Weed Responses to Crop Residues Management in a Summer Maize Cropland in the North China Plain

: Crop residues management has great effects on weeds in croplands. To understand the weed responses to crop residues management and weeds impact on crop yield, a ﬁeld trial with three crop residues management strategies has been conducted in the North China Plain since 2008. Weed community composition and structure across the species, morphological types, life forms, and community levels were investigated during 2019–2020. The results show that the ﬁeld with crop residues retention signiﬁcantly decreased weed density than that in the ﬁeld with no crop residues retention. Furthermore, total crop residues retention signiﬁcantly decreased weed density than half crop residues retention. Compared with no crop residues retention, the weed aboveground dry matter in the ﬁeld with total and half crop residues retention signiﬁcantly decreased. Meanwhile, the maize grain yield signiﬁcantly increased, resulting from weeds decreased with crop residues retention on the ﬁeld. Negative correlations were found between maize grain yield and the density and aboveground dry matter of monocotyledonous weeds. These ﬁndings indicate that long term crop residues retention under conventional tillage might be an effective agronomic practice to retard weed growth. However, the mechanism of crop residues retention on weed control is still needed to research.


Introduction
Weeds in croplands is one of the main harmful factors influencing crop growth and productivity. High infestation of weeds in croplands has been regarded as the most severe threat to food production [1,2]. It was estimated that the yield potential of major crops losses is as much as 33.92% on an average globally as a result of weeds impacting [3]. To diminish the adverse effects of weeds on crop growth, farmers take various control measures, such as physical, chemical, biological, and ecological weed control ways. However, the extensive use of chemical herbicides has caused environmental pollution, biodiversity reduction, food quality degradation, and human health problems [4,5]. Therefore, environmental-friendly alternative methods should be considered.
Changes in the environment derived from crop rotation, tillage, fertilization, and weed control strategies directly affect weed growth, weed distribution, and competition over crops [6]. Studies showed that fields with corn-oats-hay rotations reduced weed seed density than continuous corn [7]. Similarly, wheat-safflower and wheat-canola rotations can reduce weed species and increase the total revenue compared with continuous wheat [8]. It was reported that tillage had a larger effect on weed density and diversity than crop rotation [9]. Weed species, species richness, and diversity indices were higher under conservation tillage than conventional tillage [10,11]. Fertilization is a vital agronomy practice to ensure food production, but it would affect weed growth. Generally, fertilizer addition, especially nitrogen fertilization, could significantly increase the biomass of the 3 of 12 in June. Other agronomy practices (sow, fertilization, irrigation, harvest, etc.) were the same for the three treatments since 2008. No herbicides have been used since 2008.

Weed Community Evaluations
The weed community was investigated in 2019-2020. There were two winter wheat seasons and two summer maize seasons during our observation period. The weed community evaluations were conducted once a year in late July in the summer maize season when most weed species arrived at their maximum growth. At each weed evaluation, four 0.25 m 2 split herbaceous quadrats (0.5 × 0.5 m) were randomly placed in the interior of each plot. All weeds presented in the quadrats were counted, sorted by species, recorded, and oven-dried (70 • C) to measure the aboveground dry matter. In addition, the plant height and grain yield of summer maize were measured and recorded during harvest.

Statistical Analysis
Weed density was calculated as the total weeds per square meter. To examine the responses of different weed types to crop residues retention, weed species in this study were sorted into one of two weed types by the different morphological types: Monocotyledons and dicotyledons. We also sorted all the weeds into two weed types (i.e., annuals and perennials) according to the weed life cycle. The relative abundance of each weed species was calculated by the ratio between the specific weed abundance and the total abundance of all the weed species. The relative abundances of weed types were summed as the relative abundances of all weed species of each weed type. Weed species richness and the Pielou evenness index were described as the weed community biodiversity. Species richness (S) was evaluated as all weed species presented in the four quadrats per plot. Pielou evenness index (J) was computed as J = (−∑ P i ln P i )/ln S, where P i is the relative abundance of weed species i and S is weed species richness.
Linear mixed-effects models with repeated measures were used to evaluate the effects of year, crop residues retention, and its interaction on weed density, weed aboveground dry matter, weed types, weed community diversity, plant height and grain yield of summer maize. Variables were log-transformed before analysis to meet homogeneity of variance assumptions when necessary. The differences among the crop residues retention treatments were compared with a Tukey's HSD test at p < 0.05 when the effect of treatment was found. Statistical analyses were performed using SPSS (version 20.0), and all figures were drawn with SigmaPlot (version 12.5).

Density and Aboveground Dry Matter of Weeds
Total weed density ranged from 9.67 ± 3.28 individuals m −2 in TR to 323.00 ± 34.59 individuals m −2 in NR in 2019 and ranged from 13.00 ± 3.21 individuals m −2 in TR to 300.33 ± 39.31 individuals m −2 in NR in 2020, respectively (Table 1). Total weed density significantly decreased with the increase of the number of crop residues (p < 0.001) ( Figure 1a, Table S2). The weed density of monocotyledons significantly reduced with crop residues retention (p < 0.05), while no differences were found for the dicotyledons density in the treatments (p > 0.05) both in 2019 and 2020 (Table 1).
Overall, aboveground dry matter of total weeds and monocotyledonous weeds decreased significantly with crop residues retention (p < 0.001) (Figure 1b, Table 2), but were unchanged by the year and the interaction of the year and crop residues retention (p > 0.05) ( Figure 1b, Table S3). However, the aboveground dry matter of dicotyledonous weeds varied significantly with the year (p < 0.001), but was unchanged by crop residues retention and the interaction of the year and crop residues retention (p > 0.05) (Table S3).   Values were means ± SE (n = 3), and numbers in parentheses referred to the percentage of the specific weed types. Different letters indicate significant differences at the 0.05 level between the treatments. Table 2. Results of one-way analysis of variance of weed aboveground dry matter (g m −2 ) of treat-  Values were means ± SE (n = 3), and numbers in parentheses referred to the percentage of the specific weed types. Different letters denote significant differences at the 0.05 level between the treatments.

Weed Community Biodiversity
The relative abundance of the different morphological types (monocotyledons and dicotyledons) is unchanged by the year, crop residues retention, and their interaction (p > 0.05) ( Figure 2a, Table S4). TR showed a lower relative abundance of monocotyledons than NR and HR in 2020 (p < 0.05). On the contrary, we found that both HR and TR showed a significantly higher relative abundance of dicotyledons than NR in 2019 (p < 0.05). Furthermore, the relative abundance of dicotyledons was higher in TR than NR and HR in 2020 (p < 0.05).
NR and HR in 2020 (p < 0.05). On the contrary, we found that both HR and TR showed a significantly higher relative abundance of dicotyledons than NR in 2019 (p < 0.05). Furthermore, the relative abundance of dicotyledons was higher in TR than NR and HR in 2020 (p < 0.05).
The relative abundance of the life cycle (annuals and perennials) was similar to the relative abundance of the morphological types. The relative abundance of annual weeds decreased, while the relative abundance of perennial weeds increased with crop residues retention ( Figure 2b), but the differences between treatments were not significant (p > 0.05) (Table S4).
Weed species richness varied significantly with the year (p = 0.001) and crop residues retention (p < 0.05) ( Figure 3a, Table S5). Pielou evenness index varied significantly with the year (p < 0.01) (Figure 3b, Table S5). On average, species richness was higher in NR than in HR and TR during the experimental period, but the difference was not significant in 2020 (p > 0.05). TR showed a significantly higher Pielou index than NR in 2020 (p < 0.05).  The relative abundance of the life cycle (annuals and perennials) was similar to the relative abundance of the morphological types. The relative abundance of annual weeds decreased, while the relative abundance of perennial weeds increased with crop residues retention ( Figure 2b), but the differences between treatments were not significant (p > 0.05) (Table S4).
Weed species richness varied significantly with the year (p = 0.001) and crop residues retention (p < 0.05) (Figure 3a, Table S5). Pielou evenness index varied significantly with the year (p < 0.01) (Figure 3b, Table S5). On average, species richness was higher in NR than in HR and TR during the experimental period, but the difference was not significant in 2020 (p > 0.05). TR showed a significantly higher Pielou index than NR in 2020 (p < 0.05).

Impact of Weeds on Crop Growth and Yield
Over the study period, the plant height and grain yield of summer maize varied by the year, crop residues retention, and their interactions (p < 0.01), except for the year and crop residues retention interaction on maize yield (p = 0.94) (Figure 4a,b, Table S6). The maize yield of this study followed the order: NR < HR < TR. Maize yield in HR and TR increased by 4.38% and 20.12% in 2019, and 3.73% and 9.89% in 2020 compared with NR, respectively.

Impact of Weeds on Crop Growth and Yield
Over the study period, the plant height and grain yield of summer maize varied by the year, crop residues retention, and their interactions (p < 0.01), except for the year and crop residues retention interaction on maize yield (p = 0.94) (Figure 4a,b, Table S6). The maize yield of this study followed the order: NR < HR < TR. Maize yield in HR and TR increased by 4.38% and 20.12% in 2019, and 3.73% and 9.89% in 2020 compared with NR, respectively.

Impact of Weeds on Crop Growth and Yield
Over the study period, the plant height and grain yield of summer maize varied by the year, crop residues retention, and their interactions (p < 0.01), except for the year and crop residues retention interaction on maize yield (p = 0.94) (Figure 4a,b, Table S6). The maize yield of this study followed the order: NR < HR < TR. Maize yield in HR and TR increased by 4.38% and 20.12% in 2019, and 3.73% and 9.89% in 2020 compared with NR, respectively.

Effects of Crop Residues Retention on Weed Density and Weed Aboveground Dry Matter
Both total weed density and weed aboveground dry matter was significantly decreased with crop residues retention in the present study (Figure 1, Tables 1 and 2), which is consistent with the findings derived from other croplands [2,11,15,18,25]. The density and aboveground dry matter of monocotyledons were higher in no crop residues retention than in crop residues retention treatments. (Tables 1 and 2). In addition, significant differences in the aboveground dry matter of monocotyledons were found under half crop residues and total crop residues retention treatments ( Table 2). This result could be explained by the reason that no crop residues retention treatments do not have a barrier of crop residues that impede the emergence and growth of fertile graminoids [26], so the weed seed bank may be increased after the spread of seed rain, eventually expanding the potential for severe weed infestation originating from seeds. These findings suggest that the effects of crop residues retention on weed emergence and growth vary among different amounts of crop residues [17,18]. Researchers pointed out that successful weed control by crop residues mulches was highly dependent on sufficient aboveground dry matter, so it was necessary to add additional crop residues when insufficient because the number of crop residues fluctuated over the years [15,27]. This study showed that the proportion of dicotyledons increased with the number of crop residues retention (Table 1), which could be attributed to the fact that graminoids with smaller seeds could be impeded to the retention of crop residues. In contrast, the growth of crop and broadleaf weeds with relatively big seeds might be unaffected [28]. The results of the present study showed that half

Effects of Crop Residues Retention on Weed Density and Weed Aboveground Dry Matter
Both total weed density and weed aboveground dry matter was significantly decreased with crop residues retention in the present study ( Figure 1, Tables 1 and 2), which is consistent with the findings derived from other croplands [2,11,15,18,25]. The density and aboveground dry matter of monocotyledons were higher in no crop residues retention than in crop residues retention treatments. (Tables 1 and 2). In addition, significant differences in the aboveground dry matter of monocotyledons were found under half crop residues and total crop residues retention treatments ( Table 2). This result could be explained by the reason that no crop residues retention treatments do not have a barrier of crop residues that impede the emergence and growth of fertile graminoids [26], so the weed seed bank may be increased after the spread of seed rain, eventually expanding the potential for severe weed infestation originating from seeds. These findings suggest that the effects of crop residues retention on weed emergence and growth vary among different amounts of crop residues [17,18]. Researchers pointed out that successful weed control by crop residues mulches was highly dependent on sufficient aboveground dry matter, so it was necessary to add additional crop residues when insufficient because the number of crop residues fluctuated over the years [15,27]. This study showed that the proportion of dicotyledons increased with the number of crop residues retention (Table 1), which could be attributed to the fact that graminoids with smaller seeds could be impeded to the retention of crop residues. In contrast, the growth of crop and broadleaf weeds with relatively big seeds might be unaffected [28]. The results of the present study showed that half crop residues retention (3500 Kg ha −1 wheat straws and 3750 Kg ha −1 maize stalks every year) could suppress weeds in the maize fields, so we roughly estimated that 3500 Kg ha −1 wheat straws and 3750 Kg ha −1 maize stalks retention every year might be sufficient in suppressing weeds in the winter wheat-summer maize annual double-cropping systems.

Effects of Crop Residues Retention on the Weed Community Diversity
The present study demonstrated that the shift in the composition of weed types with different crop residues management. Total crop residues retention reduced the relative abundance of monocotyledons and elevated dicotyledons for the morphological types in 2020 (Figure 2a). Overall, crop residues retention increased the relative abundance of dicotyledons, indicating that broadleaf dicotyledons may be more adaptive to the residues retention environment than monocotyledons for diverse root systems [29,30]. Previous studies found the broadleaf weeds were species with taproots, whereas the graminoids were the shallower-rooted and fibrous-rooted species [31]. Although both graminoids and broadleaf weeds were suppressed by the crop residues retention (Table 1), the species with shallower or fibrous roots may be more sensitive to the allelochemicals released by the crop residues, leading to the lower relative abundance of graminoids. Meanwhile, the taproot broadleaf weeds can stretch their roots in deep soil layers to utilizing deeper water and nutrients and avoiding being seriously affected by exudates from the crop residues, which facilitates their survival and growth in the crop residues retention conditions [32]. That would, in turn, contribute to increasing the percentage of broadleaf weeds. In addition, many studies found crop residues retention could enhance the soil water content [33]. This study revealed the relative abundance of annual weeds decreased with crop residues retention (Figure 2b), suggesting that soil water content reduction favored annual plant growth [34]. This result could be attributed to the fact that annual weeds may be rapid growth species with a short life span, growing and reproducing faster than perennials when encountering disadvantageous conditions like drought [35]. Overall, the present study found contrasting results of the relative abundance of weed types under different crop residues management, indicating that different weed types may have different adaptation strategies associated with their morphological or physiological traits to adapt to environmental changes [36].
Species richness in the present study showed an apparent decrease with crop residues retention during the observation period (Figure 3a), i.e., crop residues retention significantly decreased species richness of the weed community, consistent with other studies [2,37]. These findings suggest that long term crop residues retention may reduce the weed species richness in the summer maize field. Crop residues could hinder the emergence of seeds by decreasing the light and providing a surviving habitat to pathogens and weed predators [2,26]. Consequently, plots with crop residues could decrease weed species richness. In the present study, no difference in the Pielou evenness index was found under treatments (Figure 3b), implying that crop residues retention might not be the dominant factor affecting the Pielou evenness index. Species richness and Pielou evenness index in this summer maize field varied significantly with the year (Figure 3a,b, Table S4). This could be attributed to the high germination failure rate and seedling mortality rate, due to different annual soil temperatures and precipitations during the observation period ( Figures S1 and S2). The present study found an interesting result that the Pielou evenness index in no crop residues retention treatments was slightly lower than that of crop residues retention ones, contrasting with species richness. Although no crop residues retention treatments showed more weed species, the graminoids accounted for more than 90 percent of the total weed density (Table 1), so the Pielou evenness index was lower in no crop residues retention treatments. These findings imply that even small changes in the environment may impact the resistance of the weed community and lead to alternations in weed species composition.

Effects of Weeds on Maize Growth and Yield
This study showed that the presence of monocotyledons, favoured by no crop residues retention conditions, negatively affected summer maize grain yield ( Figure 5). This may be attributed to the fact that crop residues retention could improve soil health status, provide nutrients (C, N fractions, and other elements), and optimize soil conditions [38][39][40], facilitating maize growth. Moreover, crop residues could impede weed emergence and occurrence by changing the soil conditions [2,11], decreasing the weed competition over crops. Although summer maize grain yield is not dominated by an increase in monocotyledonous weeds solely, it should be noted that monocotyledonous weeds like Digitaria sanguinalis (L.) Scop was abundant species in this study, especially in no crop residues retention treatment. Whenever a weed becomes a dominant or troublesome species in a farming field, farmers should alter weed control strategies to mitigate the high infestation of certain weeds [41]. In the present study, the weed management practices of tillage and crop residues retention with no chemical herbicides were used for more than ten years according to the original experiment design, which may benefit certain weeds. The present study showed that crop residues retention caused a significant shift in the weed community, suggesting flexible weed control strategies should be considered in the future according to the weed flora. Previous studies showed that integrated weed management could efficiently control weeds in croplands. Specifically, compared with the chemical approach, mechanical weeding after applied atrazine as a pre-emergence herbicide reduced weed density [42,43]. Integrated weed management in no-till systems may make it possible to reduce herbicides in croplands [44]. Moreover, no-till has been popular in recent decades for many benefits to the cropland environment. Many studies observed that reduced or zero tillage increased weed density and weed aboveground biomass [24,45]. In addition, strip-tillage with crop mulch could decrease the size, but increase the diversity of the soil seed bank when comparing with conventional tillage with no crop mulching [46]. Therefore, further studies with long term observation are needed in the future to determine the mechanisms of integrated weed management related to the soil physicochemical properties, nutrient uptake mechanisms of both crops and weeds, soil microbial community, and soil enzyme activities on the weed community composition and structure.

Conclusions
Long term crop residues retention in the field under conventional tillage could change the weed community. Comparing with no crop residues retention from the field, the weed density in the field with total crop residues retention and half crop residues retention were decreased by 96.34% and 68.37%, respectively. The weed aboveground dry matter in the field with total and half crop residues retention decreased by 95.52% and 59.01%, respectively. Furthermore, maize yield was increased by 4.06% and 15.01% in half and total crop residues retention, due to weeds decreased. Negative correlations were found between grain yield and the density and aboveground dry matter of monocotyledonous weeds. Therefore, we suggest that crop residues retention under conventional tillage is an effective agronomic measure to impede weeds and improve crop yield.
Supplementary Materials: The following are available online at https://www.mdpi.com/article/ 10.3390/agriculture11080746/s1, Figure S1: Soil temperature in 5 cm depth in 2019 (a) and 2020 (b) under treatments (NR, no crop residues retention; HR, half crop residues retention; TR, total crop residues retention) during the study period, Figure S2, Soil water content in 2019 (a) and 2020 (b) under treatments (NR, no crop residues retention; HR, half crop residues retention; TR, total crop residues retention) during the study period, Table S1: List of weed species and their classification on plant functional groups under different crop residues management (NR, no crop residues retention, HR, half crop residues retention, TR, total crop residues retention) in the summer maize field in NCP during the study period, Table S2: Significance of the impacts of crop residues retention, year, and their interactions on weed density during the period of study, Table S3: Significance of the impacts of crop residues retention, year, and their interactions on aboveground dry matter of weeds during the period of study, Table S4: Significance of the impacts of crop residues retention, year, and their interactions on the relative abundance of weed types during the period of study, Table S5: Significance of the impacts of crop residues retention, year, and their interactions on the weed community-level diversity during the period of study, Table S6: Significance of the impacts of crop residues retention, year, and their interactions on the plant height and crop yield of maize during the period of study. Raw data: The data of this study.