Phosphorus Dynamics in the Conversion of a Secondary Forest Into a Rubber Tree Plantation in the Amazon Rainforest

TECHNICAL ARTICLE

Phosphorus Dynamics in the Conversion of aSecondary Forest Into a Rubber Tree Plantation

in the Amazon RainforestAdônis Moreira J Larissa Alexandra Cardoso Moraes, J Rean Augusto Zaninetti/, 3

and Bruna Trova Canizella

Abstract: During the past several decades, extensive areas of the hu-mid tropical Amazon have been cleared for the introduction of pastureor subsistence farming and later abandoned after a few years of use.An option for recovering these areas is the planting of rubber trees be-cause it is a native tree that could restore conditions similar to thoseexisting before the clearing. ln addition, the high economic value of nat-ural rubber is a source of income for small producers. In soil manage-ment under the Amazon conditions, low phosphorus (P) has been themost limiting factor. Under natural conditions, P is provided to theplants almost exclusively by organic matter mineralization. The aim ofthis study was to evaluate a chronosequence of reforestation with rubbertrees planted at different times in cleared areas, with primary forest as areference. The amounts of litter and P content in the plant, in the litter,and in soil, as well as biological indicators (acid and alkaline phospha-tase and P microbial biomass) associated with the P cycle in a XanthicFerralsol (Oxisol), in the Central Amazon were assessed. Rubber treegrowth resulted in changes in total organic carbon, with an increase of104.6% in the 45-year-old rubber trees compared with the 6-year-oldrubber tree plantation. This was also observed for acid and alkalinephosphatase activities, which were close to those of the primary forest,The formation oflitter and the P content of the microbial biomass in soilwere higher in rubber areas. The P immobilized into microbial biomasswas the main reserve to meet the plant's nutritional demand for phospho-rus. The content of P available in the soil, regardless of the extractants(Mehlich I, Mehlich 3, and Bray I) and the age ofvegetation cover werebelow the levels indicated as appropriate.

Key Words: Acid phosphatase, alkaline phosphatase, Hevea species,Amazon rainforest, phosphorus microbial biomass, organic P and total P

(SoU Sei 2013;178: 618-625)

In the humid tropical Amazon, conversion ofthe primary forestinto crops of economic interest has most often had a negative

impact on this ecosystem because of the remova I of trees ofhigh economic value, conversion of lhe remaining forest treeinto firewood, and burning the underbrush for planting subsis-tence crops or pasture (Andreux and Cerri, 1989; Schroth et al.,2002; Moreira and Fageria, 2011).

'Ernbrapa Soybean, Paraná State, Brazil.2Federal University of Amazonas, Amazonas State, Brazil.3Estadual University ofLondrina, Paraná State, Brazil.Address for correspondence: Dr. Adônis Moreira, Embrapa Soybean,Caixa Postal 231, 86001-970, Londrina, Paraná State, Brazil. E-mail:[email protected] Disclosures/Conflicts of Interest: This study was supported byCNPq (National Counsel ofTechnological and Scientific Development).Received August 30,2013.Accepted for publication December 12, 2013.Copyright © 2014 by Lippincott Williams & WilkinsISSN: 0038-075XDOI: 1O.1097/SS.0000000000000025

618 I www.soilsci.com

Often, there is no application of fertilizers and productsused to control soil pH in these areas. Thus, there is a sharp de-crease in soil fertility with time; thus, the sites are deserted(Cravo and Smyth, 1997), giving rise to a secondary forest(capoeira) ofno economic interest and little biodiversity (Souzaet al., 2010).

The introduction of agroforestry systems and monocul-tures of perennial plants are economic options for restoringthe vegetation, as well as a source of income for small pro-ducers of the region, based on the ecological aspects involvedin the progressive accumulation of litter and the production ofitems of economic interest for small producers in the region(Schroth et al., 2002). For example, Bi and Omont (1987) foundthat, with fully grown rubber trees, there was litter deposition inthe soil from 5.0 to 7.0 Mg ha-1, which corresponded to 90 to95% ofthe litter deposed in primary forests located in areas un-der the same climatic and soil conditions. The decomposition ofplant residues in soil is influenced broadly by substrate quality(CIN index), climatic conditions, and decomposer biota (Tian,1998). Concerning the economic aspect, rubber cultivation de-mands family labor because the latex frorn the trees is manuallycollected, providing an income for the population, which, in thecase ofthe Amazon, occurs almost every month ofthe year.

Besides the fact that the weather conditions ofthe Amazonregion provide rapid mineralization of organic matter (OM) incleared areas, the soils have low fertility, with high levels of ex-changeable AI, low base saturation, and 83% ":ith available P-Mehlich 1 extractant-below 5.4 mg kg (Moreira andFageria, 2009), reducing the period of nutrient uptake. In thisecosystem, microorganisms play a key role in the decreasedavailability ofP in climax plant communities especially becausethe nutrient fiows through the microbial mass where it isretained, which reduces soil P content. Microorganisms andplant roots also convert organic forms of P (Po) to plant-available inorganic forms of P (Pj) because of acid and alkalinephosphatase activities (Tabatabai, 1994). According to Dick andTabatabai (1993), Carneiro et aI. (2004), and Costa and Lovato(2004), soil microbes are the main producers of phosphatasesbecause of their high metabolic activity compared with plants.

Therefore, the purpose of the present study was to esti-mate the impacts ofreforestation with rubber trees on the avail-ability of carbon (C) and the dynamics of P in soil, in litterand in the plant, as well as the biological indicators of clearedareas reforested with rubber trees planted in different periods(within 6,16,18,19,20, and 45 years) compared with the orig-inal (primary) forest in a Xanthic Ferralsol (Oxisol) in lhe cen-tral Amazon region.

MATERIAlS ANO METHOOSThe study was carried out on the research station

of Embrapa in the county of Manaus, Amazonas State, Brazil(3°8'25"LS and 59°52'LW). The predominant climate in the

Copyrighl © 2014 Lippincolt Williams & Wilkins. Unaulhorized reproduclion of lhis article is prohibiled.

Sai! Science • Volume 178, Number 11, November 2013

Soil Science • Volume 178, Number 11, November 2013 Phosphorus Dynamics

TABLE 1. Minimum and Maximum Temperatures and Monthly Rainfall in the County of Manaus, Amazonas State, Brazil(2008 and 1971-2008) (Antonio, 2009)

- - - - - - - - - - - - - - - -2008- - - - - - - - - - - - - - - - - - - - - - - - - - - - -1971-2008- - - - - - - - - - - - -Temperature Temperature

Minimum Maximum Rainfall Minimum Maximum RainfalI

Months_________ oC_________

- - - mm- --_________ oC_________

- - - rnm- - -

January 21.8 31.3 286.7 22.5 30.7 264.6February 21.0 31.6 443.1 22.6 30.5 294.8March 20.9 31.3 456.3 22.6 30.8 327.0April 21.3 31.8 391.6 22.6 31.0 327.3May 21.4 31.1 445.9 22.4 31.1 2S1.3June 21.1 31.1 178.5 21.9 30.9 167.5July 20.9 33.2 106.7 21.5 31.4 120.9August 21.4 33.9 210.4 21.5 32.6 113.SSeptember 21.7 33.7 111.1 22.1 33.0 124.1October 22.0 33.6 158.7 22.4 33.0 161.9

ovember 22.4 33.2 467.9 22.6 32.4 193.7December 22.3 32.6 298.8 22.5 31.4 242.8Total 3,555.7 2,619.SMean 21.5 32.3 296.3 22.3 31.6 21S.3

fertilizer recommendation for Embrapa Westem Amazon(CPAA) for rubber cultivation (Moraes et aI., 2008). No mainte-nance fertilization was done in the rubber plantations studiedbecause the export of nutrients from the latex is very low(Murbach et aI., 2003).

Pits (1.0 x 1.0 x 0.6 m) were prepared, and sampleswere collected from depths of O to 10 em, 11 to 20 em, and21 to 40 em at the beginning of the rainy season (October andNovember; Table 1), following removal of the litter layer ofthe soil. Sampling consisted of collection at a pit in the cen-ter of the area and in four other pits located within a 25.0-mdistance, using cardinal directions for orientation (Moraeset aI., 1996). This method was repeated twice for a total of10 pseudoreplicates. At each site, 300 g of soil were removedand divided into five subsamples of 60 g. Soil density was alsodetermined at the depth ofO to 10 em (Ajayi et aI., 2009), whichvaried from 0.80 to 1.04 Mg m-3 (average, 0.89 Mg m-3). Alisamples were taken in the rows, outside the crown area of therubber trees, to eliminate the effect of fertilization of the plant-ing hole.

Soil AnalysisAfter collection, the soils were homogenized and placed

in plastic bags and stored in a refrigerator (4 ± 0.5°C). Visi-ble residues from plants and animais were removed, and thesamples were subsequently passed through a sieve of 2.0-mmmesh. Moisture for standardization of samples was deterrninedby the gravimetric method after removal of plant material. Thedetermination of P microbial biomass (pMB) (Brookes et aI.,1982) was performed in duplicate by the method of irradiationand extraction (IE) in the microwave oven (127 V; 2,450 MHz,and 1,380 W), as proposed by Ferreira et aI. (1999) and Mendonçaand Matos (2005). Samples of wet soil corresponding to 5.0 gof dry soil were used. Phosphorus was extracted with 50 mL of0.5 moi L-I sodium bicarbonate (NaHC03) (pH 8.5). The P wasquantified by a colorimetric method using molybdate--ascorbicacid. To correct fixation of part of the P adsorbed onto theclays during extraction, the recovery rate was estimated concomi-tantly in the samples through the addition of 0.5 ug mL-1 of

region is tropical wet (Aí), according to Kôppen climate classi-fication, with relatively abundant rainfall throughout the year(average of 2,600 mm). The amount of rainfall during themonths of lowest rainfall (July to October) was always morethan 100 mm (Table 1). The average annual temperature was ap-proximately 26°C (Antonio, 2009).

The areas were first cleared from the primary forest usingheavy machinery to pile up vegetation debris and for removalof tree stamps. An experiment with rubber trees (Heveabrasiliensis) was established and was abandoned because ofSouth American leaf blight (SALB) (Microcyclus ulei) attack,a fungal disease that caused the decline of rubber production.The secondary forest that developed was manually cleared forthe establishment of tree crops. An adjacent primary forestwas retained as a control for this study.

Soil SamplingSoil samples were collected in areas of the primary forest

and adjacent rubber plantations covering an area of 2.0 ha(476 plants ha-I; 8.0 x 2.5-m spacing) with crown of Heveapauciflora (Spruce ex. Benth.) Müell Arg. and hybrid of H.pauciflora x Hevea guianensis Aubl. planted in six distinctperiods (within 6, 16, 18, 19, 20, and 45 years) in a secondaryforestland in a Xanthic Ferralsol (FAO, 1990; dystrophic Yel-low Latosol-Brazilian classification (EMBRAPA, 1999)) withclayey texture (830 g kg-I). The areas are underlain by Quater-nary sediments of the Alter-do-Chão Formation, The soil'schemical attributes are shown in Table 2.

The planting holes (60 x 60 x 60 em) were prepared30 days before the application of limestone. After planting,they were filled with 100.0 g of dolomitic limestone. Beforethe seedlings were planted, 200.0 g of simple superphosphate(20% P20S), 5.0 g of copper sulfate (13% Cu), and 30.0 g ofzinc sulfate (20% Zn) were placed in each hole. The followingfertilizers were broadcast 8 months after planting: urea, 45%N (equivalent to 696.9 kg ha-I N); ammonium sulfate, 20%N (equivalent to 464.4 kg ha-I N); and potassium chloride,60% K20 (equivalent to 696.1 kg ha-1 K20)-standard

© 2014 Lippincott Williams & Wilkins

Copyright © 2014 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.

www.soilsci.comI619

1'----- -

Moreira et al. Soil Science • Volume 178, Number 11, November 201 3

TABLE 2. Soil Chemieal Praperties ofaXanthie Ferralsol (Oxisal) at the Depths of O to 10, 11 ta 20, and 21 to 40 em in PrimaryFarest and Six Rubber Tree Plantations of Different Ages (Within 6, 16, 18, 19, 20, and 45 Years)

Depth pH P K Ca Mg AI H+Al

em Water - - - - mg kg-I ____ - - - - - - - - - - - - - - - - - - - crnol, kg-I ___________________

Forest0-10 4.1 1.7 42.0 0.10 0.17 2.1 9.911-20 4.1 1.1 27.0 0.17 0.12 1.6 7.521--40 4.4 0.8 14.8 0.06 0.08 1.1 7.0

Rubber tree plantation45 Years old

0-10 4.4 2.2 32.0 0.06 0.07 0.6 7.811-20 4.4 1.2 20.8 0.07 0.08 0.5 7.821--40 4.3 1.0 14.0 0.02 0.07 0.5 6.8

20 Years old0-10 4.1 6.2 22.8 0.02 0.07 1.7 9.411-20 4.0 4.7 14.6 0.01 0.03 1.2 9.321-40 4.1 3.0 9.4 0.01 0.02 1.0 8.1

19 Years old0-10 3.8 3.5 30.4 0.17 0.11 1.4 8.811-20 3.7 7.7 39.6 0.07 0.07 1.5 8.721--40 3.9 6.1 56.8 0.17 0.13 1.5 9.0

18 Years old0-10 4.1 1.1 12.2 0.04 0.04 1.1 4.911-20 3.9 1.8 17.6 0.03 0.02 1.1 5.321--40 3.9 1.9 31.0 0.06 0.05 1.2 5.5

16 Years old0-10 4.2 6.6 35.2 0.12 0.13 1.7 7.711-20 4.3 4.9 24.2 0.08 0.10 1.5 6.621--40 4.5 1.8 8.4 0.04 0.08 1.1 6.3

6 Years old0-10 4.0 1.8 20.0 0.06 0.06 1.2 6.511-20 3.7 5.4 26.0 0.05 0.04 1.3 6.921--40 3.8 3.7 41.8 0.10 0.07 1.6 6.4

Manaus, Amazonas State, Brazil."P and K, Mehlich 1extractant;Ca, Mg, and AI,KCI 1.0moIL-I;H +AI (potentiaIacidity), extractant SMP buffer soIution.The values correspond

to the mean of 10 subreplicates collected in each area.

P (KH2P04), which allowed the correction of the calculations.The determination was made by the difference between thesoil sample irradiated in microwave for 10 min and the nonirra-diated sample (control). For estimation of the P content in themicrobial biomass, a KEP correction factor of 0.40 was used(Brookes et aI., 1984).

Acid and alkaline phosphatase activities were deterrninedonly in the 0- to !O-cm depth soil samples in an 0.05-mol L-Ibuffer solution of p-nitrophenyl-phosphatase with pH 6.5 and11.0, respectively (Tabatabai, 1994), which is based on themeasurement of p-nitrophenol released by the action of thephosphatases afier soiI incubation in 0.05-moI L-I buffer solu-tion of p-nitrophenylphosphate. Ali microbiological analyseswere conducted with moist soil adjusted to 40% moisture. Theresults of PMB and acid and alkaline phosphatases are ex-pressed in dry soil at 105°C.

After microbiological and enzymatic analyses, the remain-ing soil sample was air-dried for determination ofpH (water), C(Walkley-Black), P and K (Mehlich I extractant - 0.05 moi L-IHCI + 0.025 mmol L-I H2S04), and Ca, Mg, and AI (extractant1.0 moi L-I) according to the methods described by Embrapa

620 I www.soilsci.com

(1997). Potential acidity (H + AI) was indirectly determined withSMP buffer solution (Moreira et aI., 2004). Total P and organic Pwere analyzed by the methodology described by Olsen andSommers (1982), in which the soil organic matter (SOM) isdestroyed by igniting the soil sample at 550°C. This rendersthe organic P acid-extractable organic P and the difference inthe acid extractable of an ignited and an unignited sample ofthe same soil as measures of the total organic P in the soil,whereas the available P was quantified by extraction withMehlich 1, Mehlich 3-0.2 moi L-I C2H402+ 0.25 moi L-INH4N03 + 0.015 moi L-I NH4F + 0.013 moi L-I HN03 +0.001 moi L-I EDTA solutions (Mehlich, 1984) and Bray I0.5 moi L-I HCI + 1.0 moi L-I NH4F (Olsen and Dean, 1976).

Ten samples of leaves were collected from the upper andmiddle thirds of rubber trees located in the forest and accordingto the established chronosequence to determine the P content inthese plants. Fine litter (without branches and fruits) was col-lected within a 1.5 x 1.00m rectangle, quickly washed with dis-tilled water, and dried in an oven at 70°C, reported as kg ha-I.The P content in the leaves and in the litter was determined ac-cording to the methodology described by Malavolta et aI. (1997).

Copyright © 2014 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.

© 2014 Lippincott Williams & Willdns

Soil Science • Volume 178, Number 11, November 2013 Phosphorus Dynamics

Data ProcessingData were analyzed by analysis ofvariance, F test, correla-

tion, and comparison for means using Scott and Knott test at 5%probability (Pimentel Gomes and Garcia, 2002). For statisticalanalyses, each sampled point was defined as a pseudoreplicate,and in biological analysis, each pseudoreplicate was consideredas the mean of two replicates.

RESULTS AND DISCUSSION

Soil C and PThe age of the rubber tree plantations caused changes in

the total organic carbon (TOC) , with an average increase inthese 1evels of 104.6% in the 45-year-old rubber trees comparedwith the 6-year-old rubber tree plantation, which correspondsto 68.9% of the TOC content in the primary forest (Table 3).Regarding depth, the TOC in the 0- to 10-cm layer variedfrom 16.9 g kg-I in the 6-year-old rubber trees to 41.3 g kg-I

in the primary forest, with an average decrease in chrono-sequences from 34.6% and 33.4% for the 11- to 20-cm and21- to 40-cm layers, respectively (Table 3). In the XanthicFerralsol of the central Amazon, developed on soft sedimen-tary material with the deep homogeneous clayey horizon, thetotal C concentration in the soil decreased near the 0- to 20-cmdepth, after which it stabilized and decreased slowly (Cerri andVolkoff, 1987).

Because of the weather conditions in the region, wherethere is continuous mineralization of OM, the formation ofthick layers of SOM occurs slowly, even in balanced ecosys-tems such as the primary forest. This negative gradient of TOCwith soil depth, with greater accumulation in the 0- to 10-cmdepth, was also reported by Moreira and Malavolta (2004) forcupuassu (Theobroma grandiflorum) and Moreira and Fageria(2011) in pastures (Urochroa brizantha) and citrus (Citrussinensis) under the same climatic and soil conditions, where asignificant reduction in TOC in the soil in areas c1eared forthe introduction of grazing and fruit plantations was ais o ob-served. These results demonstrate that, in the Amazon ecosys-tem, approximately 50% of the TOC of the soil, the main

source of nutrients, is present in the 0- to l O-cm depth, a sitewith a strong presence of Iitter.

Removal of the primary forest, which gave rise to second-ary forest pioneer species (Laetia procera, Vismia japurensis,and Bellucia grossularioides) well adapted to environmentslow in nutrients, and consequent formation of Iitter of lowernutritional quality (Silva et aI. 2006) have aiso contributed toa decrease in TOC in the early years of rubber cultivation. Amajor factor for the increase in TOe being directly proportionalto the age of the rubber trees is that the H. pauciflora and its hy-brid varieties are not deciduous trees as is H. brasiliensis (Pireset aI., 2002), with the recovery of litter in soil on a continuousbasis similarly to the primary forest, unlike the H. brasiliensiswhose defoliation occurs only in the driest period of the yearwithout continuous and permanent recovery of the SOM.Whereas the 6-month Iitter yielded from H. brasiliensis with15-year-old trees was 1.7 Mg ha-I dry weight (Murbach et a!.,2003), the lO-year-old H. Pauciflora litter accumulation was4.7 Mg ha-I year " (Moraes et aI. 2012). The TOC correlatedpositively with the acid and alkaline phosphatases and nega-tively with the availab1e P (Mehlich 3 and Bray I extractants)and the total P in the litter (Table 4). This finding corroboratesresults of Juma and Tabatabai (1978), who, in their study ofthe distribution ofthe phosphomonoesterases in soils, observedincreased phosphatase activity because ofthe greater accumula-tion of C and the inhibition of the enzymes caused by thegreater ofphosphate anions content in the soil.

Unlike what was observed by Conte et al. (2002), despitethe high levels of C in the soil (Table 3), the total organic andavailable P contents (Melhichl, Mehlich 3, and Bray I extract-ants) in the 0- to lO-cm depth were lower in the primary forestcompared with the rubber plantations (Table 5). This obser-vation is possibly caused by the application of 40 g plant " ofP205 in the plantation (Moraes et aI., 2008) that resulted in anincreased content ofthe nutrient in the soil. Despite the P fertil-ization, the content of P available in the soil remained withinthe ranges indicated for Mehlich I extractant-as low and verylow (Alvarez Venegas et aI., 1999). The small amount of totalP in the soils of the Amazon (Table 5) is the main cause oflow content of P available in the soil (Lehmann et aI., 200 I).

TABlE 3. TOC and PMB in a Xanthic Ferralsol (Oxisol) Under Different Vegetation Covers (Primary Forest and Six Rubber TreéPlantations Within 6, 16, 18, 19, 20, and 45 Years)

Vegetation Cover TOC PMB

0-10 em 11-20 em 21-40 em 0-10 em 11-20 em 21-40 em- - - - - - - - - - - - - - g kg -) - - - - - - - - - - - - - - -)-------------ilgg -------------

Primary forest 41.3aA 22.3aB 14.3aC 3045bA 0.56eB 0.79bBRubber tree p1antation

45 Years old 21.8eA 1604bB 12.6aB 2.98eA 0.96bB 0.62bB20 Years old 20.7eA 1O.3eB 7.9bB 3044bA 0.81bB 0.73bB19 Years old 23.5bA 22.2aA 8.6bB 4.55aA O.72eB 0.75bB18 Years old 26.8bA 13.6eB 9.7bB 4.04aA 1.32aB 1.04aB16 Years old 28.5bA 2004aA 12.0aB 2.85eA 1.08bB 0.68bB6 Years old 16.9dA 13.1eA 12.7aA 2.73eA 0.76eB Oo42eB

Mean 25.6 16.9 11.1 2.19 0.89 0.39S.D. 4.2 3.1 1.8 0.32 0.14 0.08CV% 1604 18.3 16.2 14.61 15.73 20.51

*Statistieal analysis considering ali treatments (n = 10). Means followedby different lowercase letters in the same column and uppercase letters inthe same line differ among themselves at 5% probability by Scott and Knott test.

CV: coefficient of variation.

© 2014 Lippincott Williams & Wilkins www.soilsci.comI621

Copyright © 2014 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.

Moreira et aI. Sai! Science • Volume 178, Number 11, November 2013

TABLE 4. Coefficient of Correlation Between the Chemieal and Biologieal Properties of Soil at the Depth of O to 10 em in AliVegetation Covers (Prirnary Forest and Six Rubber Tree Plantations Within 6, 16, 18, 19, 20, and 45 Years)

Variables C Total P Litter P Leaf P-ase alk P-ase Acid PMB P Total P Org P Bray 1 PM3

P-Ml 0.05NS 0.35* O.23NS -0.53* 0.12NS 0.02NS 0.S2* 0.76* O.3SNS 0.16NS

P-M3 -0.32* 0.21NS -0.45* -O.4S* -0.50* -0.2SNS 0.45* 0.53* 0.66*P-Bray 1 -0.51 * O.4S* -0.14NS -0.46* -0.45* -O.03NS 0.54* 0.59*P org -0.02NS 0.30* -0.04NS -0.39* -0.10NS -0.31* 0.96*P total -0.05NS 0.35* 0.05NS -0.41 * -O.OSNS -0.16NS

P-MB 0.06NS 0.30* 0.16NS O.OSNS 0.11NS

P-ase aci 0.45* 0.35* -0.04NS 0.5S*P-ase alk 0.66* -0.72* -0.19 s

P leaf 0.24NS -0.19NS

P litter -0.66*

Manaus, Amazonas State, Brazil.*Significant at 5% of probability.

S: not significant at 5% of probability; P-MI: available P-Mehlich I extractant; P-M3: available P-Mehlich 3 extractant; P-Bray I: availableP-Bray 1 extractant; P Org: P organic; P total: phosphorus total; P-MB: P microbial biomass; P-ase aci: acid phosphatase; P-ase alk: alkaline phos-phatase; P-Ieaf: P content in diagnosis leaf; P litter: P content in litter; total C: total carbono

The highest P recovery rates were observed in Mehlich3 and Bray I extractants compared with Mehlich I extractant(Table 5). The presence of ammonium f1uoride (NH4F) in thecomposition of the two extractants indicates that most of the Pextracted in the seven areas surveyed (rubber plantations andforest) is related to aluminum (P-Al), given the predominanceof kaolinite in the clay fraction in a Xanthic Ferralsol (Marqueset aI., 2010). Mehlich 3 extractant was found to be more ac-curate in determining the decrease in the availability of P insoil in the relationship to the chronosequence of the rubberplantations than were Mehlich I and Bray 1 extractants, whichranged from 11.36 mg kg -I (6 years old) to 1.03 mg kg-I

(45 years old), values similar to that of the primary forest-1.12 mg kg-I ofP (Table 3). These results are consistent withthe findings of Ye et aI. (20 11), who reported that the presenceof EDTA (ethylenediamine tetraacetic acid) in the extraction

solution increases the efficiency of recovery of the element insoils with high levels of OM.

Litter and Nutritional StatusShorrocks (1965) and Bi and Omont (1987) found that,

under the conditions of Malaysia and Iv01y Coast tropics,respectively, litter deposition under fully grown rubber trees(H. brasiliensis) varied from 5.0 to 7.0 Mg ha-I. These valuesare close to those found for rubber trees 16 years old or older(Table 6). In the primary forest, the average levels oflitter were4.7 Mg ha-I, far below the 8.2Mg ha-I year " reported byLuizão and Schubart (1987). However, the levels were similarto those found by Schroth et aI. (2002), who obtained valuesranging from 2.3 to 7.2 Mg ha-I for the accumulation oflitterduring a 7-year period in their assessment of the deposition of

TABLE 5. Total P, Organie P,and PAvailable With Mehlieh 1, Mehlieh 3, and Bray 1 Extraetants in a Xanthie Ferralsol (Oxisol) atthe Depth of O to 10 em

Total P Organic P P-Mehlich 1 P-Mehlich 3 P-Bray 1

Vegetation Cover - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -rng kg-I- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -

Primary forest 33.4b 19.5c 1.7b 1.1 c 2.5bRubber tree plantation

45 Years old 43.7b 35.4b 2.2b 1.0c 3.5b20 Years old 75.2a 45.4a 6.2a 1.5c 4.6b19 Years old 70.7a 47.3a 3.5b 7.9b 7.5a18 Years old 42.2b 25.3c 1.1 c 5.Sb 4.0bc16 Years old 65.4a 13.6c 6.6a 10.2a 7.4a

6 Years old 5S.4b 49.1a I.Sbc 11.4a 7.1aMean 55.6 33.7 3.3 5.5 5.2S.D. 12.33 6.14 0.91 1.29 1.22

CV% 22.2 18.2 27.8 23.4 22.1

Manaus, Amazonas State, Brazil.*Statistical analysis considering ali treatments (n = 10), and means followed by different letters in the same column differ among themselves at 5%

probability by Scott and Knott test.CV: coefficient of variation.

622 I www.soilsci.com © 2014 Lippincou Williams & Wilkins

Copyright © 2014 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.

I

Sai! Science • Volume 178, Number 11, November 2013 Phosphorus Dynamics

TABLE 6. Stable Accumulation of Litter, P Content in the Diagnosis Leaf and in the Litter, and Amount of P in the Litter of thePrimary Forest and Six Rubber Trees With Different Ages (Within 6, 16, 18, 19, 20, and 45 Years) Cultivated in a XanthicFerralsol (Oxisol)

Litter P Content Leaves P Content Litter Amount of P in the Litter

Vegetation Cover - - Mg ha-1 --_ __ _ __ _ _ _ _ _ _ _ g kg-1 _____________ ______ kg ha -1 ______

Primary forest 4.7a 2.35a 0.28b 1.36cRubber tree plantation

45 Years old 4.4ab 2.23a 0.51a 2.13b20 Years old 4.8a 2.25a 0.60a 2.45b19 Years old 4.8a l.92a 0.61a 2.52b18 Years old 4.7a 1.37b 0.53a 2.84b16 Years old 5.2a 1.45b 0.51a 3.17b6 Years old 3.9b 1.48b 0.49a l.35c

Mean 4.6 1.86 0.50 2.26S.D. 0.5 0.32 0.18 0.49CV% 10.9 17.2 36.0 21.7

Manaus, Amazonas State, Brazil.*Statisticalanalysis considering a11treatments (n = 10), and means fo11owedby different letters in the same column differ among themselves at 5%

probability by Snott and Knott test.CV: coefficient of variation.

litter in a secondary forest, agroforestry, and monoculture sys-tems with native species of the Amazon forest.

Total P contents in the litter were consistent with theamounts of OM in the soil (Table 6), and there were no signifi-cant differences in the total P content in the litter for the differ-ent chronosequences, but the values varied from 0.28 ± 0.01 inthe primary forest to 0.61 ± 0.02 g kg-I in the 19-year-old rub-ber plantation. Despite the statistical differences, the foliarcontents of 1.37 ± 0.14 of P in the 18-year-old rubber treesand of 2.35 ± 0.22 g kg-I in native rubber trees in the pri-mary forest (Table 6) were close to the content consideredsufficient by Moraes et aI. (20 11) in hybrid varieties ofH pauciflora (1.5 ± 0.2 g kg-I ofP) grown under similar cli-mate and soil conditions.

Other factors, such as rain, help maintain proper P contentin the leaves despite the low content of the nutrient in thelitter and in the soil of the primary forest. According to Jordan(1982) and Luizão (1989), in unchanged ecosystems such asthe Amazon, water from rainfall is another important sourceof P for plants, with an availability of 23.30 ± 12.58 kg ha-Iyear-I of PO/- -P through the atrnosphere. Pushparajah(1979) states that rubber tree branches accumulate a largeamount of P, so the amount of P that is retumed out of fallingbranches must therefore be significant. Pushparajah (1979) ais onoted that rainfall provides the soi! with a considerable amountofthe P contained in the leaves. Annual precipitation, approxi-mately 2,450 mm, is responsible for retuming 0.5 kg ha-I P tothe soi! from adult plant leaves. ln others conditions, Grahamand Duce (1979) calculated deposition rates to be approxi-mately 320 x 10 g year-I of Ponto the continents andapproximately 140 x 10 g year -I of Ponto the surface ofthe oceano

Phosphorus of the Microbial Biomass and Acidand Alkaline Phosphatase

The plantation age did not significantly affect PMB at anyof the depths sampled (Table 3). The greatest changes in thePMB, in relation to the vegetation cover (Table 3), occurred at

e 2014 Lippincott Williams & Willdns

the 0- to 10-cm depth. ln this soil surface layer, PMB con-tent progressively decreased to the 21- to 40-cm depth, with areduction of 74.1 % of the PMB from the depth of O to 10 emto a depth of l l to 20 em and a nonsignificant decrease (atP ::::;0.05) of 12.7% from the depth of l l to 20 em to 21 to40 em. At the depth of O to 10 em, PMB content varied from2.98 ± 0.37 to 4.55 ± 1.01 ug g-I (Table 3), which is compara-ble to the range of2.18 to 5.461lg g-I reported by Moreira et aI.(2011) in different plant covers of the Westem Amazon. Thedifferences observed at the depth of O to 10 em within thechronosequence of the rubber trees also showed that, even un-der the same climate and soil conditions, PMB is sensitive tochanges in soil management, showing significant correlationonly with the P content in the litter (Table 4).

On average, PMB content accounted for 10% of P0' whichis higher than levels reported by Conte et aI. (2002) under directseeding, and also higher than the Mehlich I-extractable P con-tent in the soil (Table 5). Moreira and Fageria (2011) found thatthe P immobilized by the microbial biomass in the Amazon inmost areas sampled was higher than the P available for plants;that is, the availability ofP in the soi! under unchanged systemsoccurs to a large extent through microbial activity with the re-lease of phosphatases, transforming organic Pinto inorganicP, conditioning the maintenance of the nutritional status of theplants to P, cycling, mostly through litter (Conte et aI., 2002;Oliveira et aI., 20 11).

ln the primary forest soil, acid phosphatase activity (APAc)was 164.45 ± 8.57 ug of p-nitrophenol h-I g-I of soi!, whichwas 20.4% more than that of the 6-year-old rubber planta-tion (Fig. I). This value was lower than that reported by Mendeset al. (2012) for soi! in the Cerrado region (Oxisol), rangingfrom 739 to 1,892 ug of p-nitrophenol h-I g-I of soil. Despitethe small but insignificant increase of the acid phosphatase ac-tivity reported in the 16-year-old rubber plantation, phosphataseactivity was not significantly changed during the years (Fig. I),even with a decreasing trend observed in linear regressionequation y (year of rubber plantation) = 153.46 - 2.24APAc,P > 0.05. Nevertheless, the slight increase observed was possi-bly caused by the large need for P of phosphatase extruders

Copyrighl © 2014 Lippincott Williams & Wilkins. Unaulhorized reproduclion of this article is prohibiled.

www.soilsci.comI623

I

Moreira et ai. Soil Science • Volume 178, Number 11, November 2013

~ Acid phosphatase200

o Alkaline phosphatase

aA at!- aa a a aB

B

B

B

m B

t B

t- -

160

40

o20 18 16Forest 45 19 6

Age of rubber tree plantation in years

FIG. 1. Acid and alkaline phosphatase aetivities in the primaryforest and rubber tree plantations within different periods in aXanthie Ferralsol (Oxisol). The values correspond to the meanof the samples eolleeted at the depth of O to 10 em. Meansfollowed by different lowerease and upperease letters in thesame treatment differ from eaeh other at 5% probability bySeott and Knott test.

(microbial biomass and plants) because of the decreased avail-ability of P in the soil that occurs as the rubber trees growolder and there is a stabilization of TOC, resulting in a sys-tem similar to the primary forest (Tables 3, 5). This observa-tion corroborates the results obtained by Olander and Vitousek(2000). As expected, the acid phosphatase activity correlatednegatively with the P content in the litter, indicating stabilityof enzyme activity within the process of P cycling in the soil-system plant.

Variation in alkaline phosphatase activity (APAI) that iscaused by fungal and bacterial activity (Dakora and Phillips,2002) was similar to that of acid phosphatase activity, withhigher levels in the primary forest (163.30 ± 15.55 /Jg ofp-nitrophenol h-I g-I of soil) and lower Jevels in the 6-year-old rubber trees (63.18 ± 10.81 ug of p-nitrophenol h-I g-I ofsoil) (Fig. I), with significant correlation between year of rub-ber tree growth versus APAI (Y (year of rubber plantation) =116.82 - 8.26APAI, P :::;0.05). It was also found that, after45 years of cultivation, the rubber tree plantation showed values70.8% higher than those of the primary forest, indicating apossible stabilization of P dynamics in the microbial biomassin this ecosystem (Table 3). Similar results on the behavior ofacid and aIkaline phosphatase activities were also obtained byCosta and Lovato (2004) in studies of different soil manage-ment systems.

The alkaline phosphatase activity correlated negativelywith the available P (Mehlich 1, Mehlich 3, and Bray 1 ex-tractants), organic P, total P, and P in the litter and positivelywith organic C (TOC) and acid phosphatase (Table 4), sup-porting the postulation that the activity of phosphatases in-creases with a decrease in P content in the soi!. The aIkalinephosphatase activity did not correlate significantly with P inthe leaves (Table 4), in contrast to the results of Costa andLovato (2004), indicating that P content in plant tissue evi-denced the influence of phosphatase activity on the flow and

624 I www.soilsei.eomCopyright © 2014 Lippincolt Williams & Wilkins. Unauthorized reproduction of this article is prohibited.

© 2014 Lippincott Williams & Wilkins

types of P in the soi!. This has also been found for PMB(Table 5) probably because of the low content of the nutrientin the soil, regardless of the vegetation cover (Tables 2, 5).

CONCLUSIONSIn the humid tropical Amazon, nutrient-poor soils are com-

mon, with the low P content being of greatest concern. An op-tion for recovering cleared forest areas and minimizing theimpact of nutrient dynamics in this ecosystem could be theplanting of rubber trees (Hevea species), a native tree adaptedto the local climate and soil conditions. In addition, the higheconomic value of natural rubber is a source of income forsmall producers, which shows that the conversion of abandonedor/and deforested land to rubber tree plantations is a viable eco-nomic and environmental strategy for the Amazon basin. There-fore, it is important to understand the influence ofthe biologicalindicators associated with the process ofP cycling. The refores-tation of the forest in deserted areas with rubber plantationsgives rise to secondary forest with increased TOC content,which was 68.9% higher than the TOC in the primary forest.Similar increases in acid and alkaline phosphatases activitieswere also observed. Because of the decrease in the amount oflitter and P available in the soil, the amount of P stored in themicrobial biomass decreased. The increased P levels were con-centrated in the rubber plantations. Regardless ofthe vegetationcover, the P immobilized in the microbial biomass was the mainreserve.of the nutrient in the studied chronosequences. At thedepth of O to 10 em, Mehlich 3 and Bray I extractants weremore accurate in determining P availability in soil in kaoliniticsoils, with most P bound to AI than Mehlich I extractant.

REFERENCESAjayi A. E., M. S. Dias Junior, N. Curi, C. F Araujo Junior, T. T. T. Souza,

and A. N. TndaJunior. 2009. Stretch properties and compaction suscep-tibility ofBrazilian Latosols. Soil Till. Res. 105: 122-127.

Alvarez Venegas V. H., R. F. Novais, N. F. Barros, R. B. Catarutti, andA. S. Lopes. 1999. Interpretation of soil analysis results. In: Recom-mendations for Use of Amendments and Fertilizers in Minas Gerais,5th Approximation. A. C. Ribeiro, P T. G. Guimarães, and V.H. AlvarezVenegas (eds.). Soil Fertility Commission for the State of Minas Gerais.Viçosa, Brazil, pp. 25-32.

Andreux F., and C. C. Cerri. 1989. Current trends in the research on soilchanges due to deforestation, buming and cultivation in the Braziliantropics. Toxico!. Environ. Chem. 20/22:275-283.

Antonio I. C. 2009. Agrometeorological Bulletin-2008. Embrapa WesternAmazon, Manaus, Amazonas State, Brazil, p. 26.

Bi T. T., and H. Omont. 1987. Étude des sols de parcelles dhévéa en BasseCote dIvoire. Rev. Genér. Caout. Plast. 674:111-116.

Brookes PC., D. S. Powlson, and D. S. Jenkinson. 1982. Measurement ofmicrobial biomass phosphorus in soi!. Soil Biol. Biochem. 14:319-329.

Brookes P.c., D. S. Powlson, and D. S. Jenkinson. 1984. Phosphorus in thesoil microbial biomass. Soil Biol. Biochem. 16:169-175.

Carneiro R. G., I. C. Mendes, P. E. Lovato, A. M. Carvalho, and L. J.Vivaldi. 2004. Soil biological indicators associatted to the phosphoruscycle in a Cerrado soil under no-till and conventional tillage systems.Pesq. Agropec. Bras. 39:661-669.

Cerri C. C., and B. Volkoíf 1987. Carbon content in a yellow Latosol ofcentral Amazon rain forest. Acta Geolog. 8:29-42.

Conte E., l. Anghinoni, and D. S. Rheinheirner. 2002. Phosphorus in themicrobial biomass and acid phosphatase activity by phosphates appli-cation in soi! under no tillage system. R. Bras. Cio Solo. 26:925-930.

I

Soii Science • Volume 178, Number 11, November 2013 Phosphorus Dynamics

Costa M. D., and P. E. Lovato. 2004. Phosphatase activity in soil undermycorrhizal and non-mycorrhizal cover crops. Pesq. Agropec. Bras.39:603-605.

Cravo M. S., and T. J. Smyth. 1997. Soi! fertility management for sustainablecropping on an Oxisol ofthe Central Amazon. R. Bras. CioSolo 21:607-616.

Oakora F. O., and O. A. Phillips. 2002. Root exudates as mediators of min-eral acquisition in low-nutrient environments. Plant SoiI245:35--47.

Oick W. A., and M. A. Tabatabai. 1993. Significance and potential use ofsoi! enzymes. In: Soil Microbial Ecology Applications in Agriculturaland Environmental Management. F. B. Metting Junior (ed.). M. Oekker,New York, NY, pp. 95-124.

EMBRAPA (Empresa Brasileira de Pesquisa Agropecuária). 1997. Manualfor Methods of Soil Analysis, 2nd ed. National Research Center forSoils, Rio de Janeiro, Brazil, p. 212.

EMBRAPA (Empresa Brasileira de Pesquisa Agropecuária). 1999. BrazilianSystem of Soil Classification. SPIIEmbrapa Production of Infonnation,Brasília, Brazil, p. 306.

FAO. 1990. Soil map ofthe world, rev. legend. Unesco, Paris, France, p. 59.

Ferreira A. S., F. A. O. Camargo, and C. Vidor. 1999. Use ofmicrowave ra-diation to evaluate soil microbial biomass. R. Bras. CioSolo 23:991-996.

Graham W. F., and R. A Duce. 1979. Atmospheric pathways ofthe phospho-rus cycle. Geoch. Cosmoch. Acta. 43:1195-1208.

Jordan C. F. 1982. The nutrient balance of an Amazonian rain forest.Ecology 63:647-654.

Juma N. G., and M. A. Tabatabai 1978. Oistribution of phosphomono-esterases in soils. Soil Sei, 126:101-108.

Lehmann J., M. S. Cravo, J. L. V Macêdo, A. Moreira, and G. Schroth.2001. Phosphorus management for perennial crops in central Amazonianupland soils. Plant Soi! 237:309-319.

Luizão F. J. 1989. Litter production and mineral element input to the forestfloor in a central Amazonian forest. GeoJournal 19:407--417.

Luizão F. J., and H. O. R. Schubart. 1987. Litter production and decomposi-tion in a terra-firme forest ofCentral Amazonia. Experientia 43:259-265.

Malavolta E., G. C. Vitti, and S. A. Oliveira. 1997. Evaluation ofnutritionalstatus ofplants. Potafos, Piracicaba, Brazil, p. 319.

Mehlich A. 1984. Mehlich 3 soil test extractant: A modification ofMehlich2 extractant. Commun. Soil Sci. Plant Ana!. 15:1409-1416.

Mendonça E. S., and E. S. Matos. 2005. Soi! organic matter: Methods ofanalysis. Federal University ofViçosa, Viçosa, Brazil, p. 107.

Marques J. O. O., W. G. Teixeira, A. M. Reis, O. F. Cruz Júnior, S. M.Batista, and M. A. C. B. Afonso. 2010. Chemical, physical and hydricattributes and mineralogy of clay fractions in soils of lower Amazon:Serra de Parintins. Acta Amazon. 40:1-12.

Mendes I. c., M. F. Fernandes, G. M. Chaer, and F. B. Reis Junior. 2012.Biological functioning of Brazilian Cerrado soils under different vegeta-tion types. Plant Soil 359:183-195.

Moraes J. F. L., B. Volkoff, C. C. Cerri, and M. Bernoux. 1996. Soil prop-erties under Amazon Forest and changes due to pastures installation inRondônia, Brazi!. Geoderma 70:63-81.

Moraes L. A. C., V H. F. Moraes, F. E. L. Magalhães, and A. Moreira.2012. Fluxes emission of leaves, flowering and nutrient cycling in crownclones of Hevea pauciflora. Acta Amaz. 42: 399--404.

Moraes L. A. c., A. Moreira, J. R. A. Fontes, E. R. Cordeiro, and V H. F.Moraes. 2011. Assessment of rubber tree panels under crowns resistantto South American leaf blight. Pesq. Agropec. Bras. 46:466-473.

Moraes VH.F., L.A.C. Moraes, and A. Moreira. 2008. Cultivo da seringueiracom copas enxertadas resistentes ao mal-das-folhas. Embrapa WesternAmazon, Manaus, Amazonas State, Brazil, p. 44.

© 2014 Lippincott Williams & Wilkins

Moreira A., M. P. Almeida, O. G. Costa, and L. S. Santos. 2004. Potentialacidity by pH SMP method in Amazonas State, Brazi!. Pesq. Agropec.Bras. 39:89-92.

Moreira A., and N. K Fageria. 2009. Soil chemical attributes of AmazonasState, Brazi!. Commun. Soil Sci. Plant Ana!. 40:2912-2925.

Moreira A., and N. K. Fageria. 2011. Changes in soil properties under twodifferent management systems in the Western Amazon. Commun. SoilSci. Plant Ana!. 42:2666-2681.

Moreira A., N. K. Fageria, and A. Garcia y Garcia. 2011. Soil fertility,mineral nitrogen, and microbial biomass in upland soils of the CentralAmazon under different plant covers. Commun. Soil Sei, Plant Ana!.42:694-705.

Murbach M. R., A. E. Boaretto, T. Muraoka, and E. C. A Souza. 2003. Nu-trient cycling in a RRIM 600 clone rubber plantation. Sei, Agric.60:353-357.

Oliveira L. B., T. Tiecher, F. L. F. Quadros, and O. R. Santos. 2011. Effectsof burning and grazing on microbial phosphorus in natural grasslandsoils. R. Bras. Cio Solo 35:1509-1515.

Olsen S. R., and L. A. Dean.1976. Phosphorus. In: Methods of Soil Analy-siso C. A. Black (ed.). Soil Science Society of America, Madison, WI,pp. 1035-1049.

Olsen S. R., and L. E. Sommer. 1982. Phosphorus. In: Methods of SoilAnalysis. R. H. Miller, and D. R. Kenney (eds.). Soil Science Societyof America, Madison, WI, pp. 403--430.

Olander L. P., and P M. Vitousek. 2000. Regulation of soi! phospha-tase and chitinase activity by N and P availability. Biogeochemistry49:175-191.

Pimentel Gomes F., and C. H. Garcia. 2002. Applied Statistics to Elemen-tary Agronomy and Forestry. FEALQ, Piracicaba, Brazil, p. 309.

Pires J.M., R. S. Secco, and J. I. Gomes. 2002. Taxonomia e fitogeografiadas seringueiras Hevea spp. Embrapa Amazônia Oriental, Belém,Brazil, p. 103.

Pushparajah E. 1979. Nutrient cycle in rubber plantations. In: TrainingManual on Soil, Soil Management and Nutrition of Hevea. RubberResearch lnstitute ofMalaysia (ed.). RRIM, Kuala Lumpur, Malaysia,pp. 88-97.

Schroth G., S. A. D'Angelo, S. A. W. G. Teixeira, O. Haag, and R.Lieberei. 2002. Conversion of secondary forest in agroforestry andmonoculture plantation in Amazonia: Consequences for biomass litterand soil carbon stocks afier 7 years. For. Ecol, Manage. 163:131-150.

Shorrocks V M. 1965. Mineral nutrition, growth and nutrient cycle ofHevea brasiliensis. Il. Nutrient cycle and fertilizer requirements. J.Rubber Res. Inst. Malaya 19:48-61.

Silva C. E. M., J. F. C. Gonçalves, T. R. Feldpausch, F. J. Luizão, R. R.Morais, and G. O. Ribeiro. 2006. Nutrient use efficiency for pioneerspecies grown on abandoned pasture in central Amazonia. Acta Amaz,36:503-512.

Souza C. R., C. P. Azevedo, R. M. Lima, and L. M. B. Rossi. 2010. Forestspecies behavior in plantations at full sun and secondary forest enrich-ment strips in the Amazon. Acta Amaz. 40:127-134.

Tabatabai M. A. 1994. Soil enzymes. In: Methods of Soil Analysis,Microbiological and Biochemical Properties. R. W. Weaver, S. Angle,P. Bottomley, O. Bezdicek, S. Smith, A. Tabatabai, and A. Wollum(eds.). Soil Science Society of America, Madison, WI, pp. 775-833.

Tian G. 1998. Effect of soil degradation on leaf decomposition and nutrientrelease under humid tropical conditions. Soil Sci. 163:897-906.

Ye R., A. L. Wright, and J. M. McCray. 2011. Seasonal changes in nutrientavailability for sulfur-amended everglades soils under sugarcane. J. PlanlNutr. 34:2095-2113, 2011.

Copyright © 2014 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.

www.soilsci.comI625