Genotype, Row Spacing and Environment Interaction for Productivity and Grain Quality of Pigeonpea (Cajanus cajan) in sub-tropical Australia

Published by Oriental Scientific Publishing Company © 2018 This is an Open Access article licensed under a Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International License (https://creativecommons.org/licenses/by-nc-sa/4.0/ ), which permits unrestricted Non Commercial use, distribution and reproduction in any medium, provided the original work is properly cited. Genotype, Row Spacing and Environment Interaction for Productivity and Grain Quality of Pigeonpea (Cajanus cajan) in sub-tropical Australia

The pigeonpea (Cajanus cajan (L.) Millspaugh] is a perennial legume from the family Fabaceae.It is believed to be originated from India 1 with some 13 wild species endemic to Australia 2 .Over millions of years of domestication, it is now grown as an annual food and fodder crop in Asia, Africa, and Latin America.The crop ranks sixth in the world in dryland legume production, with also a forage yield of greater than other legumes and grasses when used for livestock production 3 .Pigeonpea is cultivated on about 6.23 million hectares globally with India accounting for around 70% of the total global production followed by Myanmar, Malawi, Kenya and Tanzania 4,5,6 .
Although India is the largest producer, the average national yield remains less than 1 ton/ha 7 , while grain yield of up to 3 t/ha has been reported in other parts of the world 8,9 .
Pigeonpea grain is an excellent source of protein (20-22%), specifically enriched with amino acids that are often limiting in the human diet, including methionine, lysine, and tryptophan 10 , and consumed by more than billion people globally, as a source of vegetable protein.
Relative to other legume crops, pigeonpea is tolerant to drought and heat, being able to retain a minimum productivity levels in the regions with less than 650 mm annual rainfall 11 .However, the response to irrigation was found to be substantial in some genotypes 12 .Pigeonpea can fix atmospheric nitrogen and is noted for its adaptability to diverse soils 13 .Its deep tap root with abundance of root organic matter is known to improve soil structure, nutrients turnover and soil water storage.
In Australia, the University of Queensland (UQ) had implemented a pigeonpea breeding program for the first time in the 1970's and 1980's 14,15 , to explore the development of photo-insensitive, short duration and high yielding varieties that were amenable for mechanical cultivation.However, pigeonpea industry could not be sustained due to lack of awareness about optimum agronomic practices, including the integrated pest management practices amongst growers, and lack of market support, and thus, the varieties released by UQ in 1980's were confined for use as a mandated trap crop in Australian cotton production to help reduce the likelihood of Helicoverpa (pod borer) developing resistance to Bt cotton 16 and as a fodder crop 17 .However, there has been a renewed interest in Australian grain industry to identify appropriate high yielding pigeonpea genotypes due to rising global demand and asses their fit in the Australian cropping systems.The broad adaptation of pigeonpea to the Australian cotton cropping regions, and it's resilience to extreme climate variability, makes this crop a potential summer legume option where existing legumes are poorly adapted or not profitable.However, there have been limited studies in Australia on assessing potential productivity of pigeonpea as food crop in the northern grains region.
The objective of this study was to assess the seasonal yield potential of four selected pigeonpea genotypes under two diverse subtropical environments in Australia.The study examines the genotype (G) x environment (E) x management (M) interaction for shoot dry matter (DM), grain yield (Yld), harvest index (HI), 100 seed mass, fractional intercepted radiation (f), photosynthetically active radiation use efficiency (RUE), nitrogen content in shoot dry matter (SNC) and grain nutrient profile (GN).

Crop Management
The four genotypes of pigeonpea were selected from a total of 10, introduced from the International Crops Research Institute for the Semi-Arid Tropics (ICRISAT), following the Australian bio-safety regulations.The four genotypes (Pcv1, Pcv2, Pcv3 and Pcv4) were selected based on their thermal time to maturity, agronomic traits, which included plant height, yield, seed size, and thermal time to maturity, for further studies.A two-site field experiment was conducted at Gatton (27.55° S 152.33° E) and Kingaroy (26.55° S and 151.85°E) locations in Queensland, Australia.The soil at Kingaroy is red-ferrosol (red soil) while that of Gatton site was a self-mulching cracking clay (vertisol).Planting was done using a mechanical planter on 5/12/2016 at Kingaroy and 12/1/2017 at Gatton.Prior to planting, seeds were treated with Group "J" Rhizboium and planted in three row-spacing treatments, 25, 50 and 100 cm, at a rate of approx.300,000 plants/ha, in a randomised complete block design with three replications.The plot size was 15m L × 2 m W at Kingaroy, while it was 12m L × 2m W at Gatton.A basal fertilizer of 13.5N:15P:12.5K:1.2S was applied mechanically in the seeding row at the time of planting.The crop received two supplementary irrigations of 25mm each at 31-1-17 and 14-2-17 at Kingaroy, and 13-1-2017 and7-2-2017 at Gatton.A post-emergence herbicide (verdict ®) was applied to control grasses at Kingaroy, while at Gatton, a pre-emergence herbicide "Pendimethalin" and a post-emergent herbicide "Imazapic" were applied using a shielded sprayer.
At Kingaroy and Gatton, two sprays of "Chlorantraniliprole" (Aliticore®) was applied around flowering time to control insect pests.Plants were desiccated on 5-5-2017 & 22-5-2017 at Kingaroy and Gatton, respectively, using a foliar spray of Glyphosate® at 2 L/ha and machine-harvested approximately10-12 days after desiccation at both sites.The crop was threshed using a mechanical thrasher and seeds were collected.

Measurements Weather
The hourly data on Max T, Min T, rainfall, incident solar radiation, relative humidity and pan evaporation were obtained from an automatic weather stations installed within 0.5 km distance from the experimental sites at both locations.The data was integrated over 24 hrs period for analysis.

Fractional Intercepted radiation (f)
The dynamic changes in fractional intercepted radiation was measured at hourly intervals between 11-02-2017 (30 DAS) to 22-5-2017 (130 DAS), only at Gatton site, using tube-solarimeters (Model Delta-T).The tube solari-meters which were of 90cm length were installed perpendicular to the rows of the crop at 10 cm above the ground.Two tube solari-meters were placed only in 50cm spacing treatment for all cultivars excepting for Pcv 1 and Pcv 3 which had only one tube solari-meter in each spacing.Two reference tubes were placed 2m above the plant canopy to record incident radiation.The tubes were flushed with nitrogen gas as when necessary to remove any moisture condensation in the tube and occasionally readjusted to avoid any interference of intercepted radiation by weeds.Data from the tubes were captured using a data logger at 5min intervals and integrated over one hour.The tubes along with the logger were removed 2 or 3 days before the desiccation.
Daily totals and individual tube-calibration factors were used to calculate the fraction of the incident radiation intercepted by the crop (f).
Radiation intercepted were computed as described in equation 1. Cultivars were compared using the daily short-wave radiation intercepted which was converted into the photosynthetically active radiation (PAR) [18].The radiation use efficiency (RUE) was then calculated as a ratio of total dry mass g/m2 and PAR intercepted (MJ/m2) as described in equation 2. RUE = DM / RI ...(2) Where, RUE (g/MJ) was the ratio of the DM (g/m 2 ) accumulated and RI (MJ/m 2 ) was the total radiation intercepted by the canopy, during the measurement period (i.e.11 th Feb to 22 nd May 2017).

Shoot dry matter
Plants from 1m 2 ground area were handharvested at the ground level at the peak biomass stage (just before start of senescence) 20-05-2017 and 08-05-2017, at Gatton and Kingaroy, respectively.The harvested plants were dried in a fan-forced oven at 80 o C for 48 hrs before recording the dry weight of shoot (leaves, stems and pods) and the weights calculated based on area of harvest and expressed as t/ha.

Grain yield
The trials were harvested when 90% of plants in a plot had more than 80% mature pods as indicated by the dark colour pod wall.The effective plot area was measured before harvest.The grain samples collected from the harvester were cleaned to remove any extraneous matter before recording the grain weight.The final grain yield was calculated on the plot area basis and expressed as t/ha, at 12% seed moisture content.

Shoot and grain nitrogen
A sub-set of DM and grain samples were ground to pass through 2mm sieve and sub-sample of about 50g were used for carbon and nitrogen analysis.Another sub-sample of grain was used for other elemental analysis.

Carbon and Nitrogen analysis
The plant or grain tissue sample of 0.25 g was weighed in a tin foil capsule which was placed into the induction furnace of a LECO Truspec.The CHN combustion analyser was set at 1100 o C and calibrated with EDTA.The carbon present in the sample was combusted to CO 2 which was determined with an infra-red detection cell.The nitrogen present was combusted to N 2 , NO 2 and NO.The oxides were reduced to N 2 which was determined quantitatively using a thermal conductivity cell.Results were automatically expressed as weight percentages 19 .

Elemental analysis (acid digestion)
The grain tissue sample of 0.25 g was weighed and digested with 15 ml of 5:1 nitric: per-chloric acid mixture.The digested sample was made up to a volume of 25 ml and the elemental concentrations were then determined by ICPAES.Concentration measured in mg/L were converted into mg/kg by multiplying by the digest volume and dividing by the sample weight.The results were expressed as weight percentages 20 .

Weather
Kingaroy and Gatton received in-season rainfall of 175 mm and 368 mm respectively.However, at Gatton most of the rain fell during pre-podding stage, while at Kingaroy it was well distributed throughout the growing season (Fig. 1a  and 1b).At Gatton, the maximum temperatures ranged between 30     By the end of the season, minimum temperatures at Kingaroy also reached 5 o C.

Fractional Intercepted Radiation (f)
Genotypes differed significantly in the fractional intercepted radiation "f" throughout the growing season.The genotype Pcv1 showed the highest "f" from 50 DAS, while the "f" was the lowest for Pcv4 (<0.10) followed by Pcv2 (0.2) (Fig 2).However during 35-80 DAS, all genotypes rapidly expanded their leaf area to reach "f" of >0.6, although there were significant genotypic differences in the rate of canopy expansion as indicated by the rate of increment in "f".The rate of increment of "f" during this period was 0.01/ day (1% / day) for Pcv1 and Pcv3, respectively, while it was 0.013/day (1.3%/day) for Pcv2 and 0.017/day (1.7%/day) for Pcv4.However, Pcv1 and Pcv4 were intercepting 90% of the incident radiation ("f" = >0.9)from 80 DAS for rest of the season, while PCv3 reached 0.9 "f" from 100 DAS.It was interesting to note that the "f" remained 0.8 for Pcv2 and declined to <0.6 from 120 DAS.The total intercepted photosynthetically active radiation (PAR) between 35-120 DAS and the radiation use efficiency (RUE) of the four genotypes is presented in table 1.The Pcv 1and Pcv 4 recorded similar RUE (1.76 g/MJ) dedspite variable "f" intercepted and different shoot dry matters produced during the 35-120 DAS period.However, Pcv 2 and Pcv 3 recorded 1.67 and 1.40 RUEs respectively.

Shoot dry matter
There was a response to row spacing with DM being the highest at 25 cm spacing, compared to the 50 cm and 100 cm treatments (P<0.001) for all genotypes at both sites (Table 2).The mean DM in general was greater at Kingaroy compared to Gatton in all row spacing treatments (P<0.001) and the highest average DM of 13 t/ha was recorded by  Pcv 4 at 25 cm row spacing at Kingaroy.The site effects for DM were significant at P<0.05 with an average DM being higher at Kingaroy compared to Gatton.There was a significant (P<0.05)site × variety interaction with Pcv 4 producing the highest DM at Kingaroy and the lowest at Gatton.The site × row spacing interaction was also significant (P<0.05), with row spacing differences especially between 25cm and 50cm treatments were more consistent at Gatton compared to Kingaroy.For example, the optimum row spacing was 50 cm for Pcv 1 and Pcv 3 at Kingaroy compared to other genotypes which showed 25cm being optimum.

Grain yield
The grain yield differed significantly between the sites (P<0.001) with the average yield being higher at Gatton (2.63 t/ha) compared to 2.16 t/ha at Kingaroy (Table 2).Varieties differed significantly with Pcv 3 and Pcv 4 consistently producing higher grain yields at Gatton (2.85 t/ha) compared to 2.2 t/ha at Kingaroy (P<0.05).There was a significant effect of row spacing (P<0.001) with yield being the highest at 0.50cm row spacing at both sites.The yields declined by 29% at 25cm row spacing (P<0.001) at Kingaroy while the yield reduction between 50 cm and 25 cm treatments were marginal at Gatton (Table 2).

Harvest Index
Harvest Index (HI) indicate the proportion of economic yield in the total dry matter produced by a crop.The sites differed significantly (P<0.001) with the average HI being 0.39 at Gatton compared to 0.24 at Kingaroy (Table 2).The HI in general was higher for all the four genotypes at Gatton compared to Kingaroy.However, the Pcv 3 had the highest average HI (0.49) compared to other genotypes, while at Kingaroy Pcv 1 and Pcv 2 had slightly higher HI compared to the other two.The effects of row spacing treatment was also significant (P<0.05), at both sites, with the mean HI being highest at 50cm row spacing compared to the other two row spacing treatments.Similar to grain yield, three was a consistent decline in HI at 25cm row spacing.

Relationship between shoot dry matter and grain yield
The effect of environment on the shoot dry matter productivity and its relationship with grain yield is presented in the fig.3. It was clear that the grain yields were generally higher for a given amount of dry matter produced at the Gatton site compared to the Kingaroy site.There was trend for a curvilinear relationship between the shoot dry matter and grain yield at both sites, with R 2 being significant at Gatton (P<0.001).At Kingaroy the shoot dry matter productivity reached >14 t/ha with no apparent benefits in grain yield while at Gatton, maximum shoot dry matter reached 12 t/ ha but higher partitioning of dry matter to yield is seen throughout the range of the shoot dry matter productivity.

Hundred seed mass
Hundred seed mass is an important indicator of grain quality and it also contributes to crop yield as well as marketability.There was a significant effect of site (P < 0.001) with Gatton site recording 7% higher 100-seed mass than Kingaroy site.The differences between varieties was also significant (P<0.001) with Pcv 1 recording the highest and Pcv 3 the lowest 100-seed mass at both Gatton and Kingaroy.The row spacing did not affect the 100-seed mass.The lack of site x varieties interaction shows the variation in 100seed mass was genetic and environment did not affect genotypic ranking for the 100-seed mass (Table .3).

Shoot N and grain N content, and net N contribution from the vegetative dry matter to the soil
The varieties differed significantly for shoot N content (P<0.001).Pcv 4 had the highest shoot N content at both sites.However, site had significant effect with grain N content at Kingaroy being 11% higher than at Gatton (Table 4).Varieties differed significantly (P<0.001) with Pcv 1 having the lowest grain N content compared to other three varieties at both sites.The net N contribution (after discounting N removal from grain), was generally higher at Kingaroy (93 kg/ha) compared to Gatton (60 kg/ha) due to higher DM production at Kingaroy.Varieties also differed significantly for net N contribution through vegetative stubble following harvest of grains.Pcv 4 recorded the highest N contribution (128 kg/ha) at Kingaroy but not at Gatton.However, Pcv 1 maintained N contribution of 86-88 kg/ha at both sites (Table 4).

Mineral composition of the pigeonpea grain
There was significant effect of site and verities on mineral composition of pigeon pea grain but there was no row spacing effect and hence, the data was pooled across row spacing treatments (Table 5).The differences between the sites were significant for most minerals excepting for Boron (B) and Sodium (Na).Grain harvested from Kingaroy site generally contained a higher levels of most minerals except for potassium and phosphorus whose levels were higher in the grain harvested at Gatton (Table 5).The differences between the sites could also be noticed for nutrients such as Aluminium (Al), Copper (Cu), Iron (Fe), Magnesium (Mg), Manganese (Mn), Sodium (Na) and Zinc (Zn).However, some nutrients such as Potassium (K), Phosphorous (P) and Sulphur (S) were higher at Gatton compared to Kingaroy.

DISCUSSION
The RUE, the ratio of the total dry matter (DM) produced and the PAR absorbed by canopy, is one of the major crop attributes used to explain the variation in DM due to either genotypes or environments 21,22,23 .There were significant differences between varieties for the early canopy development as indicated by the rate of increase in fractional radiation intercepted ("f") between 35 to 60 DAS period at Gatton.The RUEs ranged from 1.40 to 1.76 g/MJ amongst varieties.These values are close to those measured for pigeonpea 24 but higher than the values (1.30 g/MJ and 1.62 g/ MJ) observed by 18 and 25 respectively.However, the variation in the RUEs in the current study accounted for 76% of the variation in DM produced by the varieties.
The "f" values during early stages of crop growth (up to 40 DAS) were low and ranged from 0.1 to 0.25, compared to >0.4 observed for other legumes 26 .The poor seedling growth of pigeonpea observed in the current study supports earlier studies 24,27 .However, the current study demonstrated significant varietal differences for the rate of early canopy development with Pcv 1 showing rapid canopy development in contrast to Pcv 4 which showed the least, during the 30-70 DAS (Fig 2).While the reasons for the poor seedling growth in pigeonpea is not clear, it could be attributed to some bio-chemical changes occurring at the molecular level in germinating pigeonpea seed compared to other legume species 28 .This phenomenon needs further investigation, as the poor seedling growth would influence the subsequent crop growth.
Row spacing treatments had significant effect with the average DM being maximum at 25cm compared to 50 and 100cm spacing treatments.However, the grain yield was maximum at 50cm spacing and there was a trend for decline in the yield with further reduction in row spacing at both the sites.
There is limited systematic efforts to study the effects of row spacing on the pigeonpea productivity, and to our knowledge this is the first attempt to study the row spacing effects, particularly for the new pigeonpea cultivars in Australia.
Among the four genotypes, mean DM ranged from 5.6 to 10.4 t/ha at the two sites which were generally higher than those reported by other researchers 3.06-5.87, 29, 3.7t/ha 30 , 6.4-12.6, 31 .Genotype Pcv 3 resulted in the lowest DM (5.6t/ ha) at Gatton (Table 1).The values are pooled for row spacing treatments as there was no significant effect The grain yield of a genotype is a principal measure of its economic performance in a cropping system as it contributes to higher whole farm profit.Average grain yield ranged between 2.0 and 3.1 t/ha at the two sites (Table 1).However, these yields are higher than the average national yield measured in the major pigeon pea producing countries such as India, 1 ton/ha 7 , 1.4 to 1.9 t/ha at a plant population of 100000 to 200000 plants/ha, 32 and matches with the maximum yield of 3 t/ha reported from different parts of the world 8,9,12,33 .
Grain yield is dependent on the efficiency with which the plant drymatter is produced and partitioned into grains.In this study the relationship between shoot dry matter and grain yield clearly showed the environmental effect in partitioning of dry matter into yield (Fig. 3).The preferential allocation of dry matter to grain at Gatton needs further investigation.The historical varietal development in pigeonpea, has largely resulted in an indirect selection for biomass production in a given environment 34 .However, the current study showed that the grain yields increased with increase in dry matter only up to 8t/ha, there was no further increase in yield with further increase in DM from 8 to 10t/ha at both sites.Unfortunately, partitioning efficiency has not been systematically assessed in pigeonpea for identifying key physiological traits underpinning this trait 35 .In general enhanced biomass production among improved varieties could also be due to longer maturity or heterosis, however, in the current study all the four genotypes were short duration indeterminate types, not hybrids, which have been planted and harvested at the same time.
The harvest index (HI) across the two sites ranged from 0.21 to 0.39 with the HI being generally higher at Gatton site compared to Kingaroy site.Low harvest index in pigeonpea has been identified a yield constraint, and attributed to extensive flower drop 36 .This flower drop could be due to the continuous nature of flowering, where the earlier formed pods out-compete the later formed flowers for attracting assimilates.
The mean 100 seed mass ranged from 9.8 to 10.9 g, which were much higher than what was reported 32 (5.41-7.95g).The higher solar radiation in the Australian environments could contribute to higher photosynthetic efficiency, which in turn would have contributed to the translocation of photosynthates to the grain.The 100 seed mass was also higher at Gatton site compared to Kingaroy site.The 100 seed mass was significantly different among sites and between genotypes (P<0.001).The increase in HI and 100 seed mass at Gatton are not clear and need further investigation.
The nitrogen contribution by the crop is also higher than those reported by other researchers, 32 (32 kg/ha), 8 (26-100 kg/ha).
The composition of Zn, Fe, Mn and Cu from the pigeonpea grain observed in the study was comparable with what was reported by 37 , 24, 39, 13 and 18 mg/kg respectively).Nutritionally, trace minerals determined in our work are higher in B, Cu, Na and Zn content and comparable in Fe and Mn compared to composition of common beans, reported by 34 , highlighting the outstanding nutritional value of pigeonpea.

CONCLUSIONS
The two-site field study evaluated seasonal productivity potential of four pigeonpea varieties with three (100, 50 and 25 cm) row spacing treatments in subtropics of Australia.The decreasing row spacing from 100 cm to 25 cm resulted in increased shoot drymatter significantly, but there were genotypic differences.The intercepted solar radiation measured in 50 cm row spacing only at Gatton showed that RUE accounted for 76% of the variation in shoot dry matter production.However, average grain yields of up to 2.9 t/ha were achieved at 50cm spacing, although some genotypes (pcv 4) yielded >3.1 t /ha at both sites.A decline in grain yields with further reductions in row spacing to 25cm was apparent at both sites.The average harvest index (HI) was higher (by 48%) at Gatton site than Kingaroy mainly due to higher DM production at Kingaroy.The varieties differed significantly for shoot and grain N content at both sites.The net N contribution was generally higher at Kingaroy (93 kg/ha) compared to Gatton (60 kg/ha).There was significant effect of site and verities on mineral composition of pigeonpea grain but there was no row spacing effect.
The study concluded that pigeonpea can provide an alternative broad leaf summer legume for Australian grain growers.The grain can be used as a source of energy-rich protein for the vegetarian population throughout the world.
Daily intercepted radiation by the crop (MJ/ m2) r = Proportion of radiation recorded by the tube Dr = Daily incident radiation (from the two tubes placed 2m above the canopy) and 35 o C and the daily incident solar radiation between 20-25 MJ/m 2 from emergence until 100 DAS after which, both the parameters declined steadily.The minimum temperatures reached below 5 o C by the end of the season.At Kingaroy, the maximum temperatures ranged between 30 and 40 o C and incident solar radiation ranged between 20-28MJ/m2, although radiation levels were 20MJ/m2 for most of the days during vegetative phase.After 120 DAS temperatures and solar radiation started declining.

Fig. 2 .
Fig. 2. Fractional intercepted radiation (f) by the pigeonpea genotypes over the growing period RSp= Row Spacing; ** = P<0.001;* = P<0.05;ns = non-singificant In the Mean column, values with the same letter are not significantly different

Table 1 .
Radiation-use efficiency of pigeonpea genotypes at Gatton #The values of net intercepted PAR were averages of 2 replications and values in the parenthesis of RUE are standard deviation from the mean of two replications.

Table 2 .
Shoot dry matter (DM), grain yield and harvest index of the four pigeonpea genotypes grown in three row spacing treatments at Gatton and Kingaroy sites

Table 3 .
Summary of 100-seed mass of four varieties at the Gatton and Kingaroy sites Site: **; Var: **; RSp: ns; Site x var: ns; Site x RSp: ns; Var x RSp: ns; Site x Var x Rsp: ns

Table 4 .
The nitrogen content in shoot, grain and net N-contribution through the vegetative matter or four pigeonpea varieties at the Gatton and Kingaroy sites

Table 5 .
Mineral composition in the grain of 4 pigeonpea genotypes (mg/kg) at the Gatton and Kingaroy sites