Comparison of Existing Legal Assessment Values for Heavy Metal Deposition in Western Europe and Calculation of Assessment Values for Luxembourg

: The protection against eco- and human-toxicological impairments caused by atmospheric deposition of heavy metals requires legally deﬁned assessment values. Since such values are missing for Luxembourg, the aim of this investigation was to evaluate different approaches to derive assessment values for the regulation of heavy metals that are in accordance with scientiﬁc and legal standards. To this end, assessment values for heavy metals were derived from the compilation of respective values implemented in European countries. In addition, (1) precipitation-related assessment values for the protection of soil for Cr, Zn, and Cu and (2) precautionary assessment values (critical loads for Cr, Zn, and Cu, as well as As, Cd, Ni, and Pb) for the protection of human health and ecosystems were calculated. The calculation of the regionally differentiated precipitation-related assessment values resulted in ranges of 17–272 g Cu ha − 1 a − 1 , 167–2672 g Zn ha − 1 a − 1 and 17–272 g Cr total ha − 1 a − 1 . The critical loads for drinking water protection vary in the ranges from 1.23 to 2.14 g Cd ha − 1 a − 1 , from 4.05 to 8.63 g Pb ha − 1 a − 1 , from 2.6 to 5.9 g As ha − 1 a − 1 , from 258 to 564 g Cu ha − 1 a − 1 , from 1292 to 2944 g Zn ha − 1 a − 1 , and from 12.9 to 29.9 g Cr total ha − 1 a − 1 . Ecosystems are signiﬁcantly more sensitive to Pb, Cu, and Zn inputs than humans. For As and Cr, humans react much more sensitively than ecosystems. For Cd, the critical loads for drinking water, ecosystems, and wheat products are about the same.


Introduction
When, in the middle of the last century, several thousand people in the Japanese cities of Minamata and Niigata became ill from damage to the nervous system and some also died, the international public became aware for the first time of the environmental pollution caused by heavy metals. The cause of this poisoning, later referred to as Minamata disease, was wastewater contaminated with Hg from plastic production that was discharged into the sea. The Hg accumulated in fish via the food chain and led to symptoms of poisoning, particularly in humans, who mainly fed on fish. By the time of the first court case in 1973, 78 deaths had already been reported [1], a total of about 3000 people were poisoned, and more than 1800 of these cases were fatal.
As trace elements, some heavy metals are essential for life processes in the biological system. On the other hand, above a certain quantity, they contribute to pollution and can have toxic or carcinogenic effects. Accumulation is possible via the food chain, which can ultimately also endanger human health. For this reason, regulations have been established at both a national and international level to limit the input of heavy metals to a tolerable level in order to protect human health. scientific and legal standards. Founded by this comprehensive evaluation, it should be worked out whether the existing regulations in Luxembourg's neighboring countries can be applied to Luxembourg or whether new limit values should be set according to the critical loads determined by the authors on the basis of models for specific regions of Luxembourg. Based on the compilation and comparative analysis of existing legal and sublegal assessment values for atmospheric deposition of As, Cd, Cr, Cu, Ni, Pb, and Zn in the Western European countries France, the Netherlands, Belgium, Switzerland, Austria, Germany, and in the EU (Schlutow et al. 2021), the second aim of this investigation was to establish regionalized assessment values for Luxembourg for the first time. The methodology applied to this end was developed and already applied by the authors for Germany [5,6] but extended and further developed for Luxembourg: 1.
Precipitation-based assessment values for soil protection for Cr, Zn, and Cu (Section 2.1).

2.
Precautionary assessment values (critical loads) for Cr, Zn, and Cu, as well as, Cd, Ni, and Pb for the protection of human health and ecosystems (Section 2.2).

Comparison of Legal Regulations
The regulations and recommendations considered below contain different categories of assessment values that differ in their protective purpose, their level of protection, and their protective objective. Therefore, this paper uses the overarching term "assessment value" but adopts the nomenclature of the regulations when quoting from them.
In the following section, we distinguish between precaution-oriented assessment values and those that serve to avert danger: Assessment values that serve to avert danger permit, in principle, higher pollutant concentrations or discharges than precaution-oriented ones. In contrast to precaution-oriented assessment values, they generally serve to assess concrete (including planned) facilities, projects, or management measures and are derived on a use-specific basis (e.g., test values and measure values in soil protection).
For the protection of human health, assessment values for concentrations in air are contained in the following Western European regulations, directives, and recommendations. These can be converted into deposition values in order to compare them with other assessment criteria. In this context, not only the depositions on arable land and grassland, which are important for humans as primary links in the food chain, but also forests where people spend time for recreation are considered. The assessment values of the recommendations, laws, and sub-legislative regulations are summarized in Table 1. Table 1. Comparison of assessment values from legal and sublegal regulations in Western Europe, critical depositions after conversion by means of deposition rates according to Schaap et al. [7].

Rulebook
Protection Target   German TA Luft (2021) [14] Project-related load Human and environment 365 7 55 15 GermanTA Luft (2021) [14] Project-related load Environment 675-6935 9.  [19] Project-related load Human and environment 365 7 55 15 Due to the methodological differences in their derivation, the assessment values compiled in Table 1 are only comparable with each other and with the critical loads to a limited extent. The partly significant differences exist due to different protection levels, protection goals, and the impact reference (Table 2).   The technical bases according to Prinz and Bachmann [23] for the derivation, including in particular the test and/or measure values as well as the assumptions on soil thickness and density, predominantly still correspond to the current status (explanatory memorandum to the new version of the TA Luft [24]. The calculation of the precipitation-limiting values is based on the following calculation procedure [23]:  (Tables 3-5), since a situation is certainly to be regarded as undesirable, in which an exceeding of test values would be foreseeable as a consequence of airborne pollutants. If, instead of the test values, only measure values are available as the starting point for the calculation, Prinz and Bachmann [23] recommend that, when converting to precipitation-limiting values, a discount be taken into account to compensate for possible uncertainties. However, the amount of the recommended discount is not specified. Therefore, no discount is applied in this paper. Table 3. Test values according to § 8 paragraph 1 sentence 2 no. 1 of the Federal Soil Protection Act [25] for the direct uptake of pollutants on children's playgrounds, in residential areas, parks and recreational facilities, and industrial and commercial sites (in mg kg −1 dry matter, fine soil).

Metal
Children  Accumulation Period, Soil Density and Thickness A period of 200 years is taken as the time period for which a still tolerable enrichment is to be calculated. The heavy metals introduced via atmospheric deposition are primarily bound in the topsoil and thus enriched there. The use of the soils results in varying degrees of mixing of the upper soil layers, so that depending on the use, soil layers of varying thickness must be assumed in which the heavy metals accumulate. Likewise, the soils have different storage densities. For the calculations, the soil thicknesses and storage densities listed in Table 6 were used in accordance with the report of the subcommittee "Impacts" of the State working group on Emission (LAI) "Emission values for mercury compounds" [26].

Determination of Background Values for Luxembourg
Measured values of heavy metal levels in Luxembourg were provided by the Administration de l'environnement, Unité stratégies et concepts, as of 5 February 2020 [27]. Soil samples were collected and analysed at 308 sites, including 113 arable, 124 grassland, 5 vineyard, and 66 forest sites, each at 2-3 depth levels. Heavy metal contents were determined both after aqua regia digestion and with ammonium nitrate extract. In the following section, however, only the values determined with aqua regia digestion are considered, because this corresponds to the methodology of Prinz and Bachmann [23] and thus to TA Luft [14,28]. In the following section, only the data sets corresponding to the assessment-relevant soil layers (Table 6) are used.
Prinz and Bachmann [23] used the 90th percentile of the contents determined in comparable studies in arable soils and forest soils in rural areas with certain widespread parent rocks of soil formation for the calculation as the background value of the ubiquitous distribution of substances in soils. Prinz and Bachmann [23] point out "that background contents could be much higher, especially for some bedrock soils, for geogenic reasons alone. In extreme cases, they can be so high that the test values are already exceeded, at least in part, by the geogenic heavy metal contents. In this case, the Soil Protection and Contaminated Sites Ordinance [29] provides for special assessment regulations. Within the framework of the calculation of precipitation-limiting values presented by Prinz and Bachmann [23], which are based on typical regulatory conditions, such extreme locations were not included, however, since this would lead to negative calculation results. The consideration of these aspects must be reserved for a special case examination." Therefore, in the present paper, a determination of outlier values and extreme values was carried out first. First, the 25th and 75th percentiles (perc.) of Cr, Cu, and Zn contents in aqua regia digestion were determined for the data sets of the soil layers relevant to the assessment within the soil-use types. Then, the outlier values were determined according to the following equations:

Calculation of Precipitation-Limiting Values
As in Prinz and Bachmann [23], we used the following input data for the soilhuman impact pathway as the basis for calculating precipitation-limiting values for settlement areas: - The test value for children's play areas from the German Federal Soil Protection Ordinance [25]. However, such a test value is only available for Cr total (Table 1) The input data used for the soil-crop impact pathway were [23]: -Test and action values for the pollutant transition soil-crop on arable land and in kitchen gardens as well as on grassland. Only the measure value for Cu on grassland can be used for this project (Table 4). With the limitation that an adequate methodological basis was only available for Pb and Cd, Prinz and Bachmann [23] additionally considered the heavy metal transport from soil to groundwater. The underlying regression analyses were performed only on data sets of arable soils for Pb and Cd; therefore, the supplementary consideration of Prinz and Bachmann [23] was limited to these two elements only.
In a very simplified way, the approach of Prinz and Bachmann [23] for Cd could also be applied to Cr, Cu, and Zn. Here, it was assumed that the input is equal to the output, i.e., uptake by vegetation is neglected. Prinz and Bachmann [23] set the determination of the leachate quantity at a flat rate of 30% of the precipitation quantity. Since a calculation of the leachate quantities for the determination of critical loads for groundwater protection was carried out on the basis of land-use-differentiated reference values depending on soil permeability and slope, it was better to fall back on these values here, especially since the leachate share of the annual precipitation total in Germany of 11-42% corresponds quite well with the average value of 30% [23].
In Luxembourg, the annual precipitation sum in the 30 year average (1971-2000) is 871 mm a −1 on average, according to Geo-Portal Luxembourg [30]. The standard deviation within the country is 36 mm a −1 , so that the use of the mean value is sufficient for this rough estimate. The seepage fraction in Luxembourg's soils [31]  The test values, with which the seepage rate was multiplied, were taken from Table 8.

Critical Loads
In this paper, the methodological approach for calculating critical loads for heavy metals follows the recommendations in the ICP Modelling and Mapping [32] (Chapter V.5).
Here, all relevant fluxes into or out of a specific soil layer, in which the main substance transformations occur or in which the receptors have their distribution focus and which is therefore relevant for the effects in the system, are compared. The consideration of heavy metal fluxes, stocks, and concentrations refer to the mobile or potentially mobilizable metals, and only they are relevant for the consideration of the substance fluxes. The mass balance equation includes as discharge pathways from the terrestrial ecosystem the uptake into the biomass with subsequent harvesting and the discharge with the leachate flux as follows: with: In accordance with the recommendations of the Expert Panel for Heavy Metals to the ICP Modelling and Mapping [33,34], there have been no changes to this approach since 2004 [35,36].

Harvest Withdrawal of Heavy Metals
The removal rate of heavy metals with biomass harvesting is derived from the yield of the biomass to be harvested and multiplied by the substance content as follows: With regard to forests, it can be assumed that in the long term that the conversion to near-natural forest management, which has already begun nationwide, in combination with the trend decrease in nitrogen inputs, regulates the potential wood yield expectation as well as the substance contents to a sustainable stable equilibrium. Therefore, conservative assumptions are made for yield and content estimation, derived from measured data at more or less unpolluted sites [5,15]. The distribution of the main tree species was derived as a rough generalization from the soil types of the 1:100,000 soil map of Luxembourg [31], including climate data [30] and elevation levels [31].
Crop yields for intensive agriculture are taken from Luxembourg's 2010-2018 crop statistics. Unlike forest management, there are no discernible trends of extensification in arable farming (except for organic farming, but its share of land is small). However, with regard to crop rotations, i.e., for the cultivation ratios of the individual crop types, it is assumed that the rules of good professional practice (in particular phytosanitary favourable, nutrient-effective and soil-preserving crop rotations) are applied. In the following, this must be assumed, since the critical loads are intended to apply in the long term. For the future, the cropping structure should be assumed to be in accordance with good professional practice in the long term. The crop rotations are derived in rough generalization from the soil types of the soil map 1:100,000 of Luxembourg [31], taking into account the climate data [30] and the altitude levels [31].
The estimation of dry matter yield in utilized grassland habitats assumes extensive use (2-3 ash mowing or grazing predominantly with cattle) with only stand-maintaining fertilizer applications. For dry slope grasslands, minimum conservation use or maintenance use (mowing or grazing primarily with sheep and goats) is assumed to prevent natural scrub encroachment. However, this necessary minimum use also depends on the biomass production potential of the respective site. The grassland types are derived as a rough generalization from the soil types of the soil map 1:100,000 of Luxembourg [31], taking into account the climate data [30] and the altitude levels [31]. Yield determination was performed according to the method of Schlutow et al. [5,15].
The annual heavy metal removal (M u ) for exploited forests is derived from the estimated biomass removal by the annual increment of rough wood and bark of the main and secondary tree species of the current stand at the site, multiplied by the average contents of heavy metals in rough wood and bark (Table 9). These contents can be regarded as sustainably tolerable and thus acceptable in the long term, since only measured values from areas not specifically contaminated were evaluated for this purpose. The M u for used grassland biotopes and arable crops results from the growth rate of above-ground grass mass in the year (dry matter) and the heavy metal contents in the harvested mass (from studies without specific contamination) according to Table 9. Table 9. Heavy metal contents (mg kg −1 ) in the dry matter of rough wood with bark of the main tree species, of arable crops and grassland [37][38][39].

Discharge of Heavy Metals with Water Runoff
The basic information for determining the water runoff from the soil layer under consideration is provided by the map of annual precipitation totals in the 30 year mean of the years 1971-2000 [30]. From the map of annual precipitation totals, 4 significantly different zones emerge.

•
Southeast: from 744 to <800 mm a −1 • Central South: from 800 to <850 mm a −1 • North and southwest: from 850 to <900 mm a −1 • Extreme southwest and west (2 small areas on the border): from 900 to 967 mm a −1 .
The seepage rate Q le(z) (subsurface runoff) results from the difference of precipitation, minus evapotranspiration rate and surface runoff. The calculation of total runoff is based on the methodology of Renger et al. [40]. In a rough generalization, taking into account the soil-type-specific permeability, the mean annual total evapotranspiration by vegetation according to BMVBS [41] and the surface runoff on slopes [41], the following ratio of infiltration rate of precipitation can be assumed (Table 10). The assignment of the soil forms (from soil map 1:100,000 Luxembourg [31]) to slope classes, in which the soil forms predominantly occur, was carried out by overlaying the layers of the Geoportal Luxembourg. The classification was performed as follows: Slope 0-10 • -flat, 10-20 • -undulating, and 20-30 • -dome-shaped >30 • -sloping. It was assumed that the protection of groundwater with regard to the exceeding of drinking water limits by anthropogenic pollutant inputs is guaranteed if the limits are not exceeded in the leachate directly below the root zone. Possible interactions of the leached metals with exchange sites in deeper layers of the water-saturated soil zone are neglected.
Soil microorganisms, invertebrates, and sensitive plant species of the herb layer are predominantly distributed or rooted in the more humus-rich O and A horizons. Therefore, for CL(M) eco and CL(Cd) food , the lower-thickness biologically active soil layer (zb) was considered, where water runoff (referred to here as soil water) is higher. The difference water flow to the seepage flow below the root zone Q le(z) is absorbed in the deeper soil layers by plant roots and is subject to transpiration. Q le(zb) was calculated as follows: where by: Q le(zb) = Seepage rate below the biologically active soil horizons (zb); Q le(z) = Seepage rate below the total rooted soil layer (z); P = Precipitation (30 year average of annual precipitation totals); f ET(zb) = Factor for determining the proportion of evapo-transpiration from the biologically active soil layer (zb); f i(zb) = Factor for calculating the shares of interception in annual precipitation.
The following generalizing assumptions were made [35]:

Critical Concentrations for the Protection of Human Health
In order to protect the groundwater as a drinking water reservoir, the limit values for heavy metal contents can be found in the Grand-Ducal Regulation on the Quality of Water Intended for Human Consumption in Luxembourg [42]. Currently, there are various legal limits or guideline values for the concentration of heavy metals in drinking water worldwide. An overview is given in Table 11. The critical limits for heavy metals in drinking water [M] crit(drink) as given in the Mapping Manual [32,35,36] with reference to the WHO guideline [43] for Pb, As, and Cr correspond to the limits of the currently valid drinking water regulations for Luxembourg [42] and Germany [45]. The Cd limit in Luxembourg is higher than according to WHO and in Germany, while the Cu limit is lower. Therefore, the respective lower limit concentrations were applied in this study (marked in bold).
In Order to protect soils for the production of plant pood, the EU limit value for Cd in wheat grain of 0.2 mg kg −1 dry matter (Commission of the European Community [46] is not derived based on effects. Therefore, in this study, the Cd limit [Cd] con for wheat used according to the recommendation of the Manual of ICP Modelling and Mapping [32,35,36] instead of the EU regulation [47] (Table 12) [34]. Since the concentration (critical limit) for the plant is given, the critical concentration in the soil solution [Cd] crit(food) can be determined iteratively with transfer functions according to Römpkens et al. [48].

Critical Concentrations (Critical Limits) for the Protection of Ecosystems and Biodiversity
The ecotoxicological effect of heavy metal ions depends on their concentration in soil water, since only free active ions are taken up into the biomass and thus interact with organisms. In a Europe-wide survey on CL(M), critical limits for a number of heavy metals were compiled from the literature in 2006/07 [39,49]. Determination of the total critical concentration of heavy metals in soil water with effect on soil microorganisms, invertebrates, and plants must be performed for each heavy metal under consideration according to its chemical properties using different approaches as follows: For a number of heavy metals (Cd 2+ , Pb 2+ , Cu 2+ , Zn 2+ , and Ni 2+ ), toxicity is highly dependent on the simultaneous presence of nontoxic cations (Na + , Ca 2+ , H + ), which limit the uptake of the toxic heavy metals into organisms and thus protect the organisms. The concentration of protective competing cations is closely correlated with pH values. Thus, the concentration of free heavy metal ions is the following function of soil water pH in connection with Table 13 [50]: [M] crit( f ree) = 10 α·pH+γ (5) Metals occur in soil water not only as free ions but also in the form of soluble complexes. Manual [32] (Chapter V.5) recommends that the transformation be performed using a chemical speciation model, e.g., the Windemere Humic Aqueous Model, WHAM [52,53]. This model (version 6) was specifically adapted to meet the requirements of the critical limit derivation for soils (W6S-MTC2). The critical concentrations of metals in leachate used in the calculation of critical loads for ecosystem protection in this study are consistent with those specified in Reinds et al. [39]. Accordingly, the total critical concentrations for Cu, Ni, Zn, Pb, and Cd were calculated based on models differentiated by their bioavailability as a function of soil-specific pH and organic matter and dissolved organic carbon content (see also [32]). The modelling was based on PNEC values (for As and Cu) or on NOEC values (for Cr, Ni, Zn, Pb, and Cd).

Soil-Man Impact Pathway
The calculation for Luxembourg analogous to Prinz and Bachmann [23] results in a precipitation-limiting value for Cr total in settlement areas of 219 g ha −1 a −1 (Table 14). The calculation for Luxembourg analogous to Prinz and Bachmann [23] results in precipitation-limiting values for copper in sheep pastures of 1117 g ha −1 a −1 and in other grassland of 8267 g ha −1 a −1 (Table 15).

Soil-Groundwater Impact Pathway
The result of the rough estimate of the precipitation-limiting assessment values for the protection of groundwater as a drinking water reservoir is shown in Table 16.

Protection of Human Health
The results of the determination of critical loads for the protection of human health through compliance with the drinking water limit concentrations in groundwater under the different vegetation types (arable land, grassland, deciduous forest, coniferous forest, and mixed forest) is differentiated according to the 27 soil types of the soil map 1:100,000 Luxembourg [31] are presented in Table 17.   The results of the determination of critical loads for Cd in wheat and wheat products for the protection of human health are presented for the 27 soil types of the soil map 1:100,000 Luxembourg [31] differentiated in Table 18. Table 18. Critical loads for Cd in wheat and wheat products for the protection of human health in Luxembourg.

Protection of Ecosystems
The results of the determination of critical loads for the protection of plants, soil invertebrates, and microorganisms for the different vegetation types (arable, grassland, deciduous, coniferous, and mixed forest), differentiated according to the 27 soil types of the Soil Map 1:100,000 Luxembourg [31] are presented in (Table 19).

Discussion
The assessment values for the protection of human health are based on different criteria. While the values of the TA Luft [14], the 39th BImSchV [12] and the EU Position Paper [8] (from which the assessment values of most other laws and regulations were derived) as well as the value for settlements according to Prinz and Bachmann [23] focus on the protection of humans in case of direct contact with soil in settlement areas, the assessment values for the soil-groundwater impact pathway of the BBodSchV [13] as well as the CL(M) drink consider the protection of humans in case of ingestion of drinking water from the groundwater reservoir. The CL(M) food considers the protection of plants for human consumption. Due to different pedo-transfer processes in soil, from soil into plants, and from soil into groundwater, the results can only be compared to a limited extent (Table 20). Nevertheless, it is noticeable that the assessment values for Cd, As, Ni, and Cr are close to each other several times. The assessment value for Cr (total) on settlement areas (children's play areas) calculated according to Prinz and Bachmann [23] for the soil-human impact pathway on the basis of the test value in relation to the background levels in Luxembourg corresponds exactly to the value in the draft TA Luft [24]. The calculation results of Prinz and Bachmann [23] show a range of 16-70 µg m −2 d −1 for Germany, so that the value for Luxembourg is validated as plausible.
The assessment value for Cu on sheep pastures for the soil-plant-animal pathway based on the measure value relative to background levels in Luxembourg (306 µg m −2 d −1 ) is in the range of 230-335 µg m −2 d −1 calculated by Prinz and Bachmann [23] for Germany. Likewise, the assessment value for other grassland for the soil-plant effect pathway with 2264 µg m −2 d −1 is within the range of 2189-2294 µg m −2 d −1 for Germany of Prinz and Bachmann [23].
Prinz and Bachmann [23] did not determine any assessment values for the soilgroundwater impact pathway. A comparison with this is therefore not possible. If the calculation results calculated according to the method of Prinz and Bachmann [23] (Section 2.1) are compared with the critical loads for the protection of groundwater as a drinking water reservoir (Section 2.2), there is an exact agreement for Cr as a whole. For Cu, the critical loads are about 40× higher and for Zn about 10× higher than the calculation results, according to the method of Prinz and Bachmann [23].
The reasons are obvious. For example, the Drinking Water Ordinance for Germany [45] contains a 40× higher limit concentration for Cu. The limit concentration for Zn (according to Health Canada [44]) is ten times higher than the test value according to BBodSchV Germany [13].
Assessment values for the protection of plants, animals, biodiversity, and ecosystems (Table 21) as a whole are based in the legal and sublegal regulations on the assumption that human toxicological threshold values protect ecosystems to a sufficient degree. Ecotoxicological thresholds do not underlie these assessment values. Thus, the assessment values are also largely identical to those for the protection of human health. A comparison with the critical loads shows that this thesis does not apply in every case. The CL(M) eco are based exclusively on ecotoxicological threshold values (PNEC, LOEC, and NOEC). The comparison of the critical loads in  shows that the sensitivity of humans cannot be equated with the sensitivity of ecosystems with their plants, animals, and microorganisms. Thus, ecosystems are significantly more sensitive to Pb, Cu, and Zn inputs than humans. In particular, Cu and Zn, as essential trace elements for humans, are rather insufficient in drinking water and in food crops, so that deficiency symptoms are commonly observed in humans. The situation is different for As and Cr. Here, humans react much more sensitively than ecosystems, especially to Cr(VI) compounds, e.g., chromate [54]. For Cd, the critical loads for drinking water, ecosystems, and wheat products are about the same.

Conclusions
From the comparison of existing legal regulations with assessment values calculated on an empirical basis, a number of indications for further scientific and political work in connection with the determination and application of assessment values for heavy metal discharges emerge.

1.
The assessment values of the considered recommendations, laws, and sub-legislative regulations are only conditionally comparable with each other and with calculated precipitation-related values or with the critical loads due to the methodological differences of their derivation. The differences, some of which are significant, are due to different levels of protection, protection goals, and the impact relationship.

2.
With regard to human health, other heavy metals are of immense importance, especially Hg [55], Tl [56], and Cr [57]. There is an urgent need for research on these metals. Comparing the calculation results calculated according to the method of Prinz and Bachmann [23] (Section 2.1) with the critical loads for the protection of groundwater as a drinking water reservoir (Section 2.2), there is an exact match for Cr total at 60 µg m −2 d −1 . This value was also included in the draft of the draft TA Luft as of 2016 [24]. Unfortunately, this value was not included in the current TA Luft [14]. Previously regulated assessment values for Zn, on the other hand, are rather superfluous, because for people in Europe there is rather a zinc deficiency than a toxic overdose.

3.
The assessment values for depositions of dusts containing heavy metals, as given in the Flemish Ordinance on Environmental Permitting [17], in the Swiss Air Pollution Control Ordinance [16], in the German Federal Soil Protection Ordinance [13], and in the German TA Luft [14], do not or not sufficiently take into account the regional and especially the geogenic differences in accumulation. As shown by the determinations of precipitation-related values based on natural background levels in soil both in Luxembourg (see Tables 17-19) and in Germany [5], the partly strong regional differentiation cannot be neglected. The derivation of a tolerable annual total input rate from the assessment values alone is not meaningful. However, it can be calculated from the difference between the background value and the assessment value, differentiated by region, as precipitation-related assessment values, as was performed for Luxembourg using the method of Prinz and Bachmann [23].

4.
Although the precipitation-related assessment values according to the method of Prinz and Bachmann [23], such as the critical loads for heavy metals, take into account all input pathways (air, management, possibly others), they differ significantly from the critical loads in their methodological approach. They assume an acceptable increase in concentrations in the soil when precautionary values have already been exceeded, whereas CL(M) are calculated assuming an equilibrium between inputs and outputs at the concentration level of the critical limits (the impact thresholds), regardless of the current concentration in the soil. The precipitation-related assessment values therefore have only a limited precautionary character in the sense of sustainable prevention of risks of adverse effects due to pollutant accumulation. They are more comparable to a de minimis threshold or irrelevance threshold.

5.
Higher safety is provided by assessment values for acceptable additional input rates that ensure a balance with the harmless discharges (critical loads). If the balanced assessment values are observed, further enrichment beyond critical concentrations can be ruled out in the long term if they are currently undercut. If the critical concentrations are already exceeded today, a depletion can also take place under favourable conditions (tolerable discharges higher than inputs).
All assessment values set by law and sub-law for the protection of natural assets are based on human toxicological threshold values. Therefore, they are only conditionally suitable for application to ecological protected goods.
The comparison of the critical loads in Luxembourg-as well as in Germany [5]shows that the sensitivity of humans cannot be equated with the sensitivity of ecosystems with their plants, animals, and microorganisms.
For the goal of the European Biodiversity Strategy for 2030 to set ecosystem-based impact thresholds for pollutants that describe the effects on biodiversity, the critical loads for the protection of ecosystems provide a very precautionary scientific basis for discussion.

Data Availability Statement:
The data presented in this study are available on request from the corresponding author.

As
Arsenic; here limited to As(V), the stable form in aerobic environment (