Hostname: page-component-cd9895bd7-hc48f Total loading time: 0 Render date: 2024-12-28T03:24:11.643Z Has data issue: false hasContentIssue false

Measuring Interference from Midseason Tall Morningglory (Ipomoea purpurea) to Develop a Model for Teaching Weed Seedbank Effects on Chile Pepper

Published online by Cambridge University Press:  10 March 2017

Brian J. Schutte*
Affiliation:
Assistant Professor, Department of Entomology, Plant Pathology, and Weed Science, New Mexico State University, P.O. Box 30003 MSC 3BE, Las Cruces, NM 88003
*
*Corresponding author’s E-mail: [email protected]
Rights & Permissions [Opens in a new window]

Abstract

Tactics that target seedbanks are important components of weed management systems; however, such tactics can be difficult to adopt because consequences of seedbank reduction are often unclear. This study developed model-based software to provide insights on the economic outcomes, in the context of chile pepper production, of additions to tall morningglory seedbanks. Data for the model were derived from this and previous studies. In this study, field experiments were conducted to determine chile pepper yield and harvest efficiency responses to mid-season tall morningglory infestations. The experimental treatments were factorial combinations of herbicide (pendimethalin-treated, nontreated) and tall morningglory density (0, 4, 8, 12, 16, 20 plants 10-m row–1). Treatments were installed 9.5 weeks after crop seeding. Data collected included fresh weight of marketable chile peppers and time required for one individual to harvest 10-m of crop row, which was used to calculate the amount of chile pepper harvested in 1 min (harvest efficiency). Results indicated that crop yield was not influenced by tall morningglory density, pendimethalin treatment and interactions between tall morningglory density and pendimethalin. Harvest efficiency was influenced by tall morningglory density but was not influenced by herbicide treatment or interactions between herbicide treatment and tall morningglory density. Each additional tall morningglory plant decreased the amount of chile pepper harvested in 1 min by 9.7 g. The results of this and previous studies were used to develop model-based software that presents tall morningglory seedbank density effects on: (1) tall morningglory seedling densities after pendimethalin, (2) time requirements for hand-hoeing after pendimethalin, and (3) time requirements for hand-harvesting to acquire yield goals. The model-based software is intended to be used in the instruction of weed seedbank management strategies. By presenting seedbank density effects on weed control outcomes and crop production expenses, the model-based software might promote adoption of seedbank reduction strategies.

Tácticas que se enfocan en los bancos de semillas son importantes componentes de los sistemas de manejo de malezas. Sin embargo, estas tácticas pueden ser difíciles de adoptar porque las consecuencias de la reducciones en el banco de semillas son comúnmente poco claras. Este estudio desarrolló un programa computacional basado en modelos para brindar una visión profunda sobre los resultados económicos de adiciones al banco de semillas de Ipomoea purpurea, en el contexto de la producción de chile picante. Los datos para el modelo fueron derivados a partir de el presente estudio y estudios previos. En este estudio, se realizaron experimentos de campo para determinar las respuestas en el rendimiento de chile picante y la eficiencia en la cosecha a infestaciones de I. purpurea en medio de la temporada de crecimiento. Los tratamientos experimentales fueron combinaciones factoriales de herbicida (tratado con pendimethalin, sin tratamiento) y densidad de I. purpurea (0, 4, 8, 12, 16, 20 plantas 10-m hilera−1). Los tratamientos fueron instalados 9.5 semanas después de la siembra del cultivo. Los datos colectados incluyeron peso fresco de chile comercializable y el tiempo requerido para que un individuo cosechara 10-m de hilera de cultivo, lo que fue usado para calcular la cantidad de chile cosechado por minuto (eficiencia de cosecha). Los resultados indicaron que el rendimiento del cultivo no fue influenciado por la densidad de I. purpurea, el tratamiento con pendimethalin ni las interacciones entre la densidad de I. purpurea y pendimethalin. La eficiencia de cosecha fue influenciada por la densidad de I. purpurea, pero no fue influenciada por los tratamientos con el herbicida o las interacciones entre la densidad de I. purpurea y pendimethalin. Cada planta adicional de I. purpurea disminuyó la cantidad de chile cosechado por min en 9.7 g. Los resultados de este y estudios previos fueron usados para desarrollar un programa computacional basado en modelos que presenta los efectos de la densidad del banco de semillas de I. purpurea sobre: (1) la densidad de plántulas de I. purpurea después de pendimethalin, (2) el tiempo requerido de control con azadón manual después de pendimethalin, y (3) el tiempo requerido de deshierba manual para alcanzar objetivos específicos de rendimiento. El programa basado en modelos tiene la intención de ser usado para capacitaciones en estrategias de manejo de banco de semillas de malezas. Al presentar los efectos de la densidad del banco de semillas sobre los resultados del control de malezas y los costos de producción del cultivo, este programa computacional basado en modelos podría promover la adopción de estrategias para la reducción del banco de semillas.

Type
Education/Extension
Copyright
© Weed Science Society of America, 2017 

Chile peppers, which are cultivars of Capsicum species that can be hot or not depending on fruit final use (Bosland et al. Reference Bosland, Bailey and Iglesias-Olivas1996), are important components of New Mexico’s agricultural economy. In 2015, chile peppers were planted on 3,360 ha in New Mexico and provided over $46 million in cash receipts (USDA NASS 2016). The chile pepper hectarage in New Mexico is approximately 42% of the total hectarage in the United States for this crop, and the chile peppers harvested in New Mexico are approximately 33% of those harvested in the United States (USDA NASS 2016).

Chile pepper production in New Mexico is challenged by tall morningglory, a broadleaf weed with a summer annual life cycle. The success of tall morningglory in chile pepper is caused by factors including physically dormant seeds that can remain viable in soil for up to 17 yr (Baskin and Baskin Reference Baskin and Baskin2014; Burnside et al. Reference Burnside, Wilson, Weisberg and Hubbard1996), tolerance to herbicides that can be applied after chile pepper emergence (Baucom and Mauricio Reference Baucom and Mauricio2010; Schutte and Cunningham Reference Schutte and Cunningham2015), a twining habit that is considered an adaptation to shading (Gianoli Reference Gianoli2003), and prolonged periods of emergence that extend through the middle and later stages of the crop-growing season (J Schroeder, personal communication). This weed is especially unwanted because tall morningglory plants significantly increase time requirements for hand-hoeing (Schutte and Cunningham Reference Schutte and Cunningham2015) and can be asymptomatic hosts for chile pepper pests including Verticillium dahliae (causal agent of Verticillium wilt; Sanogo et al. Reference Sanogo, Etarock and Clary2009) and Meloidogyne incognita (root-knot nematode; Sanogo et al. Reference Sanogo, Thomas, Schroeder and Clary2008).

For chile pepper weeds other than tall morningglory, midseason infestations are known to reduce crop yield and disrupt crop harvest. For example, a previous study determined that spurred anoda [Anoda cristata (L.) Schlecht.] infestations that emerged 9 to 10 wk after chile pepper seeding caused crop yield reductions as great as 49% and increased the amount of time required for hand-harvesting, compared to that required in weed-free conditions (Schroeder Reference Schroeder1993). Crop yield losses as great as 76% have been attributed to multi-species infestations (Palmer amaranth [Amaranthus palmeri S. Wats.], common lambsquarters [Chenopodium album L.], oakleaf datura [Datura quercifolia Kunth], and Wright groundcherry [Physalis acutifolia (Miers) Sandw.]) that emerged 8 to 9 wk after chile pepper seeding (Schroeder Reference Schroeder1992). Tall morningglory plants that emerge midseason are likely to adversely affect chile pepper production, but this hypothesis has yet to be tested.

To control and suppress midseason weed infestations, chile pepper growers in New Mexico can apply soil-residual herbicides, including halosulfuron-methyl, S-metolachlor, trifluralin, and pendimethalin. Among these herbicides, pendimethalin is popular because this herbicide provides broad-spectrum control (halosulfuron-methyl does not control grasses), does not require growers to release the manufacturer of liability and indemnification for crop damage (such is required by S-metolachlor), and does not require mechanical incorporation (such is required by trifluralin). Although widely used in chile pepper production, pendimethalin does not sufficiently control tall morningglory (Grey and Wehtje Reference Grey and Wehtje2005; Schutte and Cunningham Reference Schutte and Cunningham2015; Wilcut et al. Reference Wilcut, Jordan, Vencill and Richburg1997). For tall morningglory plants that survive pendimethalin, the effects of this herbicide on their level of interference in chile pepper are poorly understood. For many weed species, plants that escape soil-residual herbicides are less competitive than nontreated plants (Adcock and Banks Reference Adcock and Banks1991; Liphadzi and Dille Reference Liphadzi and Dille2006).

Tall morningglory management strategies benefit from tactics that target soil seedbanks, because such tactics suppress population growth rates of annual weeds (Davis Reference Davis2006) and potentially improve outcomes of herbicidal (Dieleman et al. Reference Dieleman, Mortensen and Martin1999; Schutte and Cunningham Reference Schutte and Cunningham2015) and mechanical (Davis and Williams Reference Davis and Williams2007; Dieleman et al. Reference Dieleman, Mortensen and Martin1999) control interventions. However, seedbank-targeting tactics can be difficult for growers to adopt, because near-term economic consequences of seedbank reduction are often unclear (Swanton et al. Reference Swanton, Mahoney, Chandler and Gulden2008). To address possible valuation uncertainty for weed seedbank management and help focus attention on management of soil seedbanks, the overall goal of this project was to develop model-based software for demonstrating relationships between tall morningglory seedbank density and chile pepper production expenses.

To accomplish this goal, a field study was first conducted to determine chile pepper yield and harvest efficiency responses to midseason tall morningglory infestations treated and not treated with pendimethalin. The field study was framed by the following hypotheses: 1) increasing density in midseason tall morningglory infestations reduces chile pepper yield and harvest efficiency, and 2) the effects of midseason tall morningglory on chile pepper yield and harvest efficiency are conditioned by pendimethalin applied just prior to tall morningglory emergence. The results of the field study were then combined with results from previous studies that determined seedbank density effects on pendimethalin control outcomes (Schutte and Cunningham Reference Schutte and Cunningham2015) and seedling density effects on time requirements for hand-hoeing (Schutte Reference Schutte2015). The aggregate data were used to develop model-based software that presents tall morningglory seedbank density effects on tall morningglory seedling densities after pendimethalin, time requirements for hand-hoeing after pendimethalin, and time requirements for hand-harvesting to acquire yield goals. In accordance with Wilkerson et al. (Reference Wilkerson, Wiles and Bennett2002), who proposed that bioeconomic models are more valuable as educational tools rather than predictive instruments, the model developed in this study is intended to elevate concern for weed seedbank density among chile pepper growers in New Mexico.

Materials and Methods

Field Study Procedures

A field study was conducted at the New Mexico State University, Leyendecker Plant Science Research Center (32.19°N, 106.74°W) on a Belen clay loam soil (clayey over loamy, montmorillonitic [calcareous], thermic Vertic Terrifluvent) in 2014, and was repeated in 2015. Annual experimental runs occurred on different sections of the research center. In 2014, soil organic matter at the study site was 0.5%, and in 2015, soil organic matter was 0.9%. Each year, fields were subjected to a sequence of preparatory procedures that are customary for chile pepper production in the region. Field preparations included tilling, laser leveling, and listing and shaping raised beds into rows spaced 1 m apart. The width of a raised bed was 0.8 m.

The chile pepper cultivar ‘New Mexico 6-4’ was seeded into the center of raised beds to a depth of 2 cm at 6.7 kgha−1. Seeding was performed using a mechanical planter (MaxEmerge® Plus, John Deere, Moline, IL) and took place on May 2, 2014 and April 23, 2015. At the time of seeding, the field was uniformly treated with napropamide at 1.1 kg ai ha−1 to provide PRE control of small-seeded broadleaf and grass weeds. POST control of weeds that were not being studied was accomplished with cultivation, hand weeding, and herbicide (clethodim at 0.14 kg ai ha−1). Throughout the duration of the study, fields were irrigated as needed using furrow irrigation, which is flood irrigation in the furrows between raised beds.

Study treatments were factorial combinations of herbicide (pendimethalin-treated or nontreated) and tall morningglory population density (0, 4, 8, 12, 16, or 20 plants per 10 meters of row). Treatments were arranged in a randomized complete block design with four replications. Treatments were installed by seeding tall morningglory and applying herbicide immediately after crop thinning, which occurred at 9.5 weeks after crop seeding. Crop thinning adjusted chile pepper stands to clumps (2 to 3 plants per clump) spaced 0.18 m apart. Experimental units were three beds by 10 m and are hereafter referred to as “plots”. On the center bed of each plot, approximately 300 tall morningglory seeds were sown in a row that ran parallel to, and was 20 cm away from, the chile pepper row. The tall morningglory seeding depth was 2 cm. Prior to burial in the field, tall morninngglory seeds were hand-scarified with coarse sandpaper to remove physical seed dormancy. Pre-burial seed germination rates were determined with 14-d germination assays conducted in chambers set to 35/25 C day/night temperatures with 12 hour photoperiods. The pre-burial germination rate was 83% in 2014 and 91% in 2015. When tall morningglory seeds were sown, soil was dry (<8% volumetric moisture content) and not conducive to germination.

Immediately after tall morningglory seeding, pendimethalin (456 g ai L−1, Prowl® H2O, BASF Corp., Research Triangle Park, NC) was applied at 1.6 kg ai ha−1 using a CO2-powered backpack sprayer equipped with a boom with an even-flat nozzle (TeeJet® 8002EVS, TeeJet Technologies, Wheaton, IL). Herbicide bands (30-cm width) were directed to soil areas where tall morningglory seeds were buried. Herbicide was incorporated, and tall morningglory germination stimulated, with flood irrigation occurring 4 hr after herbicide application. At 12 d after spraying (DAS), tall morningglory plants were thinned to appropriate densities. Seedlings were removed at the first leaf stage such that tall morningglory plants were spaced evenly along the length of a plot.

At 78 to 90 DAS, marketable chile peppers were harvested by hand. Marketable chile peppers were straight, free of disease symptoms, and at least 10 cm in length. Data collected at harvest included tall morningglory population density, chile pepper fresh weight (fw), and time required for one individual to harvest 10 m of crop row, which is equal to 1×10−3 ha. These data were used to calculate the amount of chile pepper harvested in 1 min from 10 m of crop row, which is hereafter referred to as “harvest efficiency”.

Field Study Data Analysis

All statistical analyses were performed with the statistical software R version 3.3.0 (The R Foundation for Statistical Computing, http://www.r-project.org). Bartlett’s test for homogeneity of variance (Zar Reference Zar1999) indicated equal variances between years for crop yield and harvest efficiency, and therefore these data were pooled across years.

The R library lme4 was used to produce linear mixed-effects models (LMMs) for crop yield and harvest efficiency responses to tall morningglory density and pendimethalin treatment. Applications and principles of LMMs have previously been reviewed in the context of weed science (Luschei and Jackson Reference Luschei and Jackson2005). In this study, mixed-effects modeling was selected as the data analysis approach because the resulting models provided parameter estimates for influential factors that were independent of year (Luschei and Jackson Reference Luschei and Jackson2005). Such generalized understanding was desired for the bioeconomic model. Fixed effects in the LMMs were tall morningglory density, pendimethalin treatment, and interactions between tall morningglory density and pendimethalin treatment. Random effects in the LMMs were the hierarchical structures of sampling, which were year and replication within year.

LMMs were assessed for parsimony by comparing Akaike’s information criteria (AIC) between models with intercepts, fixed effects, and random effects (AICfull) and models with only intercepts and random effects (AICnull) (Nakagawa and Schielzeth Reference Nakagawa and Schielzeth2013). LMMs were assessed for goodness of fit by evaluating the proportions of total variance explained by fixed effects alone (marginal R 2) and the proportions of total variance explained by both fixed and random effects (conditional R 2) (Nakagawa and Schielzeth Reference Nakagawa and Schielzeth2013). Standard errors for fixed effect estimates were used to determine 95% confidence intervals (Pinheiro and Bates Reference Pinheiro and Bates2000). Confidence intervals (95%) that included zero were indicative of nonsignificant fixed effects.

Bioeconomic Model

Software that presents possible economic implications of tall morningglory seedbank additions was created by considering a system featuring state variables and user-defined variables (Fig. 1). Rates that regulate transitions between state variables were from the current and previous studies. For the transition from seedbank density to seedling density, a previous study determined that 4% of nondormant seeds buried 1 to 2 cm produced seedlings that escaped pendimethalin applied at 0.16 kgaiha−1, and that 26% of nondormant seeds produced seedlings that escaped pendimethalin applied at 0.8 kgaiha−1 (Schutte and Cunningham Reference Schutte and Cunningham2015). It is important to note that Schutte and Cunningham (Reference Schutte and Cunningham2015) determined seedbank density effects on pendimethalin control outcomes with tall morningglory seeds that were collected from one site-year and buried to depths that were within the optimum range for emergence. Thus, seedling establishment parameters used in this model do not reflect age and depth structures of natural seedbanks and are not recommended for use in predictive instruments involving tall morningglory population dynamics. For hoeing requirements, prior research showed that each additional tall morningglory seedling increased the time required to hand hoe 10 m of crop row by 3.6 s (Schutte Reference Schutte2015). Harvest data from the current study indicated that each additional tall morningglory plant increased the amount of time to harvest 1 kg of chile pepper from 10 m of crop row by 0.84 s. User-defined variables, as well as model outputs, utilize units of measurement that are customary for chile pepper growers in New Mexico. However, all variables and outputs are herein presented in SI units.

Figure 1 Underlying model for software that presents relationships between tall morningglory seedbank density and weed-control cost parameters in chile pepper production. The model initiates with the number of nondormant tall morningglory seeds buried at optimal depths for emergence within a 10-m2 area that is hereafter referred to as a “patch”. Users interact with the model by specifying the patches per hectare, pendimethalin application rate, and expected crop yield from a tall morningglory patch. User-defined variables, which are depicted with dashed-line boxes. Solid-line boxes present state variables. Rates that regulate transitions between state variables are provided in the text. Abbreviation: PHBPU, tall morningglory (Ipomoea purpurea).

The model produces per-hectare estimates for 1) tall morningglory seedling densities after pendimethalin treatment, 2) additional time required for hand-hoeing after pendimethalin treatment, and 3) additional time required for hand-harvesting to achieve yield goals if tall morningglory escapes from pendimethalin are not controlled. Model projections are presented across tall morningglory seedbank densities ranging from 0 to 1000 seeds per patch. Additional times required for hoeing and harvesting are multiplied by the New Mexico minimum wage ($7.50 hr−1) plus taxes (Social Security, 6.20%; Medicare, 1.45% [IRS 2016]; and workers’ compensation, 6.29% [NCCI 2016]) to provide information on potential financial consequences of additions to tall morningglory seedbanks.

The model is implemented in Microsoft® Excel and is available on the World Wide Web (http://aces.nmsu.edu/faculty/schutte/index.html). The Excel file also provides nontechnical background information on tall morningglory biology, seedbank management, and model construction. The bioeconomic model does not account for environmental variability that might influence growth of tall morningglory and chile pepper. Further, the model-based software does not provide technical guidance on specific weed management practices. Thus, the model-based software is not intended for use as a decision-support system.

Results and Discussion

Field Study

Censuses at harvest indicated that tall morningglory densities were identical to those established 12 DAS (data not shown). Overall mean crop yield was 23,910 (± SE 879) kg fw ha−1, which was comparable to previous reports of crop yield for chile pepper produced under similar conditions (20,982±SE 743 kg fw ha−1; Schroeder Reference Schroeder1993). Also, the overall mean crop yield for the current study was, to some degree, similar to the average chile pepper yield for New Mexico in 2014 (17,023 kg ha−1) and 2015 (19,600 kg ha−1; USDA NASS 2016). Crop yields from this study were expected to be greater than New Mexico averages, because state-level data included weights of fresh and dry chile peppers.

Statistical models for crop yield and harvest efficiency responses to tall morningglory density and pendimethalin were more parsimonious than models that did not include these independent variables (Table 1). Further, the proportions of variances explained by statistical models in this study were consistent with models used in previous ecological studies (Calcada et al. Reference Calcada, Lenoir, Plue, Broeckx, Closset-Kopp, Hermy and Decocq2015; Garfinkel and Johnson Reference Garfinkel and Johnson2015; Levesque et al. Reference Levesque, Walthert and Weber2016). Model results indicated that crop yield was not influenced by tall morningglory density, pendimethalin treatment, or interactions between pendimethalin treatment and tall morningglory density. Harvest efficiency was diminished by tall morningglory density, but was not influenced by pendimethalin treatment or interactions between pendimethalin treatment and tall morningglory density, which indicated that the effects of tall morningglory on harvest efficiency were not conditioned by pendimethalin. Within a 10-m crop row, each additional tall morningglory plant decreased the amount of chile pepper harvested in 1 min by 9.7 g (Fig. 2).

Figure 2 a) Crop yields and b) harvest efficiencies for chile pepper plots that differed in tall morningglory density and pendimethalin treatment. Data are means (± SE) from a study that was conducted near Las Cruces, NM during 2014 and 2015 (four plots per treatment per year).

Table 1 Summary of linear mixed-effects models for crop yield and harvest efficiency responses to pendimethalin treatment and tall morningglory density.

a Akaike information criterion for models with intercepts, fixed effects, and random effects.

b Akaike information criterion for models with only intercepts and random effects.

c The proportion of variance explained by fixed effects.

d The proportion of variance explained by both fixed and random effects.

e The study included two pendimethalin treatments (pendimethalin-treated and nontreated). The coefficient for the factor “pendimethalin” represents the effects of the pendimethalin treatment.

f Abbreviation: PHBPU, tall morningglory (Ipomoea purpurea).

The results of this study, combined with results from a previous study (Schroeder Reference Schroeder1993), indicate that potential weed species in midseason infestations differ in competitiveness with chile pepper. Schroeder (Reference Schroeder1993) showed that, at densities similar to those used in this study, midseason infestations of spurred anoda caused linear reductions in chile pepper yield. Differences in interference between spurred anoda and tall morningglory might reflect species-level variation in growth habit. Specifically, spurred anoda plants feature upright habits that may be more competitive in established chile pepper than the twining habits of tall morningglory. A similar conclusion can be inferred from previous studies that compared functions for soybean yield loss in response to increasing densities of different weed species (Schutte et al. Reference Schutte, Hager and Davis2010; Stoller et al. Reference Stoller, Harrison, Wax, Regnier and Nafziger1987). In these previous studies, ivyleaf morningglory (Ipomoea hederacea Jacq.), which is a climbing summer annual closely related to tall morningglory, was found to affect crop yield less than did some weeds with upright habits (Schutte et al. Reference Schutte, Hager and Davis2010; Stoller et al. Reference Stoller, Harrison, Wax, Regnier and Nafziger1987).

The absence of tall morningglory effects on crop yield might partly reflect enhanced pollination of chile pepper plants in close proximity with tall morningglory. Tall morningglory plants attract insect pollinators, such as bumble bees (Bombus; Galetto and Bernardello Reference Galetto and Bernardello2004), that can also pollinate flowers on chile pepper plants (Tanksley Reference Tanksley1984). This study did not measure pollinator abundance and behavior, and thus, further research is required to clarify the alleged associations among midseason tall morningglory, chile pepper yield, and pollinator activity.

Tall morningglory–induced reductions in harvest efficiency were consistent with previous studies that reported severe disruptions to cotton (Gossypium hirsutum L.) harvest caused by tall morningglory (Buchanan and Burns Reference Buchanan and Burns1971; Crowley and Buchanan Reference Crowley and Buchanan1978). In commercial chile pepper production, professional laborers might not pick areas of the field that are heavily infested with weeds. By using amateur pickers to harvest all chile peppers in research plots, this study potentially underestimated the consequences of tall morningglory on crop pepper yield. Nonetheless, this field study indicated that if a producer chooses to harvest an area that contains midseason tall morningglory, increases in tall morningglory density will prolong the physical activity of harvesting.

A previous study determined that pendimethalin reduced the number of tall morningglory seedlings (Schutte and Cunningham Reference Schutte and Cunningham2015), whereas results from the current study indicated that interference characteristics of surviving tall morningglory were not affected by pendimethalin. The absence of pendimethalin effects on chile pepper yield and harvest efficiency was inconsistent with previous studies that showed that weed interference potentials were attenuated by exposure to pendimethalin. Specifically, Schmenk and Kells (Reference Schmenk and Kells1998) determined that pendimethalin at 0.6 and 1.1 kgaiha−1 reduced velvetleaf (Abutilon theophrasti Medik.) growth and competitiveness in corn (Zea mays L.). Adcock et al. (Reference Adcock, Banks and Bridges1990) determined that pendimethalin at 0.84 kgaiha−1 reduced aboveground fresh weights of tall morningglory and common cocklebur (Xanthium strumarium L.) grown with soybean (Glycine max [L.] Merr.) under greenhouse conditions. Because the interference characteristics of midseason tall morningglory in chile pepper were not conditioned by pendimethalin, projections of tall morningglory interference in this study’s bioeconomic model did not require a parameter for the presence of pendimethalin.

Bioeconomic model

Central to the bioeconomic model were projections for hoeing time and harvest efficiency responses to tall morningglory seedbank density in 10 m of crop row (Fig. 3). These patch-level relationships represent rates of increase in hoeing time and decrease in harvest efficiency caused by one additional tall morningglory plant, that were then superimposed on rates of tall morningglory escape from pendimethalin. Patch-level relationships for hoeing time and harvest efficiency were the foundations for hectare-scale projections for economic implications of tall morningglory seedbank density (Fig. 1).

Figure 3 Projected responses of a) hoeing time and b) harvest efficiency to increasing density in tall morningglory seedbanks. Projections combine rates of escape from pendimethalin with midseason tall morningglory effects on hoeing and harvesting. Midseason tall morningglory emerged approximately 10.5 weeks after chile pepper seeding. At 9.5 weeks after chile pepper seeding, chile pepper stands were thinned to clumps (2 to 3 plants per clump) spaced 0.18 m apart. Underlying rates supporting the projections were determined in this study and in previous studies. Schutte and Cunningham (Reference Schutte and Cunningham2015) determined that 4% of nondormant seeds buried 1 to 2 cm produce seedlings that escape pendimethalin applied at 0.16 kgaiha−1, and 26% of nondormant seeds produce seedlings that escape pendimethalin applied at 0.8 kgaiha−1. Schutte (Reference Schutte2015) determined that, for midseason tall morningglory, one additional plant per 10 m of crop row increased hoeing time by 3.6 s. The current study determined that one additional midseason tall morningglory plant per 10 m of crop row decreased the amount of chile pepper harvested in 1 min by 9.7 g.

To demonstrate the model-based software, a simulation was conducted with the following user-provided variables: tall morningglory infestation severity, 3 tall morningglory patches ha−1; pendimethalin application rate, 0.8 kgaiha−1; expected yield from tall morningglory infested areas, 150 kg fw patch−1 (Fig. 4). Under these conditions, the model projects that as tall morningglory seedbanks increase from 0 to 1000 seeds per patch, tall morningglory density after pendimethalin treatment increases from 0 to 780 plants ha−1. Because additional seedlings require more time for hand hoeing, hoeing is prolonged 47 min ha−1 as tall morningglory seedbanks increase from 0 to 1000 seeds per patch. If plants that survive pendimethalin are not controlled, increases in tall morningglory seedbanks from 0 to 1000 seeds per patch extend hand-harvesting by 27 hrha−1.

Figure 4 Screenshot from the Excel spreadsheet for projecting the effects of tall morningglory seedbank density on weed control outcomes and production expenses in chile pepper production. This Excel sheet allows users to customize model outputs by providing site-specific information. User inputs include patches per hectare, pendimethalin application rate, and expected chile pepper yield from a patch infested with tall morningglory. Model outputs are presented in figures that change according to the user-provided inputs. The “Click to Continue” option advances users to an Excel sheet that provides site-specific financial consequences of one-seed additions to tall morningglory seedbanks.

According to the projected additional times for hoeing and harvesting, and if field labor expenses are New Mexico minimum wage plus taxes, increases in tall morningglory seedbanks from 0 to 1000 seeds per patch add $6.67 to per-hectare expenses for hand-hoeing, and add $223.29 to per-hectare expenses for hand-harvesting. In other words, one seed added in each tall morningglory patch increases per-hectare labor expenses by $0.007 for hand-hoeing, and $0.23 for hand-harvesting. Considering the number of seeds potentially produced by individual plants (26,000 seeds per plant in the absence of competition [Crowley and Buchanan Reference Crowley and Buchanan1982], and 3500 seeds per plant in competition with chile pepper [Schutte Reference Schutte2015]), tall morningglory seed rains can be costly. It is important to note that implied costs of seed rains are likely upper limits, because the model does not account for seed dormancy, burial depths that are suboptimal for emergence, and postdispersal seed mortality. Future studies that combine this study’s bioeconomic model with models that project seed fates and movements (e.g., Renton et al. Reference Renton, Peltzer and Diggle2008; Spokas et al. Reference Spokas, Forcella, Archer and Reicosky2007) will enable detailed understanding of economic consequences of tall morningglory seed rains.

Rogers (Reference Rogers2003) indicated that the decision to adopt a new practice is a multistage process that often begins with recognition of the need for change. This initial stage of the adoption process, as indicated by Rogers (Reference Rogers2003), is the intended target for the model-based software presented in this study. The model is not meant to be a predictive tool but rather a component of broader educational effort on weed seedbank management in agricultural systems in New Mexico. Providing growers information on relationships between weed seedbank density and labor requirements for chile pepper production might promote adoption of seedbank reduction strategies, because costs and availability of labor are primary constraints on chile pepper production in New Mexico (Skaggs et al. Reference Skaggs, Decker and VanLeeuwen2000), and adoption of an integrated weed management practice is generally influenced by grower perceptions of the practice’s economic value in the context of the local cropping system (Llewellyn et al. Reference Llewellyn, Pannell, Lindner and Powles2005).

Acknowledgments

This project was funded in part by the US Department of Agriculture, National Institute of Food and Agriculture, through the Western Integrated Pest Management Center. Salaries and research support were provided by state and federal funds appropriated to the New Mexico Agricultural Experiment Station. I gratefully acknowledge Drey Clark, Ashley Cunningham, Nina Klypin, Christopher Landau, Israel Marquez, Taylor Mesman, Edward Morris, Kyra Smith, and Joseph Wood for their assistance with data collection. I thank Dr. Ram Acharya for his helpful insights on economic modeling.

Footnotes

Associate Editor for this paper: Ramon G. Leon, University of Florida.

References

Literature Cited

Adcock, TE, Banks, PA (1991) Effects of preemergence herbicides on the competitiveness of selected weeds. Weed Sci 39:5456 Google Scholar
Adcock, TE, Banks, PA, Bridges, DC (1990) Effects of preemergence herbicides on soybean (Glycine max): weed competition. Weed Sci 38:108112 Google Scholar
Baskin, CC, Baskin, JM (2014) Seeds: Ecology, Biogeography, and Evolution of Dormancy and Germination. New York: Elsevier Google Scholar
Baucom, RS, Mauricio, R (2010) Defence against the herbicide RoundUp® predates its widespread use. Evol Ecol Res 12:131141 Google Scholar
Bosland, PW, Bailey, AL, Iglesias-Olivas, J (1996) Capsicum Pepper Varieties and Classification. Las Cruces: New Mexico State University, Cooperative Extension Service Circular 530 Google Scholar
Buchanan, GA, Burns, ER (1971) Weed competition in cotton. I. sicklepod and tall morningglory. Weed Sci 19:576579 Google Scholar
Burnside, OC, Wilson, RG, Weisberg, S, Hubbard, KG (1996) Seed longevity of 41 weed species buried 17 years in eastern and western Nebraska. Weed Sci 44:7486 CrossRefGoogle Scholar
Calcada, EA, Lenoir, J, Plue, J, Broeckx, LS, Closset-Kopp, D, Hermy, M, Decocq, G (2015) Spatial patterns of water-deposited seeds control plant species richness and composition in riparian forest landscapes. Landscape Ecol 30:21332146 Google Scholar
Crowley, RH, Buchanan, GA (1978) Competition of four morningglory (Ipomoea spp) species with cotton (Gossypium hirsutum). Weed Sci 26:484488 Google Scholar
Crowley, RH, Buchanan, GA (1982) Variations in seed production and the response to pests of morningglory (Ipomoea) species and smallflower morningglory (Jacquemontia tamnifolia). Weed Sci 30:187190 Google Scholar
Davis, AS (2006) When does it make sense to target the weed seed bank? Weed Sci 54:558565 Google Scholar
Davis, AS, Williams, MM (2007) Variation in wild proso millet (Panicum miliaceum) fecundity in sweet corn has residual effects in snap bean. Weed Sci 55:502507 Google Scholar
Dieleman, JA, Mortensen, DA, Martin, AR (1999) Influence of velvetleaf (Abutilon theophrasti) and common sunflower (Helianthus annuus) density variation on weed management outcomes. Weed Sci 47:8189 Google Scholar
Galetto, L, Bernardello, G (2004) Floral nectaries, nectar production dynamics and chemical composition in six Ipomoea species (Convolvulaceae) in relation to pollinators. Ann Bot 94:269280 Google Scholar
Garfinkel, M, Johnson, M (2015) Pest-removal services provided by birds on small organic farms in northern California. Agr Ecosyst Environ 211:2431 Google Scholar
Gianoli, E (2003) Phenotypic responses of the twining vine Ipomoea purpurea (Convolvulaceae) to physical support availability in sun and shade. Plant Ecol 165:2126 Google Scholar
Grey, TL, Wehtje, GR (2005) Residual herbicide weed control systems in peanut. Weed Technol 19:560567 Google Scholar
[IRS] Internal Revenue Service (2016) Agricultural Employer’s Tax Guide 2016. US Department of Treasury Publication 51, Circular A, Cat. No. 10320R. Washington, DC: US Department of TreasuryGoogle Scholar
Levesque, M, Walthert, L, Weber, P (2016) Soil nutrients influence growth response of temperate tree species to drought. J Ecol 104:377387 Google Scholar
Liphadzi, KB, Dille, JA (2006) Annual weed competitiveness as affected by preemergence herbicide in corn. Weed Sci 54:156165 Google Scholar
Llewellyn, RS, Pannell, DJ, Lindner, RK, Powles, SB (2005) Targeting key perceptions when planning and evaluating extension. Aust J Exp Agr 45:16271633 CrossRefGoogle Scholar
Luschei, EC, Jackson, RD (2005) Research methodologies and statistical approaches for multitactic systems. Weed Sci 53:393403 Google Scholar
Nakagawa, S, Schielzeth, H (2013) A general and simple method for obtaining R 2 from generalized linear mixed-effects models. Methods Ecol Evol 4:133142 Google Scholar
[NCCI] National Council on Compensation Insurance (2016) Workers Compensation Rates by State, Class 0037, Farm: Field Crops. http://classcodes.net/workers-compensation-rates-by-state/. Accessed November 5, 2016Google Scholar
Pinheiro, JC, Bates, DM (2000) Mixed-effects Models in S and S-Plus. New York: Springer Verlag, p. 528 Google Scholar
Renton, M, Peltzer, SC, Diggle, A (2008) Understanding, predicting and managing seedbanks in agricultural systems with the Weed Seed Wizard. Pages 7779 in Proceedings of the 16th Australian Weed Conference. Cairns, Australia: Queensland Weed Society Google Scholar
Rogers, EM (2003) Diffusion of Innovations. 5th edn. New York: Free Press, p. 551 Google Scholar
Sanogo, S, Etarock, BF, Clary, M (2009) Recovery of Verticillium dahliae from tall morningglory (Ipomoea purpurea) in New Mexico and its pathogenicity on chile pepper. Plant Dis 93:428428 Google Scholar
Sanogo, S, Thomas, S, Schroeder, J, Clary, M (2008) Natural co-infection of chile pepper and tall morning glory by Verticillium dahliae and root-knot nematode. Phytopathology 98:S139 Google Scholar
Schmenk, R, Kells, JJ (1998) Effect of soil-applied atrazine and pendimethalin on velvetleaf (Abutilon theophrasti) competitiveness in corn. Weed Technol 12:4752 Google Scholar
Schroeder, J (1992) Oxyfluorfen for directed postemergence weed control in chile peppers (Capsicum annuum). Weed Technol 6:10101014 CrossRefGoogle Scholar
Schroeder, J (1993) Late-season interference of spurred anoda in chile peppers. Weed Sci 41:172179 Google Scholar
Schutte, BJ (2015) Impacts of late-season tall morningglory infestations in chile pepper production. Page 36 in Proceedings of the 55th Annual Meeting of the Weed Science Society of America. Lexington: Weed Science Society of America Google Scholar
Schutte, BJ, Cunningham, A (2015) Tall morningglory (Ipomoea purpurea) seedbank density effects on pendimethalin control outcomes. Weed Technol 29:844853 Google Scholar
Schutte, BJ, Hager, AG, Davis, AS (2010) Respray requests on custom-applied, glyphosate-resistant soybeans in Illinois: how many and why. Weed Technol 24:590598 Google Scholar
Skaggs, RK, Decker, M, VanLeeuwen, D (2000) A Survey of Southern New Mexico Chile Producers: Production Practices and Problems. Las Cruces, NM: New Mexico State University, Agricultural Experiment Station. P 68 Google Scholar
Spokas, K, Forcella, F, Archer, D, Reicosky, D (2007) SeedChaser: vertical soil tillage distribution model. Comput Electron Agr 57:6273 CrossRefGoogle Scholar
Stoller, EW, Harrison, SK, Wax, LM, Regnier, EE, Nafziger, ED (1987) Weed interference in soybeans (Glycine max). Weed Sci 3:155181 Google Scholar
Swanton, CJ, Mahoney, KJ, Chandler, K, Gulden, RH (2008) Integrated weed management: knowledge-based weed management systems. Weed Sci 56:168172 Google Scholar
Tanksley, SD (1984) High rates of cross pollination in chile pepper. Hortscience 19:580582 Google Scholar
[USDA NASS] US Department of Agriculure, National Agricultural Statistics Service (2016) Quick Stats. http://quickstats.nass.usda.gov/. Accessed May 17, 2016Google Scholar
Wilcut, JW, Jordan, DL, Vencill, WK, Richburg, JS (1997) Weed management in cotton (Gossypium hirsutum) with soil-applied and post-directed herbicides. Weed Technol 11:221226 Google Scholar
Wilkerson, GG, Wiles, LJ, Bennett, AC (2002) Weed management decision models: pitfalls, perceptions, and possibilities of the economic threshold approach. Weed Sci 50:411424 CrossRefGoogle Scholar
Zar, J (1999) Biostatistical Analysis. 4th edn. Upper Saddle River, NJ: Prentice Hall Google Scholar
Figure 0

Figure 1 Underlying model for software that presents relationships between tall morningglory seedbank density and weed-control cost parameters in chile pepper production. The model initiates with the number of nondormant tall morningglory seeds buried at optimal depths for emergence within a 10-m2 area that is hereafter referred to as a “patch”. Users interact with the model by specifying the patches per hectare, pendimethalin application rate, and expected crop yield from a tall morningglory patch. User-defined variables, which are depicted with dashed-line boxes. Solid-line boxes present state variables. Rates that regulate transitions between state variables are provided in the text. Abbreviation: PHBPU, tall morningglory (Ipomoea purpurea).

Figure 1

Figure 2 a) Crop yields and b) harvest efficiencies for chile pepper plots that differed in tall morningglory density and pendimethalin treatment. Data are means (± SE) from a study that was conducted near Las Cruces, NM during 2014 and 2015 (four plots per treatment per year).

Figure 2

Table 1 Summary of linear mixed-effects models for crop yield and harvest efficiency responses to pendimethalin treatment and tall morningglory density.

Figure 3

Figure 3 Projected responses of a) hoeing time and b) harvest efficiency to increasing density in tall morningglory seedbanks. Projections combine rates of escape from pendimethalin with midseason tall morningglory effects on hoeing and harvesting. Midseason tall morningglory emerged approximately 10.5 weeks after chile pepper seeding. At 9.5 weeks after chile pepper seeding, chile pepper stands were thinned to clumps (2 to 3 plants per clump) spaced 0.18 m apart. Underlying rates supporting the projections were determined in this study and in previous studies. Schutte and Cunningham (2015) determined that 4% of nondormant seeds buried 1 to 2 cm produce seedlings that escape pendimethalin applied at 0.16 kgaiha−1, and 26% of nondormant seeds produce seedlings that escape pendimethalin applied at 0.8 kgaiha−1. Schutte (2015) determined that, for midseason tall morningglory, one additional plant per 10 m of crop row increased hoeing time by 3.6 s. The current study determined that one additional midseason tall morningglory plant per 10 m of crop row decreased the amount of chile pepper harvested in 1 min by 9.7 g.

Figure 4

Figure 4 Screenshot from the Excel spreadsheet for projecting the effects of tall morningglory seedbank density on weed control outcomes and production expenses in chile pepper production. This Excel sheet allows users to customize model outputs by providing site-specific information. User inputs include patches per hectare, pendimethalin application rate, and expected chile pepper yield from a patch infested with tall morningglory. Model outputs are presented in figures that change according to the user-provided inputs. The “Click to Continue” option advances users to an Excel sheet that provides site-specific financial consequences of one-seed additions to tall morningglory seedbanks.