Weed and insect management alter soil arthropod densities, soil nutrient availability, plant productivity, and aphid densities in an annual legume cropping system

Applied Soil Ecology xxx (xxxx) xxx–xxx

Contents lists available at ScienceDirect

Applied Soil Ecology journal homepage: www.elsevier.com/locate/apsoil

Weed and insect management alter soil arthropod densities, soil nutrient availability, plant productivity, and aphid densities in an annual legume cropping system Ashton A. Hansena,d, Amitava Chatterjeeb, Greta Gramigc, Deirdre A. Prischmann-Voldsetha,

⁎

a

North Dakota State University, Department of Entomology 7650, Fargo, ND 58108-6050, United States North Dakota State University, Department of Soil Science, Dept. 7680, Fargo, ND 58108-6050, United States c North Dakota State University, Department of Plant Sciences, Dept. 7670, Fargo, ND 58108-6050, United States d Farm Service Agency, Lake County Office, 123 SW 2nd St., Madison, SD 57042, United States b

A R T I C LE I N FO

A B S T R A C T

Keywords: Glyphosate-resistant soybean Chlorpyrifos Collembola Soybean aphid

Transgenic glyphosate-resistant (GR) soybean (Glycine max L. Merr.) is cultivated throughout the United States. Soybean growth is influenced by the presence of weeds, although managing non-crop vegetation can potentially impact soil arthropods, which are being increasingly recognized for their impacts on soil health, plant growth, and above-ground trophic interactions. We investigated how weed management (weedy controls, hand-weeded, glyphosate herbicide) and soil insecticide (chlorpyrifos) application affected densities of soil arthropods, soil nutrient availability, soybean growth and yield, and densities of an above-ground herbivore on sandy and clayey soils for two consecutive growing seasons. The soil insecticide treatment was intended to lower densities of subterranean arthropods to gain insight into how their presence influenced other factors, although their densities were primarily reduced the first year of the study. Surprisingly, weed management and soil insecticide use had virtually no interactive effects on any response. Weed presence had a positive effect on soil K at the sandy site and on nodule density per unit root. Negative effects of weed management on plant growth and aphids were related to the presence of weeds rather than herbicide use. Reduced soybean aphid density (at the clayey site) and soil P availability (at the sandy site) were associated with insecticide treated plots. Conversely, several measures of plant productivity, including number of nodules per unit root, and root and shoot biomass increased in +insecticide plots compared to other treatments, although effect strength depended on year and location. Collembola were the dominant soil microarthropod, and their densities in 2012 were negatively associated with nodule numbers in 2013. One explanatory hypothesis is that increased plant growth in plots treated with insecticide was caused by altered soil arthropod-microorganism interactions, possibly affecting arbuscular mycorrhizae function. This work highlights the importance of management decisions that affect soil arthropods in annual legume production systems.

1. Introduction The role of soil microarthropods in plant production and soil health is being increasingly recognized (Brussaard et al., 2007; Ferris and Tuomisto, 2015; Bender et al., 2016). Although some directly affect plants via root consumption (Hopkin, 1997; Endlweber et al., 2009), many affect plants indirectly via brown food webs or by influencing organic matter decomposition and nutrient cycling (Petersen and Luxton, 1982; Seastedt, 1984; Moore et al., 1988; Neher and Barbercheck, 1998), or via the presence of their cadavers (Kos et al., 2017). Soil arthropods may be especially important in legume crops, like soybean, that depend on symbiotic bacteria to fix nitrogen ⁎

(Lussenhop, 1996). Soil microarthropods can also impact foliar herbivores (Scheu et al., 1999; Schütz et al., 2008; Megías and Müller, 2010) and their natural enemies (Scheu, 2001; A’Bear et al., 2014). Collembola (springtails, Arthropoda: Entognatha) and mites (Arachnida: Acari) are among the most abundant soil microarthropods (Seastedt, 1984; Norton, 1990; Hopkin, 1997). Collembola are found world-wide and primarily feed on fungi or decaying plant material (Hopkin, 1997; Rusek, 1998). Some taxa are thought to consume arbuscular mycorrhizal fungi symbiotically associated with plant roots (Warnock et al., 1982; Jonas et al., 2007; Caravaca and Ruess, 2014), potentially disrupting uptake of key nutrients, especially phosphorus, that are critical for plant growth and legume productivity (Keyser and

Corresponding author. E-mail address: [email protected] (D.A. Prischmann-Voldseth).

https://doi.org/10.1016/j.apsoil.2018.06.006 Received 12 December 2017; Received in revised form 6 June 2018; Accepted 7 June 2018 0929-1393/ © 2018 Elsevier B.V. All rights reserved.

Please cite this article as: Hansen, A.A., Applied Soil Ecology (2018), https://doi.org/10.1016/j.apsoil.2018.06.006

Applied Soil Ecology xxx (xxxx) xxx–xxx

A.A. Hansen et al.

weedy control) and using a soil insecticide to manipulate soil arthropods (natural level, suppressed) affected soil nutrients, growth of glyphosate-resistant soybean, and densities of an above-ground herbivore under sandy (coarse-textured) and clayey (fine-textured) soils for two consecutive growing seasons. We established the field study in two disparate soil contexts because soil structure and abiotic properties can impact the abundance, diversity, and behavior of soil organisms (Villani and Wright, 1990; Lauber et al., 2008; Birkhofer et al., 2012) and plant growth (Gliński and Lipiec, 1990; Passioura, 1991). We incorporated a hand-weeding treatment to separate effects of the herbicide from the presence/absence of weeds and used a broad-spectrum soil insecticide to suppress soil arthropod populations. Our hypotheses were that glyphosate application would (1) reduce soil nutrient availability, (2) negatively impact densities of soil arthropods, and (3) increase soybean growth and yield via weed suppression. Furthermore, we hypothesized that using a soil insecticide to reduce soil arthropod densities would have a negative impact on soybean growth and yield.

Li, 1992; O’Hara, 2001). Most oribatid mite (Acari: Oribatida) taxa feed on microbes and decaying plant material (Norton, 1990; Coleman et al., 2004). Ecologically, oribatid mites regulate organic matter decomposition and nutrient cycling, and influence soil structure (BehanPelletier, 1999; Coleman et al., 2004). To maximize agricultural production, we need to understand more about how soil arthropods affect crops, and how human activities, such as pest management practices, alter belowground processes. Weeds are common pests within agricultural systems. Although weeds compete with crops for nutrients, water and sunlight, they represent a source of vegetative biodiversity (Altieri, 1999). Soil arthropods can be influenced by plants and vice versa (De Deyn et al., 2004; Bennett, 2010), and arthropod densities are often greater in weedy than weed-free environments (Altieri et al., 1985; Wardle, 1995). Weeds modify environmental conditions within the canopy and near the soil surface, including regulating temperature, increasing humidity, and decreasing wind (Norris and Kogan, 2005). Weeds often have large root profiles (Davis et al., 1967) with active rhizospheres within which soil arthropods are associated (Curry and Ganley, 1977; House, 1989; Garrett et al., 2001), provide food resources in the form of seeds (Brust and House, 1988; Bohan et al., 2011; Kulkarni et al., 2015), and contribute residue that impacts the detritus food web (Curry, 1973; Wardle, 1995; Wardle et al., 1999). Consequently, removal of weeds may cause either direct or indirect trophic effects on soil arthropods. Herbicides are the primary weed management tool used within most agricultural systems. Glyphosate-resistant soybeans are used extensively in the United States (Cerdeira and Duke, 2006; Bonny, 2008), with 85 percent of total soybean production in 2013 relying on a glyphosate-based weed management system (NASS, 2014). The success of this technology has led to increased reliance on the herbicide glyphosate for weed management within soybean fields (Bonny, 2008). Glyphosate is a glycine derivative, and is a non-selective herbicide that kills plants by inhibiting the enzyme 5-enolpyruvyl-shikimate-3-phosphate synthase (EPSP) within the shikimate pathway, thus reducing biosynthesis of aromatic amino acids within plants (Franz et al., 1997; Duke and Powles, 2008). Glyphosate is a widely used agrochemical for many reasons, including its low ecotoxicological risk (Baylis, 2000; Giesy et al., 2000; Duke and Powles, 2008). The general consensus is that glyphosate has little impact on non-target organisms, in part because it binds to soil, is rapidly degraded by microbes (Giesy et al., 2000; Haney et al., 2000; Borggaard and Gimsing, 2008; Duke and Powles, 2008), and only plants and microorganisms have a shikimate pathway (Herrmann and Weaver, 1999). However, potential ecological impacts of glyphosate, including leaching (Vereecken, 2005) and effects on soil organisms and crop plants (Helander et al., 2012) may be worth considering. Research has suggested glyphosate negatively affects densities of beneficial soil microorganisms (Zaller et al., 2014; Druille et al., 2016) and affects the composition of the rhizosphere microorganism community (Kremer and Means, 2009), although others did not find strong effects (Liphadzi et al., 2005; Weaver et al., 2007; Barriuso and Mellado, 2012; Lane et al., 2012a,b; Nakatani et al., 2014). Several studies have found glyphosate negatively impacts legume nodulation, nodule biomass, or Nfixation (Mallik and Tesfai, 1985; Reddy and Zablotowicz, 2003; Zablotowicz and Reddy, 2004; Bohm et al., 2009; Kremer and Means, 2009; Zobiole et al., 2012), reduced plant uptake or tissue concentrations of micronutrients (Eker et al., 2006; Neumann et al., 2006; Cakmak et al., 2009), lower plant growth or biomass (Bott et al., 2008; Zobiole et al., 2010a), and altered seed composition (Zobiole et al., 2010b), although see Duke et al. (2012a,b). Regarding soil microarthropods, effects of glyphosate are thought to primarily be indirect via reduced plant cover (Brust, 1990; Wardle, 1995; Bitzer et al., 2002; Cerdeira and Duke, 2006), but Evans et al. (2010) found this herbicide altered the mobility and long-term survival of epigeal predatory arthropods. We studied how weed management (herbicide, hand-weeding,

2. Methods 2.1. Experimental sites On-farm field experiments were conducted during the 2012 and 2013 growing seasons at two sites located at a distance approximately 40 km from each other, but with differing soil textures. The soil at the first site (‘Sandy’) was a sandy loam (Leonard, North Dakota) in the Glyndon soil series (coarse-silty, mixed superactive, frigid Aeric Calciaquoll; Table 1). The soil at the second site (‘Clayey’) was a silty clay loam (Mapleton, North Dakota) with a mixed Dovray (fine, smectitic, frigid Cumulic Vertic Epiaquolls) and Bearden soil series (fine-silty, mixed, superactive, frigid Aeric Calciaquoll; Table 1). 2.2. Experimental design The field experiment was laid out in a randomized complete block design with six replications. Each plot was 9.15 m by 9.15 m with 6.1 m alleyways between each plot. During both years on May 15, field plots were prepared for planting using a John Deere cultivator (wide field cultivator, spring tooth harrow, 2.3 m wide). Soybean variety Roughrider Genetics 607 Roundup Ready® (Monsanto Company, St. Louis, Missouri) was planted using a John Deere 71 flex planter with 76.2 cm between rows and twelve rows per plot at approximately 370,600 seeds per hectare resulting in a within row spacing of approximately 3.0 cm. In 2012 the sites were planted on May 22, whereas Table 1 Location, basic soil properties, and rainfall at the two experiment sites. Sandy (Leonard)

Clayey (Mapleton)

Sand (g kg ) Silt (g kg−1) Clay (g kg−1) Texture Soil pH EC (ds m-1)

46°39′58.3560″, −097°14′32.9640″ 640 240 120 Sandy loam 5.8 0.57

46°55′42.1680″, −097°01′03.1800″ 60 560 380 Silty clay loam 7.3 1.19

Rainfall (cm)

Sandy (Leonard)

Clayey (Prosper†)

May 2012 July 2012 May-August 2012 May 2013 July 2013 May-August 2013

4.27 11.15 24.05 7.75 2.77 27.10

4.62 1.63 15.27 10.52 2.01 36.8

Coordinates −1

† Rainfall data from weather stations in the North Dakota Agricultural Weather Network; Prosper was the closest to the Mapleton field site.

2

Applied Soil Ecology xxx (xxxx) xxx–xxx

A.A. Hansen et al.

15.2 cm band over the rows within the +insecticide plots at a rate of 1.13 g/m2 and raked into the soil to a depth of approximately 4 cm. The insecticide was applied pre-planting (May 21, 2012, May 22, 2013) and again mid-season (July 12, 2012 and July 17, 2013) to keep soil arthropod densities low throughout the growing season. We did not irrigate plots due to logistical difficulties, but relied on rainfall events, which we quantified using weather stations in the North Dakota Agricultural Weather Network (Table 1; https://ndawn.ndsu. nodak.edu). In May 2012, sites received significant rain the day after insecticide application (sandy = 26.9 mm, clayey = 20.8 mm), while in July rain occurred several days prior to and after insecticide application. In May 2013, sites received significant rain in the four days prior insecticide application (sandy = 46.2 mm, clayey = 50.8 mm) as well as later in the month, and in July 2013 rain occurred several days prior to and after insecticide application. There was 34.8–58.9 mm more rain in May 2013 than May 2012, and 83.8 mm less rain in July 2013 than July 2012 at the sandy site. Overall, seasonal rainfall (May-August) was slightly higher in 2013 at the sandy site and twice as high in 2013 at the clayey site.

in 2013 the sites were planted on May 24. Field location, location of individual experimental plots, and assignment of treatments to each plot remained the same for both years in order to elucidate potential long term treatment effects. The two main factors in this experiment, weed management and insecticide, were established using a factorial arrangement with three levels of the former and two levels of the latter nested within each site. The weed management treatments were: 1) Glyphosate: +glyphosate, no weeds, 2) Hand-weeded: no glyphosate, no weeds, and 3) Weedy: no glyphosate, +weeds. This arrangement was designed to separate effects of the herbicide glyphosate from effects related to the presence or absence of weeds. Comparing glyphosate plots with hand-weeded plots should show effects of the herbicide, whereas comparing hand-weeded plots with weedy plots should demonstrate effects of the presence of weeds. 2.3. Weed management treatments Glyphosate (Buccaneer Plus®; Tenkoz, Inc; Alpharetta, Georgia) was applied at the recommended rate (1.12 kg ai ha−1) twice each summer (June 15 and July 13, 2012; June 18 and July 16, 2013) using a tractor mounted boom sprayer delivering 187 L spray solution ha−1 at 275 kPa. Weeds were removed manually in hand-weeded plots, with care taken to minimize disturbances to the soil and any adjacent soybean plants. In 2012, weeds were pulled by hand within the rows and flatedged garden hoes were used to remove weeds between rows. In 2012, plots were weeded on June 12–15, 25–26 (sandy site only) and July 2–6 and in 2013, hand-weeded plots were weeded on June 13–14, June 18–19, July 9–11, and July 24. In 2013, in addition to previous methods, we used mini-cultivators (MC 43, Earthquake, Cumberland, Wisconsin) in order to improve the degree of weed reduction. Cultivator tines were set at the highest position (resulting in a tillage depth of approximately 4 cm) to reduce soil disturbance. Weeds from the naturally existing seed bank were allowed to grow in Weedy control plots. In 2012, the weed population was so robust that weeds had to be managed in order for the soybeans to survive. Therefore, weeds between the rows were cut once with a grass trimmer (FS 45 C, Stihl®, Waiblingen, Baden-Württemberg, Germany) to a height of approximately 10 cm during the last week of June. In 2013, weeds between the rows were cut periodically with a hedge trimmer (HS 45, Stihl®, Waiblingen, Baden-Württemberg, Germany). Weed management began on July 12 and was done as needed to prevent weeds from going to seed.

2.6. Sampling and quantification of subsurface soil arthropods A key question was to determine if the soil insecticide successfully reduced densities of soil arthropods so we could explore potential impacts of their presence on soil and plant parameters. Soil samples were taken periodically throughout the growing season, with eight samples taken in 2012 (May 17, June 7, June 18, June 27, July 16, July 27, Aug 10, Aug 24) and four in 2013 [May 16, June 22 (sandy) and July 1 (clayey), July 18, Aug 12]. In both years, the first sample was taken in mid-May, prior to any insecticide or herbicide applications or weeding. On each sample date, four soil samples (15 cm deep) were collected within each plot between soybean rows using a golf cup cutter (11 cm in diameter; Par Aide Products Co., Lino Lakes, Minnesota) and combined. In the weedy plots, the soil was shaken off the roots and plant matter was discarded. In 2012, soils were stored in a refrigerator (10 °C) for a maximum of 1 month prior to being processed (mean ± SE: 17.6 ± 4.9 days), while in 2013 they were processed the day of collection. Soil from each sample date was processed at the same time, and in 2012 any variability in storage time occurred among sampling dates. To extract arthropods from soil samples, soil aggregates were broken apart and large pieces of debris and plant material removed by hand, and then 3240 mL of soil from each plot placed into separate Berlese (Tullgren) funnels. Berlese funnels allow arthropods to move through the soil in response to light and heat and are considered one of the best methods for extraction of belowground arthropods (Dindal, 1990). Extraction time varied between 9 and 11 d (mean ± SE: 2012, 9.5 ± 0.2; 2013, 9.6 ± 0.5) depending on soil moisture level. Soil arthropod data is reported as arthropods per 1.0 L of dry soil. A dissecting microscope and relevant keys were used to identify arthropods (Borror et al., 1989; Dindal, 1990; Christiansen and Bellinger, 1998). We focused on the most abundant arthropod taxa relevant to plant production, i.e., four groups of focal arthropods: springtails (Entognatha: Collembola, all species and life stages combined), predatory beetles (Insecta, Coleoptera: Carabidae + Staphylinidae, all species combined but adults only), oribatid mites (Acari, Oribatida, all species and life stages combined), and mesostigmatid mites (Acari, Mesostigmata, all species and life stages combined).

2.4. Sampling and quantification of weeds Weed pressure was quantified mid-season in both 2012 and 2013 to assess the establishment and potential effectiveness of our weed management treatments. At each site, weeds were destructively sampled within three systematically-placed 0.25 m2 quadrats per plot on July 31. A tape measure was stretched diagonally across the plots and quadrats were placed between the rows of soybeans at 3.4 m, 6.7 m and 10.1 m. All weeds within a quadrat were identified (Iowa State University Extension, 2010), the density of each species quantified, and then weeds clipped at ground level and placed in paper bags by species. Weeds were dried to a constant weight at 71 °C and weighed immediately after being removed from the drier. 2.5. Insecticide treatment

2.7. Assessment of soil parameters The insecticide treatment had two levels: with or without a soilapplied insecticide (chlorpyrifos, Lorsban® 15G Dow AgroSciences; Indianapolis, Indiana). The soil insecticide was intended to reduce densities of soil arthropods, thus allowing us to explore their potential effects on soil and plant parameters by comparing data from plots with and without the insecticide. Chlorpyrifos was distributed by hand in a

Soil from samples taken to assess densities of soil arthropods were also used to quantify % soil moisture and soil nutrients. To quantify % soil moisture, an 8–10 g subsample was taken every time soil arthropods were sampled. Samples were placed in metal cans, weighed, dried to a constant weight, reweighed, and % soil moisture calculated. 3

Applied Soil Ecology xxx (xxxx) xxx–xxx

A.A. Hansen et al.

(G-G) correction was used if P < 0.05 (Greenhouse and Geisser, 1959). When time × treatment interactions were significant, a profile analysis was done to examine effects of independent variables on each sampling date. When weed management or insecticide treatments were found to be significant (P < 0.05), preplanned contrasts were used for mean separation among treatments at each site. With regard to the former, we were primarily interested in two contrasts. The first was between weedy control plots and hand-weeded plots, which was intended to demonstrate effects of the presence of weeds (i.e., weed effect) and the second was between hand-weeded plots and plots receiving the herbicide glyphosate, which was intended to show effects of the chemical (i.e., herbicide effect). Contrasts between weedy plots and glyphosate plots represent effects of weeds plus herbicides (i.e., combined effect). Pearson correlation analyses were used to examine linear relationships between seed quality parameters. Weed density and biomass data were log X + 1 transformed. Each year, after the initial sample date that occurred prior to treatments being implemented, soil arthropod data were summed across the season and so initial densities and cumulative densities (for each focal taxa, log X + 1 transformed) were used as dependent variables in separate analyses. Percent soil moisture data were averaged across the season and arcsin transformed for analysis. Root nodule weights, root weights, and shoot weights were log transformed. Percent seed protein, oil, and fiber were arcsin transformed prior to analysis. In 2012, aphid populations were only sampled once, and so aphids per plant was the dependent variable, whereas in 2013, cumulative aphid-days (Ruppel, 1983) were calculated from density data from two sampling dates. Aphid data were log (X + 1) transformed prior to analysis. Factorial ANOVA analyses revealed that root nodules were one of the primary factors affected by the insecticide treatment at the sandy site. To further explore this finding, we used least squares multiple linear regression to determine relationships between independent variables that we felt had the greatest potential to directly impact the dependent variable (i.e., total nodule density per plant in 2013 on the last sample date), namely: weed pressure (log transformed total weed density), soil nutrients in 2012 and 2013 (NO3−-N, P, K, Zn, Fe, Mn), and log X + 1 transformed cumulative densities of the two non-predatory focal soil arthropod groups (Collembola, oribatid mites) (2012, sampling dates 1–8 combined; 2013, sampling dates 1–4 combined). We eliminated highly correlated variables (correlation coefficient > 0.7), specifically 2013 Mn, which was correlated with 2013 Fe (0.88), and total weed density in 2012 (correlated with total weed density in 2013, 0.83). We then used a best subsets regression analysis and associated model criteria (i.e., adjusted R-square, corrected AIC, Schwartz’s BIC, Mallows’ Cp vs. P) to determine which subset of independent variables to use in an additive general linear model (Quinn and Keough, 2002).

To quantify soil nutrients, a 50–100 g subsample was taken from soil used to assess densities of soil arthropods on July 16, 2012 and July 18, 2013 (i.e., after the 2nd application of glyphosate and chlorpyrifos). Subsamples were kept frozen until submitted to the North Dakota State University soil testing lab, where the soil was dried, ground, and macronutrients (NO3−-N, P, K) and micronutrients (Zn, Fe, Mn) quantified. Analysis of NO3−-N was done using water extraction and transnitration of salicyclic acid, P was analyzed using the Olson procedure, K was analyzed by 1 N ammonium acetate, and all micronutrients were determined using DTPA (diethylene triamine pentaacetic acid) (NCR, 1988). 2.8. Assessment of soybean plant parameters In 2012, soybean plant samples were taken once (July 18, R1 growth stage: beginning flowering), while in 2013 they were taken three times [June 24 (sandy) and July 1 (clayey) (V1-V3: first to third trifoliate), July 18 (R2: full flowering), Aug 14 (R5: beginning seed) (Licht, 2014)]. On each sample date five consecutive plants were cut at the soil surface in each plot. The above-ground plant material (shoot) was placed into individual bags, which were dried to a constant weight at 71 °C and weighed immediately after being removed from the drier. Roots were extracted by removing a uniform volume of soil centered around the stem (approx. 30.5 L × 30.5 W × 30.5 D cm) using a shovel and hand trowel. Research indicates that in the field, most of the soybean root matter is found 0–15 cm from the soil surface (Mayaki et al., 1976; Hicks, 1978). Excess soil was gently removed by hand, and root balls were placed in individual plastic bags and kept in coolers (approximately 4 °C) until roots were recovered by rinsing the soil from roots over a metal sieve (710 μm opening, U.S.A. standard testing sieve, no. 25). Root nodules were counted and removed by hand within 1 week. After nodules were removed, they and the remaining root material were dried to a constant weight at 71 °C and weighed separately. Soybean yield was quantified by harvesting all plants along a 3.05 m transect within the center of the middle two rows of each plot in early October. Seeds were removed by hand, dried in an oven (71 °C) for one week, and weighed immediately after removal from driers. Seeds were then sent to the Northern Crops Institute (Fargo, North Dakota) where percent protein, oil, and fiber were determined. 2.9. Above-ground herbivores Densities of phloem-feeding soybean aphids (Hemiptera: Aphis glycines Matsumura) were quantified on July 18, 2012, July 18, 2013, and August 15, 2013 (R2: full flowering). On each sampling date, wholeplant aphid counts were taken on the same plants used to assess shoot biomass immediately prior to shoots being cut.

3. Results 2.10. Statistical analyses 3.1. Weed pressure For all analyses, data from each growing season were analyzed separately using JMP software (SAS Institute, 2013). Histograms and analysis of residuals by prediction plots were used to determine if data needed to be transformed prior to analysis. For all dependent variables, multiple samples from within a plot were averaged prior to analyses. Data from each dependent variable were analyzed separately. Data from 2012, from dependent variables only sampled once per year (e.g., yield), and data summed across the season (i.e., soil arthropod data, soil moisture) were analyzed using factorial ANOVA, with site, replicate, weed management, insecticide, and weed management × insecticide as the independent variables. The last four factors were nested within site, and replicate was considered a random variable. Plant growth data from 2013 were analyzed using repeated measures ANOVA (Gotelli and Ellison, 2004). When the assumption of sphericity was not violated, Wilks’ Lambda values were reported, whereas the Greenhouse-Geisser

Common lambsquarters [Chenopodium album (L.)], redroot pigweed [Amaranthus retroflexus (L.)], water hemp [(A. tuberculatus (Moq.)], and grasses, including volunteer corn [Poaceae, Zea mays (L.)] were the dominant weed species (Table 2). Other weeds that were present at low densities included: Canada thistle (Cirsium arvense L.), horseweed (Conyza canadensis (L.) Cronquist), Venice mallow (Hibiscus trionum L.), common purslane (Portulaca oleracea L.), Eastern black nightshade (Solanum ptychanthum Dunal), dandelion (Taraxacum officinale F. H. Wigg), and Siberian elm (Ulmus pumila L.). In 2012, the sandy site had greater mean ( ± SE) density per m2 (Site, df1,10, F = 16.0, P = 0.003) and increased biomass (g m2) (Site, df1,10, F = 13.5, P = 0.004) compared to the clayey site. Weed pressure differed between years at the clayey site, but at the sandy site it was similar. In 2013, weed density tended to be greater at the clayey site 4

Applied Soil Ecology xxx (xxxx) xxx–xxx

A.A. Hansen et al.

Table 2 Density and biomass of weeds. Data are means ( ± SE). All weed species Density

A. retroflexus

A. tuberculatus

C. album

Grasses

Biomass

Density

Biomass

Density

Biomass

Density

Biomass

Density

Biomass

2012 Sandy G 31.6 ± 8.1a HW 18.2 ± 3.1a W 312.2 ± 41.0b

15.9 ± 5.0a 43.8 ± 10.3b 311.1 ± 32.2c

1.0 ± 0.4a 0.5 ± 0.2a 13.8 ± 4.6b

0.7 ± 0.4a 6.5 ± 4.0a 42.3 ± 26.0b

0 0 0

0 0 0

27.0 ± 7.5a 14.7 ± 3.0a 282.7 ± 43.2b

14.8 ± 4.8a 31.9 ± 8.9a 256.3 ± 25.2b

2.6 ± 2.3a 1.4 ± 0.8a 7.8 ± 3.2a

0.2 ± 0.2a 4.1 ± 2.5a 5.0 ± 1.9a

Clayey G 11.6 ± 4.3a HW 6.1 ± 1.8a W 47.9 ± 19.2b

0.5 ± 0.2a 12.0 ± 6.3b 311.8 ± 76.3c

0.1 ± 0.1a 2.0 ± 0.9a 13.7 ± 6.7b

< 0.01a 11.1 ± 6.2a 238.1 ± 67.3b

0 0 0

0 0 0

0a 0a 0.6 ± 0.3a

0a 0a 48.4 ± 32.1b

0a 0.3 ± 0.3a 1.6 ± 1.1a

0a 0.2 ± 0.2a 16.0 ± 11.7a

2013 Sandy G 10.1 ± 3.8a HW 10.5 ± 2.5a W 452.4 ± 47.4b

0.5 ± 0.2a 1.5 ± 0.4a 151.6 ± 14.5b

0.6 ± 0.3a 2.6 ± 1.0a 114.3 ± 34.6b

0.03 ± 0.02a 0.6 ± 0.3a 48.0 ± 10.3b

0.1 ± 0.1a 0a 54.5 ± 31.4b

0.1 ± 0.1a 0a 16.0 ± 5.8b

3.9 ± 1.4a 4.7 ± 1.5a 227.4 ± 54.2b

0.2 ± 0.1a 0.6 ± 0.3a 28.0 ± 5.6b

0a 0.1 ± 0.1a 35.8 ± 7.8b

0a 0.01 ± 0.01a 50.0 ± 9.9b

Clayey G 74.6 ± 35.3a HW 11.1 ± 3.4b W 776.5 ± 108.9c

3.0 ± 1.0a 2.9 ± 1.9a 411.6 ± 47.5b

0a 5.8 ± 2.4b 358.6 ± 65.0c

0a 1.2 ± 0.5b 188.2 ± 31.8c

0.8 ± 0.5a 0.1 ± 0.1a 111.4 ± 24.8b

0.3 ± 0.2a 0.8 ± 0.8a 78.6 ± 14.8b

0.8 ± 0.6a 0.2 ± 0.2a 286.0 ± 83.7b

0.2 ± 0.2a < 0.01a 130.3 ± 36.3b

0a 0.6 ± 0.6a 10.2 ± 6.2b

0a < 0.01a 10.7 ± 7.0b

Density is number per m2, biomass is g m2; S = sandy site, C = clayey site; W = weedy control plots, HW = hand-weeded, G = glyphosate.

sandy = 2.8 ± 0.3, clayey = 3.6 ± 0.5; 2013: P = 0.0003, sandy = 1.7 ± 0.3, clayey = 9.7 ± 2.7). Otherwise, site did not impact densities of the other arthropod groups on the first sampling date (P > 0.05), with mesostigmatid mite densities 3.1 ± 0.6 in 2012 and 0.3 ± 0.1 in 2013, and densities of predatory beetles 0.1 ± 0.04 in 2012 and 0.1 ± 0.04 in 2013. In 2012, cumulative densities (sampling dates 2–8) of Collembola after treatments were applied were greater at the sandy site than the clayey site (Site, df1,10, F = 6.2, P = 0.032; sandy = 176.8 ± 33.8, clayey = 79.8 ± 9.9), while oribatid mites were more abundant at the clayey site compared to the sandy site (Site, df1,10, F = 71.0, P < 0.0001; sandy = 32.1 ± 3.9, clayey = 91.3 ± 6.3) and there was no difference in densities of mesostigmatid mites (Site, df1,10, F = 2.1, P = 0.179; sandy = 19.2 ± 3.8, clayey = 28.7 ± 4.9) or predatory beetles (Site, df1,10, F = 0.1, P = 0.819; sandy = 0.8 ± 0.2, clayey = 0.7 ± 0.1). Similar to the previous year, in 2013 cumulative densities of Collembola after treatments were applied (sampling dates 2–4) were greater at the sandy site than the clayey site (Site, df1,10, F = 63.2, P < 0.0001; sandy = 50.7 ± 5.3, clayey = 17.5 ± 2.9), while both mite groups were more abundant at the clayey site compared to the sandy site (oribatid: Site, df1,10, F = 30.4, P = 0.0003; sandy = 18.2 ± 3.5, clayey = 54.2 ± 7.9; mesostigmatid: Site, df1,10, F = 16.8, P = 0.002; sandy = 1.8 ± 0.4, clayey = 5.4 ± 0.8), and there was no difference in densities of predatory beetles (Site, df1,10, F = 1.8, P = 0.216; sandy = 0.3 ± 0.1, clayey = 0.5 ± 0.1). In 2012, cumulative densities of the various arthropod groups were not affected by weed management (both Insecticide × Weed and Weed, P > 0.05; Fig. 1). However, at both sites mean ( ± SE) densities of most arthropod groups were lower in plots treated with the insecticide, including Collembola (Insecticide, df2,50, F = 35.3, P < 0.0001; Fig. 2), predatory beetles (Insecticide, df2,50, F = 18.7, P < 0.0001), and mesostigmatid mites (Insecticide, df2,50, F = 15.9, P < 0.0001). Densities of oribatid mites were not impacted by the insecticide in 2012 (Insecticide, df2,50, F = 1.1, P = 0.351). In 2013, there were no interactive effects of weed management and the insecticide on cumulative arthropod densities (Insecticide × Weed, P > 0.05 for all groups). However, all four arthropod groups were affected by weed management (Weed, df4,50, P < 0.05), primarily at the clayey site, with positive effects being due to the presence of weeds (Fig. 1). One exception was that Collembola were lower in weedy plots

than the sandy site (Site, df1,10, F = 4.7, P = 0.055), whereas weed biomass was significantly greater at the clayey site than the sandy site (Site, df1,10, F = 31.8, P = 0.0002; Table 2). Overall, we observed that the sandy site had numerous smaller stature weeds both years of the study. At the clayey site, the first year there were fewer, heavier (i.e., larger) weeds that went to seed at the end of the season, which appeared to result in high densities of small weeds the following year (Table 2). Weed management had the intended impact on weed density in both 2012 (Weed, df4,50, F = 29.3, P < 0.0001) and 2013 (Weed, df4,50, F = 55.4, P < 0.0001), with more weeds in weedy control plots and similar weed pressure between hand-weeded and glyphosate plots (Table 2). One exception was at the clayey site in 2013, where glyphosate plots had more weeds than hand-weeded plots (P = 0.039) at the time weed populations were assessed. Results for weed biomass were similar to that of weed density (2012: Weed, df4,48.2, F = 46.6, P < 0.0001; 2013: Weed, df4,50, F = 213.2, P < 0.0001). However, in 2012 at both sites, weed biomass was significantly reduced in glyphosate plots compared to hand-weeded plots. Insecticide application did not significantly impact weed density or biomass in either year (2012 and 2013: Insecticide × Weed, P > 0.070; Insecticide, P > 0.140). Results were similar when looking at data from individual weed species (Table 2), although there was a weak effect of insecticide on the density and biomass of A. retroflexus in 2013 (density: Insecticide × Weed, P = 0.040; Insecticide, P = 0.399; biomass: Insecticide × Weed, P = 0.040; Insecticide, P = 0.032), which was driven by slightly higher weed density and biomass in weedy plots with insecticide at the clayey site, while the opposite trend occurred in hand-weeded plots and weedy plots at the sandy site. 3.2. Soil arthropods In 2012 and 2013, arthropod densities recovered from soil cores were similar among treatments on the first sample date (d1) prior to the application of any chemicals, with P > 0.05 for all interactive and main effects of Insecticide and Weed for the four arthropod groups. On the first sampling date mean densities ( ± SE) of Collembola per 1.0 L dry soil were significantly higher at the sandy site (2012: P = 0.001, sandy = 66.2 ± 10.1, clayey = 23.9 ± 2.8; 2013: P < 0.0001, sandy = 30.1 ± 4.4, clayey = 3.7 ± 0.8), and oribatid densities were higher at the clayey site, but only in 2013 (2012: P = 0.373, 5

Applied Soil Ecology xxx (xxxx) xxx–xxx

A.A. Hansen et al.

Fig. 1. Density of the four focal arthropod groups by weed management treatment. Data is cumulative arthropod densities (per 1.0 L of dry soil) through the season after the first sampling date (i.e., 2012 data is date 2 through date 8, 2013 data is date 2 through date 4). Data are means ( ± SE).

Fig. 2. Density of four focal arthropod groups by insecticide treatment. Data is cumulative arthropod densities (per 1.0 L of dry soil) through the season after the first sampling date (i.e., 2012 data is date 2 through date 8, 2013 data is date 2 through date 4). Data are means ( ± SE). 6

Applied Soil Ecology xxx (xxxx) xxx–xxx

A.A. Hansen et al.

Table 3 Effect of treatments (P-values) on macronutrients during 2012–2013 growing seasons. NO3-N

Site W INS W × INS

P

K

2012

2013

2012

2013

2012

2013

0.025 NS NS NS

NS 0.002 NS NS

0.011 NS 0.004 NS

0.0008 0.003 NS NS

0.039 < 0.0001 NS NS

0.0003 0.024 NS NS

W = weed management treatment, INS = insecticide treatment, NS = nonsignificant.

compared to hand-weeded plots at the sandy site (P = 0.048). Predatory beetles were the only arthropod group negatively impacted by the insecticide treatment in 2013 (Insecticide, df2,50, F = 13.4, P < 0.0001).

Fig. 3. Insecticide effect on soil available phosphorus. Data are means ( ± SE).

reduced in insecticide treated plots compared to non-treated plots (Fig. 3), but in 2013, the insecticide did not affect macronutrients. The sandy site had greater concentrations (mg kg−1, mean ± SE) of soil micronutrients in both 2012 (for all Site, df1,10, P < 0.0001; Zn: sandy = 2.9 ± 0.1, clayey = 1.1 ± 0.04; Fe: sandy = 91.5 ± 4.0, clayey = 18.5 ± 1.8; Mn: sandy = 42.6 ± 2.9, clayey = 12.3 ± 1.8) and 2013 (for all Site, df1,10, P < 0.0001; Zn: sandy = 3.9 ± 0.1, clayey = 1.3 ± 0.05; Fe: sandy = 102.3 ± 4.9, clayey = 22.1 ± 1.9; Mn: sandy = 43.0 ± 2.4, clayey = 18.2 ± 2.0). There were no effects of any of the treatments on soil micronutrients in either year (all interactive and main effects of Insecticide and Weed, P > 0.05, data not shown).

3.3. Soil characteristics Soil moisture levels (%, averaged across sampling dates, mean ± SE) were consistently greater at the clayey site than the sandy site (mean ± SE: 2012, Site, df1,10, P < 0.0001, sandy = 12.6 ± 0.2, clayey = 22.1 ± 0.3; 2013, Site, df1,10, P < 0.0001, sandy = 16.6 ± 0.4, clayey = 24.0 ± 0.5). Weed management and insecticide treatments did not impact the average level of soil moisture in either year at either the sandy or clayey site (all interactive and main effects of Insecticide and Weed, P > 0.05, data not shown). For soil macronutrients, in 2012, the clayey site had lower levels of soil available N and higher levels of P and K, whereas in 2013, N was similar among sites and P and K were greater at the clayey site than the sandy site (Tables 3 and 4). There were no interactive treatment effects on availability of NO3−-N, P, or K. In 2012, weed management only impacted the availability of soil K at the sandy site, with increased levels in weedy plots compared to other plots. Impacts of weed management on soil macronutrients were more apparent the following year, when the presence of weeds generally had a negative impact on soil NO3−-N and P, although K concentration was greatest in weedy plots, again at the sandy site only. In 2012 at both sites, soil available P was

3.4. Soybean plant growth parameters In both 2012 and 2013, there were no interactive effects of insecticide application and weed management treatment on any soybean plant parameter assessed (Time × Insecticide × Weed, P > 0.05; Insecticide × Weed, P > 0.05). 3.4.1. Soybean root nodules The number of root nodules per g dry weight of root was similar among treatments in 2012 (Table 5, Fig. 4a and b; mean ± SE: sandy = 20.2 ± 1.3, clayey = 17.4 ± 1.2). In 2013, nodule density

Table 4 Levels of soil available macronutrients by weed management treatment. Data are means ( ± SE). Letters indicate significance at P > 0.05.

Table 5 Effect of treatments (P-values) on plant growth parameters during the 2012 and 2013 growing season from a repeated measures analysis.

Macronutrients NO3-N (kg/ha)

P (mg/kg)

K (mg/kg)

2012 Sandy G HW W

45.9 47.4 47.8 42.6

± ± ± ±

2.4 3.4 a 3.9 a 5.2 a

26.3 24.9 26.5 27.6

± ± ± ±

1.4 2.6 a 2.7 a 2.2 a

256.1 217.7 211.4 339.1

± ± ± ±

13.1 11.7 a 12.0 a 17.9b

Clayey G HW W

39.0 45.2 40.1 31.7

± ± ± ±

2.3 3.7 a 4.0 a 3.4 a

32.3 34.8 32.6 29.4

± ± ± ±

1.3 1.5 a 2.6 a 2.2 a

272.2 279.6 280.4 256.7

± ± ± ±

6.4 9.1 a 14.6 a 7.9 a

2013 Sandy G HW W

31.1 34.4 31.7 27.4

± ± ± ±

1.7 3.1 a 3.6 a 1.9 a

34.3 34.0 38.4 30.5

± ± ± ±

1.5 2.7 ab 2.9 a 1.6b

288.8 258.8 286.3 321.3

± ± ± ±

9.7 14.2 a 18.6 ab 13.2b

Clayey G HW W

27.2 31.7 33.5 16.4

± ± ± ±

1.5 2.6 a 5.4 a 0.7b

41.9 46.8 44.0 35.0

± ± ± ±

1.4 1.5 a 2.7 a 1.9b

347.5 354.6 345.4 342.5

± ± ± ±

5.2 9.2 a 6.0 a 11.5 a

# Root nodules per g root

Weight per nodule

Root dry wt

Shoot dry wt

NS NS NS NS

NS NS 0.040 NS

0.019 < 0.0001 NS NS

0.016 < 0.0001 NS NS

2013 Time Time × Site Time × W Time × INS Time × W × INS

< 0.0001 < 0.0001 NS NS NS

< 0.0001 < 0.0001 NS NS NS

< 0.0001 < 0.0001 0.0001 0.034 NS

< 0.0001 < 0.0001 < 0.0001 0.010 NS

Site W INS W × INS

< 0.0001† < 0.0001 0.033 NS

0.001† 0.046 NS NS

NS† < 0.0001† NS† NS

NS† < 0.0001† NS† NS

2012 Site W INS W × INS

W = weed management treatment, INS = insecticide treatment, NS = nonsignificant, P > 0.05. † Profile analysis done due to significant time × treatment interaction.

W = weedy control plots, HW = hand-weeded, G = glyphosate. 7

Applied Soil Ecology xxx (xxxx) xxx–xxx

A.A. Hansen et al.

Fig. 4. Weed management (a) and soil insecticide (b) effects on number of root nodules per g dry weight root. Data are means ( ± SE).

paralleled that of roots for both years. In 2012, the dry weight of soybean roots and shoots (g) was lower at the sandy site (Table 5, Fig. 6a and b; mean ± SE: roots, sandy = 0.93 ± 0.06, clayey = 1.20 ± 0.06; shoots, sandy = 5.08 ± 0.51, clayey = 6.63 ± 0.35). In 2013, the effect of site was not consistent, with roots and shoots lighter at the sandy site at the start of the season (Site, df1,10, P < 0.0001 for both parameters) and no difference at the end of the season (Site, df1,10, F = 0.8, P > 0.135 for both parameters). In 2012 at the sandy site, the presence of weeds had a strong negative impact on soybean root and shoot weight (Weed, df4,50, P < 0.0001 for both), with differences between glyphosate and handweeded treatments likely due to limitations of physically managing weeds resulting in greater weed pressure in the latter plots. In 2012, weed management did not impact root and shoot weight at the clayey site, which had fewer, larger weeds (all contrasts, P > 0.13). In 2013, the negative effects of weed presence on root and shoot weight intensified over the season, leading to a Time × Weed interaction. In 2012, the insecticide treatment did not impact root or shoot weight. In 2013, effects of the insecticide on root and shoot weight increased through time, and was more pronounced at the end of the season (Insecticide, df2,50, roots, P = 0.023; shoots, P = 0.004) with heavier roots and shoots in +insecticide plots at the sandy site (roots, P = 0.034; shoots, P = 0.007) and similar roots and lighter shoots in +insecticide plots at the clayey site (roots, P = 0.073; shoots, P = 0.048).

per g dry weight of root was higher at the sandy site at the start and end of the season (Site, df1,10: D1, F = 20.5, P = 0.001; D2, F = 0.3, P = 0.614; D3, F = 19.9, P = 0.001). Consistently through time, nodules per g dry weight of root were increased in weedy plots compared to other plots. At the sandy site there were consistently more root nodules per g root in +insecticide plots (P = 0.011), whereas at the clayey site values were similar between + and no insecticide plots (P = 0.540). Mean weight of individual root nodules (mg) was similar between sites in 2012 (Table 5; mean ± SE: sandy = 0.59 ± 0.15, clayey = 0.43 ± 0.06). In 2013, effects of site on weight of individual root nodules varied across time, with heavier nodules at the clayey site at the start of the season (Site, df1,10: P = 0.010, sandy = 0.11 ± 0.01, clayey = 0.18 ± 0.02) and the converse on the last sampling date (P = 0.003, sandy = 0.18 ± 0.02, clayey = 0.12 ± 0.01). Weed management did not affect nodule weight in 2012, although in 2013, nodules in weedy plots at the sandy site were consistently heavier than those from glyphosate plots (P = 0.006), although other contrasts at both sites were not significant (Fig. 5a). In 2012, nodules were marginally heavier in no insecticide plots at the clayey site (P = 0.054), but there were no differences in nodule weight between insecticide plots in 2013 (Fig. 5b). 3.4.2. Soybean root and shoot dry weight Effects of site and treatments on dry weight (g) of soybean shoots 8

Applied Soil Ecology xxx (xxxx) xxx–xxx

A.A. Hansen et al.

Fig. 5. Weed management (a) and soil insecticide (b) effects on dry weight of individual root nodules. Data are means ( ± SE).

plant ± SE: 0.22 ± 0.08), and there were no significant treatment effects (all main effects and interactions, P > 0.234). In 2013, cumulative soybean aphid-days were greater at the sandy site compared to the clayey site (Site, df1,10, F = 7.1, P = 0.024; mean ± SE: sandy = 423.6 ± 56.0, clayey = 257.5 ± 57.6). Interactive effects of weed management and insecticide on aphid-days were absent (Insecticide × Weed, df4,50, F = 1.2, P = 0.303). However, aphid-days were negatively affected by the presence of weeds at both sites (Fig. 8a and b; Weed, df4,50, F = 8.8, P < 0.0001). Aphid-days were lower in plots receiving the insecticide, but only at the clayey site (Fig. 8; Insecticide, df2,50, F = 3.9, P = 0.027).

3.4.3. Soybean yield and seed composition In 2012, yield data (g per m of row) were only available for the clayey site (mean ± SE: 146.2 ± 7.7), and in 2013 the amount of yield was similar between sites (Site, df1,10, F = 4.4, P = 0.063; sandy = 169.4 ± 13.6, clayey = 142.8 ± 13.8). The presence of weeds had a negative impact on soybean yield (2012: Weed, df2,25, F = 7.7, P = 0.003; 2013: Weed, df4,50, F = 45.4, P < 0.0001; Fig. 7), but there was no difference in yield between +insecticide and no insecticide plots (2012: Insecticide, df1,25, F = 0.5, P = 0.472; 2013: Insecticide, df2,50, F = 2.0, P = 0.144). Seed protein (% dry basis) was strongly negatively correlated with seed oil (% dry basis) (Pearson correlation coefficient, 2012: R2 = −0.857, 2013: R2 = −0.829) and positively correlated with amino acids (2012: R2 = 0.889, 2013: R2 = 0.949). Seed protein at the clayey site (mean ± SE: 43.0 ± 0.2) was similar among treatments in 2012 (Weed, df2,25, F = 1.0, P = 0.394; Insecticide, df1,25, F = 0.3, P = 0.859). In 2013, seed protein was lower at the sandy site (sandy = 41.6 ± 0.3, clayey = 42.6 ± 0.3), and the only significant treatment effect was the higher protein in seeds from weedy plots at the clayey site (Weed, df4,48.6, F = 3.1, P = 0.024; glyphosate (41.9 ± 0.3) vs weedy (44.1 ± 0.6) P = 0.008, hand-weeded (42.1 ± 0.3) vs weedy P = 0.009; Insecticide, df2,47.7, F = 0.4, P = 0.693).

3.6. Relationship between nodule density and soil and arthropod parameters At the sandy site, density of two soil arthropod groups were the independent variables from the best subsets analysis that most parsimoniously explained variation in 2013 nodule density (total nodules per plant) on the last sampling date. The general linear model generated using 2012 Collembola densities and 2013 oribatid mite densities as the independent variables explained 22% of variation in 2013 nodule density (adjusted R2 = 0.220, n = 36, P = 0.006). Nodule density in 2013 was negatively related to density of Collembola in 2012 (β = −35.52, t = -3.24, P = 0.003) and positively related to density of oribatid mites in 2013 (β = 31.13, t = 2.12, P = 0.042). The rate of change of the conditional mean of nodule density with respect to Collembola was estimated to be between −57.81 and −13.22, whereas

3.5. Above-ground herbivores In 2012, densities of soybean aphids were extremely low (mean per 9

Applied Soil Ecology xxx (xxxx) xxx–xxx

A.A. Hansen et al.

Fig. 6. Weed management (a) and soil insecticide (b) effects on root dry weight root. Data are means ( ± SE).

the rate of change with respect to oribatid mites was between 1.21 and 1.05.

management practices and insecticide application influenced soil nutrient availability, densities of below- and above-ground arthropods, and the growth of glyphosate tolerant soybeans. In general, responses were similar between sites, although many treatment effects were most pronounced at the location with sandier-textured soil. Because weeds compete with plants for resources (Zimdahl, 2004) and alter the physical environment and provide shelter and food for arthropods (Norris

4. Discussion Understanding how farming practices indirectly affect crop growth can help inform management decisions. We found that weed

Fig. 7. Weed management effect on soybean yield in 2012 (clayey site only) and 2013. Data are means ( ± SE). 10

Applied Soil Ecology xxx (xxxx) xxx–xxx

A.A. Hansen et al.

Fig. 8. Weed management (a) and soil insecticide (b) effects on density of herbivorous soybean aphids. Data are means ( ± SE).

profile in soil never exposed to glyphosate, but not in soil to which glyphosate had been applied for 10+ years. Weeds compete with plants for resources and reduce crop growth and yield (Stoller et al., 1987; Zimdahl, 2004). Thus, our expectation was that soil nutrients and soybean growth and yield would be reduced in the presence of weeds. In general, weed presence had a detrimental effect on soil macronutrients and soybean yield, with effects of weeds being more pronounced the second year of the study. Lack of weed treatment effects at the clayey site in 2012 were likely due to nonuniform weed pressure, as during that site-year there were fewer, but larger, weeds compared to the sandy site. However, weed management did not impact soil micronutrients and soil K was greater in weedy compared to weeded or herbicide-treated plots at the sandy site. Transport of K from soil to roots is reduced under dry conditions (Kuchenbuch et al., 1986); however, soil moisture levels were similar among weed management treatments. K adsorbs to clay particles and soil K is positively related to the clay and organic matter content of the soil (Mitra and Prakash, 1957; Munson and Nelson, 1963; Sharpley, 1989); therefore, K often leaches from sandy soil (Öborn et al., 2005), although this is reduced by the presence of plants (Munson and Nelson, 1963). Increases in soil K in weedy plots at the sandy site may be related to decreased leaching caused by the presence of additional plant roots. Alternatively, deep rooted weeds may have increased the amount of K within the upper portion of the soil profile (Witter and Johansson, 2001). We found that soybeans in weedy plots had more nodules per g of root and/or heavier nodules, particularly during the last year of the study. Soybeans may have compensated for increased competition for N by increasing nodulation. For example, Wahua and Miller (1978) found soybean intercropped with dwarf sorghum had more nodules than soybean grown in monoculture. Effects of weed management on soil arthropods appeared to be related to the presence of weeds rather than the use of glyphosate, as is consistent with previous studies (Brust, 1990; Bitzer et al., 2002; Cerdeira and Duke, 2006). Positive effects of weed presence on soil arthropods were inconsistent across site-years, although consistent across arthropod taxa. Our experimental treatments did not affect soil moisture, so weeds may have provided arthropods with shelter or food (Norris and Kogan, 2005), including via plant residue mixed into the

and Kogan, 2005), we expected that their presence might mitigate negative impacts of soil insecticides, or moderate effects of soil arthropods on plants (Boerner and Harris, 1991; Sabais et al., 2012). However, there were virtually no interactive effects of weed management and soil insecticide application on any response variable assessed in this study. Within each year soil arthropod data were averaged among sampling dates, however, it is important to note that sample storage time was considerably longer during 2012 versus 2013, and refrigeration time can have a negative impact on arthropod recovery using active collection techniques, particularly for soft-bodied arthropods (Lakly and Crossley, 2000). Notably, effects of weed management appeared to be related to the presence of weeds rather than use of glyphosate for all response variables, because glyphosate and hand-weeded plots responded similarly. Intermediate values in hand-weeded plots, particularly with regard to plant growth, are likely due to logistical difficulties of maintaining weed-free plots over the entire growing season. Minor soil disturbance associated with use of the mini-cultivator and potential input of soil nutrients (e.g., nitrogen, carbon) caused by cutting or trimming weeds could have impacted outcomes related to soil arthropods in handweeded plots (House and Parmelee, 1985). Others have suggested glyphosate negatively affects crop plants, including reduced: (1) legume nodulation or function (Mallik and Tesfai, 1985; Reddy and Zablotowicz, 2003; Zablotowicz and Reddy, 2004; Bohm et al., 2009; Kremer and Means, 2009; Zobiole et al., 2012), (2) plant uptake or tissue concentrations of micronutrients (Eker et al., 2006; Neumann et al., 2006; Cakmak et al., 2009), (3) plant growth or biomass (Bott et al., 2008; Zobiole et al., 2010a), and (4) altered seed composition (Zobiole et al., 2010b). However, in their review of the literature Duke et al. (2012b) concluded that most studies indicate glyphosate has a minimal impact on plant mineral nutrition and crop performance. Lack of plant responses in our study may be related to the fact the experiment was established on land historically exposed to glyphosate. Biological organisms may adapt to successive and repeated glyphosate applications such that additional applications cause no pronounced responses. For example, Dick et al. (2010) suggested the soil microbial community can adapt to long-term glyphosate exposure, as they found the herbicide altered the microbial fatty acid 11

Applied Soil Ecology xxx (xxxx) xxx–xxx

A.A. Hansen et al.

similar between+ and no insecticide plots, and at levels that would result in 90–95% maximum soybean yield (Snyder, 2004). However, we still saw effects of the soil insecticide treatment on soybean growth parameters during 2013. These observations could indicate Collembola indirectly affected plant growth via another mechanism, such as by altering densities or the community composition of other microorganisms beside AMF (Moore et al., 1988; Lussenhop, 1996). Mean Collembola populations were over 2.5 times greater at the start of year one versus year two. Density-dependent effects of Collembola on plant growth can be nonlinear, with intermediate densities being more beneficial than higher densities (Harris and Boerner, 1990). Additionally, effects of Collembola densities in year one may have had consequences for microorganisms and plant physiology in year two. This conclusion is supported by results from the best subsets regression analysis, which indicated that soybean nodule density in 2013 was negatively related to Collembola density the previous year. Use of a soil insecticide impacted densities of soybean aphids, an above-ground phloem-feeding herbivore, at the clayey site. Several studies have shown that below-ground microorganisms and arthropods can impact foliar-feeding herbivores, including aphids (Scheu, 2001; Schütz et al., 2008; Brunner et al., 2014). However, a more likely explanation is that the chemical directly reduced aphid densities, as chlorpyrifos is listed for soybean aphid control (Gomez, 2009). Aphid densities were lower in weedy plots, which could be related to reduced host-plant finding, alterations in plant quality, or effects on aphid natural enemies (Norris and Kogan, 2005). In conclusion, we found that the presence of weeds influenced the assessed parameters rather than the use of the herbicide glyphosate. Using a soil insecticide altered the soil arthropod community, decreased soil P, increased soybean productivity, and decreased densities of an above-ground herbivore. Although more direct effects of the insecticide on soil microbes and plants cannot be ruled out, our research results suggest soil arthropods significantly alter plant productivity in annual soybean production systems, similar to conclusions from perennial reduced-disturbance systems such as grasslands and forests. Understanding more about subterranean interactions can contribute to sustainable ecological intensification of agricultural systems (Bender et al., 2016).

soil (Curry, 1973; Wardle et al., 1999). The soil insecticide primarily suppressed soil arthropods the first year of the study, although there was no effect on densities of oribatid mites either year. Collembola were a dominant arthropod group in our study, and chlorpyrifos negatively effects the survival and reproduction of many Collembola species (Pereira et al., 2005; Fountain et al., 2007; Jegede et al., 2017). A field study using chlorpyrifos to suppress soil arthropods also found notable decreases in Collembola, predatory beetles, and predatory gamasid mites, but no effect on densities of oribatid mites (Eisenhauer et al., 2010). Other studies have also found chlorpyrifos reduced Collembola densities (Frampton, 1999; Endlweber et al., 2006; Frampton and van den Brink, 2007), and Wiles and Frampton (1996) found chlorpyrifos residues were more toxic on sandy soil versus sandy clay loam. Inconsistencies in effectiveness of the soil insecticide between years at our field site may have been due to variation in abiotic conditions such as rainfall (Solomon et al., 2014). Contrary to our initial expectations, applying a soil insecticide tended to have a positive effect on soybean growth parameters, which was most pronounced the second year of the study at the sandy site, although trends were similar in the first year, and so additional sampling at the end of the first season may have revealed significant treatment effects. The soil insecticide is unlikely to have directly enhanced soybean plant growth; chlorpyrifos has been shown to have neutral, or even inhibitory effects on plant growth at higher concentrations (Zhang et al., 2011; Akbar and Sultan, 2016). Chlorpyrifos did not interfere with the Bradyrhizobium japonicumsoybean symbiosis in greenhouse and field experiments (Revellin et al., 1992) and had insignificant impacts on parameters associated with soil microbial activity in another study (Singh et al., 2002). Others found chlorpyrifos temporarily reduced soil microbial functional diversity (Fang et al., 2009), decreased the total population of soil bacteria and increased the total population of fungi (Pandey and Singh, 2004), had taxa-specific effects, with decreased populations of asymbiotic aerobic nitrogen fixers, nitrifiers, and denitrifiers (Kumar et al., 2017). Akbar and Sultan (2016) found chlorpyrifos negatively affected plant growth promotion traits of specific bacterial strains that degrade the chemical. In general, it seems unlikely that direct effects of chlorpyrifos on the microbial community would result in the increased plant growth seen in our study. The soil insecticide possibly had a positive impact on soybeans by reducing densities of herbivorous soil arthropods. However, densities of specialized root herbivores (e.g., wireworms, data not shown) were low, and a similar study found chlorpyrifos did not affect densities of herbivorous fly larvae and gastropods (Eisenhauer et al., 2010). Saprophytic fungi and decaying plant material have been considered the preferred diet of most Collembola species (Hopkin, 1997; Rusek, 1998), but some taxa consume root tissue (Endlweber et al., 2009; Ngosong et al., 2009). Stevens and Jones (2006) found that chlorpyrifos use (Lorsban 15G) did not directly affect plant root growth (i.e., in the absence of soil arthropods), but the number of fine roots increased in insecticide-treated field plots when arthropods densities were suppressed. Collembola also consume arbuscular mycorrhizal fungi (Warnock et al., 1982; Jonas et al., 2007; Caravaca and Ruess, 2014), but see Potapov and Tiunov (2016), and so another hypothesis is that the enhanced plant growth seen when the soil insecticide was applied was related to altered interactions between Collembola and soybean root microorganisms. Direct or indirect interference with arbuscular mycorrhizal fungi symbiotically associated with plant roots can disrupt P transport from the soil to plants, thus negatively impacting several aspects of plant health that are dependent on adequate P, including nodulation (Israel, 1987; O’Hara, 2001) and plant biomass (Warnock et al., 1982). Although we did not assess root fungi, plant-available soil P was significantly reduced in +insecticide plots in year one, which is suggestive of increased transport into plants. The following year when the insecticide did not reduce Collembola densities, soil P levels were

Acknowledgements Thank you to the anonymous reviewers, Sadie Fiegel, Mary Gebhardt, Sarah Hewitt, Jade Monroe, Ethan Rudell, Gene Schmidt, Caitlin Stegmiller, and Stephanie Swenson for helping establish field experiments and collecting and processing samples, to Jason Harmon for input on analysis of a nested design, and to the North Dakota Soybean Council and North Dakota State Board of Agricultural Research & Education for support and funding. References A’Bear, A.D., Johnson, S.N., Jones, T.H., 2014. Putting the ‘upstairs–downstairs’ into ecosystem service: What can aboveground–belowground ecology tell us? Biol. Control 75, 97–107. Akbar, S., Sultan, S., 2016. Soil bacteria showing a potential of chlorpyrifos degradation and plant growth enhancement. Braz. J. Microbiol. 47, 563–570. Altieri, M.A., 1999. The ecological role of biodiversity in agroecosystems. Agric. Ecosyst. Environ. 74, 19–31. Altieri, M.A., Wilson, R.C., Schmidt, L.L., 1985. The effects of living mulches and weed cover on the dynamics of foliage- and soil-arthropod communities in three crop systems. Crop Prot. 4, 201–213. Barriuso, J., Mellado, R.P., 2012. Glyphosate affects the rhizobacterial communities in glyphosate-tolerant cotton. Appl. Soil Ecol. 55, 20–26. Baylis, A.D., 2000. Why glyphosate is a global herbicide: strengths, weaknesses and prospects. Pest Manage. Sci. 56, 299–308. Behan-Pelletier, V.M., 1999. Oribatid mite biodiversity in agroecosystems: role for bioindication. Agric. Ecosyst. Environ. 74, 411–423. Bender, S.F., Wagg, C., van der Heijden, M.G.A., 2016. An underground revolution: biodiversity and soil ecological engineering for agricultural sustainability. Trends

12

Applied Soil Ecology xxx (xxxx) xxx–xxx

A.A. Hansen et al.

Endlweber, K., Ruess, L., Scheu, S., 2009. Collembola switch diet in presence of plant roots thereby functioning as herbivores. Soil Biol. Biochem. 41, 1151–1154. Evans, S.C., Shaw, E.M., Rypstra, A.L., 2010. Exposure to a glyphosate-based herbicide affects agrobiont predatory arthropod behavior and long-term survival. Ecotoxicology 19, 1249–1257. Fang, H., Yu, Y., Chu, X., Wang, X., Yang, X., Yu, J., 2009. Degradation of chlorpyrifos in laboratory soil and its impact on soil microbial functional diversity. J. Environ. Sci. 21, 380–386. Ferris, H., Tuomisto, H., 2015. Unearthing the role of biological diversity in soil health. Soil Biol. Biochem. 85, 101–109. Fountain, M.T., Brown, V.K., Gange, A.C., Symondson, W.O.C., Murray, P.J., 2007. The effects of the insecticide chlorpyrifos on spider and Collembola communities. Pedobiologia 51, 147–158. Frampton, G.K., 1999. Spatial variation in non-target effects of the insecticides chlorpyrifos, cypermethrin and pirimicarb on Collembola in winter wheat. Pestic. Sci. 5, 875–886. Frampton, G.K., van den Brink, P.J., 2007. Collembola and macroarthropod community responses to carbamate, organophosphate and synthetic pyrethroid insecticides: direct and indirect effects. Environ. Pollut. 147, 14–25. Franz, J.E., Mao, M.K., Sikorski, J.A., 1997. Glyphosate: A Unique Global Herbicide. American Chemical Society, Washington D.C. Garrett, C.J., Crossley, D.A., Coleman, D.C., Hendrix, P.F., Kisselle, K.W., Potter, R.L., 2001. Impact of the rhizosphere on soil microarthropods in agroecosystems on the Georgia piedmont. Appl. Soil Ecol. 16, 141–148. Giesy, J.P., Dobson, S., Solomon, K.R., 2000. Ecotoxicological risk assessment for Roundup® herbicide. Rev. Environ. Contam. Toxicol. 167, 35–120. Gliński, J., Lipiec, J., 1990. Soil Physical Conditions and Plant Growth. CRC Press Inc., Boca Raton, FL. Gomez, L.E., 2009. Use and Benefits of Chlorpyrifos in U.S. Agriculture. Dow AgroSciences LLD, Indianapolis, IN. http://www.chlorpyrifos.com/pdf/ chlorpyrifosusebenefits_18may2009.pdf (accessed 1 July 2016). Gotelli, N.J., Ellison, A.M., 2004. A Primer of Ecological Statistics. Sinauer Associates Inc, Sunderland, MA. Greenhouse, S.W., Geisser, S., 1959. On methods in the analysis of profile data. Psychometrika 24, 95–112. Haney, R.L., Senseman, S.A., Hons, F.M., Zuberer, D.A., 2000. Effect of glyphosate on soil microbial activity and biomass. Weed Sci. 48, 89–93. Harris, K.K., Boerner, R.E.J., 1990. Effects of belowground grazing by Collembola on growth, mycorrhizal infection, and P uptake of Geranium robertianum. Plant Soil 129, 203–210. Helander, M., Saloniemi, I., Saikkonen, K., 2012. Glyphosate in northern ecosystems. Trends Plant Sci. 17, 569–574. Herrmann, K.M., Weaver, L.M., 1999. The shikimate pathway. Annu. Rev. Plant Physiol. Plant Mol. Biol. 50, 473–503. Hicks, D.R., 1978. Growth and development. In: Norman, A.G. (Ed.), Soybean Physiology, Agronomy, and Utilization. Academic Press Inc., New York, pp. 17–44. Hopkin, S.P., 1997. Biology of the Springtails: (Insecta: Collembola). Oxford University Press, Don Mills, Canada. House, G.J., 1989. Soil arthropods from weed and crop roots of an agroecosystem in a wheat-soybean-corn rotation: impact of tillage and herbicides. Agric. Ecosyst. Environ. 25, 233–244. House, G.J., Parmelee, R.W., 1985. Comparison of soil arthropods and earthworms from conventional and no-tillage agroecosystems. Soil Tillage Res. 5, 351–360. Iowa State University Extension, 2010. Weed Identification Field Guide, CSI 0003. Iowa State University of Science and Technology, Ames, IA. Israel, D.W., 1987. Investigation of the role of phosphorus in symbiotic dinitrogen fixation. Plant Physiol. 84, 835–840. Jegede, O.O., Owojori, O.J., Römbke, J., 2017. Temperature influences the toxicity of deltamethrin, chlorpyrifos and dimethoate to the predatory mite Hypoaspis aculeifer (Acari) and the springtail Folsomia candida (Collembola). Ecotoxicol. Environ. Saf. 140, 214–221. Jonas, J.L., Wilson, G.W.T., White, P.M., Joern, A., 2007. Consumption of mycorrhizal and saprophytic fungi by Collembola in grassland soils. Soil Biol. Biochem. 39, 2594–2602. Keyser, H.H., Li, F., 1992. Potential for increasing biological nitrogen fixation in soybean. Plant Soil 141, 119–135. Kos, M., Jing, J., Keesmaat, I., Declerck, S.A.J., Wagenaar, R., Bezemer, T.M., 2017. Afterlife effects: living and dead invertebrates differentially affect plants and their associated above- and belowground multitrophic communities. Oikos 126, 888–899. Kremer, R.J., Means, N.E., 2009. Glyphosate and glyphosate-resistant crop interactions with rhizosphere microorganisms. Eur. J. Agron. 31, 153–161. Kuchenbuch, R., Claassen, N., Jungk, A., 1986. Potassium availability in relation to soil moisture. II. Calculations by means of a mathematical simulation model. Plant Soil 95, 233–243. Kulkarni, S.S., Dosdall, L.M., Willenborg, C.J., 2015. The role of ground beetles (Coleoptera: Carabidae) in weed seed consumption: a review. Weed Sci. 63, 335–376. Kumar, U., Berliner, J., Adak, T., Rath, P.C., Dey, A., Pokhare, S.S., Jambhulkar, N.N., Panneerselvam, P., Kumar, A., Mohapatra, S.D., 2017. Non-target effect of continuous application of chlorpyrifos on soil microbes, nematodes and its persistence under subhumid tropical rice-rice cropping system. Ecotoxicol. Environ. Saf. 135, 225–235. Lakly, M.B., Crossley, D.A., 2000. Tullgren extraction of soil mites (Acarina): effect of refrigeration time on extraction efficiency. Exp. Appl. Acarol. 24, 135–140. Lane, M., Lorenz, N., Saxena, J., Ramsier, C., Dick, R.P., 2012a. Microbial activity, community structure and potassium dynamics in rhizosphere soil of soybean plants treated with glyphosate. Pedobiologia 55, 153–159. Lane, M., Lorenz, N., Saxena, J., Ramsier, C., Dick, R.P., 2012b. The effect of glyphosate

Ecol. Evol. 31, 440–452. Bennett, A., 2010. The role of soil community biodiversity in insect biodiversity. Insect Conserv. Divers. 3, 157–171. Birkhofer, K., Schöning, I., Alt, F., Herold, N., Klarner, B., Maraun, M., Marhan, S., Oelmann, Y., Wubet, T., Yurkov, A., Begerow, D., Berner, D., Buscot, F., Daniel, R., Diekötter, T., Ehnes, R.B., Erdmann, G., Fischer, C., Foesel, B., Groh, J., Gutknecht, J., Kandeler, E., Lang, C., Lohaus, G., Meyer, A., Nacke, H., Näther, A., Overmann, J., Polle, A., Pollierer, M.M., Scheu, S., Schloter, M., Schulze, E.-D., Schulze, W., Weinert, J., Weisser, W.W., Wolters, V., Schrumpf, M., 2012. General relationships between abiotic soil properties and soil biota across spatial scales and different landuse types. PLoS One 7, e43292. http://dx.doi.org/10.1371/journal.pone.0043292. Bitzer, R.J., Buckelew, L.D., Pedigo, L.P., 2002. Effects of transgenic herbicide-resistant soybean varieties and systems on surface-active springtails (Entognatha: Collembola). Environ. Entomol. 31, 449–461. Boerner, R.E.J., Harris, K.K., 1991. Effects of Collembola (Arthropoda) and relative germination date on competition between mycorrhizal Panicum virgatum (Poaceae) and non-mycorrhizal Brassica nigra (Brassicaceae). Plant Soil 136, 121–129. Bohan, D.A., Boursault, A., Brooks, D.R., Petit, S., 2011. National-scale regulation of the weed seedbank by carabid predators. J. Appl. Ecol. 48, 888–898. Bohm, G.M.B., Alves, B.J.R., Urquiaga, S., Boddey, R.M., Xavier, G.R., Hax, F., Rombaldi, C.V., 2009. Glyphosate- and imazethapyr-induced effects on yield, nodule mass and biological nitrogen fixation in field-grown glyphosate-resistant soybean. Soil Biol. Biochem. 41, 420–422. Bonny, S., 2008. Genetically modified glyphosate-tolerant soybean in the USA: adoption factors, impacts and prospects. A review. Agron. Sustain. Dev. 28, 21–32. Borggaard, O.K., Gimsing, A.L., 2008. Fate of glyphosate in soil and the possibility of leaching to ground and surface waters: a review. Pest Manag. Sci. 64, 441–456. Borror, D.J., Johnson, N.F., Triplehorn, C.A., 1989. An Introduction to the Study of Insects, sixth ed. Saunders College Publishers, Philadelphia, PA. Bott, S., Tesfamariam, T., Candan, H., Cakmak, I., Römheld, V., Neumann, G., 2008. Glyphosate-induced impairment of plant growth and micronutrient status in glyphosate-resistant soybean (Glycine max L.). Plant Soil 312, 185–194. Brunner, S.M., Goos, R.J., Swenson, S.J., Foster, S.P., Schatz, B.G., Lawley, Y.E., Prischmann-Voldseth, D.A., 2014. Impact of nitrogen fixing and plant growth-promoting bacteria on a phloem-feeding soybean herbivore. Appl. Soil Ecol. 86, 71–81. Brussaard, L., Pulleman, M.M., Ouédraogo, É., Mando, A., Six, J., 2007. Soil fauna and soil function in the fabric of the food web. Pedobiologia 50, 447–462. Brust, G.E., 1990. Direct and indirect effects of four herbicides on the activity of carabid beetles (Coleoptera: Carabidae). Pest Manage. Sci. 30, 309–320. Brust, G.E., House, G.J., 1988. Weed seed destruction by arthropods and rodents in lowinput soybean agroecosystems. Am. J. Altern. Agric. 3, 19–25. Cakmak, I., Yazici, A., Tutus, Y., Ozturk, L., 2009. Glyphosate reduced seed and leaf concentrations of calcium, manganese, magnesium, and iron in non-glyphosate resistant soybean. Eur. J. Agron. 31, 114–119. Caravaca, F., Ruess, L., 2014. Arbuscular mycorrhizal fungi and their associated microbial community modulated by Collembola grazers in host plant free substrate. Soil Biol. Biochem. 69, 25–33. Cerdeira, A.L., Duke, S.O., 2006. The current status and environmental impacts of glyphosate-resistant crops. J. Environ. Qual. 35, 1633–1658. Christiansen, K.A., Bellinger, P.F., 1998. The Collembola of North America, North of the Rio Grande: A Taxonomic Analysis. Grinnell College, Grinnell, IA. Coleman, D.C., Crossley, D.A., Hendrix, P.F., 2004. Fundamentals of Soil Ecology, second ed. Elsevier Academic Press, Burlington, MA. Curry, J.P., 1973. The arthropods associated with the decomposition of some common grass and weed species in the soil. Soil Biol. Biochem. 5, 645–657. Curry, J.P., Ganley, J., 1977. The arthropods associated with the roots of some common grass and weed species of pasture. Ecol. Bull. 25, 330–339. Davis, R.G., Johnson, W.C., Wood, F.O., 1967. Weed root profiles. Agron. J. 59, 555–556. De Deyn, G.B., Raaijmakers, C.E., Van Ruijven, J., Berendse, F., Van Der Putten, W.H., 2004. Plant species identity and diversity effects on different trophic levels of nematodes in the soil food web. Oikos 106, 576–586. Dick, R., Lorenz, N., Wojno, M., Lane, M., 2010. Microbial dynamics in soils under longterm glyphosate tolerant cropping systems. In: 19th World Congress of Soil Science, 1–6 August 2010, Brisbane, Australia, pp. 153–156. Dindal, D.L., 1990. Soil Biology Guide. John Wiley & Sons, New York. Druille, M., García-Parisi, P.A., Golluscio, R.A., Cavagnaro, F.P., Omacini, M., 2016. Repeated annual glyphosate applications may impair beneficial soil microorganisms in temperate grassland. Agric. Ecosyst. Environ. 230, 184–190. Duke, S.O., Powles, S.B., 2008. Glyphosate: a once-in-a-century herbicide. Pest Manag. Sci. 64, 319–325. Duke, S.O., Lydon, J., Koskinen, W.C., Moorman, T.B., Chaney, R.L., Hammerschmidt, R., 2012a. Glyphosate effects on plant mineral nutrition, crop rhizosphere microbiota, and plant disease in glyphosate-resistant crops. J. Agric. Food Chem. 60, 10375–10397. Duke, S.O., Reddy, K.N., Bu, K., Cizdziel, J.V., 2012b. Effects of glyphosate on the mineral content of glyphosate-resistant soybeans (Glycine max). J. Agric. Food Chem. 60, 6764–6771. Eisenhauer, N., Sabais, A.C.W., Schonert, F., Scheu, S., 2010. Soil arthropods beneficially rather than detrimentally impact plant performance in experimental grassland systems of different diversity. Soil Biol. Biochem. 42, 1418–1424. Eker, S., Ozturk, L., Yazici, A., Erenoglu, B., Romheld, V., Cakmak, I., 2006. Foliar-applied glyphosate substantially reduced uptake and transport of iron and manganese in sunflower (Helianthus annuus L.) plants. J. Agric. Food Chem. 54, 10019–10025. Endlweber, K., Schädler, M., Scheu, S., 2006. Effects of foliar and soil insecticide applications on the collembolan community of an early set-aside arable field. Appl. Soil Ecol. 31, 136–146.

13

Applied Soil Ecology xxx (xxxx) xxx–xxx

A.A. Hansen et al.

Entomol. 76, 375–377. Rusek, J., 1998. Biodiversity of Collembola and their functional role in the ecosystem. Biodivers. Conserv. 7, 1207–1219. Sabais, A.C.W., Eisenhauer, N., König, S., Renker, C., Buscot, F., Scheu, S., 2012. Soil organisms shape the competition between grassland plant species. Oecologia 170, 1021–1032. SAS Institute, 2013. JMP®, Version 11.0.0. SAS Institute Inc., Cary, NC. Scheu, S., 2001. Plants and generalist predators as links between the below-ground and above-ground system. Basic Appl. Ecol. 2, 3–13. Scheu, S., Theenhaus, A., Jones, T.H., 1999. Links between the detritivore and the herbivore system: effects of earthworms and Collembola on plant growth and aphid development. Oecologia 119, 541–551. Schütz, K., Bonkowski, M., Scheu, S., 2008. Effects of Collembola and fertilizers on plant performance (Triticum aestivum) and aphid reproduction (Rhopalosiphum padi). Basic Appl. Ecol. 9, 182–188. Seastedt, T.R., 1984. The role of microarthropods in decomposition and mineralization processes. Annu. Rev. Entomol. 29, 25–46. Sharpley, A.N., 1989. Relationship between soil potassium forms and mineralogy. Soil Sci. Soc. Am. J. 53, 1023–1028. Singh, B.K., Walker, A., Wright, D.J., 2002. Degradation of chlorpyrifos, fenamiphos, and chlorothalonil alone and in combination and their effects on soil microbial activity. Environ. Toxicol. Chem. 21, 2600–2605. Snyder, C.S., 2004. Got Enough P and K For Your Soybeans? News and Views. Potash & Phosphate Institute, Norcross, GA. Solomon, K.R., Williams, W.M., Mackay, D., Purdy, J., Giddings, J.M., Giesy, J.P., 2014. Properties and uses of chlorpyrifos in the United States. In: Giesy, J.P., Solomon, K.R. (Eds.), Ecological Risk Assessment for Chlorpyrifos in Terrestrial and Aquatic Systems in the United States. Springer, New York, pp. 13–34. Stevens, G.N., Jones, R.H., 2006. Patterns in soil fertility and root herbivory interact to influence fine-root dynamics. Ecology 87, 616–624. Stoller, E.W., Harrison, S.K., Wax, L.M., Regnier, E.E., Nafziger, E.D., 1987. Weed interference in soybeans (Glycine max). Rev. Weed Sci. 3, 155–181. Vereecken, H., 2005. Mobility and leaching of glyphosate: a review. Pest Manage. Sci. 61, 1139–1151. Villani, M.G., Wright, R.J., 1990. Environmental influences on soil macroarthropod behavior in agricultural systems. Annu. Rev. Entomol. 35, 249–269. Wahua, T.A.T., Miller, D.A., 1978. Effects of intercropping on soybean N2-fixation and plant composition on associated sorghum and soybeans. Agron. J. 70, 292–295. Wardle, D.A., 1995. Impacts of disturbance on detritus food webs in agro-ecosystems of contrasting tillage and weed management practices. Adv. Ecol. Res. 26, 105–185. Wardle, D.A., Nicholson, K.S., Bonner, K.I., Yeates, G.W., 1999. Effects of agricultural intensification on soil-associated arthropod population dynamics, community structure, diversity and temporal variability over a seven-year period. Soil Biol. Biochem. 31, 1691–1706. Warnock, A.J., Fitter, A.H., Usher, M.B., 1982. The influence of a springtail Folsomia candida (Insecta, Collembola) on the mycorrhizal association of leek Allium porrum and the vesicular-arbuscular mycorrhizal endophyte Glomus fasciculatus. New Phytol. 90, 285–292. Weaver, M.A., Krutz, L.J., Zablotowicz, R.M., Reddy, K.N., 2007. Effects of glyphosate on soil microbial communities and its mineralization in a Mississippi soil. Pest Manage. Sci. 63, 388–393. Wiles, J.A., Frampton, G.K., 1996. A field bioassay approach to assess the toxicity to insecticide residues on soil to Collembola. Pestic. Sci. 47, 273–285. Witter, E., Johansson, G., 2001. Potassium uptake from the subsoil by green manure crops. Biol. Agric. Hortic. 19, 127–141. Zablotowicz, R.M., Reddy, K.N., 2004. Impact of glyphosate on the symbiosis with glyphosate-resistant transgenic soybean. J. Environ. Qual. 33, 825–831. Zaller, J.G., Heigl, F., Ruess, L., Grabmaier, A., 2014. Glyphosate herbicide affects belowground interactions between earthworms and symbiotic mycorrhizal fungi in a model ecosystem. Sci. Rep. 4 (5634), 1–8. Zhang, Z.-Y., Shan, W.-L., Song, W.-C., Gong, Y., Liu, X.-J., 2011. Phytotoxicity and uptake of chlorpyrifos in cabbage. Environ. Chem. Lett. 9, 547–552. Zimdahl, R.L., 2004. Weed-Crop Competition: A Review, second ed. Blackwell Publishing, Ames, IA. Zobiole, L.H.S., de Oliveira Jr., R.S., Huber, D.M., Constantin, J., Castro, C., de Oliveira, F.A., de Oliveira Jr., A., 2010a. Glyphosate reduces shoot concentrations of mineral nutrients in glyphosate-resistant soybeans. Plant Soil 328, 57–69. Zobiole, L.H.S., Oliveira Jr., R.S., Visentainer, J.V., Kremer, R.J., Bellaloui, N., Yamada, T., 2010b. Glyphosate affects seed composition in glyphosate-resistant soybean. J. Agric. Food Chem. 58, 4517–4522. Zobiole, L.H.S., Kremer, R.J., de Oliveira, R.S., et al., 2012. Glyphosate effects on photosynthesis, nutrient accumulation, and nodulation in glyphosate-resistant soybean. J. Plant Nutr. Soil Sci. 175, 319–330.

on soil microbial activity, microbial community structure, and soil potassium. Pedobiologia 55, 335–342. Lauber, C.L., Strickland, M.S., Bradford, M.A., Fierer, N., 2008. The influence of soil properties on the structure of bacterial and fungal communities across land-use types. Soil Biol. Biochem. 40, 2407–2415. Licht, M., 2014. Soybean Growth and Development, PM 1945. Iowa State University Extension and Outreach, Ames, IA. Liphadzi, K.B., Al-Khatib, K., Bensch, C.N., Stahlman, P.W., Dille, J.A., Todd, T., Rice, C.W., Horak, M.J., Head, G., 2005. Soil microbial and nematode communities as affected by glyphosate and tillage practices in a glyphosate-resistant cropping system. Weed Sci. 53, 536–545. Lussenhop, J., 1996. Collembola as mediators of microbial symbiont effect upon soybean. Soil Biol. Biochem. 28, 363–369. Mallik, M.A.B., Tesfai, K., 1985. Pesticidal effect on soybean-rhizobia symbiosis. Plant Soil 85, 33–41. Mayaki, W.C., Teare, I.D., Stone, L.R., 1976. Top and root growth of irrigated and nonirrigated soybeans. Crop Sci. 16, 92–94. Megías, A.G., Müller, C., 2010. Root herbivores and detritivores shape above-ground multitrophic assemblage through plant-mediated effects. J. Anim. Ecol. 79, 923–931. Mitra, S.P., Prakash, D., 1957. Adsorption of potassium as influenced by concentration and pH of the solution. Clay Miner. Bull. 3, 151–153. Moore, J.C., Walter, D.E., Hunt, H.W., 1988. Arthropod regulation of micro-and mesobiota in below-ground detrital food webs. Annu. Rev. Entomol. 33, 419–435. Munson, R.D., Nelson, W.L., 1963. Movement of applied potassium in soils. J. Agric. Food Chem. 3, 193–201. Nakatani, A.S., Fernandes, M.F., de Souza, R.A., da Silva, A.P., dos Reis-Junior, F.B., Mendes, I.C., Hungria, M., 2014. Effects of the glyphosate-resistance gene and of herbicides applied to the soybean crop on soil microbial biomass and enzymes. Field Crops Res. 162, 20–29. (NASS) National Agriculture Statistics Service, 2014. Agricultural Resource Management Survey: U.S. Soybean Industry. NASS, Washington, D.C. (NCR) North Central Region, 1988. Recommended Chemical Soil Tests for North Central Region, North Central Regional Publication, No. 221 (Revised). Missouri Agricultural Experiment Station SB1001. http://extension.missouri.edu/explorepdf/specialb/ sb1001.pdf (accessed 15 March 2018). Neher, D.A., Barbercheck, M.E., 1998. Diversity and function of soil mesofauna. In: Qualset, C.O., Collins, W.W. (Eds.), Biodiversity in Agroecosystems. CRC Press, Boca Raton, FL, pp. 27–47. Neumann, G., Kohls, S., Landsberg, E., Stock-Oliveira Souza, K., Yamada, T., Römheld, V., 2006. Relevance of glyphosate transfer to non-target plants via the rhizosphere. J. Plant Dis. Prot. 20, 963–969. Ngosong, C., Raupp, J., Scheu, S., Ruess, L., 2009. Low importance for a fungal based food web in arable soils under mineral and organic fertilization indicated by Collembola grazers. Soil Biol. Biochem. 41, 2308–2317. Norris, R.F., Kogan, M., 2005. Ecology of interactions between weeds and arthropods. Annu. Rev. Entomol. 50, 479–503. Norton, R.A., 1990. Acarina: Oribatida. In: Dindal, D.L. (Ed.), Soil Biology Guide. John Wiley & Sons Inc., New York, pp. 779–804. Öborn, I., Andrist-Rangel, Y., Askekaard, M., Grant, C.A., Watson, C.A., Edwards, A.C., 2005. Critical aspects of potassium management in agricultural systems. Soil Use Manage. 21, 102–112. O’Hara, G.W., 2001. Nutritional constraints on root nodule bacteria affecting symbiotic nitrogen fixation: a review. Aust. J. Exp. Agric. 41, 417–433. Pandey, S., Singh, D.K., 2004. Total bacterial and fungal population after chlorpyrifos and quinalphos treatments in groundnut (Arachis hypogaea L.) soil. Chemosphere 55, 197–205. Passioura, J.B., 1991. Soil structure and plant growth. Soil Res. 29, 717–728. Pereira, J., da Silva, A., Picanço, M., Barros, E., Jakelaitis, A., 2005. Effects of herbicide and insecticide interaction on soil entomofauna under maize crop. J. Environ. Sci. Health, Part B 40, 45–54. Petersen, H., Luxton, M., 1982. A comparative analysis of soil fauna populations and their role in decomposition processes. Oikos 39, 288–388. Potapov, A.M., Tiunov, A.V., 2016. Stable isotope composition of mycophagous collembolans versus mycotrophic plants: Do soil invertebrates feed on mycorrhizal fungi? Soil Biol. Biochem. 93, 115–118. Quinn, G.P., Keough, M.J., 2002. Experimental Design and Data Analysis for Biologists. Cambridge University Press, New York. Reddy, K.N., Zablotowicz, R.M., 2003. Glyphosate-resistant soybean response to various salts of glyphosate and glyphosate accumulation in soybean nodules. Weed Sci. 51, 496–502. Revellin, C., De Canson, B., Catroux, G., 1992. Effect of a mixture of chlorpyrifos and lindane on the symbiosis of Bradyrhizobium japonicum and soybean (Glycine max (L.) Merrill). Pestic. Sci. 36, 69–74. Ruppel, R.F., 1983. Cumulative insect-days as an index of crop protection. J. Econ.

14