The Ecotoxicity of Pesticides Used in Conventional Apple and Grapevine Production in Austria Is Much Higher for Honeybees, Birds and Earthworms than Nature-Based Substances Used in Organic Production

Abstract

It is debated whether the ecotoxicity of active substances (ASs) contained in synthetic pesticides applied in conventional agriculture (conASs) differs from nature-based ASs used in organic agriculture (orgASs). Using the official pesticide use statistics, we evaluated the ecotoxicity of ASs used in apple and grapevine production in Austria. In 2022, 49 conASs and 21 orgASs were authorized for apple production and 60 conASs and 23 orgASs were authorized for grapevine production in Austria. Based on the latest publicly available data on the actual use of pesticides in apple and grapevine production (from the year 2017), we evaluated their ecotoxicity based on information in the freely accessible Pesticide Properties and Bio-Pesticides Databases. The results showed that although the amount of ASs applied per hectare of field was higher in organic farming, the intrinsic toxicities of ASs used in conventional farming were much higher. The number of lethal toxic doses (LD₅₀) of ASs applied in conventional apple orchards was 645%, 15%, and 6011% higher for honeybees, birds, and earthworms, respectively, than in organic apple production. In conventional vineyards, lethal doses for honeybees, birds, and earthworms were 300%, 129%, and 299% higher than in organic vineyards. We conclude that promoting organic farming would therefore contribute to the better protection of biodiversity on agricultural land and beyond.

1. Introduction

Biodiversity loss in agroecosystems and beyond has many drivers, and pesticides contribute to it through their mode of action [1]. Moreover, pesticide residues are often found outside the target area, even in nature reserves set up to protect biodiversity [2,3,4,5]. Reducing pesticide use is high on the international political agenda, for example, with the European Green Deal with its ‘Farm to Fork’ and ‘Biodiversity’ strategies, which aim, among other aspects, to halve the use and risk of chemical pesticides in the European Union (EU) by 2030 and to expand organic farming to 25% of the agricultural land in the EU [6].

Agricultural systems differ in terms of the active substances (ASs) used to control pests and diseases. In integrated and conventional agriculture, ASs are used that are contained in synthetic chemical pesticides [7], the production of which releases large amounts of greenhouse gases [8], whereas the ASs used in organic farming are derived from plant extracts, pheromones, microorganisms or elements [9]. Most synthetic ASs act through a targeted interaction with biochemical processes that are important for the survival of populations of pests and pathogens [10]. In contrast, microorganisms and pheromones, as well as the vast majority of plant extracts and mineral ASs allowed in organic agriculture, exert their effect via a non-specific chemical, physicochemical, or physical mode of action or by deterring or confusing pests [11]. Regardless of the origin of the substance, an AS may only be used in a pesticide in the EU if its efficacy and safety have been assessed in environmental risk assessments (ERAs). These ERAs are conducted on non-target surrogate pollinators, such as honeybees, or pest predators, such as birds, or soil organisms, such as earthworms, as well as various other organismic groups [12]. The EU currently has the most comprehensive and protective pesticide regulations of any major agricultural producer in the world [13]. All approved ASs are listed in Annex 1 of Directive 91/414/EEC and published in the EU pesticide database [14].

The intensity of pesticide use varies greatly between individual agricultural crops. Among the most pesticide-intensive crops in the temperate zones are apples, with about 30 ASs, and grapevines, with about 17 ASs reported in 2022 from integrative/conventional farming in Germany [15]. In Italy, up to 41 ASs were reported in apple production and 38 ASs in grapevines [16]. This intensive use of pesticides is particularly problematic because both apple orchards and vineyards are structurally rich perennial agroecosystems for many species [17,18]. Pollinator insects such as honeybees are the most important pollinators in apple orchards [19] and influence both yields and fruit quality [20]. Birds play a very important role as natural pest controllers in apple orchards and vineyards, as they eat different types of insect pests and feed them to their offspring [21,22], thus providing free, environmentally friendly pest control in the landscape [23]. Earthworms improve soil fertility by converting organic matter into humus [24]. Their burrowing activity affects soil hydrology by improving soil aeration and water infiltration [25,26]. However, the use of pesticides in agroecosystems has adverse effects on non-target organisms such as insects, birds, and earthworms, and the overall biodiversity [1,4,27,28,29,30,31].

The aim of this study was to assess the potential ecotoxicological impact of ASs in conventional and organic apple and grapevine production. For this purpose, we used data on pesticide application in Austria from 2017, the most recent dataset. We assessed the potential impact on honeybees, birds, and earthworms, as these are common in European agroecosystems and are often used as surrogate species in environmental risk assessments during the approval process of ASs. This approach is novel as we used the rarely available crop-specific pesticide use data to make an evaluation on three important non-target organismic groups that have different functions in agroecosystems [32].

2. Materials and Methods

2.1. Pesticide Use Data

Firstly, we collected data on the area under apples and grapevines in Austria (

Table 1. Apple and grapevine area in Austria in 2017 and 2021. Acreage from Grüner Bericht 2022 [35]; pesticide usage data from 2017 from the Austrian Agency of Nutrition and Health AGES [33].

Crop	Farming System	Area Cultivated (ha)			Amount Applied (kg)
		2017	2021	Area Change 2017–2021 (%)	Fungicides	Herbicides	Insecticides	Others
Apple	Conventional	7675	4820	−37.2	118,749	3910	18,500	3728
	Organic	1619	1529	−5.6	74,337	0	19,887	791
Grapevine	Conventional	40,618	35,859	−11.7	783,440	14,978	8862	6
	Organic	5717	6976	+22.0	201,526	0	7074	20

). Secondly, the ASs approved in Austrian apple and grapevine production were taken from the publicly accessible Austrian plant protection product register of the Austrian Authority for Food Safety (https://psmregister.baes.gv.at; accessed on 22 February 2023). Thirdly, the dosages applied in conventional or organic apple and grapevine production were gathered from the Austrian Federal Agency of Health and Food Safety [33]. The most recent report is from 2017 and contains data on pesticide use in the field from 940 Austrian farms growing 16 different crops on a total area of 28,200 ha; however, it was not specified how many apple and grapevine farms were included in the dataset. Furthermore, most of the documented crops only consider conventional cultivation; a distinction between conventional and organic production was only available for apple and grapevine production [33]. Unfortunately, however, these data on pesticide use do not indicate the amounts of specific ASs, but only for groups of ASs, e.g., for fungicides only, the groups carbamates and dithiocarbamates, imidazole and trizoles, other organic fungicides, and inorganic fungicides (Table 1). Information on which of the approved ASs belong to which pesticide group for conventional and organic apple and grapevine productions was collected from information in VO 1185/2009 from the European Commission [34].

All authorized orgASs in Austria are summarized online under “EASY-CERT services”, a reference for organic farmers on officially certified organic plant protection products (https://www.betriebsmittelbewertung.at/; accessed on 8 August 2024). The EU pesticide database distinguishes between four paraffin oils (CAS 64742-46-7, CAS 72623-86-0, CAS 8042-47-5 and CAS 97862-82-3; [14]), while the Austrian plant protection register summarizes them as one AS: paraffin oil. The five fatty acids (methyl decanoate, methyl octanoate, caprylic acid, lauric acid, and oleic acid) differentiated in the EU pesticide database are also listed as one AS group in Austria: fatty acids. The various Bacillus thuringiensis subsp. Kurstaki strains (ABTS 351, PB 54, SA 11, SA12, and EG 2348) are also listed as one AS in Austria. Thus, 49 conASs authorized for conventional apple cultivation and 21 orgASs for organic apple cultivation were considered in our analysis.

In addition, the orgAS Metarhizium brunneum strain Ma 43, which is authorized for grapevine production, was not included because it should be mixed into the soil before planting and ecotoxicological data were missing. Lambda-cyhalothrin, a conAS in pre-planting treatments and indolylbutyric acid, which is only applied under glass, were also not included in the assessment regarding grapevine production. Our analysis therefore included 60 conASs and 23 orgASs that are approved for use in grapevine production in Austria.

We did not include neonicotinoids and organophosphates in our assessment, as these have been banned for outdoor use in Europe since 2018/2019 and not a single AS from these groups is authorized for use in apple and grapevine production in Austria. Therefore, this exclusion reflects current legislation and ensures that our study only focuses on substances that are currently approved.

2.2. Potential Toxic Loads in Conventional vs. Organic Apple and Grapevine Production

The toxicity data of the ASs were taken from the Pesticide Properties Database and the Bio-Pesticides Database [36]. Since the use data were only available for chemical groups, but the LD₅₀ data were for single ASs, we had to assign each AS authorized for apple and grapevine production to the corresponding chemical group to calculate the mean LD₅₀ of a chemical group. Which AS belongs to which chemical group is defined in Regulation no. 1185/2009 [34]. Substances that we could not assign to a specific group in the regulation were categorized as ‘other organic fungicide’, ‘other organic herbicide’ or ‘other organic fungicide’ based on their mode of action.

We extracted ecotoxicity data for honeybees (Apis mellifera), birds (Colinus virginianus, Corturnix japonica and Anas platyrhynchos), and earthworms (Eisenia fetida). Specific parameters that we extracted from the PPDB and the BPDB were the contact acute LD₅₀ for honeybees (µg bee⁻¹), acute dietary LD₅₀ for birds (mg kg⁻¹), and the LC₅₀ (mg kg soil⁻¹) for earthworms. LD₅₀ values in the database with a ‘>’ sign were processed as equal to the value, thus likely overestimating their potential ecotoxicity.

An overview of the data acquisition and calculations is provided in Figure 1.

Honeybee TL was calculated by dividing the amount applied per crop by the corresponding acute LD₅₀ values for honeybees [37,38].

Bird TL was calculated for the European serin as the reference species, as it occurs in agroecosystems and is in sharp decline in Austria [39]. As the LD₅₀ values in the pesticide databases include those for a range of bird species like Colinus virginianus, Coturnix japonica, or Anas platyrhynchos [36], we used Equation (1) to calculate the LD₅₀ of the European serin for each AS [40]:

$$\log \left({\mathrm{LD}}_{50}2\right)=\log \left({\mathrm{LD}}_{50}1\right)+\left(\log \left(\mathrm{W}2\right)-\log \left(\mathrm{W}1\right)\right)\ast 1.239$$

(1)

W1 is the reference bodyweight of the surrogate species based on the literature. Thus, we had 3 different values for W1 based on the bird species in the PPDB. This means that for our analysis, we chose either the mean bodyweight of 189 g of C. virginianus (167–214 g) (Mach, 2002), 95 g of C. japonica (90–100 g) [41], or 1115.5 g of A. platyrhynchos (967–1264 g) [42]. The LD₅₀1 is the original LD₅₀ given in the PPDB for each AS. The LD₅₀2 and the W2 are the estimated LD₅₀ and reference weight of the European serin. Based on the literature, we have assumed an average body weight of 12 g (=W2) [43]. For the specific example of spinosad, with an LD₅₀ of 2231 mg per A. platyrhynchos (2000 mg kg bodyweight⁻¹), this results in an LD₅₀ estimate for the European serin of 8.97 mg bird⁻¹.

We used a scaling factor of 1.239 to refine the extrapolation of the ecotoxicity in birds based on the weight based on 130 pesticides [44].

To calculate the TL for earthworms, we converted the LC₅₀ values to LD₅₀ values by dividing the LC₅₀ value of each AS by the average mass of earthworms in Austrian arable farmland [37] using Equation (2).

$${\mathrm{LD}}_{50 \mathrm{worm}}={\displaystyle \frac{{\mathrm{LC}}_{50 \mathrm{worm}}}{{\mathrm{n}}_{\mathrm{worms}}{ \mathrm{per} \mathrm{kg}}_{\mathrm{soil}}}}$$

(2)

Firstly, we calculated the mass in kg by multiplying the volume of one m² by the soil density. The depth for the volume calculation was 0.28 m. We therefore multiplied 2.8 × 10⁵ cm⁻³ by an assumed soil density of 1.52 g cm⁻³. Finally, we divided the earthworm density by the total mass. Earthworm density was taken from two Austrian field studies [45,46] with 68 subplots: 208 ± 90 earthworms m⁻² (nworms). The determined number of earthworms kg⁻¹ soil was 0.49 kg⁻¹ (Equation (2)).

As no application data are available for individual ASs, we have calculated the TLs on the basis of the manufacturers’ application recommendations. For the calculation, we used the highest recommended dosages given in the safety data sheets of the authorized products for organic and conventional apple and grapevine cultivation. We extracted data on the amount of product per hectare in l ha⁻¹ or kg ha⁻¹ and the amount of AS in the product in g l⁻¹ or g kg⁻¹.

Finally, we obtained the amount of pure AS ha⁻¹ by multiplying the amount of the AS by the recommended dosage of the pesticide product (Equation (3)).

$$\mathrm{Amount} \mathrm{AS} \left(\mathrm{kg}/\mathrm{ha}\right)={\displaystyle \frac{\mathrm{pure} \mathrm{AS} \left({\mathrm{g} \mathrm{l}}^{-1}{\mathrm{or} \mathrm{g} \mathrm{kg}}^{-1}\right)\ast \mathrm{recomm}. \mathrm{dosage} \mathrm{of} \mathrm{pesticide} \mathrm{product} \left({\mathrm{l} \mathrm{ha}}^{-1}{\mathrm{or} \mathrm{kg} \mathrm{ha}}^{-1}\right)}{1000}}\times \mathrm{application} \mathrm{frequency}$$

(3)

As some ASs need to be applied more frequently during the season, we multiplied them by the maximum allowable application frequency. If there were products with the same AS, we calculated the average amount ha⁻¹ for each AS. The number in our calculation is therefore the recommended amount of AS ha⁻¹. Finally, we multiplied the amount of AS ha⁻¹ by the total acreage per crop in Austria (Table 1).

For each AS, we extrapolated application data based on the manufacturers’ recommendations and calculated the toxic load by dividing the quantity by the corresponding LD₅₀ value [47] and finally determined the toxic loads of each AS used for wine and apple cultivation. As the preliminary analyses showed that the highest toxic load in apple orchards is delivered to honeybees and in vineyards to European serins, we focused our specific calculations on these species. Thus, for each AS authorized for apple cultivation, the maximum number of acute LD₅₀ doses (based on contact acute LD₅₀ µg bee⁻¹) delivered to A. mellifera per year and the maximum number of acute LD₅₀ doses (based on acute LD₅₀ mg kg⁻¹) delivered to S. serinus per year of each AS approved for grapevine cultivation were determined.

Even though the LD₅₀ captures lethality, the calculated number of LD₅₀ doses per AS does not correspond to the actual number of non-target organisms killed. It is important to note that the effects may not be directly lethal, but also indirect and sublethal [48], or that there may be no effects at all. Furthermore, we have not considered interactions between the different substances or products in the pesticide cocktail itself, as there is insufficient data on such interactions for a well-founded analysis.

3. Results

3.1. Pesticides Used in Apple Production in Austria and Their Toxic Loads

In the year 2022, 49 conASs and 21 orgASs were authorized for apple cultivation in Austria. The conASs were dominated by fungicides (43% of conAS), herbicides (27%), plant regulators (14%), and insecticides (10%). The orgASs were dominated by insecticides (67% of orgAS) and fungicides (29%); no plant regulators or herbicides were authorized for organic apple production. One conAS (2%) and one orgAS (5%) was authorized as acaricides. One conAS was authorized as a molluscicide and one conAS was authorized as a rodenticide (each 2%) (Figure 2).

Fungicides were authorized to control grey mold (Botrytis cinerea), grey mold (Botryotinia fuckeliana), bitter rot (Gloeosporium spp.), Nectria canker (Nectria galligena), fruit rot (Monilia fructigena), apple collar rot (Phytophthora cactorum), apple powdery mildew (Podosphaera leucotricha), blue mold (Penicillium expansum), flyspeck (Schizothyrium pomi), and apple scab (Venturia inaequalis). Insecticides were authorized to control summer fruit tortrix moth (Adoxophyes orana), apple blossom weevil (Anthonomus pomorum), apple aphids (Aphis spp.), fruit tree tortix moth (Archips podana), apple-Campylomma bug (Campylomma verbasci), apple coddling moth (Cydia pomonella), apple mites (Eriophyes spp.), apple sawfly (Hoplocampa testudinea), winter moth (Operophtera brumata), apple brown tortix (Pandemis heparana), apple sucker (Psylla mali), Rhynchites weevil (Rhynchites aequatus), and apple spider mites (Tetranychus spp.). Herbicides were authorized to control monocotyledonous and dicotyledonous plant species under and between apple trees.

The ASs used in Austrian apple production in 2017 contributed to a potential toxic load (TL) to honeybees of 4.1 × 10¹² LD₅₀ doses in conventional apple production versus 0.55 × 10¹² LD₅₀ in organic apple production (Figure 3A).

In conventional apple production, about 1.1 × 10¹¹ lethal doses to honeybees came from carbamates, 5.8 × 10⁹ LD₅₀ from imidazoles and triazoles, 5.2 × 10¹¹ LD₅₀ from inorganic pesticides, and 2.9 × 10¹¹ LD₅₀ from other fungicides. Insecticides in conventional production included pyrethroids (the single substance lambda-cyhalothrin) with the highest TL of 9.5 × 10¹¹ LD₅₀. Other insecticidal classes in conventional production included organophosphorus insecticides (TL = 3.4 × 10¹¹ LD₅₀ doses) and carbamates (chlorantraniliprole and the oximcarbamates pirimicarb; TL = 1.2 × 10¹⁰ LD₅₀ doses). The TL for honeybees from herbicides was 3.5 × 10¹⁰ LD₅₀ (Figure 3A). In organic apple production, the TL for honeybees resulted from inorganic fungicides (5.3 × 10¹¹ LD₅₀) (Figure 3A). The TL of insecticides used in organic farming came from biological insecticides (8.6 × 10⁸ LD₅₀ doses).

Pesticide use in Austrian apple production contributed a TL of 1.5 × 10¹⁰ LD₅₀ doses for birds (S. serinus) per year in conventional and 1.41 × 10¹⁰ LD₅₀ doses in organic apple production (Figure 3B). In conventional production, fungicides contributed the highest TL, with a total 1.3 × 10¹⁰ LD₅₀ doses (including carbamates, inorganic fungicides, other fungicides, imidazoles, and triazoles). Insecticides used in conventional apple fields contributed 1.6 × 10⁹ LD₅₀ doses. Herbicides and all other ASs contributed lower LD₅₀ doses to birds than did fungicides and insecticides. In organic apple production, 1.4 × 10¹⁰ LD₅₀ doses were delivered through inorganic fungicides, 1.3 × 10⁶ LD₅₀ doses through biological and botanical product-based insecticides, and 1.1 × 10⁸ LD₅₀ doses through mineral and vegetal oils (Figure 3B).

The TL for earthworms in apple production was 2.2 × 10⁹ LD₅₀ through conASs and was 0.037 × 10⁹ LD₅₀ for orgASs (Figure 3C). In conventional production, the highest TL to earthworms was delivered by fungicides: 0.021 × 10¹¹ LD₅₀ doses from carbamates and dithiocarbamates, 3.4 × 10⁷ LD₅₀ doses from inorganic fungicides, 2.7 × 10⁷ LD₅₀ from other fungicides, and 4.3 × 10⁵ LD₅₀ doses from imidazoles und triazoles. In summary, a potential 2.16 × 10⁹ LD₅₀ doses for earthworms were delivered from fungicides in conventional production, while 0.0344 × 10⁹ LD₅₀ doses were delivered in organic production. Insecticides (TL = 9.2 × 10⁵ LD₅₀), herbicides (1.6 × 10⁶ LD₅₀), plant growth regulators (4.5 × 10⁵ LD₅₀), and rodenticides (3 × 10³ LD₅₀) delivered much lower TLs to earthworms than did fungicides (Figure 3C). In organic apple production, fungicides delivered 6.5 × 10⁴ LD₅₀: 2.8 × 10⁶ through mineral and vegetal oils, 4.3 × 10⁴ through biological and botanical insecticides, and 3.4 × 10⁷ through inorganic fungicides. Insecticides delivered 3.7 × 10⁷ LD₅₀ doses (Figure 3C).

Considering honeybees, earthworms, and birds together, a total of 4.1 × 10¹² LD₅₀ doses per year were delivered under conventional and 0.56 × 10¹² LD₅₀ doses under organic apple production (Figure 3D).

All LD₅₀ values used to calculate the toxic loads can be found in Supplementary Table S1.

3.2. Specific ASs Used in Apple Production in Austria Affecting Honeybees

Honeybees were the organismic group that potentially received the highest TL from apple production. In conventional apple production, the highest TL for honeybees was caused by the insecticide emamectin, with 7.9 × 10¹² LD₅₀ doses, which corresponds to 46% of the total TL of all the conASs (Figure 4). The second most toxic conAS, accounting for 39% of the total TL, comes from the insecticidal AS milbemectin (6.7 × 10¹² LD₅₀ doses). The third most toxic conAS contribution of 6% to the total TL was the insecticide lambda-cyhalothrin (9.5 × 10¹¹ LD₅₀ doses). Pirimicarb, fosetyl, and penconazole each contributed less than 2% to the total TL (Figure 4). The ten most hazardous ASs for honeybees in conventional apple production also included two fungicides, two herbicides, and one plant growth regulator. The contribution of the herbicidal ASs was less than 1% (Figure 4).

In organic apple production, 70% of the total TL for honeybees was accounted for by fatty acids (4.7 × 10¹² LD₅₀ doses), 15% of the TL by the fungicide sulphur (9.8 × 10¹¹ LD₅₀), and 5.0% of the TL by paraffin oils authorized as acaricides (Figure 4). The contribution of the three fungicides copper oxychloride (3%), potassium hydrogen carbonate (2%), and copper hydroxide (2%) followed in their share of the total TL in organic production (Figure 4).

Supplementary Table S2 provides a detailed list of the contributions to the total TL of the individual ASs.

3.3. Pesticides Used in Grapevine Production in Austria and Their Toxic Loads

A total of 60 conASs and 23 orgASs were authorized for grapevine production in Austria (Figure 5). ConASs were dominated by fungicides (72% of conAS), herbicides (13%), insecticides (7%), and plant growth regulators (3%). Molluscicides, rodenticides, and acaricides each made up 2% of the conASs. The orgASs were dominated by fungicides (47% of orgASs) and insecticides (43%); acaricides and rodenticides delivered 5% to orgASs.

3.4. Potential Toxic Loads in Austrian Grapevine Production

In conventional grapevine production, 7.7 × 10⁹ LD₅₀ doses were delivered to honeybees by the use of fungicides. Within the fungicides, the highest TL was attributable to inorganic fungicides (6.1 × 10⁹ LD₅₀), other fungicides (1.1 × 10⁹ LD₅₀), and carbamates (0.31 × 10⁹ LD₅₀). Herbicides in conventional vineyards contributed 0.15 × 10⁹ LD₅₀ doses to honeybees; pyrethroid insecticides accounted for 0.16 × 10⁹ LD₅₀ doses. In organic grapevine production, the entire TL came from inorganic fungicides (2.0 × 10⁹ LD₅₀).

The potential TL released to birds was 1.1 × 10¹¹ LD₅₀ doses in conventional compared to 0.48 × 10¹¹ LD₅₀ in organic grapevine production. In conventional production, 87% of the TL caused by fungicides was from inorganic fungicides (0.98 × 10¹¹ LD₅₀), 9% from other fungicides (0.1 × 10¹¹), 3% from carbamates (0.028 × 10¹¹ LD₅₀), 1% from imidazoles and triazoles (0.0096 × 10¹¹ LD₅₀), 0.3% from morpholines (0.0038 × 10¹¹ LD₅₀), and 0.1% from biological- and botanical-based fungicides (0.00051 × 10¹¹ LD₅₀). A proportion of 97% of the TL caused by herbicides (1.1 × 10⁹ LD₅₀) was from organophosphorus herbicides (1.1 × 10⁹ LD₅₀), 2% was from phenoxy-phytohormones (2.1 × 10⁷ LD₅₀), 1% was from other herbicides (1.2 × 10⁷ LD₅₀), and 0.2% was from derivates of urea, uracil, or sulfonylurea (1.4 × 10⁶ LD₅₀). The TL caused by plant growth regulators in conventional viticulture equated to 5.5 × 10⁵ LD₅₀ doses (Figure 6C). Only inorganic fungicides were used in organic viticulture, as no other pesticide classes were used in 2017 (Figure 6C).

In conventional grapevine production, 4.8 × 10⁸ LD₅₀ doses were delivered to earthworms from fungicides: 3.4 × 10⁸ LD₅₀ doses were via inorganic fungicides, 1.0 × 10⁸ LD₅₀ doses were via other fungicides, 0.28 × 10⁸ LD₅₀ doses were via carbamates, 0.082 × 10⁸ LD₅₀ doses were via imidazoles and triazoles, 0.051 × 10⁸ LD₅₀ doses were via morpholines, and 9.9 × 10⁴ LD₅₀ doses to earthworms were via biological- and botanical-based fungicides. Insecticide use delivered 0.000059 × 10⁸ LD₅₀ doses in conventional viticulture. Herbicidal ASs contributed 0.024 × 10⁸ LD₅₀ doses to earthworms in conventional grapevine production. Organic grapevine production delivered 1.2 × 10⁸ LD₅₀ doses via inorganic fungicides. No insecticidal or herbicidal orgASs were used in organic vineyards in Austria in 2017 (Figure 6C).

Considering honeybees, earthworms, and birds, a total of 1.2 × 10¹¹ LD₅₀ doses per year were delivered in conventional grapevine production, whereas 0.5 × 10¹¹ LD₅₀ doses were delivered in organic production (Figure 6D).

The LD₅₀ values on the basis of which the toxic loads were calculated can be found in Supplementary Table S3.

3.5. Specific ASs Used in Vineyards in Austria Affecting Birds

In conventional grapevine production, the most toxic AS for birds was a fungicide from the chemical class of quinones (dithianone) with 4.2 × 10¹⁰ LD₅₀ doses and a share of 14% of the total TL (Figure 7). The conAS zinc phosphide, mainly used to control voles, contributed 12% of the total TL (3.5 × 10¹⁰ LD₅₀) because it had one of the lowest LD₅₀ doses for birds (Colinus virginianus), 11% (3.4 × 10¹⁰ LD₅₀) of the total TL was contributed by the conASs potassium phosphonates (inorganic fungicides). In total, 7 of the 10 most toxic conASs were fungicides, only 1 AS was a herbicide, 1 was a rodenticide, and 1 was a molluscicide (Figure 7).

In organic grapevine production, 6 of the 10 most toxic ASs were fungicides, while only 2 insecticides, 1 acaricide, and 2 molluscicides are placed in the top 10. The top five ASs of the ranking accounted for 99% of the total toxic load for birds in viticulture. The orgAS with the highest TL for birds was tribasic copper sulphate (2.1 × 10¹¹ LD₅₀ doses), contributing 65% of the total TL. Ranking second was copper oxychloride, contributing 21% of the TL (6.8 × 10¹⁰ LD₅₀ doses), thirdly ranked, with 8%, was the orgAS copper hydroxide, delivering 2.5 × 10¹⁰ LD₅₀ doses to birds per year in Austrian organic vineyards. The contribution of sulphur to the total TL was 4%, with 1.2 × 10¹⁰ LD₅₀ doses. The fifth highest ranking orgAS contributed less than 1% of the total toxic load (Figure 7).

The top ten Ass are listed in Supplementary Table S4.

4. Discussion

When analyzing the ASs used in conventional or organic apple and grapevine production in Austria, we found that the overall ecotoxicity was substantially higher in conventional than in organic production: the ecotoxicity was 632% higher in apple and 140% higher for honeybees, birds, and earthworms in conventional than in organic production. Overall, the number of LD₅₀ doses released was lowest for earthworms compared to birds and bees in Austrian apple and vine cultivation. This result was independent of the production system (org and con). Honeybees received the highest toxic loads from ASs used in apple cultivation and birds from ASs in grapevine production. Although our analysis was based on ASs used in Austrian farms, all of these ASs are also authorized in other countries of a similar agroclimatic zone as that in the European Union [49,50]. Our finding of higher ecotoxicological hazards from ASs used in conventional agriculture is consistent with the results of an EU-wide analysis of ecotoxicological harm [51], as well as aquatic toxicity and effects on human health [52].

In addition to the various ASs used in conventional and organic agriculture, plant protection in organic farming also differs conceptually and practically from that in conventional farming. In contrast to conventional farming, preventive measures form the basis of measures in organic farming and all practises are regularly monitored [53]. Organic farming regulations allow the use of pesticide ASs only as a last resort and include provisions on increasing biodiversity, crop rotation, soil conservation, and health and area-based animal husbandry [53]. Indeed, only around 5% to 10% of organically farmed land is actually treated with ASs [54]. Furthermore, organic agriculture benefits the overall biodiversity [55,56]. Although our analysis has shown that orgASs have a much lower hazard potential than conASs, a further reduction in these novel substances in our environment seems imperative [57].

4.1. Potential Ecotoxicological Impact of Apple Production in Austria

In 2017, conventional apple farms in Austria applied 18.9 kg ha⁻¹ of ASs, and organic farms applied 58.7 kg ha⁻¹. Nevertheless, 8 times as many LD₅₀ doses were delivered to honeybees, birds, and earthworms when conASs were applied (4.1 × 10¹² LD₅₀ doses) than with the application of orgASs (0.56 × 10¹² LD₅₀ doses). The largest differences in ecotoxicity between conASs and orgASs were found for earthworms (61 times more LD₅₀ doses from conASs than from orgASs), followed by honeybees (7.5 times more LD₅₀ doses from conASs than from orgASs) and birds (15% more LD₅₀ doses from conASs than from orgASs).

More than twice as many ASs were approved for conventional apple production (49 conASs) than for organic apple production (21 orgASs). In addition, all orgASs can also be used in conventional farming, but not vice versa. Overall, fungicides, herbicides, plant growth regulators, and insecticides dominate conventional apple production, while insecticides and fungicides dominate organic apple production. This difference is counterintuitive, since conventionally and organically produced apples are threatened by the same pests and diseases.

The predominant use of fungicides in the apple orchards was typical for this perennial crop [58,59]. Without knowing the predominant pests and diseases, we assume that fungicides were most likely used against apple scab, the most common apple disease caused by the ascomycete fungus Venturia inaequalis [59,60]. Insecticides were probably used against various moth, aphid, and weevil species [58,59]. Herbicides and plant growth regulators are only used in conventional apple production [61], while in organic apple orchards, weeds are partly tolerated or controlled by mowing, grubbing, or tillage.

Although our data do not include information on the number of pesticide products applied (containing the ASs) or on the frequency of spraying during the season, we know from other studies that about 228 pesticide products are applied in about 19.9 ± 4.4 sprays in Austrian apple fields [62]. The number of pesticides used is similar to that in northern Italy (215 pesticide products), one of the most intensive apple growing regions in Europe [63]. Other countries report spraying frequencies on apple farms in this range: 16.9 sprayings in the UK, 22.7 sprayings in Poland, and 25.6 sprayings in Italy [16].

The use of plant growth regulators is only allowed in conventional production. They are used to thin out the fruit, to regulate growth, and to adjust the harvest periods for apples. Since organic apple cultivation does not permit the use of plant growth regulators and can nevertheless produce high-quality apples, it seems that plant growth regulators can also be dispensed with.

In general, beneficial organisms have been shown to benefit from lower toxic exposure to sprays in apple orchards [64], and lower toxic exposure also generally reduces risks to aquatic and terrestrial organisms as well as to humans [65,66]. Apart from apple orchards, many studies show lethal, sublethal, and combined effects of pesticides on bees [67,68], which also include effects of non-insecticide pesticides [69,70,71].

Since honeybees are the most important pollinators in apple orchards, we have analyzed the toxicities of the ASs used in these locations in more detail. To reduce the toxic load for honeybees in apple orchards, the most efficient approach would be to reduce the use of the insecticide ASs emamectin and milbemectin, which accounted for 46% and 39% of the toxic load for honeybees in conventional apple orchards. In organic apple orchards, the toxic loads for honeybees could be reduced by reducing the use of the insecticide ASs fatty acids, which contributed 70%, and the fungicide AS sulphur, which contributed 15% of the toxic load for honeybees. However, it should be noted that emamectin is classified as highly toxic, delivering 7.9 × 10¹¹ LD₅₀ doses (0.036 µg bee⁻¹) and fatty acids are classified as moderately toxic, delivering 4.7 × 10¹² LD₅₀ doses (25 µg bee⁻¹). Emamectin was highly toxic not only to honeybees but also to birds and is one of the conASs with the lowest acceptable exposure level (AOEL) for humans [52]. The fatty acids were moderately toxic to the predatory mite Typhlodromus pyri [72], honeybees, and aquatic organisms, mildly toxic to birds, but non-toxic to humans. Fatty acids are approved as herbicides at the EU level, but no herbicide product containing fatty acids as an AS has been authorized in the market in Austria [73]. We therefore recommend not to apply the products automatically at the maximum frequency recommended by the manufacturer, but only as often as it is actually necessary.

To reduce the toxic exposure to birds in apple orchards, the use of the ASs most hazardous to birds would have to be reduced: inorganic fungicides in both conventional and organic production; the insecticides carbamates and dithiocarbamates in conventional production; biological insecticides in organic production; and the herbicides phenoxy-phytohormones and plant growth regulators in conventional production; these pesticide groups are not used in organic apple production.

To reduce the toxic load for earthworms in apple orchards, it would be necessary to reduce the use of the most earthworm-toxic ASs. The most earthworm-toxic fungicides were carbamates and dithiocarbamates in conventional production and inorganic fungicides in organic farming. The most earthworm-toxic insecticides were organophosphates in conventional and biological and botanical insecticides in organic production. The earthworm toxicity of herbicides and plant growth regulators is only a problem in conventional production. Inorganic fungicides in organic production mainly comprise ASs based on copper. However, copper products are also applied in conventional fields and 60% of the 118749 kg of fungicidal ASs applied in conventional apple fields in Austria in 2017 was copper-based products [33]. Nevertheless, the highest number of LD₅₀ doses of conASs delivered to earthworms was not from inorganic fungicides, but from the chemical groups of carbamates and dithiocarbamates (2.1 × 10⁹). The conAS mefentrifluconazole is currently the only authorized AS in Austrian apple production and is classified as acutely and highly toxic to earthworms (Eisenia fetida LD₅₀ = 2.65 mg kg soil⁻¹).

If the goal is to reduce the use of pesticides in apple orchards, the most effective measure would be to avoid herbicides, as these are not used in organic apple production. It is known that many management methods can promote functional agrobiodiversity and limit the use of pesticides in apple orchards [74,75,76]. Perennial flower strips between apple tree rows can promote natural enemies of pests, reduce important apple pests and their associated fruit damage, and help reduce insecticide use [64,74,77,78,79]. It has been shown that herbicides can be replaced by mulching, flaming, or steam weeding without any loss of apple yields or fruit quality [80,81]. In addition, cover crops and weed harrowing in apple fields showed similar yields to herbicide treatment, with positive effects on earthworms [82]. Pesticide use was also positively influenced by the frequency of fertilizer application [62], and fertilizer use has been associated with increased fungal disease severity in apples [83,84]. In contrast, groups of natural enemies of pests in apple orchards responded positively to organic fertilizers and composts [85]. Tillage can also influence pesticide use, and tillage should not be too intensive, as more intensive tillage was associated with higher pesticide use [62]. In addition, the frequency of management practises could also have an indirect effect on pesticide use, as moving machinery through apple orchards can serve as a vector for the transmission of fungal pathogens [86]. The knowledge and behaviour of individual farmers was also an important factor [62]. In addition, it was found that meteorological parameters related to human-made climate change can challenge pesticide reduction targets: fewer pesticides are used when the number of hot and warm humid days increases, but more pesticides are used when the number of frost days and the number of sunny days increase [62].

However, it should be noted that the responsibility for reducing pesticide use lies not only with apple farmers, but also with retailers. The acceptance of less-than-perfect-looking apples [30], moving away from current monoclonal orchards and growing a wider range of more robust cultivars [87], and a fair price that acknowledges environmental efforts would encourage farmers to take greater steps to reduce pesticides. Other measures to reduce pesticides in apple production, e.g., in the Chinese provinces Shaanxi and Shandong, have shown that 71% of apple farms use excessive amounts of pesticides and that a reduction could be achieved by implementing measures to promote crop insurance, expand acreage, and increase pesticide prices [88].

When talking about pesticide reduction, it is also important to consider ecotoxicity. In Austria, for example, herbicide sales decreased by 24% between 2010 and 2019, while the toxic loads to honeybees increased by 487% (oral exposure), lethal toxic loads to earthworms increased by 498%, and to birds, they increased by 580% [37]. This is attributed to a shift towards the use of more acutely toxic and more persistent ASs.

4.2. Potential Ecotoxicological Impact of Grapevine Production in Austria

In conventional vineyards in Austria, 19.9 kg ha⁻¹ of conASs was applied in 2017, compared to 36.5 kg ha⁻¹ in organic vineyards. Nevertheless, the ecotoxicological burden on honeybees, birds, and earthworms was 2.4 times higher for conASs (1.2 × 10¹¹ LD₅₀ doses) than for orgASs (0.5 × 10¹¹ LD₅₀ doses). The largest differences between the ecotoxicity of conASs and orgASs were observed in earthworms (4 times greater LD₅₀ from conASs than from orgASs), followed by honeybees (4 times greater LD₅₀ from conASs than from orgASs) and the least difference was observed for birds (2.7 times greater LD₅₀ from conASs than from orgASs).

Overall, 2.6 times more ASs are approved for conventional grapevine production (60 conASs) than for organic grapevine production (23 orgASs). As in apple farming, orgASs can also be used in conventional farming but not vice versa. Fungicides, herbicides, insecticides, and plant growth regulators dominate in conventional viticulture, while fungicides dominate in organic viticulture.

Here, too, fungicides were the most important pesticide category. In 2017, a total of 601.226 kg of inorganic fungicides was used in conventional viticulture, compared to only 201.526 kg in organic grapevine production. Inorganic fungicides were the only pesticides that could potentially be toxic to earthworms in organic vineyards, as no other pesticides were used. In conventional vineyards, other fungicides, morpholines, imidazoles and triazoles, carbamates, and dithiocarbamates as well as fungicides based on biological and botanical products were also used.

Since the TL in Austrian grapevine production is the highest for birds, we analyzed which specific AS can cause the highest toxic loads. We found that the contribution to the total TL from conASs, based on the maximum of the recommended applications, was highest for the fungicide dithianon, at 14%; the plant growth regulator zinc phosphide, at 12%; and the fungicides potassium phosphonates, at 11%. In organic vineyards, the highest ecotoxicity to birds was caused by the fungicide tribasic copper sulphate, at 65%; copper oxychloride, at 21%; and copper hydroxide, at 8%.

In acute oral testing, birds receive doses directly from the crop or via their proventriculus [89]. It is unclear whether birds in fields exposed to the LD₅₀ doses, determined by direct oral exposure, are exposed to lower doses or whether they receive any doses at all. We focused on the European serin, a small songbird that uses vineyards (and apple orchards) as a frequent habitat. The number of individuals has declined sharply in the last 20 years and the population size has decreased by about 80% [90]. However, it should be noted that pesticides and their non-target effects are not the only reason for this decline [91].

The finding that the toxic load for birds was higher in conventional vineyards than in organic vineyards, even though more ASs are used, emphasizes that the amount of pesticides is not a good parameter to compare the effects of pesticides and that reducing the amount alone is not sufficient to protect non-target organisms [37].

The use of copper in organic agriculture is a frequently discussed problem because copper accumulates in the soil. Due to their persistence and impact on organisms, copper products are ‘candidates for substitution’ of pesticides in the European Union [92]. However, it should be noted that substances containing copper are also used in conventional agriculture.

A higher risk for S. serinus was found in vineyards (organic and conventional) than in apple orchards, as the vineyard area in Austria is larger. As a result, a larger amount of pesticides is used and thus the number of LD₅₀ released is also higher. Again, the number of lethal doses delivered by herbicides and insecticides was much lower compared to fungicides. In particular, because herbicides are prohibited in organic farming, obviously, no insecticides were used and no other pesticides were used either. The risk of the death of S. serinus caused by these substances was zero. On the other hand, 1.1 × 10⁹ lethal LD50 doses were released by herbicidal conASs and 3.4 × 10⁵ from conASs in pyrethroids. Within the group of pyrethroids, only a single substance was authorized for grapes: lambda-Cyhalothrin, against grape phylloxera. At the beginning of the 20th century, phylloxera was the most dangerous pest for grapes, as it attacked the roots and caused the vines to die. Although it is still a pest on vines, it has become less important since the European variety Vitis vinifera was grafted onto a phylloxera-tolerant rootstock. The insecticidal ASs rapeseed oil and azadirachtin have been authorized in organic farming against phylloxera aphis, albeit with less efficacy [93]. However, based on the data used in the current study, it was not used in vineyards in Austria in 2017 [33].

Inorganic fungicides again delivered a very high number of LD₅₀ doses to S. serinus in both production systems. We found that the highly toxic orgAS tribasic copper sulphate also delivered the highest number of lethal LD₅₀. Although it is an orgAS, we do not know whether organic farmers use products with tribasic copper sulphate at all or if conventional farmers use it, as it is permitted in conventional farming too. In another study conducted in Austria, tribasic copper sulphate is not among the 10 most commonly used ASs in apple orchards [62]. Sulphur is the most commonly used, while copper oxychloride was in fifth place. However, if products with tribasic copper sulphate were used at the recommended frequency, 2.06 × 10¹¹ S. serinus could potentially die. Tribasic copper sulphate is authorized against ‘roter brenner’ and ‘downy mildew’ (Peronospora). Moreover, also, copper oxychloride and copper hydroxide are authorized against Peronospora. But since tribasic copper sulphate has an LD₅₀ of 72.4 mg kg bw⁻¹, copper oxychloride has an LD₅₀ of 173 mg kg bw⁻¹, and copper hydroxide has an LD₅₀ of 223 mg kg bw⁻¹, copper sulphate showed the highest bird toxicity within the copper compounds. An Italian study analyzed the efficacy of various copper compounds against Peronospora and found that copper hydroxide was the most effective [94].

Another little-studied AS group is rodenticides, which delivered 1.0 × 10⁷ LD₅₀ doses to birds in conventional orchards, but zero LD₅₀ doses in organic orchards. The only approved AS with the mode of action of a rodenticide is zinc phosphide and it is highly toxic to birds. In the worst case, 3.5 × 10¹⁰ highly toxic doses could be delivered. In 2015, 150 wild geese died in Germany after oral exposure to zinc phosphide (agrarheute, 2015). In organic farming in Austria (and the rest of Europe), however, no anticoagulant rodenticides are allowed [73].

To reduce the toxic load for honeybees in vineyards, the use of the most hazardous ASs for honeybees would have to be reduced: inorganic fungicides in both conventional and organic production and pyrethroids as insecticides in conventional vineyards. Meanwhile, no insecticides are used in organic vineyards. Regarding herbicides, the use of organophosphorus ASs and plant growth regulators should be reduced in conventional vineyards, while in organic vineyards, neither herbicides nor plant growth regulators are allowed anyway.

To reduce the toxic load for earthworms in vineyards, the use of the most harmful ASs to earthworms would have to be reduced: in terms of fungicides, morpholines in conventional and inorganic fungicides in organic production; in terms of insecticides, pyrethroids in conventional vineyards, while no insecticides are used in organic vineyards; in terms of herbicides, organophosphorus ASs and plant growth regulators in conventional production, while no herbicides or plant growth regulators are used in organic vineyards anyway.

4.3. Limitations of Our Analysis

Although our study was based on application data from pesticide-intensive crops in Austria, we recognize some limitations in our calculations.

Firstly, the calculated TLs are only the number of LD₅₀ doses applied to non-target organisms. During application, a large proportion of the pesticides are distributed in the field and in the environment, and in most cases, only a small proportion comes into direct contact with the non-target organism. Whether and when a pesticide comes into contact with a non-target organism depend on many factors, e.g., the target crop, the timing of application, the type of application, and the persistence of the pesticides in the environment [40].

Secondly, like official environmental risk assessments, our assessment was based on a few surrogate species. The shortcomings of the surrogate species concept have been widely discussed and do not take into account the impact on the overall biodiversity or ecosystem functions and services [1,95,96].

Thirdly, LD₅₀ values describe direct effects, but apart from direct effects on organisms, the effects of pesticides on organisms can often be indirect [48]. Such effects can occur at very low doses, but are not considered in our approach using LD₅₀ values. Nevertheless, we believe that this is a useful approach to describe the potential adverse effects in the environment.

Fourthly, we have not considered the effects of multiple contaminations. However, simultaneous exposure to a variety of pesticides in low concentrations may trigger additive or synergistic effects, especially when the exposure is over a long period of time [97] or when certain substances (e.g., azole fungicides) are involved [98]. Other studies from apple- and grapevine-producing regions have shown that the contamination of sites with multiple pesticide residues is frequent and, therefore, non-target organisms and ecosystems are exposed to multiple pesticides [5,99], even via the inhalation of contaminated air [4]. Nevertheless, pesticide risk assessments still focus on toxicity in terms of the exposure to a single pesticide [100]. The actual ecotoxicological risk assessments performed by the EU underestimate the real effects on non-target organisms, as they do not take into account indirect effects, not to mention the already known interactions between individual pesticides [1,101].

Fifthly, we only considered the ecotoxicological effects and did not include other factors in the difference between conventional and organic production such as the energy intensity and thus the greenhouse gas emission of the production of synthetic pesticides [8].

5. Conclusions

Our results have shown that the highest toxic loads for honeybees, birds, and earthworms were caused by fungicides, both in conventional and organic apple and grapevine production systems. As the harmful effects of pesticides on non-target organisms, including humans, are undisputed, a reduction in pesticide use seems imperative. Fortunately, numerous conservation biological control methods have been developed for both apple farming and vineyards, which have shown that by reducing pesticide use in these production systems, yields and income can be maintained and agroecosystem integrity and agronomic performance can be preserved [62,102].

Conservation biological control can reduce pesticide use by generally extensifying farming and/or providing alternative habitats and food sources for natural enemies of pest species [17]. In grapevine production, extensive management of vegetation has increased the above- and belowground biodiversity and associated ecosystem services [103]. Interestingly, no trade-off between yields and biodiversity was found for both apple and grapevine production, a caveat that is often raised to avoid biodiversity enhancement measures. Viable biodiversity above and below ground forms the basis of sustainable production with an impact on fruit quality [18]. Future assessments of the ecotoxicity of agricultural production could be supplemented by an analysis of pesticide residues in apples, grapes, and soil.

The results of the current study show that the promotion of organic farming would reduce ecotoxicological hazards in agricultural systems in Europe, as formulated in the EU strategy Farm-to-Fork [6]. This would have a positive impact on biodiversity and human health [51,52].