1 Introduction

Restoration of degraded Atlantic rainforest sites is often slowed by poor physical, chemical, and biological properties of soils resultant from previous intensive uses, such as grazing and agriculture (Bezerra et al. 2006). Application of organic amendments such as biosolids (i.e., treated municipal sewage sludge) can potentially restore soil properties associated with native forests more quickly than natural ecosystem processes of organic matter and nutrient accumulation over time. Biosolid application for native forest restoration could also potentially utilize organic waste that otherwise might require an alternate disposal method with negative environmental impacts and high cost. Avoiding the use of mineral fertilizers also conserves energy and natural resources (Henry et al. 1993; Henry et al. 1994; Harrison et al. 1994; Andreoli et al. 1999). Despite the potential benefits of the addition of organic matter and nutrients from biosolids to soil, contaminants, including trace metals, may cause problems related to plant toxicity and environmental contamination (Brallier et al. 1996; Dongsen et al. 1992; Yang et al. 1998; Bettiol and Camargo 2000).

Much of the degraded land in developing countries such as Brazil is located in tropical regions, and studies of the interaction of biosolids with tropical soils are few in number. Understanding the potential negative impact of biosolid application would ideally utilize long-term studies, since some negative mechanisms may require time to be realized and detectable (Tsutiya et al. 2002; Harrison et al. 1994; Harrison et al. 1996). Generally, movement of metals resulting from biosolid application is limited and primarily observed in studies where acidic sandy soils with low organic matter content were treated with high rates of biosolids (Hue 1995; Brallier et al. 1996). In these cases, movement of metals into the soil profile was relatively rapid in the first few years, and did not increase greatly with additional time periods up to 20 years.

Plant species tolerant to high rates of biosolid application generally accumulate trace metals in roots rather than in the aerial portion (Marques 2000). and in most cases where toxicity was noted, accumulation occurred relatively quickly after biosolid application. In Brazil, the few studies that have utilized biosolids as a soil amendment have shown promising results both in use for seedling production (Trigueiro and Guerrini 2003) and in soil improvement for reforestation (Gonçalves et al. 2000; Poggiani et al. 2000). In those studies, results were similar to the more abundant research performed in temperate regions; however, the potential for the migration of metals through the soil profile was not investigated. The purpose of this study was to investigate the potential for tree uptake and mobililization and movement of trace metals in the soil profile of a degraded Atlantic rainforest soil during the first year after application of municipal biosolids.

2 Material and Methods

2.1 Site Description

The experiment was conducted at the Entre-Rios Farm of the Suzano Bahia Sul Papel e Celulose Company, in the region of Itatinga, São Paulo, Brazil, at approximately 23.3° S, 48.5° W, where the average elevation is 636 m. The climate of this region is warm and temperate with average annual temperature of 19.1 °C and average annual rainfall of 1324 mm. The primary wet season ranges from October to March, with a drier season April to September. The variation in the precipitation between the driest and wettest months is 195 mm. The variation in temperatures throughout the year is 6.4 °C with the mean monthly temperature ranging from 22.0 °C in January to 15.6 °C in July. The area is characterized by relatively level, degraded soils (highly-compacted without topsoil), due to previous use for agriculture and timber storage. The soil of the experimental area is classified as a typic Quartzipsamment, according to the Brazilian and American Systems of Soil Classification (EMBRAPA 2006).

2.2 Experimental design

The field experimental design utilized four replicated blocks with six plots each, and one biosolid application rate for each plot (0 “control,” 2.5, 5, 10, 15, and 20 dry Mg ha−1). Biosolids from the sewage treatment plant of the city of Jundiaí, São Paulo State, were utilized for this study. Jundiaí treats wastewater initially in completely mixed aerated lagoons, followed by further treatment in a settling pond. Table 1 shows the metal composition of the Jundiaí biosolids compared to regulatory levels (NRC 1996) in Brazil and the USA and a range of biosolids and soils from the USA. Note that the maximum permitted concentration (MPC) in Brazil for the metals is the same as the level for class A (“exceptional quality”) biosolids in the USA, which indicates a lower legal application rate of metals in Brazil than the national standards for the USA (US EPA 2000). The class A level was developed by the US EPA to identify biosolids that are unlikely to cause problems when applied to land for the metals under consideration. Unlike Brazil, US law provides for application of biosolids at the higher concentration levels noted with additional restrictions.

Table 1 Composition of biosolids compared to Brazilian and US standards
Full size table

Each plot received a different level of biosolids, assigned randomly. Each plot was split into a 9 × 9 grid, for a total of 81 subplots (Fig. 1). The border row was planted with Cytharexyllum myrianthum (Pau-viola) forming a treated border buffer between plants. The buffer trees were not analyzed. Each of the 49 internal subplots was planted with one of nine species of native trees, spaced at 3.0 × 2.0 m between plants for a total area of 486 m2 per plot. Six application levels (treatments) per block and four blocks yield a total study area of 1.17 ha. The number of plants per group in each plot included eight pioneer, six secondary, and three climax species.

Fig. 1
figure 1

Distribution of plant species in each of six plots within a block

Full size image

Tree species were selected based on their ecological and silvicultural importance to the Atlantic forest region, as well as being relatively common in native riparian forests. The Brazilian Atlantic Rainforest is considered to be one of, if not the most, biodiverse forest in the world in terms of the variety of tree species, so these nine species certainly cannot represent the natural biodiversity of a typical native rainforest, but this is among the first studies using these native species coupled with biosolid amendment of degraded soil. The species used were from three groups: pioneer, secondary, or climax. Pioneer species included capixingui (Croton floribundus Spreng) and aroeira-pimenteria (Schinus terebinthifolius Raddi). The secondary tree species used were canafístula (Peltophorum dubium (Spreng.) Taub.), cedro-rosa (Cedrela fissilis Vell.), mutamba (Guazuma ulmifolia Lam), and angico-vermelho (Anadenanthera macrocarpa (Benth.) Brenan). Three climax tree species were also planted: copaíba (Copaifera langsdorffii Desf.), jatobá (Hymenaea courbaril L.), and jequitibá (Cariniana estrellensis (Raddi) Kuntze). All plots were planted using the same pattern (Fig. 1) where pioneer, secondary, and climax species were alternated, and each subplot had one tree planted in its center.

The experiment started in March 2005, by ground clearing to remove invasive Brachiaria (signalgrass) vegetation and plowing to a depth of 40–50 cm to ameliorate previous soil compaction. Biosolids were applied in June 2005, and immediately incorporated into the soil by tillage to 20-cm total depth. Tree seedlings were planted in August 2005.

2.3 Soil Analysis

Composite soil samples of 24 individual samples within each plot were collected uniformly from 0- to 20-cm (same as incorporation) depths 6 and 12 months after the biosolid application. Total arsenic (As), cadmium (Cd), chromium (Cr), lead (Pb), and nickel (Ni) were analyzed with digestion according to EPA method 3051 (US EPA 1994). and atomic absorption spectrophotometry (Perkin Elmer, Model 1100B, Perkin Elmer Corporation, 520 South Main St, Akron, OH 44311). To calculate the means for each treatment level at month 12, we used values of 249 μg kg−1 soil for nondetectable Pb (<250 μg kg−1 soil), calculated from the detection limits of the procedures and equipment utilized. Total As, Cd, Cr, and Ni were above detection levels and were quantified. Soil extractable As and Cd concentrations were also determined by extraction of soil with a 1 N HCl and atomic absorption spectroscopy (Tedesco et al. 1995).

2.4 Foliage Analysis

For each species, 6 and 12 months after planting, recently matured leaves were collected from the upper third of the tree crowns in the four cardinal directions, avoiding sampling both very young and senescent leaves. Foliage of all trees of the same species were mixed and analyzed together (giving one observation per species per plot). Foliage was dried at 65 °C in forced air ovens and ground in a Wiley-type mill to 0.15 mm. Chemical analysis of As, Cd, Cr, Pb, and Ni was performed by nitric-perchloric digestion according Malavolta et al. (1997). with determination of solution metals by atomic absorption spectrophotometry.

2.5 Statistical Analysis

Soil concentrations of heavy metals over time (6 and 12 months) and for treatment levels (0, 2.5, 5, 10, 15, and 20 Mg ha−1 biosolids) was analyzed using repeated measures analysis of variance (ANOVA; SPSS 2012). including an interaction of time by treatment. Similarly, foliar concentration of heavy metals over time, according to treatment level and species was analyzed through repeated measure ANOVA including all interactions between time, treatment, and species. We only used pioneer and secondary species in the foliar analysis since there was not enough material to be sent to the laboratory for analysis of climax species (too little foliage in young plants at 6 months due to slow growth and lack of foliage because of senescence during the dry season at 12 months).

3 Results and Discussion

The biosolids used in this study contained trace metal concentrations below the standards established by the Conselho Nacional do Meio Ambiente (CONAMA 2006) and the Companhia Ambiental do Estado de São Paulo (CETESB 1999). which regulates the use of biosolids in agriculture in the Brazilian state of São Paulo (Table 1). According to Marques et al. (2006). even though the levels of trace metals found in biosolids were below both the legal maximum permitted concentration and the US EPA class A quality guidelines, studies on the application of different doses of this residue to the soil are still critical because metals present in biosolids that are applied to soil will be added either to metal existing naturally in the environment (Andreoli et al. 2006) or to metal from other anthropogenic sources, including industrial emissions, effluents, fertilizers, soil conditioners, and pesticides (Silveira 2003). Based on the highest dose of biosolids applied to the soil (20 Mg ha−1) and a measured bulk density of 1000 kg m3, the added metal quantities were approximately 0.002, 0.16, 3.30, 0.76, and 3.90 kg ha−1 of As, Cd, Cr, Ni, and Pb, respectively (Table 1), with correspondingly lower loading rates at the lower application rates.

3.1 Addition of Trace Metals to Soil

Soils were analyzed 12 months after the biosolid application for trace metal concentrations (Table 2). In all cases, the high variability of metals in soil analysis did not show a uniform and statistically significant increase in metal concentration vs. rate of biosolids applied. Due to the relatively low metal concentrations of the biosolids (Table 1), low application rates and metal loading relative to background metal concentrations, and operational difficulty in thoroughly mixing biosolids with soil such small changes in biosolid-amended soils compared to the control soil are not completely unexpected (Hamel et al. 2003).

Table 2 Total As, Cd, Cr, Ni, and Pb concentrations in soil 12 months after biosolid application
Full size table

The treatment that best followed the predicted increase in soil As was the 10 Mg ha−1 application rate; all other treatments resulted in a lower observed change than expected except in the case of Cd (Table 2). The 10 Mg ha−1 application rate also showed the greatest increase in Cr and Ni over the other application rates. Arsenic was the only trace metal that showed a trend of increasing concentration with application rate.

The actual increase in Cd due to the biosolid treatment were much higher than what was predicted from the application rate (Table 2); observed increases in soil Cd were between 2 and 13 times higher than predicted, and the highest increases were seen in the lower application rate treatments. This could be due to nonuniform mixing of biosolids into the soil profile. However, the increase was not directly linear with the application rate of biosolids, and it is not known with certainty why we observed these unexpected results.

The estimated amounts of Pb added would result in increases and total soil concentrations below the detection limit of Pb for the method used (2500 μg kg−1), so little could be inferred related to the impact of the application of biosolids on soil Pb. It has been noted in other studies that EPA 3051 sometimes does not extract all of total soil Pb (Hamel et al. 2003).

3.2 Extractable Soil Trace Metals

Since much of the concentration of As, Cd, Cr, Ni, and Pb is expected to be bound to soil solids and relatively unavailable to plants (Brallier et al. 1996). soil samples were also analyzed at 6 and 12 months after biosolid application for extractable metals. Extractable As and Cd were the only metals that had concentrations above the detection limits of the procedures and equipment used (Table 3). Extractable Cr, Ni, and Pb were below detection limits of 3000, 3000, and 25,000 μg kg−1, respectively, and are not shown. Extractable As and Cd increased in the 2.5–10 Mg ha−1 application rates from 6 to 12 months, while extractable metals decreased in the 15–20 Mg ha−1 application rates over the same time period. A repeated measures ANOVA showed that neither the time (6 vs. 12 months), nor the treatment level were statistically significant in the model, which might be due to the large variation among plots (Fig. 2). Though surface biosolid applications and tillage to 20 cm were made as uniformly as possible, it is not uncommon for high variation in subsequent soil sampling to be observed (Brallier et al. 1996).

Table 3 Mean extractable As and Cd concentrations in soil 6 and 12 months after biosolid application
Full size table
Fig. 2
figure 2

Extractable As (a) and Cr (b) according to block (site) at 6 and 12 months

Full size image

Soil extractable Cd followed a very similar trend with biosolid application rate (Table 3). For instance, available Cd measured 6 months after biosolid application increased from 10 to 330 μg kg−1 from the 0 to the 20 Mg ha−1 rates, respectively. A repeated measures ANOVA showed a statistically significant effect of treatment (p = 0.001), but no significant effect of time or the interaction time by treatment. As with As, the potential for Cd leaching out of soil was not investigated, though research in similar soils has shown low potential for leaching (Mattos et al. 1996).

3.3 Trace Metal Uptake in Foliage

Foliar uptake of metals and the potential for metals entering the food chain of grazing animals was investigated by sampling foliage at 6 and 12 months following biosolid application. Climax species (Jequitibá-branco, jattobá, and copaiba) were excluded from this analysis due to excessive number of missing values due to slow growth and small amounts of foliage early, and senescence during the dry season.

No interactions between treatment, species, and time and foliar uptake were statistically significant (Table 4). Only time (6 and 12 months) was statistically significant for foliar As (p = 0.001) and Pb (p = 0.03). In general, the level of foliar concentration of heavy metals decreased over time. Some possibilities for the observed results of very low foliar metals is that plants accumulated the available metals in their root systems (Marques 2000). that metals were diluted with uptake by the high production of organic matter or there was low metal availability in the soil.

Table 4 Foliar As, Cd, Cr, Ni, and Pb concentrations 6 and 12 months after biosolid application (including pioneer and secondary species and excluding climax tree species)
Full size table

Oliveira and Mattiazzo (2001) found that trace metal uptake in foliage by sugar cane in an Oxisol treated with biosolids was very low or undetectable, though they did observe increases with higher biosolid application rates in soil. Similarly, Velasco-Molina et al. (2006) did not observe Cd and Cr foliar uptake at biosolid application rates as high as 40 Mg ha−1. These results are also consistent with other studies of metal uptake by Brallier et al. (1996). where soluble and extractable trace metal concentrations and trace metal uptake was very low, even though soil total metal concentrations were higher. These results suggest that there will be no effect of similar biosolid applications on foliar uptake by trees and therefore little influence on grazing animals.

4 Conclusions

In this study, trace metal contamination of soil and trees by biosolid applications appears to be negligible when applied up to 20 Mg ha−1 due to the low concentrations of metals, relatively low application rates, and the strong binding of metals in soil. The results of this study show that biosolids can likely be utilized in an environmentally safe manner with little chance of metals becoming highly bioavailable, at least under the conditions utilized in this study.