SEARCH

SEARCH BY CITATION

Keywords:

  • Koeleria macrantha;
  • performance;
  • distribution;
  • calcium;
  • magnesium;
  • Carboniferous;
  • Magnesian and dolomitic limestone

Summary

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Conclusion
  8. Acknowledgements
  9. References
  • • 
    The reported sensitivity of Koeleria macrantha (Poaceae) to soil magnesium, resulting in its absence from South Wales dolomitic limestone, was examined here in relation to varying ionic Ca to Mg ratios, and by cultivation in different limestone soils.
  • • 
    In a growth cabinet experiment, shoot and root Ca and Mg concentration and dry weight yield were determined for five edaphically varied populations of K. macrantha grown from tillers, over a range of Ca : Mg concentrations. The influence of relative concentrations of other nutrients was also investigated. In addition, K. macrantha plants were cultivated on Carboniferous, Magnesian and dolomitic limestone soils.
  • • 
    Total plant Ca: Mg ranged from 0.3 to > 20 mille-equivalents. Optimal substrate ratios (from 25 : 1 to 0.1 : 1), and response to different concentrations of nutrients varied between the populations. Cultivation on dolomitic limestone soil produced the highest yields.
  • • 
    The adverse effect on yields of all populations with low substrate Ca: Mg was much lower than predicted. Results suggest K. macrantha has a higher substrate Mg tolerance than other members of the Poaceae: its reported absence from the South Wales dolomitic limestone is unlikely to be due to soil magnesium sensitivity.

Introduction

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Conclusion
  8. Acknowledgements
  9. References

The grass Koeleria macrantha is widespread on Carbonifeous limestone in Britain. It also occurs on Magnesian limestone in Derbyshire (Grime & Lloyd, 1973) and in Durham (Rodwell, 1992); it has been recorded from serpentine soils in Scotland (Proctor & Woodell, 1971), Cornwall (Malloch, 1971) and from Greece (Karataglis et al., 1982). However, Cooper & Etherington (1974) concluded that its absence from dolotomized limestone in South Wales might be attributed to a high soil magnesium level.

The hypothesis of sensitivity to magnesium can be tested at both the species and the population level. Much of the previous work on the effects of soil magnesium has involved crop plants – including cereals of the family Poaceae to which K. macrantha belongs. Plant performance has been found to be unaffected where the soil calcium to magnesium ratio (henceforth written as Ca : Mg) is greater than unity (Sanik et al., 1952) and holds for most natural soils (Kelley & Brown, 1927; Martin, 1929). Most of the dolomitic soils in Britain have a Ca : Mg greater than unity (Trefor Jones, 1951), whilst serpentine soils have a ratio between 0.02 : 1 and 0.43 : 1 mille-equivalents (Proctor, 1971). All substrate and plant Ca : Mg are given as mille-eqivalents (me) in this paper. In dolomitic limestone Mg can constitute between 3 and 15 me (Frisby, 1961), whilst estimates for serpentine vary between 4.1 and 19.7 (Walker, 1954) and 14.9 and 34.8 me (Proctor, 1971).

The work reported here tests experimenally the hypothesis that substrate Ca : Mg is sufficiently critical for K. macrantha to prevent its colonization of dolomitic soil in South Wales as suggested by Cooper & Etherington (1974). Evidence and explanation are sought for intraspecific variation in its yield response and ionic uptake to a range of Ca : Mg, at both high and low elemental levels, by means of sand culture experiments, and by cultivation in natural soils with different characteristics.

Materials and Methods

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Conclusion
  8. Acknowledgements
  9. References

Plants used in these experiments were cloned at random from field collections which had been cultivated in a glasshouse on John Innes no. 1 compost under standard conditions for 1 yr. Plants of five provenances (Table 1), encompassing a wide range of edaphic conditions in the UK, were selected for study. Details of the sites and soil analyses are provided in Table 1. Soil was collected at each site from the top 10 cm, after removal of surface vegetation and litter. Measurements of pH were made on fresh soil. Chemical analyses (cf. Allen, 1989) were carried out on air-dried soils: exchangeable calcium, magnesium and potassium concentrations were determined by atomic absorption spectrophotometry; PO4-phosphorus concentrations were obtained using Olsen's extractant for calcareous soils and molybdenum blue colourimetry; total nitrogen was estimated by the Kjeldhal method.

Table 1.  Site details and soil analyses for five provenances of Koeleria macrantha
LocationNGRAltitudeSlopeAspect
1. Kynance Cove, Cornwall Koeleria abundant in very short turf over serpentineSW686135 64 m10°SW
2. Rhum (Fonchra), Highland Base of steep, vegetated cliff over basalt. Koeleria frequentNG337003350 m28°SE
3. Holme-next-the-Sea, Norfolk Partially fixed dunes, long loose vegetation to 60 cm Koeleria occasional and very scattered, mostly at edge of board-walkTF708448  2 mflat
4. Coxhoe, Durham Dolomite quarry spoil heap. Koeleria in scattered small patchesNZ326364132 mflat
5. Bishopdale, North Yorkshire Grassy bank over Carboniferous limestone, ungrazed vegetation to 45 cm high. Koeleria frequentSD943803419 m20°S
Soil analyses
ProvenancepHTotal N (%)ExchangeableRatio
(mg 100 g−1)(me 100 g−1)Ca : Mg
PKCaMg
K6.00.880.311.18 5.3 16.40.3
R6.20.840.880.64 9.9 5.41.8
H5.50.990.480.12 3.4 1.13.1
C6.30.050.800.4612.7 7.0 1.8
B6.30.690.760.6632.4 4.8 6.8

Experiment 1

Plants were cultivated individually in 7 cm diameter plastic pots filled with acid-washed, medium grain silica sand, and their performance in relation to treatment with different culture solutions was judged after 12 wk in a growth cabinet. The cabinet was maintained at 20°C during a 14-h day and at 14°C at night. Lighting was provided by warm white fluorescent tubes and tungsten bulbs at a mean radiant flux of 1350 µmol m−2 s−1 (range 1200–1500). Trays of water in the bottom of the cabinet maintained the humidity at 60%.

Each plant at the start of the experiment was a single tiller which had been washed, stripped down to three leaves, the roots clipped to 1 mm and rooted in tap water. Before planting the leaves of the rooted tiller were clipped to 2 cm in length. The pots were watered once with a standard nutrient solution and subsequently watered twice weekly, with one of the solutions described in Table 2, and flushed fortnightly with deionized water.

Table 2.  Solutions used to produce varying calcium to magnesium ratios
TreatmentConcentration (me)Ca : Mg ratio
  1. Each treatment in experiment 1 was prepared with the basic nutrient solution, and also with one third, two thirds and X2 concentrations of the basic nutrient solution: experiment 2 used only the basic nutrient solution.

Experiment 1
 High Ca + Low Mg    25 + 125 : 1
 Low Ca + Low Mg    1 + 11 : 1
 High Ca + High Mg    1 + 11 : 1
 Low Ca + High Mg    1 + 100.1 : 1
Supplementary experiment 2
 Low Ca + High Mg0.9 + 160.05 : 1

The experiment comprised four substrate Ca : Mg treatments (using CaCl2 6H2O and MgSO4 7H2O), to give both high and low elemental levels of Ca and Mg within the range encountered in natural soils. These were tested against four concentrations of nutrient solution (Hewitt solution – see Bradshaw et al., 1958) with material of five provenances replicated six times. Each replicate was derived from a different clone of the original field collection. The pots were moved side to side and back to front, one shift in each direction at weekly intervals, to avoid position effects.

After 12 wk the plants were washed and dried, the remnants of the original leaves were removed and the plants weighed. The roots and shoots were separately subjected to a mixed acid digestion (10 ml concentrated HNO3 + 1 ml concentrated H2SO4) and the calcium and magnesium concentrations were determined by atomic absorption spectrophotometry.

Experiment 2

A further concentration of Ca and Mg was used to extend the range of conditions created in the above experiment. This provided a ratio of 0.05 : 1 (see Table 2). This experiment was carried out using only the basic nutrient solution owing to time and space constraints, but under the same conditions as experiment 1.

Experiment 3

Tillers of four of the five provenances used for experiment 1 were available and were prepared as for experiment 1. These were cultivated in calcareous soils from four sites – two being dolomitic, and also in John Innes no. 1 compost. Details of sites and soil analyses are given in Table 3. Plant material was collected from the Coxhoe site only. The soil was dried, sieved to remove stones and then used to fill 10 cm plastic pots, eight replicates of each soil type. The pots were kept in a growth cabinet, conditions as for experiment 1 and watered with tap water, as necessary. After 14 wk the plants were harvested and washed. Root and shoot lengths were recorded and the plants were then dried and weighed.

Table 3.  Site details for soil collected from various locations, and used to grow Koeleria macrantha
(a) Soil no.NGRAltitudeLocationSite details
1SD 833551 285Langcliffe, N. Yorks.Molehills over Carboniferous limestone
2ST 184855 230Cefn-on-Farm, GwentJust below Cefn-on-Farm at the edge of a wood, within the dolomite band
3NZ326364 132Coxhoe, DurhamDolomite quarry spoil heap
4NZ 458343 122Near Elwick, DurhamHedgerow over Magnesian limestone
(b) Analysis of soils used to grow Koeleria macrantha
Soil no.pHTotal N (%)ExchangeableRatio
(mg 100 g−1)(me 100 g−1)Ca : Mg
PKCaMg
16.80.3560.800.4113.74.13.3 : 1
26.90.3150.560.5616.07.92.0 : 1
36.30.050.800.4612.77.01.8 : 1
46.70.1120.920.3814.95.32.8 : 1
5 (J.I.)6.60.210.54 9.95.91.7 : 1

Results

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Conclusion
  8. Acknowledgements
  9. References

Experiments 1 and 2

Total plant and root yields are presented in Fig. 1. Root yield followed a similar pattern to total yield and separate consideration of the former is therefore unprofitable.

image

Figure 1. Mean d. wt (± SE) obtained with different calcium and magnesium treatments and nutrient solution concentrations. Provenances: K, Kynance Cove; R, Rhum; H, Holme-next-the-Sea; C, Coxhoe; B, Bishopdale. (a) One third basic nutrient solution; (b) two thirds basic nutrient solution; (c) full strength nutrient solution; (d) double strength nutrient solution concentration. LSD (P = 0.05) for total d. wt for each provenance for all combinations of treatments and nutrient solutions is: K 0.069 g; R 0.202 g; H 0.86 g; C 0.095 g; B 0.067 g. Closed columns, root; open columns, shoot.

Download figure to PowerPoint

The pattern and magnitude of yield response to substrate Ca : Mg for plants of any one provenance differed according to the concentration of the nutrient solution, but not systematically: in general the lowest yields were found with solutions with high Ca, whether Mg was low or high (Table 4). With the former, shoot Ca : Mg ranged from 3.8 to 11.3 and with the latter from 0.7 to 2.9; the related root ratios were 3.5–27.5 and 0.7–4.4, respectively (see Table 5).

Table 4.  Performance of Koeleria macrantha for expts 1 and 2 (17 treatments in total) ranked in order according to total dry weight yield (highest to lowest)
Provenance
Kynance CoveRhumHolme-next-the-SeaCoxhoeBishopdale
NCTNCTNCTNCTNCT
  1. NC, nutrient solution concentration: Basic, 0.3 (one third), 0.6 (two thirds) and X2 basic concentration. T, Treatments; H–L, High Ca-Low Mg; H-H, High Ca-High Mg; L–L, Low Ca-Low Mg; L–H, Low Ca-High Mg. *indicates all plants died during the course of the experiment.

0.3L–H0.3L–HBH–L0.3L–LX2H–L
0.3L–L0.3H–LBL–L0.3L–H0.6L–L
BL–LBH–H0.3L–L0.6L–LX2L–L
BL–HBL–HBH–H0.6L–H0.6H–H
0.6L–H0.6H–L0.3L–HBL–HX2H–H
BL–H0.3L–LBL–HBL–HX2L–H
0.3H–H0.3H–H0.6L–LX2L–LBL–L
0.6L–LBL–L0.6H–LBL–L0.6L–H
BH–HBH–L0.3H–HBH–HBL–H
0.3H–L0.6L–L0.3H–LX2L–H0.3L–L
X2L–LBL–HX2L–L0.3H–H0.3L–H
X2L–H0.6H–H0.6H–H0.3H–LBL–H
BH–LX2L–HBL–HBH–LBH–H
0.6H–LX2L–L0.6L–HBH–HBH–L
BH–HX2H–LX2H–HX2H–H0.3H–L
X2H–L*0.6L–HX2L–H0.6H–L0.3H–H
X2H–HX2H–H*X2H–LX2H–L0.6H–L
Table 5.  Mean shoot and root Ca: Mg ratios (me) in Koeleria macrantha
TNCProvenance
Kynance CoveRhumHolme-next-the-SeaCoxhoeBishopdaleLSD All NC
SRSRSRSRSRSR
  1. NC, nutrient solution concentration: Basic, 0.3 (one third), 0.6 (two thirds) and X2 basic concentration. T, Treatments; H–L, High Ca-Low Mg; H–H, High Ca-High Mg; L–L, Low Ca-Low Mg; L–H, Low Ca-High Mg. S, shoot and R, root. n = 6; P = 0.05.

H–L0.3 6.7 4.6 9.3 3.57.2 6.9 5.3 6.9 7.6 6.7  
 0.610.210.110.211.38.7 9.010.327.5 5.815.6  
 B 8.314.5 9.7 9.95.6 8.5 6.813.1 7.810.2  
 X2 7.711.24.511.6 3.824.8 4.519.1  
MeanAll 8.4 9.7 9.2 6.56.5 9.0 6.618.1 6.412.91.93.8
 NC            
H–H0.3 2.2 1.1 1.9 0.71.1 1.2 2.2 2.2 1.2 4.4  
 0.6 1.6 2.1 2.9 1.61.4 1.5 1.1 1.1 1.4 2.6  
 B 1.4 3.4 2.5 2.01.5 1.2 0.7 0.9 1.0 2.9  
 X20.9 3.3 1.1 1.3 1.1 2.4  
MeanAll 1.7 2.2 2.4 1.41.2 1.8 1.3 1.4 1.2 3.10.50.9
 NC            
L–L0.3 1.0 1.1 1.8 1.61.2 1.7 1.0 1.8 1.0 1.5  
 0.6 1.0 1.7 2.0 3.51.4 2.8 0.9 2.0 0.9 1.3  
 B 2.0 2.1 1.3 3.30.8 2.2 1.1 2.7 1.4 3.8  
 X2 1.8 3.6 3.0 4.51.0 3.0 0.6 2.5 0.9 1.4  
MeanAll 1.4 2.1 2.0 3.21.1 2.4 0.9 2.2 1.0 2.00.30.5
 NC            
L–H0.3 1.1 0.6 0.5 0.80.6 0.7 0.4 0.8 0.6 0.8  
 0.6 0.9 0.9 0.8 0.70.5 1.1 0.4 0.7 0.5 0.8  
 B 1.0 0.7 0.6 0.70.3 0.5 0.3 0.7 0.5 1.0  
 X2 0.9 1.0 0.9 0.30.4 0.8 0.3 0.5 0.4 0.6  
MeanAll 1.0 0.8 0.7 0.60.4 0.8 0.3 0.7 0.5 0.80.10.2
 NC            

The X2 nutrient solution produced adverse effects at high Ca concentrations, irrespective of Mg concentration, on all but Bishopdale material (Table 4).

In experiment 2 the magnitude of the responses to the solution with a Ca : Mg of 0.05 : 1 varied between plants of different provenances, but the total and root weights were, with the exception of Coxhoe material, lower than for corresponding material grown with a Ca : Mg of 0.1 : 1 (Table 6).

Table 6.  Mean total and root d. wt (mg), mean shoot and root Ca and Mg concentrations (me) and plant Ca : Mg for treatments low Ca-high Mg with Ca : Mg of 0.05 : 1 and 0.1 : 1, with the basic nutrient solution
Prov.Treat. Ca : MgTotal Wt.Root Wt.Shoot CaRoot CaShoot MgRoot MgShoot Ca : MgRoot Ca : Mg
  1. Prov., provenance; K, Kynance Cove; R, Rhum; H, Holme-next-the-Sea; C, Coxhoe; B, Bishopdale. *, significant difference between the pairs of treatments at P = 0.05 (t-tests). n = 6.

K0.1 : 1293.658.815.010.715.014.01.00.8
 0.05 : 1253.145.018.3 6.920.0* 9.0*0.90.8
R0.1 : 1812.398.511.718.718.825.10.60.7
 0.05 : 1490.0*59.813.314.522.135.2*0.60.4
H0.1 : 1296.059.0 8.011.122.020.30.40.5
 0.05 : 1148.3*25.2 8.016.237.0*29.1*0.20.6
C0.1 : 1321.652.6 6.510.724.615.90.30.7
 0.05 : 1396.056.0 7.312.524.823.00.30.5
B0.1 : 1180.234.120.523.439.222.70.51.0
 0.05 : 1157.229.212.5*19.828.1*45.0*0.40.4

Plant Ca and Mg concentrations (Figs 2 and 3) reveal a great complexity of responses among materials of the five provenances to the various Ca : Mg treatments; attention is drawn here only to salient observations. Shoot and root concentrations of Ca and Mg show similar patterns in relation to the solutions provided; root Mg concentrations were lower than root Ca concentrations throughout the experiments with the exception of the low Ca and high Mg treatments. All plants (apart from Coxhoe material) had their highest Ca concentrations with high Ca and low Mg treatments; this contrasted significantly with low Ca concentrations produced when high Ca and high Mg were offered. The former treatment always resulted in shoot Mg concentrations significantly lower than Ca concentrations, as did the high Ca and high Mg with two exceptions (Coxhoe with basic and Holme with X2 nutrient solutions).

With the exception of Bishopdale material the Ca concentrations in plants grown with the 0.1 : 1 and the 0.05 : 1 treatments (Table 6) were not significantly different for either shoots or roots, whilst shoot Mg concentration with the 0.05 : 1 treatment was significantly greater for Kynance, Holme and Bishopdale plants and root Mg was significantly greater for all except Coxhoe plants at this ratio.

Experiment 3

Results from the soil cultivation experiment are presented in Table 7. Total and root yields followed the same pattern, with material of all provenances giving high yields with the dolomitic soil from Cefn-on-Farm. Only Bishopdale material gave a significantly greater yield with a Carboniferous limestone soil (Langcliffe). However, the Coxhoe dolomitic soil was associated with low yields of all materials including that from Coxhoe itself. The Magnesian limestone soil produced low total yields for Bishopdale and Holme material (both from low Mg soils), but also for Coxhoe (from high Mg soil), but not for Kynance material (from serpentine soil).

Table 7.  Mean (± 1 SE) measurements for Koeleria macrantha grown in a range of soils
Soil no.Shoot length(cm)Root length (cm)Total d. wt mgRoot d. wt (mg)
  1. Soils: 1, Langcliffe (Carboniferous limestone); 2, Cefn-on-Farm (dolomitic limestone); 3, Coxhoe (dolomitic limestone); 4, Elwick (Magnesian limestone); 5, John Innes no. 1 compost. n = 8; P = 0.05.

Provenance Kynance Cove
1 9.0 (0.6) 8.3 (0.6) 747 (249)212 (73)
2 9.4 (0.4)10.1 (0.8) 933 (122)296 (54)
3 6.6 (0.5) 8.5 (0.7) 593 (86)225 (42)
4 7.7 (0.4) 9.2 (0.3)1044 (57)469 (46)
5 7.4 (0.4) 8.8 (0.6) 701 (117)242 (147)
LSD 1.5 2.0 372.4161.8
Provenance Holme-next-the-Sea
114.2 (0.5) 7.7 (0.8)1757 (223)411 (61)
217.0 (1.1)11.2 (0.3)1561 (109)502 (38)
312.3 (0.2)11.1 (1.0) 672 (52)276 (30)
410.2 (0.4) 9.5 (0.2) 953 (122)444 (71)
513.6 (0.4)10.1 (0.2)1631 (174)659 (20)
LSD 2.1 2.0 479.4172.6
Provenance Coxhoe
111.4 (0.6) 7.6 (0.8)1431 (199)348 (49)
212.6 (0.5) 9.3 (0.2)1770 (133)586 (55)
3 9.3 (0.2) 9.5 (0.2) 583 (104)274 (44)
4 9.5 (0.4) 8.4 (0.3) 886 (60)383 (31)
510.4 (0.1) 8.7 (0.2)1617 (262)674 (119)
LSD 1.6 1.5 656.9252.9
Provenance Bishopdale
117.3 (0.2) 9.9 (0.3)2405 (205)806 (100)
215.3 (0.4) 9.7 (0.2)1488 (201)571 (75)
312.6 (0.4) 9.2 (0.2)1131 (64)469 (30)
412.4 (0.3) 9.4 (0.3)1544 (143)774 (85)
512.4 (0.2) 9.9 (0.3)1515 (159)668 (208)
LSD 1.8NS 645.4231.7

The Cefn-on-Farm dolomitic and the Langcliffe Carboniferous soils produced the longest shoots, and the Coxhoe dolomitic and Elwick Magnesian limestone soils the shortest for all provenances. The Carboniferous limestone soil was associated with the shortest roots for Coxhoe and Holme material; root lengths were not significantly different for any soils for the other two provenances.

Discussion

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Conclusion
  8. Acknowledgements
  9. References

There was a much more adverse effect of high substrate Ca concentration and a high substrate Ca : Mg on the yields of plants of all five provenances than that of the high Mg concentration and low Ca : Mg which had been expected from the work of Cooper & Etherington (1974). There was no recurrent pattern between substrate Mg concentration and yield throughout the four nutrient solution concentrations, but plant Mg was usually lower and less variable than plant Ca concentration. Table 4 shows that while Bishopdale plants responded well to the high levels of nutrients in the X2 solution, with four of the top six yields at this concentration, plants of the other four provenances all produced low total yields in this solution. Jeffries & Willis (1964) observed that increasing the concentration of nutrients led to the restriction of plants to cultures where the ratio of Ca to other cations is low. This observation is supported here: plants of one provenance died at the most concentrated nutrient solution and high Ca – low Mg treatment and three of the others produced very low yields (Fig. 1c).

The results of a taxonomic study (Dixon, 2001) suggest that K. macrantha is a variable species, with numerous ecotypes showing morphological and growth adaptations to a wide range of ecological conditions. It is not therefore, surprising to find variation between plants of five provenances, varying in soil chemistry, in their response to soil Ca and Mg. For example Kynance plants from serpentine, with a native soil Ca : Mg of 0.3 : 1, produced their highest yields (Table 4) where both shoot and root Ca : Mg were low (0.6–2.0 : 1) – (see Table 5), while material from Coxhoe with a native dolomitic soil Ca : Mg of 1.8 : 1, also produced its highest yield with low Ca treatments, but with plant Ca : Mg generally lower than those of Kynance plants (0.3–1.1 : 1). Plants from Rhum, on basalt, and on Holme, from coastal sand, showed no pattern in their response to either substrate Ca : Mg or nutrient solution concentration (Fig. 1): Bishopdale material, from Carboniferous limestone, was also variable in its response to substrate Ca : Mg but uniquely produced high yields with X2 nutrient solutions, despite the fact that its native soil has only moderate levels of major nutrients.

None of the Ca : Mg treatments used were outside the range within which K. macrantha was able to grow, but the high Ca – low Mg with a ratio of 25 : 1 had the most detrimental effect overall, and the resulting plant Ca : Mg of around 7.0 : 1 or greater was often associated with low yields (Tables 4 and 5). Natural soils with ratios as high as 25 : 1 may be uncommon: Allen (1989) records values for the Ca and Mg concentrations of chalk and limestones soils which give ratios of around 11 : 1, but nevertheless our plants from basalt, sand and Carboniferous limestone all produced high yields with at least one of the high Ca – low Mg treatments.

Material of all provenances grew well at substrate Ca : Mg down to 1 : 1. This ratio is sometimes quoted as being the minimum for normal plant growth but Hunter (1949) found that a range of substrate Ca : Mg from 32 : 1 to 1 : 4 had no significant effect on the yield of alfalfa, although the highest yield was at a ratio of 1 : 1.

Our plants from serpentine and dolomite grew well at the very low Ca : Mg treatment ratio of 0.05 : 1, and with high plant Mg concentrations (Fig. 3), and even plants from basalt, sand and Carboniferous limestone produced moderate yields at the low ratio of 0.1 : 1. This ratio is just within the range found in soils with high Mg levels, such as those over serpentine, but much lower than those of the native soils of the other four provenances (Table 1).

image

Figure 3. Mean shoot and root magnesium concentrations (± SE) obtained with different calcium and magnesium treatments and nutrient solution concentrations. Closed columns, shoots; open columns, roots. HH, high Ca-high Mg; HL, high Ca-low Mg; LH, low Ca-high Mg; LL, low Ca-low Mg. LSD (P = 0.05) for shoots of each provenance for all combinations of treatments and nutrient solutions is (me 100 g−1): K 4.00; R 3.71; H 5.33; C 5.33; B 8.21.LSD (P = 0.05) for root of each provenance for all combinations of treatments and nutrient solutions is (me 100 g−1): K 2.46; R 2.45; H 2.80; C 4.90; B 4.79.

Download figure to PowerPoint

Although K. macrantha has produced high yields in sand culture with high concentrations of Mg here, these do not necessarily reflect conditions under natural growth in soil. Thus in experiment 3, plants of all four provenances grew very well on the dolomitic soil from Cefn-on-Farm, which has the highest Mg concentration and the lowest Ca : Mg of the soils used (see Table 3). Plants of all four provenances produced higher total and root yields on this soil than on the John Innes compost which had been their substratum for the previous year, and on which all had flowered abundantly. However, K. macrantha of every provenance gave a low yield on the dolomitic soil from Coxhoe. This has a very low total N of 0.05% and it seems likely that poor growth here was due to nitrogen deficiency. Data referring to plant dimensions are not at variance with yield and require no separate interpretation.

Cooper & Etherington (1974) in their work on the vegetation of the limestone soils of South Wales also grew K. macrantha (from a Carboniferous limestone site) on soil collected from Cefn-on-Farm and found that the species produced a lower yield (33% reduction) on this than on soil mixtures of the dolomitic soil and soil from Carboniferous limestone: the reduction however, was not statistically significant. In another experiment they used dolomitic soil and the same soil with added MgCO3 plus NPK fertilizer, and although K. macrantha performed less well on the latter again the decline was not significant. Proctor (1971) reported that the addition of sodium nitrate to a serpentine soil reduced growth in oats. This was ascribed to increased solubilization of Mg by the added nitrate. Therefore the Mg toxicity described by Cooper & Etherington (1974) might be explained by the addition of NPK fertilizer to a soil that had had its naturally high Mg concentration supplemented with extra MgCO3.

Soil analyses by Cooper & Etherington (1974) gave Mg levels which ranged from around 2 me 100 g−1 on the Carboniferous limestone to between 4 and 15 me 100 g−1 on the dolomitic limestone, but their values for Ca are all extremely low, not exceeding 3 me 100 g−1 even on the Carboniferous limestone. These low values were attributed to the method of soil analysis used, but the authors considered that this would not influence the comparison made between the soil types. However, the Ca : Mg ratios for the dolomitic limestone, using their figures, are all less than unity and as low as 0.01 in some cases, which is less than that recorded from serpentine (Proctor, 1971), and they concluded that K. macrantha is absent from the dolomitic limestone in South Wales as a result of the low Ca : Mg present, and the sensitivity of K. macrantha to high soil Mg.

It seems unlikely that their Ca values accurately reflect the exchangeable Ca present in the limestone soils of South Wales; values given in Allen (1989) for Carboniferous limestone soils, by Findlay (1965), by Ball (1960), and by Frisby (1961) for Magnesian limestone soils and by the current authors are all much higher. If these revised figures are used to determine the Ca : Mg ratio for the dolomitic soil then this is greater than unity (see Table 3).

Conclusion

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Conclusion
  8. Acknowledgements
  9. References

The sand culture results presented suggest that, when compared with other members of the Poaceae (see Key et al., 1962; Proctor, 1971; Nagy & Proctor, 1997), K. macrantha has a relatively high tolerance of substrate magnesium. Thus, considering the occurrence of K. macrantha on serpentine, where the Ca : Mg is less than unity, and from our results, where K. macrantha from populations native not only to serpentine and dolomite, but to dune sand and Carboniferous limestone from North Yorkshire, have grown equally well on dolomitic soil, it would appear that its absence from the dolomitic limestone of South Wales is not due to a sensitivity to soil Mg.

Acknowledgements

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Conclusion
  8. Acknowledgements
  9. References

I am grateful to Dr D. J. Hambler for critical reading of the manuscript and to Dr A. D. Headley for helpful discussion and advice.

References

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Conclusion
  8. Acknowledgements
  9. References
  • Allen S ( ed. ). 1989. Chemical analysis of ecological materials, 2nd edn. Blackwell Scientific Publications. Oxford, UK.
  • Ball DF. 1960. The soils and land use of the district around Rhyl and Denbigh. Memoirs of the soil survey of Great Britain, England and Wales. Harpenden, UK: Agricultural Research Council.
  • Bradshaw AD, Lodge RW, Jowett D, Chadwick MJ. 1958. Experimental investigations into the mineral nutrition of several grass species Part I. Calcium level. Journal of Ecology 46: 749757.
  • Cooper A & Etherington JR. 1974. The vegetation of Carboniferous limestone soils in South Wales. I Dolomitization, soil magnesium status and plant growth. Journal of Ecology 62: 179190.
  • Dixon JM. 2001. On the status of the genus Koeleria Pers in Britain. Watsonia 23: 377390.
  • Findlay DC 1965. The soils of the Mendip District of Somerset. Memoirs of the soil survey of Great Britain, England and Wales. Harpenden, UK: Agricultural Research Council.
  • Frisby ME. 1961. The Rendzina soils of the Magnesian Series in Durham. MSc thesis, University of Durham.
  • Grime JP & Lloyd PS. 1973. An ecological atlas of grassland plants. London, UK: Edward Arnold.
  • Hunter AS. 1949. Yield and composition of alfalfa as affected by variations in the calcium- magnesium ratio in the soil. Soil Science 67: 5362.
  • Jeffries RL & Willis AJ. 1964. Studies on the calcicole-calcifuge habit II. The influence of calcium on the growth and establishment of four species in soil and sand culture. Journal of Ecology 52: 691707.
  • Karataglis S, Babalonas D, Kabasakalis B. 1982. The ecology of plant populations growing on serpentine soils. II Ca/Mg ratio and the Cr,Fe,Ni, Co concentrations as development factors of Buxus sempervirens L. Phyton (Austria) 22: 317327.
  • Kelley WP & Brown SM. 1927. Replaceable bases in soils. Department of Agriculture Experimental Station Technical Paper 15. Berkeley, CA, USA: University of California.
  • Key JL, Kurtz LT, Tucker BB. 1962. Influence of ratio on exchangeable calcium-magnesium on yield and composition of soybeans and corn. Soil Science 93: 265270.
  • Malloch AJC. 1971. Vegetation on the maritime cliff-tops of the Lizard and Land's End Peninsulas West Cornwall. New Phytologist 70: 11551197.
  • Martin JC. 1929. Effect of crop growth on replaceable bases of some California soils. Soil Science 27: 123136.
  • Nagy L & Proctor J. 1997. Soil Mg and Ni as causal factors of plant occurrence and distribution at the Meikle Kilrannoch site in Scotland. New Phytologist 135: 561566.DOI: 10.1046/j.1469-8137.1997.00681.x
  • Proctor J. 1971. The plant ecology of serpentine. II. Plant response to serpentine soils. Journal of Ecology 59: 397410.
  • Proctor J & Woodell SRJ. 1971. The plant ecology of serpentine. I. Serpentine vegetation of England and Scotland. Journal of Ecology 59: 375395.
  • Rodwell JS ( ed. ). 1992. British plant communities, Vol. 3. Cambridge, UK: Cambridge University Press.
  • Sanik J Jr, Perkins AT, Schrenk WG. 1952. The effect of the calcium-magnesium ratio on the solubility and availability of plant nutrients. Soil Science Society of America Proceedings 16: 263267.
  • Trefor Jones H. 1951. Magnesium as a plant nutrient. Chemistry and Industry 7: 11081110.
  • Walker RB. 1954. The ecology of serpentine rocks. II. Factors affecting plant growth on serpentine soils. Ecology 35: 259266.