Effect of sugarcane cropping systems on herbicide losses in surface runoff

Effect of sugarcane cropping systems on herbicide losses in surface runoff

Science of the Total Environment 557–558 (2016) 773–784 Contents lists available at ScienceDirect Science of the Total Environment journal homepage:...

948KB Sizes 0 Downloads 52 Views

Science of the Total Environment 557–558 (2016) 773–784

Contents lists available at ScienceDirect

Science of the Total Environment journal homepage: www.elsevier.com/locate/scitotenv

Effect of sugarcane cropping systems on herbicide losses in surface runoff Gunasekhar Nachimuthu a,b,⁎, Neil V. Halpin b, Michael J. Bell c a b c

NSW Department of Primary Industries, Australian Cotton Research Institute, 21888 Kamilaroi Highway, Narrabri, NSW 2390, Australia Department of Agriculture and Fisheries (QLD), Bundaberg Research Facility, 49 Ashfield Road, Kalkie, QLD 4670, Australia School of Agriculture and Food Sciences, The University of Queensland, Gatton, QLD 4343, Australia

H I G H L I G H T S

G R A P H I C A L

A B S T R A C T

• We evaluated four sugarcane cropping systems for herbicide loss in runoff • Soil and trash management effects on herbicide losses were of primary importance • The physico-chemical properties of herbicide on losses were less pronounced • Improved practices reduced Atrazine losses by 62% relative to Conventional Practice

a r t i c l e

i n f o

Article history: Received 21 January 2016 Received in revised form 15 March 2016 Accepted 15 March 2016 Available online 17 April 2016 Editor: D. Barcelo Keywords: Atrazine Diuron Metolachlor Pendimethalin Metribuzin Paddock to reef

a b s t r a c t Herbicide runoff from cropping fields has been identified as a threat to the Great Barrier Reef ecosystem. A field investigation was carried out to monitor the changes in runoff water quality resulting from four different sugarcane cropping systems that included different herbicides and contrasting tillage and trash management practices. These include (i) Conventional – Tillage (beds and inter-rows) with residual herbicides used; (ii) Improved – only the beds were tilled (zonal) with reduced residual herbicides used; (iii) Aspirational – minimum tillage (one pass of a single tine ripper before planting) with trash mulch, no residual herbicides and a legume intercrop after cane establishment; and (iv) New Farming System (NFS) – minimum tillage as in Aspirational practice with a grain legume rotation and a combination of residual and knockdown herbicides. Results suggest soil and trash management had a larger effect on the herbicide losses in runoff than the physicochemical properties of herbicides. Improved practices with 30% lower atrazine application rates than used in conventional systems produced reduced runoff volumes by 40% and atrazine loss by 62%. There were a 2-fold variation in atrazine and N10-fold variation in metribuzin loads in runoff water between reduced tillage systems differing in soil disturbance and surface residue cover from the previous rotation crops, despite the same herbicide application rates. The elevated risk of offsite losses from herbicides was illustrated by the high concentrations of diuron (14 μg L−1) recorded in runoff that occurred N2.5 months after herbicide application in a 1st ratoon crop. A cropping system employing less persistent non-selective herbicides and an inter-row soybean mulch resulted in no residual herbicide contamination in runoff water, but recorded 12.3% lower yield compared

⁎ Corresponding author at: NSW Department of Primary Industries, Australian Cotton Research Institute, 21888 Kamilaroi Highway, Narrabri, NSW 2390, Australia. E-mail address: [email protected] (G. Nachimuthu).

http://dx.doi.org/10.1016/j.scitotenv.2016.03.105 0048-9697/© 2016 Elsevier B.V. All rights reserved.

774

G. Nachimuthu et al. / Science of the Total Environment 557–558 (2016) 773–784

to Conventional practice. These findings reveal a trade-off between achieving good water quality with minimal herbicide contamination and maintaining farm profitability with good weed control. © 2016 Elsevier B.V. All rights reserved.

1. Introduction In Australia, sugarcane and grazing lands have been a major focus for export of nutrients, sediment and agricultural chemical residues from the Great Barrier Reef (GBR) catchments (Brodie et al., 2012; King et al., 2013; Masters et al., 2008; Masters et al., 2013; Mitchell et al., 2005; O'Reagain et al., 2005). The Burnett-Mary catchment discharges into the southern part of GBR lagoon. The main crops grown in this region include sugarcane, vegetables, tree crops for fruits and nuts and grain legumes. An initial investigation carried out in the Burnett-Mary Region (Stork et al., 2008b) identified the presence of nutrients and herbicides in runoff water from vegetable, macadamia and sugarcane production systems. The sugar industry in the Burnett-Mary has a gross value of $196 M. The sugarcane crop is typically grown for four years, with one plant crop and three ratoon crops followed by a break crop (vegetables or grain legumes) or a bare fallow. Planting of sugarcane usually happens in spring (late Aug–Sept) or autumn (March–April). Several herbicides are applied to sugarcane, primarily from planting to just after fill-in (canopy closure), although occasionally herbicide sprays may occur at a later stage in weedy fields. Applications during the planting to fill-in period coincide with the summer rainfall season for the spring-planted crop, which represents the period of highest risk of losing herbicides in runoff. The sugarcane industry in Australia is currently developing an industry best management practices (BMP) to improve the sustainability of the cropping systems (Sugarcane BMP, 2015). Any additional information on improved herbicide management practices will lead to improvement in the Smartcane BMP program. It has been clearly demonstrated that herbicides used on sugarcane farms, especially residual herbicides such as diuron, ametryn, atrazine and metribuzin, can move into waterways in runoff (Masters et al., 2013; Stork et al., 2008a; Stork et al., 2008b). Such herbicide contamination of waterways discharging into the GBR lagoon has been reported to impact on aquatic life in a number of studies (Jones, 2005; King et al., 2013; Råberg et al., 2003; Smith et al., 2012), so minimizing such losses through improved management practices (especially for residual herbicides, such as atrazine and diuron) has been a clear focus of research and best practice implementation programs. The Australian Government has implemented the Reef Rescue program (Reef Rescue program, 2008) to help growers and managers improve farming practices in an attempt to improve water quality in the GBR lagoon by reducing the amount of nutrient, herbicides and sediment leaving farms in runoff. This program was based on the hypothesis that a reduction in pollutant loads could be directly correlated with the adoption of improved farming practices, although only a few studies have quantified this link. Most of the research that compares different herbicide management practices in northern Australia used small plot rainfall simulators to assess herbicide losses (Masters et al., 2013), with only limited field monitoring having been conducted (Oliver et al., 2014; Stork et al., 2008a). In some cases that monitoring was for a very limited duration (e.g. from a single irrigation event (Oliver et al., 2014)). In addition to the shortage of seasonal or annual monitoring of farm scale herbicide losses from sugarcane cropping systems in GBR catchments, there are also few field programs (other than rainfall simulation studies) comparing grower standard practices with other management approaches. Given the general scarcity of field scale data on herbicide loss from irrigated sugarcane cropping systems, the ability to reliably predict the impact of changed management practices on water quality in GBR catchments is seriously limited.

Our study is one of the first of its kind for the sugarcane industry and is aimed at addressing this current knowledge gap. We have measured the herbicide concentrations and calculated loads leaving adjacent management strips in a cropping field. The concentrations of selected herbicides (diuron, atrazine, metribuzin, metolachlor and pendimethalin) in soil and trash were also quantified over time to assess herbicide persistence and risk of off-farm impacts in the longer term. 2. Materials and methods A three-year field investigation was conducted to quantify the impact of different land management practices on offsite water quality generated during the fallow (vegetable or grain legume) and sugarcane production phases of regionally significant intensive cropping systems. The grain legume and intensive vegetable systems were assessed during a one-year rotation break, during which crops of soybean or sequential crops of capsicum and zucchini were grown, followed by a return to a sugarcane cycle monitored during the plant and subsequent 1st ratoon crops. Management practices were assessed for their capacity to reduce sediment, nutrient and herbicide movement from fields to streams. This paper focuses on the herbicide data monitored during the sugarcane plant cane (2011−12) and first ratoon crop (2012−13) crops. 2.1. Site description The site was established in the Burnett Mary region of Queensland, Australia, in a well-drained field containing a mixture of Yellow Brown Chromosol or Dermosol soils (Isbell, 2002) depending on location. The average sand, silt and clay fractions of the soil were 77%, 16% and 8%, respectively. The area has a subtropical climate where long term average mean maximum and minimum temperatures were 17 and 27 °C and the long-term average rainfall of 1019 mm sees N50% of rain fall in the summer months (Jan–Mar). Soil pH and organic carbon were 6.5 and 1%, respectively. The site was a commercial cane field that was very uniform in the top 70 cm of the soil profile. The field was split into four management units with each unit being 280 m long and 9 m wide (i.e. 5 × 1.83 m cane rows), with contrasting management systems randomly allocated to each strip. The 280 m length was subdivided into two subunits of approximately 120 m and 160 m based on a highpoint in the middle of the field, with drainage in either direction in response to a 1% slope. Runoff flumes were installed to quantify the runoff, sediment and herbicide movement from the smaller 120 m × 9 m blocks (Fig. 1), while the 160 m × 9 m blocks were used for nutrient investigations (Nachimuthu et al., 2013). The management systems and their various management practices are described in Table 1. Key features of these systems were as follows (i) Conventional practices - current commercial practice consisting of full tillage after an intensive vegetable rotation and the application of traditional residual herbicides; (ii) Improved practices - only the beds were tilled after the vegetable phase (zonally tilled with the inter-space left undisturbed) and residual herbicide rates were reduced; (iii) Aspirational practices – a minimum tillage system (one pass of a single tine ripper in the bed zone prior to the vegetable and sugarcane phases), where vegetative trash mulch was maintained during cane planting, no residual herbicides were used and a legume intercrop was sown after cane establishment; and (iv) New Farming System (NFS) – a minimum tillage system (as in Aspirational practice) with grain legume rotation crops, retention of a surface trash mulch and a combination of residual and knockdown herbicides.

G. Nachimuthu et al. / Science of the Total Environment 557–558 (2016) 773–784

775

Fig. 1. Field lay out of the experimental plot. The field was split into four management units with each unit being 280 m long and 9 m wide (i.e. 5 × 1.83 m cane rows), with contrasting management systems randomly allocated to each strip. The 280 m length was subdivided into two subunits of approximately 120 m and 160 m based on a highpoint in the middle of the field, with drainage in either direction in response to a 1% slope. Runoff flumes were installed to quantify the runoff, sediment and herbicide movement from the smaller 120 m × 9 m blocks, while the 160 m × 9 m blocks were used for nutrient investigation not reported in this paper.

equation that we used in this experiment based on our 200 mm San Dimas flume was

2.2. Runoff sampling and analysis San Dimas flumes (Wilm et al., 1936) (200 mm) were used to measure the runoff volume from each management practice. The galvanized steel flumes were manufactured as per standard specifications outlined in Walkowiak (2008). The flumes were installed approximately 5 m away from the end of the sugarcane rows, outside the actual cropping area. Steel and rubber belting was used as a barrier to collect runoff from four inter-rows and direct the runoff water into the flume for flow measurement and sample collection (Fig. 1). The flow discharge

Q ðflow=unit timeÞ ¼ ½4:09  ðFlow height ðftÞÞˆ1:34: The flow height was measured using a Teledyne ISCO 730 bubbler module connected to a Teledyne ISCO 6712 standard portable water sampler which logs the flow height and rainfall in minute intervals. The module uses a differential pressure transducer and a flow of bubbles to measure liquid levels to determine flow height. The samplers were

Table 1 A comparative summary of management practices investigated. Systems

Conventional

Improved

Previous management First crop Trash management Cultivation Fallow management

Ground cover

Capsicum Removed/burnt Full tillage Bare fallow (knockdown herbicides) Zucchini No tillage Cane Full tillage in beds and inter-rows Nil

Herbicide (refer Table 2A)

Residual - traditional

Second crop Cultivation Third crop Cultivation

Fourth crop Groundcover Herbicide (refer Table 2A)

Aspirational

New Farming System

Sugarcane

Ratoon Residual- traditional rate

Capsicum Capsicum Removed/burnt Retained Full tillage Strip Forage sorghum grown Forage sorghum grown and slashed and slashed Zucchini Zucchini No tillage No tillage Cane Cane Tillage only in beds – 3 No tillage/minimum tine zonal tillage disturbance Forage sorghum and Forage sorghum and zucchini residues (~1 t/ha) zucchini residues (~1.5 t/ha) Knock- down and modified Knock-down residual Ratoon Ratoon Sugarcane trash blanket

Soybean Retained Strip Soybean

Knock-down and low rate residual

Knock-down and low rate residual

Knock-down

Fallow No tillage Cane No/minimum tillage Soybean and cane trash residues (~2 t/ha) Knock-down and modified residual Ratoon

776

G. Nachimuthu et al. / Science of the Total Environment 557–558 (2016) 773–784

Fig. 2. Cumulative monthly rainfall totals at the site during the monitoring period.

programmed to carry out flow weighted sampling (50 mL of samples taken after every 125 L of flow and combined for each runoff event), and the samplers were fitted with 4 × 3.6 L glass sample bottles. After each runoff event, samples were retrieved within 12 h and stored in the dark in a cool room (4–6 °C) until transported overnight on ice to the laboratory. 2.3. Herbicide analysis in runoff water Herbicide analyses (for active ingredients and their metabolites) were performed on runoff samples from selected runoff events up to 30 days

from each herbicide application, with samples collected from events after a second herbicide application were also analysed for herbicides applied in the first application. Sample extraction and analysis was conducted by the Queensland Health Forensic and Scientific Services (QHFSS) organics laboratory, which is accredited through the National Association of Testing Authorities. Atrazine and its metabolites, metolachlor and diuron were extracted through solid phase extraction cartridges and analysed by LCMS. Pendimethalin and metribuzin were solvent extracted followed by GC– MS analysis while 2,4-D was extracted through solid phase extraction and analysed by GC–MS. It should be noted that results from the solid

Table 2A Summary of herbicide applied to different management systems during plant cane and ratoon crop (PS II herbicides in bold). Date/crop stages

Management practices

Plant cane crop (August 2011 to Sept 2012) 19/09/2011 (Planting) Aspirational Improved

Conventional

NFSb

11 and 12/01/2012a (Fill-in)

First ratoon (Sept 2012 to Aug 2012) 08/11/2012

NFSb, Improved & Conventional

Aspirational

Improved

Conventional

NFSb

a b

Herbicide

Rate/ha

Active ingredient

a.i./ha

Powermax 2,4-D Spray seed 250H Dual gold Atradex Atradex Stomp-extra Spray seed 250H Spray seed 250H Dual gold Atradex Spray seed 250H Soccer

1.5 L/ha 1.5 L/ha 2.4 L/ha 1.3 L/ha 1.5 kg/ha 2.2 kg/ha 2.2 L/ha 2.4 L/ha 2.4 L/ha 1.3 L/ha 1.5 kg/ha 2.4 L/ha 1.8 kg/ha

Glyphosate 2,4-D Paraquat/diquat Metolachlor Atrazine Atrazine Pendimethalin Paraquat/diquat Paraquat/diquat Metolachlor Atrazine Paraquat/diquat Metribuzin

810 g 937.5 g 324/276 g 1248 g 1350 g 1980 g 1001 g 324/276 g 324/276 g 1248 g 1350 g 324/276 g 1350 g

Gramoxone Starane advance 2,4-D Gramoxone Soccer 2,4-D Gramoxone Diuron 2,4-D Gramoxone Starane advance 2,4-D

1.6 L/ha 0.78 L/ha 1 L/ha 1.6 L/ha 1 L/ha 1 L/ha 1.6 L/ha 1.8 L/ha 1 L/ha 1.6 L/ha 0.78 L/ha 1 L/ha

Paraqaut Fluroxypyr 2,4-D Paraquat Metribuzin 2,4-D Paraquat Diuron 2,4-D Paraqaut Fluroxypyr 2,4-D

400 g 260 g 625 g 400 g 750 g 625 g 400 g 1620 g 625 g 400 g 260 g 625 g

Aspirational practice was companion-planted with soybeans after fill-in and was not sprayed with any herbicide. NFS - New farming systems.

G. Nachimuthu et al. / Science of the Total Environment 557–558 (2016) 773–784

777

Table 2B Selected physico-chemical properties of herbicides monitored in runoff in this trial (adopted from University of Hertfordshire, pesticide properties database, viewed online on 26/02/ 2016). Analyte

Solubility in water at 20 °C (mg/L)

Soil half-lives and persistence interpretation

Soil sorption coefficient (Koc) (mL/g)

GUS leaching potential indexa

Atrazine Diuron Metolachlor 2,4-D Metribuzin Pendimenthalin

35 (low) 35.6 (low) 530 (high) 24,300 (high) 1165 (high) 0.33 (low)

29 (non persistent) 89 (moderately persistent) 21 (non persistent) 28.8 (non persistent) 19 (non persistent) 90 (moderately persistent)

100 (moderately mobile) 813 (slightly mobile) 120 (moderately mobile) 39.3 (mobile) 95 (mobile) 17,581 (non mobile)

3.2 (high) 1.83 (transition state) 2.10 (transition state) 1.69 (low) 2.57 (transition state) −0.41(low)

a Calculated from the soil half-life and partition coefficient. GUS = log t1/2 × (4 − log Koc) where Koc is the soil sorption coefficient (mL/g) and t1/2 is the half-life in soil (days). Higher values indicate greater potential for leaching. GUS values lower than 1.8 and higher than 2.8 indicate, respectively, non-leacher and leacher pesticide compounds; for GUS values between 1.8 and 2.8 the pesticide is considered in a transition zone (Gustafson, 1989).

phase extraction method will not include pesticides sorbed to sediments present in the sample, while the solvent extraction approach employed for the GC–MS analysis will.

2.4. Calculation of herbicide active ingredient loads The herbicide active ingredient load per rainfall event was calculated by multiplying volumes of runoff water (L per ha) by the concentration herbicide (μg L−1), while, herbicide losses for each event are presented as event mean concentrations (EMC - herbicide active ingredient load in an event/flow volume in same event), total active ingredient loads (g/ha) and active ingredient loads expressed as a percentage of the active ingredient applied. The results were presented for individual events (Tables 4, 5 and 6) and also for a cumulative total for the events monitored (Table 3) similar to the farm scale water quality results presented by Hollinger et al. (2001).

2.5. Soil sampling and herbicide analyses To study the degradation of herbicides, sampling was undertaken on a log time scale with a minimum of 5 times (e.g. pre-spray and then day 0.5, 5, 10 and 30 days after spraying during the plant cane crop and prespraying, 0.5, 4, 13, 18 and 27 days after spraying during the ratoon crop) for trash and the 0–25 mm soil layer. Samples were collected from at least three replicates in each management practice (to capture the field variability), pooled and analysed for herbicides. A quadrant (96 cm2) was placed over a section of trash/soil to be sampled and pushed into the soil to a depth of 25 mm. A knife was used to cut trash from the area inside the quadrant. All the trash was placed in an alfoil-lined bag pre-rinsed internally with acetone. A trowel was used to dig a hole on the outside of the quadrant and the trowel was slid under the quadrant to collect the top 25 mm of soil. Soil samples were again packed into an alfoil-lined bag that had been rinsed with acetone. All the samples were placed in the dark in a cool room (4–6 °C) immediately after collection until sent to the laboratory for analysis.

2.5.1. Analysis of atrazine and metolachlor in soil during the plant cane crop Ammonium chloride solution and acetone were added to the sample and shaken for approximately 12 h. The solvent portion was filtered and evaporated to the aqueous solution. Acetonitrile was added to the solution and a QuEChERS pre-mix salt was added to the solution and shaken vigorously for 1 min. An aliquot of the supernatant was transferred to a vial and internal standards were added; the sample was then analysed for herbicides by LCMSMS. A set of blank soil sample, spike soil sample and non-extracted spike and surrogate (NESS) sample were included with each batch of analysis. A duplicate sample, if available, was also included. 2.5.2. Metribuzin and pendimethalin analysis in soil/trash during plant cane crop Ammonium chloride solution, acetone and hexane was added to the sample and shaken for approximately 12 h. The solvent portion was filtered and evaporated to a smaller volume; which was then subjected to gel permeation chromatography (GPC) as a sample clean-up. The eluent collected from the GPC was then passed through a second clean-up procedure using Florisil. The eluent from the Florisil clean-up was then evaporated and finally analysed by gas chromatography–mass spectrometry (GC–MS) for pesticides. A set of blank soil/trash samples, spiked soil/trash samples and NESS sample were included with each batch of analysis. A duplicate sample, if available, was also included. 2.5.3. Herbicide analysis in soil and trash during the ratoon crop Soil/trash samples were accurately weighed to 5.00 g/2.00 g and recorded to 2 decimal points. Samples were extracted in acidified methanol and sonicated for 30 min. The extracts were then filtered through a 0.45 μm Teflon membrane filter to remove sediment and particulate matter. The filtered extract was pH and volume adjusted then analysed by Ultra performance liquid chromatography - tandem mass spectrometer (UPLC MSMS). Samples were quantitated against standardised calibration curves using freshly and accurately prepared standards. As well as freshly prepared standards, stock standards from a second source were also analysed to provide added confidence. Blanks, spikes and

Table 3 Summary of herbicide active ingredient loads (g/ha) and event mean concentrations (EMC) in different management systems during the sugarcane plant cane crop in 2011–12. Management practices Loads NFSb Conventional Improved EMC NFSb Conventional Improved a b

No of sampled events 4 4 3 4 4 3

Sampled runoff

Atrazine

Desethyl atrazine

Desisopropyl atrazine

Metolachlor

Pendimethalin

Metribuzina

mm 55.5 56.1 31.7 mm 55.5 56.1 31.7

g/ha 19.22 24.08 9.27 μg L−1 34.6 42.9 29.3

g/ha 1.16 2.09 1.28 μg L−1 2.09 3.73 4.03

g/ha 0.46 1.06 0.61 μg L−1 0.83 1.89 1.94

g/ha 8.11 0.23 6.24 μg L−1 14.61 0.41 19.70

g/ha 0.00 14.83 0.00 μg L−1 0.00 26.43 0.00

g/ha 7.04 1.50 0.53 μg L−1 14.26 3.61 1.68

New farming systems. Only two sampled events due to later application times (see Table 2A).

778

G. Nachimuthu et al. / Science of the Total Environment 557–558 (2016) 773–784

Table 4 Herbicide active ingredient loads (g/ha) in runoff water in New Farming Systems (trash retained) during the sugarcane plant crop. Herbicide application rates were atrazine-1350 g a.i./ha and metolachlor – 1248 g a.i./ha applied 19/09/2011, metribuzin-1350 g a.i./ha applied 12/01/2012. Date

Days after application

Runoff (mm)

Atrazine

Desethyl atrazine

Desisopropyl atrazine

Metolachlor

Metribuzin

25/09/2011 9/10/2011 16/01/2012 24/01/2012 Total (% applied)

5 20 118 126

5.4 0.8 1.8 47.6

18.79 0.39 0.003 0.04 1.42

0.97 0.17 0.001 0.02

0.38 0.07 0.001 0.01

7.52 0.37 0.008 0.21 0.65

0.86 6.19 0.52

Refer Table 3 for runoff total volume.

duplicates analysis was conducted so that each batch of samples included 5% blank analysis, 10% duplicate analysis and 10% high/low spikes. 2.6. Statistical analysis The field (each management unit) was divided into three replicates along the slope (top, middle and bottom). Final yield was estimated for three replicates of each practice and the results were presented as mean and standard error (Table 8). However, the herbicide active ingredient analysis of soil and trash were carried out in the composite sample of these three replicates to monitor the degradation. The herbicide active ingredient loads in runoff from different management practices were compared using relative percentage differences between management practices. 3. Results and discussion 3.1. Rainfall and runoff Total rainfall received during the sugarcane plant crop was 1603 mm (Aug 6, 2011 to Sep 13, 2012) which was higher than long term average of 1030 mm. Monthly cumulative rainfall is presented in Fig. 2 which covers both the plant cane and first ratoon crop growth periods. After herbicide application, significant runoff-generating rainfall or irrigation events (in all the management systems) occurred on 25 September and 9 October in 2011; and 16, 17, 24, 25, 26 and 28 January 2012 during the herbicide monitoring period. Among these events, runoff samples were analysed for herbicides on 25 September, 9 October, 16 January and 24 January, which corresponded to 5, 20, 118 and 126 days after first herbicide application (Table 2A). The first runoff event after herbicide application occurred due to irrigation (33 mm). Runoff volumes of selected events are presented in Tables 4 to 6, with a cumulative summary in Table 3. The sugarcane first ratoon crop experienced an extreme rainfall event during Jan 2013. Rainfall volumes recorded at the site on 25, 26 and 27 January 2013 were 131, 134 and 286 mm respectively. This resulted in extreme runoff events above the flume capacity and the herbicide loads were not calculated due to low confidence in the flow volumes. 3.2. Herbicide dynamics in soil and runoff The runoff losses and degradation of herbicides are influenced by several factors which include microbial break down, volatilisation, physical and chemical properties of the herbicides themselves, photodegradation, soil organic carbon, water table depth, temperature, soil

moisture content, soil texture, antecedent days and time after application (Baker and Mickelson, 1994; Kadian et al., 2008; Koskinen and Banks, 2008; Ng et al., 1995; Topp and Smith, 1994). Of particular interest in field studies is the relative solubility of the different products in water, which can be related to both leaching and runoff losses. The solubility of atrazine, diuron, metolachlor, metribuzin and pendimethalin in water have been reported as 35, 35.6, 530, 1165 and 0.33 mg/L (Table 2B) with the pronounced differences in solubility and soil sorption co-efficient of these herbicides likely to have had an impact on leaching losses and persistence in our study. However, the results suggest soil and trash management had a larger effect on the herbicide losses in runoff than the physico-chemical properties of herbicides. 3.2.1. Atrazine, its metabolite and metolachlor losses in runoff The four different sugarcane cropping systems produced different concentrations and loads of herbicides in runoff water. For example, while there were similar runoff amounts generated from the NFS and Conventional practices, there was a 20% reduction in atrazine loads in the runoff water from the NFS which could be attributed to the 30% reduction in application rate of atrazine (c.f. Conventional practice). Similarly, Improved practice produced 40% less runoff which, coupled with the reduced atrazine application rates, resulted in 62% less atrazine in the runoff water compared to Conventional practice (Tables 2A and 3). A recent study reported a twofold reduction in herbicide loss from using conservation tillage practices compared to conventional tillage (Potter et al., 2015). The avoidance of use of residual herbicides in favour of knock downs completely eliminated residual herbicides in runoff in the Aspirational practices, but was still able to control weeds successfully in the plant cane crop. This was consistent with previous studies which examined banded application of herbicides, where results showed that more strongly sorbed products like glyphosate were a good alternative to residuals such as diuron and atrazine for use in furrows (Oliver et al., 2014). While both Improved and NFS practices received similar herbicides and application rates (Table 2A), NFS always recorded higher loads of applied herbicide active ingredients in runoff water after the first herbicide application in September (Tables 4 and 5). The significant and interesting finding from this study was the 50% lower atrazine losses from Improved practices compared to NFS, despite both the systems receiving the same application rates. This result was likely due to the retention of applied herbicides on trash (retained from the previous soybean crop), increasing the vulnerability of the herbicide to being washed off the trash and into runoff water. The trend was

Table 5 Herbicide active ingredient loads (g/ha) in runoff water in Improved practice (trash removed, zonal tillage in beds) during the sugarcane plant crop. Herbicide application rates were atrazine-1350 g a.i./ha, metolachlor – 1248 g a.i./ha, metribuzin-1350 g a.i./ha. Date

Days after application

Runoff (mm)

Atrazine

Desethyl atrazine

Desisopropyl atrazine

Metolachlor

Metribuzin

25/09/2011 9/10/2011 24/01/2012 Total (% applied)

5 20 126

2.5 6.0 23.1

6.61 2.65 0.009 0.69

0.43 0.84

0.24 0.37

3.31 2.90 0.04 0.50

0.53 0.04

Refer Table 3 for runoff total volume.

G. Nachimuthu et al. / Science of the Total Environment 557–558 (2016) 773–784

779

Table 6 Herbicide active ingredient loads (g/ha) in runoff water in Conventional (trash removed, full tillage) during the sugarcane plant crop. Herbicide application rates were atrazine-1980 g a.i./ ha, pendimethalin – 1001 g a.i./ha, metribuzin-1350 g a.i./ha. Date

Days after application

Runoff (mm)

Atrazine

Desethyl atrazine

Desisopropyl atrazine

Pendimethalin

25/09/2011 9/10/2011 16/01/2012 24/01/2012 Total (% applied)

5 20 118 126

4.6 9.8 4.4 37.3

21.34 2.73 0.001 0.01 1.22

1.11 0.98 0.00

0.65 0.41

13.45 1.37 0.009 1.48

Metribuzin

0.57 0.93 0.11

Refer Table 3 for runoff total volume.

still reflected 126 days after spray application, with herbicide active ingredient loads of both atrazine and metolachlor greater in runoff from NFS than Improved practices – although in both cases herbicide active ingredient loads were by then very low (0.21 and 0.04 g/ha respectively, Tables 4 and 5). Atrazine runoff losses were 1.4% of the mass applied in the NFS, with trash retained, compared to 0.69% and 1.22% for the Improved and Conventional practices where the soil surface was bare. The reasons for greater runoff loss of atrazine from the management practices with trash retained could be related to the timing of runoff events after application, the differential sorption behaviour of soil and trash and antecedent conditions. Residual herbicides are known to be easily washed off crop residues (Martin et al., 1978), so herbicide runoff can be increased or decreased with surface trash depending on the timing of runoff events after the herbicide application (Masters et al., 2013; Thorburn et al., 2013). The time course concentrations of atrazine and metolachlor in the soil in the NFS system indicate an increase in their concentrations (Fig. 5A and 5C) 10 days after application, suggesting herbicides have washed off the trash and onto the soil. This biphasic compartment behaviour between soil and trash for these herbicides suggests a future investigation on impact of rainfall intensity on partitioning of herbicide movement between soil and runoff from trash is warranted. The atrazine concentrations recorded (Fig. 3a) in the first runoff event after application were several hundred times greater than the trigger value of 0.7 μg L−1 for protection of freshwater species (ANZECC, 2000). However, the peak concentration recorded for atrazine in any single event and the total loads of atrazine were much lower than recorded in other studies in irrigated row cropping systems (Davis et al., 2013; Oliver and Kookana, 2006a; Oliver and Kookana, 2006b). Along with atrazine itself, the principle metabolites desethyl atrazine and deisopropyl atrazine (Koskinen and Banks, 2008) were also present in runoff water (Tables 3 to 6). Other studies have reported either similar (Gerritse et al., 1996; Oliver et al., 2003) or lower (Moreau and Mouvet, 1997; Seybold and Mersie, 1996) sorption affinity of these metabolites to the soil compared to atrazine itself. Considering the free draining nature of the soil used in this experiment (Isbell, 2002), weakly sorbed or moderately mobile atrazine metabolites could easily reach the groundwater, as reported in earlier studies (Oliver et al., 2003; Topp and Smith, 1994). Though solubility of atrazine in water is low, its Groundwater Ubiquity Score (GUS) leaching potential index is high (Table 2B) and the off-site movement risk is higher for the days immediately after application. The half-life of atrazine (29 days) reported in the pesticide properties database (Table 2B) and the time course herbicide Table 7 Herbicide event mean concentrations (EMC) in different management systems during the sugarcane first ratoon crop on 26 Jan 2013. Management practices

Herbicides

Concentrations (μg L−1)

Days after spray

Aspirational Improved Conventional Conventional Conventional

2,4-D Metribuzin Diuron Atrazine 2,4-D

0.37 0.40 14 0.01 0.20

79 79 79 79 79

concentration in soil (Fig. 5A) suggest the maximum risk is within the first 10 days after application. Metolachlor losses in Improved and NFS practices were 0.003% and 0.6% of the applied herbicide in the first runoff event, with values dropping to 0.002% and 0.03%, respectively, in the second runoff event (Tables 4 and 5). These large decreases with time are shown in Fig. 3. Metolachlor event mean concentrations (Fig. 3) in our studies were much lower than those detected in the Fitzroy basin central Queensland, Australia (Murphy et al., 2012). The GUS leaching potential index (Gustafson, 1989) indicates metolachlor is in intermediate category in terms of leachability (Table 2B) and the soil sorption co-efficient suggests a moderate mobility for metolachlor.

3.2.2. Pendimethalin losses in runoff This study observed some unusual pendimethalin dynamics in soil and runoff. Initial soil concentrations of pendimethalin on day 0.5 after application were 0.91 mg kg−1, which was b1/3rd of the applied rate of 3.3 mg kg−1 (Fig. 5D), suggesting a very rapid degradation in 12 h. Inspite of the low initial pendimethalin concentration in soil and the pesticide properties database suggesting this product had a high soil sorption co-efficient (Koc = 17.581), was non-mobile and had a low GUS leaching potential index (Gustafson, 1989) (Table 2B), the total losses of pendimethalin were 1.48% of the applied herbicide. This was higher than atrazine (1.22% of applied herbicide). These apparently contradictory herbicide behaviours may be related to the type of analytical method employed on the runoff samples. The pendimethalin analysis included a solvent extraction which would account for any herbicide in the sediment fraction, but atrazine analysis did not account for any herbicide in that sediment fraction. The time course dissipation of pendimethalin in soil shows the concentration dropped by N60% within 10 days after spraying, suggesting a rapid dissipation in this sub-tropical climate. Previous studies have reported a wide variation in half-life for pendimethalin, ranging from 0 to 200 days, with the pesticide properties database reporting its halflife as 90 days in field (Table 2B). Under controlled conditions, the half-life of pendimethalin from the soil taken from this field experimental site was found to be 34 days (Shaw et al., 2013), while another study in Australia reported the half-life of pendimethalin as 33 and 38 days in the top 2.5 and 5 cm soil layers, respectively (Silburn, 2003). Strandberg and Scott-Fordsmand (2004) reported up to 20% volatilisation of pendimethalin in the first week after application. The same authors revealed lower temperatures and prolonged drought increased the dissipation time of pendimethalin to as long as 2094 days, although other field studies in the United States estimated that the concentrations of Table 8 Sugar yield of different management systems. Management practices

Sugar yield (t/ha) (mean ± standard error)

Aspirational Improved Conventional New Farming Systems

19.01 ± 1.075 20.10 ± 0.178 21.68 ± 1.450 18.57 ± 1.133

780

G. Nachimuthu et al. / Science of the Total Environment 557–558 (2016) 773–784

Fig. 3. Concentrations (μg L−1) of atrazine (a) and metolachlor (b) in runoff water for the sugarcane plant crop during 2011–2012.

pendimethalin in soil, water, and sediment were below the level of ecological concern (Patricia et al., 2002).

Fig. 4. Time course of diuron loads in soil and trash starting at the time of application in first ratoon sugarcane crop.

3.2.3. Metribuzin losses in runoff Metribuzin was applied 114 days after the other residual herbicides and experienced a different sequence of rainfall immediately after application. There were N10-fold variation in metribuzin loads in runoff water between reduced tillage systems (NFS and Improved) differing in soil disturbance and surface residue cover from the previous rotation crops, despite the same herbicide application rates. Metribuzin has also been reported to be quite mobile (Table 2B) and able to leach down the soil profile in previous studies (Jebellie and Prasher, 1998), so leaching losses of metribuzin may also have been significant. Time course dissipation of metribuzin in soil suggest the concentration dropped by N 10 and 30 times for conventional and NFS practices, respectively, within 5 days after spraying (Fig. 5), suggesting a rapid dissipation in this sub-tropical climate. The half-life of metribuzin derived from a dissipation study conducted simultaneously using the same soil under controlled glasshouse condition was found to be 16 days (Shaw et al., 2013), which closely aligns with the pesticide property database halflife reported for metribuzin (Table 2B). While metribuzin has a

G. Nachimuthu et al. / Science of the Total Environment 557–558 (2016) 773–784

781

Fig. 5. Time course of herbicide concentrations in soil starting at the time of application in plant cane crop. Data are shown for (a) atrazine, (b) metribuzin, (c) metolachlor in the respective management systems and (d) pendimethalin in Conventional systems.

relatively low half-life compared to other residual herbicides (atrazine, diuron) used in sugarcane, a recent finding suggests the ecotoxicological profile of metribuzin is similar to atrazine and careful management is warranted if using metribuzin as an alternative to atrazine (Davis et al., 2014). 3.3. Herbicide losses in runoff during the first ratoon crop During the ratoon crop, there were no runoff-generating rainfall events for ca. 2.5 months after herbicide application (8 November 2012), resulting in substantial dissipation of applied herbicides and low concentrations in events when runoff was finally recorded (late Jan 2013). For example, 2,4-D concentrations in Aspirational practice on 25th Jan and 28th Jan 2013 were 0.37 and 0.19 μg L−1, while 2,4-D, atrazine and diuron concentrations in Conventional practice on 28th Jan 2013 were 0.20, 0.01 and 14 μg L− 1, respectively (Table 7). The metribuzin concentration in Improved practice on the same date was 0.40 μg L − 1 . The 2,4-D concentrations of 0.2 to 0.37 μg L − 1 were much lower than the ANZECC trigger value of 140 μg L− 1 for 2,4-D (ANZECC, 2000). Unfortunately, those events were extreme with very high rainfall (Fig. 2) and reliable estimates of runoff volumes and herbicide loads could not be derived. However, the presence of diuron at concentrations of 14 μg L− 1 2.5 months after application was notable (Table 7), confirming the persistence of diuron reported in other studies (Davis et al., 2012; Mitchell et al., 2005; Stork et al., 2008a). It should be noted there was evidence of diuron being washed off trash onto soil contributing a biphasic component to the degradation pattern (Fig. 4). This again suggests the herbicide partitioning dynamics from trash to soil and runoff warrant further investigation. The DT50 values of diuron in soil and trash were 56 and N100 days reported from controlled experiments with the same soil reported by Shaw et al. (2013). Other reports by Silburn (2003) showed the half-

life of diuron in soil was 21 and 23 days in the top 25 and 50 mm of the soil, respectively – much lower than the 89 days (Table 2B) reported in the pesticide properties database under field situations. Shallow surface soils used in glasshouse and field studies are exposed to variable climatic factors and so variability in DT50 values is probably not unexpected. However, sampling depth is one aspect that needs to be considered when comparing half-lives of herbicides in different studies. The slow breakdown of diuron in soil in our study was consistent with the detection of diuron in runoff from a sugarcane crop in the Mackay-Whitsunday region of Queensland, despite that chemical not being applied in that year (Rohde et al., 2011). Continuous monitoring of PS II herbicides such as diuron may be required given the lack of suitable models in Australia incorporating the various environmental parameters for agricultural lands of the GBR catchments (Holmes, 2014). 3.4. Influence of timing of runoff event on herbicide losses The time between the herbicide application and rainfall or irrigation significantly influenced the herbicide loads measured in runoff water in our study (Tables 2A, 4, 5 and 6), while trash cover (Table 1) also played a role in herbicide loss in runoff (Tables 4 and 5). Avoiding irrigationinduced runoff five days after herbicide application would have resulted in 98% and 95% lower loads in runoff of atrazine and metolachlor in NFS and ~ 87% and 90% lower loads of atrazine and pendimethalin in Conventional practice, when the next runoff event occurred on day 20 (Tables 4 and 6). While these estimates do not take into account the losses incurred in that irrigation event, we expect that any additional residual herbicide present would have had only a small impact on losses at day 20. The rapid reduction in runoff concentrations with time after application for both chemicals is illustrated in Fig. 3, with concentrations at day 20 ca. 10% (atrazine) to 30% (metolachlor) of those in the event at day 5. The physico-chemical properties of various

782

G. Nachimuthu et al. / Science of the Total Environment 557–558 (2016) 773–784

pesticide types (solubility, binding characteristics and persistence, Table 2B) are regarded as a major contributor to the management of risks of off-site movement after application (Davis et al., 2013). A number of the relatively soluble and moderately persistent herbicides (Table 2B) investigated in this study have low sorption affinity for soil particles and are inherently susceptible to off-site movement in solution. Our field scale monitoring results were consistent with these expectations. 3.5. Implications for water quality Concentrations of the PS II herbicides atrazine (Fig. 3) and diuron (Table 7) in runoff were much lower than the safe drinking water guideline values (NHMRC and NRMMC, 2004), suggesting no risk for human consumption in fresh water bodies. However, the event mean concentrations of atrazine (Fig. 3, Table 3) and diuron (Table 7) were higher than the safe guideline values indicating risks for marine and fresh water aquatic life (ANZECC, 2000; Chesworth et al., 2004; Davis et al., 2008; Haynes et al., 2000; Stork et al., 2008a). A previous study monitoring the herbicide concentrations from eight catchments in the GBR reported similar evidence of concentrations of PS II herbicides such as diuron and atrazine exceeding water quality guideline trigger values (ANZECC, 2000) in 8 of the 11 sites monitored (Smith et al., 2012). Runoff samples were collected as water left the cropping field in these studies and the herbicide concentrations would therefore be expected to be higher than those found in adjacent creeks and rivers. Earlier studies assessing the concentrations of herbicides in the field and nearby creeks found the herbicide concentrations in nearby creek systems were invariably an order of magnitude or more lower than values collected in the field, highlighting the significant dilution that takes place over relatively short distances (Davis et al., 2013). Policy makers need to account for this dilution factor when extrapolating from the results of this study. Similarly, geographical extrapolation of data on the environmental fate of these herbicides needs to account for management and climatic factors (Riaz and Kookana, 2007). 3.6. Cropping system comparison, herbicides and weed control Sugarcane cropping systems receiving the same (Conventional) or lower (Improved and NFS practices) than the recommended herbicide rates achieved similar levels of weed control in the sugarcane plant crop (no notable weed population observed), suggesting there may be scope for reducing the current conventional herbicide rates to achieve both economic and environmental benefits. The higher runoff loss in NFS (trash present) compared to Improved practice (no trash) with similar application rates suggest that herbicide application rates should be reduced further where surface residues are present. A precautionary approach would be to apply the 30% lower herbicide rates (or similar) to sugarcane trash blankets to achieve a desired water quality outcome, although further testing of such an approach would be required to both confirm these findings and also determine whether the desired level of weed control could be obtained. In the case of a trade-off between effective weed control and water quality outcomes for a ratoon crop with a trash blanket, alternate options such as less persistent knock-down herbicides need to be explored. Earlier studies have also reported reduced herbicide runoff from controlled traffic plots (Silburn et al., 2013), and location-specific application methods such as banding (Masters et al., 2013; Oliver et al., 2014) that can be achieved in these systems may be another way of achieving improved water quality outcomes. Although Improved practice recorded slightly higher sugar yield than NFS practice (Table 8), the weed control in those two systems were similar. However, there was a marked differences in atrazine and metribuzin loads in runoff water, suggesting that even fine modifications to tillage and trash management practices may make a significant impact to runoff water quality. Clearly a better understanding of the

mechanisms underlying the resultant herbicide loads needs to be gained before any recommendations can be made, with differences in residue cover on the beds (more cover in NFS) and possibly in the interspaces (residual inter row mulch grown during the preceding vegetable rotation phase in Improved practice) possibly contributing to these effects. The strategies of employing knock-down herbicides and an interrow soybean mulch in Aspirational practice, instead of residual herbicides, resulted in adequate weed control in the plant cane crop and a complete absence of residual herbicides detected in runoff. However there was a significant decline in crop yield (12.3% lower sugar yield compared to Conventional practice, Table 8), possibly due to the soybean companion crop effectively acting as a weed and constraining cane growth. Further research to determine ways of limiting these negative companion planting effects, while retaining any soil health, weed suppression or fixed nitrogen benefits, needs to be undertaken before such strategies could be adopted commercially. Unlike the experience in the plant crop, replacing residual herbicides with knockdown herbicides (Table 2A) in Aspirational practice and NFS in the 1st ratoon crop resulted in poor weed control and additional herbicide applications were necessary at a later stage of the crop. The presence of the trash layer in the ratoon crops adds a degree of complexity to weed control strategies. A recent review suggested a strong relationship between the depth or thickness of trash and weed control (Marble, 2015), with any trash layer with a thickness of 7 cm or more giving satisfactory weed control without herbicide application. However, in thinner trash layers where herbicide applications are necessary, interactions between the herbicide and the trash can be pronounced (Marble, 2015). While some sugarcane ratoons may start with a cane trash mulch of 7 cm or more, the thickness will rapidly decline during the first few months of a ratoon crop, resulting in herbicide-mulch interactions that have implications for weed control. In particular, the delayed emergence of weeds due to early trash cover will limit the effectiveness of knockdown applications while crop access is still possible, as observed in Aspirational and NFS practices in this study. Also, while it is true fast degradation of alternate residual herbicides may result in less offsite impact, this may also lead to poor residual weed control as reported in earlier studies (Krutz et al., 2010). Thus, faster dissipation of residual herbicides will have further implications on agronomic weed management strategies and a trade-off will exist between better weed management to achieve profitability and reducing the off-site impacts from herbicide contamination.

4. Conclusions This study compared Conventional practices with a range of other practices in measuring offsite herbicide losses in a sugarcane cropping system and provided some clear outcomes on reducing herbicide active ingredient loads in runoff water under improved management practice. However, these reductions in loads were achieved at the expense of some reduction in productivity. This needs further investigation before a commercial recommendation can be made. The findings clearly suggest it is not only the rate of herbicide applied but also the soil and trash management practices (tillage and mulching) that influence the offsite movement of herbicides. A future investigation on effect of rainfall intensity, trash levels and choice of herbicide on herbicide partitioning from trash to soil and runoff under field situations is warranted. There is also potential to explore alternative options such as new products with less environmental risk and different application methods such as spot spraying and banding. Furthermore, other onfarm and off-farm herbicide contamination mitigation strategies such as buffer strips on the edges of fields and riparian buffer strips, vegetated ditches and constructed wetlands need to be considered (Vymazal and Březinová, 2015), in addition to improved herbicide management practices, to reduce off-site impacts.

G. Nachimuthu et al. / Science of the Total Environment 557–558 (2016) 773–784

Funding source This paper is a contribution to the Paddock to Reef Monitoring, Modelling and Reporting Program. Funding from the Queensland Department of Natural Resources and Mines, Burnett Mary Regional Group and the Australian Government's Caring for our Country Program (Grant number- A11016 Burnett Mary Paddock Monitoring Program) is gratefully acknowledged. Acknowledgement Many thanks go to the numerous technical, administrative and professional staff at the Bundaberg Research Facility at Kalkie, who assisted in the establishment and conduct of this study. Special mention goes to Steve Ginns, Bill Rehbein and Sherree Short. We would also acknowledge the assistance provided by Burnett Mary Regional group staff including Sally Fenner, Cathy Mylrea and Fred Bennett. Don Halpin (and his son Andrew Halpin) permitted us to use their farm to conduct the trial and we acknowledge their generosity in time and in the conduct of crop management operations. Help and suggestions from Melanie Shaw and Mark Silburn and the other P2R monitoring team members regarding sampling protocols are acknowledged. We acknowledge Mark Silburn, Michael Rose and Rai Kookana for their valuable feedback and suggestions. References ANZECC, 2000. Australian and New Zealand Guidelines for Fresh and Marine Water Quality (2000) Australian and New Zealand Environment and Conservation Council. NHMRC, NRMMC, 2004. Australian Drinking Water Guidelines 2004, National Water Quality Management Strategy. Nation Health and Medical Research Council and the Natural Resource Management Ministerial Council. Australian Government, Canberra. Baker, J.L., Mickelson, S.K., 1994. Application technology and best management practices for minimizing herbicide runoff. Weed Technol. 8, 862–869. Brodie, J.E., Kroon, F.J., Schaffelke, B., Wolanski, E.C., Lewis, S.E., Devlin, M.J., Bohnet, I.C., Bainbridge, Z.T., Waterhouse, J., Davis, A.M., 2012. Terrestrial pollutant runoff to the Great Barrier Reef: an update of issues, priorities and management responses. Mar. Pollut. Bull. 65, 81–100. http://dx.doi.org/10.1016/j.marpolbul.2011.12.012. Chesworth, J.C., Donkin, M.E., Brown, M.T., 2004. The interactive effects of the antifouling herbicides Irgarol 1051 and Diuron on the seagrass Zostera marina (L.). Aquat. Toxicol. 66, 293–305. http://dx.doi.org/10.1016/j.aquatox.2003.10.002. Davis, A., Lewis, S., Bainbridge, Z., Brodie, J., Shannon, E., 2008. Pesticide residues in waterways of the lower burdekin region: challenges in ecotoxicological interpretation of monitoring data. Aust. J. Ecotoxicol. 14, 89–108. Davis, A.M., Lewis, S.E., Bainbridge, Z.T., Glendenning, L., Turner, R.D.R., Brodie, J.E., 2012. Dynamics of herbicide transport and partitioning under event flow conditions in the lower Burdekin region. Aust. Mar. Pollut. Bull. 65, 182–193. http://dx.doi.org/10. 1016/j.marpolbul.2011.08.025. Davis, A.M., Thorburn, P.J., Lewis, S.E., Bainbridge, Z.T., Attard, S.J., Milla, R., Brodie, J.E., 2013. Environmental impacts of irrigated sugarcane production: herbicide run-off dynamics from farms and associated drainage systems. Agric. Ecosyst. Environ. 180, 123–135. http://dx.doi.org/10.1016/j.agee.2011.06.019. Davis, A.M., Lewis, S.E., Brodie, J.E., Benson, A., 2014. The potential benefits of herbicide regulation: a cautionary note for the Great Barrier Reef catchment area. Sci. Total Environ. 490, 81–92. http://dx.doi.org/10.1016/j.scitotenv.2014.04.005. Gerritse, R., Beltran, J., Hernandez, F., 1996. Adsorption of atrazine, simazine, and glyphosate in soils of the Gnangara Mound. Western Aust. Soil Res. 34, 599–607. http://dx.doi.org/10.1071/SR9960599. Gustafson, D.I., 1989. Groundwater ubiquity score: a simple method for assessing pesticide leachability. Environ. Toxicol. Chem. 8, 339–357. Haynes, D., Ralph, P., Prange, J., Dennison, B., 2000. The impact of the herbicide diuron on photosynthesis in three species of tropical seagrass. Mar. Pollut. Bull. 41, 288–293. http://dx.doi.org/10.1016/S0025-326X(00)00127-2. Hollinger, E., Cornish, P.S., Baginska, B., Mann, R., Kuczera, G., 2001. Farm-scale stormwater losses of sediment and nutrients from a market garden near Sydney. Aust. Agric. Water Manag. 47, 227–241. http://dx.doi.org/10.1016/S03783774(00)00107-4. Holmes, G., 2014. Australia's pesticide environmental risk assessment failure: the case of diuron and sugarcane. Mar. Pollut. Bull. 88, 7–13. http://dx.doi.org/10.1016/j. marpolbul.2014.08.007. Reef Rescue program, 2008. http://www.nrm.gov.au/national/continuing-investment/ reef-programme. Isbell, R., 2002. The Australian Soil Classification. CSIRO, Canberra, Australia. Jebellie, S.J., Prasher, S.O., 1998. Role of water table management in reducing metribuzin pollution. Trans. Am. Soc. Agric. Eng. 41, 1051–1060. Jones, R., 2005. The ecotoxicological effects of Photosystem II herbicides on corals. Mar. Pollut. Bull. 51, 495–506. http://dx.doi.org/10.1016/j.marpolbul.2005.06.027.

783

Kadian, N., Gupta, A., Satya, S., Mehta, R.K., Malik, A., 2008. Biodegradation of herbicide (atrazine) in contaminated soil using various bioprocessed materials. Bioresour. Technol. 99, 4642–4647. King, J., Alexander, F., Brodie, J., 2013. Regulation of pesticides in Australia: the Great Barrier Reef as a case study for evaluating effectiveness. Agric. Ecosyst. Environ. 180, 54–67. http://dx.doi.org/10.1016/j.agwat.2009.01.001. Koskinen, W.C., Banks, P.A., 2008. Chapter 24 - soil movement and persistence of triazine herbicides. In: LeBaron, H.M., McFarland, J.E., Burnside, O.C. (Eds.), The Triazine Herbicides. Elsevier, San Diego, pp. 355–385 http://dx.doi.org/10.1016/B978044451167-6.50027-1. Krutz, J.L., Shaner, D.L., Weaver, M.A., Webb, R.M.T., Zablotowicz, R.M., Reddy, K.N., Huang, Y., Thomson, S.J., 2010. Agronomic and environmental implications of enhanced s-triazine degradation. Pest Manag. Sci. 66, 461–481. http://dx.doi.org/ 10.1002/ps.1909. Marble, C., 2015. Herbicide and mulch interactions: a review of the literature and implications for the landscape maintenance industry. Weed Technol. 29, 341–349. Martin, C.D., Baker, J.L., Erbach, D.C., Johnson, H.P., 1978. Washoff of herbicides applied to corn residue. Trans. Am. Soc. Agric. Eng. 21, 1164–1168. Masters, B., Rohde, K., Gurner, N., Higham, W., Drewry, J., 2008. Sediment, Nutrient and Herbicide Runoff from Canefarming Practices in the Mackay Whitsunday Region: A Field-based Rainfall Simulation Study of Management Practices. Queensland Department of Natural Resources and Water for the Mackay Whitsunday Natural Resource Management Group, Australia. Masters, B., Rohde, K., Gurner, N., Reid, D., 2013. Reducing the risk of herbicide runoff in sugarcane farming through controlled traffic and early-banded application. Agric. Ecosyst. Environ. 180, 29–39. http://dx.doi.org/10.1016/j.agee.2012.02.001. Mitchell, C., Brodie, J., White, I., 2005. Sediments, nutrients and pesticide residues in event flow conditions in streams of the Mackay Whitsunday Region. Aust. Mar. Pollut. Bull. 51, 23–36. Moreau, C., Mouvet, C., 1997. Sorption and desorption of atrazine, deethylatrazine, and hydroxyatrazine by soil and aquifer solids. J. Environ. Qual. 26, 416–424. Murphy, T., Dougall, C., Burger, P., Carroll, C., 2012. Runoff water quality from dryland cropping on Vertisols in Central Queensland. Aust. Agr. Ecosyst. Environ. 180, 21–28. Nachimuthu, G., Halpin, N., Bell, M., 2013. Paddock scale water quality monitoring of vegetable-sugarcane and legume-sugarcane farming systems- final technical report submitted to Australian and Queensland Government. Department of Agriculture, Fisheries and Forestry (QLD) & Queensland Alliance for Agriculture and Food Innovation (QAAFI). The University of Queensland, p. 153. Ng, H.Y.F., Gaynor, J.D., Tan, C.S., Drury, C.F., 1995. Dissipation and loss of atrazine and metolachlor in surface and subsurface drain water: a case study. Water Res. 29, 2309–2317. Oliver, D.P., Kookana, R.S., 2006a. Minimising off-site movement of contaminants in furrow irrigation using polyacrylamide (PAM). I. Pestic. Soil Res. 44, 551–560. http://dx.doi.org/10.1071/SR05197. Oliver, D.P., Kookana, R.S., 2006b. On-farm management practices to minimise off-site movement of pesticides from furrow irrigation. Pest Manag. Sci. 62, 899–911. Oliver, D.P., Kookana, R.S., Salama, R.B., Correll, R., 2003. Land use effects on sorption of pesticides and their metabolites in sandy soils. II. Atrazine and two metabolites, deethylatrazine and deisopropylatrazine, and prometryne. Aust. J. Soil Res. 41, 861–874. http://dx.doi.org/10.1071/SR02095. Oliver, D.P., Anderson, J.S., Davis, A., Lewis, S., Brodie, J., Kookana, R., 2014. Banded applications are highly effective in minimising herbicide migration from furrow-irrigated sugar cane. Sci. Total Environ. 466–467, 841–848. http://dx.doi.org/10.1016/j. scitotenv.2013.07.117. O'Reagain, P.J., Brodie, J., Fraser, G., Bushell, J.J., Holloway, C.H., Faithful, J.W., Haynes, D., 2005. Nutrient loss and water quality under extensive grazing in the upper Burdekin river catchment, North Queensland. Mar. Pollut. Bull. 51, 37–50. Patricia, J.R., Gary, D.M., Maximilian, M.S., 2002. Determining the dissipation and potential offsite movement of pendimethalin through a monitoring survey and environmental modeling, terrestrial field dissipation studies. American Chemical Society 273–286 http://dx.doi.org/10.1021/bk-2002-0842.ch018. Potter, T.L., Bosch, D.D., Strickland, T.C., 2015. Tillage impact on herbicide loss by surface runoff and lateral subsurface flow. Sci. Total Environ. 530–531, 357–366. http://dx. doi.org/10.1016/j.scitotenv.2015.05.079. Råberg, S., Nyström, M., Erös, M., Plantman, P., 2003. Impact of the herbicides 2,4-D and diuron on the metabolism of the coral Porites cylindrica. Mar. Environ. Res. 56, 503–514. http://dx.doi.org/10.1016/S0141-1136(03)00039-4. Riaz, A., Kookana, R.S., 2007. Geographical extrapolation of pesticide environmental fate data: challenges, risks, and opportunities, rational environmental management of agrochemicals. Am. Chem. Soc. 100–119 http://dx.doi.org/10.1021/bk-2007-0966. ch007. Rohde, K., Bush, A., Agnew, J., 2011. Paddock to Sub-catchment Scale Water Quality Monitoring of Sugarcane Management Practices. Interim Report 2010/11 Wet Season. Mackay Whitsunday Region. Department of Environment and Resource Management, Queensland Government for Reef Catchments (Mackay Whitsunday Isaac) Limited, Australia. Seybold, C.A., Mersie, W., 1996. Adsorption and desorption of atrazine, deethylatrazine, deisopropylatrazine, hydroxyatrazine, and metolachlor in two soils from Virginia. J. Environ. Qual. 25, 1179–1185. Shaw, M., Silburn, D.M., Ponce, S.R., Lewis, S., Davis, A., 2013. Herbicide degration on Queensland cropping soils and crop residue -half-lives measured in a controlled environment. Queensland Department of Natural Resources and Mines and James Cook University, p. 66. Silburn, D.M., 2003. Characterising Pesticide Runoff from Soil on Cotton Farms using a Rainfall Simulator Phd thesis. Faculty of Agriculture, Food and Natural Resources, The University of Sydney.

784

G. Nachimuthu et al. / Science of the Total Environment 557–558 (2016) 773–784

Silburn, D.M., Foley, J.L., deVoil, R.C., 2013. Managing runoff of herbicides under rainfall and furrow irrigation with wheel traffic and banded spraying. Agric. Ecosyst. Environ. 180, 40–53. Smith, R., Middlebrook, R., Turner, R., Huggins, R., Vardy, S., Warne, M., 2012. Large-scale pesticide monitoring across Great Barrier Reef catchments – paddock to reef integrated monitoring, modelling and reporting program. Mar. Pollut. Bull. 65, 117–127. Stork, P.R., Bennett, F.R., Bell, M.J., 2008a. The environmental fate of diuron under a conventional production regime in a sugarcane farm during the plant cane phase. Pest Manag. Sci. 64, 954–963. http://dx.doi.org/10.1002/ps.1593. Stork, P.R., Bennett, F.R., Bell, M.J., Moody, P.W., 2008b. Benchmarking pesticides and nutrients in horticulture and new sugar cane farming systems. Department of Primary Industries & Fisheries (DPI&F), Bundaberg and Kingaroy. Strandberg, M., Scott-Fordsmand, J.J., 2004. Effects of pendimethalin at lower trophic levels—a review. Ecotoxicol. Environ. Saf. 57, 190–201. http://dx.doi.org/10.1016/j. ecoenv.2003.07.010. Sugarcane BMP, 2015. Smartcane Best Management Practice (BMP).

Thorburn, P.J., Wilkinson, S.N., Silburn, D.M., 2013. Water quality in agricultural lands draining to the Great Barrier Reef: a review of causes, management and priorities. Agric. Ecosyst. Environ. 180, 4–20. http://dx.doi.org/10.1016/j.agee.2013.07.006. Topp, E., Smith, W.N., 1994. Soil Persistence of Atrazine, Metolachlor, and Metribuzin as Influenced by Temperature, Soil Moisture, and Soil Characteristics. Centre for Land and Biological Resources Research. Research Branch, Agriculture and Agri-Food Canada, Ottawa, Ontario, Canada. Vymazal, J., Březinová, T., 2015. The use of constructed wetlands for removal of pesticides from agricultural runoff and drainage: a review. Environ. Int. 75, 11–20. http://dx.doi. org/10.1016/j.envint.2014.10.026. Walkowiak, D.k., 2008. ISCO Open Channel Flow Measurement Handbook. 6 ed. TELEDYNE ISCO. Wilm, H.G., Cotton, J.S., Storey, H.C., 1936. Experiments with critical-depth flumes for measurement of flow in debris-laden streams. Eos. Trans. AGU 17, 528-528. http:// dx.doi.org/10.1029/TR017i002p00528-1.