Breeding erect plant type sweetpotato lines using cross breeding and gamma-ray irradiation.

Few sweetpotato (Ipomoea batatas Lam.) cultivars with erect plant type are available despite their advantages over spreading type, such as simplicity of cultivation and ability to adapt to limited space. One of the reasons is insufficiency of their agronomic characteristics for table use. So, it is important to overcome these drawbacks of ER-type lines. We attempted to breed new erect plant type sweetpotato lines having good agronomic traits using cross breeding and mutation breeding with gamma-ray irradiation. With cross breeding we successfully developed new erect plant type lines with almost equal levels of yield as compared to 'Beniazuma', one of the leading cultivars in Japan. However, mutation breeding failed to develop any promising lines because we could not obtain distinct erect plant type lines. In the future larger numbers of plants should be used for mutation breeding, and irradiation methods should be improved.

Introduction

Sweetpotato (Ipomoea batatas Lam.) lines with erect plant type (abbreviated as ER-type), also known as dwarf sweetpotato lines, have advantages. One advantage is the simplicity of cultivation. Vines of ER-type sweetpotato lines are short and suitable for machine transplanting. Another advantage is that weeds growing between rows can be removed by machinery since the inter-row space is not covered by vines, as is the case with normal spreading lines. Moreover, sweetpotato cultivars of a vigorous ER-type are less affected by weeds than cultivars of a spreading plant type (Harrison and Jackson 2011). In addition, ER-type sweetpotatoes are suitable for a kitchen garden because their vines do not grow extensively, facilitating cultivation in a limited space. However, only a few cultivars with ER-type growth habit have been bred in Japan, such as ‘Tsurusengan’ (Shiga ) for feed use and ‘Tama-akane’ (Sakai 2009) for brewing. ‘Benikomachi’ (Sakai ), bred for table use, possesses short, thick vines and is classified as a compact plant type (Tarumoto ). In spite of the high palatability of its storage roots, the cultivation area of ‘Benikomachi’ is very limited due to the irregular shape of its storage roots and its susceptibility to stem rot. An ER-type sweetpotato line may be marketable if these drawbacks are minimized. So, it is important to improve agronomic characteristics of ER-type lines.

Field trials revealed that ER-type sweetpotato lines were not as competitive as normal lines and they were eliminated during the selection process (Sakai and Sirasaka 1957). Hence, the selection of seedlings should not be focused on yield in preliminary small-scale trials. Field performance tests for ER-type lines should be carried out in sufficiently large plots with at least four rows, in order to estimate yield correctly (Sakai and Sirasaka 1957).

Generally, cross breeding using ER-type parent lines has been used to breed ER-type cultivars (Toyama , 2005). On the other hand, the mutation breeding method could be useful because new ER-type lines with good agronomic traits could be bred in a short amount of time by mutating major cultivars.

In this study, both breeding methods were used to develop ER-type sweetpotato lines. We then checked the productivity and quality characteristics of the new lines to select ones that were practical. Furthermore, we tested a new index (plant height/plant width ratio) that would be useful in the selection of plant type.

Materials and Methods

We distinguished seven plant types; 1: very erect; 2: erect; 3: compact; 4: semi-compact; 5: moderate; 6: semi-spreading; and 7: spreading (Tarumoto ). The detailed description of these plant type is shown in Table 1. We judged plant type index of ‘Beniazuma’ in early-July to mid-July for 5 (moderate) and the index slightly increased after August by the elongation of vines. Plants with indices lower than four were defined as ‘ER-type’ and used for screening. Different plant types of lines are shown in Fig. 1.

Table 1

Plant type and traits of parent lines/cultivars and Japanese standard cultivar ‘Beniazuma’

Lines and cultivars	Plant type indexa	Number of primary branchesb	Plant height (cm)	Plant width (cm)	Division	JP numberc
Boniato 14	1	7.0	30.0	52.0	introduced line	174863
Tsurusengan	3	10.7	34.0	75.3	bred cultivar	168605
Benikomachi	3	5.0	28.8	82.2	bred cultivar	168603
Tokyokintoki	3	–	–	–	native line	–
Beniazuma	5	4.3	18.0	104.7	bred cultivar	168606

1: very erect (Internode length and vine length are very short. Vines grow vertically.); 2: erect (Internode length and vine length are short. Vines grow nearly vertically.); 3: compact (Internode length and vine length are short. Vines grow slant ways.); 4: semi-compact (The middle of compact and moderate.); 5: moderate (Internode length and vine length are moderate. Vines grow nearly horizontally.); 6: semi-spreading (The middle of moderate and spreading.); 7: spreading (Internode length and vine length are long. Vines grow horizontally.).

Number of primary branches longer than 1 cm.

Accession number of NIAS Genebank, Japan.

The investigation of plant type index, primary branches, plant height and plant width were carried out on early-July 2001.

Fig. 1

Different plant types of sweetpotato lines. A: Erect (Plant type index: 2), B: Moderate (Plant type index: 5). Photos were taken on 22 July 2011 in Tsukubamirai, Ibaraki.

1. Breeding of ER-type sweetpotato lines by cross breeding

Cross breeding using ER-type sweetpotato germplasm was initiated in 2000 in the glasshouses of the National Agriculture Research Center (Tsukubamirai, Ibaraki) and the National Agricultural Research Center for Kyushu Okinawa Region, Japan (Miyakonojo, Miyazaki). The characteristics of parent lines/cultivars are summarized in Table 1. Starting in 2007, we used parent lines possessing good agronomic traits to breed the practical lines. The breeding process consisted of four steps, as follows.

Year 1 (selection of seedlings): Seeds were sown in the nursery beds in late-March. Vines were cut from the seedlings in mid-May and vertically transplanted to an experimental field (Tsukubamirai, Ibaraki). They were planted in rows, with intervals of 25 cm between plants and 100 cm between rows. The rows were covered with plastic film mulch. Before transplanting, 60 g m−2 of chemical fertilizer (3N-12P-10K) was applied. The amount of chemical fertilizer was reduced to 40 g m−2 starting in 2007. In mid-July, plant type was checked, and ER-type individuals were marked and selected. At the end of October, storage roots were harvested, and further selection was made based on the root shape of each plant. Plants with irregular root shape were excluded.

Year 2: In mid-March, selected storage roots were planted individually in a nursery bed. Eight to ten vines per storage root were cut and planted in the experimental field. The cultivation methods from year 1 were used. Plant type was determined in mid-July. Storage roots were harvested and selected for quality in mid-October and the lines were designated.

Year 3: Five storage roots per line were planted in a nursery bed. Thirty-two vines were obtained from each line and planted in the field. Each experimental plot consisted of four rows, each with eight plants. The cultivation conditions and schedule from year 1 were used. Some ER-type lines found in the breeding program for table and processing use were also included.

Year 4 and after: All experimental procedures were the same as years 2 and 3, except there were two replications. ER-type lines were selected for table use and processing use. The plant height and the plant width were measured in July, three times from 2013 to 2014. The plant height was the length from the plant foot to the tip of the highest leaf, and the plant width was the distance that was the longest between the tips of the vines.

2. Breeding through mutation by gamma-ray irradiation

Cultivars and lines that received gamma-ray irradiation were ‘Beniazuma’ (irradiated in 2006 and 2007), ‘Purple Sweet Lord’ (2006), ‘Quick Sweet’ (2007), and ‘Himeayaka’ (2006 and 2007).

Gamma-ray irradiation was conducted at the Institute of Radiation Breeding (National Institute of Agrobiological Sciences) in mid-June. In 2006, two doses of gamma-ray irradiation (2 and 10 Gy hr−1) were applied to sweetpotato vines for 20 hours. Each vine received 5 Gy hr−1 of gamma-ray irradiation for 20 hours in 2007. Vines were planted in the field immediately after treatment (Tsukubamirai, Ibaraki). Cultivation methods followed the methods used for cross breeding. The vines were examined for plant type and bud growth. Mutated plants were visually selected.

Vines of selected plants were planted in pots and grown in a glasshouse. A maximum of 20 vines per plant were cut and planted in the field in late-May of the following year. Cultivation conditions were the same as the previous year. Plant type was evaluated in mid-July. Storage roots were harvested and yield characteristics were investigated in late-October.

Results

1. Cross breeding

(1) Breeding trials in 2001–2003

The plant type of the seedlings from each cross was determined in 2001 and is summarized in Table 2. Although growth habit of the seedlings tended to be spreading type rather than ER-type, erect type offspring (plant type 2) also segregated from crosses even when both parents were of compact type (plant type 3). Some crosses produced moderate type offsprings (plant type 5), but no semi spreading type plants (plant type 6) were produced.

Table 2

Plant type index of seedlings from the cross between parental lines in 2001

Cross designation	Seed parent	Pollen parent	No. of seedlings	No. of seedlings in each plant typea						Average of all seedlings	No. of selected seedlings	Average of selected seedlings
				1	2	3	4	5	6–7
00S01	Boniato-14 (1)	Benikomachi (3)	29	0	15	10	4	0	0	2.6 ± 0.7	3	2.0
00S02	Benikomachi (3)	Boniato-14 (1)	10	0	6	4	0	0	0	2.4 ± 0.5	1	2.0
00S03	Boniato-14 (1)	Tsurusengan (3)	28	0	15	13	0	0	0	2.5 ± 0.5	4	2.3
00S04	Tsurusengan (3)	Benikomachi (3)	65	0	22	30	13	0	0	2.9 ± 0.7	4	2.3
00S05	Benikomachi (3)	Tokyokintoki (3)	14	0	2	7	5	0	0	3.2 ± 0.7	2	2.5
00S06	Tokyokintoki (3)	Benikomachi (3)	15	0	2	9	4	0	0	3.1 ± 0.6	1	2.0
00S07	Tsurusengan (3)	Tokyokintoki (3)	29	0	7	15	5	2	0	3.1 ± 0.8	1	2.0
00S08	Tokyokintoki (3)	Tsurusengan (3)	23	0	5	11	6	1	0	3.1 ± 0.8	2	2.5

1: very erect to 7: spreading, see Table 1.

Figures in parentheses indicate plant type on early-July 2001.

Average data are shown as average ± standard deviation.

The investigation was carried out on 6 July 2001.

Seedlings of plant type 2, in addition to type 3 seedlings with good root shape, were selected and evaluated again for their growth habit in 2002. The plant type of the offsprings was almost unchanged (nine lines remained unchanged; four declined, and one improved).

Lines ‘00S01-21’ (plant type 3) and ‘00S04-10’ (plant type 4) were selected and compared with the leading cultivars, ‘Beniazuma’ and ‘Kokei No. 14’, for table use characteristics, in 2003. These lines were inferior to ‘Beniazuma’ in yield and taste.

(2) Breeding trials in 2008–2013

The ER-type seedlings were selected from the 2008 to 2010 trials. These seedlings were used for further selection the next year. The segregating rate of ER-type seedlings in 2008 was 4.4% (40 ER-type/899 all) when ER-type parents were used. On the other hand, only 1.0% (108 ER-type/10,593 all) ER-type seedlings were obtained from the crosses between non-ER-type parents. Selection rate was low in 2008, 0.2% (18 seedlings were selected from 11492 seedlings) since seedlings from non-ER-type parents were included. From 2009 to 2010, the selection rates were higher than in 2008 and were 8.7% and 4.6% respectively. These seeds were obtained from the crosses using ER-type parents.

Erect-type lines ‘07074-1d’, ‘08138-20d’, and ‘09121-16d’ were selected from the seedlings and compared with reference cultivars and lines, in terms of their important traits (Table 3).

Table 3

Yield and taste of the steamed roots of erect-compact type sweetpotato lines

Year	Cultivars and lines	Plant type indexa		Storage root weightb (kg a−1)	Average weight of a storage root (g)	Characteristics of steamed roots
		Early	Late			Flesh color	Texturec	Tasted
2011–2014	07074-1d	2.4 ± 0.4	2.5 ± 0.4	299 ± 50	140 ± 22	light yellow white	moderately mealy	moderate
	Beniazuma	5.0 ± 0.0	5.0 ± 0.0	290 ± 21	222 ± 25	light yellow	moderately mealy	moderately good
	Kyukei 281	2.6 ± 0.4	2.8 ± 0.2	175 ± 25	113 ± 17	yellow white	moderate	moderate
2012–2014	08138-20d	3.0 ± 0.0	3.2 ± 0.2	286 ± 12	157 ± 16	light orange	moderate–moderately mealy	moderately good
	Beniazuma	5.0 ± 0.0	5.0 ± 0.0	301 ± 11	234 ± 15	light yellow	moderately mealy	moderately good
	Kyukei 281	2.5 ± 0.4	2.8 ± 0.2	172 ± 28	119 ± 15	yellow white	moderate	moderate
2013–2014	09121-16d	3.0 ± 0.0	4.0 ± 0.0	333 ± 53	119 ± 14	yellow white	moderate	moderate–moderately good
	Beniazuma	5.0 ± 0.0	5.0 ± 0.0	308 ± 5	225 ± 7	light yellow	moderately mealy	moderately good
	Kyukei 281	2.5 ± 0.5	3.0 ± 0.0	154 ± 14	110 ± 10	light yellow–yellow white	moderate	moderate

1: very erect to 7: spreading, see Table 1. The investigation was carried out on mid-July (early) and early-August (late).

Roots weighing less than 50 g were eliminated.

sticky, moderately sticky, moderate, moderately mealy, mealy.

poor, moderately poor, moderate, moderately good, good.

Average data are shown as average ± standard deviation in each trial years. 32 plants were planted in each experimental plot.

Line ‘07074-1d’, originating from a cross between ‘Beniharuka’ and ‘Tama-akane’ (compact plant type), was estimated as an ER-type line in seedling selection in 2008, and was subjected to a performance test in 2009. As a result, the plant type index of this line was determined to be 2.4 (mid-July) and it was selected as a promising ER-type line (Fig. 2). Line ‘07074-1d’ remained erect until its late growth stage. The taste of the steamed roots was moderate.

Fig. 2

Growth habit of the erect-plant type line ‘07074-1 d’ (right) and the leading cultivar for table use, ‘Beniazuma’ (left). Photos were taken on 10 July 2013 in Tsukubamirai, Ibaraki.

Line ‘08138-20d’, originating from a cross between ‘Himeayaka’ and ‘Tama-akane’, was estimated as an ER-type line in seedling selection in 2009, and was subjected to a performance test in 2010. As a result, the plant type index of this line was determined to be 3.0 (mid-July) and it was selected as an ER-type line. This line contains carotenoid in the roots, and the taste of the steamed roots was moderately good.

Line ‘09121-16d’, originating from a cross between ‘Kanto No. 134’ and ‘Tama-akane’, was estimated as an ER-type line in seedling selection in 2010, and was subjected to a performance test in 2011. As a result, the plant type index of this line was determined to be 3.0 (mid-July) and it was selected as an ER-type line. The taste of the steamed roots was moderate to moderately good.

The yields of these three ER-type lines were almost equal to that of ‘Beniazuma’ and were higher than that of ‘Kyukei 281’, an ER-type line bred by the NARO Kyushu Okinawa Agricultural Research Center.

These features of ‘07074-1d’, ‘08138-20d’ and ‘09121-16d’ made these lines promising.

The plant height (PH) of these ER-type lines was almost equal to that of ‘Kyukei 281’ in early-July, while that of ‘Beniazuma’ was almost half (Fig. 3). The plant height of all lines were elongated in mid-July, especially in ‘09121-16d’. The plant width (PW) of these ER-type lines was almost equal to that of ‘Kyukei 281’. The plant width of the lines increased approximately two fold in mid-July, but the elongation was not clear in ‘07074-1d’. On the other hand, it was very clear in ‘Beniazuma’, the standard cultivar.

Fig. 3

Plant height (PH), plant width (PW) and PH/PW ratio of selected lines. The investigation was carried out in early- mid-July in 2013–2014. Error bars indicate standard deviations. 6 plants were observed in each plot.

The PH/PW ratio of these ER-type lines was almost 0.5 in early-July and equal to that of ‘Kyukei 281’. That of ‘Beniazuma’ was about 0.3. The ratio of ‘08138-20d’ was lower in mid-July, while that of ‘07074-1d’ and ‘Kyukei 281’ remained unchanged. The ratio for ‘Beniazuma’ was low and decreased in mid-July, to 0.2.

2. Mutation breeding through gamma-ray irradiation

Only a few plants were selected from all of the cultivars (Table 4) since gamma-ray irradiation resulted in growth inhibition and deformity of leaves. The influence was obvious in plants that received radiation of more than 100 Gy. When ‘Purple Sweet Lord’ received 200 Gy radiation, more than 90% of the plants were negatively affected, showing limited growth or death. When ‘Beniazuma’ and ‘Purple Sweet Lord’ received 40 Gy radiation, no plant was selected as ER-type. Vines of some plants were selected from 4 cultivars respectively to examine plant type. Variability in plant type was not observed within plants from any cultivars. We could not recognize typical chimera plant.

Table 4

Change in growth type after gamma-ray irradiation and the number of selected plants

Year	Cultivars	Amount of irradiation (Gy)	No. of vines	Types of growth		No. of selected plants
				Almost normal	Stopped or dead
2006	Beniazuma	40	100	95	5	0
		200	100	53	47	5
	Purple Sweet Lord	40	96	93	3	0
		200	98	6	92	1
	Himeayaka	200	201	87	114	2
2007	Beniazuma	100	187	66	121	5
	Quick Sweet	100	190	137	53	5
	Himeayaka	100	188	98	90	2

Plant type and yield of the lines selected in 2006 and 2007 were investigated from 2008 to 2010 in the field. Plant type index was at first scored lower by 1 to 2 in the selected lines than in their original cultivars/lines but they became indistinguishable until the late growth stage (Table 5). The selected lines were inferior in yield to their original cultivars/lines except the lines selected from ‘Himeayaka’ in 2007.

Table 5

Yield of the selected lines created with gamma-ray irradiation

Year	Cultivars and linesa	Plant type indexb		Storage root weight (kg a−1)	Average weight of a root (g)
		Early	Late
2008–2009	Beniazuma 06Gy-1	4.0	5.0	272	222
	Beniazuma 06Gy-2	3.0	5.0	255	243
	Beniazuma (original)	4.5	5.0	373	278
2009–2010	Beniazuma 07Gy-1	3.5	4.5	326	215
	Beniazuma (original)	4.5	5.0	361	258
2008, 2010	Himeayaka 06Gy-1	4.0	4.0	196	126
	Himeayaka 06Gy-2	3.0	4.0	88	107
	Himeayaka (original)	4.0	4.5	297	154
2009–2010	Himeayaka 07Gy-1	4.0	4.5	241	123
	Himeayaka 07Gy-1-4	4.0	4.3	243	122
	Himeayaka (original)	4.5	5.0	200	123
2008	Purple Sweet Lord 06Gy-1	4.0	4.0	198	127
	Purple Sweet Lord (original)	5.0	5.0	544	266
2009	Quick Sweet 07Gy-1	4.0	4.0	291	264
	Quick Sweet (original)	4.0	4.0	307	241

06 Gy: Irradiated in 2006. 07 Gy: Irradiated in 2007.

1: very erect to 7: spreading, see Table 1.

Early: estimated in early-July; Late: estimated in early-August to mid-October.

Discussion

ER-type sweetpotatoes could be developed for both table use and processing, since we found some promising lines in the breeding program. Breeding of sweetpotatoes has emphasized the root characteristics but not the plant type. Yamakawa and Sakamoto (1978) investigated the effect of seedling lines on plant type and found that ER-type offspring occurred only rarely. We found the same trend in the crosses between erect/compact parents and spreading ones, but confirmed that the ER-type were more likely to occur from crosses using these ER-type parents than from crosses between non-ER-type parents. These findings indicate the need to choose the right parents. We have selected some promising lines from the progenies of some new Japanese cultivars, for example, ‘Beniharuka’, ‘Himeayaka’, and ‘Tama-akane’. It is important to use parent lines that possess good agronomic traits.

As Sakai and Sirasaka (1957) pointed out, erect seedlings should be carefully selected since they are not as competitive as spreading ones. McLaurin and Kays (1993) grafted five vine length types onto the same rootstock to investigate the relationship between plant type and yield. They found that yield was low in the erect plant type. Sakamoto (1984) reported that vegetative growth in the early stage was related to plant type. On the other hand, Yamakawa and Sakamoto (1978) suggested that breeding ER-type sweetpotato lines with a high yield is possible through the selection of plants with a high growth rate in the early stage. We successfully developed erect/compact type lines with almost equal levels of yield as ‘Beniazuma’, the leading cultivar for table use. Therefore, it is possible to breed high quality ER-type cultivars with high root yield.

The judgement of plant type index should be done in early-July to mid-July because of elongation of vines after August. We selected the ER-type lines using the plant type index, but it may be possible to select lines more clearly using the PH/PW ratio, because the PH/PW ratio indicates the difference of plant type between ER-type plant and non-ER-type plant objectively, and the ratios of ER-type plants are generally stable from early-July to mid-July. Further examination is needed to confirm the effect of screening using the PH/PW ratio.

ER-type sweetpotatoes, such as ‘Tsurunasi-genzi’ and ‘Tachi-kagoshima’ were presumably bred through mutation of native cultivars at the beginning of the 19th century (Fujise 1965). Spontaneous mutation is less likely to produce erect plant type. We used gamma-ray irradiation and we could not find any promising ER-type plants using a limited number of plant materials, but we obtained some moderately erect plants. When the plants received 40 Gy radiation, no plant was selected. So, there is a possibility that the plants must receive more radiation to obtain ER-type lines. We did not distinguish each buds of irradiated plants, but it might be necessary to do that to obtain ER-type plants efficiently. In the future, larger numbers of plants should be used and irradiation methods should be improved to obtain ER-type lines.

The elongation of vines and petioles is regulated by different physiological and morphological systems (Suge 1979). We failed to find plants with short petioles in ER-type lines. Elongation of the petioles should contribute to the three dimensional position of leaves allowing for more efficient photosynthesis, especially in ER-type lines. These lines would be expected to also have the potential for high yield and good quality making them marketable.