Agrochemicals and Shade Complexity Affect Soil Quality in Coffee Home Gardens

: Soil quality can directly inﬂuence the health, yield, and quality of a particular crop species, and agrochemicals are often used to boost soil micro- and macro-nutrients. The excessive application of agrochemicals, however, is often the cause of imbalances in acidity and nutrient concentration and can cause soil to deteriorate. The presence of multiple shade trees in farmland can positively inﬂuence soil quality. Here, we evaluate the effect of agrochemical use (i.e., organic, mixed, and intensive) and shade tree complexity (i.e., sun, low, and high) on soil quality (i.e., pH, macronutrients, and micronutrients) in 56 coffee home gardens in Indonesia. We found that Al, Fe, K, and Mn were signiﬁcantly higher in farms that used agrochemicals, and pH was more acidic in ﬁelds with intensive use of agrochemicals. C:N ratio and Mn were higher in soils with high shade complexity than in sun-exposed soils. The use of agrochemicals, however, is not sustainable as it increases the Al concentration and decreases pH, both of which are associated with poor coffee growth and reduced soil quality. Shade tree removal and the use of invasive, non-native species, such as eucalyptus, can also negatively inﬂuence soil quality, and thus the maintenance of complex shade cover with native trees should be prioritised.


Introduction
The global human population is expected to increase to 10 billion by 2050, with demand for food crops exceeding this population inflation [1][2][3]. As a result, the attention of researchers has turned to the ability of farmers not only to produce enough food to feed the population and meet demand but ensure that generations to come are able to utilize the soils that are left to them. Agroforestry, i.e., the integration of natural forest and farmland, has been suggested as a possible solution to ensure the longevity of smallholder farmland by preserving biodiversity, enhancing ecosystem services and, in turn, securing the livelihoods of farmers [4][5][6]. Shade trees within farmland provide both habitat for wildlife and a means of income for farmers and enrich soil previously damaged through monoculture farming [7][8][9]. The quality of soil plays a vital role in providing ideal conditions for crop growth [8,10]. Through the optimal composition of nutrients and minerals, achieved using both artificial and organic inputs, soil can directly influence the health, yield and quality of a particular crop species [11]. Additionally, through interactions between the composition of micro-and macro-nutrients within soil and the diversity of fauna and flora that exist within and around it, soil can influence and benefit entire ecosystems [10,[12][13][14]. By ensuring

Soil Sample Collection
We collected soil samples from 56 coffee home gardens in May 2019 in the municipalities of Cipaganti and Pangauban, West Java (a detailed description of the study site can be found in [5,27]). Soil samples consisted of 15 sub-samples taken by soil probe from 30 cm depth for a total of 1 kg for each home garden. Sub-samples were taken randomly within each home garden, mixed, and stored in one plastic bag labelled with a code of each garden. Before taking soil samples we ensured that the soil was not wet from rain, that the vegetation covering the soil was removed and that the probe was clean from the previous sampling. Once we collected all soil samples, we assessed the shade complexity and agrochemical use within each garden. Agrochemical use (i.e., fertilizer urea and NPK and pesticide Endosulfan) was established through ad hoc interviews with farmers (more details in [6,24]). Following these interviews, we categorised farms into one of three categories detailing their level of agrochemical usage: chemicals (intensive use of chemicals); mixed (used a mix of chemicals and organic methods); and organic (did not use any chemical practices). Shade complexity was calculated using both shade cover and tree species richness. Shade cover was estimated using the Canopeo App that uses photos of the canopy to produce an approximate percentage shade cover [24,28]. We obtained shade tree richness by counting the number of tree species present within and along the perimeter of each farm [24]. Using these data, we then assigned farms a level of shade complexity: sun (less than or equal to 10% shade cover); low shade complexity (above 10% shade cover and less than five tree species present); and high shade complexity (above 10% shade cover and more than or equal to five tree species present). Although there is a relationship between shade farms and organic practices, it is not evenly distributed throughout the study area, and sun farms use a range of agrochemicals and organic alternatives.

Soil Sample Analysis
We stored the soils sample at room temperature for a maximum of three days after the collection and analyzed them at the Balai Penelitian Tanaman Sayuran (laboratorium penguji terpadu) in Lembang, Bandung. This laboratory follows strict procedures to ensure a high standard of analytical quality control and is accredited ISO/IEC 17025. We calculated: soil pH via by potentiometry (H 2 O and KCl); organic carbon (C) as % via the Kumies spectrophotometric method; nitrogen (N) as % via the Kjeldhal method; available potassium (K) as % via the Morgan Venema method; iron (Fe), manganese (Mn), copper (Cu), and zinc (Zn) as ppm via the Morgan Venema method; calcium (Ca) and magnesium (Mg) as cmol/kg via Buffer NH4OAC 1 M pH 7; and aluminium (Al) as cmol/kg via the titrimetric method [29].

Data Analysis
We tested whether soil quality (i.e., soil pH and concentration of individual nutrients) was influenced by agrochemical use and shade complexity via Generalised Linear Models (GLMs). We used the "glmmTMB" package and the "glmmTMB" function to produce the models due to the availability of several fit families within the package [30]. We tested each component within the soil analysis against each family, using the QQ residual plots and residual vs. predicted plots produced via the "DHARMa" package to establish the best fit [31]. We used the "emmeans" package and "emmeans" function to produce estimated marginal means and pairwise ratios for each component [32]. We considered p < 0.05 as a significant result and <0.001 as a highly significant result. We used R v.4.2.0 to run all of the analyses [33].

Agrochemical Usage
Al, Fe, K, and Mn were all significantly higher in farms that used agrochemicals, with significant differences observed between organic and chemicals, organic and mixed,  Figure 1). Al was particularly affected by chemical usage, exhibiting significant differences between each of the agrochemical use treatments (Table 1; estimated marginal means ratio (95% CI): C:M: 0.895 (0.788-1.017); C:O: 12.13 (2.837-51.8); M:O: 3.02 (0.785-11.6)). In all instances, pH was lower in farms that used agrochemicals, with significant differences observed between those intensively using chemicals and those using organic farming practices (Table 1; Figure

Shade Complexity
Shade complexity significantly influenced pH, C:N ratio, and Mn concentration (  Figure 3; estimated marginal means ratio (95% CI): L:S: 0.914 (0.836-1.00)). C:N ratio was higher in farms with low shade complexity, with significant differences observed between high shade complexity and sun farms, and low shade complexity and sun farms (Table 1; Figure 4; estimated marginal means ratio (95% CI): H:S: 1.15 (1.035-1.28); L:S: 1.08 (1.005-1.15)). Mn concentration was reduced with decreased shade complexity, with the only significant difference being between high shade complexity and sun farms (Table 1; Figure 5; estimated marginal means ratio (95% CI): H:S: 1.70 (1.025-2.83)). We found no significant association between shade complexity and agrochemical use (χ 2 = 4.29, p = 0.368).  6)). In all instances, pH was lower in farms that used agrochemicals, with significant differences observed between those intensively using chemicals and those using organic farming practices (

Discussion
We found that pH was significantly lower in farms that used agrochemicals (both intensive and mixed) than in organic farms, bolstering the extensive published literature documenting the harmful effects of inorganic farming practices on soil pH [10,11,18,34]. The pH of the soil has wide-reaching effects on the soil as an ecosystem, affecting both the life within it and the growing potential of the crops. Soil acidification, caused by the uptake of nitrogen through chemical fertilizers, creates an unfavourable environment for microbial communities to exist and, as a result, harbours fewer nematodes and reduced nematode diversity [13]. In addition to hampering diversity, soil acidification has been found to lead to reductions in yield due to the decrease in soil base cations, Ca 2+ , and Mg 2+ , which are essential nutrients for coffee growth, and increases in non-base cations, Fe 3+ , and Al 3+ [11,12,35].
Whilst pH was generally lower in sun farms and farms that had lower shade complexity, low pH was also observed in low and high shade complexity farms. This difference will likely be due to the dominating presence of eucalyptus trees (Eucalyptus spp.) amongst the shade tree species used at the study site. Eucalyptus trees are often used as shade trees due to their fast growth and income-generating properties [27,36,37]. Numerous studies detail the applicability of eucalyptus trees as shade trees in coffee farms with varying results due to competition for root growth between eucalyptus and crop plants. Competition for root growth has been found to impact the productivity of coffee, although the extent to which productivity is affected seems to vary depending on annual rainfall [36,37]. In addition to affecting yield, eucalyptus trees release a toxin to inhibit the growth of other plants in their vicinity, and this toxin has been proven to cause soil acidification [38]. Therefore, if a farm has high tree species richness yet many of the trees are eucalyptus, acidity is likely to still be higher than in other organic farms. The observed variation in pH in sun farms could be attributed to a wider variation of fertilizers used, both organic and chemical, whereas in high shade complexity farms, farmers use more organic inputs.

Discussion
We found that pH was significantly lower in farms that used agrochemicals (both intensive and mixed) than in organic farms, bolstering the extensive published literature documenting the harmful effects of inorganic farming practices on soil pH [10,11,18,34]. The pH of the soil has wide-reaching effects on the soil as an ecosystem, affecting both the life within it and the growing potential of the crops. Soil acidification, caused by the uptake of nitrogen through chemical fertilizers, creates an unfavourable environment for microbial communities to exist and, as a result, harbours fewer nematodes and reduced nematode diversity [13]. In addition to hampering diversity, soil acidification has been found to lead to reductions in yield due to the decrease in soil base cations, Ca 2+ , and Mg 2+ , which are essential nutrients for coffee growth, and increases in non-base cations, Fe 3+ , and Al 3+ [11,12,35].
Whilst pH was generally lower in sun farms and farms that had lower shade complexity, low pH was also observed in low and high shade complexity farms. This difference will likely be due to the dominating presence of eucalyptus trees (Eucalyptus spp.) amongst the shade tree species used at the study site. Eucalyptus trees are often used as shade trees due to their fast growth and income-generating properties [27,36,37]. Numerous studies detail the applicability of eucalyptus trees as shade trees in coffee farms with varying results due to competition for root growth between eucalyptus and crop plants. Competition for root growth has been found to impact the productivity of coffee, although the extent to which productivity is affected seems to vary depending on annual rainfall [36,37]. In addition to affecting yield, eucalyptus trees release a toxin to inhibit the growth of other plants in their vicinity, and this toxin has been proven to cause soil acidification [38]. Therefore, if a farm has high tree species richness yet many of the trees are eucalyptus, acidity is likely to still be higher than in other organic farms. The observed variation in pH in sun farms could be attributed to a wider variation of fertilizers used, both organic and chemical, whereas in high shade complexity farms, farmers use more organic inputs.
We found significantly higher concentrations of Al, Zn, Fe, and K in farms that used agrochemicals intensively and those that used both agrochemicals and organic practices. It is true that these minerals are essential for growing Arabica coffee, yet the significant concentrations of each within farms that use agrochemicals suggests that this is due to the excessive use of chemical fertilizers. Whilst high concentrations of Zn and K have been found to influence coffee cup aroma [26,39], high nutrient concentrations through the application of chemical fertilizers can lead to nutrient leaching and lower CEC, rendering the presence of such nutrients less beneficial due to their impact on the soil's ability to hold them [10,11]. Studies surrounding the nutrients necessary for coffee cup quality (i.e., consistency, freshness, and flavours) are contradictory, with several researchers documenting the opposite effect of Zn [39]. In contrast to the aforementioned nutrients, the toxicity of excessive Al in coffee farms has been well-documented for several decades, with increased Al concentrations associated with poor coffee growth [35]. Not only do increased Al concentrations limit root function and growth, but, similarly to low pH, they have been shown to limit the retention of Ca and Mg [35,[40][41][42]. This is in line with our findings, as Ca and Mg were the only nutrients that were found to be lower in farms that used chemicals.
The C:N ratio of soil samples was significantly higher in farms with high shade complexity, with the C:N ratio decreasing at the same rate between farms with high, low, and sun shade complexity. This difference is due to the increased availability of organic residue from the litter of surrounding shade trees [43]. The fact that C:N ratio was higher in farms with high shade complexity is in line with research by Teixeria et al. [25], who found that increased plant diversity positively influenced soil carbon stocks, showing that it is not just the presence of shade cover that is important, but the diversity of tree species that are used for shade. Whilst shade complexity did positively influence the C:N ratio and soil pH, it was found to be less influential than agrochemical use with regards to soil composition overall. This is in contrast to existing literature reporting the benefits of a diversity of shade tree species for soil quality, with Sauvadet et al. [13] finding that shade complexity was as influential on soil quality as agrochemical use [25]. Our findings are likely due to the presence of eucalyptus trees within our study site affecting soil composition, nutrient retention and soil acidity [36,37].
Whilst the concentrations of particular nutrients within soil are crucial for successful coffee growth, the importance of a diversity of nutrients has been well documented, with researchers highlighting the benefits of organic farming for maintaining this diversity [10,13,44]. Although farms classified as organic, mixed, and agrochemical were using such practices prior to our study, soil samples were taken during the same year that training for the implementation of high-quality organic practices, such as natural pest control, rabbit manure, and microorganisms (EM4) used to boost plant growth, were implemented [27]. Therefore, it is likely that we were not able to observe the benefits of these practices on nutrient concentration and diversity to the full extent.
By using organic, wildlife-friendly farming practices that promote biodiversity through the use of natural fertilizers and agroforestry, farmers are ensuring the longevity of their farmland through the protection of ecosystem services at all levels [45]. Less acidic soil harbours a diversity of soil microbiota, nematode species, and invertebrates, improving soil structure and plant growth, and, in turn, influencing the presence of birds and mammals [46]. This diversity ensures the direct and indirect provision of ecosystem services, such as pollination, seed dispersal, and natural pest management [5,6,13,15,16,47]. Whilst there is some debate as to whether organic farming practices improve crop yield, researchers have documented decreases in coffee yield under acidic conditions, and in turn, Ca 2+ and Mg 2+ depleted soils [8,11,48]. Furthermore, farmers can increase the premium they receive from their crops through organic and environmental certification, with consumers saying they are willing to pay more for products that are pesticide-free and made using organic practices [49]. Finally, by creating a natural optimal environment for coffee, farmers increase their crops' resilience to climate change. Environmental variables, such as temperature, humidity, and altitude, significantly affect coffee growth regardless of agrochemical use Earth 2022, 3 862 and shade complexity [10,50]. Therefore, by tailoring their farms to the needs of coffee through natural interventions, such as adding shade trees to lower temperature, they are increasing the likelihood of successful harvests despite a changing climate [51][52][53].

Conclusions
We aimed to evaluate the effect of agrochemicals and shade complexity on soil quality in coffee home gardens in West Java, Indonesia. We found that both agrochemical use and shade complexity significantly affected soil pH, causing soil to become more acidic in farms that exclusively used agrochemicals and those with low shade complexity. In addition to low soil pH, farms that exclusively used agrochemicals saw higher concentrations of Al, Zn, Fe, and K. Whilst high Al concentrations are directly damaging to coffee growth, the high concentrations of Zn, Fe, and K are due to the excessive application of fertilizers, which will eventually lead to nutrient leaching and low CEC [10,11,34].
Moving forward, we will disseminate the results of this work to farmers from our study site and help promote the ongoing use of high-quality organic practices that will increase pH and decrease Al toxicity, such as the use of rabbit manure and microorganism EM4 [27]. Another way of achieving this is through "liming", where lime calcium oxide, calcium hydroxide, calcium carbonate, or calcium magnesium carbonate are used to reduce the availability of H and Al, leading to a higher pH and reduced Al toxicity [35]. The waste from coffee processing, i.e., coffee husks and pulp, has been found to be effective in liming soil and increasing the presence of Mg and K, helping to reduce waste, reduce costs, and reduce farmers' dependence on chemical substitutes [54,55]. Furthermore, we will promote the use of native fruit tree species as shade trees instead of eucalyptus trees due to the increased micro-organism activity of soil with coffee intercropped with other plant species [56,57]. Although eucalyptus trees are fast-growing and act as a source of income for farmers, as environmental variables begin to fluctuate due to climate change (i.e., fluctuations in rainfall), the long-term sustainability of farmland should be prioritised over short-term gain.

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