*2.2. Statistical Analysis*

A repeated-measures analysis of variance (ANOVA) was used to determine the effects of crop residue treatment, cycle, incubation period, and their interactions on NO3-N mineralization. As the measurements of NO3-N mineralization collected from the same experimental tube unit are related over time, the model imposed a covariance structure on the error term of the model. Akaike's information criteria (AIC) was used to determine the appropriate covariance structure, and the smaller the AIC value, the better. Throughout this paper, AR(1) covariance structure was used due to it always producing the smallest AIC value. The least square mean (LS mean) of each level of the freeze and thaw periods was estimated, and the significance of the difference between all possible pairs of these LS means was identified with Tukey's honest significant difference (HSD) at a significance level of 0.05. In addition, LS mean of each individual crop residue treatment within each freeze and thaw cycle was estimated, and an HSD test with a significance level of 0.05 was performed to test the significant difference among all possible pairs of the LS means. Moreover, LS mean of each individual crop residue treatment for each incubation period within each freeze and thaw cycle was estimated, and an HSD test with a significance level of 0.05 was used to find the LS means that are significantly different from each other. In order to detect the change of NO3-N mineralization for the bare, unamended soil over incubation cycles for each freeze and thaw period, a simple one-way ANOVA with a repeated measurements model was fitted. LS mean of each incubation period within each freeze and thaw cycle was estimated, and an HSD test was used to find the LS means that are significantly different from each other. All analyses were conducted using SAS version 9.4 [40].

#### **3. Results and Discussion**

Mean soil NO3-N mineralization capacity for individual crop residue treatments varies among freeze and thaw periods from 1.88 mg NO3-N kg−<sup>1</sup> to 58.97 mg NO3-N kg−<sup>1</sup> (Table S1, Supplementary Materials). The mean soil NO3-N mineralization capacity (range in means) peaks during cycle 3 while displaying the narrowest range in mean soil NO3-N mineralization capacity in cycle 5. Freeze and thaw period means varied from 4.95 mg NO3-N kg−<sup>1</sup> to 17.84 mg NO3-N kg<sup>−</sup>1, with cycles 4 and 5 significantly higher than cycles 1, 2, and 3. For all freeze and thaw periods, the control (bare, unamended soil) means show net N mineralization varying from 1.74 mg NO3-N kg−<sup>1</sup> to 22.72 mg NO3-N kg<sup>−</sup>1. Both cycle 4 and cycle 5 display higher mean soil NO3-N mineralization values for the bare, unamended soil (22.72 mg NO3-N kg−<sup>1</sup> and 18.45 mg NO3-N kg<sup>−</sup>1, respectively), and it is important to note that the higher control values contribute to the narrowing of the range mineralization capacity in the later freeze and thaw periods. This increase observed in the bare, unamended soil may be due to a natural microbial shift as an adaptation to the fact that no new carbon source was added to the soil only controls [41]. In other words, microbes in the bare, unamended soil recognize no carbon source addition and start to consume SOM in those controls as an energy source whereby NO3-N is being released.

When examining the five incubation cycles, the forage radish and pea (narrow C/N ratio) are the only crop residue treatments that show a significant increase in the soil NO3-N mean from the control (Figure 1).

**Figure 1.** NO3-N mineralization means and ranges in value for soil control and corn, pea, radish, soybean, spring wheat, and winter wheat crop residue treatment over five incubation cycles.

During cycle 1, the forage radish is the only crop residue treatment significantly different from the other crop residues treatment (Figure 2). In cycle 2, the soil NO3-N mean increases in both the bare, unamended soil (control) and the pea crop residue treatment, although they are still significantly different from the forage radish crop residue treatment (Figure 2). For cycle 3, the forage radish crop residue treatment still remains significantly different from the control, although both are similar to the soil NO3-N mean for the pea

crop residue treatment (Figure 2). In cycle 4 and cycle 5, the bare, unamended soil (control) and the pea crop residue treatment are not significantly different from the forage radish, nor were they significantly different from the corn, soybean, spring wheat, and winter wheat crop residue treatments (Figure 2). By cycle 5, it is evident that the corn, soybean, spring wheat, and winter wheat crop residue treatments were not mineralizing soil NO3-N and were always significantly different from the forage radish crop residue treatment. A reason for the observations in these mineralization patterns is the C/N ratio for each crop residue treatment influences the N mineralization characteristic of the residue [26].

**Figure 2.** First incubation (**A**), second incubation (**B**), third incubation (**C**), fourth incubation (**D**), and fifth incubation (**E**) cycle mean soil NO3-N mineralization patterns for control, corn, pea, radish, soybean, spring wheat, and winter wheat crop residue treatments. Different letters within each graph indicate significant difference at the 0.05 level using Tukey's comparison test.

For example, the forage radish (C/N = 8) NO3-N mineralization mean is significantly higher when compared to each crop residue treatment and the bare, unamended soil during all five freeze and thaw periods. Pea (C/N = 18) shows increased N mineralization over the bare, unamended soil in all five freeze and thaw periods, with a steady increase in mineralization beginning in cycle 3 and beyond. All other crop residue treatments with higher C/N ratios (corn (C/N = 73), soybean (C/N = 53), spring wheat (C/N = 76), and winter wheat (C/N = 101)) exhibit patterns of soil NO3-N immobilization.

A closer examination of the mean soil NO3-N mineralization patterns over freeze and thaw periods (Figure 1) indicates a wide range in soil NO3-N variation from the mean that is consistent for the forage radish crop residue treatment. This is because the forage radish crop residue possesses available N and other nutrients required by microorganisms, in addition to what the soil harbors itself; therefore, the microbes have plentiful food options based upon their wants and needs and N contained in the plant material. Conversely, the wider C/N ratio crops (corn, soybean, spring wheat, and winter wheat) exhibit narrower soil NO3-N variation from the mean N mineralization from the soil and possess limited nutrient availability in the soil, creating an environment where it is much harder for microbes to extract N, illustrating the inverse relationship with the C/N ratio (i.e., as the C/N ratio decreases, range increases). With this information alone, it appears that freeze and thaw effects on mean soil NO3-N mineralization vary based on crop residue treatment and the C/N ratio of the material.

The bare, unamended soil, absent of any crop residue treatment, shows a cumulative soil NO3-N mineralization pattern that increases with freeze and thaw periods. Cumulative values of soil NO3-N mineralization of the five cycles are 11, 19, 19, 147, and 122 mg NO3-N kg<sup>−</sup>1, respectively (Table S1). From this, it is evident that soil NO3-N mineralization in the control itself is increasing in the absence of any crop residue on the soil surface. This cumulative build-up may be evidence of the microbial shift mentioned previously that is preferentially attacking the native SOM. Because there is no carbon source added to the soil, microbes compete for a relatively stable (C/N = 10–12) nutrient source, whereby the N is then released from stable SOM. When examining soil NO3-N mineralization further over the incubation cycles for each crop residue treatment, mineralization shifts from cycles representing the early growing season towards the mid-growing season cycles with increased freeze and thaw periods. For example, during the first incubation cycle, on day 14, soil NO3-N mineralization values are significantly higher than the other incubation cycles (days) for each crop residue treatment (Figure 3). During the second incubation cycle, on day 28, soil NO3-N mineralization is significantly higher for all crop residue treatments and day 14 shows the lowest soil NO3-N mineralization values overall (Figure 3). In cycle 3, no significant differences occur among incubation cycles for all crop residue treatments (Figure 3). The shift becomes evident in the fourth incubation cycle, where soil NO3-N mineralization is significantly different on days 42, 56, and 70 from the other incubation cycles (Figure 3). In the fifth incubation cycle, soil NO3-N mineralization is significantly higher in the mid-growing season (day 42) from all other incubation cycles, while day 84 indicates significantly lower soil NO3-N mineralization values (Figure 3). This pattern of NO3-N mineralization release with the increasing number of freeze and thaw periods is of importance to pinpoint when nutrients might be available to plants. From this analysis, increased freeze and thaw periods can contribute to nutrient availability in the mid-growing season period, when plants often need it the most.

When evaluating the individual crop residue treatments and their contributions to soil NO3-N mineralization averaged over all incubation cycles these can be ranked as forage radish > pea > bare, unamended soil ≥ corn = winter wheat = spring wheat = soybean (47.80 > 19.86 > 10.57 > 0.94 = 0.76 = 0.70 = 0.66 mg NO3-N kg−1, respectively). Forage radish is significantly different from all crop residues treatments and the bare, unamended soil. The pea crop residue treatment is significantly different from the bare, unamended soil. The corn, soybean, spring wheat, and winter wheat crop residue treatments are always less than mg NO3-N kg−1, are similar, and significantly different from the forage radish, pea, and unamended soil control.

**Figure 3.** First incubation (**A**), second incubation (**B**), third incubation (**C**), fourth incubation (**D**), and fifth incubation (**E**) cycle mean soil NO3-N mineralization patterns for control, corn, pea, radish, soybean, spring wheat, and winter wheat crop residue treatments by incubation period.

Figure 4 shows the average daily trends of soil NO3-N mineralization and immobilization from the crop residue treatments during each incubation cycle (*n* = 5) to determine whether NO3-N in the crop residue contributes to N availability in the soil. The bare, unamended soil is represented by the zero line. Above the zero line indicates N mineralization, while below the zero line suggests N immobilization. Mineralization/immobilization quantities differ based on the number of incubation cycles. Daily mineralization and immobilization NO3-N values ranged from −0.20 to 2.53 mg NO3-N kg−<sup>1</sup> for the first cycle, from −0.34 to 18.61 mg NO3-N kg−<sup>1</sup> for the second cycle, from −0.30 to 7.74 mg NO3-N kg−<sup>1</sup> for the third cycle, from −2.32 to 4.60 mg NO3-N kg−<sup>1</sup> for the fourth cycle, and from −2.15 to 4.96 mg NO3-N kg−<sup>1</sup> for the fifth cycle. From these figures, it is evident that forage radish and pea are the only crops mineralizing at or above the bare, unamended soil control levels. All other crops (corn, soybean, winter wheat, and spring wheat) show N immobilization for each leaching of every incubation cycle.

**Figure 4.** First incubation (**A**), second incubation (**B**), third incubation (**C**), fourth incubation (**D**), and fifth incubation (**E**) cycle daily mean mineralization/immobilization of soil NO3-N.

These N mineralization/immobilization trends are consistent with the results provided earlier from crops with a narrower C/N ratio being the only ones to show mineralization above the bare, unamended soil. Patterns of the average daily NO3-N mineralization/immobilization trends are also similar to the trends observed when examining the overall mean for each freeze and the cycle, where increases are observed in the earlier growing season for cycles 1 and 2 and where increases are observed in the mid to late growing season in cycles 4 and 5. These patterns are also generally consistent regardless if above or below the control, although they are heavily influenced by the forage radish results.
