Short-term responses of soil carbon, nitrogen, and microbial biomass to cover crop mixtures and monocultures

Abstract

Increasingly, cover crops are being adopted for the purpose of improving soil health, yet the timescale and magnitude by which living annual cover crops might modify soil chemical and biological aspects of soil health is not well understood. At the same time, there is growing interest among farmers in cover crop mixtures due to perceptions that species-rich cover crop communities will enhance soil health relative to monocultures. In a field experiment in southeast New Hampshire, we investigated how groups of cover crops grown as monocultures and mixtures for specific seasonal niches (winter/spring, summer, and fall) influenced levels of soil nitrogen (N) and carbon (C), microbial biomass carbon (MBC), and nitrogen (MBN). Soils were sampled at cover crop maturity (winter/spring group), and at seeding, mid-season, and maturity (summer and fall groups). In the winter/spring group, average total soil N ranged from 0.192 to 0.215 mg g⁻¹ dry soil; highest total soil C content was 2.66 mg g⁻¹ dry soil; and average MBC ranged from 304.8 to 387.3 μg C g⁻¹ dry soil. In the summer group soil MBC decreased from 909.5 μg C g⁻¹ dry soil at mid-season to 644.9 μg C g⁻¹ dry soil at the end of the growth cycle. In the fall group MBC fell and rose over the season in the range of 236.0–808.3 μg C g⁻¹ dry soil. We found little evidence that cover crops influenced soil C and N parameters during the cover crop growth period relative to a weedy control or that mixtures differed from monocultures. MBC and MBN were more influenced by seasonality than the composition or diversity of the cover crop stand.

Abbreviations
• C:N
• carbon/nitrogen ratio
• MBC
• microbial biomass carbon
• MBN
• microbial biomass nitrogen

INTRODUCTION

Maintenance of soil health will be critical to efforts to sustainably intensify agricultural production systems. Cover crops in particular have been promoted as a means of improving soil health, in addition to providing a myriad of other ecosystem services (Blanco-Canqui et al., 2015). As a key component of integrated crop-livestock systems, cover crops have additional value, including as a high-quality supplemental forage (Franzluebbers & Stuedemann, 2015) and to enhance soil nutrient availability to subsequent feed and forage crops (Thapa, Mirsky et al., 2018). As a consequence, farmer adoption of cover crops continues to grow in the United States (Basche et al., 2020; Wittwer et al., 2017).

Cover crops can improve soil nutrient status by decreasing nitrogen (N) losses to the environment (Delgado et al., 2011) and, in the case of legumes, by fixing atmospheric N (Kambauwa et al., 2015). Biological N inputs by legumes increase the availability of N in soils (Lee et al., 2014; Oelmann et al., 2007), leading to increases in both above- and below-ground plant biomass (Fornara & Tilman, 2008), as well as increases in the metabolic activity of soil microorganisms (Tiemann et al., 2015).

Cover crops are also increasingly being promoted as a tool for sequestering carbon (C) and increasing soil organic matter (Lal, 2002; McDaniel et al., 2014). The C inputs from cover crops can be expected to influence soil microbial biomass and activity (Buyer et al., 2010; Nair & Ngouajio, 2012), both of which play a fundamental role in soil organic matter dynamics (Grandy & Neff, 2008).

Increasingly, farmers are planting cover crops as species-diverse mixtures, often under the assumption that mixtures confer greater or additional benefits to soil health compared to planting cover crops as monocultures (CTIC, 2017; Groff, 2008; Murrell et al., 2017). Several lines of evidence underpin this assumption. Previous research, primarily conducted in perennial grasslands, has demonstrated positive relationships between plant diversity and important ecosystem functions, including plant biomass production and resource utilization and retention (Hooper et al., 2005; Reich et al., 2012; Tilman et al., 2014).

Further, plant diversity has been shown to influence root biomass and the activity and composition of soil microbial communities (Kulmatiski et al., 2012; Lange et al., 2015; Thapa, Poffenbarger et al., 2018). Increasing plant diversity may also increase the quantity and diversity of root exudates, which can influence the activity and composition of the soil microbial community (Bais et al., 2006). Plant diversity, in combination with other factors, can alter the metabolic activity of soil microbes and storage of soil C (Eisenhauer et al., 2017; Lange et al., 2015). Lastly, plant diversity has also been linked to soil water storage, which can buffer soil microbial biomass from variation in precipitation and temperature (Eisenhauer et al., 2013; Holden et al., 2016; Lange et al., 2014).

Core Ideas

• There is little evidence that cover crops or weedy fallow differentially affected soil parameters during cover crop growth periods.
• Cover crop mixtures did not differ from monocultures.
• There is strong temporal variation in soil variables within cover crop growth periods.

While evidence from unmanaged ecosystems for positive effects of plant diversity on soil health parameters is compelling, there are also reasons to suspect these effects may be lessened in annual cropping systems and that cover crop mixtures may not necessarily provide additional benefits beyond those provided by cover crop monocultures. This reasoning flows from the fact cover crops are typically grown for only a fraction of the overall growing season each year and the short duration of cover crop growth may preclude many of the facilitative and other positive species interactions that are thought to contribute to diversity effects observed in perennial ecosystems (Smith et al., 2020).

A recent review of 27 cover crop mixture studies found little evidence that mixtures outperformed the best performing cover crops grown as monocultures for a variety of ecosystem service metrics, including biomass production, N scavenging, soil water conservation, or stimulation of soil biology (Florence & McGuire, 2020). Hence, despite the increasing adoption of cover crop mixtures, questions remain as to the magnitude and timescale of their effects on soil chemical and biological factors that regulate soil health and other ecosystem services (Tully & McAskill, 2020; Crystal-Ornelas et al., 2021), especially relative to cover crop monocultures.

To address these questions, we quantified N, soil C, and microbial biomass in cover crop mixtures and monocultures over their growth period in three field experiments in southeast New Hampshire. Each experiment was conducted during one of three seasonal niches for cover cropping (winter/spring, summer, and fall). The winter/spring period is the most common cover cropping niche in annual row crops (e.g., corn), while the summer and fall niches are more often utilized by vegetable growers. Examining all three of these niches enables us to determine whether the timing of cover crop planting and growth may influence their relative effects on the soil properties of interest. We hypothesized that cover crops would vary in their effects on total soil C and N, microbial biomass carbon (MBC), and microbial biomass nitrogen (MBN) and that mixtures would increase these parameters more than monocultures regardless of seasonal niche. Further, we hypothesized that the magnitude of effects would increase over the cover crop growth period.

MATERIALS AND METHODS Site description

The research was conducted at the University of New Hampshire Kingman Research Farm in Madbury, NH (43°11′N 70°56′W), during spring, summer, and fall of 2015. The dominant soil type is a sandy loam, with an average annual temperature of 8.1°C and mean annual precipitation of 1108 mm. Monthly precipitation and temperatures during the study year are displayed in Figure 1. The chemical and physical characteristics of the soil at the experimental site are shown in Table 1.

FIGURE 1. Monthly rainfall and monthly average maximum and minimum temperature over the year-long study period (2015) at the UNH Kingman Research Farm in Madbury, NH.

TABLE 1 Average chemical and physical characteristics of soil samples collected at 0–15 cm depth in 2015 from the experimental site at the UNH Kingman Research Farm in Madbury, NH.

[Table omitted. See PDF]

Note: Analyses were conducted by the Penn State Soil Analysis Laboratory.

Abbreviations: BD, bulk density; BS, base saturation; C, carbon; Ca, calcium; CEC, cation-exchange capacity; K, potassium; Mg, magnesium; N, inorganic nitrogen; P, phosphorus.

For several years prior to the study, the fields used for the experiments were part of a vegetable breeding program and managed as a conventional vegetable-winter rye (Secale cereale L.) cover crop rotation. Each experiment corresponded to one of three seasonal niches for cover cropping and involved either five or six cover crop species appropriate for that growing period. The seasonal niches were spring planting-summer termination (hereafter "summer group"), summer planting-fall termination (hereafter "fall group"), and fall planting-subsequent spring termination (hereafter "winter/spring group"). The cover crop species used across the three experiments included eight species of annual cool-season and warm-season grasses, three legumes, and three nonlegume forbs (Table 2). Treatment levels included in each experiment were each of the five or six cover crop species seeded as monocultures at the full rate recommended for that species and a mixture of all five or six species, depending on the experiment, each sown at one-fifth or one-sixth of their recommended rate, respectively (Table 2).

TABLE 2 Cover crop species and seeding rates used in the monocultures and mixture treatments in the summer, fall, and winter/spring seasonal cover crop groups.

			Treatments
	Species	Season group	Monoculture (kg seed ha−1)	5 species (kg seed ha−1)	6 species (kg seed ha−1)	14 species (kg seed ha−1)
Barley	Hordeum vulgare L	Winter/spring	123.29	24.66	–	8.81
BMR sorghum	Sorghum bicolor (L.) Moench (BMR)	Summer	39.23	–	6.54	2.80
Buckwheat	Fagopyrum esculentum Moench	Summer	100.88	–	16.81	7.21
Canola	Brassica napus L. cv. Dwarf Essex	Fall	11.21	–	1.87	0.80
Cereal rye	Secale cereal L	Winter/spring	134.5	26.90	–	9.61
Chickling vetch	Lathyrus sativus L cv Ac Greenfix	Summer	78.46	–	13.08	5.60
Forage radish	Raphanus sativus L. var. Niger J. Kern	Fall	11.21	–	13.08	0.80
Hairy vetch	Vicia villosa Roth	Winter/spring	44.83	8.97	–	3.20
Japanese millet	Echinochloa esculenta (A.Braun) H.Scholz	Summer	33.63	–	5.61	2.40
Oats	Avena sativa L cv Streaker	Summer, Fall	123.29	–	20.55	8.81
Crotalaria	Crotalaria juncea L	Fall	44.83	–	7.47	3.20
Teff	Eucroschus tef (Zucc.) Trotter cv. Tiffany	Summer	8.97	–	1.50	0.64
Triticale	×Triticosecale Wittmack	Fall, Winter/Spring	168.13	33.63	28.02	12.01
Wheat	Triticum aestivum L. cv. Frederick	Fall, Winter/Spring	168.13	33.63	28.02	12.01

Two of the experiments (summer group and fall group) also included a 14-species mixture made up of all of the species used across all three experiments, each sown at 1/14 of their recommended rate. All experiments included a treatment in which no cover crop was sown, which served as the weedy control.

All cover crop monoculture and mixture treatments were planted with an eight-row precision planter with rows spaced 17.5-cm apart (ALMACO). Prior to planting each experiment, the experimental site was plowed, harrowed, and then rolled to create a firm seedbed. The control was prepared as described above; however, no cover crops were sown. Each experiment was a randomized complete block design with treatment levels (cover crop monocultures and mixtures) replicated across four blocks. Individual replicate plots were 1.4-m wide and 12.2-m long. Each experiment was conducted in a different field of the farm. Planting for the summer group experiment occurred on June 12, 2015. The fall group experiment was planted on August 20, 2015, and the winter/spring group experiment was planted on October 1, 2014. After planting, cover crop treatments were allowed to grow for approximately 1.5 or 2.25 months (summer and fall groups, respectively) or 7.5 months (winter/spring group), after which time cover crops were terminated. No fertilizer was applied at any time.

Soil sampling

Soil sampling for the winter/spring group experiment occurred on May 14, 2015, just prior to cover crop termination. To assess how soil parameters might vary over the course of the cover crop growing period, soils were collected at three time points in both the summer and fall group experiments. In each case, samples were collected soon after cover crop seeding, mid-season, and just prior to termination. In the summer group, soil samples were collected on June 16, 2015, July 8, 2015, and July 31, 2015, corresponding to 3, 25, and 48 days after seeding. In the fall group, soil samples were collected on August 23, September 25, and October 25, 2015, corresponding to 3, 36, and 66 days after seeding.

At each sampling period, three replicate soil cores to a depth of 15 cm were collected from each plot. Individual soil cores were sealed in airtight bags with sufficient air to maintain oxygenation and transported to the laboratory on ice. In the lab, the three replicate cores from each plot were combined into one composite sample, with the four blocks providing four replicates per treatment. Composite samples were sieved to 2 mm and immediately stored at 4°C for further testing.

Total carbon and total nitrogen

Subsamples were air-dried and ground to a fine powder in a ball mill. Following this, 35 mg soil per sample was packed in tin capsules for the determination of C and N using an elemental combustion analyzer (Costech ECS 4010; Costech Analytical Technologies Inc.).

Soil microbial biomass carbon and nitrogen

MBC and MBN were determined using the chloroform fumigation and extraction method (Vance et al., 1987). Briefly, for each soil sample, two sets of 5 g of fresh soil were weighed into 50 mL tubes. A cotton ball was placed in the upper third of the tube without touching the soil sample. Next, 3 mL of chloroform was applied to the cotton ball in only one set of the tubes and immediately capped. After 24 h of incubation at room temperature, samples were uncapped and exposed to open air in a fume hood to allow the chloroform to evaporate from fumigated samples. Then both fumigated and unfumigated sets were extracted by adding 25 mL of a 0.5 molar potassium sulfate (K₂SO₄) solution to each tube, shaking for 1 h at 120 rpm on an orbital shaker, and then filtering through paper #01. Extracts were frozen in borosilicate flasks at −20°C until analysis on a TOC TN analyzer (TOC-V-CPN; Shimadzu Scientific Instruments Inc.). The extraction efficiency factors 0.45 (Joergensen, 1996) and 0.54 (Brookes et al., 1985) were used for microbial C and N, respectively.

Statistical analyses

Analyses of variance (ANOVA) were performed for the variables evaluated, according to a randomized complete block design. For the winter/spring group, the treatments were the cover crop monocultures and the five-species mixture and only the fixed effects of block and treatment were considered in the model. For the summer and fall groups, ANOVAs were performed in a split-plot scheme, in which the cover crop treatments were considered as a main plot, and the sampling period a subplot. In this case, the fixed effects of block, cover crop treatment, collection time, their interactions, as well as the random effect of block nested within cover crop treatment, were considered in the mixed model. A compound symmetry covariance structure was used in all analyses. When a significant interaction between cover crop treatment and collection period was detected, slicing was performed to verify differences within factor levels. Tukey’s honest significant difference test was applied to discriminate the least square means. Significance was assessed at p < 0.05 for ANOVA, slicing, and testing of means. The ANOVAs were performed using the MIXED procedure of the SAS software version 9.4 (SAS Institute, 2013).

Effects of treatments and sampling periods on soil proprieties were further analyzed with a canonical variates analysis (CVA), also known as discriminant analysis (see McCune & Grace, 2002 for a thorough description of the approach). All multivariate analyses were performed using the R Studio version 4.1.2 (R Core Team, 2021) using the “BiplotGUI” library (La Grange et al., 2009), using the settings: R > data(“dataset”); R > dataset; R > Biplots(Data = dataset[, −1], groups = dataset, 1]). Results were graphically displayed as CVA biplots.

RESULTS Winter/spring group

There were no differences in total soil N, C, or microbial biomass C or N between any of the cover crop treatments when these variables were analyzed separately (Table 3). Across the treatments N content ranged from 0.192 to 0.215 mg g⁻¹, C ranged from 2.25 to 2.66 mg g⁻¹, MBC ranged from 304.8 to 387.3 μg C g⁻¹ dry soil, and MBN ranged from 15.0 to 22.2 μg N g⁻¹ dry soil.

TABLE 3 Average total nitrogen (N), total carbon (C), carbon/nitrogen ratio (C:N), and microbial biomass carbon (MBC), and microbial biomass nitrogen (MBN) of soil samples collected at the time of termination of the winter/spring group of cover crop treatments.

	Cover crop treatment
Variable	Wheat	Triticale	Barley	Cereal Rye	Hairy Vetch	Five-species mixture	Fallow control	SEM	p value
N (mg g−1 dry soil)	0.215	0.200	0.195	0.200	0.187	0.195	0.192	0.0088	0.4672
C (mg g−1 dry soil)	2.66	2.37	2.35	2.43	2.25	2.35	2.34	0.119	0.3512
C:N	12.4	12.0	12.1	12.1	12.0	12.1	12.1	0.16	0.5963
MBC (μg g −1 dry soil)	311.9	387.3	323.9	342.1	316.4	307.5	304.8	40.33	0.7848
MBN (μg g−1 dry soil)	22.2	15.0	18.6	20.4	18.7	18.1	19.3	3.5269	0.9268

Abbreviation: SEM, standard error of the mean.

When the soil variables were analyzed together with CVA, soils from the wheat (Triticum aestivum L.) and triticale (× Triticosecale Wittmack) monocultures separated from those of the other treatments; wheat was associated with higher soil C:N while triticale was associated with higher MBC and lower MBN (Figure 2). The barley (Hordeum vulgare L.) monoculture and five-species mixture did not differ from the control.

FIGURE 2. Canonical variate analysis biplots of the total nitrogen (N), total carbon (C), the C:N ratio, and microbial biomass carbon (MBC) and nitrogen (MBN). Samples are grouped by cover crop treatment (centroids) in the winter/spring group (panel A; green shading denotes 50% and 95% confidence intervals). Samples (cover crop treatment centroids) in the summer (panel B) and fall (panel C) groups are grouped by sampling period (dotted lines delineate the range of each sampling point).

Summer group

To investigate the effect of vegetative growth on soil parameters, we sampled the cover crop treatments at three time points over their growth period. There were no interactive effects of cover crop treatment and sampling period for any of the measured variables, nor any independent effect of cover crop treatment (Table 4). We did observe a significant independent effect of sampling period for total N, C:N, MBC, and MBN, but not for total C (Table 5, Figure 2).

TABLE 4 Average of total nitrogen (N), total carbon (C), carbon/nitrogen ratio (C:N), microbial biomass carbon (MBC), and microbial biomass nitrogen (MBN) of soil samples collected from cover crop treatments in the summer group.

Cover crop treatments
Variable	BMR sorghum	Buckwheat	Teff	Millet	Oats	Chickling vetch	Six-species mixture	14-species mixture	Fallow	SEM	p value
N (mg g−1 dry soil)	0.145	0.142	0.134	0.139	0.147	0.140	0.140	0.147	0.149	0.0056	0.6788
C (mg g−1 dry soil)	1.48	1.57	1.43	1.50	1.59	1.52	1.54	1.66	1.64	0.057	0.1377
C:N	10.7	11.2	10.8	10.9	10.9	10.9	11.2	11.4	11.1	0.30	0.8415
MBC (μg g−1 dry soil)	794.7	662.1	847.6	760.4	828.7	810.9	722.4	817.7	836.0	48.18	0.1747
MBN (μg g−1 dry soil)	11.6	12.7	11.5	12.0	11.4	11.5	13.9	12.6	14.4	1.0	0.4066

Abbreviation: SEM, standard error of the mean.

TABLE 5 Average of total nitrogen (N), total carbon (C), carbon/nitrogen ratio (C:N), microbial biomass carbon (MBC) and microbial biomass nitrogen (MBN) of soil samples collected at three time points from cover crop treatments in the summer group.

	Sampling period Days after seeding
Variable	3	25	48	SEM	p value
N (mg g−1 dry soil )	0.139b	0.150a	0.138b	0.0030	0.0078
C (mg g−1 dry soil)	1.57a	1.52a	1.55a	0.023	0.1253
C:N	11.4a	10.3b	11.3a	0.17	<0.0001
MBC (μg g−1 dry soil)	805.7b	909.5a	644.9c	27.82	<0.0001
MBN (μg g−1 dry soil)	19.8a	12.4b	5.0c	1.5	<0.0001

Note: Means followed by different letters in the same row differ at 5% significance by the Tukey test.

Abbreviation: SEM, standard error of the mean.

Soil N content increased from 0.139 mg g⁻¹ at 3 days after seeding to 0.150 mg g⁻¹ at day 25. By 48 days after seeding, however, soil N content declined to 0.138 mg g⁻¹ (Table 5). Soil C content ranged from 1.43 to 1.66 mg g⁻¹ across treatments (Table 4) but did not differ between sampling periods (Table 5, Figure 2). The soil C:N ratio increased from 10.3 at 25 days after seeding to 11.3 at 48 days after seeding (Table 5).

MBC content in soils ranged from 662.1 to 847.6 μg C g⁻¹ dry soil across treatments (Table 4). The highest MBC content was measured at 25 days after seeding, at 909.5 μg C g⁻¹ dry soil, which was significantly higher than the other two periods (Table 5). MBN content ranged from 11.4 to 14.4 μg N g⁻¹ dry soil across treatments (Table 4). The largest MBN content was observed 3 days after seeding (19.8 μg N g⁻¹ dry soil), and decreased to 5.0 μg N g⁻¹ dry soil by the last sampling period (Table 5, Figure 2).

The multivariate CVA indicated samples grouped strongly by sampling date. Samples collected 3 days after seeding were associated with intermediate MBN and, compared to the other sampling periods, lower C but higher N and MBC (Figure 2). Soil samples collected 25 and 48 days after seeding were distinguished by relatively high and low MBN concentrations, respectively.

Fall group

Similar to the summer group, we did not detect a significant interaction between cover crop treatment and the sampling period for any of the soil variables, or an independent effect of cover crop treatment in the fall group of treatments (Table 6). There was, however, a significant effect of sampling period on all measured variables except for total soil N (Table 7, Figure 2).

TABLE 6 Average of nitrogen (N), carbon (C), carbon/nitrogen ratio (C:N), microbial biomass carbon (MBC), and microbial biomass nitrogen (MBN) of soil samples collected in the fall group of cover crop treatments.

	Cover crop treatments
Variable	Forage radish	Oats	Canola	Wheat	Triticale	Crotalaria	Six-species mixture	14-species mixture	Fallow	SEM	p value
N (mg g−1 dry soil)	0.139	0.142	0.151	0.137	0.137	0.143	0.146	0.151	0.147	0.0037	0.0881
C (mg g−1 dry soil)	1.60	1.56	1.67	1.48	1.55	1.60	1.64	1.71	1.61	0.044	0.0580
C:N	11.6	11.2	11.3	11.0	11.4	11.2	11.4	11.5	11.2	0.16	0.3241
MBC (μg g−1 soil)	554.9	559.9	595.5	540.2	482.5	563.7	533.0	550.9	553.4	36.45	0.6997
MBN (μg g−1 soil)	25.5	25.2	25.6	25.2	19.2	19.7	24.7	25.1	23.4	2.4560	0.4359

Note: Means followed by different letters in the same row differ at 5% significance by the Tukey test.

Abbreviation: SEM, standard error of the mean.

TABLE 7 Average of total nitrogen (N), total carbon (C), carbon/nitrogen ratio (C:N), microbial biomass carbon (MBC), and microbial biomass nitrogen (MBN) of soil samples collected at three time periods in the fall group of cover crop treatments.

	Sampling period (days after seeding)
Variable	3	36	66	SEM	p value
N (mg g−1 dry soil)	0.143a	0.144a	0.143a	0.0019	0.8767
C (mg g−1 dry soil)	1.54b	1.65a	1.62a	0.020	0.0002
C:N	11.0b	11.5a	11.4a	0.09	0.0003
MBC (μg g−1 soil)	600.4b	236.0c	808.3a	21.07	<0.0001
MBN (μg g−1 soil)	1.86b	3.05b	66.30a	1.418	<0.0001

Note: Means followed by different letters in the same row differ at 5% significance by the Tukey test.

Abbreviation: SEM, standard error of the mean.

Soil C content was lowest at 3 days after seeding (1.54 mg g⁻¹) and significantly higher at 36 (1.65 mg g⁻¹) and 66 (1.62 mg g⁻¹) days after seeding (Table 7, Figure 2). MBC decreased from 600.4 μg C g⁻¹ dry soil at 3 days after seeding to 236.0 μg C g⁻¹ dry soil at 36 days after seeding, and then increased to 808.3 μg C g⁻¹ dry soil at 66 days after seeding (Table 7). The average content of soil MBN ranged from 25.6 to 19.2 μg N g⁻¹ dry soil across treatments (Table 6). Soil MBN was low 3 and 36 days after seeding (1.86 and 3.05 μg N g⁻¹ dry soil, respectively) and was significantly higher at 66 days after seeding (66.3 μg N g⁻¹ dry soil) (Table 7, Figure 2).

Similar to the summer group, the predictive CVA biplot indicated that soil samples grouped strongly by sampling period (Figure 2). Samples collected 3 days after seeding were associated with low MBC but high soil C, N, and C:N; the inverse was true for samples collected 36 days after seeding, while samples collected just before cover crop termination (66 days after seeding) were associated with intermediate values of these parameters. Samples collected 3 and 36 days after seeding were associated with low MBN while samples collected just before termination were associated with high MBN.

DISCUSSION

We quantified soil C and N, and microbial biomass during the cover crop growing period in three experiments. Each experiment was conducted in a distinct cover cropping season (winter/spring, summer, and fall) and included cover crops belonging to several different functional groups (grasses, legumes, non-legume forbs, and winter and summer annuals) planted in monocultures and mixtures.

Across all three experiments, we found little evidence that individual cover crops or the associated weedy fallow treatment varied with respect to their influence on the soil parameters we measured. Where treatment-level variation was detected (winter/spring group only) the differences were subtle, and apparent only via analysis of the combined variation in multiple of the parameters. This relative lack of differences was despite substantial differences in cover crop performance and aboveground biomass production across treatments. For example, biomass varied from 1300 (weedy fallow) to 1000 kg ha⁻¹ (wheat and triticale) to less than 500 kg ha⁻¹ (hairy vetch, Vicia villosa Roth, and barley) among the winter/spring group of cover crop treatments; from over 4500 kg ha⁻¹ (buckwheat, Fagopyrum esculentum Moench) to less than 500 kg ha⁻¹ (chickling vetch, Lathyrus sativus L.) among the summer group; and from over 3000 kg ha⁻¹ (oats, Avena sativa L., and wheat) to less than 1000 kg ha⁻¹ (Crotalaria) among the fall group of treatments (Smith et al., 2020).

These results are somewhat surprising given that previous studies have demonstrated substantial differences in these parameters, often correlated with differences in aboveground biomass production, between different cover crop species when measured post-termination (Blanco-Canqui et al., 2015; Martinez-Garcia et al., 2018; Mortensen et al., 2021; Olson et al., 2014; Poeplau & Don, 2015). For example, Hubbard et al. (2013) found that Crotalaria (sunn hemp) increased soil C and N more than did a cover crop of crimson clover (Trifolium incarnatum L.) over a period of 3 years, likely due to greater total biomass inputs to the soil. Similarly, a study of the cumulative effects of long-term cover crop use (12 years) found that grass cover crops increased soil organic matter concentrations by 11% while legumes did not appreciably alter organic matter levels (Blanco-Canqui & Jasa, 2019).

Our study differed from many of these previous studies in that our measurements were taken during the period of active cover crop growth, before cover crop residues are subject to decomposition. Hence, it is likely the only cover-crop derived inputs to the soil during the study period were in the form of fine root turnover and root exudates. This suggests cover crop species-specific differences in the quantity and quality of root inputs and exudates may not be strong drivers of C and N in the soil, or microbial communities, while cover crops are alive and actively growing. In contrast to this interpretation, however, Finney et al. (2017) demonstrated that cover crops can have species-specific effects on soil microbial community structure and function, including microbial biomass, within 2 months of planting. Similarly, Strickland et al. (2019) found greater mineralizable soil C and higher soil microbial biomass in soils under cover crops at the time of termination compared to soils with no cover crops, suggesting that relative to bare fallow, living cover crops can alter these parameters in the short term in some cases. The fact that our fallow was weedy, and therefore experienced root-derived inputs from the living weeds, may explain the lack of differences observed between this treatment and the cover cropped treatments in our study.

Our study also differs from many others in that our measurements were made over a single cover cropping period rather than multiple years of cover cropping. Hence, it is possible that multiple years of cover crop growth would be necessary to generate differences in our soil parameters of interest. That said, several previous studies have observed relatively rapid changes in soil parameters after a single season of cover crop growth (e.g., Finney et al., 2017; Nguyen et al., 2022). For example, Nguyen et al. (2022) reported increases in microbial biomass and enhanced soil aggregation after a single year of a winter hardy cover crop mixture interseeded into corn compared to several other cover crop treatments and a no cover crop control.

We also found little evidence that cover crop mixtures affected our soil parameters of interest differently compared to their component species grown as monocultures. Given that we observed few species-specific differences in these parameters, the fact that differences were also not observed between the mixtures and monocultures is perhaps not surprising.

Congruent with our findings, in a similarly short-term study, Romdhane et al. (2019) reported that cover crop mixture diversity and composition had relatively little influence on soil microbial community composition and function after a single year of cover cropping compared to the effects of other factors, such as cover crop termination strategy. In contrast, Finney et al. (2017) reported cover crop mixture composition-specific effects on several components of soil microbial community composition and function.

Lastly, in the two experiments where we made measurements over multiple time periods (summer and fall groups), we observed temporal changes in most of the soil parameters. This result was not unexpected. Similar temporal dynamics in these soil parameters have been reported in previous cover crop and cash crop studies (e.g., Kaiser & Heinemeyer, 1993; Romdhane et al., 2019) and may have been due to the effects of soil disturbance from tillage used to prepare the seed bed prior to establishing the cover crops (Calderon & Jackson, 2002), temporal changes in growth, resource use, and C input quantity and quality by the cover crops (Zhao et al., 2021), changes in temperature or precipitation over the time course of the study (Muhammad et al., 2021; Sawada et al. et al., 2019 ), or interactions between these or other factors.

These results provide evidence that the rapid and cover crop species-specific alterations of microbial biomass C and N reported in previous studies (e.g., Finney et al., 2017) may not occur in all cover cropping situations. Our results also highlight the temporal variability in these soil parameters over cover crop growth periods, variability that can be nondirectional and occur irrespective of cover crop species or mixture composition or species richness. Additional research will be necessary to determine the causal factors responsible for this variation.

AUTHOR CONTRIBUTIONS

Igor Alexandre de Souza: Conceptualization; data curation, formal analysis, investigation; methodology; validation, writing–original draft; writing–review and editing. Amanda B. Daly: Data curation; validation; writing–review and editing. Jörg Schnecker: Data curation; methodology; validation; writing–review and editing. Nicholas D. Warren: Conceptualization; data curation, methodology; validation, writing–review and editing. Adalfredo Rocha Lobo Júnior: Data curation; formal analysis; validation; writing–review and editing. Richard G. Smith: Conceptualization; data curation; funding acquisition; investigation; methodology; project administration; resources; supervision; validation; visualization; writing–review and editing. André F. Brito and A. Stuart Grandy: Conceptualization; funding acquisition; investigation; methodology; project administration; resources; supervision; validation; visualization; writing–review and editing.

ACKNOWLEDGMENTS

The authors would like to thank Evan Ford and the late John McLean of the New Hampshire Agricultural Experiment Station for help with site logistics and land management. The authors would also like to thank the anonymous reviewers who provided insightful comments that improved this manuscript. Partial funding was provided by the New Hampshire Agricultural Experiment Station (Durham, NH; this is the scientific contribution number 2924). This work was further supported by Northeast SARE (award number LNE13-323) and Hatch Multistate NC-2042 (project number NH00670-R; project accession number 1017808). This study was financed in part by the Coordenação de Aperfeiçoamento de Pessoal de Nível Superior—Brasil (CAPES)—Finance Code 001.

CONFLICT OF INTEREST STATEMENT

The authors declare no conflicts of interest.