- Email: [email protected]
Physiological determinants of high yielding ultra-narrow row cotton: Canopy development and radiation use efficiency
Field Crops Research 148 (2013) 86–94
Contents lists available at SciVerse ScienceDirect
Field Crops Research journal homepage: www.elsevier.com/locate/fcr
Physiological determinants of high yielding ultra-narrow row cotton: Canopy development and radiation use efficiency R. Brodrick a,c,∗ , M.P. Bange a , S.P. Milroy b , G.L. Hammer c a b c
CSIRO Plant Industry, Cotton Catchment Communities Cooperative Research Centre, LMB 59, Narrabri, NSW 2390, Australia CSIRO Plant Industry, Centre for Environment and Life Sciences, Private Bag 5, Wembley, WA 6913, Australia School of Land and Food Sciences, the University of Queensland, Brisbane, QLD 4072, Australia
a r t i c l e
i n f o
Article history: Received 16 May 2011 Received in revised form 15 May 2012 Accepted 16 May 2012 Keywords: Gossypium hirsutum Leaf area index Light interception Canopy extinction coefficient Radiation use efficiency Row spacing
a b s t r a c t Ultra-narrow row cotton (UNR, with rows spaced less than 40 cm apart) has long been proposed to have the potential to increase yields while reducing the time to crop maturity. Investigations have shown that biomass accumulation in high-input, high yielding UNR (25 cm spaced rows with yields greater than 1800 kg lint ha−1 ) cotton is similar to conventionally spaced rows (100 cm) despite a three-fold increase in plant density, indicating a limitation on individual plant growth. This study investigates whether the increased plant density in UNR crops (36 plants m−2 ) leads to differences in canopy development, radiation use efficiency (RUE) and light interception contributing to plant growth limitations. Three experiments over three years compared UNR treatments to conventionally spaced treatments in high-input production systems and found that early canopy development (leaf area index LAI) and consequently early interception was higher in the UNR crops in two of the three experiments. This resulted in a 17% higher seasonal canopy extinction coefficient (k) in UNR crops over the season. However, seasonal RUE of the UNR crop was 19% lower as increases in light interception were not accompanied by increased total dry matter. Light distribution through the canopy was poorer (higher k) in the UNR crop and LAI continued to increase in the UNR crop after maximum light interception was reached, which combined with a lower leaf nitrogen concentration may have reduced the photosynthetic efficiency of the UNR crop. We conclude that differences in canopy light interception and the efficiency of conversion of light to biomass were the primary factors responsible for differences in the pattern of biomass accumulation between UNR and conventionally spaced cotton. © 2012 Elsevier B.V. All rights reserved.
1. Introduction Ultra-narrow row cotton (UNR, with rows spaced less than 40 cm apart) has long been proposed to have the potential to increase yields while reducing the time to crop maturity (Nichols et al., 2003). This paper forms part of a detailed investigation into the growth and development of ultra-narrow row cotton in high-input production systems. We found that there was potential for higher yields when compared to conventionally spaced rows (100 cm row spacing) in high yielding, high-input productions systems in Australia, however there was no difference between row spacings in time to crop maturity (Brodrick et al., 2010). A detailed growth analysis and physiological determinants framework (Charles-Edwards et al., 1986; Coleman et al., 1994) was used to identify key differences in the factors influencing yield and
∗ Corresponding author at: CSIRO Plant Industry, Cotton Catchment Communities Cooperative Research Centre, LMB 59, Narrabri, NSW 2390, Australia. Tel.: +61 2 6799 1500; fax: +61 2 6793 1186. E-mail address: [email protected] (R. Brodrick). 0378-4290/$ – see front matter © 2012 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.fcr.2012.05.008
maturity of UNR and conventionally spaced cotton crops in highinput, high yielding production systems. Brodrick et al. (in press) found that the increased yield in UNR crops was not due to differences in final biomass accumulation but associated with more bolls per unit area and increased partitioning to reproductive growth in the UNR crops. Initially crop growth rate and biomass production was higher but crop growth rate slowed earlier in the UNR crop and final biomass was not different from the conventionally spaced crop despite a three-fold increase in plant density (Brodrick et al., in press). The UNR plants were smaller, with less biomass produced per plant indicating limitations in assimilates for growth due to the increased number of plants competing for resources in the UNR crop. If competition between plants in the UNR treatments reduced the availability of light, water and nutrients to individual plants, this would have limited their biomass production and growth, and as total crop water use and final nutrient uptake were not different (Brodrick et al., in press), it is likely that competition for light in the higher density UNR crop slowed biomass accumulation. Few studies have compared the leaf area development and light interception characteristics of cotton grown in UNR and conventionally spaced rows in low-input systems, and none to our
R. Brodrick et al. / Field Crops Research 148 (2013) 86–94
knowledge in high-input, high yielding (>1800 kg lint ha−1 ) systems. In low-input UNR cotton systems the higher number of plants can lead to greater early leaf area index (LAI) accumulation compared to conventionally spaced cotton (Darawsheh et al., 2009; Gwathmey and Clement, 2010; Jost and Cothren, 2000; Kreig, 1996). Closer plant spacing means that plants do not need to be as large for the crop to achieve maximum light interception. An increase in light interception has been reported for narrower row spacings in cotton (Heitholt et al., 1992; Peng and Krieg, 1991) and other species such as corn (Andrade et al., 2002), chickpea (Leach and Beech, 1988), sorghum (Flenet et al., 1996), and soybean (Board and Harville, 1992; Savoy et al., 1992). This rapid canopy closure may also lead to reductions in weed competition (Forcella et al., 1992) and decreased soil evaporation (Kreig, 1996; Nunez and Kamprath, 1969). This increased early light interception has been thought by many researchers to be the primary reason for increases in yield in narrower row spacings in many indeterminate and determinate crops (Andrade et al., 2002; Flenet et al., 1996; Shibles and Weber, 1966). Constable (1975) found a significant negative relationship between boll growth rate and LAI in UNR spaced cotton. Heitholt et al. (1992) found that narrow rows (0.5 m spaced rows) had a higher canopy extinction coefficient (k) compared to conventionally spaced rows and concluded that the optimal LAI for maximum light interception and yield for narrow rows was between 3 and 4 compared with an LAI of 4–5 for conventionally spaced rows. Other authors have hypothesized that the denser canopy in UNR spaced cotton could lead to reduced light penetration leading to increased shedding or smaller bolls (Baker, 1976; Constable, 1975; Jost and Cothren, 2001). Changes in biomass production can also be a result of differences in a crop’s ability to convert solar radiation into biomass, as represented by the radiation use efficiency (RUE) of the crop (Monteith, 1977). No studies have compared radiation use efficiency in UNR to conventionally spaced rows in cotton. Savoy et al. (1992) found that narrower rows in soybean (0.36 m spaced rows) had higher light interception, greater biomass accumulation and higher radiationuse-efficiency compared to wide rows (1.02 m spaced rows). Understanding how decreasing row spacing in high-yielding, high-input cotton systems affects the physiological determinants of biomass production is important in realizing any benefits of these systems. In this paper we will test the following hypotheses, (i) the increased plant population in the UNR crop increases light interception in the UNR crop; (ii) canopy development leads to a higher canopy extinction coefficient in the UNR crop and (iii) radiation use efficiency is lower resulting in no differences in final biomass production in the UNR crop.
Experiments were fully irrigated according to crop requirements. Management followed current commercial practices for high-input management, irrigation and insect control as described by Hearn and Fitt (1992). Crop water use and nitrogen uptake details are reported in Brodrick et al. (in press). 2.2. Measurements 2.2.1. Leaf area index On approximately 12 occasions in each experiment, biomass, leaf area, specific leaf area (SLA) and LAI were measured by destructive sampling (see Brodrick et al., in press). Leaf area was determined by measuring the leaf area of the sub-sample with a LiCor planimeter (Model LI-3100, LiCor Inc., Lincoln, NB, USA). This sample was dried and weighed and specific leaf area determined (m2 g−1 ). LAI was calculated as the product of specific leaf area and amount of leaf dry matter (g m−2 ). 2.2.2. Light interception Total daily incoming radiation was measured using a calibrated pyranometer at the Australian Cotton Research Institute weather station less than 2 km from the experimental fields. In Exp. 1, solar radiation intercepted by the canopies was measured using tube solarimeters (Model TSL Delta-T Devices Ltd, Cambridge, UK). A single tube solarimeter was placed across one bed in each plot (in a north–south orientation) to measure transmitted radiation. One tube solarimeter was placed above the crop in the middle of the experiment to measure incident solar radiation. The solarimeters were calibrated against the solarimeter positioned above the crop before and after each experiment. The solarimeters were programmed to scan at 5-min intervals, recording average hourly readings on a programmable datalogger (Model DL Delta-T Devices Ltd, Cambridge, UK). In Exps. 2 and 3 solarimeter data were not collected for the full season or for all plots due to problems with dataloggers. For these experiments intercepted solar radiation was calculated from weekly measurements of intercepted photosynthetically active radiation (PAR) using a sunfle0ck ceptometer (SF-80, DeltaT Devices Ltd, Cambridge, UK). Incident radiation was recorded between 11:00 and 13:00 h (Australian Eastern Standard Time) above each plot averaging three readings. Transmitted radiation was recorded by average readings taken at ground level in three random areas in each plot from the center of the furrow to the center of the bed. Solarimeter data from experiment 1 was converted to PAR. To estimate canopy closure, the proportion of PAR intercepted by the crop at midday (LII ) was calculated as: LII =
2. Methods 2.1. Site and climate description UNR and conventionally spaced cotton crops were compared in three experiments grown near Narrabri, in a semi-arid environment of north-west New South Wales, Australia. The experiments are described in detail in Brodrick et al. (in press). Briefly, three experiments were sown using the cultivar Sicala V-3RRi (Reid, 2001). Exp. 1 was sown 16 November 2001, Exp. 2 was sown 10 October 2002, and Exp. 3 was sown 23 October 2003. These experiments correspond with Exps. 1, 2 and 5 in Brodrick et al. (2010). In the 25 cm UNR treatment, the row configuration was six rows spaced 25 cm apart on a 2 m bed sown with 36 plants m−2 . In the conventionally spaced treatment, the row configuration was two rows spaced 1 m apart on a 2 m bed sown with 12 plants m−2 . A randomized complete block design with four replicates was used.
87
(incident radiation − transmitted radiation) incident radiation
An exponential function was fitted to LII over DAS to allow interpolation between measurement dates:
LII = a 1 − e(−bDAS) + c where a, b and c are fitted coefficients (Charles-Edwards and Lawn, 1984). To calculate cumulative intercepted daily radiation and the light extinction coefficient (k), total daily intercepted radiation (LID ) was calculated from instantaneous measurements by adjusting the measurements using the relationship (Charles-Edwards et al., 1986): LID =
2LII 1 + LII
To allow interpolation between dates of measurement LID was also regressed over DAS using the same equation as for LII .
88
R. Brodrick et al. / Field Crops Research 148 (2013) 86–94
Cumulative intercepted solar radiation (CLID ) was calculated using total incident daily radiation (LID ) and the measured daily proportion intercepted for the period of measurement in each experiment. For each experiment, CLID was calculated up to the biomass harvest with the highest average LAI, after which LAI began to drop off due to leaf senescence (Exp. 1 – 118 DAS, Exp. 2 – 123 DAS and Exp. 3 – 112 DAS). In all experiments, this period covered the period of maximum growth and light interception. 2.2.3. Light extinction coefficient The light extinction coefficient (k) is a parameter that indicates the effectiveness of a crop canopy at intercepting PAR (CharlesEdwards et al., 1986). The light extinction coefficient was derived for the whole season by regressing light interception (LID ) on LAI for each plot, using a modified form of Beer’s Law:
LID = a 1 − e−kLAI
where LID is the proportion of intercepted PAR; k is the light extinction coefficient; and a represents the maximum value of light interception that can be attained by the crop canopy. This analysis, however, assumes that light interception characteristics are constant throughout crop development, thereby ignoring changes in leaf angle, the condition of leaf surface and the overall canopy structure (Charles-Edwards et al., 1986). 2.3. Radiation use efficiency Seasonal crop radiation use efficiency (RUEg ) (g MJ−1 ) was derived from the gradient of the linear regression of accumulated total dry matter (glucose equivalent) and cumulative intercepted total solar radiation (CLID ) (Monteith, 1977). RUEg between harvest dates was calculated by a similar linear regression for each time period. 2.4. Specific leaf nitrogen Dried leaf samples for each plot from each biomass harvest were ground, mixed, and analyzed for nitrogen content using Kjeldahl digestion. Specific leaf nitrogen (SLN (g N/m−2 per unit leaf area)) was calculated as the quotient of leaf nitrogen content and specific leaf area (Muchow, 1988). SLN for UNR treatments was then plotted against SLN for conventionally spaced treatments using data from all experiments. 2.5. Data analysis To test for differences between UNR and conventionally spaced systems a combined analysis across all experiments was undertaken using generalized linear modeling (GLM). In this analysis the main factors were row spacing and experiment (Exp. × row spacing treatment), and the random factors replicate and experiment (Exp. × replicate). For those variables that differed in their response across experiments an analysis of variance (ANOVA) for a randomized block design was used for comparing row spacings. Simple linear regression analysis was performed to test for differences between row spacings in the regressions for k and RUEg . To test for differences between SLN of the two treatments the linear relationship was compared against unity (1:1 line). All statistical analyses were conducted using Genstat software (VSN International, Rothamsted, UK). Unless stated otherwise significant differences were considered at 95% confidence intervals (P < 0.05). Where shown graphically, the standard errors are ±one standard error of the treatment means from the associated ANOVA. In some data sets, the means of all harvests are presented
Fig. 1. Cumulative rainfall over the growing season for Exps. 1 to 3.
graphically, however an individual ANOVA was performed for each harvest date. 3. Results 3.1. Climatic conditions More detail about the climatic conditions over the three years of experiments is presented in Brodrick et al. (in press). Exps. 1, 2 and 3 experienced similar temperature patterns resulting in similar cumulative day degrees at the end of the growing season. The later sowing of Exp. 1 meant that day degree and solar radiation accumulation was slightly slower at the end of the season. Early in-crop rainfall was extremely low in Exp. 2 (Fig. 1). Total in-crop rainfall was lower in Exps. 1 and 2 compared with Exp. 3; however, 155 mm of the rainfall during Exp. 3 occurred in a single 5-day period during early flowering. 3.2. Leaf area index Responses of LAI to row spacing were not consistent across the three experiments (Fig. 2). Early LAI (prior to 100 days after sowing (DAS)) tended to be significantly higher in the UNR spaced treatments compared to the conventionally spaced treatments in Exps. 1 and 3 but in Exp. 2 there were no significant differences in LAI at any of the harvest dates. 3.3. Light interception The proportion of intercepted photosynthetically active radiation (LII ) was derived from measurements of PAR over the period of greatest growth (Exp. 1 – 118 DAS, Exp. 2 – 123 DAS and Exp. 3 – 112 DAS) (Fig. 2). In all three experiments, the UNR treatments reached 80% light interception (approximation of canopy closure) before the conventionally spaced treatments (Fig. 2). The UNR treatment reached 80% LII 35 days earlier than the conventional spacing in Exp. 1 and 11 days earlier in Exp. 3. In Exp. 2, however, there was not a significant difference between the row spacings. 3.4. Crop light extinction coefficient A combined analysis of the relationship between LII and LAI showed a significantly higher k in the UNR treatments compared to conventionally spaced treatments across all three experiments (Figs. 3 and 4; Table 1).
R. Brodrick et al. / Field Crops Research 148 (2013) 86–94
89
Fig. 2. Mean leaf area index (LAI) versus days after sowing for ultra-narrow rows (UNR) and conventionally spaced treatments in Exps. 1 (a), 2 (b) and 3 (c). Proportion of intercepted photosynthetically active radiation (LII ) versus days after sowing for UNR (solid line) and conventionally spaced (broken line) treatments in Exps. 1 (d), 2 (e) and 3 (f). Error bars are ± one standard error of the mean. * = 95% significance level; ** = 99% significance level. Lines are included to assist comparison between treatments.
Table 1 Summary of significant differences from combined analysis of experiments for radiation use efficiency (RUEg ) and the crop light extinction coefficient (k). Period of measurement for Exp. 1 was to 118 DAS, Exp. 2 to 123 DAS and Exp. 3 to 112 DAS. Error df = 13. Variable
UNR
Conventionally spaced
RUEg (g MJ−1 ) k
1.03 0.81
1.23 0.69
n.s. = no significant difference.
P-value Row spacing
P-value Experiment
P-value Experiment × row spacing
df = 1 <0.001 <0.001
df = 2 <0.001 n.s.
df = 2 n.s. n.s.
90
R. Brodrick et al. / Field Crops Research 148 (2013) 86–94
Fig. 4. Relationship between light interception and leaf area index for ultra-narrow rows (UNR) (solid line) and conventionally spaced (broken line) treatments for all data from Exps. 1, 2 and 3.
3.6. Specific leaf area and specific leaf nitrogen
Fig. 3. Relationship between light interception and leaf area index for ultra-narrow rows (UNR) (solid line) and conventionally spaced (broken line) treatments for Exps. 1 (a), 2 (b) and 3 (c).
There were few significant differences in SLA between row spacing experiments (Fig. 7). In Exp. 1, SLA in the UNR treatments was significantly higher than the conventionally spaced treatments at 35 DAS and significantly lower at 59 DAS (Fig. 7a). The only significant difference in SLA in Exp. 2 was significantly lower in the UNR treatments compared to the conventionally spaced treatments at 55 DAS (Fig. 7b). In Exp. 3, SLA was significantly lower in the UNR treatments compared to the conventionally spaced treatments at 54 and 60 DAS (Fig. 7c). Differences in specific leaf nitrogen (SLN) between row spacings were not consistent across the three experiments (Fig. 7d–f). Although SLN was generally numerically higher in the conventionally spaced treatments, there were few significant differences. SLN in individual plots in the conventionally spaced treatments was regressed against SLN in UNR plots. In Exps. 1 and 3, the majority of data lay above the 1:1 line, reflecting the numerically higher observation seen in the plots against DAS (Fig. 7g–i). The intercept was significantly greater than zero, indicating that when the canopies had a low N status, canopies of conventionally spaced crops had higher SLN. However SLN of the UNR crops was more responsive when leaf N status was generally higher (Fig. 7g–i)
4. Discussion 3.5. Radiation use efficiency A seasonal crop radiation use efficiency (RUEg ) was derived from the slope of the linear regression of cumulative total dry matter versus cumulative intercepted total solar radiation (Fig. 5). RUEg was significantly lower in the UNR treatment compared to the conventionally spaced treatment in Exp. 1, but there was no significant difference between treatments in Exps. 2 and 3 (Fig. 5). A combined analysis of RUEg across the three experiments showed significantly lower RUEg in the UNR treatments compared to the conventionally spaced treatments (Table 1). There were few significant differences in RUEg calculated for the intervals between sampling dates (Fig. 6).
4.1. Leaf area index, light interception and light extinction coefficient were greater in UNR Early canopy development and consequently light interception was higher in the UNR crops in two of the three experiments but this did not lead to a significant difference in total cumulative intercepted solar radiation during the period of measurement compared with the conventionally spaced crops. Early canopy development and light interception followed a similar pattern to biomass accumulation and crop growth rate (Brodrick et al., in press). This pattern of early light interception and higher crop growth rate in UNR is consistent with the results of Kreig (1996).
R. Brodrick et al. / Field Crops Research 148 (2013) 86–94
91
Fig. 5. Relationship between total dry matter and cumulative intercepted solar radiation for ultra-narrow rows (UNR, solid line) and conventionally spaced (broken line) treatments in Exps. 1 (a), 2 (b) and 3 (c).
There was very little early in-crop rainfall in Exp. 2 and leaf expansion and canopy development were much lower in both treatments. This appears to be a direct effect of water stress (Constable and Rawson, 1982) on leaf expansion as early SLA was lower and SLN tended to be higher during early development in this experiment. In addition, SLN of the two treatments did not differ in this experiment, suggesting that expansion in the UNR crop was more strongly affected than in the conventionally spaced crop. LAI in the UNR crops continued to develop after maximum LII had been reached, whereas in the conventionally spaced treatments peak LAI was more aligned with maximum LII. Elevated LAIs can be detrimental to the crop in a number of ways. Low light intensity in the lower canopy reduces photosynthetic rates of those leaves and hence assimilate production to support boll development on the associated branches (Constable and Rawson, 1982;
Fig. 6. Mean radiation use efficiency (RUE) versus days after sowing for ultra-narrow rows (UNR) and conventionally spaced treatments in Exps. 1 (a), 2 (b) and 3 (c). * = 95% significance level; ** = 99% significance level. Error bars are ± one standard error of the mean. Lines are included to assist comparison between treatments.
Zhao and Oosterhuis, 2000). This may have resulted the lower fruit retention and smaller boll size measured in the UNR crops (Brodrick et al., 2010). In addition, there is evidence that changes in light intensity and quality can influence boll shedding directly (Board et al., 1994; Constable, 1981). Beyond the physiological influences, excessive LAI impedes penetration of insecticides and has been shown to have negative effects on insect control (Marois et al., 2004) and increases in boll rot (Andries et al., 1969).
92
R. Brodrick et al. / Field Crops Research 148 (2013) 86–94
Fig. 7. Mean specific leaf area (SLA) and mean specific leaf nitrogen (SLN) versus days after sowing for ultra-narrow rows (UNR) and conventionally spaced treatments in Exps. 1 (a and d), 2 (b and e) and 3 (c and f). Comparison of SLN between UNR and conventionally spaced treatments in Exps. 1 (g), 2 (h) and 3 (i). To test for differences between SLN of the two treatments the relationship was compared against unity (1:1 line). * = 95% significance level; ** = 99% significance level. Error bars are ± one standard error of the mean. Lines are included to assist comparison between treatments.
Through both increased plant density and reduced row spacings the UNR (36 plants m−2 ) and conventionally spaced (12 plants m−2 ) treatments have different spatial arrangements and plant structure, difference in the light extinction coefficient (k) is likely due to differences in morphology and arrangement of plants. A higher k in the UNR crop, and hence, greater light capture at low LAI, did not increase final total biomass production most likely because of reduction in RUEg . A dense canopy with overlapping leaves may mean that although there was more light intercepted per unit LAI in the UNR treatments, the vertical distribution of light in the canopy was not uniform and likely result in more light being intercepted by the upper part of the canopy (Constable, 1986). Higher k generates less uniform light distribution in the canopy so that overall conversion efficiency is reduced, especially at high LAI (Duncan et al., 1967). In the UNR crop, with LAI continuing to increase after maximum light interception had been reached, the higher k would have only been of benefit to biomass accumulation of the UNR crops while light interception was still increasing.
4.2. Radiation use efficiency was lower in the UNR crop Lower seasonal RUEg indicated that the UNR treatment was less efficient in converting intercepted solar radiation into biomass production. The UNR canopy had higher early LAI and high light interception in Exps. 1 and 3 but no differences in total dry matter production. An average RUEg of 1.0 g MJ−1 for UNR crops when converted to total dry matter is 0.54 g MJ−1 and is at the lower end of the range of reported values from 0.62 to 0.75 g MJ−1 (converted to solar radiation if the authors measured PAR) for seasonally measured RUE in cotton (Bange and Milroy, 2004; Constable et al., 1990; Rosenthal and Gerik, 1991; Sadras, 1996). The RUEg in the UNR crop could be lower due to the poor light distribution through the canopy (higher seasonal k) or due to the lower canopy SLN in the UNR crop reducing photosynthesis and ultimately biomass production; or a combination of both. The UNR crop may have had less light penetrating to the lower parts of the canopy due to the combination of higher LAI and higher
R. Brodrick et al. / Field Crops Research 148 (2013) 86–94
k. At high LAI, canopies with better light distribution through the canopy (lower k) have better photosynthetic efficiency and hence increased RUEg because more of the leaf surface in the canopy is intercepting light (Charles-Edwards et al., 1986; Sadras, 1996). Radiation use efficiency is closely linked to leaf N concentration (Sinclair et al., 2000; Sadras, 1996; Milroy and Bange, 2003) because of the dependence of photosynthesis on leaf N. The lower canopy SLN may also be as a result of poorer light distribution in the UNR canopies; research for conventionally spaced cotton has found that SLN is positively correlated with light intensity throughout the canopy (Milroy et al., 2001). However, in Exp. 1 there are indications that SLN played a role in the lower RUEg in the UNR crop, as both SLN and RUEg were significantly lower before 60 DAS when LAI and light interception would not have been influencing distribution of SLN in the canopy. As SLA was not different between row spacings, lower SLN in the UNR crop indicates that there is a dilution of N in the leaves, and may suggest there was greater demand for N in the UNR crop, although total N uptake remained the same (Brodrick et al., in press). In addition to the UNR crop having lower RUEg and SLN accompanied by higher k, one of the experiments in this study (Exp. 2) indicated that the UNR crop may also be more susceptible to early water stress. Investigations are continuing to determine whether changing canopy arrangement through different plant populations or rowing spacings; or increasing early nitrogen and water to increase SLN or mitigate any early water stress; can alter leaf area development, light interception and radiation use efficiency in UNR crops.
5. Conclusion The results of this study suggest that differences in canopy light interception characteristics and the efficiency of conversion of light to biomass may be the primary factors responsible for differences in the pattern of biomass accumulation between high yielding UNR and conventionally spaced crops. Early canopy development (LAI) and consequently early light interception was higher in the UNR crops in two of the three experiments. This resulted in a 17% higher canopy extinction coefficient (k) in the UNR crops over the season. However as total dry matter was not higher despite increased light interception, RUEg was lower in the UNR crop. The 19% lower RUEg in the UNR crop was most likely due to poor light distribution through the canopy, combined with lower leaf nitrogen concentration, reducing the photosynthetic efficiency of the leaves in the UNR crop. Hence, the similar total final biomass of the two systems is a consequence of two compensating factors. Limitations in assimilates to individual fruit in the UNR crops indicated by a lower RUEg and increased shading of the lower part of the canopy may also explain why boll size was smaller in the UNR treatments as boll size is related closely to carbohydrate supply, especially from nearby leaves. The high plant densities in this study in the UNR crop may be too high for optimal plant growth. Further studies are needed to determine whether a lower plant density might increase crop growth and biomass production in high yielding UNR crops.
Acknowledgments This work was partly funded by the Cotton Research and Development Corporation. We gratefully acknowledge Mr. Darin Hodgson, Mrs. Jane Caton and Ms. Joanne Price for their technical assistance; Drs Greg Constable and Warren Conaty for helpful discussions on this manuscript.
93
References Andrade, F.H., Calvino, P., Cirilo, A., Barbieri, P., 2002. Yield responses to narrow rows depend on increased radiation interception. Agron. J. 94, 975–980. Andries, J.A., Jones, J.E., Sloane, L.W., Marshall, J.G., 1969. Effects of okra leaf shape on boll rot, yield, and other important characters of upland cotton, Gossypium hirsutum L. Crop Sci. 9, 705–710. Baker, S.H., 1976. Response of cotton to row patterns and plant populations. Agron. J. 68, 85–88. Bange, M.P., Milroy, S.P., 2004. Growth and dry matter partitioning of diverse cotton genotypes. Field Crops Res. 87, 73–87. Board, J.E., Harville, B.G., 1992. Explanations for greater light interception in narrowvs wide-row soybean. Crop Sci. 32, 198–202. Board, J.E., Harville, B.G., Kamal, M., 1994. Radiation-use efficiency in relation to row spacing for late-planted Soybean. Field Crops Res. 36, 13–19. Brodrick, R., Bange, M.P., Milroy, S.P., Hammer, G.L., 2010. Yield and maturity of ultranarrow row cotton in high input production systems. Agron. J. 102, 843–848. Brodrick, R., Bange, M.P., Milroy, S.P., Hammer, G.L. Physiological determinants of high yielding ultra-narrow row cotton:biomass accumulation and partitioning. Fields Crop Res., http://dx.doi.org/10.1016/j.fcr.2012.05.007, in press. Charles-Edwards, D.A., Doley, D., Rimmington, G.M., 1986. Modeling Plant Growth and Development. Academic Press, Sydney. Charles-Edwards, D.A., Lawn, R.J., 1984. Light interception by grain legume row crops. Plant Cell Environ. 7, 247–251. Coleman, J.S., McConnaughay, K.D.M., Ackerly, D.D., 1994. Interpreting phenotypic variation in plants. Trends Ecol. Evol. 9, 187–191. Constable, G.A., 1975. Growth, development and yield of cotton as influenced by cultivar and row spacing. Masters Thesis. Faculty of Agriculture, University of Sydney, Sydney, p. 175. Constable, G.A., 1981. Carbon fixation and distribution in cotton:implications of single leaf measurements to plant performance. PhD Thesis. Dept of Environmental Biology, The Australian National University, Canberra, p. 303. Constable, G.A., 1986. Growth and light receipt by mainstem cotton leaves in relation to plant density in the field. Agric. For. Meteorol. 37, 279–292. Constable, G.A., Rawson, H.M., 1982. Distribution of 14 C label from cotton leaves:consequences of changed water and nitrogen status. Aust. J. Plant Physiol. 9, 735–747. Constable, G.A., Rochester, I.J., Hodgson, A.S., 1990. A comparison of drip and furrow irrigated cotton on a cracking clay soil 1. Growth and nitrogen uptake. Irrigation Sci. 11, 137–142. Darawsheh, M.K., Khah, E.M., Aivalakis, G., Chachalis, D., Sallaku, F., 2009. Cotton row spacing and plant density cropping systems I. Effects on accumulation and partitioning of dry mass and LAI. J. Food Agric. Environ. 7, 258–261. Duncan, W.G., Loomis, R.S., Williams, W.A., Hanau, R., 1967. A model for simulating photosynthesis in plant communities. Hilgardia 38, 181–205. Flenet, F., Kiniry, J.R., Board, J.E., Westgate, M.E., Reicosky, D.C., 1996. Row spacing effects on light extinction coefficients of corn, sorghum, soybean, and sunflower. Agron. J. 88, 185–190. Forcella, F., Westgate, M.E., Warnes, D.D., 1992. Effects of row width on herbicide and cultivation requirements of row crops. Am. J. Altern. Agric. 7, 161–167. Gwathmey, C.O., Clement, J.D., 2010. Alteration of cotton source-sink relations with plant population density and mepiquat chloride. Field Crops Res. 116, 101–107. Hearn, A.B., Fitt, G.P., 1992. Cotton cropping systems. In: Pearson, C.J. (Ed.), Field Crop Ecosystems. Elsevier, Amsterdam, pp. 85–142. Heitholt, J.J., Pettigrew, W., Meredith, W., 1992. Light interception and lint yield on narrow-row cotton. Crop Sci. 32, 728–733. Jost, P.H., Cothren, J.T., 2000. Growth and yield comparisons of cotton planted in conventional and ultra-narrow row spacings. Crop Sci. 40, 430–435. Jost, P.H., Cothren, J.T., 2001. Phenotypic alterations and crop maturity differences in ultra-narrow row and conventionally spaced cotton. Crop Sci. 41, 1150–1159. Kreig, D.R., 1996. Physiological aspects of ultra narrow row cotton production. In: Dugger, P., Richter, D.A. (Eds.), Proc. Beltwide Cotton Conf. Nashville, TN, January 9–12. Natl. Cotton Counc. Am., Memphis, TN, p. 66. Leach, G.J., Beech, D.F., 1988. Response of chickpea accessions to row spacing and plant density on a vertisol on the Darling Downs, south-eastern Queensland. 2. Radiation interception and water use. Aust. J. Exp. Agric. 28, 377–383. Marois, J.J., Wright, D.W., Wiatrak, P.J., Vargas, M.A., 2004. Effect of row width and nitrogen on cotton morphology and canopy microclimate. Crop Sci. 44, 870–877. Milroy, S.P., Bange, M.P., 2003. Nitrogen and light responses of cotton photosynthesis and implications for crop growth. Crop Sci. 43, 904–913. Milroy, S.P., Bange, M.P., Sadras, V.O., 2001. Profiles of leaf nitrogen and light in reproductive canopies of cotton (Gossypium hirsutum). Ann. Bot. 87, 325–333. Monteith, J.L., 1977. Climate and the efficiency of crop production. Br. Philos. Trans. R. Soc. Lond. Ser. B 218, 277–297. Muchow, R.C., 1988. Effect of nitrogen supply on the comparative productivity of maize and sorghum in a semi-arid tropical environment. I. Leaf growth and leaf nitrogen. Field Crops Res. 18, 1–16. Nichols, S.P., Snipes, C.E., Jones, M.A., 2003. Evaluation of row spacing and mepiquat chloride on cotton. J. Cotton Sci. 7, 148–155. Nunez, R., Kamprath, E., 1969. Relationships between N response, plant population and row width on growth and yield of corn. Agron. J. 61, 279–282. Peng, S., Krieg, D.R., 1991. Single leaf canopy photosynthesis response to plant age in cotton. Agron. J. 83, 704–708. Reid, P., 2001. Sicala V-3RRi. Plant Var. J. 14, 39–40.
94
R. Brodrick et al. / Field Crops Research 148 (2013) 86–94
Rosenthal, W.D., Gerik, T.J., 1991. Radiation use efficiency among cotton cultivars. Agron. J. 83, 655–658. Sadras, V.O., 1996. Cotton responses to simulated insect damage:radiation-use efficiency, canopy architecture and leaf nitrogen content as affected by loss of reproductive organs. Field Crops Res. 48, 199–208. Savoy, B.R., Cothren, J.T., Shumway, C.R., 1992. Soybean biomass accumulation and leaf area index in early-season production environments. Agron. J. 84, 956–959. Shibles, R., Weber, C.R., 1966. Interception of solar radiation and dry matter production by various planting patterns. Crop Sci. 6, 55–59.
Sinclair, T.R., Pinter Jr., Kimbali, P.J., Adamsen, B.A., LaMorte, F.J., Wall, R.L., Hunsaker, G.W., Adam, D.J., Brooks, N., Garcia, T.J., Thompson, R.L., Leavitt, T., Matthias, S.A., 2000. Leaf nitrogen concentration of wheat subjected to elevated CO2 and either water or N deficits. Agric. Ecosyst. Environ. 79, 53–60. Zhao, D., Oosterhuis, D., 2000. Cotton responses to shade at different growth stages:growth, lint yield and fibre quality. Exp. Agric. 36, 27–39.
Contents lists available at SciVerse ScienceDirect
Field Crops Research journal homepage: www.elsevier.com/locate/fcr
Physiological determinants of high yielding ultra-narrow row cotton: Canopy development and radiation use efficiency R. Brodrick a,c,∗ , M.P. Bange a , S.P. Milroy b , G.L. Hammer c a b c
CSIRO Plant Industry, Cotton Catchment Communities Cooperative Research Centre, LMB 59, Narrabri, NSW 2390, Australia CSIRO Plant Industry, Centre for Environment and Life Sciences, Private Bag 5, Wembley, WA 6913, Australia School of Land and Food Sciences, the University of Queensland, Brisbane, QLD 4072, Australia
a r t i c l e
i n f o
Article history: Received 16 May 2011 Received in revised form 15 May 2012 Accepted 16 May 2012 Keywords: Gossypium hirsutum Leaf area index Light interception Canopy extinction coefficient Radiation use efficiency Row spacing
a b s t r a c t Ultra-narrow row cotton (UNR, with rows spaced less than 40 cm apart) has long been proposed to have the potential to increase yields while reducing the time to crop maturity. Investigations have shown that biomass accumulation in high-input, high yielding UNR (25 cm spaced rows with yields greater than 1800 kg lint ha−1 ) cotton is similar to conventionally spaced rows (100 cm) despite a three-fold increase in plant density, indicating a limitation on individual plant growth. This study investigates whether the increased plant density in UNR crops (36 plants m−2 ) leads to differences in canopy development, radiation use efficiency (RUE) and light interception contributing to plant growth limitations. Three experiments over three years compared UNR treatments to conventionally spaced treatments in high-input production systems and found that early canopy development (leaf area index LAI) and consequently early interception was higher in the UNR crops in two of the three experiments. This resulted in a 17% higher seasonal canopy extinction coefficient (k) in UNR crops over the season. However, seasonal RUE of the UNR crop was 19% lower as increases in light interception were not accompanied by increased total dry matter. Light distribution through the canopy was poorer (higher k) in the UNR crop and LAI continued to increase in the UNR crop after maximum light interception was reached, which combined with a lower leaf nitrogen concentration may have reduced the photosynthetic efficiency of the UNR crop. We conclude that differences in canopy light interception and the efficiency of conversion of light to biomass were the primary factors responsible for differences in the pattern of biomass accumulation between UNR and conventionally spaced cotton. © 2012 Elsevier B.V. All rights reserved.
1. Introduction Ultra-narrow row cotton (UNR, with rows spaced less than 40 cm apart) has long been proposed to have the potential to increase yields while reducing the time to crop maturity (Nichols et al., 2003). This paper forms part of a detailed investigation into the growth and development of ultra-narrow row cotton in high-input production systems. We found that there was potential for higher yields when compared to conventionally spaced rows (100 cm row spacing) in high yielding, high-input productions systems in Australia, however there was no difference between row spacings in time to crop maturity (Brodrick et al., 2010). A detailed growth analysis and physiological determinants framework (Charles-Edwards et al., 1986; Coleman et al., 1994) was used to identify key differences in the factors influencing yield and
∗ Corresponding author at: CSIRO Plant Industry, Cotton Catchment Communities Cooperative Research Centre, LMB 59, Narrabri, NSW 2390, Australia. Tel.: +61 2 6799 1500; fax: +61 2 6793 1186. E-mail address: [email protected] (R. Brodrick). 0378-4290/$ – see front matter © 2012 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.fcr.2012.05.008
maturity of UNR and conventionally spaced cotton crops in highinput, high yielding production systems. Brodrick et al. (in press) found that the increased yield in UNR crops was not due to differences in final biomass accumulation but associated with more bolls per unit area and increased partitioning to reproductive growth in the UNR crops. Initially crop growth rate and biomass production was higher but crop growth rate slowed earlier in the UNR crop and final biomass was not different from the conventionally spaced crop despite a three-fold increase in plant density (Brodrick et al., in press). The UNR plants were smaller, with less biomass produced per plant indicating limitations in assimilates for growth due to the increased number of plants competing for resources in the UNR crop. If competition between plants in the UNR treatments reduced the availability of light, water and nutrients to individual plants, this would have limited their biomass production and growth, and as total crop water use and final nutrient uptake were not different (Brodrick et al., in press), it is likely that competition for light in the higher density UNR crop slowed biomass accumulation. Few studies have compared the leaf area development and light interception characteristics of cotton grown in UNR and conventionally spaced rows in low-input systems, and none to our
R. Brodrick et al. / Field Crops Research 148 (2013) 86–94
knowledge in high-input, high yielding (>1800 kg lint ha−1 ) systems. In low-input UNR cotton systems the higher number of plants can lead to greater early leaf area index (LAI) accumulation compared to conventionally spaced cotton (Darawsheh et al., 2009; Gwathmey and Clement, 2010; Jost and Cothren, 2000; Kreig, 1996). Closer plant spacing means that plants do not need to be as large for the crop to achieve maximum light interception. An increase in light interception has been reported for narrower row spacings in cotton (Heitholt et al., 1992; Peng and Krieg, 1991) and other species such as corn (Andrade et al., 2002), chickpea (Leach and Beech, 1988), sorghum (Flenet et al., 1996), and soybean (Board and Harville, 1992; Savoy et al., 1992). This rapid canopy closure may also lead to reductions in weed competition (Forcella et al., 1992) and decreased soil evaporation (Kreig, 1996; Nunez and Kamprath, 1969). This increased early light interception has been thought by many researchers to be the primary reason for increases in yield in narrower row spacings in many indeterminate and determinate crops (Andrade et al., 2002; Flenet et al., 1996; Shibles and Weber, 1966). Constable (1975) found a significant negative relationship between boll growth rate and LAI in UNR spaced cotton. Heitholt et al. (1992) found that narrow rows (0.5 m spaced rows) had a higher canopy extinction coefficient (k) compared to conventionally spaced rows and concluded that the optimal LAI for maximum light interception and yield for narrow rows was between 3 and 4 compared with an LAI of 4–5 for conventionally spaced rows. Other authors have hypothesized that the denser canopy in UNR spaced cotton could lead to reduced light penetration leading to increased shedding or smaller bolls (Baker, 1976; Constable, 1975; Jost and Cothren, 2001). Changes in biomass production can also be a result of differences in a crop’s ability to convert solar radiation into biomass, as represented by the radiation use efficiency (RUE) of the crop (Monteith, 1977). No studies have compared radiation use efficiency in UNR to conventionally spaced rows in cotton. Savoy et al. (1992) found that narrower rows in soybean (0.36 m spaced rows) had higher light interception, greater biomass accumulation and higher radiationuse-efficiency compared to wide rows (1.02 m spaced rows). Understanding how decreasing row spacing in high-yielding, high-input cotton systems affects the physiological determinants of biomass production is important in realizing any benefits of these systems. In this paper we will test the following hypotheses, (i) the increased plant population in the UNR crop increases light interception in the UNR crop; (ii) canopy development leads to a higher canopy extinction coefficient in the UNR crop and (iii) radiation use efficiency is lower resulting in no differences in final biomass production in the UNR crop.
Experiments were fully irrigated according to crop requirements. Management followed current commercial practices for high-input management, irrigation and insect control as described by Hearn and Fitt (1992). Crop water use and nitrogen uptake details are reported in Brodrick et al. (in press). 2.2. Measurements 2.2.1. Leaf area index On approximately 12 occasions in each experiment, biomass, leaf area, specific leaf area (SLA) and LAI were measured by destructive sampling (see Brodrick et al., in press). Leaf area was determined by measuring the leaf area of the sub-sample with a LiCor planimeter (Model LI-3100, LiCor Inc., Lincoln, NB, USA). This sample was dried and weighed and specific leaf area determined (m2 g−1 ). LAI was calculated as the product of specific leaf area and amount of leaf dry matter (g m−2 ). 2.2.2. Light interception Total daily incoming radiation was measured using a calibrated pyranometer at the Australian Cotton Research Institute weather station less than 2 km from the experimental fields. In Exp. 1, solar radiation intercepted by the canopies was measured using tube solarimeters (Model TSL Delta-T Devices Ltd, Cambridge, UK). A single tube solarimeter was placed across one bed in each plot (in a north–south orientation) to measure transmitted radiation. One tube solarimeter was placed above the crop in the middle of the experiment to measure incident solar radiation. The solarimeters were calibrated against the solarimeter positioned above the crop before and after each experiment. The solarimeters were programmed to scan at 5-min intervals, recording average hourly readings on a programmable datalogger (Model DL Delta-T Devices Ltd, Cambridge, UK). In Exps. 2 and 3 solarimeter data were not collected for the full season or for all plots due to problems with dataloggers. For these experiments intercepted solar radiation was calculated from weekly measurements of intercepted photosynthetically active radiation (PAR) using a sunfle0ck ceptometer (SF-80, DeltaT Devices Ltd, Cambridge, UK). Incident radiation was recorded between 11:00 and 13:00 h (Australian Eastern Standard Time) above each plot averaging three readings. Transmitted radiation was recorded by average readings taken at ground level in three random areas in each plot from the center of the furrow to the center of the bed. Solarimeter data from experiment 1 was converted to PAR. To estimate canopy closure, the proportion of PAR intercepted by the crop at midday (LII ) was calculated as: LII =
2. Methods 2.1. Site and climate description UNR and conventionally spaced cotton crops were compared in three experiments grown near Narrabri, in a semi-arid environment of north-west New South Wales, Australia. The experiments are described in detail in Brodrick et al. (in press). Briefly, three experiments were sown using the cultivar Sicala V-3RRi (Reid, 2001). Exp. 1 was sown 16 November 2001, Exp. 2 was sown 10 October 2002, and Exp. 3 was sown 23 October 2003. These experiments correspond with Exps. 1, 2 and 5 in Brodrick et al. (2010). In the 25 cm UNR treatment, the row configuration was six rows spaced 25 cm apart on a 2 m bed sown with 36 plants m−2 . In the conventionally spaced treatment, the row configuration was two rows spaced 1 m apart on a 2 m bed sown with 12 plants m−2 . A randomized complete block design with four replicates was used.
87
(incident radiation − transmitted radiation) incident radiation
An exponential function was fitted to LII over DAS to allow interpolation between measurement dates:
LII = a 1 − e(−bDAS) + c where a, b and c are fitted coefficients (Charles-Edwards and Lawn, 1984). To calculate cumulative intercepted daily radiation and the light extinction coefficient (k), total daily intercepted radiation (LID ) was calculated from instantaneous measurements by adjusting the measurements using the relationship (Charles-Edwards et al., 1986): LID =
2LII 1 + LII
To allow interpolation between dates of measurement LID was also regressed over DAS using the same equation as for LII .
88
R. Brodrick et al. / Field Crops Research 148 (2013) 86–94
Cumulative intercepted solar radiation (CLID ) was calculated using total incident daily radiation (LID ) and the measured daily proportion intercepted for the period of measurement in each experiment. For each experiment, CLID was calculated up to the biomass harvest with the highest average LAI, after which LAI began to drop off due to leaf senescence (Exp. 1 – 118 DAS, Exp. 2 – 123 DAS and Exp. 3 – 112 DAS). In all experiments, this period covered the period of maximum growth and light interception. 2.2.3. Light extinction coefficient The light extinction coefficient (k) is a parameter that indicates the effectiveness of a crop canopy at intercepting PAR (CharlesEdwards et al., 1986). The light extinction coefficient was derived for the whole season by regressing light interception (LID ) on LAI for each plot, using a modified form of Beer’s Law:
LID = a 1 − e−kLAI
where LID is the proportion of intercepted PAR; k is the light extinction coefficient; and a represents the maximum value of light interception that can be attained by the crop canopy. This analysis, however, assumes that light interception characteristics are constant throughout crop development, thereby ignoring changes in leaf angle, the condition of leaf surface and the overall canopy structure (Charles-Edwards et al., 1986). 2.3. Radiation use efficiency Seasonal crop radiation use efficiency (RUEg ) (g MJ−1 ) was derived from the gradient of the linear regression of accumulated total dry matter (glucose equivalent) and cumulative intercepted total solar radiation (CLID ) (Monteith, 1977). RUEg between harvest dates was calculated by a similar linear regression for each time period. 2.4. Specific leaf nitrogen Dried leaf samples for each plot from each biomass harvest were ground, mixed, and analyzed for nitrogen content using Kjeldahl digestion. Specific leaf nitrogen (SLN (g N/m−2 per unit leaf area)) was calculated as the quotient of leaf nitrogen content and specific leaf area (Muchow, 1988). SLN for UNR treatments was then plotted against SLN for conventionally spaced treatments using data from all experiments. 2.5. Data analysis To test for differences between UNR and conventionally spaced systems a combined analysis across all experiments was undertaken using generalized linear modeling (GLM). In this analysis the main factors were row spacing and experiment (Exp. × row spacing treatment), and the random factors replicate and experiment (Exp. × replicate). For those variables that differed in their response across experiments an analysis of variance (ANOVA) for a randomized block design was used for comparing row spacings. Simple linear regression analysis was performed to test for differences between row spacings in the regressions for k and RUEg . To test for differences between SLN of the two treatments the linear relationship was compared against unity (1:1 line). All statistical analyses were conducted using Genstat software (VSN International, Rothamsted, UK). Unless stated otherwise significant differences were considered at 95% confidence intervals (P < 0.05). Where shown graphically, the standard errors are ±one standard error of the treatment means from the associated ANOVA. In some data sets, the means of all harvests are presented
Fig. 1. Cumulative rainfall over the growing season for Exps. 1 to 3.
graphically, however an individual ANOVA was performed for each harvest date. 3. Results 3.1. Climatic conditions More detail about the climatic conditions over the three years of experiments is presented in Brodrick et al. (in press). Exps. 1, 2 and 3 experienced similar temperature patterns resulting in similar cumulative day degrees at the end of the growing season. The later sowing of Exp. 1 meant that day degree and solar radiation accumulation was slightly slower at the end of the season. Early in-crop rainfall was extremely low in Exp. 2 (Fig. 1). Total in-crop rainfall was lower in Exps. 1 and 2 compared with Exp. 3; however, 155 mm of the rainfall during Exp. 3 occurred in a single 5-day period during early flowering. 3.2. Leaf area index Responses of LAI to row spacing were not consistent across the three experiments (Fig. 2). Early LAI (prior to 100 days after sowing (DAS)) tended to be significantly higher in the UNR spaced treatments compared to the conventionally spaced treatments in Exps. 1 and 3 but in Exp. 2 there were no significant differences in LAI at any of the harvest dates. 3.3. Light interception The proportion of intercepted photosynthetically active radiation (LII ) was derived from measurements of PAR over the period of greatest growth (Exp. 1 – 118 DAS, Exp. 2 – 123 DAS and Exp. 3 – 112 DAS) (Fig. 2). In all three experiments, the UNR treatments reached 80% light interception (approximation of canopy closure) before the conventionally spaced treatments (Fig. 2). The UNR treatment reached 80% LII 35 days earlier than the conventional spacing in Exp. 1 and 11 days earlier in Exp. 3. In Exp. 2, however, there was not a significant difference between the row spacings. 3.4. Crop light extinction coefficient A combined analysis of the relationship between LII and LAI showed a significantly higher k in the UNR treatments compared to conventionally spaced treatments across all three experiments (Figs. 3 and 4; Table 1).
R. Brodrick et al. / Field Crops Research 148 (2013) 86–94
89
Fig. 2. Mean leaf area index (LAI) versus days after sowing for ultra-narrow rows (UNR) and conventionally spaced treatments in Exps. 1 (a), 2 (b) and 3 (c). Proportion of intercepted photosynthetically active radiation (LII ) versus days after sowing for UNR (solid line) and conventionally spaced (broken line) treatments in Exps. 1 (d), 2 (e) and 3 (f). Error bars are ± one standard error of the mean. * = 95% significance level; ** = 99% significance level. Lines are included to assist comparison between treatments.
Table 1 Summary of significant differences from combined analysis of experiments for radiation use efficiency (RUEg ) and the crop light extinction coefficient (k). Period of measurement for Exp. 1 was to 118 DAS, Exp. 2 to 123 DAS and Exp. 3 to 112 DAS. Error df = 13. Variable
UNR
Conventionally spaced
RUEg (g MJ−1 ) k
1.03 0.81
1.23 0.69
n.s. = no significant difference.
P-value Row spacing
P-value Experiment
P-value Experiment × row spacing
df = 1 <0.001 <0.001
df = 2 <0.001 n.s.
df = 2 n.s. n.s.
90
R. Brodrick et al. / Field Crops Research 148 (2013) 86–94
Fig. 4. Relationship between light interception and leaf area index for ultra-narrow rows (UNR) (solid line) and conventionally spaced (broken line) treatments for all data from Exps. 1, 2 and 3.
3.6. Specific leaf area and specific leaf nitrogen
Fig. 3. Relationship between light interception and leaf area index for ultra-narrow rows (UNR) (solid line) and conventionally spaced (broken line) treatments for Exps. 1 (a), 2 (b) and 3 (c).
There were few significant differences in SLA between row spacing experiments (Fig. 7). In Exp. 1, SLA in the UNR treatments was significantly higher than the conventionally spaced treatments at 35 DAS and significantly lower at 59 DAS (Fig. 7a). The only significant difference in SLA in Exp. 2 was significantly lower in the UNR treatments compared to the conventionally spaced treatments at 55 DAS (Fig. 7b). In Exp. 3, SLA was significantly lower in the UNR treatments compared to the conventionally spaced treatments at 54 and 60 DAS (Fig. 7c). Differences in specific leaf nitrogen (SLN) between row spacings were not consistent across the three experiments (Fig. 7d–f). Although SLN was generally numerically higher in the conventionally spaced treatments, there were few significant differences. SLN in individual plots in the conventionally spaced treatments was regressed against SLN in UNR plots. In Exps. 1 and 3, the majority of data lay above the 1:1 line, reflecting the numerically higher observation seen in the plots against DAS (Fig. 7g–i). The intercept was significantly greater than zero, indicating that when the canopies had a low N status, canopies of conventionally spaced crops had higher SLN. However SLN of the UNR crops was more responsive when leaf N status was generally higher (Fig. 7g–i)
4. Discussion 3.5. Radiation use efficiency A seasonal crop radiation use efficiency (RUEg ) was derived from the slope of the linear regression of cumulative total dry matter versus cumulative intercepted total solar radiation (Fig. 5). RUEg was significantly lower in the UNR treatment compared to the conventionally spaced treatment in Exp. 1, but there was no significant difference between treatments in Exps. 2 and 3 (Fig. 5). A combined analysis of RUEg across the three experiments showed significantly lower RUEg in the UNR treatments compared to the conventionally spaced treatments (Table 1). There were few significant differences in RUEg calculated for the intervals between sampling dates (Fig. 6).
4.1. Leaf area index, light interception and light extinction coefficient were greater in UNR Early canopy development and consequently light interception was higher in the UNR crops in two of the three experiments but this did not lead to a significant difference in total cumulative intercepted solar radiation during the period of measurement compared with the conventionally spaced crops. Early canopy development and light interception followed a similar pattern to biomass accumulation and crop growth rate (Brodrick et al., in press). This pattern of early light interception and higher crop growth rate in UNR is consistent with the results of Kreig (1996).
R. Brodrick et al. / Field Crops Research 148 (2013) 86–94
91
Fig. 5. Relationship between total dry matter and cumulative intercepted solar radiation for ultra-narrow rows (UNR, solid line) and conventionally spaced (broken line) treatments in Exps. 1 (a), 2 (b) and 3 (c).
There was very little early in-crop rainfall in Exp. 2 and leaf expansion and canopy development were much lower in both treatments. This appears to be a direct effect of water stress (Constable and Rawson, 1982) on leaf expansion as early SLA was lower and SLN tended to be higher during early development in this experiment. In addition, SLN of the two treatments did not differ in this experiment, suggesting that expansion in the UNR crop was more strongly affected than in the conventionally spaced crop. LAI in the UNR crops continued to develop after maximum LII had been reached, whereas in the conventionally spaced treatments peak LAI was more aligned with maximum LII. Elevated LAIs can be detrimental to the crop in a number of ways. Low light intensity in the lower canopy reduces photosynthetic rates of those leaves and hence assimilate production to support boll development on the associated branches (Constable and Rawson, 1982;
Fig. 6. Mean radiation use efficiency (RUE) versus days after sowing for ultra-narrow rows (UNR) and conventionally spaced treatments in Exps. 1 (a), 2 (b) and 3 (c). * = 95% significance level; ** = 99% significance level. Error bars are ± one standard error of the mean. Lines are included to assist comparison between treatments.
Zhao and Oosterhuis, 2000). This may have resulted the lower fruit retention and smaller boll size measured in the UNR crops (Brodrick et al., 2010). In addition, there is evidence that changes in light intensity and quality can influence boll shedding directly (Board et al., 1994; Constable, 1981). Beyond the physiological influences, excessive LAI impedes penetration of insecticides and has been shown to have negative effects on insect control (Marois et al., 2004) and increases in boll rot (Andries et al., 1969).
92
R. Brodrick et al. / Field Crops Research 148 (2013) 86–94
Fig. 7. Mean specific leaf area (SLA) and mean specific leaf nitrogen (SLN) versus days after sowing for ultra-narrow rows (UNR) and conventionally spaced treatments in Exps. 1 (a and d), 2 (b and e) and 3 (c and f). Comparison of SLN between UNR and conventionally spaced treatments in Exps. 1 (g), 2 (h) and 3 (i). To test for differences between SLN of the two treatments the relationship was compared against unity (1:1 line). * = 95% significance level; ** = 99% significance level. Error bars are ± one standard error of the mean. Lines are included to assist comparison between treatments.
Through both increased plant density and reduced row spacings the UNR (36 plants m−2 ) and conventionally spaced (12 plants m−2 ) treatments have different spatial arrangements and plant structure, difference in the light extinction coefficient (k) is likely due to differences in morphology and arrangement of plants. A higher k in the UNR crop, and hence, greater light capture at low LAI, did not increase final total biomass production most likely because of reduction in RUEg . A dense canopy with overlapping leaves may mean that although there was more light intercepted per unit LAI in the UNR treatments, the vertical distribution of light in the canopy was not uniform and likely result in more light being intercepted by the upper part of the canopy (Constable, 1986). Higher k generates less uniform light distribution in the canopy so that overall conversion efficiency is reduced, especially at high LAI (Duncan et al., 1967). In the UNR crop, with LAI continuing to increase after maximum light interception had been reached, the higher k would have only been of benefit to biomass accumulation of the UNR crops while light interception was still increasing.
4.2. Radiation use efficiency was lower in the UNR crop Lower seasonal RUEg indicated that the UNR treatment was less efficient in converting intercepted solar radiation into biomass production. The UNR canopy had higher early LAI and high light interception in Exps. 1 and 3 but no differences in total dry matter production. An average RUEg of 1.0 g MJ−1 for UNR crops when converted to total dry matter is 0.54 g MJ−1 and is at the lower end of the range of reported values from 0.62 to 0.75 g MJ−1 (converted to solar radiation if the authors measured PAR) for seasonally measured RUE in cotton (Bange and Milroy, 2004; Constable et al., 1990; Rosenthal and Gerik, 1991; Sadras, 1996). The RUEg in the UNR crop could be lower due to the poor light distribution through the canopy (higher seasonal k) or due to the lower canopy SLN in the UNR crop reducing photosynthesis and ultimately biomass production; or a combination of both. The UNR crop may have had less light penetrating to the lower parts of the canopy due to the combination of higher LAI and higher
R. Brodrick et al. / Field Crops Research 148 (2013) 86–94
k. At high LAI, canopies with better light distribution through the canopy (lower k) have better photosynthetic efficiency and hence increased RUEg because more of the leaf surface in the canopy is intercepting light (Charles-Edwards et al., 1986; Sadras, 1996). Radiation use efficiency is closely linked to leaf N concentration (Sinclair et al., 2000; Sadras, 1996; Milroy and Bange, 2003) because of the dependence of photosynthesis on leaf N. The lower canopy SLN may also be as a result of poorer light distribution in the UNR canopies; research for conventionally spaced cotton has found that SLN is positively correlated with light intensity throughout the canopy (Milroy et al., 2001). However, in Exp. 1 there are indications that SLN played a role in the lower RUEg in the UNR crop, as both SLN and RUEg were significantly lower before 60 DAS when LAI and light interception would not have been influencing distribution of SLN in the canopy. As SLA was not different between row spacings, lower SLN in the UNR crop indicates that there is a dilution of N in the leaves, and may suggest there was greater demand for N in the UNR crop, although total N uptake remained the same (Brodrick et al., in press). In addition to the UNR crop having lower RUEg and SLN accompanied by higher k, one of the experiments in this study (Exp. 2) indicated that the UNR crop may also be more susceptible to early water stress. Investigations are continuing to determine whether changing canopy arrangement through different plant populations or rowing spacings; or increasing early nitrogen and water to increase SLN or mitigate any early water stress; can alter leaf area development, light interception and radiation use efficiency in UNR crops.
5. Conclusion The results of this study suggest that differences in canopy light interception characteristics and the efficiency of conversion of light to biomass may be the primary factors responsible for differences in the pattern of biomass accumulation between high yielding UNR and conventionally spaced crops. Early canopy development (LAI) and consequently early light interception was higher in the UNR crops in two of the three experiments. This resulted in a 17% higher canopy extinction coefficient (k) in the UNR crops over the season. However as total dry matter was not higher despite increased light interception, RUEg was lower in the UNR crop. The 19% lower RUEg in the UNR crop was most likely due to poor light distribution through the canopy, combined with lower leaf nitrogen concentration, reducing the photosynthetic efficiency of the leaves in the UNR crop. Hence, the similar total final biomass of the two systems is a consequence of two compensating factors. Limitations in assimilates to individual fruit in the UNR crops indicated by a lower RUEg and increased shading of the lower part of the canopy may also explain why boll size was smaller in the UNR treatments as boll size is related closely to carbohydrate supply, especially from nearby leaves. The high plant densities in this study in the UNR crop may be too high for optimal plant growth. Further studies are needed to determine whether a lower plant density might increase crop growth and biomass production in high yielding UNR crops.
Acknowledgments This work was partly funded by the Cotton Research and Development Corporation. We gratefully acknowledge Mr. Darin Hodgson, Mrs. Jane Caton and Ms. Joanne Price for their technical assistance; Drs Greg Constable and Warren Conaty for helpful discussions on this manuscript.
93
References Andrade, F.H., Calvino, P., Cirilo, A., Barbieri, P., 2002. Yield responses to narrow rows depend on increased radiation interception. Agron. J. 94, 975–980. Andries, J.A., Jones, J.E., Sloane, L.W., Marshall, J.G., 1969. Effects of okra leaf shape on boll rot, yield, and other important characters of upland cotton, Gossypium hirsutum L. Crop Sci. 9, 705–710. Baker, S.H., 1976. Response of cotton to row patterns and plant populations. Agron. J. 68, 85–88. Bange, M.P., Milroy, S.P., 2004. Growth and dry matter partitioning of diverse cotton genotypes. Field Crops Res. 87, 73–87. Board, J.E., Harville, B.G., 1992. Explanations for greater light interception in narrowvs wide-row soybean. Crop Sci. 32, 198–202. Board, J.E., Harville, B.G., Kamal, M., 1994. Radiation-use efficiency in relation to row spacing for late-planted Soybean. Field Crops Res. 36, 13–19. Brodrick, R., Bange, M.P., Milroy, S.P., Hammer, G.L., 2010. Yield and maturity of ultranarrow row cotton in high input production systems. Agron. J. 102, 843–848. Brodrick, R., Bange, M.P., Milroy, S.P., Hammer, G.L. Physiological determinants of high yielding ultra-narrow row cotton:biomass accumulation and partitioning. Fields Crop Res., http://dx.doi.org/10.1016/j.fcr.2012.05.007, in press. Charles-Edwards, D.A., Doley, D., Rimmington, G.M., 1986. Modeling Plant Growth and Development. Academic Press, Sydney. Charles-Edwards, D.A., Lawn, R.J., 1984. Light interception by grain legume row crops. Plant Cell Environ. 7, 247–251. Coleman, J.S., McConnaughay, K.D.M., Ackerly, D.D., 1994. Interpreting phenotypic variation in plants. Trends Ecol. Evol. 9, 187–191. Constable, G.A., 1975. Growth, development and yield of cotton as influenced by cultivar and row spacing. Masters Thesis. Faculty of Agriculture, University of Sydney, Sydney, p. 175. Constable, G.A., 1981. Carbon fixation and distribution in cotton:implications of single leaf measurements to plant performance. PhD Thesis. Dept of Environmental Biology, The Australian National University, Canberra, p. 303. Constable, G.A., 1986. Growth and light receipt by mainstem cotton leaves in relation to plant density in the field. Agric. For. Meteorol. 37, 279–292. Constable, G.A., Rawson, H.M., 1982. Distribution of 14 C label from cotton leaves:consequences of changed water and nitrogen status. Aust. J. Plant Physiol. 9, 735–747. Constable, G.A., Rochester, I.J., Hodgson, A.S., 1990. A comparison of drip and furrow irrigated cotton on a cracking clay soil 1. Growth and nitrogen uptake. Irrigation Sci. 11, 137–142. Darawsheh, M.K., Khah, E.M., Aivalakis, G., Chachalis, D., Sallaku, F., 2009. Cotton row spacing and plant density cropping systems I. Effects on accumulation and partitioning of dry mass and LAI. J. Food Agric. Environ. 7, 258–261. Duncan, W.G., Loomis, R.S., Williams, W.A., Hanau, R., 1967. A model for simulating photosynthesis in plant communities. Hilgardia 38, 181–205. Flenet, F., Kiniry, J.R., Board, J.E., Westgate, M.E., Reicosky, D.C., 1996. Row spacing effects on light extinction coefficients of corn, sorghum, soybean, and sunflower. Agron. J. 88, 185–190. Forcella, F., Westgate, M.E., Warnes, D.D., 1992. Effects of row width on herbicide and cultivation requirements of row crops. Am. J. Altern. Agric. 7, 161–167. Gwathmey, C.O., Clement, J.D., 2010. Alteration of cotton source-sink relations with plant population density and mepiquat chloride. Field Crops Res. 116, 101–107. Hearn, A.B., Fitt, G.P., 1992. Cotton cropping systems. In: Pearson, C.J. (Ed.), Field Crop Ecosystems. Elsevier, Amsterdam, pp. 85–142. Heitholt, J.J., Pettigrew, W., Meredith, W., 1992. Light interception and lint yield on narrow-row cotton. Crop Sci. 32, 728–733. Jost, P.H., Cothren, J.T., 2000. Growth and yield comparisons of cotton planted in conventional and ultra-narrow row spacings. Crop Sci. 40, 430–435. Jost, P.H., Cothren, J.T., 2001. Phenotypic alterations and crop maturity differences in ultra-narrow row and conventionally spaced cotton. Crop Sci. 41, 1150–1159. Kreig, D.R., 1996. Physiological aspects of ultra narrow row cotton production. In: Dugger, P., Richter, D.A. (Eds.), Proc. Beltwide Cotton Conf. Nashville, TN, January 9–12. Natl. Cotton Counc. Am., Memphis, TN, p. 66. Leach, G.J., Beech, D.F., 1988. Response of chickpea accessions to row spacing and plant density on a vertisol on the Darling Downs, south-eastern Queensland. 2. Radiation interception and water use. Aust. J. Exp. Agric. 28, 377–383. Marois, J.J., Wright, D.W., Wiatrak, P.J., Vargas, M.A., 2004. Effect of row width and nitrogen on cotton morphology and canopy microclimate. Crop Sci. 44, 870–877. Milroy, S.P., Bange, M.P., 2003. Nitrogen and light responses of cotton photosynthesis and implications for crop growth. Crop Sci. 43, 904–913. Milroy, S.P., Bange, M.P., Sadras, V.O., 2001. Profiles of leaf nitrogen and light in reproductive canopies of cotton (Gossypium hirsutum). Ann. Bot. 87, 325–333. Monteith, J.L., 1977. Climate and the efficiency of crop production. Br. Philos. Trans. R. Soc. Lond. Ser. B 218, 277–297. Muchow, R.C., 1988. Effect of nitrogen supply on the comparative productivity of maize and sorghum in a semi-arid tropical environment. I. Leaf growth and leaf nitrogen. Field Crops Res. 18, 1–16. Nichols, S.P., Snipes, C.E., Jones, M.A., 2003. Evaluation of row spacing and mepiquat chloride on cotton. J. Cotton Sci. 7, 148–155. Nunez, R., Kamprath, E., 1969. Relationships between N response, plant population and row width on growth and yield of corn. Agron. J. 61, 279–282. Peng, S., Krieg, D.R., 1991. Single leaf canopy photosynthesis response to plant age in cotton. Agron. J. 83, 704–708. Reid, P., 2001. Sicala V-3RRi. Plant Var. J. 14, 39–40.
94
R. Brodrick et al. / Field Crops Research 148 (2013) 86–94
Rosenthal, W.D., Gerik, T.J., 1991. Radiation use efficiency among cotton cultivars. Agron. J. 83, 655–658. Sadras, V.O., 1996. Cotton responses to simulated insect damage:radiation-use efficiency, canopy architecture and leaf nitrogen content as affected by loss of reproductive organs. Field Crops Res. 48, 199–208. Savoy, B.R., Cothren, J.T., Shumway, C.R., 1992. Soybean biomass accumulation and leaf area index in early-season production environments. Agron. J. 84, 956–959. Shibles, R., Weber, C.R., 1966. Interception of solar radiation and dry matter production by various planting patterns. Crop Sci. 6, 55–59.
Sinclair, T.R., Pinter Jr., Kimbali, P.J., Adamsen, B.A., LaMorte, F.J., Wall, R.L., Hunsaker, G.W., Adam, D.J., Brooks, N., Garcia, T.J., Thompson, R.L., Leavitt, T., Matthias, S.A., 2000. Leaf nitrogen concentration of wheat subjected to elevated CO2 and either water or N deficits. Agric. Ecosyst. Environ. 79, 53–60. Zhao, D., Oosterhuis, D., 2000. Cotton responses to shade at different growth stages:growth, lint yield and fibre quality. Exp. Agric. 36, 27–39.