Steady agronomic and genetic interventions are essential for sustaining productivity in intensive rice cropping

Abstract

Intensive systems with two or three rice (Oryza sativa L.) crops per year account for about 50% of the harvested area for irrigated rice in Asia. Any reduction in productivity or sustainability of these systems has serious implications for global food security. Rice yield trends in the world's longest-running long-term continuous cropping experiment (LTCCE) were evaluated to investigate consequences of intensive cropping and to draw lessons for sustaining production in Asia. Annual production was sustained at a steady level over the 50-y period in the LTCCE through continuous adjustment of management practices and regular cultivar replacement. Within each of the three annual cropping seasons (dry, early wet, and late wet), yield decline was observed during the first phase, from 1968 to 1990. Agronomic improvements in 1991 to 1995 helped to reverse this yield decline, but yield increases did not continue thereafter from 1996 to 2017. Regular genetic and agronomic improvements were sufficient to maintain yields at steady levels in dry and early wet seasons despite a reduction in the yield potential due to changing climate. Yield declines resumed in the late wet season. Slower growth in genetic gain after the first 20 y was associated with slower breeding cycle advancement as indicated by pedigree depth. Our findings demonstrate that through adjustment of management practices and regular cultivar replacement, it is possible to sustain a high level of annual production in irrigated systems under a changing climate. However, the system was unable to achieve further increases in yield required to keep pace with the growing global rice demand.

Keywords: food security; intensive cropping; long-term productivity trends; rice; sustainability.

Fig. 1.

Annual rice production (observed and potential) in the LTCCE over the period from 1968 to 2017 and average production gap from 1968 to 2017. Annual production potential was computed as the sum of climatic yield potential simulated for DS, EWS, and LWS for each year averaged across three cultivars. Observed annual production was computed as the sum of observed yield for DS, EWS, and LWS for each year averaged across the cultivars and the two highest fertilizer-N levels used in the LTCCE. The annual production gap is the difference between the annual production potential and the observed production expressed in % of the annual production potential. Slopes of linear regression of potential and observed annual production with years during the phase 1 period (1968 to 1990) and during phase 2 (1996 to 2017) are presented with level of significance labeled as *** (P < 0.001) or ns (nonsignificant). In 1985, the observed yield for LWS was excluded in the total annual production because the crop was damaged by a typhoon.

Fig. 2.

Trends of climatic yield potential with climatic variables namely minimum temperature (A) and total radiation (B) during DS cropping in the LTCCE. The blue dots are values for phase 1 (1968 to 1990), and the brown dots are values for phase 2 (1996 to 2017). The open circle dots are for the transition period (1991 to 1995). The values are slope of the linear regression of climatic yield potential with variables variation [in kg ⋅ ha⁻¹ ⋅°C⁻¹ in response to minimum temperature (A) and in kg ⋅ ha⁻¹ ⋅ MJ ⋅ m⁻² in responses to total radiation (B)] with significance level of *** at P < 0.001. The black lines show the linear regression trends.

Fig. 3.

Rice yield trends in the LTCCE by season and time period. Trends are shown separately for each N rate-season combination in separate panels: (A) no-NF and (B) high-NF. Within a panel, each point is the least-square mean yield of cultivars estimated per trial. Years within phase 1 and phase 2 are blue and orange, respectively. The solid black trend lines show the rate of change in yields per phase. The upward arrows denote a statistically significant increase in yield due to the management changes starting at the beginning of the transition period (1991). Estimates of the rates of change in yields over time (kg ⋅ ha⁻¹ ⋅ y⁻¹) are written under the trend lines along with SEs (± values) and P values, which are in parenthesis. Whenever there was a significant difference in the rate of yield change between phase 1 and 2, both rates are listed under the respective trend line; otherwise, an average rate of change is listed in the center of the panel. When the rate of yield decline is not listed under the trend line panel, it was not significantly different from zero.

Fig. 4.

Nongenetic yield trends under constant management practices and climate (N rate, sowing date, cultivar, average minimum temperature, and solar radiation). Trends are shown separately for each N rate-season combination in separate panels: (A) no-NF and (B) high-NF. Within a panel, each point is the average last-square mean of year estimated using a mixed model analysis combining data across all trials within N rate-season predicted at a constant N rate excluding effect of cultivar. Years within phase 1 and phase 2 are colored in blue and orange, respectively. The solid black trend lines show the rate of change in the predicted yields per phase over time assuming cultivar, N rate, sowing date, minimum temperature, and solar ration are constant. The dotted black lines show the fitted values from the statistical model used to analyze yield trends. Estimates of the rates of change in yields over time (kg ⋅ ha⁻¹ ⋅ yr⁻¹) are written under the trend lines along with SEs (± values) and P values which are in parenthesis. Whenever there was a significant difference in the rate of yield change between phase 1 and 2, both rates are listed under the respective trend line; otherwise, an average rate of change is listed in the center of the panel. When the rate of yield decline is not listed under the trend line panel, it was not significantly different from zero.

Fig. 5.

Trends in breeding values and pedigree depth of cultivars across years. (A) The relationship in the DS between the breeding values for yield of cultivars, centered at zero, and the year of cultivar introduction in the trial ($\mathit{Yield}=11.38+1131 x \mathit{year}-748 x \mathit{yea}{r}^2$). An increase in breeding values for yield over time can be attributed to genetic improvement due to breeding ($R^2=0.16; P \mathrm{value} = 0.003$). (B) The relationship between the pedigree depth (PD) of the cultivars and the year of cultivar introduction in the trial ($PD=4.41+12.30 x \mathit{year}-7.99 x \mathit{yea}{r}^2+1.02 x \mathit{yea}{r}^3-2.03 x \mathit{yea}{r}^4$). Pedigree depth is measured as the number of the equivalent complete generations which also indicates the number of cycles of breeding that have occurred within a population ($R^2=0.78; P \mathrm{value} = 2.2\text{E}-16$). A cultivar with higher equivalent complete generations is the product of more breeding cycles compared to one with lower equivalent complete generations. In both A and B, solid black lines depict the best-fitting polynomial curve.