckleton et al. 2003) and on predation of fruits and seeds (Nodari et al. 2000). Similarly, these data are in line with Conte et al. (2000) who demonstrate that high M rates occur in early years. Paulilo (2000) argues that if seedlings at the stage of seed depletion reserves do not possess satisfactory photosynthetic capacity, they will probably not be successful in establishment. In this case, photosynthetic inability can be determined by leaf area reduction from HE, phytopathological problems, and low availability of light, nutrients, and water. Despite the large amplitude of percentages, M of E. edulis in this study (Fig. 2) was lower compared with most studies of this species (Bovi et al. 1987; Nodari et al. 2000; Ruschel et al. 1997; among others). M of young individuals of palms can be associated with HE; activity of microorganisms; deficits of water, light, and nutrients; and physical damage occurring from fallen trees, branches, or other plant structures. For banana plots, M seems to be mainly associated with increased HE (Fig. 2), leaf necrosis as caused by direct solar incidence (Fig. 4), and leaf damage due to management practices; for forest plots, it may be related to the activity of microorganisms, HE (Table 1), excessive shade, and competition for nutrients. Incidence of HE on E. edulis in banana plots can be explained by several hypotheses. First, according to the resource availability hypothesis (Coley et al. 1985), high availability of soil nutrients and light may lead plants to invest more in growth in proportion to their defenses, thus increasing the amount of young leaves and consequently, susceptibility to HE. Second, banana plantations show lower saturation of herbivores by food resources (Coley 1983) in comparison with

123

Agroforest Syst (2010) 80:303–313

forests that have natural populations of E. edulis and greater quantity of other species. Third, according to the plant vigor (Price 1991) or plant architecture (Lawton 1983) hypotheses, changes in processes of interaction with natural enemies of herbivores (Wirth et al. 2007) and larger individuals of E. edulis (Fig. 1) may also increase incidence of HE. Finally, more conspicuous individuals (Feeny 1976) in open environments of banana plantations may lead to increased HE. HE on young leaves of E. edulis can be attributed to several organisms, like a coleopterous species (personal information provided by a participating farmer), or Antirrhaea archaea Hu¨bner (Morphinae– Lepidoptera) larvae (Silva Matos 2000). For banana plots, a negative correlation between H and V suggests a negative effect of increased V or other related effects (i.e., HE) on growth of E. edulis. For forest plots, this effect was not observed. Exaggerated quantity or imbalance in proportions of nutrients can cause injury to the growth of species (Malavolta 2006). The amount of factors and complexity of interaction among them represent challenges to the proper execution and interpretation of field studies. The variety of associations found between variables of canopy, soil, individuals of E. edulis, and plots indicates the need for experiments to elucidate specific aspects of responses of E. edulis in relation to biotic and abiotic variables. Before and during the execution of this study, certain farmers reported that the Juc¸ara palm takes more than 15 years to get fruit while others reported that its growth is faster in homegardens or banana plantations. Through local ecological knowledge (Olsson and Folke 2001), these farmers make independent and apparently opposite but correct observations: growth of E. edulis depends on environmental conditions. The results of this study confirm the observations of local farmers and demonstrate the importance of involving local communities in field research. Economic, social, and cultural aspects of local people that manage environments must be considered when conducting research. The behavior of palms in banana plots (fast growth even under high incidence of HE) suggests that E. edulis can show high productivity in managed environments. Variations in quality of heart of palm may also occur due to differences in growth rates. Sustainable extraction rates of E. edulis in managed

Agroforest Syst (2010) 80:303–313

sites may be different from those already verified in natural forest populations (Orlande et al. 1996). Hence, the financial gains from heart of palm extraction may also be higher, even without considering the range of other possible products of E. edulis and other species in agroforestry systems. However, the palm may affect the productivity of banana plantations through competition for resources, depending on the density and growth stage; future research should examine this issue, checking appropriate plant densities in agroforestry systems. The fact that young individuals of E. edulis were acclimatized in banana plots, with higher growth rates in comparison with forest plots, is in agreement with Clement (1999) who stated that a population of a species with little or no degree of domestication, such as E. edulis, can establish and reproduce in managed or non-managed landscapes. These results allow us to consider that many species previously considered as slow-growing or even non-domesticable can be domesticated if managed in agroforestry systems. Agroforestry may promote microclimatic conditions similar to forest gaps, allowing for faster growth of many shade-tolerant species compared to conditions of strong shade such as in forest or full sunlight monocultures. At the same time, agroforestry can promote advantages in yields from intercropping by the increase of the ‘‘Land equivalent ratio’’ (Mead and Willey 1980). In sum, the observed behavior in managed environments suggests that E. edulis is potentially domesticable by creating agroforestry systems in the Atlantic Forest biome, contributing to income generation as well as to conservation of this threatened species. Acknowledgments We are grateful to the National Council for Scientific and Technological Development (CNPq) for its financial support, to the farmers who allowed the operation of this research on their properties, and to Matthew Rehbein for revision of the manuscript. This study is part of Favreto’s PhD thesis for the Post-Graduate Program in Botany—Federal University of Rio Grande do Sul (UFRGS).

References Augspurger CK (1984) Light requirements of neotropical tree seedlings: a comparative study of growth and survival. J Ecol 72(3): 777–795. http://www.jstor.org/stable/2259531 Bovi MLA, Godoy G Jr, Sa´es LA (1987) Pesquisas com os geˆneros Euterpe e Bactris no Instituto Agronoˆmico de Campinas. O Agronoˆmico 39(2):129–174

311 Brasil, Ministe´rio da Agricultura (1973) Levantamento de reconhecimento dos solos do estado do Rio Grande do Sul. Conveˆnio MA/DPP—SA/DRNR, Recife (Boletim Te´cnico, 30) Brokaw NVL (1985) Gap-phase regeneration in a tropical forest. Ecology 66(3):682–687. doi:10.2307/1940529 Chazdon RL, Fetcher N (1984) Photosynthetic light environment in a lowland tropical rainforest in Costa Rica. J Ecol 72(2):553–564. doi:10.2307/2260066 Clement CR (1999) 1492 and the loss of Amazonian crop genetic resources. II. Crop biogeography at contact. Econ Bot 53(2):203–216. doi:10.1007/BF02866499 Coley PD (1983) Intraspecific variation in herbivory on two tropical tree species. Ecology 64(3):426–433. doi:10. 2307/1939960 Coley PD, Bryant JP, Chapin FS (1985) Resource availability and plant antiherbivore defense. Science 230(4728):895– 899. doi:10.1126/science.230.4728.895 Conte R, Reis A, Mantovani A (2000) Dinaˆmica da regenerac¸a˜o natural de Euterpe edulis Mart. (Palmae) na Floresta Ombro´fila Densa da Encosta Atlaˆntica. In: Reis MS, Reis A et al (eds) Euterpe edulis Mart. (palmiteiro)–biologia, conservac¸a˜o e manejo. Herba´rio Barbosa Rodrigues, Itajaı´, pp 106–130 Denslow JS (1987) Tropical rainforest gaps and tree species diversity. Ann Rev Ecol Syst 18(1):431–451. doi:10.1146/ annurev.es.18.110187.002243 Feeny P (1976) Plant apparency and chemical defenses. In: Wallace JW, Mansell RL (eds) Biochemical interactions between plants and insects, vol 10. Plenum Press, New York, pp 1–40 Fortes AB (1959) Geografia fı´sica do Rio Grande do Sul. Globo, Porto Alegre Frazer GW, Canham CD, Lertzman KP (1999) Gap light analyzer (GLA): imaging software to extract canopy structure and gap light transmission indices from truecolour fisheye photographs, users manual and program documentation. Simon Fraser University, Burnaby, British Columbia, and the Institute of Ecosystem Studies, Millbrook, New York. http://www.ecostudies.org/gla Freckleton RP, Silva Matos DM, Bovi MLA et al (2003) Predicting the impacts of harvesting using structured population models: the importance of density-dependence and timing of harvest for a tropical palm tree. J Appl Ecol 40(5):846–858. doi:10.1046/j.1365-2664.2003.00842.x Givnish TJ (1999) On the causes of gradients in tropical tree diversity. J Ecol 87(2):193–210. doi:10.1046/j.1365-2745. 1999.00333.x Henderson A (2000) The genus Euterpe in Brazil. In: Reis MS, Reis A (eds) Euterpe edulis Mart. (palmiteiro)— biologia, conservac¸a˜o e manejo. Herba´rio Barbosa Rodrigues, Itajaı´ IBGE (2006) Produc¸a˜o Agrı´cola Municipal, Sistema IBGE de Recuperac¸a˜o Automa´tica—SIDRA, Instituto Brasileiro de Geografia e Estatı´stica. http://www.sidra.ibge.gov.br. Accessed 23 Nov 2008 Illenseer R, Paulilo MTS (2002) Crescimento e eficieˆncia na utilizac¸a˜o de nutrientes em plantas jovens de Euterpe edulis Mart sob dois nı´veis de irradiaˆncia, nitrogeˆnio e fo´sforo. Acta Bot Bras 16(4):385–394. doi:10.1590/S010233062002000400002

123

312 Iman˜a Encinas J, Silva GF, Pinto JRR (2005) Idade e crescimento das a´rvores. Comunicac¸o˜es te´cnicas florestais, vol 7, no 1. Departamento de Engenharia Florestal, Brası´lia IUSS Working Group WRB (2006) World reference base for soil resources 2006. World Soil Resources Reports 103. FAO, Rome Kahn F, Castro A (1985) The palm community in forest of Central Amazonia, Brazil. Biotropica 17(3):210–216. doi: 10.2307/2388221 Laurance WF, Merona JMR, Andrade A et al (2003) Rainforest fragmentation and the phenology of Amazonian tree communities. J Trop Ecol 19(3):343–347. doi:10.1017/ S0266467403003389 Lawton JH (1983) Plant architecture and the diversity of phytophagous insects. Annu Rev Entomol 28(1):23–39. doi:10.1146/annurev.en.28.010183.000323 Lee DW, Baskaran K, Mansor M et al (1996) Irradiance and spectral quality affect Asian tropical forest tree seedling development. Ecology 77(2):568–580. doi:10.2307/226 5631 Macedo AC, Kageyama PY, Costa LGS (1993) Revegetac¸a˜o de matas ciliares e de protec¸a˜o ambiental. Fundac¸a˜o Florestal, Secretaria do Meio Ambiente, Sa˜o Paulo Malavolta E (2006) Manual de Nutric¸a˜o Mineral de Plantas. Ceres, Sa˜o Paulo Mantovani W (1993) Estrutura e dinaˆmica da Floresta Atlaˆntica na Jure´ia, Iguape, SP. Thesis, Instituto de Biocieˆncia, USP Marcos CS, Silva Matos DMS (2003) Estrutura de populac¸o˜es de palmiteiro (Euterpe edulis Mart.) em a´reas com diferentes graus de impactac¸a˜o na Floresta da Tijuca, RJ. Flor Amb 10(1):27–37 McGill R, Tukey JW, Larsen WA (1978) Variations of box plots. Am Stat 32(1):12–16 Mead R, Willey RW (1980) The concept of a ‘Land Equivalent Ratio’ and advantages in yields from intercropping. Exp Agric 16(3):217–228. doi:10.1017/S0014479700010978 Model NS, Sander GR (1999) Produtividade e caracterı´sticas do fruto de abacaxizeiro em func¸a˜o do preparo do solo e te´cnicas de plantio. Pesqui Agropecu Gau´ch 5(2):209–216 Moreira RS (1987) Banana: teoria e pra´tica de cultivo. Fundac¸a˜o Cargill, Campinas Nakazono EM, Costa MC, Futatsugi K et al (2001) Crescimento inicial de Euterpe edulis Mart em diferentes regimes de luz. Rev Bras Bot 24(2):173–179. doi:10.1590/ S0100-84042001000200007 Nodari RO, Fantini AC, Reis A et al (2000) Restaurac¸a˜o de populac¸o˜es de Euterpe edulis Martius (Arecaceae) na Mata Atlaˆntica. In: Reis MS, Reis A. Euterpe edulis Martius (palmiteiro) – biologia conservac¸a˜o e manejo. Herba´rio Barbosa Rodrigues, Itajaı´, pp 189–201 Olsson P, Folke C (2001) Local ecological knowledge and institutional dynamics for ecosystem management: a study of Lake Racken Watershed, Sweden. Ecosystems 4(2):85– 104. doi:10.1007/s100210000061 Orlande T, Laarman J, Mortimer J (1996) Palmito sustainability and economics in Brazil’s Atlantic coastal forest. For Ecol Manag 80(1–3):257–265. doi:10.1016/03781127(95)03569-9 Paulilo MT (2000) Ecofisiologia de plaˆntulas e plantas jovens de Euterpe edulis: comportamento em relac¸a˜o a variac¸a˜o

123

Agroforest Syst (2010) 80:303–313 de luz. In: Reis MS, Reis A (eds) Euterpe edulis Mart. (palmiteiro)—biologia, conservac¸a˜o e manejo. Herba´rio Barbosa Rodrigues, Itajaı´, pp 93–105 Pillar VD (2006) MULTIV 2.3.20: multivariate exploratory analysis, randomization testing and bootstrap resampling. Copyright Ó 2004 by Vale´rio DePatta Pillar. Universidade Federal do Rio Grande do Sul, Porto Alegre. http:// ecoqua.ecologia.ufrgs.br/ecoqua/software.html Pillar VD, Orlo´ci L (1996) On randomization testing in vegetation science: multifactor comparisons of releve´ groups. J Veg Sci 7(4):585–592. doi:10.2307/3236308 Podani J (1994) Multivariate data analysis in ecology and systematics. SPB Academic Publishing, The Hague Price PW (1991) The plant vigor hypothesis and herbivore attack. Oikos 62(2): 244–251. http://www.jstor.org/pss/ 3545270 Reis A, Kageyama PY (2000) Dispersa˜o de sementes de Euterpe edulis Mart. Palmae. In: Reis MS, Reis A (eds) Euterpe edulis Mart. (palmiteiro)—biologia, conservac¸a˜o e manejo. Herba´rio Barbosa Rodrigues, Itajaı´, pp 60–92 Reis MS, Fantini AC, Nodari RO et al (2000) Management and conservation of natural population in Atlantic rainforest: the case study of palm heart (Euterpe edulis Martius). Biotropica 32(4b):894–902. doi:10.1111/j.1744-7429. 2000.tb00627.x Ruschel A, Reis MS, Nodari RO (1997) Genetic variation of Euterpe edulis progenies under different light intensities In: Abstracts of the 43rd Congresso Nacional de Gene´tica, Goiaˆnia, 1997. Braz J Genet 20 (3 Suppl.): 326 Sanchez M, Pedroni F, Leita˜o-Filho HF et al (1999) Composic¸a˜o florı´stica de um trecho de floresta ripa´ria na Mata Atlaˆntica em Picinguaba, Ubatuba, SP. Rev Bras Bot 22(1):31–42. doi:10.1590/S0100-84041999000100006 Sa´-Rocha J, Lima AM, Terra G et al (2002) Avaliac¸a˜o do enriquecimento por dispersa˜o natural de palmiteiro Euterpe edulis—Martius em trecho de Floresta Ombro´fila Densa Montana na regia˜o de Miguel Pereira—RJ, Brasil. Rev Univ Rural, Ser Cieˆnc Vida 22(2):205–209 Schorn LA, Galva˜o F (2006) Dinaˆmica da regenerac¸a˜o natural em treˆs esta´gios sucessionais de uma Floresta Ombro´fila Densa em Blumenau, SC. Floresta 36(1):59–74 Sevegnani L, Baptista LRM (1996) Composic¸a˜o florı´stica de uma floresta secunda´ria, no aˆmbito da Floresta Atlaˆntica, Maquine´. Sellowia 45–48:47–71 Silva Matos DM (2000) Herbivore and plant demography: a case study in a fragment of semi-deciduous forest in Brazil. J Trop Ecol 16(1):159–165. doi:10.1017/S02664 67400001334 SPSS Inc 1999. Statistical Package for Social Sciences, release 10.0.1, standard version Steel RGD, Torrie JH, Dickey DA (1997) Principles and procedures of statistics: a biometrical approach, 3rd edn. McGraw-Hill, New York Stenberg P, Linder S, Smolander H et al (1994) Performance of the LAI-2000 plant canopy analyzer in estimating leaf area index of some Scots pine stands. Tree Physiol 14(7–9): 981–995. doi:10.1093/treephys/14.7-8-9.981 Tedesco MJ, Gianello C, Bissani CA et al (1995) Ana´lises de solo, plantas e outros materiais, 2nd edn. Universidade Federal do Rio Grande do Sul, Porto Alegre

Agroforest Syst (2010) 80:303–313 Venturi S, Paulilo MTS (1998) Esgotamento das reservas na semente de Euterpe edulis Mart e efeito da nutric¸a˜o mineral nas plaˆntulas. Acta Bot Bras 12(3):215–220 Vivan JL (2002) Bananicultura em sistemas agroflorestais no Litoral Norte do RS. Agroecol Desenv Rur Sustent 3(2):17–26

313 Wirth R, Meyer ST, Almeida WR et al (2007) Increasing densities of leaf-cutting ants (Atta spp.) with proximity to the edge in a Brazilian Atlantic forest. J Trop Ecol 23(4):501–505. doi:10.1017/S0266467407004221

123