Introduction
Use of synthetic nitrogenous fertilizers has improved crop yield but has also resulted in soil acidification, that is, reduced soil pH below 5.5 in the surface horizons of 0 to 20 cm (Li et al. Reference Li, Cui, Chang and Zhang2019). Estimates suggest that 30% of global land surface is affected by acidity (Dai et al. Reference Dai, Zhang, Tang, Muhammad, Wu, Brookes and Xu2017). Soil acidity reduces plant growth by increasing the availability of toxic mineral elements, including aluminum and ferric iron, and decreasing availability of essential elements like calcium, magnesium, phosphorus, sulfur, molybdenum, and nitrogen (Li et al. Reference Li, Cui, Chang and Zhang2019; Moore Reference Moore2001; Sumner and Noble Reference Sumner and Noble2003). Application of lime is the most commonly used management practice to increase soil pH and improve crop production in acidic soils (Li et al. Reference Li, Cui, Chang and Zhang2019; Moore Reference Moore2001). Increasing pH can significantly increase yield of most crops grown in Western Australia, including wheat (Triticum aestivum L.), barley (Hordeum vulgare L.), canola (Brassica napus L.), legume species, and pasture legume species (Li et al. Reference Li, Cui, Chang and Zhang2019; Moore Reference Moore2001). The prevalence of soil acidity and interaction of soil amelioration with other aspects of agronomic management is a significant issue for grain production. The interactions between soil pH and nutrient supply are relatively well understood (Slattery et al. Reference Slattery, Conyers, Aitken, Peverill, Sparrow and Reuter1999). There has also been research into the relationship of liming and herbicide application (Aladesanwa and Akinbobola Reference Aladesanwa and Akinbobola2008; Hashem and Borger Reference Hashem and Borger2018). However there has been limited investigation into the interactions of pH, soil amelioration (such as lime application or incorporation), and weed management.
The most prominent weed in Australia is rigid ryegrass (Lolium rigidum Gaudin) (D’Emden and Llewellyn Reference D’Emden and Llewellyn2006; Llewellyn et al. Reference Llewellyn, Ronning, Clarke, Mayfield, Walker and Ouzman2016). Growth of this species is favored by widespread herbicide resistance and adoption of conservation tillage (D’Emden and Llewellyn Reference D’Emden and Llewellyn2006). Conservation tillage has less soil disturbance, reducing physical weed control and the effectiveness of PRE herbicides that require soil incorporation (D’Emden and Llewellyn Reference D’Emden and Llewellyn2006). Liming may influence the growth of this weed, although the soil pH preference of L. rigidum is unknown. Application of lime to acidic soils increased the dry biomass production of other Lolium species, including Italian ryegrass [Lolium perenne L. ssp. multiflorum (Lam.) Husnot ‘Aristocrat’) and perennial L. perenne (Bolland et al. Reference Bolland, Rengel, Paszkudzka-Baizert and Osborne2001; Meharg and Killham Reference Meharg and Killham1990). By comparison, applying lime to a subterranean clover (Trifolium subterraneum L.) and L. rigidum pasture on acidic soil did not increase L. rigidum biomass (Hochman et al. Reference Hochman, Osborne, Taylor and Cullis1990). However, increasing soil pH increases wheat growth, and a competitive crop is a highly effective weed control tactic (Lemerle et al. Reference Lemerle, Verbeek and Coombes1995; Li et al. Reference Li, Cui, Chang and Zhang2019). Lolium rigidum has been shown to be highly sensitive to crop competition (Borger et al. Reference Borger, Hashem and Powles2016a, Reference Borger, Riethmuller and D’Antuono2016b). Preliminary research indicated that lime application and subsequent increase in soil pH would reduce weed growth (Gazey and Andrew Reference Gazey and Andrew2010; Hashem and Borger Reference Hashem and Borger2018).
Incorporation of lime may also affect weed growth. An international review of liming experiments indicated that the application method of lime (surface, plow, subsoiling) did not affect the change in soil pH achieved through liming, although the review did not specifically consider varying soil depths (Li et al. Reference Li, Cui, Chang and Zhang2019). Rainfall can be low in Australian agricultural areas, and both field and pot studies indicate that in weathered soils, lime will not move farther than 2.5 cm through the soil profile without incorporation (Nunes et al. Reference Nunes, Denardin, Vaz, Karlen and Cambardella2019). Lime incorporation through cultivation will bury weed seed, changing the emergence patterns of the weed in subsequent seasons (Chauhan et al. Reference Chauhan, Gill and Preston2006a). While the impact of changes to the vertical distribution of weed seed in the soil are not well understood, seedling recruitment is usually increased if seed is buried near the soil surface and decreased where seed is buried at depth (Chauhan et al. Reference Chauhan, Gill and Preston2006a). Soil amelioration to incorporate lime is done with a wide range of implements, but aims to disrupt soil to a depth of at least 10 to 20 cm and up to 60 to 70 cm depending on soil type (Pluske et al. Reference Pluske, Boggs and Leopold2017). It is likely that cultivation to depth to incorporate lime will bury a proportion of weed seed too deeply for emergence, reducing seedling recruitment in subsequent years.
There is little research on soil acidity and crop–weed competition or the impact of incorporation of lime (through cultivation) on L. rigidum growth. We tested the hypotheses that (1) application of lime would increase the initial growth of L. rigidum and wheat (before the impact of crop weed competition), (2) application of lime to acidic soils would reduce biomass and seed production of L. rigidum within a wheat crop due to improved competitive ability of the crop, and (3) incorporation of lime would not affect recruitment of L. rigidum seeds in a pot experiment at a uniform depth but would reduce seedling recruitment in the field. These hypotheses were tested through field experiments examining the long-term impact of lime application and incorporation on L. rigidum density, biomass, and seed production in wheat crops, and a pot experiment under controlled conditions that examined the impact of lime application and incorporation on early growth of L. rigidum seedlings in a wheat crop.
Materials and Methods
Screen House Experiment
A pot experiment was established with treatments of lime at 0, 500, 1,000, 1,500, and 2,000 kg ha−1, on the surface of the soil or fully incorporated, in a randomized block design with three replications. Lime rates were low compared with rates used in the field, because meta-analysis indicated that the overall effect of liming on soil pH was 36% greater in pot conditions than in the field (Li et al. Reference Li, Cui, Chang and Zhang2019). Soil for the experiment (Yellow orthic acidic Tenosol soil, naturally acidic to a depth of over 1 m; Isbell Reference Isbell2016), was obtained from the Department of Primary Industries and Regional Development Merredin Research Station (31.482°S, 118.215°E). Soil was collected on August 31, 2018, from the A1 horizon (0 to 10 cm; see Table 1 for soil properties) and B1 horizon (20 to 40 cm). The A2 horizon (10 to 20 cm) was removed, because it is very thin in this soil type and could not be collected in bulk without significant contamination from the other soil horizons. On September 3, 2018, soil was sieved to 4 mm to remove crop residue and rocks while retaining soil microaggregates (Augé et al. Reference Augé, Stodola, Arnold and Saxton2001) For the non-incorporated treatment, pots of 2.5-kg capacity (20-cm height) were lined with paper towel and filled with 2 kg of soil from the B1 horizon and 0.5 kg of soil from A1 layered on top to artificially create the separate horizons. Lime (neutralizing value of 94.9%, 99.2% particles <0.5mm) was manually spread on the surface. For the incorporated treatments, soil from the A1 and B1 horizon and lime at the various rates were mixed in a rotary mixer and 2.5 kg of the resulting soil was placed in each pot. All pots were packed to a bulk density of 1.5 g cm−3, as this was the average bulk density of the soil in the Merredin experiment discussed later. The soil was allowed to absorb water through capillary action (to avoid soil compaction resulting from overhead watering), by placing pots in trays containing 4 cm of water for 3 h. Fifteen holes of 3-mm width were punched in the soil of each pot. Five seeds of wheat (‘Mace’) were sown at 2-cm depth (i.e., 60 kg ha−1 of wheat), five seeds of L. rigidum at 0.5-cm depth (i.e., 200 plants m−2), and fertilizer (80 kg ha−1 K-Start10 Trace, 12.0 N:13.1 P:10.0 K:3.0 S:0.1 Cu:0.2 Zn w/w%, bulk density 1.1, Landmark, 16 Yilgarn Avenue, Northam, WA 6401, Australia) was placed in the final five holes at 4-cm depth. Pots were maintained on benches in a screen house and natural rainfall was supplemented with 6 mm of irrigated water every 3 d.
Table 1. The pH and aluminum (Al) of the A1 and B1 horizon in the non-incorporated pots and in the single horizon in the incorporated pots, as well as dry biomass of the wheat and Lolium rigidum shoots and roots, and plant density, averaged over the incorporated and non-incorporated treatments, at varying rates of lime (0 to 2,000 kg ha−1).a
a The P-values indicate the significance of the linear contrast of the lime treatment, or the significance of the lime treatment in the case of plant density, with standard error of the mean (SEM) and LSD.
Plant density was assessed on October 4, 2018, when plants reached the 2- to 4-leaf stage. A single soil sample was taken from the pots with incorporated soil and two soil samples (from the A1 and B2 horizon) were taken from the pots with no incorporation. Soil samples were used to assess pHCaCl2 and extractable aluminum (Bromfield Reference Bromfield1987; Rayment and Lyons Reference Rayment and Lyons2011). Bulk root and shoot biomass of wheat and L. rigidum were harvested from each pot by washing away the remaining soil. The plant material was dried at 60 C for 3 d and weighed to determine dry biomass.
An exponential model was fit to the pH-extractable aluminum concentration data using Genstat (VSN International 2019). A linear model was used to compare wheat root or shoot dry biomass with that of L. rigidum. An ANOVA was used to assess lime rate, incorporation, and the interaction between these treatment factors on pH, aluminum, plant density, and dry root and shoot biomass. Residual plots indicated the data were normally distributed. A linear contrast was applied to the lime factor, and means are presented with standard errors of the means (SEMs). Fisher’s protected LSD was used to compare means (with significance at 0.05).
Merredin Experiment
A field experiment was conducted from 2016 to 2018 at the Merredin Research Station. The site was adjacent to the area used to collect soil for the screen house experiment, with the same soil type. The experiment was a split-split-plot design with four replicates and a plot size of 20 m by 1.54 m. The main plot factor was crop rotation; continuous wheat (2016 to 2018) or wheat–chemical fallow rotation (i.e., wheat in 2016, chemical fallow in 2017, and wheat in 2018). The subplot factor was cultivation (before seeding in 2016) to incorporate lime with offset disks to a depth of 15 cm; with soil testing to confirm that cultivation reached 15 cm. The lime (neutralizing value of 90.0%, 99.0% particles <0.6mm) at 0, 2,000, 4,000, or 6,000 kg ha−1 was sub-subplot factor, applied directly before cultivation in 2016.
In each year wheat (Mace) was sown at 60 kg ha−1, with 22-cm row spacing, at a depth of 3 cm, using a no-tillage seeding system (knife points and press wheels), with 60 kg ha−1 of fertilizer (Agras, 16.1:9.1:14.3:0.5:0.06% N:P:S:Ca:Zn, CSBP, Goldfields Rd, Merredin, WA 6415, Australia) banded 4 cm below the seed. The fallow was maintained by spraying a nonselective, nonresidual herbicide (paraquat/diquat, 270/230 g ai ha−1, Spray. Seed®, 135/115 g ai L−1, SL, Syngenta, North Ryde?, NSW, Australia) as required to ensure no weed growth. In 2018, sowing was on June 1, 2018, directly after application of paraquat/diquat (270/230 g ai ha−1), pyroxasulfone (102 g ai ha−1; Sakura® 850WG, 850 g ai kg−1, WG, Bayer Crop Science, Bentley, WA, Australia), and trifluralin (480 g ai ha−1; Trifluralin, 480 g ai L−1, EC, Nufarm, Kwinana Beach, WA, Australia?). Nitrogen fertilizer (Flexi-N, 32% N, CSBP) was applied at 50 L ha−1 on July 27, 2018 and harvest occurred on December 3, 2018.
Lolium rigidum and wheat density were assessed on July 12, 2018, from two 50 by 50 cm quadrats per plot (i.e., two 50-cm rows of wheat per quadrat). On October 10, 2018, wheat tillers were counted in two quadrats of 50 by 50 cm, and then L. rigidum biomass was harvested from the same quadrats by cutting plants off at ground level. Biomass samples were dried at 60 C for 3 d and weighed, and the entire sample was threshed. A subsample of approximately 5 g was taken from each sample to count L. rigidum seeds, and seed number per subsample was used to determine seed production per meter. Wheat grain yield was assessed by harvesting the entire plot. A subsample was taken from each plot and used to assess clean yield (with 2-mm sieves used to remove chaff and screenings).
The continuous wheat and wheat–fallow rotations were analyzed separately because of an order of magnitude difference in mean and variability of L. rigidum density, biomass, and seed production. Differences between rotations were examined based on whole-plot results and analyzed by t-test allowing for unequal variances (VSN International 2019). For wheat density, tillers, and yield, a paired t-test was used, as the variances were similar between rotations. The ANOVA applied to the separate crop rotation data sets used incorporation as the main plot factor and lime as the subplot factor. The variates included L. rigidum density, biomass, seed production, and crop density, tiller number, and yield. A linear contrast was applied to the lime factor. All means are presented with SEMs, and LSDs (0.05) are presented to separate the means. Residual plots indicated that the L. rigidum density, biomass, and seed production variates were not normally distributed, and a square- or cube-root transformation was performed, with data presented as back-transformed means. A complication of back-transforming is that the LSD values on the original scale vary depending on which treatments are being compared, and so in situations of more than two treatment factors, an LSD range was calculated. The LSD from the ANOVA was subtracted from the largest mean (x 1 − LSD = a) and added to the smallest mean (x 2 + LSD = b) before transformation. Following back-transformation of these two values (a and b) and the means (x 1 and x 2), the difference between the largest mean and value a or the smallest mean and value b are presented as the maximum and minimum LSD values. To apply this LSD range, the minimum LSD value is relevant for comparisons to the smallest means and the maximum LSD for comparisons to the largest means. In a similar way the SEM from the ANOVA was added and subtracted to each mean before transformation, and then the SEM was calculated as the average of the lower and upper SEMs. While the SEM from the ANOVA is a single pooled estimate, the back-transformation process results in different SEM values for the various factor levels of each treatment.
Wongan Hills Experiment
The Wongan Hills experiment was established in 1994, also on a Yellow orthic acidic Tenosol soil (30.854°S, 116.741°E) (Isbell Reference Isbell2016). Soil pHCaCl was 5.0 at 0 to 10 cm and 4.2 at 10 to 30 cm. At the same depths, aluminum concentration was <1.0 mg kg−1 and 7 mg kg−1. This soil is not naturally acidic like the Merredin soil. Low pH values resulting from agricultural practices extend to a depth of 40 cm; the B2 horizon beyond 40 cm is not acidic. In 1994, the experiment was a completely randomized block design with four replications (paired plots of 1.8 m by 30 m), arranged in two banks of two blocks each. The treatment factor was lime (neutralizing value of 94.9%, 99.2% particles <0.5mm) applied at 0, 500, 1,000, 2,000, and 4,000 kg ha−1. In 1998, the design was changed to a split-plot design with four replications (single plots of 1.8 m by 30 m) by the addition of 0 or 1,500 kg ha−1 of lime to one of each pair of plots. In 2014, the design changed to a split-split-plot design with four replications (plots of 1.8 m by 15 m) by the addition of 0 or 3,000 kg ha−1 of lime to each plot. In 2018, each plot was split into three (plots of 1.8 m by 5 m) when a rotary hoe was run perpendicular to the original plots to incorporate lime to a depth of 0, 15, or 25 cm before seeding.
On May 29, 2018, wheat (Mace) was sown at 80 kg ha−1, with 22-cm row spacing, at a depth of 3 cm, using a zero-tillage seeding system (coulters, knife points, and press wheels), with 80 kg ha−1 of fertilizer (Macro Pro Extra, 9.7:11.2:11.2:10.2:0.1:0.2% N:P:K:S:Cu:Zn, CSBP) banded 4 cm below the seed. Weeds were controlled over summer (i.e., in February), but no herbicide was applied at seeding due to dry conditions. Nitrogen fertilizer (Flexi-N, 32% N, CSBP) was applied at 50 L ha−1 on July 18, 2018, and harvest occurred on November 29, 2018. On June 29, 2018, crop density was assessed in two 50 by 50 cm quadrats per plot (i.e., two 50-cm lines of wheat per quadrat). Initial L. rigidum density was low (0 to 5 plants m−2), but more L. rigidum cohorts emerged as the growing season progressed. On October 19, 2018, L. rigidum density and tillers were recorded in two 50 by 50 cm quadrats per plot, before harvesting L. rigidum biomass from the same quadrats. Note that crop and L. rigidum were not assessed in every plot; treatments of interest were those plots that had 0, 2,000, or 4,000 kg ha−1 of lime in 1994. Lolium rigidum samples were processed, and clean yield was determined as described for the Merredin field experiment. An additional 25% of the data were excluded from the analysis of crop yield because plots were impacted by wheel tracks. The single wheel track evident on some plots appeared after measurements of wheat and L. rigidum density were performed. The track only affected one row of wheat, and so wheat tillers and L. rigidum biomass measurements could still be taken from those plots. The 25% of affected plots were evenly spaced though the trial.
ANOVA was used to assess L. rigidum density, biomass, seed production, and crop yield variates with blocking and treatment structures used that matched the design (VSN International 2019). The main effect of tillage was assessed separately to account for the strips going across the experiment. Residual plots indicated that the L. rigidum plant density, biomass, and seed production data were not normally distributed, and a cube-root transformation was performed to improve the distribution. Back-transformed means, SEM, and LSD were determined using the process described for the Merredin field experiment.
Results and Discussion
Lime and Soil Incorporation in Controlled Conditions
Lime rates of 0 to 2,000 kg ha−1 increased soil pH from 4.2 to 6.6 in pots with full incorporation (Table 1). Pots with no incorporation had pH values of 5.0 to 7.0 in the A1 horizon, and a consistent pH of 3.9 in the B1 horizon. Extractable aluminum concentration increased rapidly once pH was below 4.3 (Table 1; Figure 1). As a result, aluminum concentration was only high in the B1 horizon of the non-incorporated pots, where pH remained consistently low (Table 1). These results supports the findings of Nunes et al. (Reference Nunes, Denardin, Vaz, Karlen and Cambardella2019) that incorporation is required to change soil pH at depth. Australian soils can be highly acidic to depths of 40 to 100 cm, naturally or due to agronomic practices (Isbell Reference Isbell2016). It is likely that tillage is required to change the pH at a depth beyond 0 to 20 cm, following application of lime (Nunes et al. Reference Nunes, Denardin, Vaz, Karlen and Cambardella2019).
Figure 1. The pH and aluminum concentration in each pot. The dotted line indicates the exponential regression. Al = 0.210 + 13004017e−3.448 * pH (R2 = 97.1, P < 0.001).
Increasing rates of lime increased the above- and belowground dry biomass of wheat and L. rigidum, with the biomass of L. rigidum roots or shoots increasing in those pots where biomass of wheat increased, but did not affect plant density (Table 1; Figure 2). This finding supports the hypothesis that the growth of L. rigidum, like that of wheat and L. perenne, is favored by increased soil pH (Bolland et al. Reference Bolland, Rengel, Paszkudzka-Baizert and Osborne2001; Meharg and Killham Reference Meharg and Killham1990). In the current study, L. rigidum plants were deliberately grown in a simulated cropping system (to create an environment similar to that in the field) and then harvested at the seedling stage when the plants were too young for biomass to be affected by inter- or intraspecies competition. However, further research on plants grown to maturity is required to determine optimal pH requirements for L. rigidum and other common weeds. If optimal soil pH for weed species is not identical to that of crops, there may be scope to manipulate soil pH to reduce weed competitive ability.
Figure 2. The dry biomass of Lolium rigidum roots compared with that of wheat roots (left, where the line indicates the relationship y = 0.0351x – 0.0063, R2 = 41.8, P < 0.001) and the dry biomass of Lolium rigidum shoots compared with that of wheat shoots (right, where the line indicates the relationship y = 0.0352x – 0.0031, R2 = 23.2, P = 0.004).
The incorporation treatment significantly reduced L. rigidum density compared with the non-incorporated treatment (2.73 and 4.07 seedlings per pot, P = 0.009, LSD = 0.95), and also reduced dry weight of L. rigidum roots (0.0069g for incorporated and 0.0231 g for non-incorporated, P < 0.001, LSD = 0.0051) and shoots (0.0053 g for incorporated and 0.0245 g for non-incorporated, P ≤ 0.001, LSD = 0.0056). By comparison, incorporation had no significant effect on wheat density compared with the non-incorporation treatment (average of 4.9 wheat seedlings per pot), but still reduced the dry weight of wheat roots (0.487g for incorporated and 0.724 g for non-incorporated, P ≤ 0.001, LSD = 0.121) and shoots (0.442 g for incorporated and 0.584 g for non-incorporated, P ≤ 0.012, LSD = 0.106). There were no significant interactions between lime and incorporation for plant density or biomass. The reduced biomass of L. rigidum shoots and roots may have resulted from reduced plant number or a combination of reduced plant number and reduced plant growth, as observed for the wheat seedlings. Because reduced growth of wheat and reduced emergence of L. rigidum were apparent in both the limed and unlimed pots, and there was no herbicide used in the screen house, the change to early wheat growth and L. rigidum emergence was potentially related to altered soil properties following incorporation. For example, the mixing (incorporation) process may have reduced soil aggregate size, which can influence emergence and root growth of some species (Braunack and Dexter Reference Braunack and Dexter1989). We do not fully understand the interaction of incorporation and lime on weed emergence and early plant growth. Because plants were harvested early in this study, and L. rigidum seeds that failed to emerge were not retrieved and tested for viability, we did not determine whether seeds germinated and failed to emerge, or whether incorporation induced short- or long-term seed dormancy. If weeds germinated and failed to emerge, this would improve weed control. However, if dormancy was induced, seeds are likely to emerge later in the season or in subsequent years. If seedling recruitment is delayed by incorporation, the late seedlings will escape exposure to nonselective or PRE herbicides (Storrie Reference Storrie2014). Delayed recruitment of L. rigidum seedlings following incorporation may reduce their competitive ability in the crop (Pannell et al. Reference Pannell, Stewart, Bennett, Monjardino, Schmidt and Powles2004). However, this would not be an advantage if early growth of the crop (and subsequent competitive ability of the crop) is also reduced by incorporation. The current experiment was conducted in controlled conditions, and the incorporation achieved in the pots was not equivalent to field conditions. Even the non-incorporated soil in the pots experienced considerable disturbance being transferred from the field to pots. It is clear that we need more research on plant germination, emergence, and growth following soil incorporation in field conditions to determine whether reduced emergence is a consistent result of soil incorporation.
Wheat–Fallow Compared with Continuous Wheat at Merredin
At Merredin, L. rigidum density, biomass, and seed production were lower in the wheat–fallow rotation compared with the continuous wheat rotation (Table 2). Within the wheat–fallow rotation there were no effects of lime rate or incorporation on each L. rigidum variate, due to uniformly low weed density. Wheat density and tiller number were not affect by rotation, although the fallow rotation increased wheat yield. While a fallow is a highly effective weed control technique, it is not economically beneficial, except where weed density is exceptionally high (Monjardino et al. Reference Monjardino, Pannell and Powles2004). However, fallow has additional benefits, such as stored soil moisture and increased nitrogen mineralization, and may be a profitable rotation choice for a low-rainfall area like Merredin (Oliver and Sands Reference Oliver and Sands2013).
Table 2. The Lolium rigidum density, biomass, and seed production and wheat tillers and yield for the continuous wheat rotation compared with the wheat–fallow rotation at Merredin in 2018.a
a Means (and standard errors of the means [SEMs]) are based on the raw data. Significance is indicated by the P-values.
Lime and Incorporation at Merredin and Wongan Hills
Within the continuous wheat rotation at Merredin, increasing rates of lime were linearly correlated to reduced density, biomass, and seed production of L. rigidum and increased wheat tiller number and yield (Table 3). For each variate, the deviations from the linear relationship were not significant. Incorporation of lime in the continuous wheat rotation reduced L. rigidum density, but the reduction to biomass and seed production was not significant (Table 4). Incorporation also did not affect wheat tiller number or yield. Wheat density averaged 109 plants m−2 and was not affected by treatments (data not shown).
Table 3. The Lolium rigidum density, biomass, and seed production and wheat tillers and yield at lime rates of 0 to 6,000 kg ha−1 for the continuous wheat rotation at Merredin in 2018.a
a The LSD and P-values indicate the significance of the linear contrast of lime rate, with standard errors of the means (SEMs) presented in parentheses. Due to back-transformation of density, biomass, and seed production, SEMs vary between means, and LSD is presented as a range.
Table 4. Lolium rigidum density, biomass, and seed production and wheat tillers and yield following initial incorporation or no incorporation of lime in 2016 in a continuous wheat cropping system at Merredin in 2018.a
a The P-values and LSD values indicate the significance of the incorporation treatment and standard errors of the means (SEMs; presented in parentheses) vary between treatment levels of density, biomass, and seed production due to back-transformation.
At Wongan Hills, L. rigidum density, biomass, and seed production in the 2018 crop were reduced with application of lime in 1994 and 2014, but the reduction in L. rigidum density and seed production due to the 1998 lime application was not significant (Table 5). Wheat density was not affected by lime, but tiller number increased with increasing rates of lime applied in 1994 and 2014. The slight increase to wheat yield following application of lime was not significant in 1994, 1998, or 2014. Incorporation of lime in 2018 did not affect L. rigidum or wheat. However, the interaction between lime application in 2014 and incorporation of lime in 2018 was significant for L. rigidum. Lolium rigidum density, biomass, and seed production decreased with increasing depth of incorporation in those plots where lime was not applied in 2014 (0 kg ha−1), but was not significantly affected in those plots where 3,000 kg ha−1 of lime was applied in 2014 (Table 6).
Table 5. Lolium rigidum density, biomass, and seed production and wheat tillers and yield at Wongan Hills in 2018, following application of lime in 1994 (0, 2,000, or 4,000 kg ha−1), 1998 (0 or 1,500 kg ha−1), and 2014 (0 or 3,000 kg ha−1) or incorporation of lime in 2018 to a depth of 0, 15, or 25 cm.a
a Significance is indicated by the P-values and LSD values or LSD range. Standard errors of the means (SEMs; presented in parentheses) and LSD values vary between treatment levels due to back-transformation (density, biomass, and seed production data) or unequal replication (yield). For wheat tillers, there was no transformation, and the design was not unbalanced, so the SEMs were consistent, and a single LSD is presented.
Table 6. Lolium rigidum density, biomass, and seed production interaction at Wongan Hills in 2018 between application of lime in 2014 (0 or 3,000 kg ha−1) and incorporation of lime in 2018 to a depth of 0, 15, or 25 cm.a
a The significance of interactions is indicated by the P-values and LSD values. Standard errors of the means (SEMs; presented in parentheses) and LSD values vary between treatment levels due to back-transformation.
The hypothesis that the addition of lime to acidic soil is correlated to a reduction in the growth and seed production of L. rigidum was supported in both field experiments. Because lime application also increased wheat tillers in both field trials and yield at Merredin, and the screen house experiment indicated that initial growth of L. rigidum was increased by lime application, the reduction in L. rigidum growth is likely due to improved competitive ability of the crop, rather than a direct result of lime application on L. rigidum growth. Lime application did not significantly affect yield at Wongan Hills, although tiller number increased from both the 1994 and the 2014 lime applications. However, the most recent lime application at Wongan Hills (i.e., in 2014) was 4 yr before the current assessment of yield or L. rigidum growth compared with Merredin, where lime was applied 2 yr before the current assessment (in 2016). In this study the soil pH data in the year of liming and the year of assessment are not presented. However, it is common for soil in Western Australia to reacidify over 4 yr due to annual application of nitrogen fertilizer, and so lime applied in 2014 would have reduced impact on crop yield in 2018 (Mason et al. Reference Mason, Porter and Cox1994a, Reference Mason, Porter and Cox1994b). These findings support prior work indicating the sensitivity of L. rigidum to long-term competition from wheat crops and reduction in L. rigidum growth following lime application to crops in acidic soils (Borger et al. Reference Borger, Riethmuller and D’Antuono2016b; Gazey and Andrew Reference Gazey and Andrew2010). At Merredin, lime application may have altered soil chemistry, which can also influence PRE herbicide performance. However, other research found no interaction between liming and efficacy of PRE herbicides like prosulfocarb, pyroxasulfone, or simazine (Hashem and Borger Reference Hashem and Borger2018).
As stated, incorporation of lime in the screen house (where seeds were planted at a consistent depth) reduced L. rigidum recruitment and increased soil pH at depth. At Merredin, incorporation reduced L. rigidum density in the continuous wheat rotation. Weed density in the wheat–fallow rotation was too low to be affected by lime or incorporation. Likewise, at Wongan Hills, incorporation reduced L. rigidum density, biomass, and seed production in the plots where lime application was 0 kg ha−1 in 2014. There was no significant effect of incorporation in plots where 3,000 kg ha−1 of lime was applied in 2014, but weed density at all depths of incorporation was much lower in these limed plots. Reduced weed density following incorporation in the field may have resulted from altered soil properties, as for the screen house experiment, but may also have resulted from burial of the L. rigidum seed beyond a depth suitable for emergence (Braunack and Dexter Reference Braunack and Dexter1989; Chauhan et al. Reference Chauhan, Gill and Preston2006a). At Merredin, reduced weed density may have resulted from increased efficacy of the PRE herbicide (Chauhan et al. Reference Chauhan, Gill and Preston2006b). The efficacy of nonsoluble PRE herbicides is reduced if they bind to crop residue on the soil surface, but incorporation buries a proportion of the crop residue and improves herbicide performance (Chauhan et al. Reference Chauhan, Gill and Preston2006b). However, no PRE herbicide was applied at Wongan Hills. Incorporation did not affect wheat yield in the field experiments, indicating that if wheat growth is initially impacted by altered soil properties (as observed in the screen house), plants can recover during the course of the season. We do not fully understand the impact of incorporation on weed emergence or early crop and weed growth, soil properties, and herbicide performance. It is clear that we need more research on weed recruitment following soil disturbance.
Lolium rigidum growth is reduced by competition from a wide range of cereal crops, but most cereal crops have reduced growth in acidic soils (Lemerle et al. Reference Lemerle, Verbeek and Coombes1995; Paynter and Hills Reference Paynter and Hills2009; Slattery et al. Reference Slattery, Conyers, Aitken, Peverill, Sparrow and Reuter1999). In Australia, the revenue loss from residual L. rigidum in-crop is estimated to cost growers A$93 million per year, due to an annual yield loss of 346,618 t of grain (Llewellyn et al. Reference Llewellyn, Ronning, Clarke, Mayfield, Walker and Ouzman2016). The current research confirms that application of lime to increase soil pH in regions with acidic soils will increase the competitive ability of crops. Further work is required to determine the competitive advantage to other crop species following lime application in acidic soils.
Acknowledgments
The research was supported by the Grains Research and Development Corporation (Soil Constraints Initiative—Innovative Approaches to Managing Subsoil Acidity, DAW00252). The authors thank staff at the Department of Primary Industries and Regional Development Merredin and Wongan Hills Research stations, Gavin Sarre, Pete Gray, and Nerys Wilkins. The authors have no conflicts of interest to declare.
