The current Australian sugarcane industry transition towards adoption of an ‘alternative’ herbicide strategy as part of improved environmental stewardship is increasingly complicated by recent farming system, regulatory and herbicidal product changes. This study quantified and compared the efficacy, economic costs and environmental risk profile s of a range of established, emerging, and recently registered pre -emergent herbicides across field trials in the Wet Tropics region of North Queensland. Several herbicides were effective on certain weed species, but lacked broad spectrum control. Better efficacy results from products with multiple active ingredients (i.e., imazapic -hexazinone) demonstrated the benefits of using mixtures of active ingredients to widen the spectrum of weed control efficacy. All tested pre – emergent herbicides behaved quite similarly in terms of their propensity for off-site movement in water (surface runoff losses generally > 10 % of active applied), with their losses largely driven by their application rate . Herbicides with lower application rates consistently contributed less to the total herbicide loads measured in surface runoff. Results demonstrated alternative choices from the more environmentally problematic herbicides (such as diuron) are available with effective alternative formulations providing between 4 and 29 less risk than the traditional diuron -hexazinone ‘full rate’. However, considerable challenges still face canegrowers in making cost -effective decisions on sustainable herbicide selection. Additional research and effective grower extension are required to address information gaps in issues such as specific weed control efficacy of alternative herbicides and potential blending of some herbicides for more effective broad spectrum weed control, while also minimizing environmental risks.

Introduction

The decline of fresh and marine water quality associated with land -based runoff from adjacent agricultural catchments is a major cause of the poor state of many of the coastal ecosystems of Australia’s Great Barrier Reef (GBR) World Heritage Area (Waterhouse et al., 2017). Pesticides have been specifically identified as among the most important diffus e source pollutants from catchment areas of the GBR (Brodie et al., 2012). High value freshwater, estuarine and inshore waters of the GBR lagoon are regularly exposed to pesticide runoff from agricultural lands, particularly during wet season riverine flood events (December to April) (e.g. Lewis et al., 2009 ; Davis et al., 2012). The pesticides that have been most commonly detected in the GBR lagoon are herbicides that inhibit electron transport at photosystem II (PSII) in plants and include diuron, atrazine, hexazinone, ametryn, simazine and tebuthiuron (Haynes et al., 2000a ; Shaw and Müller, 2005 ; Shaw et al., 2010). Concentrations during elevated stream flow events can at times exceed ecological water quality protection guideline trigger values for up to several weeks for some of these PSII herbicides (Lewis et al., 2009 ; 2012 ; Brodie et al., 2012 ; Davis et al., 201 3 ; O’Brien et al., 2016; Novic et al., 2018).

Viable seeds of numerous plant species are not capable of germinating immediately after harvest under conditions suitable for the germination process. Such seeds are termed primarily dormant. Primary dormancy is established during seed development and maturation on the mother plant. This type of dormancy is particularly common in wild plants. Dormancy is a very important phenomenon which prevents germination on mother plants, facilitates seed dispersal, ensures plant survival of natural catastrophes, and reduces intra-specifc competition (Bewley et al. 2013). The phenomenon has turned out to be non-obligatory, its expression depending on environmental factors, e.g., temperature. Seeds can be fully dormant; such seeds are not able to germinate at any temperature (Hilhorst 2007). There are also seeds which are not capable of germination only within a certain temperature range, whereas they germinate at temperatures outside that range. Thus, the expression of dormancy in such seeds depends on temperature, and dormancy release is associated with widening the range of germination temperature. Under natural conditions, primarily dormant seeds are exposed to fuctuating environmental conditions, e.g., light, temperature, moisture, and the presence of gases, which leads to dormancy state cyclicity (Finkelstein et al. 2008). Primary dormancy can be removed also by cold stratifcation, dry storage, light, or chemicals (Bewley et al. 2013). It is commonly accepted that the balance between abscisic acid (ABA) and gibberellins (GAs) and/or sensitivity to these hormones are responsible for regulation of the dormancy state and germination of seeds in response to environmental signals (Finkelstein et al. 2008; Rodríguez-Gacio et al. 2009). ABA is considered as the most important hormone responsible for the establishment of dormancy during seed development and for maintenance of dormancy during seed imbibition. In turn, GAs have been cited as factors involved in dormancy release and/or germination. Dormancy release has been shown as involving a decline in the ABA content and an increase of the GAs level (Bewley et al. 2013).

Sensors for pesticides with high sensitivity have been urgently required to control food safety, protect ecosystem and prevent disease. In this review, we provide an overview of recent advances and new trends in optical sensors for the detection of pesticide based on fluorescence, colorimetric and surface enhanced Raman scattering, surface plasmon resonance and chemiluminescent strategies. These methods will be classified by the types of recognition elements, including enzyme, antibody, molecularly-imprinted polymers, aptamer and host-guest reaction. This review explores the basic features of established strategies through assessment of their performance. In addition, we provide brief summary of the entire review, the drawbacks of present sensor and future prospects, as well as the ongoing efforts to pesticide optical sensors.

Introduction

Pesticide are primarily used to prevent, control, or eliminate pests and weeds for boosting agricultural productivity in modern agricultural practices [1,2]. According to the literatures [3,4], the use of pesticides helps in securing almost one-third of crop production globally. However, the residue of pesticide even at trace levels not only seriously cause food contamination, but also severely breakdown the ecosystem, posing a great danger to people’s daily life [5e7]. As a result, pesticide pollution has attracted more and more concern and become one of the most alarming challenges. For proper management of pesticide, Governments have set lots of policies for guiding pesticide use and have regulated maximum residue levels on foods and agricultural commodities [2,8,9]. Although most pesticide were detected to be within recommended limits, the bioaccumulation effect and continuous exposure can rise safety risks to human health [10]. In addition, some new types of pesticides with highly effective activity, whose toxic mechanism have not clear understood, are being continuously brought into market [11]. Therefore, the analysis of pesticide residues is an urgent demand to ensure food quality and safety, safeguard the ecosystem and protect human health from possible hazards. Pesticide detection have traditionally been carried out by employing conventional chromatographic techniques, including high-performance liquid chromatography (HPLC) [12e14], gas chromatography (GC) [15e17] and mass spectrometry (MS) [18e20]. Although these techniques offer powerful trace analysis with excellent sensitivity and high reproducibility, many drawbacks, such as sophisticated equipment, time consuming, tedious sample preparation and purification steps, obviously limited their on-site and real-time application, particularly emergency cases. Thus, vast endeavors have been devoted to investigating alternative strategies for realizing pesticide in a facile, speedy, sensitive, selective, accurate and user-friendly manner. In fact, significant attention has been drawn to the fabrication of optical sensors for pesticide detection. For pesticide analysis, myriad optical strategies have been established utilizing recognition elements, such as enzyme, antibody, molecularly-imprinted polymers, aptamer and host-guest recognizer, which employed to directly capture and identify the target pesticide. Moreover, the integration of recognition elements and nanomaterials possess high sensitivity and excellent selectivity in terms of real-time analysis, which is in high demand for pesticide detection. This Review focuses on the recent development of sensitive pesticide optical sensor, with a particular emphasis on the fluorescence (FL), colorimetric (CL), surface-enhanced Raman scattering (SERS), and other strategies including surface plasmon resonance (SPR) sensor and chemiluminescence strategy (Fig. 1), which provide comprehensive coverage of current standings of pesticide detection. Rather than summarizing either enzyme-based sensors [1,2,21,22] or nanomaterials-based strategies [4,11,23e26] as performed in other excellent reviews, we highlight the latest achievements in pesticide optical sensor and provided readers with a high-impact recent advances in the developing field from our point of view. Enzyme, antibody, and host-guest chemistry as recognizer have been frequently employed in pesticide optical sensor to achieve high sensitivity and good selectivity, which exhibited great superiority.

The agricultural use of pesticides leads to environmentally relevant pesticide concentrations that cause adverse effects on stream ecosystems. These effects on invertebrate community composition can be identified by the bio-indicator SPEARpesticides. However, refuge areas have been found to partly confound the indicator. On the basis of three monitoring campaigns of 41 sites in Central Germany, we identified 11 refuge taxa. The refuge taxa, mainly characterized by dispersal-based resilience, were observed only nearby uncontaminated stream sections and independent of the level of pesticide pressure. Through incorporation of this information into the revised SPEARpesticides indicator, the community structure specifically identified the toxic pressure and no longer depended on the presence of refuge areas. With regard to ecosystem functions, leaf litter degradation was predicted by the revised SPEARpesticides and the median water temperature at a site (R2 = 0.38, P = 0.003). Furthermore, we designed the bio-indicator SPEARrefuge to quantify the magnitude of general recolonization at a given stream site. We conclude that the taxonomic composition of aquatic invertebrate communities enables a specific indication of anthropogenic stressors and resilience of ecosystems.

Introduction

The intensive use of pesticides worldwide results in high pesticide concentrations in streams (Malaj et al., 2014; Stehle and Schulz, 2015) and causes negative effects on the structure (Liess and von der Ohe, 2005; Münze et al., 2017), ecological functions (Schäfer et al., 2012) and biodiversity (Beketov et al., 2013) of freshwater communities. This widespread degradation of aquatic ecosystems is further accelerated by climate change, increasing exposure to agricultural pesticides (Kattwinkel et al., 2011) and increasing pesticide vulnerability of populations by unfavorable temperatures (Dinh et al., 2016). The use of effect-based indicators is a powerful tool for identifying the toxic pressure of pesticides. The invertebrate-based indicator SPEARpesticides uses trait information of taxa to identify pesticide pressure and the ecological effects in streams. SPEARpesticides was developed by Liess and von der Ohe (2005) and successfully applied to indicate pesticide pressure in streams of different geographical regions worldwide including Europe (Schäfer et al., 2007; Liess et al., 2008; Orlinskiy et al., 2015; Münze et al., 2015; Münze et al., 2017), Russia (Beketov and Liess, 2008), Australia (Schäfer et al., 2011), USA (Chiu et al., 2016) and South America (Hunt et al., 2017). However, as described in several studies (Orlinskiy et al., 2015; Liess and von der Ohe, 2005), the indicator value of SPEARpesticides not only is related to the toxic pressure but also is partly confounded by the presence of uncontaminated stream sections. Such stream sections serve as refuge areas for those invertebrates that recolonize affected stream sections mainly by downstream drift. The influence of refuge areas on toxicant effects has been observed particularly under weak to medium pesticide pressure (Orlinskiy et al., 2015; Liess and von der Ohe, 2005). In contrast, high pesticide pressure impeded the successful recolonization of refuge taxa. Consequently, the presence of refuge areas was taken into account to accurately predict the toxic pressure of pesticides with the SPEARpesticides indicator system. The influence of external sources of recolonizing organisms on disturbed communities has also been recognized in experimental freshwater studies. For example, in a mesocosm experiment, Caquet et al. (2007) have shown that isolation from external populations highly delays the recovery of sensitive insects from pesticide exposure. Trekels et al. (2011) have concluded from a semi-field study with the insecticide endosulfan that external recovery via adult dispersal is particularly important for univoltine compared with multivoltine aquatic invertebrates. Despite the knowledge on the general effects of refuge areas, it is unknown which taxa in freshwater streams are strongly linked to the presence of refuge areas under pesticide exposure and which traits mainly characterize those taxa. However, this knowledge is crucial to increase the stressor specificity and spatial independence of indicators such as SPEARpesticides.

Abstract

Government regulators of agricultural pesticides use multidisciplinary computer models with which they make product licencing decisions. Hydrology engineering and chemistry components are integrated into the multidisciplinary models. Although the scientific quality of the hydrology components is reasonably good, the scientific quality of the chemistry component does not match this good quality. These models often use chemically meaningless parameters and assumptions that are not supported by experiments. During the past 25 years chemical kinetics and mechanisms based on experiments and stoichiometry have been published for most agricultural pesticides in soil and water. Thus, the chemistry components should now be replaced with those that use the kinetics and mechanisms currently available. They should next be continuously upgraded as the chemical mechanisms research evolves.

Introduction

Regulatory agencies in North America, the United Kingdom, and the European Union use multidisciplinary computer models for the regulation of agricultural pesticides. The PRZM GW (pesticide root zone model, groundwater) model is an example. More than three dozen of these models have been reported, used for predicting the risk of pesticides to groundwater. They are based mainly on hydrology engineering but attempt to account for the chemical effects of soils on the pesticides. The multinational corporations that market agricultural pesticides submit applications with test data, for product registration. Licences for the sale and use of the pesticides depend on the submitted test data and computer predictions used by the regulatory agencies. However, the legally binding decisions are only as good as the model predictions. The safety of both people and their environment ultimately depends on the quality of the science from which the regulatory technology is derived. The hydrology and chemistry components of the computer models are discussed here.

Modern agriculture needs high crop productivity in order to cope with the increasing food demand, and this in order to meet the necessities of the growing world population (Del Buono et al., 2016). In cultivated fields, weeds compete with crops for nutrients and other resources (Gaba et al., 2017). In order to manage this situation, herbicides are used worldwide for crop protection against weeds (Mimmo et al., 2015). However, the massive use of these chemicals can cause resistance in weeds and affect the environment (Gaba et al., 2017). In particular, they can impact on non-target organisms and bring about hazards to human and animal health (Boily et al., 2013; van der Meulen and Singh Chauhan, 2017). Some of these chemicals can also accumulate and persist in soils (Arias-Estevez et al., 2008; Magne et al., 2006 ), and thus could represent a risk for crops in the following cultural cycle (Del Buono et al., 2015). Despite their selectivity in targeting weeds, some herbicides can be phytotoxic to crop species at morphological, physiological and biological levels (Vercampt et al., 2017). Recently, some authors found significant interferences of herbicides on plant mineral nutrition. For example, glyphosate was shown to be able to reduce the uptake of certain micronutrients in soybeans (Zobiole et al., 2010) and sunflower (Eker et al., 2006). In particular, some of these studies showed reductions on iron (Fe) assimilation and distribution in various crop species (Bartucca et al., 2017; Bellaloui et al., 2009; Del Buono et al., 2015; Eker et al., 2006; Ozturk et al., 2008). In general, interferences on the capacity of plants to acquire Fe can result in Fe-chlorosis. This disorder is characterised by interveinal to complete yellowing and browning of leaves which could, in extreme situations, lead to plant death (Singh and Dayal, 1992). Fe has a fundamental role in photosynthesis, DNA and chlorophyll biosynthesis, mitochondrial respiration and nitrogen fixation (Nagajyoti et al., 2010). This element is also a key constituent of many important proteins such as ferredoxin, catalase, peroxidase and superoxide dismutase (Nagajyoti et al., 2010). Therefore Fe is a limiting factor for plant biomass production and quality (Briat et al., 2015).

Antibiotics are widely used in public health and animal husbandry in the treatment and prevention of bacterial infections. The US Food and Drug Administration (FDA) reported that 14.6 million kilograms of annual domestic sales of antimicrobial drugs were used in food-producing animal husbandry in 2014 (FDA. 2014). Particularly, tetracyclines accounted for 71% of these sales. In many countries, the use of antibiotics both for animal disease treatment and growth promotion is unsupervised, which often leads to abuse of antibiotics. Most of veterinary antibiotics were poorly absorbed within the gastrointestinal tract of livestock, causing the presence of biologically active antibiotics in animal waste. It was not unusual to detect high concentrations of antibiotic residues in animal manures (Qiao et al., 2012; Pan et al., 2011;Nordenholt et al., 2016). In an agricultural field at Hannover, the tetracycline (TTC) level in the dried manure from the soil surface was 4 mg/kg (Gerd Hamscher, 2002). Similarly, Martinez analyzed pig manure and turkey dung samples from livestock farms in Austria, Vienna. They revealed that the content of chlortetracycline (CTC), oxytetracycline (OTC), and TTC up to 46 mg/kg, 29 mg/kg and 23 mg/kg respectively (Martinez-Carballo et al., 2007). In China, the antibiotic residues in animal feces are also commonly detected. In a research work, tetracycline antibiotics in feces of livestock breeding farms were investigated in seven provinces. The average content of OTC, TTC and CTC was 5.9 mg/kg, 2.6 mg/kg and 1.3 mg/kg, respectively (ZHANG et al., 2005). Besides, the occurrence of veterinary antibiotics in organic vegetable bases in northern China were analyzed. Strikingly, the highest concentration of antibiotics was up to 183.5 mg/kg for OTC (Hu et al., 30 2010).

Ecologists have recognized two important successions: initial succession, which is the occurrence of ecosystem expansion on the sites that have not been previously occupied by living organisms (such as bare rocks and frozen surfaces); and secondary succession, which happens in the ecosystems, which have been previously occupied by organisms. Tilled felds, clear-cut forests, and burned areas are all exposed to secondary succession (Letcher 2010; Podda et al. 2011; Rezvani et al. 2013). During the succession, vegetative variations are not coincidental, and spread of vegetation is regularly carried out with the following functional variations (Dölle et al. 2008; Schmid et al. 2017). The early stages in succession include species, which have light seeds (Rockwood 1985), high seed production (Hyatt and Casper 2000), extensive distribution (Samuel and Hart 1994), and rapid growth rate (Lambers et al. 2008). Species with large seeds have longer life period, meaning they are more competitive and dominant in the fnal stages of succession (Fenesi et al. 2014). In such stages, perennial weed species replace the annual ones (Corbin and D’Antonio 2004) in a period of one (Brown et al. 1987) to 40 years (Bornkamm 1981). The rate of these events depends on factors such as soil (Dzwonko and Loster 1997; Fanelli and Lucchese 1998), climate (Calvo et al. 2005), weed seed bank (Bekker et al. 2000), and the initial density of perennial weed species (Turner et al. 1998). The establishment of juvenile plants, bush, and weed under the tree shadow idiomatically is called ‘understory’ (Leopold and Solozar 2008) that have interference with trees. The interactions between weed growth and tree behaviour have been addressed by diferent studies (Cantarelli et al. 2006; Rondon 2006; Bonari et al. 2017). For example, Larpkern et al. (2010) have identifed specifc species of weeds under the trees after tillage. It is preferable that such weeds be controlled at the early stage of growth, or the use of repair trends become difcult. Increased frequency of species is largely controlled by seed production, species growth rate (Chapin et al. 1997), and competitiveness (Tilman 1988). The chemicals characteristics of soil such as nutrients, salts, and organic matter may cause the appearance of specifc weed species (Garnier et al. 2004; Salama et al. 2017). Some research work has shown a correlation between variation of soil nutrients and plant composition (Shaltout et al. 2014; Angiolini et al. 2017). For example, when soil nitrogen and organic matter reduces, bushes are replaced with grasses (Reynolds et al. 1999; Suding et al. 2008).

Biological control of weeds, based on applications of ecological, evolutionary and experimental science, is a highly successful discipline with a rich history, going back over 115 years. It is a global practice: the development of every weed biological control organism, and its subsequent release in the country of introduction, requires collaborative research on at least two continents and typically includes the involvement of several countries. Weed biocontrol practitioners, comprising researchers, land managers and policy makers, constitute a community with a shared goal, namely the suppression of invasive alien plants in agriculture and in conservation areas, through the deployment of suitably-tested and environmentally safe (i.e. sufficiently host-specific), plant-feeding insects, mites and pathogens. A complete record of every deliberate instance of weed biocontrol, worldwide, up to December 2012 (see Table 1 in Winston et al. 2014; and updated at http://www.ibiocontrol.org/catalog/) has been compiled in the expansive 5th edition of ‘Biological control of weeds—A world catalogue of agents and their target weeds’. This compilation records successes and failures, with accounts of the circumstances in each case, and a summary of the information by Schwarzla¨nder et al. (2018) is the subject of the opening paper in this Special Issue of BioControl.

The increasing interest in organic and low-input farming systems has renewed attention toward alternative methods of weed management, such as the development of innovative mechanical solutions (Avola et al., 2008; Pannacci and Tei, 2014; Melander et al., 2015). Organic and low-input farming systems mainly relied for its crop nutrients on legume crops (De Ponti et al., 2012). In general, increasing legume cultivation could bring benefits for the environment and resource use at a range of scales, from the field to the global; their precrop effect, nitrogen provision, and potential to improve nutrient conservation and soil structure add to the sustainability of farm productivity while saving resources and reducing emissions (Covarelli et al., 2010; Reckling et al., 2014). Among the grain legume crops, soya bean and faba bean are considered very important, although due to different reasons. In fact, soybean (Glycine max (L.) Merr.) is one of the most important grain legume and oilseed crops in the world, accounting for more than 50% of the global oilseed production (Datta et al., 2017). Faba bean (Vicia faba L.) is grown world-wide as a protein source for food and feed, offering ecosystem services such as renewable inputs of nitrogen (N) into crops and soil via biological N2 fixation, and a diversification of cropping systems (Jensen et al., 2010). It is well known that prolonged weed interference not only causes heavy crops yield losses, but increases production costs and reduces the quality of produce, thus requiring early-season weed management to achieve economically acceptable yields (Knezevic et al., 2003; Sardana et al., 2017). In particular, the presence of weeds up to beginning of seed stage of soya bean (R5) may cause 8–55% reduction in yield (Van Acker et al., 1993). Weeds are managed in soya bean primarily by herbicides (Niekamp and Johnson, 2001; Datta et al., 2017), although mechanical and cultural weed control methods showed to be effective (Chauhan and Opeña, 2013; Pannacci and Tei, 2014). Faba bean is known to compete weakly against weeds in the early growth phase (Lee and Lopez-Ridaura, 2002); so the pre-emergence herbicides are commonly used in order to control weeds until the crop is big enough to suppress any additional emerging weeds (Köpke and Nemecek, 2010). However, over the last twenty years, environmental and human health impact of herbicides use, increasing of herbicide resistance, the scarce availability of herbicides for minor crops and the increased of organic farming were the main factors that stimulated the interest to develop alternative methods to chemical weed control, such as mechanical weed control (Melander et al., 2005; Pannacci et al., 2017).

Reduction in economic losses in agricultural production due to abiotic and biotic factors is of utmost importance in modern day inputintensive agricultural systems. Sustaining the production levels demands devising newer strategies for mitigating the ill-effects of these adverse factors. As with abiotic causes, especially the lack or excess of moisture in the growth season, extreme temperatures, high or low irradiance and nutrient supply, biotic stresses have the potential to reduce yields substantially (Oerke, 2006). Among the major biotic constraints, weeds are considered as the most harmful to agricultural production besides affecting agrobiodiversity and natural water bodies. They also affect the crop production indirectly, by competing with the crop for resources, sheltering crop pests, interfering with water management, reducing the yield and quality, and subsequently increasing the cost of processing (Zimdahl, 2013). Therefore, weed management is the major and important part of crop production. In India, reduction in crop yield was estimated as 31.5% (22.7% in winter and 36.5% in summer and rainy seasons) by weeds (Bhan et al., 1999). Whereas, the economic losses due to weeds in India was estimated as INR 20 to 28 billion about two decades ago (Sahoo and Saraswat, 1988; Sachan, 1989). In another study, it was reported that loss in agricultural production due to weeds amounts to INR 1050 billion per annum (NRCWS, 2007; Varshney and PrasadBabu, 2008). In general, the yield loss due to weeds is almost always caused by a group of different weed species, and these weeds may have substantively different competitive ability (Weaver and Ivany, 1998; Milberg and Hallgren, 2004). Practically, it is very difficult to estimate the yield loss due to single weed species and therefore, it is estimated as the collective efforts by all the weeds. Overall, weeds produced the highest potential loss (34%), with animal pests and pathogens being less important (losses of 18 and 16%) worldwide (Oerke, 2006). As far as studies on yield loss at global level is concerned, Milberg and Hallgren (2004) explored the large-scale patterns in yield loss in cereals due to weeds in Sweden and mentioned that weed biomass explained 31% of the variation in yield loss due to weeds. Whereas, O’Donovan et al. (2005) developed various regression equations in western Canada to estimate the effects of weeds on yield loss of field crops. Oliveira et al. (2014) also presented that insect, pests cause an average annual production loss of 7.7% in Brazil, which is a reduction of approximately 25 million tonnes of food, fibre and biofuels.