Effect of cutting depth during sugarcane harvest on root characteristics and yield

Ratooning is an important cultivation practice in sugarcane production around the world, with underground buds on the remaining stalk acting as the source for establishment of a subsequent ratoon crop. However, the optimal depth of cutting during harvest in terms of yield and root growth remains unknown. We carried out a two-year field study to determine the effects of three cutting depths (0, 5 and 10 cm below the surface) ratoon cane root and yield. Results showed that cutting to a depth of 5 cm increased the root fresh weight and root volume by 32-40% and 49-85%, respectively, compared to cutting depths of 0 and 10 cm. Remarkably, cutting to a depth of 5 cm also had a significant effect on the development of fine roots, which is closely linked to cane yield. The effect was particularly noticeable in terms of two root traits, root volume and the surface area of roots with a diameter of 1.0-2.0mm, and root length and the number of root tips in roots with a diameter of 0-0.5mm. As a result, a cutting depth of 5 cm below the surface increased cane yield by 35 and 25% compared to depths of 0 and 10 cm below the surface, respectively. Overall, these findings suggest that a cutting depth of 5 cm is optimal in terms of sugarcane yield, largely due to the enhanced effect on root traits, especially the development of fine roots. These findings will help optimize sugarcane ratoon management and improve the ratoon cycle.


88
stalk positions belowground, thereby promoting root development on these nodes, and increasing 89 the absorption of water and nutrients from deeper rainfed soil. We therefore varied cutting depth 90 to examine changes to bud germination, root growth, and root morphology, and the effects of root 91 development on ratoon yield. The findings will help optimize ratoon crop management and 92 improve the ratoon cycle in this region. Yunnan Province, Southwest China. The soil is sandy loam with 50.0% sand, 34.0% silt, and 97 16.0% clay, and a pH of 7.5. The soil has a high organic matter content of 1.8%, with a total 98 potassium content of 0.76%, total nitrogen content of 0.142%, and total phosphorus content of 99 0.12%. The content of alkaline hydrolysis nitrogen is 78.2 mg kg -1 , available phosphorus is 26.6 100 mg kg -1 , and available potassium is 49.0 mg kg -1 . Organic matter was determined using the 101 potassium dichromate method, soil nutrients were tested using flame atomic absorption 102 spectrophotometry, and soil mechanical composition was determined using the sedimentation 103 method. The field experiment was conducted under rainfed conditions, with a groundwater level 104 of approximately 2 m below the ground surface. 105 The climate in this region is characterized by hot summers and long winters, with severe 106 droughts in early spring. The highest precipitation and temperatures occur in June and July. Mean 107 annual precipitation is 983 mm, and winter and summer precipitation are 20.3 and 174.6 mm per 108 month, respectively. The mean air temperature is 20.0 °C, and the mean winter and summer 109 temperatures are 14.4 and 24.1 °C, respectively. These annual figures are also of intrinsic value 110 since they behave differently with respect to each other during the growing season from June to 111 October (Fig. 1). The experiment was conducted from 2017 to 2018 using sugarcane cultivar YZ081609 [24]. 117 Planting was carried out on 22 April 2016 at a conventional planting density of 120,000 buds ha -1 118 (two-bud cane setts). The planting depth (cane sett bed to soil surface) was approximately 10 cm, 119 with about 10-15 cm of earthing up before the grand growth phase. Thus, at harvest, 20-25 cm of 120 soil covered the original cane setts. NPK fertilizer (20: 12: 18) was applied during the early 121 elongation period at a rate of 1200 kg ha -1 on 10 March 2017 and 10 March 2018. All other 122 cultivation and crop management procedures were in line with conventional cropping practices. 123 Based on the density of underground buds on the ratoon stool and the germination potential, 124 we divided the underground buds into three types (Fig. 2). Type 1: terminal buds in the top soil. 125 Least abundant, but fastest to germinate, with the roots distributed mainly in the upper soil layer. 126 Type 2: distributed at a depth of 5-10 cm, mostly active and fast to germinate, with deeper roots 127 than type 1. Type 3: distributed below 10 cm, mostly dormant with the poorest rate of germination. 128 Accordingly, cutting treatment was carried out at the following depths: the soil surface (0 cm, T1), 129 and 5 (T2) and 10 cm below the surface (T3) (Fig. 2 were calculated based on the area of the plots and the cultivar used. 157

Statistical analyses
158 When the P value of the correlation between root biomass and dry weight or shoot biomass 159 and dry weight was greater than that of the correlation with fresh weight, the dry weight was used 160 for analysis. In all other cases, the fresh weight was used. Mean differences in sugarcane yield 161 and yield components were compared separately each year using one-way analysis of variance 162 (ANOVA)  in an increase in net benefits of 10,900 Yuan ha -1 (64.8%) and 8,000 Yuan ha -1 (40.9%) compared 194 to T1 and T3, respectively, in both ratoon crops (Fig. 3.).
195 Fig. 3. Mean input-output values of the two ratoon crops at each cutting depth (0 cm (T1), 5 cm 196 below the ground (T2), and 10 cm below ground (T3)). The net benefits of each treatment were 197 calculated based on the following inputs in RMB(¥): 120 Yuan ha -1 for harvesting, 2,250 Yuan 198 ha -1 for trash pulverizing, 4,800 Yuan ha -1 for fertilizer, 750 Yuan ha -1 for fertilizer application 199 and hilling up, 900 Yuan ha -1 for pesticides, and 450 Yuan ton -1 for output of millable canes. 200 There was little change in the input price between study years. Costs were calculated according to 201 the area size and the sugarcane cultivar. Error bars, SD. 202

Relationship between root and shoot traits
203

Root fresh weight and shoot biomass under each cutting depth
204 Cutting depth had a significant impact on shoot and root biomass in both ratoon crops, 205 especially the second (Fig. 4). Shoot biomass and root fresh weights followed a consistent trend, 206 increasing then decreasing in both the first and second ratoon crops, and were highest under T2 at 207 all three elongation stages (Fig. 4a, 4b). The highest root fresh weight was also observed under 208 T2 compared to T1 and T3, and a similar trend was observed in terms of the shoot fresh biomass 209 and root fresh biomass on the same DAH, especially in the second ratoon crop. Overall, there 210 was little difference between the changes in shoot and root biomass between the two ratoon crops, 211 with an increase in shoot biomass and a decrease in root biomass with increasing DAH in the 212 first ratoon crop.

220
Due to the similar trends in shoot and root biomass at each stage in both ratoon crops (Fig. 3), 221 correlation analysis was carried out between shoot fresh weight and root morphology. As a result, 222 shoot biomass was found to be positively correlated with root biomass, root length, root surface 223 area, root volume, root tip number, root fork number and root crossings in both ratoon crops (P < 224 0.01). Moreover, the r values of these correlations in the first and second ratoon crops respectively 225 were in the order of root biomass ( (Table S1).
A linear relationship was observed between the root morphological characteristics and shoot 229 biological yield in both ratoon crops, and between the root fresh weight and shoot fresh biomass, 230 the root volume and shoot fresh biomass, and the root fork number and shoot fresh biomass ( Fig.  231 5). R 2 values were much lower in the first ratoon crop than the second crop, and the scatterplots of 232 the first crop were much more concentrated than in the second crop, but the correlation coefficient 233 (r) was extremely significant (P＜0.01) as shown in Table S1. Moreover, the root tip number and 234 fresh shoot biomass, the root length and fresh shoot biomass, and the root surface area and fresh 235 shoot biomass all showed the same patterns (Fig. S1). This was not because the fresh weight of the 236 first year's shoots was poorly correlated with the root parameters; on the contrary, the correlations 237 were extremely significant in the first ratoon, but because the distribution of the shoot fresh weight 238 was more concentrated with the root parameters compared with the second ratoon crop. This may 239 be attributed to the cumulative effect of the stalk cutting position from the first ratoon crop to the 240 second, and also the difference in rainfall in June and July between the two seasons.  The results showed that the greatest root volume and root surface area were observed in roots 253 with a diameter of 1.5-2.0 and 1.0-1.5 mm, respectively. Moreover, values were greater under T2 254 than T1 and T3. The greatest root length and root tip number were observed in roots with a 255 diameter of 0-0.5 mm, with root length decreasing much slower than the root tip number as the 256 root diameter increased. Root volume and root surface area were therefore considered a similar 257 morphological index, since values reached a maximum at a middle root diameter of 1.0-2.0 mm. 258 Meanwhile, root length and root tip number reached a maximum in roots with the finest diameter 259 (Fig. 5,6). in the first (FR) and second ratoon crop (SR). Root volume was highest in roots with a diameter of 265 1.5-2.0 mm in both FR and SR, while the root surface area was highest at a diameter of 1.0-1.5 266 mm plant -1 : the amount in each cluster/ millable canes in the cluster. * P < 0.05, ** P < 0.01. Error 267 bars, SD. 268 The average root diameter was similar under all three cutting depths, at 0.87, 0.89 and 0.90 269 mm under T1, T2 and T3, respectively. Each root trait showed differences among treatments 270 within different root diameter ranges, although there were only slight differences among 271 treatments in roots with a diameter of 0-3.5 mm (Table 2). That is, the treatments did not change 272 the percentage of each part of root in each plant. In terms of root volume, 25.9%, 29.4%, and 273 15.6% of roots had a diameter of 1.0-1.5, 1.5-2.0 and 2.0-2.5 mm, respectively. The root surface 274 area of roots with a diameter of 0.5-1.0, 1.0-1.5 and 1. We also estimated the root traits per square meter and a 0.3 m soil depth. Accordingly, the 286 root fresh weight ranged from 167-246 g m -2 , root volume from 79-164 m 3 m -2 , root surface area 287 from 3628-7329 cm 2 m -2 , root length from 13,430-18,390 cm m -2 , root tip number from 288 36,538-68,473 m -2 , root forks number from 114,364-215,035 m -2 , and root crossings from 9,059 289 -12,038 m -2 . The amount of root forks was highest, followed by the number of root tips, possibly 290 because of the specific morphology of the root tip, the diameter of which decreases in a fitted 291 ellipse [26]. Some root tips will also have been destroyed or damaged during harvest. 292

Root morphological traits are affected most by cutting depth
293

294
The analysis of root morphology revealed that root volume was highest in roots with a 295 diameter of 1.5-2.0 mm in both ratoon crops, with significantly higher values under T2 (Fig. 6 a,  296 b). Root volume under T2 reached 6.98 and 11.85 cm 3 in the first and second ratoon crop, 297 respectively, in roots of 1.5-2.0 mm, 2.17 and 3.06 cm 3 higher than under T1 and T3, respectively 298 (P < 0.01). A significantly higher root volume was also observed under T2 in roots with a 299 diameter of 0-2.5 mm. However, in the first ratoon crop, no significant differences were observed 300 among treatments in roots with a diameter > 2.5 mm, and at a diameter of > 3.0 mm in the second 301 ratoon crop. 302 The root surface area was highest in roots with a diameter of 1.0-1.5 mm in both ratoon crops, 303 with significantly higher values under T2 (P < 0.01). Moreover, a significantly higher root surface 304 area was also observed under T2 in roots with a diameter of 0-2.5 mm. Meanwhile, no significant 305 differences were observed between treatments in roots > 2.5 mm in the first ratoon crop and > 3.0 306 mm in the second ratoon crop (Fig. 6 c, d). Under T2, the root surface area reached 200.8 and 307 341.75 cm 2 plant -1 in roots with a diameter of 1.0-1.5 mm in the first and second ratoon crops, 308 respectively, 63.6 and 119.42 cm 2 plant -1 higher than under T1 and T3, respectively. 309 A higher root volume and root surface area were therefore observed under T2 compared to 310 T1 and T3 in roots of different diameters, especially in the second ratoon crop, with an increase in 311 the maximum value as well as the total amount. A significant change was also observed at a root 312 diameter of 2.0-2.5 and 2.5-3.0 mm. Overall, between the first and second crops, only slight 313 increase was observed in T1 compared to T1 and T3 in terms of both root volume and root surface 314 area. 315

Effect of cutting depth on root length and root tip number according to root diameter
316 Cutting depth also affected the root length and root tip number in both ratoon crops. A 317 significant increase in both parameters was observed under T2 in roots with a diameter < 2.5 mm, 318 while no differences were observed at a diameter of 3.0-4.5 mm. Values were highest in roots with 319 a diameter of 0-0.5 mm (Fig. 7) and were significantly higher under T2 than the other treatments 320 in both ratoon crops (P < 0.01). Overall, both the root length and root tip number were greater in 321 T2 than T1 and T3 at different root diameters, especially in the second ratoon crop, with increase 322 in the maximum value as well as the total amount. Moreover, a significant change was also 323 observed at a diameter of 2.0-2.5 mm and 2.5-3.0 mm. Between the first and second crop, only a 324 slight increase was observed in T1 compared to T3 in terms of both root length and root tip 325 number.
326 Fig. 7. Effect of cutting depth on root length and root tip numbers. At the diameter of 0~0.5mm, 327 the root length and root tips number of each plant reached the highest, in both FR and SR. /plant: the amount of each cluster/ millable canes of the cluster. * P < 0.05. ** P < 0.01. Error bars, SD. 329 The longest root length and maximum number of root tips was also determined in roots with 330 a diameter of 0-0. 5 mm (Fig. 7.), with a clear decrease in root tip number compared to root length 331 at diameters > 0.5 mm. Under T2, the root length was 1218 and 1790 cm plant -1 in roots with a 332 diameter of 0-0.5 mm in the first and second ratoon crop, 283 and 389 cm plant -1 greater than 333 under T1 and T3, respectively. Moreover, the root tip number was 7184 and 10,608 plant -1 at a 334 diameter of 0-0.5 mm in the first and second crops, 1768 and 2315 plant -1 more than under T1 and 335 T3, respectively. 336

337
The above results show that cutting depth affected the root biomass, and the values of four 338 root components (root volume, root surface area, root length and root tip number) in roots within a 339 diameter range of 0-2.5 mm. However, there was little effect on any of the four root components 340 in roots with a diameter > 2.5. We therefore analyzed the relationship between shoot biomass and 341 the above root components. 342 The results showed better linear relationships than under the total range of root diameters 343 between the root morphological characteristics and shoot biomass in both ratoon crops. Clear 344 linear relationships were observed between the root volume and shoot fresh biomass, root surface 345 area and shoot fresh biomass, root length and shoot fresh biomass, and root tip number and shoot 346 fresh biomass in roots with a diameter of 0-2.5 mm (Fig. 8). The R 2 values were much lower in the 347 first ratoon crop compared to the second, and the scatterplots were more concentrated. Although 348 the above relationships all showed a similar pattern, the R 2 values were greater between the fine 349 root volume and root surface area plus the shoot fresh weight. This may be attributed to the 350 cumulative effect of stalk cutting position from the first ratoon crop season to the second. rhizosphere and efficiently absorb water and nutrients from the soil [29,30]. Root length, root 358 surface area, and root volume are therefore commonly used to evaluate the performance of a root 359 system [31,32]. In this study, root development was significantly impacted by cutting depth. 360 Cutting to depth of 5 cm below the surface (T2) increased activity of buds and deeper roots, 361 enhancing the ability to absorb nutrients and water from the soil, and thereby increasing 362 germination and root development. In contrast, cutting at a depth of 0 cm (T1) resulted in shallow 363 root development and a poor ability to absorb nutrients and moisture, while cutting to 10 cm 364 below the surface (T3) resulted in mostly dormant buds, whose poor activity resulted in relatively 365 poor root development. 366 Morphological attributes of the fine roots affect nutrient and water uptake as well as biomass 367 accumulation [33], and the length and diameter of fine roots play a pivotal role in nutrient and 368 water acquisition [34][35][36]. In this study, roots with a diameter of 0-2 mm comprised about 54% of 369 the total root length, about 69% of the root volume, and 73% of the root surface area. Meanwhile, 370 roots with a diameter of 0-0.5 mm represented about 90% of the root tip number, while those with 371 a diameter > 2mm contributed to only a small proportion. The proportion of fine roots in 372 sugarcane differs from that in wheat and pulse crops. For example, in pulse crops, roots with a 373 diameter of 0-0.2mm comprise about 60% of the total root length [37]. The classification of fine 374 roots also differs between crops and, in general is characterized by roots with a diameter of 0-0.2 375 mm in rice [38,39], wheat, and pulse crops [37]. However, in some species, such as poplar, switch 376 grass, cool-season pasture grasses, corn, soybean [40], cotton, maize and sorghum [41], the 377 difference between fine and thick or coarse roots is classified by a diameter of 2 mm. In terms of 378 the proportion of roots ( Table 2) and amount of roots (Figs. 5 and 6), the root volume, root surface 379 area and root length of roots with a diameter of 0-2.5mm played an important role in this study, while root tip number was important in roots with a diameter of 0-0.5 mm. 381 Positive correlations were also observed among root traits and shoot biomass, with traits in 382 roots with a diameter of 0-2.5 mm having the greatest impact on shoot biomass. We therefore 383 removed the data of root traits in roots with a diameter > 2.5 mm, and reanalyzed the linear 384 relationships between shoot biomass and root volume, root surface area, root length, and root tip 385 number in those with a diameter of 0-2.5 mm. As a result, we found that R 2 values were highest at 386 a diameter of 1.5-2.0 mm in both ratoon crops. Moreover, there were significant differences in 387 root traits at a diameter < 2.5 mm, and the number of root tips was significantly different at a root 388 diameter < 0.5 mm. As the root diameter increased, the differences among cutting depths rapidly 389 declined, suggesting that the fine roots are essential for growth and development of the 390 underground root system in perennial sugarcane. That is, the more the number of fine roots, the 391 greater the absorption of water and nutrients, and the higher the yield. 392 Studies suggest that sugarcane stools contain about 10-19 underground buds, and together with 393 the roots, these underground buds and stalks combine to form the ratoon stool or crown [42][43][44][45][46], 394 distributed about 15 cm below the soil surface. The upper section of the sugarcane stubble is 395 mainly composed of surface buds, with about 2-3 living buds. After germination, the roots remain 396 in shallow soil, and nutrient and water absorption is poor. The mid-section is the active bud 397 section, with about 4-5 active buds. After germination, the roots extend deep into the soil, 398 improving nutrient and water absorption, and promoting survival under drought conditions. The 399 lower buds are mainly dormant, and tend to remain so. Relatively deep cutting is therefore thought 400 to eliminate apical dominance of the upper buds and promote the germination of dormant buds on 401 lower underground stalks. 402 In this study, the position of the terminal buds decreased with increasing cutting depth from 0 403 to 10 cm. As a result, germination of these lower buds resulted in more nodes for root 404 development, which, in turn, increased opportunities for water and nutrient absorption compared 405 to germination of the upper buds. Thus, the shoot and root biomass were greater under T2 than T1. 406 Theoretically, the shoot and root biomass should also have been greater under T3 compared to T2; 407 however, the opposite was observed. One reason for this is thought to be the time required to come 408 out of dormancy and for development of active buds, with the majority of buds at this depth 409 remaining dormant as in Betula pendula [47]. This is supported by a previous study by Harrell et 410 al. [17] in ratoon rice, whereby a decrease in stubble height from 40 to 20 cm affected growth of 411 the ratoon crop by shifting the panicle point of origin during early growth and delaying maturity. 412 In our study, plant height was lowest under T3 (Table 1); however, the amount of millable cane 413 was highest, and the diameter was greater than under T1. This also suggests that lower terminal 414 buds benefit from a better growth environment after germination, although the drawback is 415 delayed growth of aboveground shoots. It was previously reported that sugarcane stems 416 originating from deeper soil produce a stronger root system [25], which is the basis of a good crop 417 stand. In this study, T2 resulted in the highest root biomass, which in turn, allowed sufficient 418 uptake of nutrients and water, increasing the shoot biomass compared with other treatments. 419 Overall, the findings of this study suggest that a cutting depth of 5 cm below the soil surface 420 promotes growth the root system, especially in terms of the root volume, root surface area and root 421 length of roots with a diameter of 0-2.5mm, thereby significantly improving the shoot biomass and 422 cane yield. This cultivation practice is a simple addition to ratoon cutting management, increasing 423 overall production of ratoon cane. This could therefore be applied successfully to rain fed slopes 424 in China and other Southeast Asian countries in line with manual harvesting techniques. 425 Meanwhile, in terms of mechanical harvesting, improvements could also be made to the cutting 426 machines, such as installing a plow tooth above the disc blade to allow removal soil with the cane 427 stalk or lowering of the disc blade to reduce abrasion. In line with this, technology has already 428 been extended to a production area of 25,000 hectares in the sugarcane district of Dehong State, 429 Yunnan Province, China, in 2018.

431
Deep cutting at depths of 5 and 10 cm below the soil surface (T2 and T3) considerably 432 increased yield compared to traditional cutting at the soil surface (0 cm, T1). The highest yield 433 increase was observed under T2, with an increase in production of 32.74 (34.9%) and 25.5 ton ha -1 434 (25.2%) compared to T1 and T3, respectively. These results suggest that the improvements in all 435 root traits across all root diameters supported a greater increase in shoot biomass and yield. This 436 was especially true in terms of root volume and root surface area in roots with a diameter of 0-2.5 437 mm, and root length and root tip number in those with a diameter of 0-0.5 mm, both of which also 438 showed a good linear relationship. Moreover, since buds at the soil surface and dormant buds 439 remained under T1 and T3, this may have led to a shallower stool and delayed vegetative growth 440 of the shoot, respectively. Cutting at a depth of 5 cm below the soil surface is therefore 441 recommended in terms of root biomass, subsequent shoot biomass, and overall sugarcane yield.