Published online 1 August 2005
Published in Crop Sci 45:1728-1735 (2005)
© 2005 Crop Science Society of America
677 S. Segoe Rd., Madison, WI 53711 USA
CROP BREEDING, GENETICS & CYTOLOGY
Phosphorus Response Components of Different Brassica oleracea Genotypes Are Reproducible in Different Environments
D. J. Greenwooda,*,
A. M. Stellaccib,
M. C. Meachama,
M. R. Broadleyc and
P. J. Whitea
a Warwick HRI, Wellesbourne, Warwick CV35 9EF, UK
b Dep. of "Scienze delle Produzioni Vegetali", Univ. of Bari, via Amendola 165/a, 70126 (BA), Italy
c Univ. of Nottingham, Sutton Bonington, Loughborough LE12 5RD, UK
* Corresponding author (d.greenwood{at}warwick.ac.uk)
 |
ABSTRACT
|
|---|
Better criteria for selecting cultivars for their ability to grow well on low-P soils could reduce waste of fertilizer P. The objective of this study was to quantify and determine the reproducibility of two crop parameters (A and Km) responsible for differences in the response of 12 Brassica oleracea L. genotypes to P supply. Modified Michaelis–Menten equations were fitted to shoot dry weight responses to P supply for each genotype in each of three glasshouse and three field experiments. One of the fitted parameters (A) defined the maximum shoot dry weight that could be obtained with ample P, and the other (Km) defined the efficiency of root P-uptake as the concentration of extractable soil P in the rooting medium at which growth was half the maximum. The equations fitted the data well. Ranking and regression analyses showed that intergenotypic differences in A were considerable, and reproducible, but that differences in Km were small. Thus, B. oleracea genotypes yielding most on a P-sufficient soil will also yield most on a P-deficient soil. However, both the maximum yield (A) and the efficiency of root-P uptake (Km) vary considerably between different vegetable crops. Thus, a graphical procedure to select crops for highest yields on low-P soils, on the basis of a comparison of A and Km values, is described.
Abbreviations: A, maximum possible plant dry weight with sufficient P • B, the gradient of W against Px when Px
0 • CC, coefficient of concordance • F1, F2, F3, field experiments 1, 2, and 3 • G1h1, glasshouse experiment 1 harvest 1 • G1h2, glasshouse experiment 1 harvest 2 • G2, glasshouse experiment 2 • G3, glasshouse experiment 3 • Km, the concentration of extractable soil P in the rooting medium at which growth is half the maximum • K2, a growth rate coefficient • No., number of measurements in each experiment • Pe, a measure of the extractable soil P in unfertilized potting-mix • Pf, fertilizer P • Ps, the Olsen extractable soil P • PUE, phosphorus use efficiency • Px, the concentration of extractable soil P in the rooting medium • W, total plant dry weight exclusive of fibrous roots
|
INTRODUCTION
|
|---|
CROPPING WITH CULTIVARS that grow well on low-P soils could improve the efficiency of P use and minimize pollution. Growing crops in low-P soils would reduce the conversion of P into forms that are unavailable to plants (Karpinets et al., 2004) and minimize losses by erosion and run off (Sims et al., 1998). Selection of suitable cultivars is usually performed in field experiments having only one or two P treatments, and interpretation of the results has proved difficult. Numerous criteria, often referred to as P-use efficiency (PUE), have been advanced as a basis for selecting cultivars for their ability to grow on low-P soils (Gourley et al., 1994). They include the production of plant biomass on a low-P medium (e.g., Fageria and Baligar, 1997); the total plant biomass per unit of P absorbed (e.g., Elliott and Läuchli, 1985); yield response per unit of added P (e.g., Blair and Cordero, 1978); the amount of P in the rooting medium required to give a given percentage of maximum yield (e.g., Föhse et al., 1988); and the dry weight obtained with insufficient P divided by the dry weight obtained with sufficient P (e.g., Osborne and Rengel, 2002; Dechassa et al., 2003). In addition, Fageria and Baligar (1997) suggested a graphical procedure for assessing PUE from two rather than just one of the above criteria, thus enabling more information to be taken into account.
Few studies have been made on the reproducibility of any of these measures and each criterion has weaknesses. For example, the growth of cultivars on a P-deficient soil may be proportional to their growth on a P-sufficient soil, which indicates that there are no differences between cultivars in their response to P supply. Nor does the ratio of yield on a P-deficient soil to that on a P-sufficient soil give any indication of the ability of the crop to produce a yield at low-P supply because the ratio could be high even when growth is very small. Perhaps a factor of more importance is that growth and nutrient uptake are generally related to the supply of P by diminishing curves that have genotype dependent maxima. Consequently, absolute values of most of the above criteria for PUE can vary with even minor changes in low levels of P supply (Gerloff, 1976). Indeed, it is doubtful if any of these criteria are entirely satisfactory indicators of the relative abilities of different cultivars to grow on low-P soil. Thus, there is a need to improve the criteria for selection of cultivars for growth on low-P soils.
One way to select cultivars that respond well to low-P availability is to measure the dry matter yields obtained at different levels of P and then fit an equation to the resulting data. If a suitable equation is chosen, the response of a cultivar to P availability can be defined by a few parameters, all of which have physiological meanings. The Michaelis–Menten equation, as fitted by Gourley et al. (1994), is particularly useful because it defines the relationship between yield and P supply using only two parameters: the maximum yield that can be obtained with ample P (A) and the P supply at which yield is half maximal (Km). Our hypothesis is that differences between B. oleracea genotypes in their responses to P supply can be accounted for entirely by one or both of the fitted parameters of the Michaelis–Menten equation.
The objective of this study was to quantify and determine the reproducibility of the two crop parameters (A and Km) that determine the responses of 12 B. oleracea genotypes to P supply. Brassica oleracea was chosen for this study because much work has been done on its genetics (King, 2003) and because the seed of numerous genotypes was readily available.
|
MATERIALS AND METHODS
|
|---|
Plant Genotypes
Eight commercial varieties of B. oleracea var. sabellica (cv. Reflex hybrid; borecole), var. italica (cv. Marathon hybrid; broccoli), var. botrytis (cv. Fremont hybrid; cauliflower), var. capitata (cv. Impuls hybrid; Dutch white cabbage), var. gongylodes (cv. Kolibri hybrid; kohlrabi), var. alboglabra (cv. Green Lance hybrid; Oriental kale), var. sabauda (cv. Midvoy hybrid; Savoy cabbage), var. gemmifera (cv. Maximus hybrid; Brussel sprouts), and four parental lines of genetic mapping populations (Sebastian et al., 2000) var. alboglabra (A12DHd), var. italica (GD33 and ‘B’ line B188908), and var. botrytis (‘N’ line CA25) were grown in glasshouse and field experiments at Warwick HRI, Wellesbourne, UK.
Glasshouse Experiments
Three glasshouse experiments were performed between July 2002 and March 2003. In each experiment, the response of the 12 genotypes was determined with six levels of fertilizer P incorporated in a potting mix. There were eight replicates and each replicate was in a separate 1-L pot. Fertilizer-P levels were arranged as a pair of extended Trojan squares (Edmondson, 1998), so that each fertilizer-P treatment occurred once in each row and at least once in each column. Within each fertilizer-P block the 12 genotypes were randomly allocated to pots. The potting mix consisted of 250 g L–1 sand and 750 g L–1 compost (Shamrock medium grade sphagnum peat supplied by Scotts UK, Bromford, Suffolk), having an Olsen extractable P (Olsen et al., 1954) of about 8 mg L–1. The P treatments were imposed by first sieving single superphosphate (7% P) through a 500-µm mesh sieve and then mixing weighed amounts, corresponding to each of the treatments, with the potting mix. Other nutrients were incorporated in the potting mix in sufficient amounts to prevent deficiencies. Three seeds were sown in each pot and all the seedlings except one were removed shortly after emergence in the second (G2) and third (G3) experiments. For the first experiment (G1), two seedlings were left and one of them was harvested 37 d after emergence (G1h1) and the other after 56 d (G1h2). The plants were harvested after 42 and 48 d in experiments G2 and G3, respectively. Six levels of P: 0, 6, 12, 18, 36, and 108 mg L–1, were created by incorporating 0, 0.075, 0.15, 0.225, 0.45, 1.35 g of single superphosphate (7% P) per liter of compost. The glasshouse temperature was maintained between approximately 15 and 20°C and daylight was supplemented by artificial lighting to maintain 16 h d–1. Pots were irrigated with deionized water as required. Dry weights were determined after drying at 80°C.
Field Experiments
Three field experiments were performed on a sandy loam Inceptisol in Wharf Ground of Warwick HRI, Wellesbourne, UK. This is a Wick series soil in the English classification (Whitfield, 1974). The field had not received P and K fertilizers for over 20 yr but was regularly cropped during that period. It had an Olsen P (Olsen et al., 1954) of approximately 20 mg L–1. The treatments were 12 genotypes and seven levels of fertilizer P. The experiment was laid out as five adjacent strips of plots orientated in a north south direction. The most easterly was Strip 1 and the most westerly, Strip 5. In each strip, there were eight plots, each 6 m long and 1 m wide laid out in order of increasing soil P. The direction of increase, however, alternated from one strip to the next. Levels of soil P were created by incorporating in the soil the following amounts of P as triple superphosphate (21% P): 0, 0, 298, 662, 1125, 1786, 2713, and 3500 kg ha–1. In each plot, the 12 genotypes were randomly distributed. For the first experiment (F1) the Strips 1, 2, 3, and 4 were used as replicates, while for the second (F2) and third experiments (F3), Strips 3, 4, and 5 were used. Fertilizer was incorporated to a depth of about 10 cm in soil a few days before transplanting of the first experiment. There was an annual overall dressing of 285 kg N ha–1 and 207 kg K ha–1. Soil samples were collected at depths of 0 to 15 cm and 0 to 30 cm in the first and second year, respectively. Genotypes were transplanted on 30 May 2002, 15 May 2003 and 14 July 2003, spaced in beds at 0.20 m x 0.20 m, and five plants of each variety per plot were harvested. Irrigation was as required. For experiment F1, the plants were harvested on two successive occasions; half the plants (from alternate positions in the row) were harvested after 39 d, and the remaining ones were harvested at different times as each of the genotypes reached the commercial maturity. Only measurements made after 39 d were used in the subsequent data analysis. In experiments F2 and F3, there was only one harvest, 43 and 44 d after the transplanting, respectively. Fresh weights were measured, and after drying at 80°C, dry weights determined.
Data Analysis
Michaelis–Menten equations have often been found to give good fits to the dependence of plant dry weight (W) on the extractable soil P (Px) (e.g., Goodall et al., 1955; Gourley et al., 1994). It has also been found (D.J. Greenwood, unpublished data) to give good fits to the simulated dependence of W on Px by a theoretical model (Greenwood et al., 2001). The Michaelis–Menten equation may be expressed as
 | [1] |
where A is the asymptotic maximum value of W, and B is the gradient of W against Px when Px 0. Rearrangement of Eq. [1] gives
 | [2] |
which indicates that the fractional increase in W with increasing Px is defined by A/B. A/B is the value of Px at which W/A = 0.5, which, in biochemical terminology, is the Km value. The relation between W and Px is thus defined entirely by A and A/B.
Estimates of A and B were obtained for each genotype by fitting a rearrangement of Eq. [1], by GenStat (2002), to the data from each experiment. The rearranged equation was
 | [3] |
The values of W in field experiment F1 were related to Px by overturning curves and were fitted by the following modification of Eq. [3]
where 
is > Px and is inversely related to the depressive effect of Px on growth.
Px, for the glasshouse data, was taken as the amount of fertilizer P (Pf), mg P L–1, plus 3 mg P L–1 (Pe) to account for the P supplied from the potting mix; water or Olsen extractable P could not be used because they were insensitive to small applications of P that, nevertheless, increased yield substantially. For the field experimental data, Px was the Olsen extractable soil P (Ps).
To compare the growth rates obtained in the experiments with the ones normally found for arable crops grown in UK, the growth rate coefficient K2, was derived from the following formula
 | [4] |
where t is time, W0.5 is the value of W when growth rate is half the maximum. Integration of Eq. [4] gives
 | [5] |
where 
T is the interval of time in days from the initial measurement of W (W0) to the final measurement of Wh. For field crops W0.5 is approximately equal to 1 Mg ha–1. K2 was calculated by solving Eq. [5] with W0.5 = 1. The relationship has given good fits to past experimental data (Greenwood et al., 1977). The growth rate coefficient K2, unlike A, is independent of the duration of growth for cultivars, such as Brassica spp., that are harvested long before senescence. Therefore it has a greater chance than A of being reproducible and comparable with the corresponding values obtained from other field experiments.
Values of A, B, A/B, and K2 were calculated for each genotype in each of the glasshouse experiments. K2 could not be calculated for the third field experiment, as the initial weights of transplanted seedlings were not measured. Also two genotypes, B188908, and Oriental kale failed to survive until harvest in one or other of the field experiments.
|
RESULTS
|
|---|
Fitting the Michaelis–Menten Parameters
Percentage responses of plant dry weight to fertilizer P were much greater in the glasshouse than in the field experiments (Fig. 1). For each experiment the residual mean square after fitting the Michaelis–Menten equations was similar to that from the analysis of variance (Table 1), demonstrating that this equation gave good fits to the data. The range of fitted values of A and B and thus A/B (Km) over the entire range of genotypes was considerable (Table 2) and significant differences (P < 0.05) between values of A, but not of B, were frequently found for different pairs of genotypes in the different experiments.
View larger version (21K):
[in this window]
[in a new window]
|
Fig. 1. Shoot dry weight response to P, as a fraction of the maximum possible shoot dry weight for the most and least responsive B. oleracea genotypes in two experiments (F2 and G3). Values were obtained by fitting Eq. [3] to the data. Ps is the Olsen extractable soil P, Pf is the fertilizer-P application, and Pe is a measure of the plant available P in the potting mix from which P fertilizer was withheld.
|
|
View this table:
[in this window]
[in a new window]
|
Table 2. Range of parameter values for B. oleracea genotypes obtained by fitting the indicated models to the entire data sets.
|
|
Ranking of Fitted Parameter Values
To discover how the relative values of A, B, A/B (Km) and K2 for each genotype varied between experiments, genotypes were ranked in order of increasing parameter size, as illustrated in Table 3. The degree of agreement between the rankings of each of the possible pairs of rankings within and between field and glasshouse experiments was determined by calculating the coefficient of concordance (CC). Important features are given in Table 4. Within the glasshouse experiments and within the field experiments, with one exception, all the rankings of pairs of A or K2 were significant at at least P < 0.05. Thus, genotypic rankings for A and K2 were reproducible in a particular environment. About half the rankings of pairs of A and K2 between field and glasshouse experiments were significant at P < 0.05. However, only one or two of the comparable rankings for B or A/B were significant at P < 0.05 (not shown). Coefficients of concordance (CCs) were also calculated for the combined data from all the glasshouse experiments, from all the field experiments and from all the experiments (Table 5). The CCs of A and K2 for each of the three comparisons were significant at P < 0.001. Those for B were significant at P < 0.05 for both the glasshouse and for the combined glasshouse and field experimental data. The CCs of A/B were not significant for any group of experiments. Thus, intergenotypic variation in A and K2, but not of A/B, was reproducible in different environments.
View this table:
[in this window]
[in a new window]
|
Table 3. Rankings of fitted parameters, in order of increasing size, for the 12 B. oleracea genotypes grown in glasshouse Exp. G2 and G3.
|
|
View this table:
[in this window]
[in a new window]
|
Table 5. Values of the Coefficient of Concordance (CC) for the ranking of each of the fitted genotype parameters for different groups of experiments.
|
|
Regression of Fitted Parameters
The mean values of A for different genotypes averaged over all glasshouse experiments were almost proportional to the mean values of A for different genotypes averaged over all field experiments (Fig. 2a). The mean values of K2 for different genotypes averaged over all glasshouse experiments were linearly related to the mean values of K2 for different genotypes averaged over all field Exp. 1 and 2, but intergenotypic variation in K2 was smaller in the glasshouse than in the field experiments (Fig. 2b). There were substantial differences between genotypes in both A and K2; both A and K2 were generally lower for the parent lines of mapping populations than for commercial varieties (Fig. 2a and 2b). Cultural practices were likely to be optimal in the field experiments since the commercial varieties had K2 values of about 0.2 Mg ha–1 (Fig. 2b), which is similar to that found for C3 crops grown at the same time of year in the UK (Greenwood et al., 1977).
View larger version (22K):
[in this window]
[in a new window]
|
Fig. 2. (a) Relationship between values of A for different B. oleracea genotypes averaged over all glasshouse experiments and those averaged over all field experiments. (b) Relationship between K2 for different B. oleracea genotypes averaged over all glasshouse experiments and K2 averaged over field experiments F1 and F2 (K2 could not be calculated for F3 because of the absence of transplant weights). Numbers refer to the genotypes A12DHd (1), B188908 (2), borecole (3), brocolli (4), CA25 (5), Dutch white cabbage (6) cauliflower (7), GD33 (8), kohlrabi (9), Oriental kale (10), Savoy (11), and Brussel sprouts (12). Data for B188908 and Oriental kale were not available for the field experiments as these genotypes failed in one or another field experiment.
|
|
To elucidate the reasons for the failure to detect reproducible differences in the fitted values of A/B (Km) between genotypes, values of A were regressed against those of B (Fig. 3 and 4). The most important relationships are given below. The combined data of G2 and G3 with SE < 0.035 Mg ha–1/mg L–1 for B indicate a good proportional relationship between A and B in these glasshouse experiments (Fig. 3). A near proportional relationship was observed when plotting the mean values of A and B for different genotypes from the three field experiments (Fig. 4a). A similar proportional relationship was observed when plotting the mean values of A and B for different genotypes from the glasshouse experiments G2 and G3 (Fig. 4b). These observations support the view that intergenotypic differences in A/B were small and that this may be the reason for failure to detect reproducible differences in their ranking.
View larger version (19K):
[in this window]
[in a new window]
|
Fig. 3. Relationship between the fitted values of B and A for B. oleracea genotypes in glasshouse experiments G2 and G3. Only data where fitted values of B had a SE < 0.0353 (Mg ha–1/mg–1 L) are plotted. The relationship covered all genotypes except A12DHd and Oriental kale as the SE of B was high for these genotypes.
|
|
View larger version (19K):
[in this window]
[in a new window]
|
Fig. 4. Relationships between the fitted values of B for different B. oleracea genotypes and their corresponding values of A, averaged over three field experiments (a) or over glasshouse experiments G2 and G3 (b). Genotypes B188908 and Oriental kale are absent from (a), as these crops failed in one or other experiment. Oriental kale is the outlier in (b). Numbers refer to the genotypes A12DHd (1), B188908 (2), borecole (3), brocolli (4), CA25 (5), Dutch white cabbage (6) cauliflower (7), GD33 (8), kohlrabi (9), Oriental kale (10), Savoy (11), and Brussel sprouts (12).
|
|
|
DISCUSSION
|
|---|
Ranking (Tables 4 and 5) and regression analysis (Fig. 3 and 4) both support the view that A/B (Km) did not vary much among B. oleracea genotypes. Thus, the dependence of yield, as a fraction of the maximum, on Px was similar for all B. oleracea genotypes. Nevertheless, the maximum plant dry weight (A) and the growth rate coefficient (K2) varied considerably (Fig. 2), and in a reproducible way (Tables 4 and 5), between genotypes. Taken together, these observations imply that genotypes that grow fastest on a P-sufficient soil will also grow fastest on a P-deficient soil. These observations are consistent with studies on Solanum genotypes. Sattelmacher et al. (1990) grew 36 Solanum genotypes with a low and a higher level of NPK fertilization, and reported that the yields at each level were proportional to one another.
For some crops, however, there may be intergenotypic differences in both A and A/B. If this is the case, the genotype that would be expected to yield best on a low-P soil would be that which had a low value of A/B (a high value of B/A) and the highest value of A. It has not been possible to find suitable genotypes of B. oleracea with which to illustrate this. However, the response of different crops to Px appears to differ, implying differences in both A and A/B. For example, the maximum percentage yield responses of lettuce to applications of P fertilizer are greater than those of any Brassica spp., with reported values between 74 and 214% for lettuce (Lactuca spp.) compared with 21% for Brassica spp. (Greenwood et al., 1980; Alt, 1987). Thus, the possible relationships between A and A/B are considered for the marketable yields of different vegetable crops (Fig. 5). This diagram can be divided into four zones, 1, 2, 3, and 4. In Zones 1 and 4, B/A is high and thus Km is low and crops in these zones grow well on a low-P soil. Crops in Zones 1 and 2 give high yields when there is ample P. Any crop in Zone 1 gives a high yield on a low-P soil. As summer cabbage is the only crop in Zone 1, it follows that it would be expected to yield better than other crops on a low-P soil.
View larger version (19K):
[in this window]
[in a new window]
|
Fig. 5. Values of B and A for different vegetable crops obtained by fitting Eq. [3] with Px = 40 + Pf (kg P ha–1) to measurements for their harvestable parts (derived from Greenwood et al., 1980). Zone 1 includes only summer cabbage. Zone 2 includes carrot, leek, and autumn sown onion; Zone 3 includes broad beans, French beans, lettuce, spinach, and spring sown onion; Zone 4 includes Brussel sprouts, calabrese, parsnip, pea, potato, radish, summer cauliflower, sugar beet (sugar), turnip (roots), and winter cabbage.
|
|
|
CONCLUSIONS
|
|---|
The Michaelis–Menten equation gave good fits to the relationship between shoot dry weight and P supply. This equation provides a convenient means of separating the potential maximum growth (A) from the efficiency of root P uptake defined as the concentration of extractable soil P in the rooting medium at which growth was half the maximum (A/B = Km). The values of A, and of the growth rate coefficient K2, varied considerably between B. oleracea genotypes, and the relative values of A or K2 for different genotypes were generally similar both within and between field and glasshouse experiments. Intergenotypic variation in A/B (Km) was small in both glasshouse and field experiments. Thus, all B. oleracea genotypes show a similar relationship between growth, as a proportion of A, and P supply. Selecting the B. oleracea genotype that grows best under optimum conditions, therefore, would also select the one that grows best under low-P supply. For cultivars of other crops, in which both A and A/B may vary, a critical appraisal of the relationship between A and B could provide a convenient means of selecting cultivars for growth on low-P soils.
|
ACKNOWLEDGMENTS
|
|---|
We thank the UK Department of the Environment and Rural Affairs for financing much of this project. We are also grateful to John Fenlon and Andrew Mead for statistical advice and to John Hammond for help with the diagrams and the literature.
Received for publication August 9, 2004.
|
REFERENCES
|
|---|
- Alt, D. 1987. Influence of P and K fertilization on the yield of different vegetable species. J. Plant Nutr. 10:1429–1435.
- Blair, G.J., and S. Cordero. 1978. The phosphorus efficiency of three annual legumes. Plant Soil 50:387–398.[CrossRef]
- Dechassa, N., M.K. Schenk, N. Claassen, and B. Steingrobe. 2003. Phosphorus efficiency of cabbage (Brassica oleracea L. var. capitata), carrot (Daucus carota L.), and potato (Solanum tuberosum L.). Plant Soil 250:215–224.[CrossRef]
- Edmondson, R.N. 1998. Trojan and incomplete Trojan square designs for crop research. J. Agric. Sci. (Cambridge) 131:135–142.[CrossRef]
- Elliott, G.C., and A. Läuchli. 1985. Phosphorus efficiency and phosphate-iron interaction in maize. Agron. J. 77:399–403.[Abstract/Free Full Text]
- Fageria, N.K., and V.C. Baligar. 1997. Phosphorus-use efficiency by corn genotypes. J. Plant Nutr. 20:1267–1277.
- Föhse, D., N. Claassen, and A. Jungk. 1988. Phosphorus efficiency of plants.1. External and internal requirements and P uptake efficiency of different plant species. Plant Soil 110:101–109.[CrossRef]
- GenStat. 2002. GenStat for windows. 6th ed. Lawes Agricultural Trust, VSN International, Oxford, UK.
- Gerloff, G.C. 1976. Plant deficiencies in the use of nitrogen, phosphorus and potassium. p. 161–173. In M.J. Wright (ed.) Plant adaptation to mineral stress in problem soils. Proc. Workshop on Plant Stress. 1976. USDA, Beltsville, MD.
- Goodall, D.W., A.E. Grant Lipp, and W.G. Slater. 1955. Nutrient interactions and deficiency diagnosis in the lettuce. 1. Nutritional interaction and growth. Aust. J. Biol. Sci. 8:301–329.
- Gourley, C.J.P., D.L. Allan, and M.P. Russelle. 1994. Plant nutrient efficiency: A comparison of definitions and suggested improvement. Plant Soil 158:29–37.[CrossRef]
- Greenwood, D.J., T.J. Cleaver, S.H.M. Loquens, and K.B. Niendorf. 1977. Relationship between plant weight and growing period for vegetables in the United Kingdom. Ann. Bot. (London) 41:987–997.[Abstract/Free Full Text]
- Greenwood, D.J., T.J. Cleaver, M.K. Turner, J. Hunt, K.B. Niendorf, and S.H.M. Loquens. 1980. Comparison of the effects of phosphate fertilizer on the yield, phosphate content and quality of 22 different vegetable and agricultural crops. J. Agric. Sci. (Cambridge) 95:457–469.
- Greenwood, D.J., T.V. Karpinets, and D.A. Stone. 2001. Dynamic model for the effects of soil P and fertilizer P on crop growth, P uptake and soil P in arable cropping: Model description. Ann. Bot. (London) 88:279–291.[Abstract/Free Full Text]
- Karpinets, T.V., D.J. Greenwood, and J.T. Ammons. 2004. Predictive mechanistic model of soil phosphorus dynamics with readily available inputs. Soil Sci. Soc. Am. J. 68:644–653.[Abstract/Free Full Text]
- King, G.J. 2003. Using molecular allelic variation to understand domestication processes and conserve diversity in Brassica crops. Acta Hortic. 598:181–186.
- Olsen, S.R., C.V. Cole, F.S. Watanabe, and L.A. Dean. 1954. Estimation of available phosphorus in soil by extraction with sodium bicarbonate (Circular 39) USDA, Washington, DC.
- Osborne, L.D., and Z. Rengel. 2002. Screening cereals for genotypic variation in efficiency of phosphorus uptake and utilisation. Aust. J. Agric. Res. 53:295–303.[CrossRef]
- Sattelmacher, B., R. Kuene, P. Malagamba, and U. Moreno. 1990. Evaluation of tuber bearing Solanum species belonging to different ploidy levels for its yielding potential at low soil fertility. Plant Soil 129:227–233.[CrossRef]
- Sebastian, R.L., E.C. Howell, G.J. King, D.F. Marshall, and M.J. Kearsey. 2000. An integrated AFLP and RFLP Brassica oleracea linkage map from two morphologically distinct doubled haploid mapping populations. Theor. Appl. Genet. 100:75–81.[CrossRef][ISI]
- Sims, J.T., R.R. Simard, and B.C. Joern. 1998. Phosphorus losses in agricultural drainage; historical perspective and current research. J. Environ. Qual. 27:277–293.[Abstract/Free Full Text]
- Whitfield, W.A.D. 1974. The soils of the National Vegetable Research Station, Wellesbourne. In Report of the National Vegetable Research Station for 1973:21–30.
Related articles in Crop Science:
- THIS ISSUE IN CROP SCIENCE
Crop Science 2005 45: x.
[Full Text]
This article has been cited by other articles:
|
|
|
|
 
M. R. Broadley, J. P. Hammond, G. J. King, D. Astley, H. C. Bowen, M. C. Meacham, A. Mead, D. A.C. Pink, G. R. Teakle, R. M. Hayden, et al.
Shoot Calcium and Magnesium Concentrations Differ between Subtaxa, Are Highly Heritable, and Associate with Potentially Pleiotropic Loci in Brassica oleracea
Plant Physiology,
April 1, 2008;
146(4):
1707 - 1720.
[Abstract]
[Full Text]
[PDF]
|
|
|
TOP
ABSTRACT
INTRODUCTION
