IN lSU LIGHT INTERCEPTION, DRY MATTER PRODUCTION AND PARTITIONING OF THE POTATO CROP IN TROPICAL ENVIRONMENTS A thesis submitted in fulfilment of the requirement for the Degree of Magister in Scientia. in the University of Wales By DAVID GRAHAM NELSON UNIVERSITY COLLEGE OF WALES Department of Agriculture April 1987 l Centro Internacional de la la RibHoter.a 6530 - DECLARATION I hereby.declare that the work embodied in this thesis for the degree of Magister in Scientia of the University of Wales is the result of my own investigations except where indicated in the text. Candi date .............................••... CERTIFICATE I hereby certify that the work embodied in this thesis has not already been submitted for any degree, and is not being concurrently submitted in candidature for any degree. --------------------------------- David Graham Nelson Aberystwyth, April 1987 ACKNOWLEDGEMENTS I gratefully acknowledge the support and facilities provided by the International Potato Centre which enabled this research to proceed. In particular I am deeply indebted to Dr. D. J. Midmore of the Physiology Department for his helpful suggestions, guidance and warm encouragement throughout the course of the investigations. I would also like to thank technicians who helped with data collection at field stations in Lima, Huancayo and San Ram6n. I also acknowledge the generous assistance and helpful criticism volunteered by Dr. P. D. Jenkins of the Department of Agriculture, University College of Wales, Aberystwyth during the preparation of this thesis. Finally I wish to extend special thanks to my wife for typing the manuscript and for her patience, understanding at all times. support and .!.Q ~ wife Rosa SUMMARY Four experiments, comparing a range of cultivars at three contrasting sites in Peru, are reported. Five cultivars were compared in a summer planting at Lima (240 m.a.s l.), and at Huancayo (3200 m.a.s.l.), and four cultivars, in the absence or presence of a dried grass mulch, were compared at San Ram on ( 8 5 0 m. a. s .1 . ) and in a spring planting at Li ma. In warm lowland environments, particularly in late maturing cultivars, stem growth was promoted at the expense of leaf production and consequently leaf canopies tended to lodge and leaf preserrtation was poor. Furthermore, the onset of leaf senescence began earlier at Lima (summer planting) than in the cooler environment at Huancayo. Late maturing cultivars generally intercepted most radiation but effects were not significant at Lima (summer planting). Linear relationships between total plant dry weight and intercepted radiation were demonstrated at all sites and there was no effect of cultivar or mulch on the slope of his relationship. The slope, or radiation conversion coefficient, was 1.99 g/MJ at Huancayo while about 1.4 g/MJ in other sites. Most cultivars demonstrated linear relationships between tuber dry weight and intercepted radiation up to the achievement of maximum tuber dry weight and only at Lima (spring planting) were there significant differences between cultivars in the slope of fitted regressions. In view of the close association between radiation interception following the onset of tuber .growth and tuber dry matter production, detailed investigation into the ef ects of husbandry and genotype on the size, presentation and persistance of the leaf canopy may indicate the opportunities for improvement of tuber yield in tropical environments. INDEX INTRODUCTION GENERAL METHODOLOGY SECTION A 1. Experiment l 1.1 Materials and methods 1.2 Meteorological data 1.3 Crop growth and development 2e Experiment 2 1.1 Materials and methods 1.2 Meteorological data 1.3 Crop growth and development 3. Radiation interception and conversion pag. l 20 30 30 30 31 32 46 46 46 47 61 3.1 Radiation interception 61 SECTION B 3.2 Conversion of intercepted radiation 62 to plant dry weight. 4. Discussion 1. Experiment 3 1.1 Materials and methods 1.2 Meteorological data 1.3 Crop growth and development 64 73 73 74 75 2. Experiment 4 1.1 Materials and methods 1.2 Meteorological data 1.3 Crop growth and development 93 93 94 95 3. Radiation interception and conversion 106 3.1 Radiation interception 106 3.2 Conversion of intercepted radiation 107 to plant dry weight. 4. Discussion 109 GENERAL DISCUSSION 118 BIBLIOGRAPHY 125 INTRODUCTION It was proposed by Monteith (1977) that growth of field crops may be analysed in terms of solar radiation intercepted by the leaf surface and the efficiency of its conversion to plant dry weight. Adopting this analysis for the potato crop, Allen and Scott (1980) demonstrated that variation in total and tuber dry matter yields in the UK, was due almost entirely to differences in the amount of radiation intercepted. In the absence of disease or drought the slope of the relationship between total dry matter yields and intercepted radiation was remarkably constant. With data from Sutton Bonnington, England a linear relationship was demonstrated between tuber dry weight and intercepted radiation but data from Australia (Sale, 1973 a,b) from a warm, high radiation environment, did not fit this common line (Fig.l). Allen and Scott suggested the the lower efficiency of conversion of intercepted radiation to tuber dry weight in warm environments was due to high temperature delaying tuber initation and promoting stem growth, as reported by Bodlaender (1963). Radiation Interception Although standard tables predict geographical and seasonal variation in radiation receipts at the top of the atmosphere (e.g. The Smithsonian Meteorological Tables), calculating incident radiation at ground level must take l Figure 1. Relation between tuber dry weight and intercepted radiation from Sutton Bonnington, England (closed symbols) and from Sale, Australia (open symbols), see Allen and Scott (1980). 1500 N s '--.. 01 .µ 1000 0 ..c 01 ·r-1 Q) ~ :>-t l--l 500 ro l--l Q) ..Q ::s 8 0 0 500 1000 1500 Intercepted radiation (MJ/m2 ) int o a cc o u n t di ff ere n c es in the phys i ca l c,h a r a c t er is t i cs o f the atmosphere. Factors such as cloud cover, air pollution and elevation affects the proportion of incident radiation absorbed or scattered during its passage through the atmosphere and consequently the amount which fails to reach the earths surface (Monteith, 1973). In a global analysis, Charles Edwards (1982) showed the mean annual daily radiation receipt decreased linearly as a function of increasing latitude while the relative amplitude of seasonal variation in daily radiation receipt increased linearly with latitude. In many regions at low latitudes, monsoonal rains produce cloud cover during the summer months when radiation receipts entering the atmosphere are highest and consequently seasonal variation in radiation receipts may be relatively small. Total radiation receipts are normally measured over the wave band of 0.4 to 3.0 um, but only radiation between 0.4 and 0.7 um is important for crop growth. This photosynthetically active radiation (PAR) or short wave radiation, forms a relatively constant proportion, about 50 ?~ , o f t o t a l i n c i d e n t r a d i a t i o n i n b o t h t e m p e r a t e a n d tropical regions (Monteith, 1973). The proportion of incident radiation intercepted by a crop may be directly measured by placing tube solarimeters horizontally above and below the leaf canopy (Gallagher and Biscoe, 1978). Alternatively, the proportion of radiation intercepted may be estimated from relationships with leaf area index (LAI) or percent ground cover of the crop leaf canooy (Burstall and Harris, 1983). Burstall and Harris 2 l found that percent radiation intercepted increased with LAI up to about 3.0. At this point 70 to 90 % of radiation was intercepted but further increments in LAI produced only a small increase in the proportion of radiation intercepted. In the same study, Burstall and Harris found a close correspondence between percent ground cover and percent radiation intercepted. They estimated ground cover with a modified quadrat and found this method fast, accurate and simple. It avoided destructive growth analysis, which is necessary to estimate LAI, and is cheap as only a sing e radiometer is required to measure total incident radiation. In the Netherlands, data presented by Van Der Zaag (1984) suggested the relationship between percent ground cover and radiation interception may be curvilinear. However details of the methodology employed to obtain this result are not fully described and it was based on a limited number of field observations. The data of Burstall and Harris (1983), 1Nhich showed a linear relationship between percent ground cover and percent radiation interception, is supported by an unpublished communication from Steven, Biscoe and Jaggard. They state that a linear relationship may be expected providing a high proportion of incident radiation enters the canopy at angles greater than the inclination of individual leaves. Factors which influence leaf expansion and senescence determine leaf area and consequently the. proportion of incident radiation intercepted by crop canopies. A detailed analysis of these factors was made for potato crops grown in Great Britain by Allen and Scott (1980). It was shown 3 t L that le a f are a d u_r in g May, and in some ye a rs June, was insufficient to intercept a high proportion of incident radiation. The principal constraint to leaf growth during April and May in Great Britain was low temperature, which delays emergence and restricts leaf expansion. The base temperature of leaf appearance in potatoes was found by Kirk and Marshall (1983) to be o0 c, and the rate of the leaf appearance increased up to 20°C. Nevertheless, although warm temperatures promote rapid leaf appearance they also reduce leaf longevity, that is advance senescence. In this context Marinus and Bodlaender (1975) observed that potatoes growing in controlled environments at 27°C began leaf senescence earlier than plants kept at 16° or 22°c. However because new leaves were formed over a longer period at the highest temperature, the duration of the leaf canopy was extended. In other crop species effects of high temperature upon the leaf canopy include more erect leaves (Charles Edwards, 1982) and an increase in specific leaf area (SLA), ie. leaves are thinner (Peek et al, 1977). Some effects of light intensity upon morphology of the potato canopy were reported by Bodlaender (1963). Notably high light intensities produced a more prostrate canopy type. Sale (1973 a), in shading experiments with potatoes, found that high light intensity reduced SLA, ie. produced thicker leaves. Water availability may markedly affect crop emergence, leaf expansion, senescence, and consequently radiation interception of the crop canopy. During a period of soil 4 water-logging, oxygen availability declines and toxic chemicals are produced (Russell, 1973). Under these conditions nutrient uptake and growth of wheat are generally inhibited (Belford, 1981). At the other extreme, low soil water availability also inhibits emergence and restricts growth of potato plants (Cavagnaro et al, 1971). Water demand is determined by the size of the leaf canopy and evaporative losses created by prevailing incident radiation and the saturation vapour pressure deficit of the air (Penman, 1970). When roots are unable to ta' 0 0 ro s:: ::s 0 H tJ'l .µ s:: Q) 0 H Q) p.. Ul >-i 0 0 ro Q) .µ 35 mm was significantly higher in cvs DT0-33 and Desiree than in cv Mariva (Table 14). Cvs Mariva and Mi Peru had a significantly lower percentage of tuber fresh weight >35 mm than other cultivars. Table 14. Effect of cultivar on total and graded (>35 mm) tuber fresh weight, and percent tuber fresh weight >35 mm. Character Cultivars DT0-33 Desiree Revoluci6n Mariva Mi Pera D 501 10 Total tuber fresh weight 135 mm kg/M 2 2.11 a 2.37 a l. 75 ab 0.95 b l. 35 ab l. 03 ~b T u b e r fr e s h weight >35 mm 89.7 a 89.7 a 77.l a 58.3 b 57.9 b 18.6 There were very large differences between cultivars n tuber number at final harvest (Table 15). Cv Mi Pe u produced significantly more tubers than cv Mariva, which together with cv Revolucibn, had significantly more tubers than cvs DT0-33 and Desiree. However, the proportion of tubers >35 mm was significantly lower in cvs Mariva and Mi Peru and consequently only small differences between cultivars in tuber number >35 mm were recorded. 42 Table 15. Effect of cultivar on total and graded (>35 mm) tuber number. Character Total tuber Tuber number QI Tuber number 1Q number/m 2 >35 mm/m 2 >35 mm/m 2 Cultivar DT0-33 42.7 c 24.l 56.7 a Desiree 43.3 c 20.6 48.0 ab Revoluci6n 75.8 ab 31. 4 42.8 b Mariva 72.l b 19.6 27.3 c Mi Per ti 100.2 a 24.8 25.5 c D 5 QI IQ 26.3 ns 12.6 NON-DESTRUCTIVE GROWTH ANALYSIS Both the total number of emerqed leaves per stem and number of leaves per stem lost, by senescence, are plotted against days post-emergence for cvs Desiree, Revolucim and Mariva in Fig.8 . There was a curvilinear relationship between total leaf number and time which was similar in all cultivars. The rate of appearance of leaves declined throughout the sampling period but particularly following the third sampling date, about 50 days post-emergence. The approximate rate of leaf appearance between the first and third sampling date was 0.51 leaves/day. Subsequently, up to the fourth sampling date, about 80 days post-emergence the rate of leaf appearance was 0.18 leaves/day. L e a f l o s s b e g aT;l i n a 11 c u l t i v a r s a p p r o x i m a t e l y l 0 - l 5 days post-planting and proceeded at a relatively constant rate of about 0.32 leaves/day up to the final sampling date, about 80 days post-emergence. Prior to about 50 days post- emergence the rate of leaf loss was lower than that of leaf emergence but subsequently leaves were lost 43 Figure 8. Effect of cultivar on leaf appearance (standard symbols) and leaf loss (slashed symbols). Symbols as Fig.5 . 40 '"d i:: rel s Q) 30 Cl) ..µ Q) Cl) !> rel ~ Q) Q) ....-! ~ '"d Cl) Q) Q) 20 O"i !> fa.1 rel Q) Q) s ....-! Q) '"d 4-l Q) 0 CJ Cl) ~ Q) 10 Q) c ~ Q) Cl) ::l z 0 0 20 40 60 80 Days post-emergence 100 emergence but subsequently leaves were lost faster than they were produced. SUMMARY C v s R e v o l u c i on a n d M i P e r u e m e r g e d e a r l i e s t an d produced the highest total stem number 55 days post-planting. Cv Revolucion, Mariva and Mi Peru achieved much longer mainstems but only cv Mi Peru produced a significantly greater number of nodes per mainstem than other cultivars. The pattern of production of ground cover over time was initially sigmoidal, with all cultivars demonstrating a similar rate of canopy expansion. Subsequently, during the latter half of the growing season, the leaf canopy of most cultivars lodged and produced proportionally less ground cover. Throughout the growing season no significant differences between cultivars in LAI, ground cover, or accumulated ground cover were recorded. Nevertheless, cv DT0-33 consistently produced a low LAI and ground cover and accumulated considerably fewer cover days than other cultivars, in particular cvs Revolucibn and Mi Peru. Non-destructive growth analysis indicated that cvs Desiree, Mariva, and Revolucion did not differ in their rate of leaf appearance and senescence. Leaf senescence began about 15 days post-mergence but up to about 50 days post-emergence the rate of leaf appearance was greater than leaf senescence. Subsequently the. rate of leaf senescence proceeded at a similar rate while that of leaf appearance declined. 44 There were no differences between cultivars in leaf dry weight but stem dry weight in cvs Mi Peru, Mariva and Revolucion was much higher than in cvs Desiree and, in particular, DT0-33. The leaf /stem dry weight ratio was consistently greatest in cv DT0-33, and lowest in cv Mariva. There were no differences between cultivars in total plant dry weight but cvs DT0-33, Desiree and Revolucion partitioned a higher proportion of dry weight to tubers between 55 and 77 days post-planting and at 77 and 90 days post-planting had a significantly higher tuber dry weight than cv Mi Peru. However, cv Mi Peru appeared to bulk rapidly in the 10 days prior to final harvest and at 100 days post-planting there was no significant difference between cultivars in tuber dry or fresh weight. At final harvest cvs Revolucibn, Mariva and particularly Mi Peru had a much higher tuber number than the other cultivars and although all cultivars produced a similar number of tuber >35 mm, the cvs DT0-33 and Desiree produced a significantly higher tuber fresh weight, in this size category, than cvs Mariva and Mi Per~ 45 2. EXPERIMENT 2 2.1 MATERIALS AND METHODS Experiment 2 was planted at Huancayo on 10 January 1984. Experimental design, treatments, seed and plantinq were as described for Experiment 1. Post-plantinq operations included ridqing 40 days oost-olantinq w~en a further 100 kq/ha nitroqen was apolied. Leaf beetles (sso. Ioit~~) and Early Blight (Alternaria solani) were controlled by foliar sprays of carbaryl and mc=incozeb, respectively. Emergence was recorded at 4 or 5 day intervals beginning 13 days post-planting and estimates of ground cover were made every 7 to 21 days throughout croo growth. Destructive growth analysis took place 65, 86, and 98 days post-planting. Cvs DT0-33 and Desiree were harvested l 0 6 days post - p la n t in g and c vs Rev o l u c i 6n , Mari v a and M i Peru 117 days post-planting. A non-destructive growth analysis took place 43, 65, 86, and 106 days post-planting. 2.2 METEOROLOGICAL DATA Meteoroloqical data up to 16 weel .) 0 () 40 'O ~ ::J 0 H tri .µ 20 ~ Q) () H Q) P-i 0 I I 0 60 80 100 120 30 Ul ~ ro 'O H Q) 20 :> 0 () 'O Q) .µ ro ...-l 10 ::J s ::J () () i=:i: 0 I J 0 40 60 80 100 120 Days post-planting Ground Cover Ground cover at all measuring dates is shown in Fig.9. All cultivars produced little ground cover between emergence and the first estimation of ground cover 38 days post­ planting. Canopy expansion may have been restricted by water-logging following heavy rainfall, during and shortly following emergence. Subsequent to 38 days post-planting a more rapid, linear phase of canopy expansion began which ceased first in cvs DT0-33 and Desiree at 65 days post­ planting and later (86 days post-planting) in cvs Revolucil:n, Mariva and Mi Peru. The ground cover of all cultivars declined between 86 days post-planting and final harvest. Cvs Mariva and Mi Peru consistently recorded the highest values of ground cover. At 86 days post-planting these cultivars achieved a significantly higher ground cover than all c·ultivars except Revolucibn. Conversely, cv DT0-33 produced significantly less ground cover than all other cultivars except Desiree at both 86 and 98 days post­ planting. Accumulated Cover Days. At 86 and 98 days post-planting, and at final harvest, differences between cultivars in accumulated ground cover were significant (Fig.9). Cvs Mariva and Mi Pera consistently had the most cover days and cv DT0-33 the fewest. At final harvest, cv Mariva had accumulated significantly more cover days than cvs Revoluci~, Desiree and DT0-33, while cv Mi Peru had significantly more cover days than cv DT0-33. 51 DRY MATTER PRODUCTION Leaf Dry Weight Leaf dry weight (Table 20) reached a maximum 65 days post-planting in cvs Desiree and Mariva, and 86 days post- planting in all other cultivars. Significant differences were recorded between cultivars at all sample dates. At 86 days post-planting, cvs Mariva and Mi Per6 produced significantly more leaf dry weight than all other cultivars except Revolucibn. A low leaf dry weight was recorded by cv DT0-33 at all sampling dates. Table 20. Effect of cultivar on leaf .dry weight. Character Leaf dry weight (g/plt) Days post-planting Cul ti var DT0-33 Desiree Revolucion Mariva Mi Pera D 5?~ Stem Dry Weight 65 8.0 b 15.6 ab 11. 7 ab 24.7 a 14.4 ab 15.2 86 98 12.5 c 7.0 14.5 be 7.5 19.5 ab 18.0 24.l a 17.5 24.l a 18.5 6.5 11. 0 b ab ab ab a Most cultivars recorded maximum stem dry weight at 86 days post-planting (Table 21). At all sampling dates, cvs Mariva and Mi Per~ recorded a high stem dry weight and at 86 days post-planting these cultivars produced significantly more stem dry weight than all other cultivars. The least stem dry weight was produced by cv DT0-33 at all sampling dates. 52 Table 21. Effect of cultivar on stem dry weight. Character Stem dry weight (g/plt) Days post-planting 65 86 98 --------------------------- Cul ti var DT0-33 Desiree Revoluci6n Mariva Mi Pero Stolon Dry Weight 4.2 7.3 7.3 22.6 14.9 13.9 b 6.4 b 12.2 b 13.4 a 31. 6 ab 23.6 8.5 b 6.1 c b 10.7 be b 14.7 abc a 28.7 a a 23.0 ab 16.l At all sampling dates cvs Mariva and Mi Peru produced most stolon dry weight (Table 22). Stolon production at this site was low in all varieties. Table 22. Effect of cultivar on stolon dry weight. Character Stolon dry weight (g/plt) Days post-plantinq Cultivar DT0-33 Desiree Revoluci6n Mariva Mi Pero Total Plant Dry Weight 65 0.1 0.1 0.2 l. 5 1.1 0.8 b b b a a 86 98 0.3 b 0.1 0. 2 b 0.1 0. 7 b 0.4 l. 5 a l. 0 l. 5 a l. 6 0.7 0.6 c c be ab a Total plant dry weight reached a maximum 86 days post- planting in cvs DT0-33 and Desiree and 98 days post-planting in other cultivars (Fig.10). There were significant differences at all sampling dates but notably cvs Mariva and Mi Peru produced a significantly higher plant dry weight than all other cultivars 86 days post-planting. A lowplant dry weight was consistently produced by cv DT0-33. 53 Figure 10. Effects of cultivar on total plant dry weight and tuber dry weight. Symbols as Fig.5 . 200 ......... 160 .µ r-l O.i ........... tJi .µ ..c:: 120 tJi ·.-! Q) ~ >-i i...i ro .µ 80 i:: ctl r-l O.i r-l ctl .µ 0 40 E-i I 0 I 0 60 80 100 160 .µ r-l ~ tJi '-' .µ ..c 120 tJi ·.-! (!) ::: >-i i...i ro ~ 80 (!) ,.Q ::J E-i 40 0 60 80 100 120 Days post-planting Tuber Dry Weight Tuber dry weight increased until final harvest (117 days) in cvs Mariva and Mi Peru but in other cultivars it reached a maximum at earlier sampling dates (Fig. 10). At the first sampling date cv Desiree had the highest tuber dry weight but over the rest of the sampling period much larger increases in tuber dry weight were recorded by late maturing cultivars. As a consequence at 86 and 98 days post-planting the cvs Mariva and Mi Peru had a significantly higher tuber dry weight than cv DT0-33. By final harvest cvs Mariva and Mi Peru had a significantly higher tuber dry weight than all other cultivars. DRY MATTER PARTITIONING Leaf /Stem Dry Weight Ratio. The leaf/stem dry weight ratio of all cultivars declined between 86 and 98 days post-planting (Table 23). Cv DT0-33 had significantly higher leaf/stem dry weight ratio than cvs Desiree, Mi Peru and Mariva 86 days post- planting. Cvs Mariva and Mi Peru had the lowest leaf/stem dry weight ratio at all sampling dates. Table 23. Effect of cultivar on the leaf/stem dry weight ratio. Character Days post-planting Cultivar DT0-33 Desiree Revoluci6n Mariva Mi Pera Leaf /stem dry weight ratio 65 86 98 --------------------------- l. 72 ab l. 93 a l. 21 ab 2.17 a l. 25 be 0.84 ab l. 58 ab l. 48 ab l. 27 a 1.10 ab 0.77 c 0.61 b 0.87 b l. 03 be 0.81 ab 1.11 0.54 0.65 54 Harvest Index At 65 days post-planting higher harvest indices w re r e c o r d e d b y c v s D T 0 - 3 3 , D e s i r e e a n d R e v o l u c i on ( T a b l e 2 4 ) . Subsequently the harvest index of other cultivars made large increases and at 86 and 98 days post-planting there was no significant effect of cultivar on this parameter. Table 24. Effect of cultivar on harvest index (%). Character Harvest index (%) Days post-planting Cultivar DT0-33 Desiree Revoluci6n Mariva Mi Pera D 5?~ TUBER NUMBER 65 55.l 51. 5 42.6 26.3 13.3 24.l 86 a 67.9 a 70.0 ab 62.5 be 64.8 c 67.8 ns 98 78.6 1·8. 7 76.3 76.7 75.l ns Tuber number per plant, at all sampling dates, and final harvest is presented in Table 25. Maximum tuber number was achieved in all cultivars at either 86 or 98 days post-planting. Tuber numbers in all cultivars declined between 98 days and final harvest. There were significant differences between cultivars in tuber number at all sampling dates. The highest tuber number was consistently recorded by cvs Mariva and Mi Peru Differences between other cultivars were not significant at any sampling date. 55 Table 25. Effect of cultivar on tuber number per plant. Character Tuber number/plant Days post-planting Cul ti var DT0-33 Desiree Revoluci6n Mariva Mi Pera D 5?6 65 6.0 12.3 4.8 17.8 12.8 11. 4 86 b 9.6 ab 11. 3 b 10.6 a 18.9 ab 17.5 6.5 TUBER DRY MATTER PERCENTAGE 98 b 7.9 b ab 12.4 ab b 10.4 ab a 17.9 ab a 19.5 a 11. 8 Final harvest 7.6 b 8.9 b 9.2 b 14.7 a 14.4 a 2.8 The tuber dry matter percentage of cvs DT0-33 and Desiree reached a maximum 98 days post-planting while in other culti.vars dry matters increased until final harvest (Table 26). The decline in tuber dry matter percentage recorded with the earliest maturing cultivars was not associated with regrowth of foliage or secondary growth of tubers. Tuber dry matter percentage did not differ significantly between cultivars at 65 days post-planting but at the following sampling date cv Mariva achieved a significantly higher tuber dry matter percentage than cvs DT0-33, Desiree and Revolucibn. Similarly, at 98 days post-planting and in a final harvest comparison, cv Mariva achieved the highest tuber dry matter percentage. Cvs DT0-33 and Desiree had the lowest dry matter percentage at final harvest. 56 Table 26. Effect of cultivar on tuber dry matter percen age. Character Tuber dry matter percentage Days post-planting 65 86 98 -------------------------- Cultivar DT0-33 Desiree Revoluci6n Mariva Mi Pera D 5?o FINAL HARVEST 15.3 14.7 14.l 14.5 14.7 ns 18.8 b 19.l b 18.8 b 21. 5 a 19.5 ab 2.3 Total and Graded Tuber Fresh Weight 19.9 b 21. 2 b 21. l b 26.4 a 21. l b 3.1 Final harvest ------------- 18.4 c 19.0 c 22.5 b 28.4 a 22.9 b 2.4 Cvs Mariva and Mi Peru recorded a significantly higher total, and >35 mm tuber fresh weight at final harvest than all other cultivars (Table 27). Cv Desiree had the lowest p e r c e n t a g e o f t u b e r fr e s h w e i g h t i n t h e s i z e g r a d e > 3 5m m with differences between other cultivars not reaching significance. Table 27. Effect of cultivar on total and graded (>35 mm) tuber fresh weight, and percent tuber fresh weight >35 mm. Character Cul ti var DT0-33 Desiree Revoluci6n Mariva Mi Pero D 5?o Total tuber fresh weight kg/m 2 0.79 c l. 06 be l. 69 b 2.88 a 2.75 a 0.79 57 Tuber fresh weight >35 mm Kg/m 2 0.61 c 0.74 be 1.49 b 2.47 a 2.37 a 0.80 ?o Tuber fresh weight >35 mm 77.2 ab 69.8 b 88.2 a 85.7 a 86.2 a 13.l Total and Graded Tuber Numbers At final harvest cvs Mariva and Mi Peru produced the highest total number of tubers and also most tubers >35mm (Table 28). Cvs DT0-33 and Desiree produced fewest tubers >35 mm and also had the lowest proportion of tubers in this size grade. Table 28. Effect of cultivar on total and graded (>35 mm) tuber numbers. Character Cul ti var DT0-33 Desiree Revoluci6n Mariva Mi Pero Tuber number per m.2 28.0 b 32.l b 34.2 b 54.4 a 53.3 a 10.l NON-DESTRUCTIVE GROWTH ANALYSIS Tuber number >35 mm/m 2 10.l c 9.4 c 20.3 b 31.3 a 30.6 a 10.l ~~ Tuber number >35 mm 36.l b 29.3 b 60.l a 57.5 a 57.4 a 18.l The relationship between number of emerged leaves and senesced leaves with days post-emergence is presented in Fig.11. The rate of leaf appearance tended to decline after about 45 days post-emergence in cvs Desiree and Revolucibn, but remained relatively constant until about 60 days post- emergence in cv Mariva. The mean rate of leaf emergence between the first and second sampling dates was about 0.26 leaves per day. The longer duration of the phase of rapid leaf appearance in cv Mariva led to a higher maximum leaf number in this cultivar. Leaf loss, due to senescence, at the base of the 58 Figure 11. Effect of cultivar on leaf appearance (standard symbols) and leaf loss (slashed symbols). Symbols as Fig.5 . 20 ro c 15 m s Q) Ul .µ Q) Ul :> m l--l (1) (1) H Pl ro w 10 Q) Q) tyl :> i..i m Q) (1) SH (1) ro 4-l (1) 0 u 5 Ul l--l (1) Q) c § (!) Ul z 0 0 20 40 60 80 100 Days post-emergence mainstem began in all cultivars about 45 day post-emergence. Subsequently the mean rate of the leaf loss was 0.25 leaves per day, almost identical to the rate of leaf appearance during the early part of the grawing season. SUMMARY Cultivars emerged between 16 and 21 days post-planting but little canopy expansion took place before 38 days post­ planting. Subsequently leaf area was produced most rapidly in cvs Mariva and Mi Peru and these cultivars achieved a higher LAI, ground cover, and accumulated most cover days. The longevity of the leaf canopy, that is, the period between emergence and complete senescence, was greatest in c v s M i P e r u, M a r i v a a n d R e v o 1 u c i tn . The larger canopy size of cvs Mariva and Mi Peru at later sampling dates was associated with longer mainstems possessing more nodes. Cv Mariva had most secondary stems at all sampling dates. Stem height, node number and ground cover of cvs DT0-33 and Desiree remained low throughout the sampling period.· A non-destructive growth analysis of cvs Desiree, Revolucibn and Mariva indicated that up to about 45 days post-emergence all cultivars had a similar rate of leaf appearance. Subsequently leaf production almost ceased in cvs Desiree and Revolucion but continued up to about 60 days post-emergence in cv Mariva. Leaf senescence beqan about 45 days post-emergence at a rate similar to that of the initial rate of leaf appearance. Consequently during 59 I the latter part of the growing season leaf senescence took place much faster than leaf appearance. Cvs Mariva and Mi Peru produced more leaf and stem dry weight than other cultivars and achieved a significantly higher total plant dry weight at 86 days post-planting. Tuber dry weight was also highest in later maturing cultivars at late sampling dates including final harvest but initially, at 65 days post-planting, cv Desiree had most tuber dry weight. Harvest index was initially lowest in cvs Mariva and Mi Peru but at 86 and 98 days post­ planting effects of cultivar on this parameter did not a c h i e v e s i gn i f i c an c e. At final harvest cvs Mariva and Mi Pera had a significantly higher total and graded >35 mm tuber fresh weight than other cultivars and a significantly higher number of tubers. The tuber dry matter percentage of cv Mariva was particularly high and significantly greater than recorded in other cultivars. 60 3. RADIATION INTERCEPTION AND CONVERSION 3.1 RADIATION INTERCEPTION The radiation interception of cultivars at Lima and Huancayo is shown in Fig.12. Shortly following emergence at Lima a linear relationship between the rate of radiation interception and time was observed in all cultivars. The slope of this relationship differed little between cultivars and persisted until shortly prior to final harvest when the rate of radiation interception declined. At Huancayo little radiation was intercepted by any cultivar until after the first sampling date and subsequently more radiation was intercepted by late maturing cultivars Mariva and Mi Per~ At final harvest differences between cultivars in radiation inteDception were much larger at Huancayo than at Lima. At the former site, cvs Mariva and Mi Pera intercepted significantly more radiation than all other cultivars while cv Revolucibn intercepted significantly more radiation than CV DT0-33. At Lima differences between cultivars in total radiation interception did not reach significance. Only cv Mariva intercepted a similar amount of radiation at both sites, other cultivars intercepted more radiation at Lima. It is notable that while cv DT0-33 intercepted least radiation at both sites it intercepted about three times more radiation at Lima than Huancayo. The relationship between ground cover and leaf area index (Fig.13) indicates that at early sampling dates in 61 N ~ 1-J ~ ........... c 0 ·.-! .µ (\j ·.-! rlj Ci1 H rlj (1) .µ Pi (1) 0 H OJ .µ c: H N ~ 1-J ~ - c 0 ·.-! .µ (\j ·.-! rlj (\j H rlj OJ .µ Pi (l) 0 H (l) .µ c H Figure 12. Effects of cultivar on intercepted radiation 600 500 400 300 200 100 0 0 600 500 400 300 200 100 0 0 LIMA at Lima and Huancayo. (DT0-33, O Desiree, lJ. Mariva, A ; Mi Peru, e 20 40 60 HUANCAYO 20 40 60 Days post-planting ; Revolucion, 0 ) 80 100 I l 80 100 120 l>-l Cl) ::> 0 C.) rel i:: :::s 0 )..j tri .µ i:: Cl) C.) )..j (L) P-1 Figure 13. The relationship between percent ground cover and leaf area index at Lima (open symbols) and Huancayo (closed symbols) over successive sampling dates: first, 0 •; second, D •; third, 'J '; fourth, o Regression fitted over data from first and second sampling dates excluding DA (cv Revolucion). 70 DA 50 y = 10.2 + 17.Sx • v v· 40 1 v • v <> 30 ¢ <> ' Ot. 20 <> ,. 10 ' 0 1 2 3 4 5 6 Leaf area index both experiments, differences between cultivars in percentage ground cover were, except in cv Revolucion at Lima, proportional to LAI. However, at later sampling dates, the larger leaf canopies of most cultivars at Lima tended to produce proportionately less ground cover, presumably because of lodging. At Huancayo maximum values of LAI were lower than at Lima and ground cover remained proportional to LAI throughout the sampling period. 3.2 CONVERSION OF INTERCEPTED RADIATION TO PLANT DRY WEIGHT TOTAL PLANT DRY WEIGHT Plant dry weight was linearly related to intercepted radiation at both sites (Fig.14). The response tended to decline in most cultivars towards the end of the growing season as dry matter losses, associated with senescence, caused measurements of accumulated plant dry weight to be under-estimated. In view of these dry matter losses, data from the last sa~~ling date at each site was excluded from regression analysis (Fig.15). Following analysis of variance and elimination of block effects there was no significant effect of cultivar on the slope of regressions fitted at each site, but joint regression analysis indicated that the radiation conversion coefficient was significantly higher (P=D.05) at Huancayo than at Lima. Correlation analysis over the same data· indicated that intercepted radiation accounted for about 92% and 95% of the variation in plant dry weight at Lima and Huancayo, 62 Figure 14. Effect of cultivar on relationship between total plant dry weight and intercepted radiation at Lima and Huancayo. Symbols as Fig.12 . LIMA 800 0 0 N ,• ~ fl tyl 600 .µ 0 h. ..c 0 • tyl • ·.-I (J) ~ :>< 40 0 i....i 'O • .µ c m 4. ,.....j 0-i £ ,.....j 20 m 0 .µ 0 E-i 0 200 400 600 800 Intercepted radiation (MJ/m2) HU AN CAYO 800 ' N s • '-... tyl 600 ' .µ • ..c tJ'l 0 ·.-I (J) ~ ~ 400 i...i 'O D~ A .µ c m ,.....j °' 0-i ,.....j 200 m OA. .µ 0 E-i 0 0 200 400 600 Intercepted radiation (MJ/m2) 800 700 600 N 500 < 01 .µ ...c: 01 400 ·r-1 ()) ;3 !>1 lo-I ro .µ c: 300 rd ..--! 0-i ..--! rd .µ 0 8 200 100 0 Figure 15. Relationship between total plant dry weight and intercepted radiation, with final sampling excluded, at Lima, O ; and Huancayo, O 0 y= -11.9 + 1.99 x 12.5 + 1.44 x 0 0 0 0 100 200 300 400 500 2 : Intercepted radiation (MJ /m ) · respectively. TUBER DRY WEIGHT. Because tuber dry weight, in some cultivars, declined at final harvest regression analysis at both sites was undertaken only from the onset of tuber growth until the achievement of maximum tuber dry weight (Fig.16). At Lima, fitted linear regressions were significant for all cultivars except cv Mi Pera, which displayed an exponential relationship between tuber dry weight and intercepted radiation. Differences in the slope and origin of regressions fitted to cvs DT0-33 and Desiree and cvs Revolucibn and Mariva combined, were not significant. At Huancayo, the origin of joint regressions fitted to cvs DTO 33, Desiree and Revolucion and cvs Mariva and Mi Peru differed significantly but there was no difference in the slope of the same regressions. 63 N ~ °' .µ ~ °' ·r-1 (]) ~ :>-i H ro H (]) ..0 ~ E-i Figure 16. Effect of cultivar on relationship between tuber dry weight and intercepted radiation at Lima and Huancayo. Symbols as Fig.12 . LIMA DT0-33 : y -172.3 + 1.38 x 600 Desiree: y -138.4 + 1.02 x Revolucion and Mariva combined: y = -209.5 + 1.01 x 400 0 200 400 600 800 HUANCAYO 800 600 DT0-33, Desiree and Revolucion combined: y = -12.9 + 1. 42 x Mariva and Mi Peru 400 combined: y = -193.5 + 1. 89 x 200 O'T"'~~,.f&,...~-r--~--~-,r---~----~~----~- 0 200 400 600 800 Intercepted radiation (MJ/m 2 ) 4. DISCUSSION