REFORESTATION FEASIBILITY IN AREA FORMERLY USED FOR CATTLE RASING IN THE STATE OF RONDÔNIA, NORTHWEST BRAZILIAN AMAZON

– Little knowledge on initial behavior of native tree species in recovering landscapes in the Amazon is a current concern for expanding reforestation in the region. Thus, the aim of this study was to evaluate the establishment of native tree species that could be used for reforestation in area previously covered by a pasture of brachiaria grass ( Brachiaria brizantha ) destined for intensive cattle rasing in the State of Rondônia. For this, there were performed previous diagnostic of landscape changes and the election of tree species based on the ecological group information. Some of the critical macronutrients for plant growth were supplied in the holes to alleviate nutrient deficiencies. In addition, growth and survival parameters were taken to evaluate the initial behavior of species. Six native tree species planted with different combinations (10mx10m, 5mx5m and 3mx3m) had survival rate and growth (total height, girth stem and crown projection area) measured in different intervals: 6-month, 12-month and 24-month after planting. All the species presented survival rate over 90% at 24 months and comparable growth indices to other native species under similar situation and in the region. Overall, Schizolobium amazonicum (bandarra), the non-identified legume tree 1 (acácia grande) and Colubrina glandulosa (sóbrasil) averaged over 90% the highest girth stem growth all over the area. S. amazonicum and the non-identified legume tree 1 (acácia grande) presented the best results for height and canopy area growth parameters, respectively. The combination among native tree species from initial successional ecological groups and fertilizer was favorable to promote reforestation in the conditions of the study area in Rondônia.


INTRODUCTION
The expansion of areas for cattle raising together with slash-and-burn agriculture in the State of Rondonia has contributed to keep high deforestation rates since the strategy of regional occupation and development under colonization projects was initiated by the Brazilian government in the region (FEARNSIDE;GRAÇA, 2006;FEARNSIDE, 2007).
By the late 1970s, this portion of the Amazon comprised intensive land-use changes, mainly due to dynamic of human activities, that caused a hard replacement of its ample measures of natural forests (SHONE; CAVIGLIA-HARRIS, 2006), mainly with crop cultivation and pastures for livestock, especially dairy and beef cattle.
Forty years after the settlement, the causes of deforestation in this pioneer frontier are quite the same (ALMEIDA et al., 2006) and they are still reinforced by the expansion of new crop farming that may negatively affect the maintainence of of Legal Reserve (LR) and Permanent Protected (PP) areas in the majority of farms in the State of Rondônia.
The Amazon region accounts for nearly 80 million heads associated with a strong meat and milk industry (POCCARD-CHAPUIS et al., 2010).Regarding the Brazilian milk market, the State of Rondonia is the seventh largest producer and the largest one in the North Region -706.6 million liters (IBGE, 2013), with around 90% of the production coming from small-scale/family-farm business, influencing 40% of the Gross Domestic Product -GDP (SANTANA, 2003).However, cattle production in the Amazon rainforest is strongly linked to few environmental benefits due to the modest productivity rates and doubtful sustainability.
One of the major challenges for sustainable livestock production in the Amazon region is to minimize the impact of deforestation accomplished with environmental regulations to keep forest standing.Moreover, reconciling production and environment under co-friendly production systems (POCCARD-CHAPUIS et al., 2010) may be considered a new dimension for cattle farming in the region.
As a result, over the last few years, initiatives in reforestation with native species in the Amazon rainforest, as a means of reducing negative environmental impact due to intensive production and towards the provision of environmental services, have increased (BARBOSA et al., 2003;ALMEIDA et al., 2006;ALVINO-RAYOL et al., 2011).
Methods for increasing the use of native species in reforestation in the Amazon rainforest have also been tested in recent years (TONINI; ARCO-VERDE, 2005;MELOTTO et al., 2007MELOTTO et al., , 2009;;LIMA et al, 2009;ALVINO-RAYOL, 2011).Despite the lack of an efficient spreading of technical information (propagation, establishment and tending) of suitable and promising native tree species for recovering and, or, revegetation is still a barrier for technicians and landowners to invest in reforestation practices.
Moreover, the use of native tree species is a recommendation to be followed according to the Brazilian Forestry Code, within forestation and or, reforestation activities in Legal Reserve and Permanent Protected areas of farms in the Brazilian Amazon (BRASIL, 2012).
Therefore, the objective of this study was to evaluate the feasibility of establishing six native tree species in area intensively used by cattle farms in the State of Rondônia, as an aid to reforestation initiatives under similar conditions.
Reforestation feasibility in area formerly used for cattle...

Study area
The study was carried out at the research station of Embrapa located in the municipality of Presidente Médici (Figure 1 -11º17'S and 61º55'W and 390 m.a.s.l), Rondonia Sate, 400 km far from the capital.The climate is tropical with a dry season -Aw (Köppen classification) with a rainfall ranging from 2.000 mm to 2.300 mm (TOWNSEND et al., 2003).Rainy season is from November to April, when 80.3% of total annual rainfall of the region has been recorded (SCERNE et al., 2000).Open Ombrophilous Forest is the dominant vegetation type, characterized by open canopy with emergent individuals and high frequency of palm trees (BRASIL , 1978;IBGE, 2012).The landscape presents smooth relief and has suffered alteration by cattle raising activity which is the main activity of land use in this region since the 1980's.Annual average temperature is 24.5°C with an average relative air humidity of 80% and Podzolic Red-Yellow as the predominant soil in the area (TOWSEND et al., 2000).
The research station of Embrapa has 95.63 ha (INCRA, 2008) and part of this area, as most native forest cover in the municipality, is highly fragmented and restricted to small-medium isolated spots of natural vegetation.The Embratel stream belongs to the hydrographic basin of Machado River and plays an important role as the main water source inside the research station.Deforestation surrounding the property is the main cause of the negative effects in the stream flow, which has been intensified with the presence of the dairy cattle, since the 1980s (PEREIRA; TAVARES, 2000).In the last five years, it has been observed that the Embratel stream flow presents high variation in volume of water, reaching critical levels mainly from May to October, which is the dry period in the region.

Site selection and preparation for planting
A preliminary survey along the Embratel stream was carried out in December, 2007 to evaluate the main effects of the current cattle farming management system.By then, the reforestation of the surrounding area of the riparian forest was taken as an urgent task to attend the environmental law requirements for Legal Reserve and Permanent Protected areas.At that time, around 3.86 ha of Marandu grass (Brachiaria brizantha) was the main vegetation surrounding the edge of Embratel stream.
The species were classified into successional groups according to Carvalho (2003Carvalho ( , 2006) ) and Rede (2004).Thus, six target species -canopy and understory trees -frequent and adapted to the edafoclimatic conditions of Northwest Brazilian Amazon were chosen (Table 1).The plant strategy of species followed the characteristics of each successional group, and it is described according to Gris et al. ( 2012) as follows: i) pioneer: light demanding species during most of their lifetime, with very fast initial growth and short lifetime (from four to 30 years); important in the formation of seed banks; ii) secondary species: shade tolerant species in the beginning of their lifetime, with a moderate to rapid initial growth and a lifetime from 25 to 100 years; iii) climactic: strong shading tolerant species, may remain in the understory during their entire lifetime, slow groth, and a moderate to very long lifetime, forming plantlet banks.The elected species are also commonly used in reforestation campaigns under the Rehabilitation Land Program held by the Environmental Policy Quarters of Candeias do Jamari, settled 20 km far from the capital, Porto Velho.
Because of the relief conditions on the farm, the reforestation trial was settled in 0.65 ha in the beginning of the rainy season in January,2008.Different plant densities (10mx10m, 5mx5m and 3mx3m) formed 31 rows, set according to the number of plants and its ecological group (GRIS et al., 2012;LIMA et al., 2009;RODRIGUES and GANDOLFI, 1996).Seedlings with 0.30 m of average height were planted in holes of 50 cm depth x 50 cm diameter.Mineral fertilization with some of the critical macronutrients for plant growth were supplied in the holes only to soothe nutrient deficiencies, as follows: 100g.kg - dolomite calcarium (70% PRNT) + 1 liter of dry bovine manure + 10g.kg -1 simple superphosphate (sps) + 5g.kg -1 potassium chloride + 4g.kg -1 FTE (micronutrients).
All holes were set 30m away from the edge of the stream to avoid contamination by the fertilizers.Electric isolation protected the perimeter of the planted area.Weeding and plants coronation were made as cultural tracts twice a year.Survival (%) and growth rates (total height in m, and girth in cm), with adaptions from Melotto et al. (2009), of a sample of 224 measurable seedlings were taken in three different intervals, 6, 12 and 24 months, after planting.In addition, all seedlings had crown projection area (m 2 ) assessed at the ages of 12 and 24 months, following Hemery et al. (2005).Plants were recorded as dead when the stem was dry and had no leaves.

Statistical analysis
Aiming to quantify the effects and iteration, the species growth parameters at the three plant densities were considered according to: Cada grupo de espécies representa uma estratégia diferente de plantio.Reforestation feasibility in area formerly used for cattle... where: Y ij = i th specie within j th plant density, U = general mean of the experiment, A i = i th plant density, E j: j th species, AE = iteraction between the i th plant density and the j th species, and The PROC GLM procedure of SAS PROC GLM Statistical Analysis System (SAS, 1990) was used to adjust unbalanced experiments.Turkey's t test was conducted on significant variables by ANOVA to test differences among the means (CALLEGARI-JACQUES, 2003).
Regarding seedling height growth, no density effects in the age of evaluation were found.S. amazonicum (bandarra) and the non-identified 1 species (acácia grande), pioneer species, sustained growth over the subsequent 2 years and exhibited the highest values when compared to the other species in the area (Figure 2).Much of height growth was observed at the end of the 24 th -month, with more than 60% of species reaching over 3.10 m, including C. glandulosa and the nonidentified 2 species (acácia pequena), secondary species whereas H. courbaril, with expressive survival rate, did not present sufficiently good growth among the tested species.
The highest values for girth stem were observed for the non-identified 1 species (acácia grande) and S. amazonicum (bandarra), respectively (Figure 3).The light demanding species T. serratifolia (ipê amarelo) did not present a superior performance.
Estimates of stand density had no effects on the growth behavior of the evaluated species.Considering the crown projection as a landscape recovery indicator, results showed that more than 50% of the species were able to promote important protective functions in the area (Figure 3) in the 24-month period, as weed suppression by shadowing the soil, mainly represented by the non-identified 1 legume tree (acácia grande) and S. amazonicum (bandarra).
In this context and even with lower values, results found observed for H. courbaril can be considered satisfactory when compared to Tonini and Arco-Verde (2005), who observed 60.8 m 2 of crown projection area for this species at the age of seven years in the State of Roraima, under 2.5m x 2.0m spacing.

DISCUSSION
Environmental adaptative capacity is one of the main factors for species survival in forestation and reforestation programs.The establishment of the native tree species in this trial tended to favorable results, due to the plasticity of the species (ALVINO-RAYOL et al., 2011), and although growth behavior should be monitored over time (WISHNIE et al., 2007), the highest mortality rates for all species were mostly found only after the critical establishing phase for tropical seedlings, and in the beginning of the third year of the trial.
The low mortality rate for some species, as for example the pioneer leguminous tree species S. amazonicum -bandarra (5%), also found in Alvino-Rayol et al. (2011) in pure stands with the species, also contributes to corroborate this species as a good source of environmental services on farms, even under modified natural forest cover conditions and deficient water availability during the dry season in the region.

BENTES GAMA, M.M. et al.
The height growth for most species followed a trend found in other areas in the Amazon region, such as C. glandulosa, the non-identified 2 species (acácia pequena), both secondary species.S. amazonicum presented expressive growth in height, which reached over 3.10 m in the late evaluation of this study at the age of 24 months in Presidente Médici.Such results represent 30% of the height found by Rossi et al. (2003) in the Central Amazon for the species under 3m x 2m spacing and 48 months of age.According to Vieira et al. (1998), wide initial densities might avoid the effects of mutual competition among species.Few species as for example H. courbaril did not present sufficiently good growth among the tested species, partially due to its ecological characteristics, suggesting that it may be suitable for reforestation under the same site conditions, however.These findings were similar in Lima et al. (2010), who observed plants of H. courbaril with average height of 41.67 cm at 4 months of age in the Municipality of Ji-Paraná, 90 km away from Presidente Médici.
The light demanding species T. serratifolia did not present higher superior performance as expected.A similar behavior for this genera was found in Melotto et al. (2007), who observed low values of girth stem for Tabebuia sp. in a mixed plantation with Brachiaria brizantha under the age of 12 months.
Crown projection can be used as another important parameter to follow species behavior in reforestation initiaves (TONINI; ARCO-VERDE, 2005), mainly in this region in the Amazon forest, where understanding ecological and functional characteristics of target native tree species (MELLOTO et al., 2009) are of major need for promoting reforestation for structural and functional diversity in areas where cattle farming needs to fit to environmental requirements.

CONCLUSION
The low mortality and initial growth characteristics (plant height, girth stem, crown diameter) of native tree species from initial successional ecological groups adapted to Amazon soil and climate conditions, with the supplying of critical macronutrients by artificial fertilization tends to favor the recovery in areas previously used for cattle farming as their values continually increased as plant age increased.Despite the intrinsic ecological and functional characteristics of each native tree species used in the experiment, the pioneer Schizolobium amazonicum and the secondary Colubrina granulosa aachieved the highest growth values in all growing seasons.

ACKNOWELDGEMENTS
The authors thank Neilton Santos Pereira, Giovana Fiorella Zamora López and Poliana Heloisa da Silva Capelasso for their important contributions in the different stages of this initiave sponsored by Embrapa, and coordinated by its Animal Production Scientific Core in Rondônia.

Figure 1 -
Figure 1 -Localization of the study area at the research station of Embrapa in the Municipality of Presidente Médici, State of Rondônia.Figura 1 -Localização da área de estudo na estação experimental da Embrapa no município de Presidente Médici, Estado de Rondõnia.

Table 1 -
Tree species and spacing in the reforestation trial in the Municipality of Presidente Médici, State of Rondonia.Each group represents a different tree planting strategy.Tabela 1 -Espécies arbóreas e densidades do reflorestamento no município de Presidente Médici, Estado de Rondônia.