Plant Soil (2010) 328:313–325 DOI 10.1007/s11104-009-0111-1
REGULAR ARTICLE
Genotypic differences in nitrogen efficiency of white cabbage (Brassica oleracea L.) Gunda Schulte auf’m Erley & Elsa Rakhmi Dewi & Olani Nikus & Walter J. Horst
Received: 10 March 2009 / Accepted: 14 July 2009 / Published online: 8 August 2009 # Springer Science + Business Media B.V. 2009
Abstract In vegetable production, N balance surpluses are especially high which increases the risk of environmental pollution. The cultivation of Nefficient cultivars may contribute to alleviate the problem. A 2-year field experiment was conducted with eight white cabbage cultivars of three different maturity groups at two N fertilization levels. Genotypes differed both in N efficiency (head fresh weight at low N supply) and in yield at high N supply. These differences were not related to N uptake but to N utilization efficiency. At low N supply, harvest index was the main determining factor for genotypic yield differences. For earlier maturing cultivars a slower leaf emergence was responsible for the low harvest index. The response of the cultivars to low N supply was dependent on the weather conditions, particularly temperature, (highly significant year × cultivar × N supply interaction) at early growing stages. This suggests that breeding of cultivars with generally low-temperature tolerance could contribute to enhancing N utilization. Especially at high N supply, a high N harvest index was important for yield formation
Responsible Editor: Hans Lambers. G. Schulte auf’m Erley (*) : E. R. Dewi : O. Nikus : W. J. Horst Institute for Plant Nutrition, Leibniz University of Hannover, Herrenhäuser Str. 2, 30419 Hannover, Germany e-mail:
[email protected]
due to its effect on head water accumulation. For late cultivars, a high N retranslocation from leaves to the heads was related to yield both at low and high N supply. The study suggests that breeding of Nefficient cultivars may reduce N release to the environment by reducing the necessary N input and reducing the N content remaining in the crop residues. Keywords N limitation . N uptake . N utilization efficiency . Head fresh weight . Head water accumulation . Brassica
Introduction In vegetable production environmental pollution by nitrogen (N) losses into the atmosphere and hydrosphere is especially high (Greenwood 1990). This is caused by the high N fertilization levels which are common in commercial production and often exceed official recommendations (Booij et al. 1996). Especially for Brassica cabbage species, high fertilizer rates are recommended; for white cabbage in a range of 250 kg N ha−1 to 350 kg N ha−1, depending on maturity group (Scharpf and Weier 1994). During the past years, efforts have been made to decrease fertilization levels without affecting yield (head fresh weight per unit surface area). Experiments with varying N fertilization techniques, e.g. by split application or band placement, revealed only a limited reduction potential of fertilizer N for different
314
Brassica vegetables (Everaarts et al. 1996; Everaarts and De Moel 1998). Another possibility to save N fertilizer is a more precise prediction of the N demand. For this purpose, models predicting the time course of cauliflower growth (Alt et al. 2000), soil N availability for cauliflower N uptake (Kage et al. 2003) or the time course of N uptake of white cabbage and Brussel sprouts (Fink and Feller 1998, 2001) have been developed. A further approach which was hardly investigated yet is the breeding of N-efficient genotypes, which are characterized by a low susceptibility in yield to reduced N fertilization levels (Schenk 2006). A high N efficiency can be achieved either by a high N uptake or by an efficient N utilization for yield formation (Sattelmacher et al. 1994). Nitrogen recovery from the soil is generally high for Brassica vegetables (Everaarts 1993). At harvest, soil Nmin contents below white cabbage are normally less than 40 kg N ha−1 at 0 cm and 90 cm soil depth with almost no N leaching during the vegetation period (Everaarts and Booij 2000 and references therein). The efficient N uptake of white cabbage is caused by the large and evenly distributed root system, reaching down to a soil depth of 2–2.5 m, and by a high ability of the roots to absorb N in all soil layers (Kristensen and Thorup-Kristensen 2004). Total dry matter production and head yields as well as plant N contents are strongly influenced by the level of N supply (Peck 1981). Ontogenesis, in contrast, does not seem to be influenced by N rate (Peck 1981), which was also found for cauliflower (Everaarts and De Moel 1995). The number of leaves developed until heading and the harvest index (Hara et al. 1982; von Brandis and Scharpf 1987) were not found to be affected by the N rate. However, sucrose accumulated in outer leaves under N-limiting conditions together with enhanced sucrose concentrations in the heads (Hara 1989) indicating a limitation for the utilization of C for growth. Head quality is impaired under excess N supply because of burst heads, tipburn (Peck 1981), lower sucrose concentration in the heads and inferior taste (Nilsson 1988; Hara 1989). Although uniformity of the heads was slightly improved and head shape was not affected by increasing N fertilization, relative core length was increased leading to a decline in head quality (Everaarts and De Moel 1998). Dry matter content of the heads usually declines with increasing N
Plant Soil (2010) 328:313–325
fertilization (Peck 1981; Everaarts and De Moel 1998). However, the ratio of marketable heads and storage losses are only slightly affected by N fertilization and do not counterbalance the yield increase (von Brandis and Scharpf 1987; Freyman et al. 1991). Summarizing, the potential for an improvement of N efficiency by increasing N uptake seems to be rather limited, but there might be potential to improve N utilization efficiency. Due to the small effects of N rate on harvest index it is difficult to predict if a higher total biomass production or an enhanced head growth might be more effective. Severe yield limitations by impaired head quality are not to be expected. The objective of this study was, therefore, to explore genotypic differences in N efficiency for cabbage cultivars of differing maturity groups and to investigate the main factors contributing to yield at limiting and optimum N supply. This is expected to help finding selection criteria for the breeding of Nefficient cultivars.
Materials and methods Field experiments were conducted in 2004 and 2005 on a loamy silt soil (FAO 2006) at the experimental station of the Faculty of Natural Sciences, Leibniz University of Hannover, Germany in Ruthe located 20 km south of Hannover. The experiments were designed as split plots with N fertilization rates as main plots and cultivars as sub-plots with four replicates. N rates comprised no N fertilization and an N supply of 300 kg N ha−1 (fertilizer plus soil mineral N content at transplanting). Eight white cabbage cultivars from different maturity groups were used in the study: Parel (Bejo, Warmenhuizen, The Netherlands) as an early cultivar, Toughma (Rijk Zwaan, Welver, Germany), Castello (NickersonZwaan, Edemissen, Germany) and Perfecta (Bejo, Warmenhuizen, The Netherlands) as mid-early cultivars and Lennox, Bartolo (Bejo, Warmenhuizen, The Netherlands), Bloktor and Novator (Syngenta, Enkuizen, The Netherlands) as late cultivars. Castello and Bartolo are used as reference cultivars in variety testing for the registration of new varieties by the Bundessortenamt, Hannover, Germany, for mid-early and late white cabbage cultivars, respectively. The other cultivars were chosen because of recommendations of plant breeders and were all considered as N-efficient.
Plant Soil (2010) 328:313–325
315
mid-October). Head formation was defined to start when the newly developing leaves had a curved shape so that they covered the shoot apex. The harvest at maturity was performed when the heads reached their final size and maximum density. Replicate plots were harvested at the same day. Six plants per plot were randomly chosen for each harvest. Plants were separated into outer leaves and head including the stem. In 2005 the stem and head were separately harvested at maturity. Leaves were counted for each plant individually and then bulked for fresh weight determination. Sub-samples were taken for dry weight determination. The samples were dried at 70°C until they attained a constant weight. Nitrogen concentrations of the dried and ground plant fractions were determined using a CNS analyzer (Vario EL, Elementar Analysensysteme, Hanau, Germany). At heading, stems and heads were pooled for six plants, weighed, dried and ground for N analysis. At maturity, heads were weighed individually for each plant. Three heads per plot were sub-divided for leaf counting, then pooled together before sub-samples were taken for dry weight determination and N analysis. Water-soluble carbohydrate concentrations of outer leaves and heads were analysed using dried plant material. One hundred mg dry matter was extracted for 1 h in 20 ml distilled water and analysed by the anthrone method following the description of Yemm and Willis (1954). For nitrate analysis 100 mg dry matter was extracted for 1 h in 20 ml distilled water and nitrate concentration was determined according to Cataldo et al. (1975).
Seeds were sown in peat cubes and raised in a greenhouse for the first 3 weeks. Thereafter, the seedlings were placed in an open greenhouse to adapt the plants to the natural climatic conditions. After 5 weeks, plants were transplanted into the field on May 13 and May 19 in 2004 and 2005, respectively. Irrigation was supplied to ensure growth of the young plants. Plant spacing was 0.60 m by 0.48 m in 2004 and 0.55 m by 0.48 m in 2005 giving a plant density of 3.3 and 3.8 plants m−2 in 2004 and 2005, respectively. Individual plots were 2 m wide and 8 m long. Soil mineral N contents (0–0.9 m) at transplanting were 83 kg N ha−1 and 73 kg N ha−1 in 2004 and 2005, respectively. The high-N plots were fertilized with 150 kg N ha−1 as calcium ammonium nitrate prior to planting, and with 70 kg N ha−1 and 80 kg N ha−1 on 10 June 2004 and 22 June 2005, respectively. Weeds were controlled by hand. Oxydemeton-methyl (Metasystox) and Bacillus thuringiensis (Dipal/Turex) were sprayed for pest control in both years. Weather conditions varied between the two growing seasons studied (Fig. 1). In comparison to the long-term average, 2004 was cool and wet between May and July and warm and dry in September and October. 2005 was warmer than usual except in August, while precipitation was high in May and July, very low in June and close to normal from August to October. Harvests were performed at the beginning of heading (from mid-June to end of July, according to maturity group) and at maturity (from end of July to
20
150 2004 2005 long-term
100
10
50 5
0
0
May
June
July
Aug
Sept
Oct
Temperature [˚C]
15
Precipitation [mm]
Fig. 1 Total monthly precipitation (bars) and monthly temperature (symbols) at the experimental station Ruthe during the 2004 and 2005 growing seasons and as long-term average
316
Plant Soil (2010) 328:313–325
In addition to the leaf counting during harvests, leaf numbers were determined non-destructively in 2005. Six plants per plot were chosen at heading and all outer leaves were numbered using a water proofed marker. In weekly intervals from heading to maturity, the number of the oldest leaf still attached to the plant and of the youngest unfolded leaf was recorded. From these values the ratio of lost leaves was calculated by dividing the number of leaves shed from the plant by the total number of outer leaves formed until maturity. Nitrogen efficiency was defined in this study as yield (head fresh weight) at limiting N supply (Craswell and Godwin 1984). Factors contributing to N efficiency are N uptake efficiency (shoot-N uptake at limiting N supply) and N utilization efficiency (yield per unit shoot-N taken up). N utilization efficiency was subdivided into biomass production efficiency (Ortiz-Monasterio et al. 1997), harvest index and head fresh weight/dry weight ratio. Biomass production efficiency was defined as the ratio of shoot dry weight and shoot-N uptake at maturity. Harvest index was calculated as head dry weight divided by total shoot dry weight at maturity. Since yield of white cabbage is given by head fresh weight, the ratio between head fresh and dry weight was introduced as a third factor of N utilization efficiency. Nitrogen harvest index was calculated as head nitrogen at maturity divided by total shoot-N uptake. Data were analysed using SAS version 9. Analysis of variance was performed by PROC MIXED. Years, N rate and cultivars were considered as fixed factors with years as main-plot factor, N rates as sub-plot
factor and cultivars as sub-sub-plot factor. Multiple comparisons of means were made by the LSMEANS/ PDIFF statement for N rate by cultivar interactions. Regression lines were calculated using PROC REG in SAS.
Results Maturity times differed between cultivars and N rates (Table 1). Cultivar differences were according to their maturity group. Within mid-early cultivars, cv Toughma matured earlier than cv Castello, while cv Perfecta matured later. Low N supply delayed maturity, especially for the early cultivars. This effect was stronger in 2004 than in 2005. Heading dates did not differ between N rates. Cultivars differed in yield (head fresh weight) at both N rates (Table 2). At low N supply, cv Perfecta had the highest yield of all cultivars. Within the late maturing cultivars, cv Lennox and cv Novator had a significantly higher yield than cv Bartolo. The early cultivar Parel could not be compared to another cultivar of the same maturity group. However, yield was not decreased by low N supply. The cultivars Parel, Perfecta, Lennox and Novator were classified as N-efficient, while cvs Toughma, Castello and Bartolo as N-inefficient. Cultivar Bloktor had an intermediate N efficiency. At high N supply, yield differences were more pronounced between cultivars with different maturity times. Again, cv Perfecta had the highest yield of all cultivars. Mid-early cultivar
Table 1 Heading and maturity dates (days after planting) of eight cabbage cultivars grown in Ruthe at two N supplies (N1: no fertilization and N2: 300 kg N ha−1) in 2004 + 2005 Cultivar
2004 Heading N1 + N2
Parel
31
2005 Maturity N1 75
Maturity N2 62
Heading N1 + N2 34
Maturity N1 67
Maturity N2 62
Toughma
31
105
99
39
91
85
Castello
40
113
105
46
98
91
Perfecta
56
126
117
54
111
104
Lennox
76
159
152
67
140
139
Bartolo
76
159
152
67
140
139
Bloktor
77
159
154
69
140
139
Novator
77
159
154
69
140
139
9151 10172 13128
11294 10119 10632 10533
Toughma Castello Perfecta
Lennox Bartolo Bloktor Novator ns *** *** *** ns * *
10843 9275 10140 9452
8865 9153 12551
6359
8761 7287 7961 7880
7822 7612 10145
5908
B D C CD
E D A
F
b de cd bc
e e a
f
39.2 39.5 36.8 38.2
19.9 20.3 23.6
14.9
21.3 21.6 21.4 23.2
13.3 13.1 17.8
7.7
ns *** *** *** ns *** ***
33.2 32.7 31.1 29.3
24.7 29.5 32.6
16.8
21.9 20.5 20.7 20.4
18.6 18.4 19.7
11.5
A A AB B
E D C
F
ab ab ab a
c c b
d
45.0 45.0 46.0 42.8
33.9 31.2 37.7
23.3
64.6 61.4 64.2 59.2
39.0 46.7 52.9
40.4
2004
ns *** *** ns ns * ns
50.2 52.6 47.5 45.2
36.2 38.4 44.0
30.4
65.9 70.7 56.4 59.2
49.8 50.8 52.0
40.2
2005
Biomass production efficiency (g g−1)
AB A AB BC
D D C
E
a a b b
d c c
d
0.58 0.52 0.57 0.58
0.57 0.52 0.43
0.43
0.62 0.51 0.60 0.60
0.39 0.40 0.52
0.43
2004
*** ns *** *** ns *** ***
0.55 0.49 0.54 0.50
0.58 0.56 0.63
0.65
0.55 0.48 0.53 0.51
0.62 0.57 0.59
0.61
2005
Harvest index (g g−1)
AB D ABC BCD
A BCD CD
ABC
a de ab bc
de e bc
cd
11.0 10.9 11.0 11.3
23.9 31.1 34.5
38.5
10.5 11.3 9.7 11.0
32.1 29.7 21.2
46.1
2004
*** ns *** *** ns *** ***
11.9 11.0 12.9 14.5
17.2 14.5 13.9
19.1
11.2 10.6 13.0 13.0
13.7 14.2 17.0
20.9
2005
Head fresh weight/dry weight (g g−1)
D D D D
C BC B
A
d d d d
b b c
a
0.59 0.53 0.62 0.63
0.54 0.58 0.53
0.37
0.64 0.57 0.66 0.63
0.41 0.46 0.63
0.48
2004
** ns *** ** ns *** ***
0.57 0.53 0.61 0.56
0.55 0.56 0.68
0.66
0.57 0.51 0.61 0.62
0.60 0.60 0.69
0.61
2005
Nitrogen harvest index (g g−1)
BCD E A ABC
DE CD AB
E
b cd ab ab
d cd a
c
*, ** and *** = significant at P<0.05, 0.01 and 0.001, respectively
ns non significant
Different letters (lower and upper case letters for low and high N supply, respectively) denote significant differences (means over the years) between cultivars within columns at P< 0.05
Year (Y) N rate (N) Cult. (Cv) N × Cv Y×N Y × Cv Y × N × Cv
5326
8899 7577 7880 8583
Lennox Bartolo Bloktor Novator
Parel
6497 7092 9854
Toughma Castello Perfecta
N2
6182
Parel
N1
2005
2004
2004
2005
N uptake (g m−2)
Cultivar
N supply
Head fresh weight (g m−2)
Table 2 Head fresh weight, total shoot N uptake, biomass production efficiency, harvest index, ratio between head fresh and dry weight and N harvest index of eight cabbage cultivars grown in Ruthe at two N supplies (N1: no fertilization and N2: 300 kg N ha−1) in 2004 + 2005
Plant Soil (2010) 328:313–325 317
318
Castello had a higher yield than cv Toughma. Within the late cultivars, cv Lennox followed by cv Bloktor had significantly higher yields than cv Bartolo. Cultivar Novator had a medium yield, mainly due to its poor performance in 2005. There was no significant year effect on yield; however, mid-early cultivars had lower yields only at low N supply in 2004 compared to 2005 which was mainly responsible for the significant Y × Cv and Y × N × Cv interactions. Among the factors contributing to N efficiency, total shoot-N uptake differed between cultivars according to their maturity times. Within the maturity groups N uptake did not differ much. Nitrogen uptake was significantly greater and varied more between cultivars under high N compared to low N supply. Late cultivars accumulated more N than earlier maturing cultivars explaining most of the highly significant effect of the cultivar and the N × Cv interaction. Total shoot-N uptake did not generally differ between years; however, similar to what was observed for yield, the early and mid-early cultivars had a lower N uptake under low N supply in 2004 than in 2005. Biomass production efficiency was higher for late than for early maturing cultivars. Low N supply clearly increased biomass production efficiency of all cultivars. Within mid-early and late maturing cultivars, respectively, cvs Perfecta and Bartolo had high biomass production efficiency both at low and at high N supply. There was no N × Cv interaction for this trait. Mean harvest index ranging from 0.39–0.65 clearly differed between cultivars, also within maturity groups. At low N supply, cv Perfecta had a significantly higher harvest index than the other midearly cultivars. Within late maturing cultivars, cv Bartolo had a significantly lower harvest index than the other cultivars. Harvest index was not influenced by N supply; however, there was a highly significant N × Cv interaction. In 2004, the harvest index was lower than in 2005, especially for the earlier maturing cultivars at low N supply. The late cultivars had a higher harvest index in 2004 than in 2005 (highly significant Y × Cv interaction). The ratio between head fresh and dry weight reflects the water content of the heads and ranged from 46.1 to 9.7. Cultivars differed significantly in head fresh weight/dry weight ratio with the early cultivars showing higher values than the late maturing ones. There was not much variation within maturity
Plant Soil (2010) 328:313–325
groups. The N rate did not generally affect the head fresh weight/dry weight ratio. However, cultivar ranking within the earlier maturing cultivars was affected by N rate, at least in 2004 (highly significant Y × N × Cv interaction). The early cultivars Parel and Toughma had a significantly higher head fresh weight/dry weight ratio under low N than under high N supply, while cv Perfecta had a higher head water accumulation under high N. Cultivar differences in N harvest index were generally related to those in harvest index but with exceptions. At low N supply, cv Perfecta had the highest N harvest index and cv Bartolo had the lowest N harvest index. At high N supply, cv Perfecta still had a high N harvest index despite a comparatively low harvest index, while cv Parel had a low N harvest index despite a high harvest index. Total shoot dry weights at heading and maturity differed between cultivars according to their growing times (Table 3). Within the late maturing cultivars cvs Lennox and Bartolo had higher shoot dry weights at maturity than cvs Bloktor and Novator. At low N supply shoot dry weight was significantly lower than at high N supply, except for the early cultivars Parel and Toughma. Shoot dry weight was lower in 2004 than 2005. This effect was greater for the early and mid-early than for the late cultivars, particularly under low N supply (significant Y × N × Cv interaction). The outer leaves are important for the C assimilation and thus growth of the plants, and leaf numbers of outer leaves may also be decisive for the biomass distribution between leaves and head and thus harvest index. Late maturing cultivars developed more leaves until heading than early maturing cultivars. Cultivar Bartolo had the highest leaf number at heading under both N rates. N rate had no major effect on leaf number. A significantly decreased leaf number under low N supply was found for cvs Parel, Toughma and Castello, but not for the other cultivars. Leaf number between heading and maturity partly increased for the earlier maturing cultivars and it decreased for the late cultivars. Overall, leaf number at maturity was higher at low N that at high N supply, suggesting higher leaf losses at high N supply. Leaf number at maturity differed between but also within different maturity groups. Within maturity groups at low N supply, it was highest for those cultivars with a low harvest index. The cultivars Toughma, Castello and Bartolo had the highest leaf numbers at maturity. At high N
Plant Soil (2010) 328:313–325
319
Table 3 Shoot dry weight and leaf number of the outer leaves at heading and maturity of eight cabbage cultivars grown in Ruthe at two N supplies (N1: no fertilization and N2: 300 kg N ha−1) in 2004 + 2005 N supply
N1
N2
Cultivar
Parel
Shoot dry weight at heading (g m−2)
Shoot dry weight at maturity (g m−2)
Leaf number at heading
Leaf number at maturity
2004
2005
2004
2005
2004
2005
2004
2005
18
71
310
460
e
12.0
13.5
e
12.7
14.2
e
d
Toughma
14
124
d
516
925
d
10.0
13.3
f
17.3
17.5
b
Castello
30
202
c
602
933
d
14.3
18.8
d
16.1
18.0
b
Perfecta
161
330
b
896
1025
c
20.4
21.3
c
14.7
15.3
d
Lennox
521
524
a
1366
1425
a
21.7
21.9
ab
15.0
14.9
d
Bartolo
501
513
a
1303
1442
a
22.0
22.9
a
19.8
20.0
a
Bloktor
445
519
a
1345
1156
b
20.4
21.5
bc
16.4
15.8
c
Novator
484
533
a
1298
1204
b
20.9
20.9
c
14.8
15.7
d
42
92
D
349
509
G
15.3
13.7
E
15.3
11.6
D
Parel Toughma
30
171
D
675
894
F
13.0
13.9
F
14.5
17.0
B
Castello
59
263
C
633
1131
E
16.7
18.9
D
12.8
17.8
BC
Perfecta
281
394
B
885
1432
D
18.5
21.5
C
15.1
15.9
BC
Lennox
751
661
A
1764
1660
A
21.4
21.5
B
14.3
15.2
C
Bartolo
626
711
A
1774
1721
A
22.4
23.5
A
16.0
19.3
A
Bloktor
666
686
A
1690
1474
B
21.5
23.0
AB
15.3
15.7
BC
Novator
728
623
A
1624
1323
C
19.3
21.0
C
12.8
14.6
D
Year (Y)
**
**
**
N rate (N)
***
***
ns
** *
Cult. (Cv)
***
***
***
***
N × Cv
***
***
***
***
Y×N
ns
ns
ns
ns
Y × Cv
***
***
***
***
Y × N × Cv
**
***
***
***
Different letters (lower and upper case letters for low and high N supply, respectively) denote significant differences (means over the years) between cultivars within columns at P<0.05 ns non significant *, ** and *** = significant at P<0.05, 0.01 and 0.001, respectively
supply, the mid-early maturing cultivars and cvs Bartolo and Bloktor had high leaf numbers. Thus, the relationship to harvest index was less clear at high N supply. The ratio between head fresh and dry weight was highly significantly related to the N concentration in the heads (Fig. 2a, c). Cultivars with a high head fresh weight/dry weight ratio also had a high head N
concentration. This relationship was true for both N rates, however, under high N supply all cultivars showed higher head N concentrations without concomitant changes in the head fresh weight/dry weight ratio. In addition, cultivar differences in head fresh weight/dry weight ratio under high N supply were accompanied by greater differences in head N concentration than this was the case under low N
50
50
a
b
40
40
30 20
20
10
N2: y = -35.3 + 2.0x 2 r = 0.91***
0 25
y = 14.0 + 11.7x 2 r = 0.45**
30
N1: y = -38.4 + 2.77x 2 r = 0.88***
10 0
0 Head fresh weight/dry weight (g g-1)
Fig. 2 Relationship between head N concentration and the ratio of head fresh weight to dry weight in 2004 (a) and 2005 (c) and between head nitrateconcentration and the ratio of head fresh weight to dry weight in 2004 (b) and 2005 (d) for eight cabbage cultivars grown in Ruthe at two N supplies (N1: no fertilization, open symbols and N2: 300 kg N ha-1, filled symbols). ** and *** = significant at P<0.01 and 0.001, respectively
Plant Soil (2010) 328:313–325
Head fresh weight/dry weight (g g-1)
320
10
20
30
40
0
1
2
3
4
25
c
d
20
20 N1: y = -5.4 + 0.98x 2 r = 0.84**
15
15
10
10 N2: y = -1.6 + 0.63x 2 r = 0.89***
5
5
0
0 0 10 20 30 40 Nitrogen concentration (mg g-1 DM) Parel Toughma
supply, resulting in a steeper regression line for high N supply. The head nitrate concentrations did not differ between the N rates (Fig. 2b, d). The relationship between head nitrate concentration and head fresh weight/dry weight ratio was significant in 2004 but less close than between head fresh weight/dry weight ratio and head N concentration. In 2005 no significant relationship was observed. The comparison of head-N accumulation and total shoot-N uptake between heading and maturity (Fig. 3) allows an estimation of the N retranslocation from the outer leaves to the heads during that period. Net N retranslocation is thus given by head-N accumulation between heading and maturity minus total shoot-N uptake between heading and maturity. Early and midearly cultivars took up more N between heading and maturity than they accumulated in the heads within the same period. For the late cultivars net N
Castello Perfecta
0 1 2 3 4 Nitrate-N concentration (mg g-1 DM) Lennox Bartolo
Bloktor Novator
retranslocation from the outer leaves occurred, especially under high N and in 2004. At low N supply, across all cultivars more N was taken up than was allocated to the heads, while at high N supply, there was a net N retranslocation from the outer leaves to the heads (P<0.01 for N rate comparison). Within the mid-early cultivars, Toughma and Castello allocated significantly lower N amounts from shoot-N uptake to the heads than Perfecta at both N rates. For Perfecta, shoot-N uptake and head-N accumulation were equal. The late cultivars Lennox and Novator retranslocated more N from the outer leaves than cv Bartolo at low N supply (P<0.01 in both cases). At high N supply cv Novator had a higher N retranslocation than cv Bartolo (P<0.05). Leaf losses demonstrate a high N retranslocation from the outer leaves since N is withdrawn from the leaves before they are shed. Averaged over cultivars, the ratio between lost leaves and total number of outer
Plant Soil (2010) 328:313–325
321
2004
2005
-2
Nitrogen uptake (g m )
25
low N
a
b
c
d
l re ma llo cta ox olo tor tor Pa ugh aste erfe enn Bart Blok ova L N C P To Cultivar
l re ma llo cta ox olo tor tor Pa ugh aste erfe enn Bart Blok ova L N C P To Cultivar
20 15 10 5 0
high N Shoot Head
20 15 10 5 0
Fig. 3 Shoot (head + outer leaves) N uptake and head-N uptake between heading and maturity (means with standard deviation) of eight cabbage cultivars grown in Ruthe without N
leaves was higher under high N than low N supply (Fig. 4), but there was a highly significant N × Cv interaction. Generally, earlier maturing cultivars had lower ratios of lost leaves than late maturing cultivars. However, cv Parel lost comparatively many leaves at high N supply. Among the mid-early cultivars cv Perfecta was close to the late maturing cultivars. Among the late maturing cultivars, cv Bartolo had a significantly lower ratio of lost leaves than all other cultivars at low N supply and than cvs Bloktor and Novator at high N supply. Soluble carbohydrate concentrations in outer leaves and head leaves may give indications about possible C limitations for head growth. Late maturing cultivars had higher leaf carbohydrate-concentrations than the early and mid-early cultivars (Table 4). Within maturity groups, cv Bartolo had higher leaf carbohydrateconcentrations than the other cultivars, but no significant cultivar difference was observed within mid-early cultivars. Low N supply increased carbohydrate concentrations of the outer leaves. Averaged over culti-
fertilization in 2004 (a) and 2005 (b) and at 300 kg ha−1 N supply in 2004 (c) and 2005 (d)
0.6 N1 N2
0.5
A
A a B
Ratio of lost leaves
-2
Nitrogen uptake (g m )
25
C
BC
bc
ab
cd BC
0.4
F-Test: N rate: * Cult: *** N x Cv: ***
de e
0.3
f
D g E
0.2 0.1 0.0 a lo ta or or llo ox r el Pa ughm aste erfec Lenn Barto Blokt Novat C P To
Cultivar
Fig. 4 Ratio between lost leaves and total number of outer leaves (means with standard deviation) of eight cabbage cultivars grown in Ruthe without N fertilization (N1) and at 300 kg ha−1 N supply (N2) in 2005. Values denoted by different letters (lower and upper case letters for low and high N supply, respectively) are significantly different between cultivars at P< 0.05. ** and *** = P<0.01 and 0.001, respectively
322
Plant Soil (2010) 328:313–325
Table 4 Water-soluble carbohydrate concentrations at maturity in outer leaves and heads of eight cabbage cultivars grown in Ruthe at two N rates (N1: no fertilization and N2: 300 kg N ha−1) in 2004 + 2005 Cultivar
N1
N2
Parel
Leaves (mg g−1)
Heads (mg g−1)
2004
2005
2004
2005
27
36
149
234
f
d
Discussion The cultivars investigated differed in N efficiency, i.e. in yield at low N supply. Based on the results of this study the cultivars Parel, Perfecta, Lennox and Novator can be classified as N-efficient, while cvs Toughma, Castello and Bartolo were N-inefficient (Table 2). Cultivar Bloktor had an intermediate N efficiency. The early maturing cv Parel cannot be compared to another cultivar of the same maturity group. However, a rough estimation of N efficiency was performed by comparing yield at low N supply with high N supply. Yield of this cultivar was not decreased at low N supply, but tended to be even higher than at high N supply. This was not due to the fact that N supply was not limiting for this cultivar at the lower N rate. Shoot growth (Table 3) and N uptake (Table 2) were decreased to a similar extent as for the other cultivars under low N compared to high N supply. The cultivar differences in N efficiency were not due to differences in N uptake at low N supply, since this trait hardly differed between cultivars of the same maturity group (Table 2). This was probably due to the generally very efficient N uptake of white cabbage (Kristensen and Thorup-Kristensen 2004) with low soil N losses during the growing period (Everaarts and Booij 2000). Since N uptake efficiency did not differ between N-efficient and -inefficient cultivars, the cultivar differences in N efficiency can be attributed to differences in N utilization efficiency. The separate listing of the three factors determining N utilization efficiency revealed that N-efficient and -inefficient cultivars differed in harvest index, but not in biomass production efficiency and head fresh weight/dry weight ratio (Table 2). Biomass production efficiency reflects the efficiency of N use in plant carbon (C) assimilation. Genotypic differences in this trait were found within late maturing cultivars (Table 2). However, this was of low importance for N efficiency. Obviously, total
Toughma
36
88
e
169
252
cd
Castello
66
106
cd
183
303
bcd
Perfecta
62
80
de
235
157
d
Lennox
121
105
b
318
224
abc
Bartolo
144
131
a
286
289
ab
Bloktor
109
76
c
363
286
a
Novator
115
76
c
238
233
bcd
Parel
22
27
D
129
113
C
Toughma
32
87
C
297
238
A
Castello
33
100
C
95
208
C
Perfecta
31
69
C
80
261
BC A
Lennox
92
87
B
244
330
Bartolo
103
124
A
190
267
AB
Bloktor
111
72
B
250
231
A
Novator
103
79
B
199
275
A
Year (Y)
*
ns
N rate (N)
**
ns
Cult. (Cv)
***
***
N × Cv
ns
**
Y×N
m
ns
Y × Cv
***
*
ns
***
Y × N × Cv
generally higher for late maturing cultivars than for the early and mid-early cultivars. Cultivar Toughma had a significantly higher head carbohydrate-concentration than the other mid-early cultivars under high N supply. N rate and year had no significant effect on head carbohydrate-concentrations.
Different letters (lower and upper case letters for low and high N supply, respectively) denote significant differences (means over the years) between cultivars within columns at P<0.05 ns non significant *, ** and *** = significant at P<0.05, 0.01 and 0.001, respectively
vars, carbohydrate concentrations in outer leaves were higher in 2005 than in 2004. However, this was restricted mainly to the mid-early cultivars. Carbohydrate concentrations in the heads showed a highly significant Y × N × Cv interaction. They were
Plant Soil (2010) 328:313–325
shoot dry matter production, although greatly decreased at low N, is not the limiting factor for head yield in cabbage under N limitation. This conclusion is corroborated by the fact that carbohydrates accumulated in the outer leaves at low N supply (Table 4), suggesting a limitation for the utilization of C for growth. The partitioning of dry matter to the heads, i.e. the harvest index, was the main determining factor for N efficiency. The high harvest index of the N-efficient cultivars was accompanied by a low number of outer leaves at maturity (Table 3). Two different mechanisms seem to be responsible for high harvest indices of early and late maturing cultivars. The N-inefficient mid-early cultivars Toughma and Castello kept on forming outer leaves after heading, especially at low N and in 2004 (Table 3). This decreased the proportion of shoot dry matter that was allocated to the head and thus head dry weight. In addition, the proportion of shoot-N uptake after heading that was allocated to the heads was low for cvs Toughma and Castello (Fig. 3), which might also have decreased head growth. The ongoing leaf formation of these two cultivars was partly due to their low leaf numbers at heading (Table 3). This was not a pure genotype effect, but appeared only during the relatively cool temperature conditions in 2004 (Fig. 1) and then especially at low N supply. While reduced leaf growth is a common response of crops to N limitation, slower leaf emergence and increased duration of leaf expansion appear to be crop-specific responses (Gastal and Lemaire 2002). Strong interactions between temperature and N rate have also been found for leaf growth of wheat (Lawlor et al. 1988). However, not much is known about the underlying mechanisms. This study shows that there are also genotypic effects additionally modifying the temperature and N-rate effects and that these interactions influence head yield of cabbage under low N supply. Beside the lower rate of leaf formation, also reduced rates of leaf-area expansion might contribute to a continuous leaf unfolding. A low leaf area of the surrounding leaves may prevent shading of the shoot apex which seems to be important for head formation (Hara et al. 1982). Apart from leaf emergence between heading and maturity also the number of leaves shed from the plant until maturity determined total leaf number at maturity. This played a role especially for late maturing cultivars (Table 3, Fig. 4). High leaf losses
323
can increase harvest index by an improved retranslocation of dry matter and nitrogen of the senescing leaves to the heads. Especially an improved nitrogen supply to the heads will promote sink strength for C assimilates and thus growth. Major cause for leaf losses was most probably the shading of lower leaves, because leaf losses were mostly higher under high than under low N supply. Cultivar differences in leaf orientation could be observed in the field, which probably caused the genotypic differences in leaf losses. Cultivars Bloktor, Novator and also cv Parel had more horizontally orientated leaves than the other cultivars. The significantly higher leaf losses of cvs Castello and Lennox under low compared to high N supply (Fig. 4) might reflect a higher susceptibility of leaf senescence to N limitation in these cultivars. Head fresh weight/dry weight ratio was higher for earlier maturing cultivars, but hardly varied within maturity groups (Table 2). Therefore, this parameter had no major influence on genotypic differences in N efficiency. However, the comparatively high yield of cv Parel under low compared to high N supply could be attributed to an increased head fresh weight/dry weight ratio. A high head water accumulation is usually found under high N supply (Peck 1981; Everaarts and Booij 2000). However, in the present study the N-rate effect clearly depended upon cultivar (Table 2). Generally, cultivar differences in head fresh weight/dry weight ratio were related to high head N concentrations (Fig. 2a, c). A high N availability can lead to a high water accumulation in a tissue because it increases the size of individual cells (McDonald 1989; Kano et al. 2007). The N effect on cell size seems to be mediated by an improved cell wall extensibility and not by an increased turgor pressure due to high amounts of osmotica like nitrate (Palmer et al. 1996). Therefore, it is not surprising that nitrateN concentrations were hardly related to head fresh weight/dry weight ratio (Fig. 2b, d). However, head N concentration alone does not explain the differences in head water accumulation between N rates, since head fresh weight/dry weight ratios were lower at high N supply than could be expected from their head N concentrations (Fig. 2a, c). This may be due to higher protein storage in head-leaf cells at high N supply. High amounts of stored proteins will increase the N concentration of the heads, but it will not affect the water accumulation of cells that are already fully expanded. Relatively high protein amounts may be
324
stored instead of being used for current growth when N import into the head is generally high and at late head growth, when growth rates are already decreasing. During this developmental stage soil-N availability will be still comparatively high at high N supply and additionally N will be retranslocated from outer senescing leaves (Fig. 3). Therefore, a high N supply to the heads during late head growth might be comparatively less effective in enhancing yield. Apart from the parameters determining head yield at maturity, it also has to be considered that maturity time was delayed for the earlier maturing cultivars at low N supply especially in 2004 (Table 1). Maturity delay under N limitation was also reported in a Canadian study using a mid-early cabbage cultivar (Westerveld et al. 2003). In that study a strong year effect and year by N rate interaction were observed. The delay in maturity might have contributed to the relatively high shoot dry weight at maturity of the early cultivars (Table 3). Maturity delay was quite similar among the earlier maturing cultivars and seemed to be governed by the decreased growth rates under low N and at the lower temperatures in 2004 and not by a delay in ontogeny. Therefore, although maturity delay might have had an impact on yield at limiting N supply, it may not play a role for genotypic differences in N efficiency. Selection of N-efficient cultivars under limiting N supply is necessary only if the decisive traits for yield differ from those determining yield under optimum N supply. The most important factor for genotypic differences in yield under low N supply was harvest index. This trait also played a role for yield under high N supply, however, the relationship was less clear (Table 2). For the mid-early cultivars head fresh weight/dry weight ratio was more important for yield differences than harvest index. This was related to cultivar differences in N harvest index (Table 2). Similarly to the conditions under low N supply, a high N harvest index was related to a high proportion of shoot-N uptake after heading that was allocated to the head (Fig. 3). This was also related to the formation of outer leaves between heading and maturity (Table 3). However, cultivar differences in leaf formation were greater under low than under high N supply, indicating differential susceptibility in this trait to N limitation. Therefore, selecting for cultivars without delayed head formation will be more beneficial for N efficiency than for high yield under optimum N
Plant Soil (2010) 328:313–325
supply. For late maturing cultivars, differences in harvest index were more similar between low and high N supply, indicating a similar mechanism responsible for this trait under both N rates. In conclusion, it appears that a major effect of the N supply response lies in an alleviation of a decline in overall growth due to low temperatures, especially for early and mid-early cultivars. Therefore, breeding of cultivars with a low susceptibility in leaf appearance to low temperature might be effective in enhancing N utilization. An additional trait for adaptation to low N conditions is an increase of direct N allocation to the head leading to increased head-water accumulation. Since a high head water content might reduce the storage capability of the heads, this is a useful approach only for cultivars that are produced for the fresh market. Although no specific adaptations of late maturing cultivars to low N conditions were found, the genotypic variation in N harvest index shows that there is potential to decrease N amounts in crop residues by cultivar selection. High leaf losses due to shading are of special importance here. High N retranslocation to the head might also improve head growth and thus N utilization efficiency. A selection for cultivars with high harvest indices and N harvest indices will be advantageous without compromising storage ability of the heads by an increased water content. Testing the cultivars under stronger Nlimiting conditions than in these experiments might reveal further genotypic strategies of N efficiency. Acknowledgements Cultivar recommendations and supply of seeds by Sjaak van der Ploeg (Syngenta Seeds) are gratefully acknowledged.
References Alt C, Kage H, Stützel H (2000) Modelling nitrogen content and distribution of cauliflower (Brassica oleracea L. botrytis). Ann Bot 86:963–973 Booij R, Kreuzer ADH, Smit AL, van der Werf A (1996) Effect of nitrogen availability on dry matter production, nitrogen uptake and light interception of Brussels sprouts and leeks. Neth J Agric Sci 44:3–19 Brandis Av, Scharpf C (1987) Steuerung der Weißkohlkultur mit Stickstoff. Gemüse 23:66–67 Cataldo DA, Haroon M, Schrader LE, Youngs VL (1975) Rapid colorimetric determination of nitrate in plant tissue by nitration of salicylic acid. Commun Soil Sci Plant Anal 6:71–80 Craswell ET, Godwin DC (1984) The efficiency of nitrogen fertilizers applied to cereals in different climates. In:
Plant Soil (2010) 328:313–325 Tinker PB, Läuchli A (eds) Advances in plant nutrition, vol 1. Praeger, New York, pp 1–55 Everaarts AP (1993) General and quantitative aspects of nitrogen fertilizer use in the cultivation of Brassica vegetables. Acta Hortic 339:149–160 Everaarts AP, De Moel CP (1995) The effect of nitrogen and the method of application on the yield of cauliflower. Neth J Agric Sci 43:409–418 Everaarts AP, De Moel CP (1998) The effect of nitrogen and the method of application on yield and quality of white cabbage. Eur J Agron 9:203–211 Everaarts AP, Booij R (2000) The effect of nitrogen application on nitrogen utilization by white cabbage (Brassica oleracea var. capitata) and on nitrogen in the soil at harvest. J Hortic Sci Biotech 75:705–712 Everaarts AP, De Moel CP, Van Noordwijk M (1996) The effect of nitrogen and the method of application on nitrogen uptake of cauliflower and on nitrogen in crop residues and soil at harvest. Neth J Agric Sci 44:43–55 FAO (2006) World reference base for soil resources 2006. 2nd edition. World Soil Resources Reports No. 103. FAO, Rome Fink M, Feller C (1998) An empirical model for describing growth and nitrogen uptake of white cabbage (Brassica oleracea var. capitata). Sci Hortic 73:75–88 Fink M, Feller C (2001) Nitrogen uptake of Brussels sprouts— validation of a model. J Hortic Sci Biotech 76:615–619 Freyman S, Toivonen PM, Perrin PW, Lin WC, Hall JW (1991) Effect of nitrogen fertilization on yield, storage losses and chemical composition of winter cabbage. Can J Plant Sci 71:943–946 Gastal F, Lemaire G (2002) N uptake and distribution in crops: an agronomical and ecophysiological perspective. J Exp Bot 53:789–799 Greenwood DJ (1990) Production or productivity, the nitrate problem? Ann Appl Biol 117:209–231 Hara T (1989) Effects of nitrogen, phosphorus and potassium in culture solution on the head yield and free sugar composition of cabbage. J Japan Soc Hort Sci 58:595–599 Hara T, Nakagawa A, Sonoda Y (1982) Effects of nitrogen supply and removal of outer leaves on the head development of cabbage plants. J Japan Soc Hort Sci 50:481–486 Kage H, Alt C, Stützel H (2003) Aspects of nitrogen use efficiency of cauliflower I. A simulation modelling based
325 analysis of nitrogen availability under field conditions. J Agric Sci 141:1–16 Kano Y, Nakagawa H, Sekine M, Goto H, Sugiura A (2007) Effect of nitrogen fertilizer on cell size and sugar accumulation in the leaves of cabbage (Brassica oleracea L.). HortScience 42:1490–1492 Kristensen HL, Thorup-Kristensen K (2004) Uptake of 15N labeled nitrate by root systems of sweet corn, carrot and white cabbage from 0.2–2.5 meters depth. Plant Soil 265:93–100 Lawlor DW, Boyle FA, Keys AJ, Kendall AC, Young AT (1988) Nitrate nutrition and temperature effects on wheat: a synthesis of plant growth and nitrogen uptake in relation to metabolic and physiological processes. J Exp Bot 39:329–343 McDonald AJS (1989) Nitrate availability and shoot area development in small willow (Salix viminalis). Plant Cell Environ 12:417–424 Nilsson T (1988) Growth and carbohydrate composition of winter white cabbage intended for long-term storage. I. Effects of late N-fertilization and time of harvest. J Hortic Sci 63:419–429 Ortiz-Monasterio JI, Sayre KD, Rajaram S, McMahon M (1997) Genetic progress in wheat yield and nitrogen use efficiency under four nitrogen rates. Crop Sci 37:898–904 Palmer SJ, Berridge DM, McDonald AJS, Davies WJ (1996) Control of leaf expansion in sunflower (Helianthus annuus L.) by nitrogen nutrition. J Exp Bot 47:359–368 Peck NH (1981) Cabbage plant responses to nitrogen fertilization. Agron J 73:679–684 Sattelmacher B, Horst WJ, Becker HC (1994) Factors that contribute to genetic variation for nutrient efficiency of crop plants. Z Pflanzenernähr Bodenk 157:215–224 Scharpf HC, Weier U (1994) Kalkulatorische Ermittlung des Nmin-Sollwertes im Gemüsebau unter besonderer Berücksichtigung der Stickstoffimmobilisierung bzw. -fixierung. Z Pflanzenernähr Bodenk 157:11–16 Schenk MK (2006) Nutrient efficiency of vegetable crops. Acta Hort 700:21–33 Westerveld SM, McDonald MR, McKeown AW, Scott-Dupree CD (2003) Optimum nitrogen fertilization of summer cabbage in Ontario. Acta Hortic 627:211–215 Yemm EW, Willis AJ (1954) The estimation of carbohydrates in plant extracts by anthrone. Biochem J 57:508–514