1 Introduction

Reusing organic waste as a soil fertilizer offers a number of advantages over other management alternatives because it reduces the use of other fertilizers and eliminates the necessity of its subsequent treatment or disposal (Bruun et al. 2006; Hargreaves et al. 2008). Sewage sludge and manure are the most common organic wastes applied either raw or composted (i.e., humification of the organic matter under controlled conditions). The application of such wastes to soil provides nutrients, increases organic matter, improves soil structure, and enhances nutrient absorption by plants (Weber et al. 2007; Singh and Agrawal 2008). Therefore, the use of different types of organic waste in agriculture or farming activities instead of using conventional chemical fertilizers should be preferred in terms of sustainability. These residues can also be used as amendments to regenerate infertile soils and for improving plant cover (Soliva and Paulet 2001).

However, the European legislation has become more restrictive on the content of priority pollutants in residues that are used as raw materials for the production of fertilizers or as fertilizers themselves (European Commission 2004), ultimately limiting waste reuse in agriculture. Currently, there are several types of organic waste and compost, classified according to the origin of its raw materials (European Community 2006): urban residues, agricultural and forest residues, wastewater treatment sludge, residues resulting from terrestrial remediation activities, residues from industrial processes, and mixtures of these. Depending on the raw material, toxicity due to the presence of persistent organic pollutants or heavy metals may become important (Hua et al. 2008; Oleszczuk 2008). The application of organic waste (i.e., compost, sludge or manure) to land, especially agricultural crops, represents a significant input of nutrients (i.e., nitrogen, sulfur, and phosphorus), but also of metals, some of them being toxic like cadmium or lead (Pichtel and Anderson 1997; Pinamonti et al. 1997; Lipoth and Schoenau 2007; Madrid et al. 2007). Thus, organic waste likely to be used as fertilizer must contain metal levels that are suitable for soil application in accordance with Directive 86/278/EEC (European Community 1986), which regulates the use of sewage sludge in agriculture. However, pollutant concentration should be considered a unique criterion for waste reuse. Repeated application over extended periods of time and an increase in application frequency favor metal accumulation and biotransfer. Depending on soil composition and the presence of metals in the reused waste, specific chemical and physical associations can cause the accumulation of these pollutants in soil. This soil build-up might cause severe adverse effects to animal and human health through their incorporation into the food chain, with the intake of food grown in contaminated areas as the most direct route of exposure (Lǎcǎtuşu et al. 1996; Khan et al. 2008; Sridhara Cari et al. 2008; Smith et al. 2009; Zhuang et al. 2009). Environmental risk assessment (ERA) could assist in establishing safety conditions for organic waste application as fertilizer to agricultural crops and pasture production (Franco et al. 2006). In this type of analysis, it is important to consider the proper mechanisms of transfer, accumulation, and exposure for a reliable estimation of human exposure to heavy metals, according to the waste-reuse scenario under consideration.

There are numerous research studies related to the metal contents of different types of organic waste, such as manure (Bolan et al. 2004) and compost (Ciavatta et al. 1993; Ayuso et al. 1996; Ihnat and Fernandes 1996; Goi et al. 2006; Cai et al. 2007; Chen et al. 2008; Farrell and Jones 2009a; Haroun et al. 2009), and the potential biotransfer to soil and crops (Pinamonti et al. 1997; Bazzoffi et al. 1998; Cole et al. 2001; Korboulewsky et al. 2002; Casado-vela et al. 2007; Kidd et al. 2007; Bose and Bhattacharyya 2008; Odlare et al. 2008; Achiba et al. 2009). Many of these authors have stressed both the consequences of the presence of metals for both humans and the environment and the need for controlled agricultural activities.

In this work, a wide inventory of the heavy metal content of different types of organic waste was taken. Data collected in the inventory was used to estimate the possible risk derived from the reuse and application of these residues as fertilizers in agriculture. A multi-compartment fate and exposure model was used. This was the basis of a decision support tool for organic waste management (Río et al. 2011), to evaluate the transfer of heavy metals into the food chain and the possible impacts on human health. The influence of model parameterization on the results obtained was assessed by developing a sensitivity analysis to evaluate the contribution of the different variables considered in the model to uncertainty, especially those related to soil properties. The information and results provided in this work are intended to contribute to the current body of knowledge on the reuse of different types of organic waste as fertilizers within the field of environmental management and safety.

2 Materials and methods

2.1 Data inventory

An exhaustive review of studies presenting the heavy metal content of organic waste was collected from the scientific literature. The resulting inventory included 194 cases of different types of residues, which were classified into three main categories: compost (83 cases, Table 1), sludge and other uncomposted wastes (81 cases, Table 2), and manure (30 cases, Table 3). The inventory focused on residues of domestic origin, assuming a final fate of reuse in agriculture. Special attention was paid to works developed during the last decade, although previous studies were also considered. A higher number of studies involving compost or sludge were considered since, in general, reusing this residue might be more problematic due to its higher metal content compared to other types of organic waste. More cases were included in the inventory to better reflect the effect of possible variations in metal concentration among different sludges (domestic and industrial origin). Even though some studies presented data on several metals, only the five most commonly analyzed (i.e., Cd, Cu, Ni, Pb, and Zn) were considered in the inventory for calculating risk indexes. Another criterion for selecting these metals was to reflect different levels of toxicity in the inventory: high (Cd and Pb), mid (Ni), and low (Cu and Zn).

Table 1 Metal content inventory, metal hazard quotient (HQ), and hazard index (HI) of composts
Full size table
Table 2 Metal content inventory, metal hazard quotient (HQ), and hazard index (HI) of sludge and other wastes
Full size table
Table 3 Metal content inventory, metal Hazard Quotient (HQ) and hazard index (HI) of manure
Full size table

2.2 Environmental risk assessment model

An ERA was used to estimate the potential adverse effects on human health resulting from the application of organic waste containing heavy metals as fertilizer in the production of forage. The importance of the different metals’ distribution mechanisms in the environment varies depending on soil characteristics (e.g., pH, organic matter, and texture), climatic conditions (e.g., rainfall), and agricultural practices (e.g., intensity and frequency).

The accumulation of heavy metals in soil was assessed by establishing a dynamic mass balance between input and output fluxes according to Boekhold and van der Zee (1991) and Moolenaar et al. (1997). The input of metals to the agricultural soil surface may have several contributors: addition of organic waste (i.e., sewage sludge, manure, or compost), irrigation with wastewater, application of commercial fertilizers, or atmospheric deposition. Considering the scope of this work, only the application of organic waste was considered as an input to the model. Output fluxes from soil included leaching from plough to deeper soil layers by precipitation and plant uptake. Data corresponded to areas with different soil types/characteristics, climatology, and precipitation rates. Since metal concentration in solution is usually correlated with soil properties (e.g., pH, metal soil concentration, metal transfer by soil erosion, organic matter, cation exchange capacity, and fulvic and humic acid concentration) and climatology characteristics (e.g., precipitation rate), the leaching of heavy metals into groundwater may be more important in some areas than in others (Sauvé et al. 1997, 2000; Krishnamurti and Naidu 2002; Keller and Schulin 2003; Carlon et al. 2004). Plant absorption rate is related to metal concentration in solution and, therefore, is also dependent on soil type. With the aim of analyzing the effect of organic waste metal content on total risk regardless of soil location, the parameterization of the fate model (i.e., initial soil concentrations, waste application rates, and soil characteristics) was the same for all cases included in the inventory (Table 4). This criterion was also adopted due to the lack of data for these parameters in the majority (>60%) of studies.

Table 4 Parameter values for the distribution model
Full size table

Human exposure was estimated by taking into account five exposure pathways according to the scenario evaluated: (1) intake of meat from cattle grazing in the area, (2) ingestion of milk from cattle grazing in the area, (3) dermal absorption from soil, (4) ingestion of soil, and (5) inhalation of resuspended soil particles. Some of the exposure routes were selected based on the primary activities of the population inhabiting in the study area (e.g., farming). Minor contributions from pathways with a soil exposure source were also expected.

Cattle are exposed to metals through ingestion of contaminated food (i.e., soil, vegetation, and water), by inhalation of resuspended soil particles, or by absorption through the skin. However, only the ingestion pathways were considered to evaluate cattle exposure because dermal contact and inhalation are generally not as significant (ORNL 2004). The equations and empirical multicorrelation models used to estimate metal concentrations in solution (Sauvé et al. 2000), plants (Efroymson et al. 2001), and soil can be found in a previous work (Franco et al. 2006), as along with the exposure model equations and their parameterization.

Quantification of the potential non-carcinogenic risk was determined by a hazard quotient (HQ), which was calculated by dividing the individual doses (milligrams contaminant per kilogram of body weight per day) of each metal by the corresponding reference dose (RfD, milligrams contaminant per kilogram of body weight per day) as shown in Eq. 1.

$$ {\hbox{HQ}} = \frac{\text{Individual dose}}{{\hbox{RfD}}} $$
(1)

Route-to-route extrapolations were needed when no specific dose–response data were available (IRIS database, US EPA 2010). A hazard index (HI) was obtained for each case in the inventory by aggregating the HQs corresponding to the different metals contained in each of the organic wastes considered, reflecting the global risk (Eq. 2).

$$ {\hbox{HI}} = {\sum {\hbox{HQ}}_{\rm{metal}}} $$
(2)

A HI higher than 1.0 indicates that adverse human health effects are expected to occur.

2.3 Sensitivity analysis

A Monte Carlo simulation of 10,000 iterations was developed using the commercial software, Crystal Ball, Version 7 (Decisioneering). This numerical technique propagates parameter uncertainty through the model equations. In this particular case, the sensitivity analysis was only performed on the fate model’s parameters to evaluate the influence that different locations with different soil characteristics and climatology might have on both the HQ and HI. Probability distributions with a standard deviation of 50% around the nominal value were assigned to average production, soil organic matter, and soil infiltration (Table 4). A standard deviation of 100% was assigned to the precipitation rate to observe the effect of precipitation absence in arid locations. Finally, soil pH was allowed to vary between 5.0 and 7.5.

3 Results and discussion

3.1 Risk indexes

The data compiled on heavy metals content in compost, sludge, and manure are shown in Tables 1, 2, and 3 (inventory tables), respectively. It can be seen that sludge contained the highest values of average heavy metal concentration, 50–90% higher than in compost (depending on the metal) and considerably higher than in manure (almost 20 times higher for toxic metals like Cd or Pb). Sludge composition primarily depends on the origin of the effluent treated in the biological reactor. Metal concentrations of concern are typically found in sludge (or compost) coming from a wastewater treatment plant that collects industrial effluents (Soliva and Paulet 2001; Bose and Bhattacharyya 2008), although high concentrations can also be found in domestic sewage depending on the country of origin (Kandpal et al. 2004; Chen et al. 2008; Hua et al. 2008; Egiarte et al. 2009; Lasheen and Ammar 2009).

In general, our metal content values in sludge are within the ranges of those compiled in other works (Pathak et al. 2009). More specifically, average contents of Cu, Pb, and Zn in Table 2 agreed well with sludge values proposed by the EU, while mean values for Ni and Cd were in accordance with those reported by the USA (Stylianou et al. 2008). In Table 2, it should be highlighted that other uncomposted wastes like municipal solid waste or green waste were considered in addition to sludge. Although composting can effectively reduce the availability of metals (García et al. 1995; Smith 2009), it has proved difficult to significantly reduce the total metal content of the initial residue (Manios et al. 2003; Nomeda et al. 2008; Oleszczuk 2008). In fact, this content can be even higher in compost than in the initial waste for certain metals due to the weight loss suffered through mineralization (García et al. 1995). Intermediate metal levels between sludge and manure can be found in compost because composted waste can be either sludge or manure.

On the other hand, the presence of metals in manure is due to animal (e.g., cattle, pig, and poultry) excretion of trace elements contained in their diet or other health supplements (Petersen et al. 2007; European Commission (2003)). Thus, the concentration of metals in manure is generally moderate, especially for toxic Cd and Pb. Micronutrients like Cu and Zn can reach substantial levels because the animal is usually overdosed with these oligoelements to increase productivity and disease resistance (Nicholson et al. 1999).

The metal HQ and HI were calculated for each of the 194 cases in the inventory tables using the multi-compartment risk assessment model described in the previous section. It can be seen in Tables 1, 2, and 3 that the HI value exceeded the recommended ERA safety limit of 1.0 in 14% of sludge cases, with an average value of 0.64. The percentage of cases above 1.0 was lower for compost (4%), with an average value 0.42. However, it is important to note that the risk estimated is incremental in that it only reflects one of the possible routes of metal exposure for humans, and the obtained HI values for sludge and compost become of greater concern within this context despite being lower than 1.0 in most cases. Regarding manure, its reuse as agricultural fertilizer could be considered a safer practice (0.25 average HI). Note that only total metal contents in waste were used to calculate HQs and the HI, and aspects like bioavailability were not assessed in this work. This fact could reduce the final value of the HI because some metals may be strongly complexed with organic matter (García et al. 1995; Zheng et al. 2004; Nomeda et al. 2008). Hence, it is possible that taking bioavailability into account would result in the reduction of the HI for organic wastes. However, metal bioavailability depends not only on metal content, but also on the chemical properties of organic waste (Smith 2009).

Average metal-specific HQs and an average HI were calculated for each type of waste (Fig. 1). The highest contribution to the HI was the essential trace element Zn, and typical toxic elements like Cd and Pb posed a minor contribution to total risk. Although a very low dose (RfD) of these metals can result in severe adverse effects to human health, it is necessary to take into account each evaluated case. From the original organic waste applied on land, metals have to be transferred to vegetation and cattle, then to humans. Thus, the biotransfer potential, rather than the toxicity potential, would be the best indicator of the magnitude of risk in this particular scenario. According to the Risk Assessment Information System (ORNL 2010), biotransfer factors (BTFs) to meat and milk for Cd, Cu, and Pb ranged between 1·10−03 and 1·10−04 in magnitude, while for Zn, the values were 1·10−01, and 1·10−02 for meat and milk, respectively. Thus, although the ingestion RfDs of Zn was significantly higher in comparison with the other metals (i.e., the dose a human ingests must be high to produce any adverse effect on health), significant concentrations of Zn in either type of organic waste and high BTFs resulted in large HQs, exceeding the safety limit for several cases of compost and sludge. Ni also contributed significantly to the HI because of its high BTF to milk (1.6∙10−01). An analysis of the exposure pathways considered in the scenario revealed that ingestion of meat, followed by milk ingestion, represented between 75% and 90% of the total risk on average in all cases inventoried. As expected, pathways involving direct absorption from soil contact and inhalation had a minor effect on the risk index, and both the Cd and Pb HQ were low.

Fig. 1
figure 1

Influence of metal and organic waste type on risk indexes, hazard quotient (HQ), and hazard index (HI)

Full size image

The HQs of metals for each type of organic waste were proportional to their concentration. The contribution of Ni to the HI was approximately 10–12% for compost and sludge and 6% in manure. In the case of Zn, the opposite trend occurred, with a contribution to manure of 68% and to compost and sludge of 64%. So, although some authors have indicated that levels of Zn in manure are generally lower than in other types of organic waste (Soliva and Paulet 2001; Achiba et al. 2009), we found similar levels in manure, compost, and sludge for the cases included in the inventory. Together with Cu, Zn content was higher than that of other metals in manure due to excretion of these oligoelements after supplementation in cattle. Zn concentration was also highest in compost and sludge, but a more significant presence of the other metals was also found, especially for the toxic Cd and Pb. The average level of Zn in sludge calculated from the studies in the inventory was 1,200 mg·kg−1, while in manure it was 300 mg·kg−1.

Zn can end up in wastewater and sludge from several different sources: excretion by humans from ingested food or water, use of galvanized materials, car emissions, car washes, metallurgy, mining, painting, and any applications that involve high levels of Zn in domestic and industrial wastewaters (Sörme and Lagerkvist 2002). Zn is an essential element for humans, with a recommended dietary intake of approximately 0.16 mg·kg−1·day−1 for men and 0.13 mg kg−1 day−1 for women (ATSDR (Agency for Toxic Substances and Disease Registry) 2005). However, prolonged oral exposure to zinc at high levels (~2 mg kg−1 day−1 Zn) may cause severe symptoms of copper deficiency, including anemia and neutropenia (Ramadurai et al. 1993).

3.2 Legislative limits

Proposed limits for heavy metals in organic soil fertilizer amendments are given in Table 5, and HIs for each specified-use class (A, B, and C) have been calculated. Considering metal content, class A was the most appropriate for cultivating crops intended for direct human consumption. The resulting HI after 100 years of applications of this type of organic waste was 0.23, but a low percentage of compost (20%) and sludge (10%) considered in the inventory can be classified within this category. This percentage increased to 45% of cases adequate to be applied according to class A guidelines in manure. Sixty percent of compost and 40% of sludge fell into the type B classification, which is more adequate to fertilize land for forage or fruit production. Finally, despite its higher metal content, fertilizers classified under type C had HQs and a global HI that were similar to type B because of its limited application rate, which must be lower than 5 t ha−1 year−1.

Table 5 Limit values of heavy metals content in compost according to Legislation and its correspondent HQ and HI
Full size table

In general, countries presented similar values of maximum permissible contents in compost for each metal, providing, an acceptable HI as a first approximation. However, different soil properties and climate could influence the final value of the risk index, which was evaluated with a sensitivity analysis. Finally, although legislation allows the use of sludge containing much higher concentrations of heavy metals (Goi et al. 2006; Stylianou et al. 2008), its application in agriculture is usually strongly constrained to low application rates and frequencies, as well as to specific times of the year. These restrictions were not considered in the estimation of sludge HI, although they could result in a decrease of metal risk indexes. Despite this worst-case scenario, incremental risk cannot be considered negligible, and metal limits in organic waste should be decreased, as stated previously in literature (Madrid et al. 2007).

3.3 Sensitivity analysis

Figure 2 illustrates the influence of soil properties and climate in the HQ of each metal and in the total HI. Soil pH played a key role in the magnitude of total risk for Cd, Ni, and Zn because an increase in the value of this parameter provoked a significant reduction in HQ and HI. Low pH values enhance metal solubility, mobility, and bioavailability in soil (Smith 1994; Planquart et al. 1999), as reflected in certain countries’ legislation that establishes a different organic waste application rate depending on the pH value (i.e., lower or higher than 7).

Fig. 2
figure 2

Influence of soil and climate characteristics (pH), organic matter (OM), average production (AP), precipitation rate (PR), and infiltration factor (IF) on metal hazard quotient (HQ), and hazard index (HI)

Full size image

Soil organic matter only influenced the HQ of Pb significantly (70.9% of variance). It had a lower effect on Cd and Ni and was negligible for Cu and Zn. Figure 2 shows that an increase in soil organic matter resulted in an increase in the Pb HQ (i.e., positive effect). Pb is one of the most strongly adsorbed metals by organic matter and, thus, may be effectively retained and accumulated in the soil matrix (Schroth et al. 2008). Lead’s low biotransfer potential implies that the direct soil exposure pathways contributed more to its HQ. Organic matter can fix and increase the Pb concentration in soil and increase its HQ accordingly, although this value was very low compared with the total HI. Therefore, the influence of organic matter could be significant in scenarios where direct and prolonged contact with Pb-contaminated soil is expected.

Finally, the HQ of Cu was primarily affected by climatic conditions (i.e., precipitation rate) and was less sensitive to pH changes (Smith 1994). In contrast to the behavior of the other metals, an increase in precipitation would result in a decrease in risk due to Cu according to the sensitivity analysis. Enhanced leaching of Cu through the soil matrix (Kidd et al. 2007) escapes metal biotransfer from soil solution to vegetation and cattle, and subsequently to humans, leading to a low HQ.

The high influence of pH on the global HI can also be seen in Fig. 2. This influence is due to the high contribution of Zn, followed by Ni, because both metals significantly depend on pH. Precipitation rate is the second most influential variable at 20%, due to the contribution of Cu (after Zn and Ni). Thus, soil and climate properties (i.e., location) can significantly vary the magnitude of risk depending on the metal. For example, the sensitivity analysis revealed that in the case of organic waste reuse, locations with acidic soils and high precipitation rates would be more affected by Zn exposure. These two scenarios can be found within the same country, Spain, where the Mediterranean area has basic soils and low precipitation rates, but the Atlantic area (NW) has acidic soils and high precipitation rates.

4 Conclusions

In this study, a wide inventory of the heavy metal content in three types of organic wastes (i.e., compost, sludge, and manure) was taken. Health risks due to the reuse of these residues as agricultural fertilizers were determined by an ERA. The results indicated that sludge contained the highest concentrations of metals, and the presence of toxic metals like Cd and Pb was more significant than in compost and manure. As expected, sludge reuse in the proposed scenario resulted in the highest incremental risk. Surprisingly, the metal with the greatest risk contribution to the three types of organic waste was Zn, making the presence of toxic Cd and Pb almost negligible in terms of risk. Although Zn presents a very low level of toxicity as an essential element to life, its high biotransfer potential may create in significant concentrations that exceed the recommended doses in organic matrices like plants, cattle, and humans. Therefore, specific measures should be taken to regulate the Zn content of organic waste depending on its final management solution. The origin of the Zn should also be established for proper reduction measurements in emissions, especially in sludge. However, a worst-case scenario approach was selected, and the risk may be overestimated because legislation restrictions on the application of sludge were not considered. Another key aspect, bioavailability, was not addressed in the present work. Future efforts should be focused on assessing metal speciation in the soil solution, either as inorganic complexes or bound to humic and fulvic acids.