Hostname: page-component-78c5997874-dh8gc Total loading time: 0 Render date: 2024-11-04T19:29:55.698Z Has data issue: false hasContentIssue false

Weed community differences in row crops with varying input levels in Ghana

Published online by Cambridge University Press:  09 October 2024

Fernando H. Oreja*
Affiliation:
Assistant Professor, Department of Plant and Environmental Sciences, Clemson University, Clemson, SC, USA
Stephen Arthur
Affiliation:
Principal Technologist, Council for Scientific and Industrial Research–Crops Research Institute, Kumasi, Ghana
Grace Bolfrey-Arku
Affiliation:
Senior Research Scientist, Council for Scientific and Industrial Research–Crops Research Institute, Kumasi, Ghana
Moses B. Mochiah
Affiliation:
Professor, Council for Scientific and Industrial Research–College of Science and Technology, Kumasi, Ghana
Victoria Klutse
Affiliation:
Research Technician, Council for Scientific and Industrial Research–Crops Research Institute, Kumasi, Ghana
Maxwell Yorke
Affiliation:
Research Technician, Council for Scientific and Industrial Research–Crops Research Institute, Kumasi, Ghana
Solomon Hukporti
Affiliation:
Research Technician, Council for Scientific and Industrial Research–Crops Research Institute, Kumasi, Ghana
Israel K. Dzomeku
Affiliation:
Professor, University for Development Studies, Tamale, Ghana
Georgie Y. Mahama
Affiliation:
Research Scientist, Council for Scientific and Industrial Research–Savanna Agricultural Research Institute, Tamale, Ghana
Jerry A. Nboyine
Affiliation:
Senior Research Scientist, Council for Scientific and Industrial Research–Savanna Agricultural Research Institute, Tamale, Ghana
Ahmed Seidu
Affiliation:
Crop Protectionist, Council for Scientific and Industrial Research–Savanna Agricultural Research Institute, Tamale, Ghana
Richard Akromah
Affiliation:
Professor, Kwame Nkrumah University of Science and Technology, Kumasi, Ghana
Joseph Sarkodie-Addo
Affiliation:
Associate Professor, Kwame Nkrumah University of Science and Technology, Kumasi, Ghana
David L. Jordan
Affiliation:
William Neal Reynolds Distinguished Professor, Department of Crop and Soil Sciences, North Carolina State University, Raleigh, NC, USA
Ramon G. Leon
Affiliation:
William Neal Reynolds Distinguished Professor and University Faculty Scholar, Department of Crop and Soil Sciences, North Carolina State University, Raleigh, NC, USA
*
Corresponding authors: Fernando H. Oreja; Email: [email protected]; Ramon Leon; Email: [email protected]
Rights & Permissions [Opens in a new window]

Abstract

Peanut (Arachis hypogaea L.) and maize (Zea mays L.) are essential crops for Ghana’s economy and food security, but weed infestation poses a significant threat to their cultivation. Crop rotations influence weed communities, but little is known about these processes in peanut-cropping systems in West Africa. This study investigated the impact of different crop rotations and input levels on weed communities in Ghana over 3 yr. Results showed that low inputs (absence of herbicide and fertilization) favored species richness, while higher input levels (weed control with herbicides and fertilizer use) reduced it. Diversity and evenness were also affected by inputs, with varying patterns across locations and seasons. Weed population growth rates (λ) varied significantly by location and treatment; all management programs resulted in increasing weed populations. Principal component analysis revealed distinct associations between weed species and crop management. The majority of weed species exhibited a generalist behavior and did not associate with a particular management. However, billygoat weed (Ageratum conyzoides L.) and Benghal dayflower (Commelina benghalensis L.) were positively associated with high-input systems, while purple nutsedge (Cyperus rotundus L.) exhibited strong associations with low and medium inputs. The weed–crop rotation dynamics described here demonstrate how management drives the selection of weed species that are more pervasive and interfere with important food crops in Ghanaian agriculture.

Type
Research Article
Creative Commons
Creative Common License - CCCreative Common License - BY
This is an Open Access article, distributed under the terms of the Creative Commons Attribution licence (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted re-use, distribution and reproduction, provided the original article is properly cited.
Copyright
© The Author(s), 2024. Published by Cambridge University Press on behalf of Weed Science Society of America

Introduction

Peanut (syn.: groundnut; Arachis hypogaea L.) and maize (Zea mays L.) are very important crops for Ghana’s economy, nutrition, and food security. Their cultivation, postharvest handling, and processing serve as a source of income and employment to many smallholder farmers, processors, retailers, and other workers in the value chain. In Ghana, more than 70% of farmers in the three northern regions and transitional zones cultivate peanuts (Owusu-Adjei et al. Reference Owusu-Adjei, Baah-Mintah and Salifu2017). Apart from serving as an important source of energy in animal fodder, peanut haulms help improve soil structure and fertility. While the kernels are eaten raw, boiled, or roasted, they are also a major ingredient for the preparation of most ready-to-use supplementary food and therapeutic food to combat malnutrition (Manary and Callaghan-Gillespie Reference Manary and Callaghan-Gillespie2018). The kernels are also processed into peanut flour and confectionary. Maize also accounts for over half of the total cereal production in Ghana (Kankam-Boadu et al. Reference Kankam-Boadu, Sarkodie-Addo and Amagloh2018; Manary and Callaghan-Gillespie Reference Manary and Callaghan-Gillespie2018). More than 15% of the maize produced in the country is utilized as animal feed (mostly for poultry and fish), and the demand for maize keeps increasing (Kankam-Boadu et al. Reference Kankam-Boadu, Sarkodie-Addo and Amagloh2018; Koffi et al. Reference Koffi, Kyerematen, Eziah, Osei-Mensah, Afreh-Nuamah, Aboagye, Osae and Meagher2020; Wongnaa et al. Reference Wongnaa, Awunyo-Victor and Bakang2018). Weeds are important pests in peanut and maize fields in Ghana, causing yield losses of between 50% and 80% in peanut fields in West Africa (Dzomeku et al. Reference Dzomeku, Abudulai, Brandenburg and Jordan2009).

Among the biological factors reducing yield, weeds are the most important in most extensive crops (Oerke 2006). Farmers from sub-Saharan Africa (SSA) considered weeds the most important constraint for crop production (Adesina et al. Reference Adesina, Johnson and Heinrichs1994; Ogwuike et al. Reference Ogwuike, Rodenburg, Diagne, Agboh-Noameshie and Amovin-Assagba2014). This holds particular significance in regions with socioeconomic and environmental fragility, such as SSA, where weeds have the potential to exacerbate challenges related to food security (Bàrberi Reference Bàrberi2019). Additionally, manual labor is often the major input in these countries, and its availability is critical for timing operations to maintain productivity. Many weed management methods currently used to prevent yield reductions require extensive manual labor (Adesina et al. Reference Adesina, Johnson and Heinrichs1994; Ogwuike et al. Reference Ogwuike, Rodenburg, Diagne, Agboh-Noameshie and Amovin-Assagba2014). In addition, smallholder farmers frequently face time constraints and shortages in labor availability for weeding (Dahlin and Rusinamhodzi Reference Dahlin and Rusinamhodzi2019; Rodenburg and Johnson Reference Rodenburg and Johnson2009). To reduce this labor demand and increase productivity, it is essential to enhance overall crop management and implement complementary weed control strategies, such as utilizing herbicides, fertilizers, and fungicides to enhance crop competitiveness against weeds and other pests (Ogwuike et al. Reference Ogwuike, Rodenburg, Diagne, Agboh-Noameshie and Amovin-Assagba2014). However, intensive agricultural practices may have ecological and environmental costs (Stoate et al. Reference Stoate, Baldi, Beja, Boatman, Herzon, Van Doorn, de Snoo, Rakosy and Ramwell2009). While effective weed management is crucial for maintaining crop yield, fostering more diverse weed communities with low plant densities is a desirable objective owing to the stability and resilience they confer on agroecosystems (Smith and Gross Reference Smith and Gross2006). Diverse weed communities have the capacity to diminish the prevalence of a few highly competitive species that tend to increase the control costs and reduce crop yield (Storkey and Neve Reference Storkey and Neve2018).

Crop rotation affects key weed processes such as weed establishment, growth, and fecundity, all of which modulate weed population growth rate. Although integrated weed management (Swanton and Weise Reference Swanton and Weise1991) comprises multiple weed control tools, crop rotation has a major impact on the selection, timing, and intensity of practices such as planting, irrigation, fertilization, herbicide application, tillage, cultivation, and harvesting, all influencing weed–crop interactions (Andrade et al. Reference Andrade, Satorre, Ermácora and Poggio2017). Such diversity of practices when implemented in variable ways creates an unstable environment that reduces the reproductive success of weeds (Gaba et al. Reference Gaba, Fried, Kazakou, Chauvel and Navas2014; Weisberger et al. Reference Weisberger, Nichols and Liebman2019). In simplified and less variable environments, repetitive weed management practices tend to favor a few species over others, reducing plant diversity (Oreja et al. Reference Oreja, Inman, Jordan and Leon2021). Diversification of crop rotation can be a suitable option to prevent weed communities becoming less diverse and dominated by a few aggressive and difficult to manage species (Adeux et al. Reference Adeux, Vieren, Carlesi, Bàrberi, Munier-Jolain and Cordeau2019; Storkey and Neve Reference Storkey and Neve2018).

The level of application of external inputs (e.g., fertilizers, pesticides) to fields is another important factor that may shape weed communities (Booth and Swanton Reference Booth and Swanton2002). Weeds that take advantage of those inputs and exhibit lower sensitivity to pesticides will increase their reproductive output and disproportionately increase their populations compared with less fit species. For example, high levels of herbicide use can favor species with longer seed longevity and persistence in soil (Oreja et al. Reference Oreja, Inman, Jordan and Leon2021), as well as species with low herbicide sensitivity (Hyvönen and Salonen Reference Hyvönen and Salonen2002; Oreja et al. Reference Oreja, Inman, Jordan, Vann, Jennings and Leon2022). Additionally, systems with high fertilizer usage may favor species with more intense nutrient acquisition and utilization (Ryan et al. Reference Ryan, Smith, Mirsky, Mortensen and Seidel2010; Yin et al. Reference Yin, Cai and Zhong2006).

Despite the adoption of high-input systems that would reduce the dependence on manual labor for Ghanaian farmers and the yield reduction risks, a better understanding of how these systems affect weed community structures could facilitate their adoption in a more sustainable manner. This information is scarce in SSA and completely absent in Ghana. Therefore, the objective of this study was to assess how different crop rotations and input levels influence the diversity and balance of weed communities in Ghanaian farms.

Materials and Methods

Experimental Site

An experiment was conducted at three locations in Ghana during the years 2019, 2020, and 2021: Council for Scientific and Industrial Research (CSIR)–Crops Research Institute (CRI), Fumesua Station (Kumasi), in the Ashanti Region (6.7139°N, 1.5313°W); and Savannah Agricultural Research Institute (SARI) stations at Nyankpala (Tamale) in the Northern Region (9.3908°N, 1.0068°W), and Tanina (Wa) in the Upper West Region (9.8858°N, 2.4624°W). Kumasi is wetter (∼33% more rainfall) and cooler (∼5 C lower temperature) than Tamale and Wa (Table 1). Kumasi falls within the semi-deciduous rainforest agroecological zone and experiences a bimodal rainfall pattern, with a long rainy season occurring between March and July and a short season between September and November, and the soil is Ferric Acrisol (IUSS Working Group WRB 2015). Tamale and Wa fall within the Guinea Savannah ecological zone, which experiences a unimodal rainfall pattern between May and October, and the soils are Plinthosols and Lixisols/Luvisols, respectively (IUSS Working Group WRB 2015). The three locations had distinct soil properties and marked differences in soil textures (Table 2).

Table 1. Annual rainfall and average annual mean, minimum and maximum temperatures for locations in Ghana

a n.a., on-site data not available, but historical records are 879 mm yr−1.

Table 2. Chemical and physical properties of soil profile (0–20cm) in experimental locations in Ghana

a 1:1 (H2O).

b Mehlich 3.

Experimental Design and Treatments

A 3-yr-rotation experiment was conducted in three locations from 2019 to 2021. In Tamale and Wa, each having a single growing season per year, one experiment was established annually. In Kumasi, experiments were conducted during both the long and short seasons (Table 3). The experiment was a factorial design with three factors (i.e., crop rotation, input level, and crop variety) arranged in split plots with three replications. There were two different rotations with maize and peanut with opposite crop phases to avoid confounding effects with year (i.e., maize–peanut–maize and peanut–maize–peanut. These crops were selected because they are two of the most important food crops in the area and are commonly rotated (Table 3). Treatment input levels were low input (LI), medium input (MI), and high input (HI) (Tables 4 and 5). Peanut varieties were ‘Yenyawso’ and ‘Shi Tao Chi’, and maize varieties were CRI ‘AhoↄfꜪ’ as a hybrid and CRI ‘Omankwa’ as an open-pollinated variety; peanut and maize were planted at optimum dates for the region (Table 6). Rotations were designated as the main plots, while pairwise combinations of management program and varieties were randomly assigned to subplots within the rotations. Each plot was 5 m by 5 m with a 2-m alley between replications and a 1-m alley between plots. Peanuts were planted with one seed per hill at a spacing of 50 cm between rows and 20 cm within rows, while maize was planted with two seeds per hill at 80 cm between rows and 40 cm within rows. The integrity of each plot was maintained to ensure consistent rotations and the application of the same input intensity or crop variety throughout the experiment (2019 to 2021). In the first year, the experimental fields in all locations were plowed and harrowed; however, in subsequent seasons/years, plots were manually tilled/turned to preserve plot integrity.

Table 3. Crops included on each rotation and years in locations in Ghana: peanut–maize and maize–peanut in Kumasi and peanut–maize–peanut and maize–maize–peanut in Tamale and Wa

Table 4. Input intensity and details for peanut and maize production in Kumasi, Tamale, and Wa, Ghana

Mixture of vegetable oils extracted from Butyrospermum parkii (G. Don) Kotschy (shea butter), Cocos nucifera L. (coconut Oil), Elaeis guineensis Jacqu. (palm kennel), Theobroma cacao L. (cocoa) sprayed foliarly.

Table 5. Active ingredient(s), rate and time of application of inputs for peanut and maize production in Kumasi, Tamale, and Wa, Ghana

Mixture of vegetable oils extracted from Butyrospermum parkii (G. Don) Kotschy (shea butter), Cocos nucifera L. (coconut Oil), Elaeis guineensis Jacqu. (palm kennel), Theobroma cacao L. (cocoa) sprayed foliarly.

a Bt, Bacillus thuringiensis; DAP, days after planting; PRN, when needed; WAP, weeks after planting.

Table 6. Planting and harvesting dates in different locations and seasons in Ghana

Measurements

Weed density was evaluated at 3 wk after planting (WAP) by randomly placing a quadrat measuring 0.5 m by 0.3 m three times within the central three rows of maize plots and five times within the central four rows of peanut plots. The number of individuals per weed species within each quadrat was recorded.

Synthetic diversity indices were used to characterize the communities (Magurran Reference Magurran2013). Species richness (S) was the total number of species on each treatment; Shannon’s diversity index (H′) was calculated as:

([1]) $$H' = \mathop \sum \nolimits_{i = 1}^S {p_i}\;ln\left( {{p_i}} \right)$$

where p i is the proportion of individuals of species i in relation with the total number of individuals of all species combined. The relative contribution of each species to the community was determined as evenness (J):

([2]) $$J = {{H'} \over {ln\left( S \right)}}$$

J ranges from 0 to 1, with lower values indicating dominance of one or few species and higher values indicating a more equitable representation of each species within the community. Moreover, the population growth rate of all weed species together on each treatment, lambda (λ), was estimated as:

([3]) $${\rm{\lambda }} = \root t \of {{{{n_{t + 1}}} \over {{n_t}}}} $$

where t is the number of seasons, n t is the total number of plants registered at 3 WAP, and n t+1 is the total number of plants registered the next season at 3 WAP.

Statistical Analysis

To examine the impact of different rotations, input levels, and crop varieties on weed community richness, diversity, evenness, and population growth rate values, ANOVA was conducted. This was followed by Tukey’s honestly significant difference (HSD) test based on a general linear model using the nlme package (Pinheiro et al. Reference Pinheiro, Bates, DebRoy and Sarkar2018) in R (R Core Team 2020). ANOVA assumptions (homoscedasticity and normal distribution of residuals) were tested, and when these assumptions were not met, the data were arcsine square-root transformed before the analyses. In cases in which transformations were not enough to achieve the homogeneity of variance, data were analyzed using the generalized least squares method, which corrected for heterogeneity of variances.

As the effects of crop management decisions on a particular species cannot be captured by a single index score, a principal component analysis (PCA) (Krzanowski Reference Krzanowski2000) was performed using PC-ORD Multivariate Analysis of Ecological Data v. 5.0. (Grandin Reference Grandin2006; McCune and Mefford Reference McCune and Mefford1999) for each location. The response variable used was the density at 3 WAP of each species in the last season (i.e., 2021), and the explanatory variables were the treatments. Treatment codes, weed scientific names, and codes of weed species in each location are listed in Table 7.

Results and Discussion

Community Structure

Since no differences on species richness (S) were observed among varieties and rotations in any location (P = 0.74, 0.81, and 0.05 for Kumasi, Wa, and Tamale, respectively), results were analyzed by pooling rotation data. In Kumasi, the number of species varied throughout the duration of the experiment (Figure 1A). Thus, S was lowered during long seasons compared with short seasons (P < 0.0001). In Tamale and Wa, S doubled from 2019 to 2021 (P < 0.0001) regardless of input levels and rotation (Figure 1B and 1C). Input management had a major impact on S, which was consistently lower (P < 0.0001) in treatments with high inputs compared with medium and low inputs across all locations and seasons (Figure 1).

Figure 1. Richness (S) estimated during three seasons and for different input levels (high, medium, and low) in locations in Ghana: (A) Kumasi, (B) Tamale, and (C) Wa. Different uppercase letters indicate differences among seasons, and different lowercase letters among input levels according to Tukey’s HSD (P < 0.05). Errors bars represent standard error of the mean.

Since no differences in H′ were observed due to variety and rotation in any location (P = 0.21, 0.81, and 0.89 for Kumasi, Wa, and Tamale, respectively), the analysis was done by pooling rotation data. In Kumasi, no differences were observed in H′ among seasons, with an average of 1.45 (Figure 2A). However, H′ increased in Tamale and Wa from season 2019 to season 2021 (P < 0.0001), rising 64% and 44%, respectively (Figure 2). Regarding input levels, there was a notable difference (P < 0.0001) in H′, with high-input levels resulting in lower H′ values (1.19 in Tamale and 0.89 in Wa) than medium- and low-input levels (1.39 in Tamale and 1.38 in Wa) (Figure 2). In contrast, Kumasi had more variation, with H′ being higher with high-input levels during the short seasons (2019 and 2020) compared with medium- and low-input levels, but it was lower at high-input levels during long seasons and the last short season compared with medium- and low-input levels (Figure 2).

Figure 2. Shannon index (H′) in response to input level (high, medium, and low) in different seasons in (A) Kumasi, (B) Tamale, and (C) Wa, Ghana. Different uppercase letters indicate significant differences among seasons, and different lowercase letters indicate significant differences among input levels according to Tukey’s (P < 0.05). Errors bars represent standard error of the mean.

Regarding evenness (J), no differences were observed among varieties and rotations in any location (P = 0.09, 0.28, and 0.39 for Kumasi, Wa, and Tamale, respectively). Therefore, seasons and inputs results were analyzed with data combined across rotations. In Kumasi, J was lower (P < 0.0001) in the last two seasons of the experiment compared with the second one (Figure 3A). On the other hand, in Tamale, J increased from season 2019 (0.64 on average) to season 2020 (0.75 on average) and season 2021 (0.74 on average) (Figure 3B). In Wa, a reduction in J was observed from season 2019 to season 2020, from 0.76 to 0.66 on average, but in season 2021, J increased (P < 0.0001) to the former level (0.74 on average) (Figure 3C). No differences were observed among input levels for J in Tamale (P = 0.27) and Wa (P = 0.31). In Kumasi, J was higher (P < 0.0001) in the high-input treatment than in the medium- and low-input treatments in long seasons. However, the reverse was observed in short seasons, when J was higher (P < 0.0001) in the medium- and low-input treatments than in the high-input treatment (Figure 3A), exhibiting an inverse pattern to that observed for H′.

Figure 3. Evenness (J) estimated at different seasons and for different input levels (high, medium, and low) in locations in Ghana: (A) Kumasi, (B) Tamale, and (C) Wa. Different uppercase letters mean significant differences among seasons, and different lowercase letters mean significant differences among input levels according to Tukey’s (P < 0.05). Errors bars represent standard error of the mean.

In general, high-input treatments resulted in low S and H′ across all locations and seasons. Additionally, a lower J, which results from the dominance of one or few species, was also associated with high inputs in comparison to medium and low inputs, although patterns varied across locations and seasons. This could be attributed to the use of herbicides in the high-input system, in contrast to the medium- and low-input levels, which did not employ herbicides. Herbicides are among the most important filters shaping weed community structures at field scale (Grundy et al. Reference Grundy, Mead, Bond, Clark and Burston2011; Owen Reference Owen2008). In environments where the same herbicide program is repeatedly used, less-sensitive species are favored over more-sensitive ones, leading to a reduction in the H′ and J of the community (Oreja et al. Reference Oreja, Inman, Jordan and Leon2021). Similarly, the use of fertilizers can influence weed composition due to variation in nutrient use among species (Pinke et al. Reference Pinke, Blazsek, Magyar, Nagy, Karácsony, Czúcz and Botta-Dukát2016). It is important to mention that those factors do not act individually, so the joint effects of herbicides, fertilizers, and fungicides included in the high-input treatment (Table 4) likely determined how the potential for both crop competitive ability and weed interference were expressed. Gaba et al. (Reference Gaba, Caneill, Nicolardot, Perronne and Bretagnolle2018) reported that nitrogen fertilization increased the competitive ability of the crop, which resulted in lower weed biomass, including that of dominant species. Additionally, weeds with small seedbanks may experience significant reductions in their performance, potentially leading to eradication (Oreja et al. Reference Oreja, Inman, Jordan, Vann, Jennings and Leon2022). At the same time, weeds are not only competing with crops, but they are also competing with other weeds. In some cases, weed–weed competition can have a higher impact on weed growth than crop–weed competition (Romillac et al. Reference Romillac, Piutti, Slezack-Deschaumes and Gaba2023). In Tamale and Wa, there was a dramatic increase in S and H′ from the beginning to the end of the experiment, regardless of input levels, rotations, and varieties. The lack of effects of different rotations on weed community structures at each location may be related to the similarity between the rotations evaluated. The rotations did not differ in crop types but only in the years the crops were planted. Weed species that are suppressed or favored in one season might experience the opposite in the next, thus compensating for any differences. Similarly, varieties with comparable cycle lengths or plant architectures could contribute to the lack of differences in weed community composition. For example, Haden et al. (Reference Haden, Duxbury, DiTommaso and Losey2007) found no effect on weed community structure in response to the presence of either of two rice (Oryza sativa L.) cultivars differing in height and presumably weed-suppression potential. Similarly, Schöb et al. (Reference Schöb, Hortal, Karley, Morcillo, Newton, Pakeman, Powell, Anderson and Brooker2017) reported that there were no differences among barley (Hordeum vulgare L.) genotypes on the establishment of new weed species and the richness of the community. They concluded that very small variation in growth traits among genotypes was the main cause of a lack of differential responses by weeds.

Evenness fluctuated across seasons but tended to increase toward the end of the experiment, indicating a balance in abundance among weed species. However, in Kumasi, the number of species and the H′ remained relatively constant throughout the experiment, with fluctuations observed between long and short rainy seasons. At this location, J decreased over time until the last two seasons, indicating a disproportionate impact of management on the abundance of different weed species (Nkoa et al. Reference Nkoa, Owen and Swanton2015). Kumasi is located in a different agroecological zone with distinct rainfall patterns and soil types compared with Tamale and Wa (Table 1). These differences may modulate the community structure differently among locations. Pinke et al. (Reference Pinke, Blazsek, Magyar, Nagy, Karácsony, Czúcz and Botta-Dukát2016) reported that environmental variables, such as temperature and precipitation, which are highly dependent on location, were strong determinants in shaping species composition in the fields. In addition, Kumasi tended to have higher values of richness and diversity than Wa and Tamale throughout the seasons, except for Wa in season 2021. Kumasi has five crops in 3 yr compared with three crops in 3 yr in Wa and Tamale, meaning the surface was covered with a crop during more time. Andrade et al. (Reference Andrade, Satorre, Ermácora and Poggio2017) found that a field with a higher number of days with a high crop cover reduced the frequency of common weeds, which leads to reductions in the dominance of a few species and increases the diversity and evenness.

Weed Population Growth Index (ʎ)

Because no interaction was observed between Tamale and Wa for λ, data of these two locations were pooled for the analysis. Also, data were pooled for varieties due to the lack of responses to this factor. In Kumasi, a higher λ (1.12; P = 0.02) was observed in the maize–peanut–maize rotation than in the peanut–maize–peanut rotation (λ = 1.05) (Figure 4A), but no differences were observed between rotations in Tamale and Wa (Figure 4). Regarding input levels, a higher λ (P < 0.0001) was observed for treatments with medium- and low-input levels in Kumasi (1.15 for medium-input level and 1.14 for low-input level) compared with high-input levels (0.97). On the contrary, in Tamale and Wa, higher λ values (P < 0.0001) were registered in high-input levels (1.98 and 2.08 for Wa and Tamale, respectively) than in medium- (1.43 and 1.79 for Wa and Tamale, respectively) and low-input (1.32 and 1.76 for Wa and Tamale, respectively) levels (Figure 4).

Figure 4. Population growth rate (lambda or λ) estimated at different crop rotations and for different input levels (high, medium, and low) in locations in Ghana: (A) Kumasi, (B) Tamale, and (C) Wa. Different uppercase letters indicate significant differences among crop rotations and different lowercase letters indicate significant differences among input levels according to Tukey’s (P < 0.05). M, maize; P, peanut. Errors bars represent standard error of the mean.

The effects on population growth rate were very different depending on the location; in Kumasi, certain rotations and lower input levels were associated with higher λ, while the opposite trend was observed in Tamale and Wa. Despite these differences, all the locations and treatments showed λ values above 1, except for the peanut–maize rotation at high-input level (λ = 0.89). These λ values above 1 mean that weed populations increased from the beginning of the experiment to the end. The efficacy of manual weeding is highly dependent on weed density, Thus, if λ is low, manual weeding may be superior to herbicides. However, if weed density is very high, then manual weeding efficacy decreases (Beltran et al. Reference Beltran, Pannell and Doole2012). On average among seasons and treatments, weed density was 2.6 and 2.1 times higher in Tamale and Wa than in Kumasi, respectively. This difference could explain the higher efficacy of treatments without herbicides compared with those with herbicides in Kumasi. Except for the peanut-maize rotation at high-input level, none of the evaluated programs were successful in reducing λ. In other words, all evaluated management programs had trajectories that will result in progressively more serious weed problems. Therefore, more weed management programs must be evaluated in order to reduce λ by maintaining weed populations at sustainable values and supporting diverse weed communities.

Species Analysis

In the PCA for Kumasi, axes 1 and 2 accounted for 70.6% and 16.8% of the total variation, respectively. Two distinct groups of explanatory variables were identified (Figure 5). One comprised treatments with high inputs (including residual herbicides) and the other group consisted of treatments with low and medium inputs (without herbicides). No associations were observed for rotations or crop genotypes. On average across all treatments, most species with low densities were positioned in the opposite direction of the treatment vectors. Conversely, species with high densities across treatments tended to be in the same direction as the treatment vectors (Figure 5; Table 7). The positioning of these species corresponded to their association with specific treatments. For instance, species such as carpetweed (Mollugo verticillata L.), southern crabgrass [Digitaria ciliaris (Retz.) Koeler], purple nutsedge (Cyperus rotundus L.), and flat-top mile (Oldenlandia corymbosa L.) were associated with low- and medium-input levels. Conversely, billygoat weed (Ageratum conyzoides L.) and wild poinsettia (Euphorbia heterophylla L.) were positively associated with high inputs (Figure 5).

Figure 5. Principal component analysis (PCA) ordination of density values registered 3 wk after planting date in the last season in Kumasi, Ghana. Vectors represent the explanatory variable and are the treatments. See Table 7 for nomenclature. Rotations are the first two numbers; crop varieties are the letter and the last two letters for the input level.

Table 7. (A) Abbreviations (codes) for treatments representing the explanatory variables (vectors) and (B) weed codes representing species used for a principal component analysis (PCA) as shown in Figure 5

In Tamale, axis 1 accounted for 85.0% and axis 2 for 7.4% of the total variation (Figure 6). As was observed in Kumasi, two distinct groups of explanatory variables were identified, one comprised treatment with high inputs (including residual herbicides) and the other group consisted of treatments with low and medium inputs (without herbicides) (Figure 6). No associations were observed for rotations or crop varieties. Most species with low densities, on average across all treatments, were positioned in the opposite direction of the treatment vectors. Ageratum conyzoides was the only species associated with high inputs. Cyperus rotundus, linear primrose-willow [Ludwigia hyssopifolia (G. Don) Exell], kodomillet (Paspalum scrobiculatum L.), pignut [Hyptis suaveolens (L.) Poit.], arrasa con todo (Gomphrena celosioides auct. non Mart.; syn.: Gomphrena serrata L.), and D. ciliaris were more clearly associated with low and medium inputs (Figure 6).

Figure 6. Principal component analysis (PCA) ordination of density values registered 3 wk after planting date in the last season in Tamale, Ghana. Vectors represent the explanatory variable and are the treatments. See Table 7 for nomenclature. Rotations are the first two numbers; crop varieties are the letter and the last two letters for the input level.

In Wa, axis 1 accounted for 75.4% and axis 2 for 15.3% of the total variation. Two distinct groups of explanatory variables were identified but were not as separated as in the other two locations (Figure 7). No associations were observed for rotations or crop varieties. Only C. benghalensis was associated with treatments with high inputs. Regarding low- and medium-input levels, tropical girdlepod [Mitracarpus villosus (Sw.) Cham. & Schltdl. ex DC.] and C. rotundus exhibited the strongest associations, while Indian sandbur (Cenchrus biflorus Roxb.), marubio (Hyptis spicigera Lam.), Cyanotis lanata Benth., H. suaveolens, and D. ciliaris were also associated but to a lesser extent (Figure 7).

Figure 7. Principal component analysis (PCA) ordination of density values registered 3 wk after planting date in the last season in Wa, Ghana. Vectors represent the explanatory variable and are the treatments. See Table 7 for nomenclature. Rotations are the first two numbers; crop varieties are the letter and the last two letters for the input level.

Species associations with input levels were consistent across locations, with certain species being more prevalent in low- to medium-input systems, while others were associated with high-input systems. The differences observed among locations in community structures were also reflected in weed compositions. Out of the 47 different species identified in our study, only 10 (21%) were found in all three locations. This highlights the significant influence of environmental variables, such as temperature and precipitation, in shaping weed community compositions (Pinke et al. Reference Pinke, Blazsek, Magyar, Nagy, Karácsony, Czúcz and Botta-Dukát2016). Most of the species registered in all the locations in this work, such as Cleome viscosa L., C. benghalensis, C. rotundus, D. ciliaris, E. heterophylla, Euphorbia hirta L., P. scrobiculatum, Phyllanthus amarus Schumach. & Thonn., Talinum triangulare (Jacq.), and Tridax procumbens L., were previously reported in other crop field surveys in SSA (Kent et al. Reference Kent, Johnson and Becker2001; Salaudeen et al. Reference Salaudeen, Daniya, Olaniyi, Folorunso, Bala, Abdullahi, Nuhu, Adedigba, Oluwole, Bankole and Macarthy2022).

Regarding the associations of species with input levels, some species were negatively associated with high input–level scenarios with herbicide and fertilizer applied. Most of those species, including D. ciliaris (Ahmed and Chauhan Reference Ahmed and Chauhan2015), M. verticillata (Kharel et al. Reference Kharel, Devkota, MacDonald and Tillman2022), P. scrobiculatum (Hossain et al. Reference Hossain, Timsina, Johnson, Gathala and Krupnik2020), H. suaveolens (Imoloame Reference Imoloame2017), and C. biflorus (Ahmed et al. Reference Ahmed, Mostafa, Reda, El-Adawy and Khozimy2017), are effectively controlled with pendimethalin. Conversely, the use of this herbicide may be favoring other species such as E. heterophylla, C. benghalensis, and A. conyzoides, because they are less sensitive (Jowers et al. Reference Jowers, Breman and Fletcher1986; Stoller et al. Reference Stoller, Wax and Slife1979; Yadav et al. Reference Yadav, Bhullar, Kaur, Kaur and Jhala2017). This differential control favors the displacement of sensitive by less-sensitive species (Grundy et al. Reference Grundy, Mead, Bond, Clark and Burston2011). In the case of C. rotundus, this weed is poorly controlled with pendimethalin (Ahmed and Chauhan Reference Ahmed and Chauhan2015), and hand weeding has been found to be more effective than herbicides (Stoller et al. Reference Stoller, Wax and Slife1979). Therefore, the evaluated herbicide program was not the most important factor determining its presence in the community. Instead, it is possible that the higher competitive ability of the crop in these treatments may be affecting the performance of this species. Previous research demonstrated the effectiveness of crop competition reducing the growth of this species (Iqbal et al. Reference Iqbal, Ditommaso, Rehmani, Jabran, Hussain, Nasim, Fahad, Shehzad and Ali2019; Neeser et al. Reference Neeser, Aguero and Swanton1997), especially once a crop closes its canopy, a process that is enhanced by the addition of fertilizer, ultimately shading weeds (Keeley Reference Keeley1987).

The present study indicates that fertilizer use and weed control intensity were more important than rotation structure and varieties in influencing weed species community composition. Herbicides may reduce manual labor and minimize risks of yield reduction (Ogwuike et al. Reference Ogwuike, Rodenburg, Diagne, Agboh-Noameshie and Amovin-Assagba2014), but the fact that the evaluated production programs were unable to reduce weed population growth rates highlights the need for expanding the range of weed control practices. Special attention must be paid to the limited differences between maize and peanut phases of the rotation despite their contrasting morphology and growth. Therefore, there is a need for integration of more cultural practices and the design and implementation of a more diverse cropping system to favor the sustainability of weed management in Ghanaian farms (Leon and Wright Reference Leon and Wright2018; Weisberger et al. Reference Weisberger, Nichols and Liebman2019).

Acknowledgments

Appreciation is expressed to staff at the Council for Scientific and Industrial Research–Savanna Agricultural Research Institute, Tamale and Wa, Ghana; the University for Development Studies, Tamale, Ghana; and the Council for Scientific and Industrial Research–Crops Research Institute, Kumasi, Ghana, for assistance with this research.

Funding statement

This research was supported by the Office of Agriculture, Research and Policy, Bureau of Food Security, U.S. Agency for International Development, under the terms of award no. AID-ECG-A-00-07-0001 to the University of Georgia as a management entity for U.S. Feed the Future Innovation Lab for Peanut. The opinions expressed herein are those of the authors and do not necessarily reflect the views of the U.S. Agency for International Development.

Competing interests

The authors declare no conflicts of interest.

Footnotes

Associate Editor: Gulshan Mahajan, Punjab Agricultural University

References

Adesina, AA, Johnson, DE, Heinrichs, EA (1994) Rice pests in the Ivory Coast, West Africa: farmers’ perceptions and management strategies. Int J Pest Manag 40:293299 CrossRefGoogle Scholar
Adeux, G, Vieren, E, Carlesi, S, Bàrberi, P, Munier-Jolain, N, Cordeau, S (2019) Mitigating crop yield losses through weed diversity. Nat Sustain 2:10181026 CrossRefGoogle Scholar
Ahmed, S, Chauhan, BS (2015) Efficacy and phytotoxicity of different rates of oxadiargyl and pendimethalin in dry-seeded rice (Oryza sativa L.) in Bangladesh. Crop Prot 72:169174 CrossRefGoogle Scholar
Ahmed, YM, Mostafa, AMA, Reda, LA, El-Adawy, AM, Khozimy, AM (2017) Comparison between chemical and mechanical treatments for controlling peanut weeds. J Appl Plant Prot 6:913 Google Scholar
Andrade, JF, Satorre, EH, Ermácora, CM, Poggio, SL (2017) Weed communities respond to changes in the diversity of crop sequence composition and double cropping. Weed Res 57:148158 CrossRefGoogle Scholar
Bàrberi, P (2019) Ecological weed management in sub-Saharan Africa: prospects and implications on other agroecosystem services. Adv Agron 156:219264 CrossRefGoogle Scholar
Beltran, JC, Pannell, DJ, Doole, GJ (2012) Economic implications of herbicide resistance and high labour costs for management of annual barnyardgrass (Echinochloa crus-galli) in Philippine rice farming systems. Crop Prot 31:3139 CrossRefGoogle Scholar
Booth, BD, Swanton, CJ (2002) Assembly theory applied to weed communities. Weed Sci 50:213 CrossRefGoogle Scholar
Dahlin, AS, Rusinamhodzi, L (2019) Yield and labor relations of sustainable intensification options for smallholder farmers in sub-Saharan Africa: a meta-analysis. Agron Sustain Dev 39:32 CrossRefGoogle Scholar
Dzomeku, IK, Abudulai, M, Brandenburg, RL, Jordan, DL (2009) Survey of weeds and management practices in peanut (Arachis hypogaea L.) in the savanna ecology of Ghana. Peanut Sci 36:165173 CrossRefGoogle Scholar
Gaba, S, Caneill, J, Nicolardot, B, Perronne, R, Bretagnolle, V (2018) Crop competition in winter wheat has a higher potential than farming practices to regulate weeds. Ecosphere 9:e02413 CrossRefGoogle Scholar
Gaba, S, Fried, G, Kazakou, E, Chauvel, B, Navas, ML (2014) Agroecological weed control using a functional approach: a review of cropping systems diversity. Agron Sustain Dev 34:103119 CrossRefGoogle Scholar
Grandin, U (2006) PC-ORD version 5: A user-friendly toolbox for ecologists. J Veg Sci 17:843844 Google Scholar
Grundy, AC, Mead, A, Bond, W, Clark, G, Burston, S (2011) The impact of herbicide management on long-term changes in the diversity and species composition of weed populations. Weed Res 51:187200 CrossRefGoogle Scholar
Haden, VR, Duxbury, JM, DiTommaso, A, Losey, JE (2007) Weed community dynamics in the system of rice intensification (SRI) and the efficacy of mechanical cultivation and competitive rice cultivars for weed control in Indonesia. J Sustain Agric 30:56 CrossRefGoogle Scholar
Hossain, K, Timsina, J, Johnson, DE, Gathala, MK, Krupnik, TJ (2020) Multi-year weed community dynamics and rice yields as influenced by tillage, crop establishment, and weed control: implications for rice-maize rotations in the eastern Gangetic plains. Crop Prot 138:105334 CrossRefGoogle ScholarPubMed
Hyvönen, T, Salonen, J (2002) Weed species diversity and community composition in cropping practices at two intensity levels—a six-year experiment. Plant Ecol 159:7381 CrossRefGoogle Scholar
Imoloame, EO (2017) Evaluation of herbicide mixtures and manual weed control method in maize (Zea mays L.) production in the Southern Guinea agro-ecology of Nigeria. Cogent Food Agric 3:1375378 CrossRefGoogle Scholar
Iqbal, J, Ditommaso, A, Rehmani, MIA, Jabran, K, Hussain, S, Nasim, W, Fahad, S, Shehzad, MA, Ali, A (2019) Purple nutsedge (Cyperus rotundus) control through interference by summer crops. Int J Agr Bio 21:10831088 Google Scholar
IUSS Working Group WRB (2015) World Reference Base for Soil Resources 2014, update 2015. International Soil Classification System for Naming Soils and Creating Legends for Soil Maps. World Soil Resources Reports No. 106. Rome: FAO. 203 pGoogle Scholar
Jowers, HE, Breman, JW, Fletcher, JW (1986) Effects of several herbicide treatments on wild poinsettia (Euphorbia heterophylla L.) control in soybean. Proc Soil Crop Sci Soc Fla 45:115117 Google Scholar
Kankam-Boadu, I, Sarkodie-Addo, J, Amagloh, FK (2018) Profitability of maize production in the northern region of Ghana. Int J Dev Res 8:2286122869 Google Scholar
Keeley, PE (1987) Interference and interaction of purple and yellow nutsedges (Cyperus rotundus s C. esculentus) with crops. Weed Technol 1:7481 CrossRefGoogle Scholar
Kent, R, Johnson, DE, Becker, M (2001) The influences of cropping system on weed communities of rice in Côte d’Ivoire, West Africa. Agric Ecosyst Environ 87:299307 CrossRefGoogle Scholar
Kharel, P, Devkota, P, MacDonald, G, Tillman, B (2022) Evaluating herbicide programs for peanut response and weed control. Crop Forage Turfgrass Manage 8:e20191 CrossRefGoogle Scholar
Koffi, D, Kyerematen, R, Eziah, VY, Osei-Mensah, YO, Afreh-Nuamah, K, Aboagye, E, Osae, M, Meagher, RL (2020) Assessment of impacts of fall armyworm, Spodoptera frugiperda (Lepidoptera: Noctuidae) on maize production in Ghana. J Integr Pest Manag 11:17 CrossRefGoogle Scholar
Krzanowski, WJ (2000) Principles of Multivariate Analysis: A User’s Perspective. Oxford: Oxford University Press. 608 pCrossRefGoogle Scholar
Leon, RG, Wright, DL (2018) Recurrent changes of weed seed bank density and diversity in crop-livestock systems. Agron J 110:10691078 CrossRefGoogle Scholar
Magurran, AE, ed (2013) Ecological Diversity and Its Measurement. Princeton, NJ: Princeton University Press. 192 pGoogle Scholar
Manary, M, Callaghan-Gillespie, M (2018) Ready-to-use therapeutic food (RUTF) and ready-to-use supplementary food (RUSF). Stunting as an Indicator. Sight Life 32:4044 Google Scholar
McCune, B, Mefford, MJ (1999) PC-ORD. Multivariate Analysis of Ecological Data, Version 4. Gleneden Beach, OR: MjM Software Design Google Scholar
Neeser, C, Aguero, R, Swanton, CJ (1997) Incident photosynthetically active radiation as a basis for integrated management of purple nutsedge (Cyperus rotundus). Weed Sci 45:777783.CrossRefGoogle Scholar
Nkoa, R, Owen, MD, Swanton, CJ (2015) Weed abundance, distribution, diversity, and community analyses. Weed Sci 63(SP1):6490 CrossRefGoogle Scholar
Ogwuike, P, Rodenburg, J, Diagne, A, Agboh-Noameshie, AR, Amovin-Assagba, E (2014) Weed management in upland rice in sub-Saharan Africa: impact on labor and crop productivity. Food Secur 6:327337 CrossRefGoogle Scholar
Oreja, FH, Inman, MD, Jordan, DL, Leon, RG (2021) Population growth rates of weed species in response to herbicide programme intensity and their impact on weed community. Weed Res 61:509518 CrossRefGoogle Scholar
Oreja, FH, Inman, MD, Jordan, DL, Vann, M, Jennings, KM, Leon, RG (2022) Effect of cotton herbicide programs on weed population trajectories and frequency of glyphosate-resistant Palmer amaranth (Amaranthus palmeri). Weed Sci 70:587594 CrossRefGoogle Scholar
Owen, MDK (2008) Weed species shifts in glyphosate-resistant crops. Pest Manag Sci 64:377387 CrossRefGoogle Scholar
Owusu-Adjei, E, Baah-Mintah, R, Salifu, B (2017) Analysis of the groundnut value chain in Ghana. World J Agric 5:177188 Google Scholar
Pinheiro, J, Bates, D, DebRoy, S, Sarkar, D, R Core Team (2018) nlme: Linear and Nonlinear Mixed Effects Models. R Package Version 3.1–137. http://CRAN.R-project.org/package=nlme. Accessed: March 11, 2024Google Scholar
Pinke, G, Blazsek, K, Magyar, L, Nagy, K, Karácsony, P, Czúcz, B, Botta-Dukát, Z (2016) Weed species composition of conventional soyabean crops in Hungary is determined by environmental, cultural, weed management and site variables. Weed Res 56:470481 CrossRefGoogle Scholar
R Core Team (2020) R: a language and environment for statistical computing [Computer software]. R Foundation for Statistical Computing. https://www.Rproject.org Google Scholar
Rodenburg, J, Johnson, DE (2009) Weed management in rice-based cropping systems in Africa. Adv Agron 103:149218 CrossRefGoogle Scholar
Romillac, N, Piutti, S, Slezack-Deschaumes, S, Gaba, S (2023) Intraspecific functional trait variation in weeds: a strategy in response to competition with crop and weed plants. Weed Res 63:283296 CrossRefGoogle Scholar
Ryan, MR, Smith, RG, Mirsky, SB, Mortensen, DA, Seidel, R (2010) Management filters and species traits: weed community assembly in long-term organic and conventional systems. Weed Sci 58:265277 CrossRefGoogle Scholar
Salaudeen, MT, Daniya, E, Olaniyi, OM, Folorunso, TA, Bala, JA, Abdullahi, IM, Nuhu, BK, Adedigba, AP, Oluwole, BI, Bankole, AO, Macarthy, OM (2022) Phytosociological survey of weeds in irrigated maize fields in a southern Guinea savanna of Nigeria. Front Agron 4:985067 CrossRefGoogle Scholar
Schöb, C, Hortal, S, Karley, AJ, Morcillo, L, Newton, AC, Pakeman, RJ, Powell, JR, Anderson, IC, Brooker, RW (2017) Species but not genotype diversity strongly impacts the establishment of rare colonisers. Funct Ecol 31:14621470 CrossRefGoogle Scholar
Smith, RG, Gross, KL (2006) Weed community and corn yield variability in diverse management systems. Weed Sci 54:106113 CrossRefGoogle Scholar
Stoate, C, Baldi, A, Beja, P, Boatman, ND, Herzon, I, Van Doorn, A, de Snoo, GR, Rakosy, L, Ramwell, C (2009) Ecological impacts of early 21st century agricultural change in Europe—a review. J Environ Manag 91:2246 CrossRefGoogle Scholar
Stoller, EW, Wax, LM, Slife, FW (1979) Yellow nutsedge (Cyperus esculentus) competition and control in corn (Zea mays). Weed Sci 27:3237 Google Scholar
Storkey, J, Neve, P (2018) What good is weed diversity? Weed Res 58:239243 CrossRefGoogle ScholarPubMed
Swanton, CJ, Weise, SF (1991) Integrated weed management: the rationale and approach. Weed Technol 5:657663 CrossRefGoogle Scholar
Weisberger, D, Nichols, V, Liebman, M (2019) Does diversifying crop rotations suppress weeds? A meta-analysis. PLoS ONE 14:e0219847 CrossRefGoogle ScholarPubMed
Wongnaa, AC, Awunyo-Victor, D, Bakang, AJ-E (2018) Factors affecting adoption of maize production technologies: a study in Ghana. J Agric Sci 13:8199 Google Scholar
Yadav, R, Bhullar, MS, Kaur, S, Kaur, T, Jhala, AJ (2017) Weed control in conventional soybean with pendimethalin followed by imazethapyr+ imazamox/quizalofop-p-ethyl. Can J Plant Sci 97:654664 Google Scholar
Yin, L, Cai, Z, Zhong, W (2006) Changes in weed community diversity of maize crops due to long-term fertilization. Crop Prot 25:910914 CrossRefGoogle Scholar
Figure 0

Table 1. Annual rainfall and average annual mean, minimum and maximum temperatures for locations in Ghana

Figure 1

Table 2. Chemical and physical properties of soil profile (0–20cm) in experimental locations in Ghana

Figure 2

Table 3. Crops included on each rotation and years in locations in Ghana: peanut–maize and maize–peanut in Kumasi and peanut–maize–peanut and maize–maize–peanut in Tamale and Wa

Figure 3

Table 4. Input intensity and details for peanut and maize production in Kumasi, Tamale, and Wa, Ghana

Figure 4

Table 5. Active ingredient(s), rate and time of application of inputs for peanut and maize production in Kumasi, Tamale, and Wa, Ghana

Figure 5

Table 6. Planting and harvesting dates in different locations and seasons in Ghana

Figure 6

Figure 1. Richness (S) estimated during three seasons and for different input levels (high, medium, and low) in locations in Ghana: (A) Kumasi, (B) Tamale, and (C) Wa. Different uppercase letters indicate differences among seasons, and different lowercase letters among input levels according to Tukey’s HSD (P < 0.05). Errors bars represent standard error of the mean.

Figure 7

Figure 2. Shannon index (H′) in response to input level (high, medium, and low) in different seasons in (A) Kumasi, (B) Tamale, and (C) Wa, Ghana. Different uppercase letters indicate significant differences among seasons, and different lowercase letters indicate significant differences among input levels according to Tukey’s (P < 0.05). Errors bars represent standard error of the mean.

Figure 8

Figure 3. Evenness (J) estimated at different seasons and for different input levels (high, medium, and low) in locations in Ghana: (A) Kumasi, (B) Tamale, and (C) Wa. Different uppercase letters mean significant differences among seasons, and different lowercase letters mean significant differences among input levels according to Tukey’s (P < 0.05). Errors bars represent standard error of the mean.

Figure 9

Figure 4. Population growth rate (lambda or λ) estimated at different crop rotations and for different input levels (high, medium, and low) in locations in Ghana: (A) Kumasi, (B) Tamale, and (C) Wa. Different uppercase letters indicate significant differences among crop rotations and different lowercase letters indicate significant differences among input levels according to Tukey’s (P < 0.05). M, maize; P, peanut. Errors bars represent standard error of the mean.

Figure 10

Figure 5. Principal component analysis (PCA) ordination of density values registered 3 wk after planting date in the last season in Kumasi, Ghana. Vectors represent the explanatory variable and are the treatments. See Table 7 for nomenclature. Rotations are the first two numbers; crop varieties are the letter and the last two letters for the input level.

Figure 11

Table 7. (A) Abbreviations (codes) for treatments representing the explanatory variables (vectors) and (B) weed codes representing species used for a principal component analysis (PCA) as shown in Figure 5

Figure 12

Figure 6. Principal component analysis (PCA) ordination of density values registered 3 wk after planting date in the last season in Tamale, Ghana. Vectors represent the explanatory variable and are the treatments. See Table 7 for nomenclature. Rotations are the first two numbers; crop varieties are the letter and the last two letters for the input level.

Figure 13

Figure 7. Principal component analysis (PCA) ordination of density values registered 3 wk after planting date in the last season in Wa, Ghana. Vectors represent the explanatory variable and are the treatments. See Table 7 for nomenclature. Rotations are the first two numbers; crop varieties are the letter and the last two letters for the input level.