The Weed-Suppressive Ability of Summer Cover Crops in the Northern Grains Region of Australia

Abstract

Pressure is mounting on the agricultural sector to reduce reliance on herbicides for weed control leading to increased interest in the potential of cover crops to control weeds in summer fallows. The weed suppression ability of three summer cover crop species, buckwheat, millet and teff, was evaluated in field trials at two sites near Camden, NSW in 2021. Buckwheat, millet and teff reduced weed biomass by 65%, 77% and 95%, respectively at Bringelly and by 94%, 92% and 90%, respectively at Lansdowne. Following cover crop desiccation, teff residues reduced weed emergence in subsequently planted wheat by 73% and 26% at Bringelly and Lansdowne, respectively. Overall cover crops were found to be effective in suppressing weed emergence, growth, and reproductive capacity. These studies identified teff grass as an important summer crop option for the northern grains region.

1. Introduction

The summer fallow period in the northern grain-based farming systems of Australia is the time between winter crop harvest (December) and the following winter crop planting (May). This fallow period enables beneficial outcomes of the accumulation of mineral nutrients, increased soil water storage and improved soil health [1]. As the northern grains region experiences summer-dominant rainfall (Table 1), the production of winter crops such as wheat, chickpeas and canola is dependent on effective fallow management to ensure maximum soil water storage. Fifty percent of winter crop yield potential can be directly attributed to summer rainfall and summer fallow management, in particular the increase in stored soil water and N [2], making it essential for farmers to efficiently manage summer fallow lands.

Despite the potential benefits of fallow periods, there can be negative consequences when the soil is left bare over this period. This includes increased evaporation, rainfall runoff and erosion, as well as the emergence of unwanted weeds that can persist into the following cash crop. Weeds in summer fallow are a particular concern to growers in Australia’s northern grains region as they deplete the levels of available soil water and nutrients for following crops. [3]. The lack of control of summer weeds, such as flax-leaf fleabane (Conyza bonariensis (L.) Cronq.), awnless barnyard grass (Echinochloa colona (L.) Link), and common sowthistle (Sonchus oleraceus L.), in fallows has been estimated to result in annual losses of up to 1.7 M tonnes of grain yield, equivalent to over AUD 428 million [4]. Such losses make summer fallow weed control a major focal point for northern region grain producers.

Typically, summer fallow weed control has been reliant on herbicides, in particular glyphosate, which has resulted in the evolution of glyphosate resistance in 17 weed species commonly occurring in fallow and crop situations in Qld and NSW [5]. Glyphosate resistance is now occurring at high frequencies in populations of awnless barnyard grass (36%), windmill grass (Chloris truncata R.Br.) (56%), feathertop Rhodes grass (Chloris virgata. Sw.) (68%) and common sowthistle (14%) [6].

High frequencies of glyphosate resistance have led to an increasing interest in alternative weed control options and the adoption of integrated weed management techniques. These include agronomic management decisions such as enhanced crop competition (e.g., narrow row spacing and higher plant densities) and competitive crop species and cultivars as well as the introduction of cover crops to suppress weeds [7]. During a fallow, cover crops can be used to effectively control weed populations while not negatively impacting following crop yields, particularly by sustaining and improving soil water retention [8].

Cover crop is a broad term used to describe a crop that is grown for ecological benefits other than being solely a harvestable product. Such crops are often planted in periods of fallow, when the soil would otherwise be left bare, or with some crop residue (stubble) cover. For cover crops to be viable in fallow, they must have a positive impact on the farming system and/or subsequent crop [9]. There have been extensive studies on the benefits of cover crops, including their ability to stabilise soils, fix carbon, alter nutrient availability, increase agroecosystem diversity and complexity and supress weed emergence and growth [10,11,12]. The ability of cover crops to suppress the emergence and growth of weeds is a significant benefit for northern growers as they look towards new ways to manage glyphosate-resistant weeds.

Common commercially available summer crop species suitable for cover crop use in the northern region include Japanese millet (Echinochloa esculenta A. Braun), lablab (Lablab purpureus L.) and soybean (Glycine max L. Merrill) [13], with other species such as Sudan grass (Sorghum sudanense L.), forage rape (Brassica napus L.) and buckwheat (Fagopyrum esculentum Moench.) also common [14]. These options all possess similar characteristics of fast establishment rates, rapid growth and high levels of biomass production. Although the species listed above are suitable as cover crops, there is always a need for additional and potentially superior cover crop options.

Teff grass or teff (Eragrostis tef (Zuccagni) Trotter), originally from Ethiopia, is a self-pollinating annual grass with small seeds that has been grown for human consumption for centuries in African countries while emerging as a popular forage crop in recent decades [15]. Teff could potentially be a successful summer cover crop in the northern grains region as it prefers warm temperatures; has rapid establishment rates, high biomass production and organic matter building potential; and can be incorporated with relative ease into a cropping system [16]. While currently not widely grown, there have been limited studies on the use of teff as a cover crop and its weed-suppressive ability [16,17] with hope that more research into the species could help validate it as a valuable option for northern grain production systems.

The aims of this study were to (i) determine the potential for teff as a summer cover crop species for the northern grains region by comparing biomass production and weed suppression with other cover crop species and (ii) determine if there is a relationship between cover crop biomass production and weed suppression.

2. Materials and Methods

2.1. Experimental Sites

Field experiments evaluating the weed-suppressive ability of summer cover crop species were established at two sites on 18 February 2021, at Bringelly (33.9453° S, 150.6821° E) and Lansdowne (34.0215° S, 150.6647° E) farms of the University of Sydney. The field trials were established at two sites to increase the validity of the results while determining if varying factors such as soil type and background weed population impacted cover crop establishment and growth. Both sites were prepared for crop planting through the application of glyphosate (1 L ha⁻¹). The Bringelly site has a loamy soil, in contrast with Lansdowne, which has a finer, well-draining, sandy soil. Due to their proximity, both Bringelly and Lansdowne experience similar climatic conditions to that of Camden, with a predominantly summer rainfall pattern averaging 238 mm over December, January and February. This rainfall pattern, similar to that of the northern grains region, was reflected in the 2021 climate data; however, both sites experienced higher than average rainfall in March (Table 1).

Table 1. Climate of Camden, NSW. Long-term maximum and minimum temperature data are based on averages of 1971–2021, while rainfall is based on data for 1943–2021. (Source: Bureau of Meteorology) [18].

	Jan	Feb	Mar	Apr	May	Jun	Jul	Aug	Sep	Oct	Nov	Dec
Long-term climate
Mean min temp (°C)	17	17	15	11	7	5	3	3.9	7	10	13	15
Mean max temp (°C)	30	29	27	24	21	18	17	19	22	24	26	29
Total rainfall (mm)	80	101	95	64	53	65	35	41	38	62	74	57
2021 climate *
Mean min temp (°C)	16	13	15	9	8	4	3	4	6	-	-	-
Mean max temp (°C)	30	17	25	24	20	17	17	20	23	-	-	-
Total rainfall (mm)	64	85	327	14	82	45	16	42	23	-	-	-

* Months averages shown in bold indicate experiment duration.

Table 1. Climate of Camden, NSW. Long-term maximum and minimum temperature data are based on averages of 1971–2021, while rainfall is based on data for 1943–2021. (Source: Bureau of Meteorology) [18].

	Jan	Feb	Mar	Apr	May	Jun	Jul	Aug	Sep	Oct	Nov	Dec
Long-term climate
Mean min temp (°C)	17	17	15	11	7	5	3	3.9	7	10	13	15
Mean max temp (°C)	30	29	27	24	21	18	17	19	22	24	26	29
Total rainfall (mm)	80	101	95	64	53	65	35	41	38	62	74	57
2021 climate *
Mean min temp (°C)	16	13	15	9	8	4	3	4	6	-	-	-
Mean max temp (°C)	30	17	25	24	20	17	17	20	23	-	-	-
Total rainfall (mm)	64	85	327	14	82	45	16	42	23	-	-	-

* Months averages shown in bold indicate experiment duration.

2.2. Experimental Design

The experimental design at each site consists of four treatments (three cover crop species and a crop-free control-fallow), with four replicates of each treatment arranged in a randomised complete block design. This design was repeated at both sites. Buffer areas were left around the trials to minimise the risk of external weed contamination and spray drift. Each plot including buffer zones and the four treatments measured 12 m × 48 m. The three treatment cover crops used were teff grass (E. tef), buckwheat (F. esculentum) and white French millet (P. milliaceum) (Figure 1) with a control plot used to replicate a bare fallow.

Individual plots (2 × 12 m) were planted in 6 rows, 25 cm apart on 18 February 2021 with buckwheat and millet planted using a plot seeder, model Wintersteiger Plotseed XL. Because of the small seed size, teff was planted by mixing the seed with potting mixture and spreading it by hand over the designated areas. The teff plots were then run over with the plot seeder, where the disc disturbance moved the seeds into rows. Fertiliser, MAP supreme Zn at the rate of 50 kg ha⁻¹, was applied to all plots at sowing. Trial sites were irrigated with an overhead sprinkler system as required, ensuring the proper establishment and growth of the crops. Each site had a top dressing of urea fertilizers (N) applied 6 weeks after emergence.

2.3. Data Collection

Cover crop and weed emergence counts were conducted by counting crop and weed plants in 4 × 0.25 m² quadrats placed randomly in each plot three weeks after planting. At seven weeks after planting, the crop heights of three randomly selected plants were recorded in each plot using a metre ruler to measure height from the ground to the highest standing point of the crop plant. At the same time, the biomass of cover crops and any weeds present was assessed by cutting crop and weed plants at ground level in 4 × 0.25 m² quadrats in each plot. Harvested crop and weed plant samples were placed in separate pre-labelled brown paper bags and put in a dehydrator set at 70 °C for 72 h. The dried samples were then weighed to determine dry biomass production.

2.4. Wheat Crop Establishment and Weed Emergence in Cover Crop Residues

To investigate the effects of cover crop residues on following wheat crop establishment, both cover crop trials were desiccated eight weeks after cover crop planting using glyphosate (Roundup) at the rate of 1 L ha⁻¹ plus saflufenacil (Sharpen) 25 g ha⁻¹, followed by the slashing of the trial area in preparation for wheat planting. The entire trial site area at both sites were planted with wheat at a rate of 175 seed m⁻² in early June using the plot seeder, as described above. Wheat crop establishment and weed emergence counts in the trial area were collected six weeks after planting by counting plants in 4 × 0.25 m² quadrats in each plot. Any emergent plants that were not wheat, including cover crop residual plants, were recorded as weeds.

2.5. Data Analysis

All data were analysed using GenStat (18th Edition). The data were entered into MS Excel sheets for the calculation of the averages, standard deviations and standard errors of each treatment as well as the percentages of weed density and biomass suppressed by test species compared with control at both sites. Further analysis was then carried out with weed and crop biomass data sets transformed from both sites using log₁₀ and square root variates, respectively, because the Shapiro–Wilk test for normality was not initially met (Supplementary Table S1). Analysis of variance (ANOVA) was performed on transformed data to determine differences between treatments, with a significance probability level of 0.05. Treatment means were compared using a Tukey’s test (p = 0.05). The results presented for crop and weed biomass are based on backtransformed data. Between-site comparisons were few due to high levels of variations between the sites, leading to each site’s data being analysed separately, reducing external factor variability.

3. Results

3.1. Initial Crop Establishment and Weed Emergence

Cover crop treatments, regardless of species and establishment densities, reduced average weed emergence by more than 70% at both the Bringelly and Lansdowne sites (p < 0.05). Buckwheat, millet and teff reduced initial weed emergence by 71%, 80% and 84%, respectively, at Bringelly compared with the control plots. Similarly, at Lansdowne, buckwheat, millet and teff suppressed weed emergence by 92%, 76% and 85%, respectively (

Table 2. Cover crop establishment and weed densities (± standard error values) three weeks after crop planting at the Bringelly and Lansdowne trial sites. Different letters show significant differences (p < 0.05).

Site	Treatment	Crop Plants (m−2)	Weed Density (Plants m−2)
			Grass Weeds	Broadleaf Weeds	Total Weeds
Bringelly	Buckwheat	21 ± 3	124	199	323 ± 93 a
	Millet	315 ± 23	25	150	175 ± 34 a
	Teff	109 ± 25	68	155	223 ± 39 a
	Control	-	360	757	1116 ± 197 b
Lansdowne	Buckwheat	104 ± 14	10	4	14 ± 4 a
	Millet	129 ± 10	11	15	26 ± 18 a
	Teff	178 ± 18	32	10	42 ± 5 a
	Control	-	101	76	177 ± 55 b

). Weed counts at Lansdowne were relatively low, averaging 65 plants m⁻² across all treatments, compared with Bringelly, where there were on average 459 plants m⁻². Bringelly had higher densities of broadleaf weeds than grass weeds, with 47% higher counts of broadleaf weeds, while Lansdowne had similar densities of both weed types.

There was poor establishment of buckwheat at Bringelly, where plant densities were five times lower compared with the Lansdowne site. However, both millet and teff were well established at both sites (>100 plant m⁻²). Millet establishment was higher at Bringelly than at Lansdowne.

3.2. Crop and Weed Biomass

Cover crops on average reduced weed biomass by 86% (p < 0.05) compared with control across both the Lansdowne and Bringelly field sites (Figure 2A,B). At Bringelly, buckwheat and millet reduced weed biomass by 65% and 77%, respectively, compared with the control, with teff achieving weed biomass suppression of 95% (Figure 2A). Millet and teff both had relatively high biomass,1.5 and 2.9 t ha⁻¹, respectively compared with buckwheat (0.4 t ha⁻¹) at Bringelly. Despite having higher (p < 0.05) emergence counts early in the trial (Table 2), millet biomass production was relatively low (1.5 t ha⁻¹) at Bringelly site compared with Lansdowne (3.6 t ha⁻¹) (Figure 2A). Poor buckwheat crop establishment resulted in low biomass production at Bringelly (Table 2, Figure 2A). However, despite this low crop density, buckwheat reduced (p < 0.05) weed biomass by 65% compared with control (p < 0.05) (Figure 2A). Similarly, at Lansdowne, all three cover crop treatments reduced (p < 0.05) weed biomass compared with the control treatment, with buckwheat, millet and teff crops suppressing weeds by 94%, 92% and 90%, respectively (Figure 2B). Buckwheat and millet yielded 92% and 59% higher biomass, respectively, at Lansdowne than at Bringelly (Figure 1 and Figure 2). Teff biomass production (2.9 t ha⁻¹) was consistent across both sites (Figure 2), indicating that it could be a successful inclusion in farming systems across different soil types, such as those at Bringelly and Lansdowne.

3.3. Crop Height

The heights of all cover crop species were lower (p < 0.05) at Bringelly than at Lansdowne (Figure 3). Teff was taller (p < 0.05) than other cover crop treatments at Bringelly and taller (p < 0.05) than buckwheat at Lansdowne (Figure 3).

3.4. Wheat and Weed Emergence

At both Bringelly and Lansdowne, wheat emergence was consistent across all treatments, averaging 100 plants m⁻² (Figure 4). Summer-grown cover crop residues did not interfere with wheat crop establishment as there were no differences (p > 0.05) in wheat plant emergence between the cover crop and bare fallow treatments (Figure 4 and Figure 5).

At Bringelly, millet and teff cover crop residues had the lowest weed emergence compared with buckwheat and control, averaging 58 and 25 weed plants m⁻², respectively. Teff suppressed weed emergence by 74% and 73% (p < 0.05) compared with the buckwheat and control treatments, respectively (Figure 4A and Figure 5).

The Lansdowne results were less defined. Any cover crop seedlings that emerged in the wheat crop were counted as weeds, and a high number of volunteer buckwheat seedlings resulted in high weed densities (157 plants m⁻²) in these treatments. The number of weeds in the buckwheat residues were 240% higher than those in the control plots, highlighting the importance of proper cover crop desiccation prior to viable seed production. However, millet and teff had 67% and 26% less weed emergence than the control plots and 86% and 69% less than the buckwheat plots, respectively (Figure 4B).

4. Discussion

Cover crop species that established well and produced the highest levels of biomass substantially restricted the emergence and growth of weeds in field trials at Bringelly and Lansdowne. For instance, millet with higher emergence and biomass at Bringelly provided greater levels of weed suppression, whilst those crops that were less established, were less competitive and allowed more weeds to persist. These results aligned with those of Creech [19], who stated that, “A large part of successful weed suppression lies in rapid and complete cover crop establishment”. Similarly, low initial weed counts in well-established treatments, particularly teff, correlated with Brown [16], who found that teff was able to supress weed emergence even when the weed seed bank levels were considered high, which was the case at the Bringelly site. Gebrehiwot et al. [17] had similar findings on the importance of having well-established teff treatments, concluding that a teff cover crop with suitable plant establishment had 41% less overall weed cover compared with those that failed to successfully establish. Our findings, therefore, align with other studies that attribute strong initial weed suppression levels to cover crops that possess fast establishment rates and early ground cover [14,16].

The direct relationship between cover crop biomass and weed suppression rates was evident in this study, with the highest biomass-yielding crops being teff at Bringelly and buckwheat at Lansdowne; these produced the greatest restrictions on weed growth. Similarly, previous studies by Smith et al. [20] and Alonso-Ayuso et al. [12] determined that cover crop biomass is the major influence on weed suppression. These results also corresponded with the findings of Teasdale et al. [10], who described the negative correlation between crop and weed biomass in studies on a range of cover crop species and weeds.

Cover crops offer the potential for two periods of weed control, first when cover crops are actively growing and second after their termination; then, the residues act as a mulch on the soil surface and interfere with weed emergence in the following crops. In our study, the cover crops planted not only suppressed the weed emergence and growth during the summer fallow (Feb-Apr), but their residue also suppressed the emergence of weeds in the subsequently planted wheat crop in autumn (May). Teff and millet both suppressed the weed emergence in following wheat crop by more than 70% at Bringelly, a site with a high density of summer weeds (1116 ± 197 weed plants m⁻²). Cover crop mulch reduces the quantity and quality of light and is a physical barrier to weed seedling emergence [21,22], especially those of small-seeded species. Wheat crop establishment was, however, not affected by the presence of any cover crop residue or its absence (fallow control). These results have important implications for the northern grain region, where early weed control in winter crops is critical for reliable crop yields.

5. Conclusions

The strong correlation between cover crop biomass and weed growth suppression was consistent across all trials, indicating that high cover crop biomass is imperative when selecting cover crop species. Teff has the potential to become a widely incorporated cover crop in the northern region as it proved itself a fast-establishing, high biomass-producing and weed-suppressive species that performs in a range of soil types. However, more research is required to determine the mechanisms of weed suppression by different cover crops, especially studies aiming to disentangle the physical (competition) and chemical (allelopathy) nature of interference.