Effect of intercropping on P uptake and P balance under different P application rates
Two years of field localization experiments showed that intercropping maize with soybean significantly increased crop P uptake and decreased P balance under different P application rates (Fig. 1). Compared with monoculture maize, intercropping enhanced maize P uptake by 43.6%~74.3% and 45.5%~76.8% under different P application treatments in 2018 and 2019, respectively. Regardless of intercropping and monoculture, crop P uptake initially increased and then decreased with the increasing of P rate, with P90 had the highest P uptake.
Although intercropping could reduce the P balance by 17.1%~33.4% and 19.9%~32.4% for all P application treatments compared to monoculture maize and P120 treatment had the highest P balance in 2018 and 2019, respectively, but the P balance was gradually increased with the increase of P application rate and the intercropping advantage was gradually weakened with the increase of P application rate (Fig. 1).
Effect of intercropping on soil P fractions under different P application rates
Intercropping significantly affected the content and proportion of inorganic and organic P (Fig. 2). The inorganic P was the main P source (sum of Resin-P, NaHCO3-Pi, NaOH-Pi, 1 M HCl-P, conc. HCl-Pi and Residual-P) with the greater proportion of total P, ranging from 57.7–75.3% and 59.4–76.6%, the proportion of organic P (NaHCO3-Po, NaOH-Po, and conc. HCl-Po) ranging from 24.7–42.3% and 23.4–40.6%, respectively.
Regardless of intercropping or monoculture, the proportion of inorganic P had an increasing trend while the proportion of organic P had a reducing trend with the increased of P application rate for two consecutive seasons. P application rate significantly increased the content of inorganic P by 23.6–65.8% and 58.2–97.9% in 2018 and 2019, respectively, while reduced the content of organic P by 4.4–28.8% and 8.9–13.6% (except P60 treatment in 2019).
Overall, intercropping reduced the content of inorganic P by 1.8% and 2.8%, and reduced the content of organic P by 2.9% and 3.9% compared with monoculture in 2018 and 2019, respectively. Compared to monoculture, intercropping could significantly reduce the proportion of organic P and increase the proportion of inorganic P in P60 (Fig. 2), and intercropping significantly decreased the content of organic P by 7.4% compared with monoculture at P60 in 2019.
According to the difference in available P activity in the all phosphorus fractions, we divided all 9 phosphorus fractions into 3 phosphorus pools including labile P pools, moderately-labile P pools and non-labile P pools. In two consecutive growing seasons, non-labile P pools (sum of conc. HCl-Pi, conc. HCl-Po, and Residual-P) represented the largest proportion of total P in all treatments, ranging from 60.2–69.8% and 56.7–68.9% (in 2018 and 2019), followed by the moderately-labile P pools (sum of NaOH-Pi, NaOH-Po and 1 M HCl-P), which changed slightly from 26.2–26.9% and 26.1–30.0%, and the labile-P pools (sum of Resin-P, NaHCO3-Pi, and NaHCO3-Po), which represented only 3.6–10.9% and 3.6–15.2% (Fig. 3).
Intercropping greatly affected the content of P pools and the proportion of these pools to the total P under different P application rates (Fig. 3). For all P application rates, intercropping increased labile P pools by 35.7% and 37.5% on average, and moderately-labile P pools by 4.4% and 2.9%, while reducing non-labile P pools by 8.6% and 10.0% in 2018 and 2019, respectively. Compared with monoculture maize, intercropping maize significantly increased labile P by 32.5%~38.4% and 14.4%~82.1%, while significantly reduced non-labile P by 7.4%~10.9% and 6.6%~11.6% under different P application rates in 2018 and 2019, respectively.
With the increased P application rate, the proportion of labile P pools was significantly increased while the proportion of non-labile P pools was significantly reduced and no obvious effect to moderately-labile P pools. Intercropping on a whole reduced the proportion of non-labile P pools while increasing the proportion of labile P pools and moderately-labile P pools (Fig. 3).
Over two consecutive growing seasons, intercropping significantly affected soil P fractions under different P application rates (Fig. 4). P application caused a larger Resin-P, NaHCO3-Pi, NaHCO3-Po, NaOH-Pi, 1 M HCl-Pi, conc. HCl-Pi and Residual-P fractions, while smaller NaOH-Po and conc. HCl-Po fractions regardless of monoculture or intercropping.
Intercropping maize increases the content of Resin-P, NaHCO3-Pi, NaHCO3-Po, NaOH-Pi, and 1 M HCl-Pi fractions by 19.6%, 55.7%, 31.1%, 13.8% and 94.3% on average, while it reduces the content of NaOH-Po, conc. HCl-Pi, conc. HCl-Po and residual-P fractions by 2.5%, 9.5%, 4.3% and 10.1%, respectively, compared to monoculture maize under different P application rates.
In the P0 treatment, intercropping maize significantly increased NaOH-Pi by 8.6 mg kg− 1 while decrease NaOH-Po, conc. HCl-Pi, conc. HCl-Po and Residual-P by 6.36 mg kg− 1, 7.77 mg kg− 1, 12.62 mg kg− 1, and 18.24 mg kg− 1 in 2018, and increased Resin-P, NaHCO3-Po, NaOH-Pi, 1M HCl-Pi by 0.23 mg kg− 1, 11.10 mg kg− 1, 11.07 mg kg− 1, and 2.04 mg kg− 1 and decreased NaOH-Po, conc. HCl-Po, Residual-P by 6.88 mg kg− 1, 11.84 mg kg− 1, and 20.12 mg kg− 1 in 2019, respectively, compared to monoculture maize.
In the P60 treatment, increase maize increased NaHCO3-Pi, NaOH-Pi, and 1M HCl-Pi by 9.04 mg kg− 1, 10.76 mg kg− 1, and 6.38 mg kg− 1 while decreased NaOH-Po, conc. HCl-Pi and Residual-P by 11.97 mg kg− 1, 10.52 mg kg− 1, and 14.43 mg kg− 1 in 2018, and increased NaHCO3-Pi, NaHCO3-Po, NaOH-Pi, and 1M HCl-Pi by 6.17 mg kg− 1, 6.67 mg kg− 1, 10.37 mg kg− 1, and 7.91 mg kg− 1 while decreased conc. HCl-Pi and conc. HCl-Po by 29.68 mg kg− 1 and 11.96 mg kg− 1 in 2019, compared with monoculture, respectively.
In the P90 treatment, intercropping maize significantly increased NaHCO3-Pi, NaOH-Pi, and 1 M HCl-Pi by 9.66 mg kg− 1, 6.81 mg kg− 1 and 2.46 mg kg− 1 while decreased NaOH-Po and conc. HCl-Pi by 8.29 mg kg− 1 and 13.05 mg kg− 1 in 2018, and increased NaHCO3-Pi and 1M HCl-Pi by 6.50 mg kg− 1 and 3.36 mg kg− 1 while decreased conc. HCl-Pi and Residual-P by 30.38 mg kg− 1 and 7.20 mg kg− 1 in 2019, compared with monoculture maize, respectively.
In the P120 treatment, intercropping maize significantly increased NaHCO3-Pi, NaHCO3-Po and NaOH-Pi by 7.64 mg kg− 1, 11.30 mg kg− 1 and 23.13 mg kg− 1 while reduced 1M HCl-Pi, conc. HCl-Pi and Residual-P by 1.56 mg kg− 1, 21.38 mg kg− 1 and 14.13 mg kg− 1 in 2018, and increased NaHCO3-Pi, NaHCO3-Po and NaOH-Pi by 12.24 mg kg− 1, 9.82 mg kg− 1 and 8.29 mg kg− 1 while reduced NaOH-Po, conc. HCl-Pi by 10.79 mg kg− 1 and 25.71 mg kg− 1 in 2019, compared with monoculture, respectively.
Year significantly affected all P fractions except NaOH-Po and Residual-P, P rate and plant pattern both significantly affected all P fractions, and the interaction of the plant pattern and P rate were affected all fractions except Resin-P, NaCO3-Po, and NaOH-Po, and the interaction of year and P rate affected all P fractions except conc.HCl-Po, while the interaction of year and plant pattern affected only1 M HCl-Pi and conc.HCl-Pi, and the interaction of year, cropping system, and P rate affected only on NaOH-Pi (Table 2).
Table 2
Three-way interactions ANOVA on P fractions of rhizosphere of maize at flowering stage, with year (Y, 2018 and 2019), phosphorus (P) rate (0, 60, 90, and 120 kg ha− 1), and plant pattern (PP, sole maize and intercropping maize)
Factors
|
Df
|
Resin-P
|
NaHCO3-Pi
|
NaHCO3-Po
|
NaOH-Pi
|
NaOH-Po
|
1M HCl-Pi
|
conc. HCl-Pi
|
conc. HCl-Po
|
Residual-P
|
Year(Y)
|
1
|
**
|
***
|
***
|
***
|
ns
|
***
|
***
|
***
|
ns
|
Plant pattern(PP)
|
1
|
**
|
***
|
***
|
***
|
***
|
***
|
***
|
***
|
***
|
P rate(P)
|
3
|
***
|
***
|
***
|
***
|
***
|
***
|
***
|
***
|
***
|
Y×PP
|
1
|
ns
|
ns
|
ns
|
ns
|
ns
|
**
|
*
|
ns
|
ns
|
Y×P
|
3
|
**
|
**
|
***
|
**
|
***
|
***
|
***
|
ns
|
***
|
PP×P
|
3
|
ns
|
***
|
ns
|
**
|
ns
|
***
|
**
|
**
|
*
|
Y×PP×P
|
3
|
ns
|
ns
|
ns
|
**
|
ns
|
ns
|
ns
|
ns
|
ns
|
Note: Df, degrees of freedom; ns, no significant difference * P < 0.05, ** P < 0.01, *** P < 0.001 |
The changes in P fractions between intercropping and monoculture (including 4 P rates) revealed that intercropping could deplete more non-labile P pool (including NaOH-Po) to increase crop P uptake and reduce P balance (Fig. 1, Fig. 6). Covering all P application treatments in two consecutive years, intercropping depleted NaOH-Po (3.39 ~ 11.97 mg kg− 1), conc. HCl-Pi (3.72 ~ 30.38 mg kg− 1), conc. HCl-Po (4.89 ~ 12.62 mg kg− 1 in P0, P60 and P90) and Residual-P (2.13 ~ 20.12 mg kg− 1), but obviously increased Resin-P (0.11 ~ 0.33 mg kg− 1), NaHCO3-Pi (0.98 ~ 12.24 mg kg− 1), NaHCO3-Po (2.80 ~ 11.30 mg kg− 1), NaOH-Pi (5.37 ~ 23.13 mg kg− 1) and 1 M HCl-Pi (1.22 ~ 7.91 mg kg− 1) compared with monoculture. The changes in each P fraction mentioned above were significantly affected by the P rate and plant pattern independently, and the interactions between the P rate and plant pattern were significantly affected NaHCO3-Pi, NaOH-Pi, 1 M HCl-P, conc.HCl-Pi, conc.HCl-Po and Residual-P (Table 2).
Effect of intercropping on soil available P content and P activation coefficient under different P application rates
Soil available P (including Olsen-P, Resin-P and NaHCO3-Pi) and P activation coefficient (PAC) in the rhizosphere soil were significantly affected by P application rate and/or plant pattern (Fig. 6). P application (P60, P90, and P120) significantly increased Olsen-P by 1.8 ~ 3.0 times and 1.5 ~ 2.5 times, respectively, in 2018 and 2019, compared with P0 treatment, and P application also increased Resin-P by 1.7 ~ 4.4 and 1.3 ~ 3.2 times related to no P application, as well as increased NaHCO3-Pi by 5.4 ~ 15.8 and 3.1 ~ 7.3 times, respectively. In addition, the P application increased the PAC by 1.4 ~ 2.3 times and 0.8 ~ 1.3 times in 2018 and 2019, respectively.
Intercropping had an advantage for increasing soil available P and P activation coefficient, increasing Olsen-P by 12.5% and 15.4%, increasing Resin-P by 19.6% and 16.9%, increasing NaHCO3-Pi by 55.7% and 33.7%, and increasing PAC by 11.3% and 15.3% at all P application rates relate to monoculture in 2018 and 2019, respectively. Compared with monoculture, intercropping maize significantly increased Olsen-P and Resin-P by 25.9% and 41.2% at P60 in 2019. Otherwise, intercropping significantly increased soil NaHCO3-Pi by 84.6%, 52.4%, 27.3% and 35.8%, 21.1%, 35.4% in 2018 and 2019, respectively, compared with monoculture maize (at P60, P90, and P120 treatments), respectively (Fig. 6). The intercropping advantage was gradually weakened with the increasing of P application rate.
Effect of intercropping on rhizosphere soil ALP and ACP under different P application rates
This study also found that intercropping significantly affected the ACP and ALP under different P application rate (Fig. 7). Compared with monoculture, intercropping maize increased ALP by 21.2%~42.6% and 19.9%~28.6%, respectively, and the differences were significant for all P application rates except P0 over two years. Meanwhile, intercropping also significantly increased the ACP by 13.8%~27.1% and 9.5%~13.4% than monoculture under different P application rates. Regardless of monoculture or intercropping, the ALP gradually decreases while the ACP was first to increase then decreased with the increase of P application, as well as the intercropping advantage gradually weakened.
Relationship between soil P fractions with available P and PAC
The correlations between the P fractions and available P and PAC were demonstrated through a heat map analysis (Fig. 8). Irrespective of monoculture or intercropping maize, available P and PAC were all significantly positively related to NaHCO3-Po, NaOH-Pi, 1M HCl-P, conc.HCl-Pi, and Residual-P fractions, and Labile P, Moderately-labile P, Non-labile P, and inorganic P pools (p < 0.05, p < 0.01), while significantly negatively related to NaOH-Po and conc. HCl-Po fractions and organic P pools (except PAC, which was not significantly related to conc.HCl-Po in intercropping maize). Thus, we think that organic P pools (including NaOH-Po and conc. HCl-Po) could be transformed into inorganic P pools to become available P.
Regulation of intercropping on soil phosphorus pools
We constructed a structural equation model (SEM) to explore the influence of intercropping on the increase in available P via regular acid phosphatase activity (ACP), and alkaline phosphatase activity (ALP) (Fig. 9). The most parsimonious model explained 91% (monoculture maize) and 85% (intercropping maize) of the variance in available P. Both ACP and ALP affected Olsen-P directly or indirectly through their effects on organic P turnover. In detail, regardless of intercropping or monoculture maize, ACP showed a significant positive impact on NaHCO3-Po to increase Olsen-P, and a significant negative impact on conc.HCl-Po transformation NaOH-Po or NaHCO3-Po into Olsen-P, while ALP mainly through significantly negative effect on NaOH-Po or NaHCO3-Po transformation into available P.
Although ACP had a greater influence on NaHCO3-Po in monoculture maize (p = 0.58***) than intercropping maize (p = 0.36**), the influence was sample NaHCO3-Po turnover into Olsen-P in monoculture and intercropping. While ACP had a greater effect on conc.HCl-Po in intercropping (p=-0.63***) than in monoculture (p=-0.53**). Furthermore, we found a significant negative relationship between ALP and NaHCO3-Po only in intercropping maize.