To develop this study, the Spanish registry of fertilisers regulated by the Spanish Fertilisers RD and available at the Agriculture section of the website of the Ministry for the Ecological Transition (MITECO) has been consulted on March 30 ^th of 2018. The register contains 44 product categories. However, when carrying out this study, the only categories analysed were those classified under the heading about organic, which are the follo­wing ones: (i) '6002-Organic amendment com­post' (OaC), (ii) '6003-Green compost' (GrC), (iii) '6004-Organic amendment manure compost' (MaC), (iv) '6005-Organic amendment vermicom­post' (VeC), and (v) '6009-Alperujo compost' (AlC). A database has been elaborated with all the available informa­tion in the registry for each different compost of each category (OaC, GrC, MaC, VeC and AlC). For each compost the following aspects are considered: (i) European Waste Code (EWC) of raw materials composition; (ii) content in main elements (total N, nitric N, ammoniacal N, organic N, ureic N, P ₂ O ₅ , K ₂ O), secondary elements (SO ₃ , Na ₂ O, CaO, MgO), micronutrients (B, Co, Cu, Fe, Mn, Mb, Zn), other characteristics (organic matter, organic carbon, humic extract, humic acids, fulvic acids, Cl ^- , maximum humidity, conductivity, density, pH, C/N), heavy metals (Cd, Cu, Ni, Pb, Zn, Hg, Cr, Cr(VI)) and pathogens content ( Salmonella and E. coli ).

Before initiating the analysis of the databases, the information provided by the different registries has been unified and standardised, by performing a previous consideration of some aspects in order to filter the collected samples. Once all data were explored and validated, a descriptive statistical analysis (means, standard deviations) of the physico-chemical parameters of the compost of each category was performed. In addition, frequency with which each residual raw material intervenes as an ingredient in the compost of each category was calculated and plotted.

Besides this, correlations between different phy­sical and chemical parameters and the different waste raw materials used in the formulation of each compost were computed using SPSS Statistics 23.

In order to define a quality assessment system for Spanish's compost, the study of Saha et al. (2010) was used as reference. A methodology has been developed to calculate new fertility and clean indicators adapted to the Spanish context. To do this, it has been taking into account the parameters of the quality check of compost established in the Spanish legislation (Spanish Fertiliser RD) as well as the most demanded characteristics of fertilisers in the European market.

The fertility indicator (FI) of produced composts has been set according to the formula proposed by Saha et al. (2010) :

where S _i is a factor associated to the content of each parameter considered in the assessment of fertility and W _i is their weighing factor.

In order to establish the FI, the stability degree of the organic matter has not been considered, contrary to what Saha et al. (2010) did, since it is not a parameter required by the Spanish legislation and, therefore, it is not available in the analysis of the MITECO's fertilisers registry. While trying to estimate the mea­sure of stability, which is moreover an indirect measure of the maturity ( Oviedo et al. , 2015 ), attempts were made to estimate such parameter through other connections such as humic/fulvic acids or ammonium/nitrate ( Jiménez & Pérez, 1992 ; Ko et al. , 2008 ). However, a very low percentage of samples state publicly those parameters, so the obtained results are only partial, in order to be able to consider maturity for the determination of the FI. The rest of the parameters considered in the study of Saha et al. (2010) to determine the FI (TOC, N, P, K and C/N) have been included in the present study by adapting to a greater or lesser extent the score values and the weighing factor to the European characteristics. Nitrogen is the parameter that suffers the highest variations in the score values. The low N content of the typical composts of the original study is more limited because of the low availability of wastes rich in proteins or livestock manures. Additionally, so as to achieve a higher distinction within the analysed samples, it has been considered relevant to use 6 scoring categories (from 0 to 5, where 0 is the minimum content of macronutrients) instead of the 5 established by Saha et al. (2010) (Table 2). The weighing factors (Wi) considered are the same used by the authors.

The clean indicator (CI) assesses the quality of com­post according to its heavy metals content and has been established by the following equation ( Saha et al. , 2010 ):

where S _j is a factor associated to the analytical result of the heavy metal and W _j is the weighing factor assigned to that specific heavy metal.

The score values proposed by Saha et al. (2010) have been adapted to the heavy metals content defined in the Spanish Fertilisers RD for the three quality categories (Class A, B or C). Moreover, an additional heavy metal (Hg) has been considered. The suggested values for the six scoring categories (from 0 to 5, where 0 represents the highest content of heavy metals) are equidistant. A maximum value of 5.5 is allowed for those products with a concentration in heavy metals lower than that required for the cleanest category (see Table 2). The weighing factor coincides with that used by Saha et al. (2010) , adding the maximum factor to Hg due to its negative effects in the nervous system and foetal development in mammals ( Govind & Madhuri, 2014 ).

The different compost typologies analysed in this paper have been produced from organic wastes or from by-products of different origin identified by their EWC. Figure 1 shows the main raw materials and the proportion in which they are used as ingredients for each of the analysed compost kinds. The EWC codes that most frequently appear in the composition are 020103 and 020106, which refer, respectively, to green waste and livestock manures. Those classified as green waste (020103) appear in 28.6% of OaC, in 14.8% of MaC, in 25.0% of GrC and in 28.9% of VeC reported in the MITECO's fertilisers database. For its part, waste from livestock farming (020106) is present in 30% of OaC, 100% of MaC and 85% of VeC.

With regard to organic waste mainly used in the formulation of initial mixtures for composting, OaC offers the widest variety of EWC. In this type of compost there are five kinds of waste that appear, as raw material, in more than 20% of the checked samples: waste from vegetable tissue (020103), livestock manu­res (020106), sludge from urban wastewater treat­ment plants (190805), biodegradable waste from gardens and parks (200201), and the organic fraction of municipal waste not collected separately (200301). However, the organic fraction of municipal waste not collected separately (200301) is not legally accepted anymore for the production of compost ( BOE 2011 , BOE 2017 ). Current registries containing such waste correspond to authorizations granted before and that will expire bet­ween years 2018 and 2021. It has to be highlighted that organic fraction of municipal solid waste separately collected (020108) is only employed in the 6.9% of the AoC and therefore it has not been included in Figure 1.

In the cases of MaC and GrC, the main organic wastes used in the initial mixtures are, obviously, those matching their designation: livestock manures (020106) in the first one, and waste from vegetable tissue (020103) and biodegradable waste from gardens and parks (200201) in the second one. It is remarkable that some products obtained only from manure or vegetal wastes are registered in type OaC, instead of in their specific category (MaC or GrC). Likewise, the registered VeC were produced mainly from green wastes (020103) and manures (020106). Vermicomposting is a biological transformation process that includes the action of earthworms, so it is logical to consider VeC as a product of a specific category, independently of the organic waste used in its initial formulation.

In Spain the legal use of a compost is determined by its heavy metals content, as it is described in the Spanish Fertiliser RD. Compost class A has the lowest heavy metal content and it can be even employed in organic farming, whereas Class C has the highest content in heavy metals and therefore it can only be applied at a maximum rate of 5 Mg ha ^-1 . Table 4 summarises the percentage of samples classified as Class A, B or C, for each compost type. The standard deviation of all heavy metals concentration for each product-type reflects a wide variety. OaC, with the highest variety of raw materials, presents the lower percentage of samples classified as Class A, only 24.1% of the available samples (Table 4). The compost produced from green wastes (GrC) presented the maximum percentage of samples considered as Class A (76.5%). MaC and VeC (85% of the included products are mainly elaborated from livestock manures) have almost the same percentage of samples classified as Class A and Class B, and the remaining (7.1% and 15.2% respectively) as Class C (Table 4).

It is then remarkable the lack of analytical data in the registry for some metals that have a significant influence in the legal classification of compost, such as Cd, Cr and Hg. In particular, 76 samples lack concentration data of Hg, 11 of Cr and 35 of Cd. Despite that, those products are registered in the MITECO's database and have been given a legal quality classification according to their heavy metals content, even some of them as Class A.

Raw materials employed in producing each com­post determinate its heavy metal content. The speci­fic heavy metals that have a higher impact in the deter­mination of the legal classification of compost (Spa­nish Fertilisers RD) for each typology have been studied:

• OaC: Data about the heavy metals content for this compost type evidence that Cu, Zn and Pb are the metals with the highest limitation impact in its legal quality. However Genevini et al. (1997) , when analysing thirty composts from municipal solid waste, found Cd as the main metal limiting legal compost quality. These different results could be due to the change in municipal solid waste collection system during the last 20 years. In Spain, Catalonia started the source-separated collection of biowaste and its treatment through composting, almost 20 years ago. Puig et al. (2016) studied the heavy metals content of the compost produced in the composting plants located in Catalonia to treat biowaste (EWC 200108). They reported low concentration of heavy metals, especially for Ni (17.7 mg kg ^-1 ), Pb (47.0 mg kg ^-1 ), Cu (121.8 mg kg ^-1 ) and Zn (281.5 mg kg ^-1 ). This lower concentration of heavy metals in the compost-type (OaC) produced in Catalonia was also indicated by Huerta et al. (2011) and Cerda et al. (2017) . These authors signalled that the main reason is the source-separated collection of biowaste in Catalonia.

• MaC: The heavy metals that limit the most in determining legal qualities are: Cu, Pb and Zn. Specially Cu and Zn are heavy metals related to the livestock manures and to the compost produced from them ( Hsu & Lo, 2000 , 2001 ).

• GrC: It has instead the highest percentage of products classified, as Class A. Cu and Pb are the limiting metals to determine the legal quality of the studied GrC. Nevertheless, most of the registries in this typology do not include information of Hg concentration, nor of other metals such as Cd, Ni and Cr. For this typology 76.5% of the samples are incomplete for at least one heavy metal (Table 4).

• VeC: Cd and Zn are the metals that limit the most in determining qualities. Specifically in the case of all registered VeC, the Pb concentration is lower than the limit legally established for the highest quality category (Class A).

A correlation has been found between the concen­trations of Ni and Pb in all samples of typology OaC ( r ² =0.566), MaC ( r ² =0.384) and GrC ( r ² =0.810), but not in category VeC ( r ² =0.007). Such connection had already been announced by other authors ( Saha et al. , 2010 ), but related to composts which predominantly contained vegetal remains from agriculture, that is, green waste. Among the samples in this study, this connection applies mainly to category GrC, which on the other hand has the lowest average Ni concentra­tion of the four studied typologies. In order to analyse if this bigger connection is influenced by the typolo­gies of the wastes mainly used in the production of GrC (Fig. 1; EWC 020103 and 200201), an analysis tried to find out if the connection is better between the heavy metals (Ni and Pb) of composts using such wastes: type GrC and OaC. For GrC, results showed a strong connection between Ni and Pb concentrations ( r ² =0.97) for EWC 200201 (vegetable waste from gardens and parks), and a lower one ( r ² =0.628) for EWC 020103 (waste from vegetable tissue). For type OaC, which also includes a big amount of products elaborated from those wastes, the only connection between Ni and Pb exists for samples using wastes classified under EWC 200201 ( r ² =0.668), while such connection does not happen for samples with wastes under EWC 020103 ( r ² =0.065).

EWC 200201 mainly corresponds to woody ma­te­rials that must be shredded so that they can be composted with an adequate particle size. The shred­ding machines normally used in composting facilities are industrial equipment with high-powered engines and stainless steel hammers or blades for crushing and/or cutting the wood. Both heavy metals (Ni and Pb) are related to different components of this kind of equipment. Ni is one of the heavy metals used to make iron into hard steels and can be found in stainless steels, so due to the use of blades and hammers, particles of this metal can contaminated the outflow of shredded wood material. Moreover, Pb is a normal additive for lubricant oils and even for soldering metals, so its presence in this kind of equipment is common. Since these products are related to leaking and wearing, the contamination by Pb of the shredded material should not be also unusual either. Specific tests should be done to confirm these sources of contamination for this kind of woody green wastes among other vegetal wastes.

In the specific case of VeC, there is no connection between heavy metals. That trend may be related to the bioaccumulation of heavy metals in the organism of the earthworms responsible for the degradation process of the organic waste ( Wang et al. , 2017 ). Many studies have proved this process of bioaccumulation of heavy metals in earthworms. Pb is the heavy metal that shows the highest trend to be accumulated in their organisms ( Liu et al. , 2012 ; Swati & Hait, 2017 ) and, therefore, vermicompost presents a low Pb content. Moreover, this study evidenced that the lack of connection is not related to the raw materials used. Most of the VeC (93%) included in the Spanish fertilisers registry are elaborated mainly with wastes from livestock manures (EWC 020106).

By considering average concentration values for the four typologies (Table 4), the compliance with national and international legal requirements of these products-types can be studied. Table 5 presents a comparative study of how the compost typologies would be classified by their legal quality according to different national and international guidelines.

When analysing the heavy metals concentration values resulting from the products-type (Table 3) with different legal requirements of limit con­centration values (listed in Table 1), it follows that (Table 5):

- Typologies GrC and VeC keep the strictest requi­rements in all compared European regulations.

- In the case of MaC, the concentration values of Cu and Zn exceed the legal limits required to be considered as an ecological product and to be included in the Spanish legal quality Class B.

- According to the Spanish legislation, types OaC and MaC are mostly classified as compost Class B. GrC and VeC are instead mostly considered as compost Class A.

- According only to the heavy metal concentration, the application of Portuguese legislation gives similar classification of the product-type quality to the Spanish one. However, the Portuguese compost quality classification considers other parameters like inert materials content. Moreover, compost maturity also defines its possible uses and applications of the final product.

- The application of the Austrian Compost Ordinance (2001) allows to generally distinguishing different degrees of legal quality between type OaC (corresponding to Class B) and MaC (corresponding to Class A).

- If the Canadian guidelines ( CCME, 2005 ) for compost quality were applied, all the studied compost typologies would reach the maximum category (A) and no differences among them would be visible.

- Overall, the product-type presenting the lowest legal quality would be type OaC, which would only fulfil the End-of-Waste standards and would be considered as Class B by the Spanish Fertilisers RD.

The main macronutrients content (N, P, K) of the different compost categories analysed (Table 3) shows that MaC presents the highest potential as fertiliser, given its highest and practically equal concentrations of N, P, K, which almost equate to a mineral fertiliser with formulation 2.5:2.5:2.5. VeC, on the contrary, presents the lowest concentration of macronutrients, since it is a product that requires a longer and more controlled maturity phase, which gives it a status more typical of a growing media than of a fertiliser. This longer process would also become evident in the lower content of ammonium N of this typology when compared with the others. The same happens with GrC, whose registered samples have relatively low concentrations of macronutrients, especially of P ₂ O ₅ , which is also indicative of its potential applications, more linked to gardening and the elaboration of growing media than to conventional agricultural applications.

When comparing these macronutrients contents with the concentration of the wastes used in its production (Fig. 1), a positive connection is visible between the use of animal manures and the N and potassium concentration levels in the resulting products.

Regarding other nutrients content (Table 3), howe­ver, VeC stands out due to its high concentration levels of some of them, especially Ca and Mg. This higher concentration of Ca in VeC compared to any other compost type is a common thing due to the combined effect of earthworms and fungi in the environment ( Spiers et al. , 1986 ; Das et al. , 2012 ), which provokes a higher excretion of that element to the environment. Equally, the widest use of animal manures in the production of the registered vermicomposts is also a reason for the higher concentration of Ca in the environment ( Arancon et al. , 2008 ). In this respect GrC is the product with the poorest concentration of these nutrients, with much lower levels than all the others are. Among the four analysed compost typologies, OaC has the most balanced concentration levels and the biggest quantity of nutrients.

Regarding the rest of the physico-chemical para­meters stated for these products, some differences are appreciated related to the kind of biological pro­cess and its scope, and to the initial composition of organic wastes (Table 3). In this way, it can be observed that contents of organic matter and organic carbon are lower in VeC (a more stable product requiring longer and more intense process times, so it is more degraded). On the contrary, the highest values belong to GrC, which is subjected to a more limited process due to the lower biodegradability of the original materials ( Puyuelo et al. , 2011 ) and to the influence of process parameters, such as the particle size or the greatest difficulty to control humidity of materials during the different phases.

As regards indicators of the process' duration or the stability of the final product, there are other indicative parameters, such as the content of humic substances, where it is observed that minimum and maximum levels correspond respectively to MaC and VeC. The first one corresponds to a very intense degradative process, due to the composition of the initial raw materials (high content of protein substances and low C/N) with high biodegradability. The maximum concentration values of humic acids are found in VeC (highly spread), corresponding to a more complex process deriving in a greater stability of the final product. (Jimenez & Pérez, 1992; Haddad et al. , 2015 ; Martínez et al. , 2016 ).

Other parameters, such as conductivity and chlo­ride content, are also affected by those factors related to the evolution of the biological process. Lowest values are registered in those products with more potential applications as growing media (VeC and GrC).

Figure 2 shows the obtained Fertility and Clean indicators for the composts analysed in this study and registered in the MITECO's database, being distingui­shed by typologies (OaC, MaC, GrC and VeC). As can be seen, there is no correlation ( r ² < 0.01) between the FI and the CI for any of the studied compost typologies. In type OaC, there are products with void CI (showed at left side on the graph) since, as already mentioned, some of the registered products do not include va­lues of the heavy metals content.

Table 6 shows, for each studied compost typology, the percentage of registries included under the different scores of the FI. Overall, it is noticeable that most of the samples have an average FI value of around 2-3. As OaC and MaC are the products elaborated from organic wastes with the highest proportional content in macronutrients (livestock manures, agro-industrial wastes, biowastes, etc.), they also include the largest percentage of samples with high FI values (above 4). For its part, VeC presents the lowest FI values. Nevertheless, the result of the degradative activity of earthworms, together with the great variety of microorganisms involved in this process, provoke the creation of certain compounds that facilitate germination and the development of vegetables ( Zhang et al. , 2000 ; Arancon et al. , 2008 ), which are not considered among the required analysis. Altogether, no products have been obtained, in any of the established compost typologies, whose calculated FI had a final global value of the highest category (>5).

Table 6 shows, for each studied compost typolo­gy, the percentage of registries included under the different scores of the CI. The CI proves the low heavy metals content of the samples included in types MaC, VeC and, mainly, GrC, even though this contrasts with the low FI of this last category. In OaC, even if it presents the largest number of samples with the lowest CI (3.3%), there are more than 70% of the samples with CIs in the highest categories (CI >3). It is also observed that in all typologies of compost there are products with CIs higher than 5, but mainly in GrC, where 81.2% of the registries have concentration values of heavy metals lower than the legally required minimum (Spanish Fertilisers RD) for the highest quality (Class A). In this type of compost, Cd and Hg are the most limiting heavy metals.

After the elaboration of this exhaustive analysis of the current compost registered in the Spanish database, it can be concluded that most of the composts registered in the MITECO's database (Spanish Registry of Fertiliser Products) main nutrients content is, in average, relatively low. Regarding heavy metal contents, the composts with highest percentage of samples with the maximum Class ('A' according to the Spanish RD) are, in this order, green compost (GrC), manure compost (MaC), vermicompost (VeC) and organic amendment compost (OaC). Considering the parameters registered in the MITECO's database (Spanish Registry of fertiliser products), a new methodology to assess and classify the global compost quality is proposed based on two indicators: fertility indicator (FI) and clean indicator (CI). However, in the future quality of marketed compost should be classified considering physico-chemical parameters that could determine the use of a compost, such as: compost stability and/or maturity, phytotoxicity, density or porosity.