Comparing how Lupinus angustifolius and Lupinus luteus use zinc fertilizer for seed production

Abstract

Zinc (Zn) deficiency is a common problem on the sandy acidic soils insouth Western Australia (WA) for seed (grain) production
of Lupinus angustifolius, and L. luteus. The Zn requirement of L. luteus is not known; while that of L. angustifolius has only been measured in one field experiment in WA. The effectiveness of Zn fertilizer was measured in 1997 for grain production
of L. angustifolius cv. Gungurru, and L. luteus cvv. Motiv and Teo, for Zn applied once only to plots not treated with Zn before, either in 1997 (current Zn) or in a previous
year (previous Zn, applied in one of the following years: 1983, 1984, 1986, 1990,1992). For each lupin species, the effectiveness
of previous Zn decreased relative to the effectiveness of current Zn, the decrease being larger with increasing time since
application. Fourteen years after application, the decrease was 96% for L. luteus cv. Teo compared with about 65% for the other two lupins. When no Zn was applied, L. luteus produced larger grain yields than L. angustifolius. For Zn applied in the current year, relative to L. angustifolius cv. Gungurru, L. luteus cv. Motiv used the Zn about 45% less effectively and so required about twice as much Zn to produce the same percentage of
the maximum (relative) yield. L. luteus cv. Teo used the Zn about 70% more effectively so only required about two thirds the Zn needed by L. angustifolius to produce the same relative yield. For Zn applied in each of the previous years, and relative to L. angustifolius cv Gungurru in each of those years, L. luteus cv. Motiv used the previously applied Zn from about 3 to 33% more effectively for producing grain, whereas L. luteus cv. Teo used previously applied Zn about 15 to 88% less effectively. Both currently and previously applied Zn fertilizer
increases lupin seed yields. The effectiveness of the previously applied Zn fertilizer declines with time since it was originally
applied.

Full text

Nutrient Cycling in Agroecosystems 59: 209– 217, 2001. 209
2001 Kluwer Academic Publishers.Printed in the Netherlands.
Comparing how Lupinus angustifolius and Lupinus luteus use zinc
fertilizer for seed production
1, 2 3,4
*
R.F. Brennan , M.D.A. Bolland and G. Shea
1
Agriculture Western Australia
, 444
Albany Highway
,
Albany
,
Western Australia
6330,
Australia
;
2 3
Agriculture Western Australia
,
PO Box
1231,
Bunbury
,
Western Australia
6231,
Australia
;
Plant Science
,
Faculty of Agriculture
,
The University of Western Australia
,
Nedlands
,
Western Australia
6907,
Australia
;
4
*
Agriculture Western Australia
,
PO Box
432,
Merredin
,
Western Australia
6415,
Australia
;
Author for
correspondence
Key words
:
Lupinus angustifolius,Lupinus luteus, Relative effectiveness of Zn, Residual Zn value, Zinc
Abstract
Zinc (Zn) deficiency is a common problem on the sandy acidic soils in south Western Australia (WA) for seed
(grain) production of Lupinus angustifolius, and L
.
luteus. The Zn requirement of L
.
luteus is not known; while
that of L
.
angustifolius has only been measured in one field experiment in WA. The effectiveness of Zn fertilizer
was measured in 1997 for grain production of L
.
angustifolius cv. Gungurru, and L
.
luteus cvv. Motiv and Teo, for
Zn applied once only to plots not treated with Zn before, either in 1997 (current Zn) or in a previous year (previous
Zn, applied in one of the following years: 1983, 1984, 1986, 1990, 1992). For each lupin species, the effectiveness
of previous Zn decreased relative to the effectiveness of current Zn, the decrease being larger with increasing time
since application. Fourteen years after application, the decrease was 96% for L
.
luteus cv. Teo compared with
about 65% for the other two lupins. When no Zn was applied, L
.
luteus produced larger grain yields than L
.
angustifolius. For Zn applied in the current year, relative to L
.
angustifolius cv. Gungurru, L
.
luteus cv. Motiv used
the Zn about 45% less effectively and so required about twice as much Zn to produce the same percentage of the
maximum (relative) yield. L
.
luteus cv. Teo used the Zn about 70% more effectively so only required about two
thirds the Zn needed by L
.
angustifolius to produce the same relative yield. For Zn applied in each of the previous
years, and relative to L
.
angustifolius cv Gungurru in each of those years, L
.
luteus cv. Motiv used the previously
applied Zn from about 3 to 33% more effectively for producing grain, whereas L
.
luteus cv. Teo used previously
applied Zn about 15 to 88% less effectively. Both currently and previously applied Zn fertilizer increases lupin
seed yields. The effectiveness of the previously applied Zn fertilizer declines with time since it was originally
applied.
Introduction more widely grown, it has been increasingly affected
by several diseases which L
.
luteus appears to tolerate
Lupinus angustifolius has been developed for sandy (Sweetingham et al. 1996). In addition, L
.
luteus is
acidic soils in southwestern Australia (WA) (Glad- better adapted than L
.
angustifolius to acidic soils
stones 1990): these soils comprise about 75% of the (Foy 1997). As the sandy soils of WA have become
approximate 18 million hectares used for agriculture more acidified, aluminium (Al) toxicity has become a
in WA. L
.
angustifolius is grown to improve soil problem (Dolling et al. 1994; Dolling 1995). L
.
luteus
fertility and acts as a break crop for diseases and pests appears to tolerate Al toxicity better than L
.
angus-
of cereals, as well as producing seed (grain) for tifolius (Sweetingham et al. 1996). Consequently, L
.
income (Hamblin 1987; Rowland et al. 1988; Delane luteus is being assessed as a possible alternative grain
et al. 1989). However, as L
.
angustifolius has been legume to L
.
angustifolius in WA.
Received 25 June 1999; accepted in revised form 12 May 2000

210
An estimated 8 to 9 million hectares of soils in WA tiveness of fertilizer Zn for grain production of the
are zinc (Zn)-deficient (Brennan 1998), including the two lupin species for Zn applied once only 14, 13, 11,
soils used mainly to grow lupins. When first cleared 7 and 5 years previously relative to the effectiveness
for agriculture, most of these soils were acutely of Zn applied in the current year. In addition, the
deficient in phosphorus (P), so profitable agricultural experiment also compared how effectively each lupin
production was only achieved by applying fertilizer P species used previous and current Zn. Since prelimin-
as single superphosphate (Bolland 1998), which in ary experiments have shown that seed of L
.
luteus
WA was traditionally made from rock phosphates have higher concentrations of cadmium (Cd) than L
.
containing modest amounts of Zn, i.e., from 400 to angustifolius, we also wished to test whether this was
21
600 mg Znkg of fertilizer (Williams 1974; Brennan so in this experiment.
1998). In addition to the Zn incidentally applied as
single superphosphate, many farmers in WA also
apply Zn directly to the soil as zinc oxide (Gartrell Materials and methods
and Glencross 1968). In such cases, the initial appli-
cation of Zn as zinc oxide has been sufficient to meet
the Zn requirements of wheat for many years (Anon Soil and site
1961; Ozanne et al. 1965), particularly where at least
21
150 kg ha of Zn-contaminated superphosphate was The field experiment started in 1983 on newly-
applied annually. cleared, acutely Zn-deficient soil. The experiment was
However, regular application of single superphos- located 60 km northeast of Esperance [33.478S,
phate to WA soils has increased the residual soil P, so 121.548E], about 650 km southeast of Perth [328S,
now less is applied. Since annual superphosphate 1168E]. The district has a Mediterranean climate, an
applications are declining, and single superphosphate annual average rainfall of about 500 mm, and a
in WA is increasingly being manufactured from rock growing season (May to November) rainfall of about
phosphate with lower Zn impurities, deficiency of Zn 400 mm.
is now being observed in many wheat areas (Brennan The experimental site was a gravelly sand over
1998). The increased use of imported diammonium clay, and is a yellow duplex soil, classified as Dy5.82
phosphate (DAP) fertilizer, which contains about one- (Northcote 1979) or as Fluventic Xerochrept (Soil
twelfth of the amount of Zn found in single super- Survey Staff 1975). Lupins do not tolerate seasonal
phosphate made in WA, has often resulted in immedi- waterlogging (Nelson and Delane 1990) so the site
ate Zn deficiency in wheat (Brennan 1998) and L
.
was selected to avoid this problem. Some properties
angustifolius (Riley et al. 1992). of the top 10 cm of the ,2 mm fraction of soil, as
The fertilizer Zn requirements of L
.
luteus grown measured on samples collected before the experiment
on the acidic soils in WA is not known, while the started, are: pH 5.2 (1:5 soil:solution 0.01 mol
fertilizer Zn requirements of L
.
angustifolius has been CaCl ); Clay, 5%; silt, 7% (Day 1965); organic
2
studied in only one WA field experiment (Riley et al. carbon, 1.0% (Walkley and Black 1934); cation ex-
21
1992). Farmers needs to know the length of time that change capacity, 6.8 cmol kg (Gillman and Supter
Zn application as contaminated Zn in single super- 1986); DTPA – extractable Zn ,0.2
21
phosphate or as zinc oxide applied in previous years mg Zn kg (Lindsay and Norvell 1978); bicarbonate-
21
remains fully effective for profitable grain production extractable P (Colwell 1963), 15 mg P kg .
of lupin crops.
This paper reports the results of a long-term (14 Experimental procedures
years) field experiment done to assesses whether the
original applications of Zn applied to the soil up to 14
years prior to 1997 (1983), either as the traditional History of the experiment
Zn-contaminated single superphosphate containing The experimental design was a randomized complete
21
400–600 mg Zn kg or as zinc oxide, or as zinc block with three replications. There was 0.4 m of
oxide 13, 11, 7 and 5 years previously (1984, 1986, untreated soil between each 1.4 by 50-m plot. Wheat
1990 and 1992), are still sufficient for crops of L
.
(Triticum aestivum) was sown in 1983, 1984, 1986,
angustifolius and L
.
luteus to produce profitable grain 1990, and 1992, when Zn fertilizer treatments were
yield in 1997. The experiment compared the effec- applied, and the pasture legume subterranean clover

211
(Trifolium subterraneum) was sown in 1985, and (Heliothis punctiger Wallengren) were controlled with
naturally regenerated subterranean clover grew in all pesticides as required.
the other years: lupins, were grown on all plots in
1997. Zinc treatments were applied once only, and Other details for
1997
were placed (drilled) 5 cm deep with the seed of Measurements
.
Lupin seedlings at the 2–3 leaf stage,
wheat or lupin while sowing at 5 cm. Previous Zn 28 days after sowing, were counted along 1-m lengths
treatments were applied as zinc oxide, in amount of of two adjacent rows at five random positions within
21
0.375, 0.75, 1.5, 3.0 and 6.0 kg Zn ha to separate each plot. In order to avoid contamination with Zn
plots which had never been treated with Zn fertilizer from the harvesting machine, 20– 25 plants were
until that time. Thus, Zn was applied only once to collected at random locations within the middle six
plots that had not been treated with Zn in previous or rows of each plot. Replicate samples of grain were
subsequent years. Three sources of zinc were applied ground and digested in a nitric-perchloric acid mix-
in 1983: (1) zinc oxide as described; (2) Zn present in ture (Johnson and Ulrich 1959), and the Zn con-
single superphosphate, which was made from rock centration in the digest was analyzed by atomic
phosphate contaminated with moderate amounts of Zn absorption spectrophotometry (Allen 1961).
21
(i.e. 400 to 600 mg Zn kg ) the amount of this Zn Grain yield was measured by machine harvesting
2121
fertilizer was 1.5 Zn kg ha ; and (3) 1.5 kg Zn ha grain from the middle six rows of each plot in late
21
as single superphosphate plus 1.5 kg ha as zinc November-early December and weighed. Sub-sam-
oxide. The recommended amount of Zn applied to ples of grain were grounded and analyzed for con-
21
wheat in WA is 1.5 kgZn ha (Gartrell and Glencross centration of nutrient elements other than Zn. To
1968). Basal fertilizers were applied to all plots in all determine P, the samples were digested in sulfuric
years from 1983 to 1996 to ensure no nutrient ele- acid and hydrogen peroxide (Yuen and Pollard 1954).
ment, except Zn, limited plant yield. Fertilizer phos- The P concentration in the digest was determined
21
phorus was applied each year at 18 kg P ha as colorimetrically by the vanadomolybdate method. For
21
imported DAP which contained ,50 mg Zn kg . Cd, samples were digested in nitric – perchloric acid
mixture (McQuaker et al. 1979) and the concentration
of Cd measured using inductively coupled plasma-
The experiment in
1997
atomic emission spectrometry using procedures pro-
In 1997, the following three lupins were sown in vided in the operating manual for the instrument.
separate 1.4 by 50 m plots for each previous or current
Zn treatment: L
.
angustifolius cv. Gungurru and L
.
Analysis of data
luteus cvv Motiv and Teo. Eight rows of seeds, 175
mm apart, were sown along each plot in mid-May. The relationship between grain yield and the amount
Seeds of L
.
angustifolius and L
.
luteus were inocu- of Zn applied was fitted to a Mitsherlich equation
lated with Bradyrhizobium lupini strain WU 425 and (Barrow and Mendoza 1990):
lime-pelleted immediately before sowing the seed at
21
120 kg ha . y5a2bexp(2cx)
Basal fertilizers were applied to all plots to ensure
21
that Zn was the only nutrient element limiting lupin where yis the grain yield (kg ha ), xis the amount of
2121
yield and comprised : (1) CoSO (0.25 kg ha ; 22% Zn applied (kg Zn ha ) and a
,
b
,
and care coeffi-
4
21
Co), Na BO (3.0 kg ha ; 8% B), applied to the soil cients. Coefficient aprovides an estimate of the
24
surface immediately before sowing; (2) CuSO ?5H O asymptote or maximum yield plateau. Coefficient b
42
2121
(6.0 kg ha ; 25% Cu) and NaMoO ?2H O (0.25 (kg ha ) estimates the difference between the
42
21
kg ha ; 39% Mo) were drilled with the seed (3) KCl asymptote and the intercept on the yield axis at x 5
2121
(100 kg ha ; 50% K) and gypsum (150 kg ha ; zero. That is, bindicates the maximum increase in
17% S) applied to the soil surface 4 weeks after grain yields due to the application of Zn fertilizer.
emergence (mid-June). Coefficient c(ha per kg Zn) describes the shape of the
All weeds were successfully controlled using pre relationship and governs the rate at which y(the yield
and post-emergent herbicides. Insects, red legged response) increases as x(the amount of Zn applied)
earth mite (Halotydeus destructor Tucker), aphids increases (Ratkowsky 1990). Mean data were fitted to
(Acrthosiphon kondoi Shinji) and native bud worm the equation by non-linear regression using a com-

212
puter program written in compiler BASIC (Barrow Results
and Mendoza 1990). The simplex method (Nelder and
Mead 1965) was used to locate the least squares Analysis of variance indicated highly significant (p ,
estimate of the non-linear coefficients. 0.001) effect due to levels of Zn addition and signifi-
As the three different lupins produced significantly cant (p ,0.05) effects due to year of application and
different grain yields for the nil-Zn treatment and the interaction of the amount of Zn applied by year of
different maximum yield plateaus, it is not valid to application.
use the ccoefficient of the Mitscherlich equation to
compare responses to applied Zn fertilizer (Barrow Plant density
and Campbell 1972; Barrow 1985). Instead the initial
slope of the relationship between grain yield and the Additions of fertilizer Zn in the current or previous
amount of Zn applied was used to compare the yield years did not have statistically significantly (P 50.05)
response of the species to Zn application. For the affect on the plant emergence. The plant densities
22
Mitscherlich equation, as xapproaches zero, dy/dx (plant m ) did not differ for the different species
tends to bc so therefore bc was used as an estimate of (standard errors in brackets): L
.
angustifolius 48 (8); L
the initial slope (Barrow and Campbell 1972; Barrow luteus cv. Motiv 49 (9): and L luteus cv. Teo 52 (5).
1975). These densities are adequate for maximum lupin grain
For grain production of each lupin species, the production of both species in WA (Nelson and Delane
effectiveness of Zn applied in each of the previous 1990).
years was calculated relative to the effectiveness of
Zn applied the current year (1997), to provide relative Zn deficiency symptoms
effectiveness (RE ) values. For each lupin species,
Zn
this was done by dividing bc values of Zn applied in Symptoms of Zn deficiency, as brown spots on
the current year (1997) or each of the previous years younger leaves, was observed on all three lupins, but
(1983, ’84, ’86, ’90, and ’92) by bc for Zn applied in was much more obvious on L
.
angustifolius cv.
the current year (1997); therefore, by definition, for Gungurru than the two L
.
luteus cultivars. The symp-
each lupin species, the RE for Zn applied in the toms for L
.
angustifolius occurred for plants grown
zn
current year is always 1.00. without added Zn (nil Zn) and those plots with low
Then how effectively each of the three lupins used amounts of Zn applied 9 and 14 years previously.
the Zn applied in the current year or in each of the Deficiency symptoms for L
.
luteus cv. Motiv and Teo
previous years was compared. For Zn applied in each only occurred on the nil Zn treatments. At the start of
of these years, this was done by dividing bc of each flowering (early September), many of the L angus-
lupin by bc for L
.
angustifolius cv. Gungurru, so by tifolius plants on the nil Zn treatment plots had died
definition, the RR species for L
.
angustifoius cv. due to Zn deficiency. By contrast, only a few plants of
Gungurru is always 1.00, regardless in which year the L
.
luteus had died on the nil-Zn plots.
Zn was applied. Comparisons were made relative to L
.
angustifolius because it is the major grain legume Grain yield
grown in WA and because L
.
luteus is being re-
searched as a possible alternative. Zinc as superphosphate
.
Zinc applied in 1983 at 1.5
21
Simple linear regression by a standard computer kg ha Zn as a Zn contaminant in single superphos-
program was used to examine the relationships be- phate produced lupin grain yields on the maximum
tween the concentration of Cd in seed, as the depen- yield plateau achieved for the six amounts of zinc
dent variable ( yaxis), and the concentration of Zn in oxide applied in the current year (1997). This is so
the seed, as the independent variable (xaxis). The when Zn was applied either as contaminated single
21
equation used was: superphosphate (1.5 kg Zn ha ) or as a mixture of
21
contaminated superphosphate (1.5 kg Zn ha ) and
221
y5A1Bx,rzinc oxide (1.5 kg Zn ha , i.e. a total of 3.0
21
kg Zn ha ). The Zn treatment applied as single
where yis the concentration of cadmium (mg/ kg) in superphosphate 14 years before growing the three
the seed, and xis the concentration of Zn in the seed different lupins had apparently provided adequate Zn
2
and r is the coefficient of determination. for grain production of all three lupins. The seed

213
21
yields were (kg ha ): L
.
angustifolius, 820; L
.
luteus Zinc oxide effectiveness in current and previous
cv. Motiv, 1020; L
.
luteus cv. Teo, 990. These are years
(
RE values
)
Zn
typical seed yield of lupins grown in WA (Nelson and
Delane 1990). The reason that the Zn applied in 1983 appears to be
Zinc applied as zinc oxide. For simplicity and sigmodial is probably due to the Zn in this treatment
clarity, data are only shown for the grain yield in- having been in contact with the soil for the longest
crease to Zn applied in three years, being 1983, 1985 period. There was therefore more time for the slow
and 1997 (Figure 1). Current and previously applied reactions between the soil and Zn to occur which has
Zn increased yield of L angustifolius and L luteus been shown to decrease the effectiveness of Zn added
(Figure 1). to soil (Barrow 1987; Brennan 1990; Brennan and
Gartrell 1990). In addition, the Zn applied in previous
years was increasingly incorporated throughout the
soil when sowing crops with tined machines, which
may increase retention of Zn by the soil. The Zn
fertilizer is then in contact with a greater soil volume
allowing chemical reaction to take place and further
decline the availability of the applied Zn for plant
uptake. These reactions have been demonstrated to
take place and these results are consistent with previ-
ous findings (Brennan 1996). Although, the Zn ap-
plied in 1983 was sigmodial, the Mitscherlich equa-
tion adequately described all the yield responses to
freshly and previously applied Zn (Table 1). There-
fore, the Mitscherlich equation was used to provide
the same equation to calculate bc, RE and RR
zn species
for Zn applied in different years.
Relative to current Zn, there was a steady decline in
the effectiveness of Zn as the length of time since the
Zn was applied increased; that is, the effectiveness of
the Zn decreased the longer it was in contact with soil
(Figure 2, RE values in Table 1). For example,
zn
relative to the Zn applied in the current year, the Zn
applied 14 years previously was about 34% as effec-
tive for L
.
angustifolius, about as 35% effective for L
.
luteus cv. Motiv and only 4% as effective for L
.
luteus
cv. Teo (see the RE values in Table 1).
Zn
Zinc oxide used by lupins
(
RR values
)
species
Relative to L
.
angustifolius cv. Gungurru (RE 5
zn
1.00), and for Zn applied in the current year (1997),
L
.
luteus cv. Motiv was about half as effective for
producing grain (RE 50.56, see Table 1), and so
zn
about twice as much Zn needed to be applied to cv.
Motiv in the current year for it to produce the same
percentage of the maximum relative yield as cv.
Gungurru. By contrast, relative to L
.
angustifolius cv.
Figure
1.
The relationship between grain yield and the amount of Gungurru, L
.
luteus cv. Teo was about 70% more
Zn applied for lupin grown in 1997 when Zn was applied in 1983 effective (RE 51.70) and so, for current Zn, about
zn
(♦), 1986 (m), and 1997 (j) for three lupin species: (a) L
.
30% less Zn needed to be applied for cv. Teo to
angustifolius cv. Gungurru;(b) L
.
luteus cv. Teo; and (c) L
.
luteus
cv. Motiv. produce the same relative yield as cv. Gungurru.

214
21
Table
1.
Values of the coefficients of the Mitscherlich equation fitted to the relationship between grain yield (kg ha ) and the amount of Zn
21
fertilizer (kg Zn ha ) applied, and relative effectiveness ( RE ) and relative response (RR ) values.
Zn species
a2bc
Lupinus Yrs abc rbc RE RR
Zn species
d
L
.
angustifolius 14 846 695 0.6111 0.952 424.71 0.340 1.00
L
.
Luteus cv
.
Motiv 14 1007 308 0.7947 0.863 244.75 0.35 1.03
L
.
Luteus cv Teo 14 984 277 0.317 0.9568 87.81 0.041 0.12
L
.
angustifolius 13 839 677 0.646 0.972 437.34 0.350 1.00
L
.
Luteus cv
.
Motiv 13 1030 318 1.008 0.9222 320.54 0.46 1.31
L
.
Luteus cv Teo 13 987 265 0.479 0.9556 126.94 0.060 0.17
L
.
angustifolius 11 822 664 0.908 0.96 602.91 0.483 1.00
L
.
Luteus cv
.
Motiv 11 993 295 1.5312 0.9775 451.70 0.64 1.33
L
.
Luteus cv Teo 11 960 260 1.096 0.953 285.02 0.134 0.28
L
.
angustifolius 7 818 661 1.3787 0.949 911.32 0.730 1.00
L
.
Luteus cv
.
Motiv 7 1008 290 1.9543 0.9408 566.75 0.81 1.11
L
.
Luteus cv Teo 7 966 307 1.8775 0.944 576.39 0.271 0.37
L
.
angustifolius 5 810 609 1.4998 0.986 913.38 0.731 1.00
L
.
Luteus cv
.
Motiv 5 1005 303 2.0833 0.898 631.25 0.90 1.23
L
.
Luteus cv Teo 5 986 279 4.7539 0.997 1326.32 0.623 0.85
L
.
angustifolius 0 808 603 2.0717 0.985 1249.24 1 1.00
L
.
Luteus cv
.
Motiv 0 1060 349 2.013 0.908 702.54 1 0.56
L
.
Luteus cv Teo 0 1037 337 6.313 0.996 2127.48 1 1.70
a
Years since Zn applied,199750; 1992 55 years; 1990 57 years; 1986 511years; 1984 513years; 1983 514 years
b
For each lupin species, the effectiveness of Zn applied in the current and each of the previous years was calculated relative to the effectiveness
of Zn applied in the current year (relative effectiveness or RE) to provide the RE values
Zn
c
The relative response (RR) of the different lupin species to Zn fertilizer was compared for Zn applied in the current year and each of the
previous years by dividing bc of each of the species by bc for L
.
angustifolius to provide RR species values. Therefore the RE for L
species
angustifolius is, by definition, 1.00.
d
L
.
angustifolius cv. Gungurru
How the three lupins used Zn in each of the measured for all lupin species. From the relationship
previous years was also compared. In each of these between grain yield (dependent variable or yaxis) and
years, relative to L
.
angustifolius cv. Gungurru, L
.
the concentration of Zn in the grain (independent or x
luteus cv. Motiv used the previous Zn more effective- axis), the concentration of Zn in grain that was related
ly to produce grain; that is RR values are always to 90% of the maximum yield, which is the critical
species
.1.00, ranging from 1.03 to 1.33 (see Table 1). concentration of Zn in the grain, was defined. The
However, L
.
luteus cv. Teo was less effective with critical values were similar for the three lupins, being
21
RR values being always ,1.00, decreasing (mg kg ): L
.
angustifolius cv. Gungurru
,
21; L
.
species
from 0.85 for Zn applied in 1992 to 0.12 for Zn luteus cv. Motiv, 22; and L
.
luteus cv. Teo, 22.
applied in 1983 (see Table 1).
Grain P and Cd concentration and the relationship
Zinc concentrations in the grain with Cd and Zn in seed
Zinc concentrations in the grain from the nil-Zn The concentration of P measured in the grain was
21
treatment were (mg kg ): L
.
angustifolius 12; L
.
unaffected by the amount of Zn applied in either the
luteus cv. Motiv 15, cv. Teo 14, but increased with Zn current year or in each of the previous years (Table 2).
additions. For example, in L
.
angustifolius
,
Zn values Compared with L
.
angustifolius concentration of P
21
increased from 12 mg kg for the nil Zn to 28 was about two times higher in L
.
luteus seed. Com-
21
mg kg for the highest amount of Zn applied. For L
.
pared with L
.
angustifolius concentration of Cd was
luteus cv Teo, grain Zn increased from 14 mg for the five to seven times higher in L
.
luteus seed. This new
21
nil Zn to 40 mg kg for the highest amount of Zn finding has since been confirmed in further field and
applied. The Zn concentration in grain of L
.
luteus glasshouse studies, done in WA, yet to be published.
was about 40% higher than in L
.
angustifolius grain. For the two cultivars of L
.
luteus, the concentration of
There was a good relationship between the grain Cd in the grain decreased as the concentration of zinc
concentration of Zn and grain yield (data not shown) in the grain increased. This suggests that as more

215
Table
2.
Concentration of P, Zn and Cd in grain, expressed on a dry
weight basis, for L. angustifolius cv. Gungurru and L
.
luteus cvv
Motiv and Teo. (Data are mean (6se, n 518)).
Phosphorus Zinc Cadmium
2121
Lupin % mg kg mg kg
Gungurru 0.33 (0.02 ) 18.81 (1.0) 0.03 (0.002)
Motiv 0.58 ( 0.03) 26.54 (1.3) 0.21 (0.01)
Teo 0.57 (0.03) 26.53 (1.2) 0.18 (0.01)
added in fertilizer applied in previous and the current
year. Therefore, we cannot comment on the effects of
P sources on Cd concentration in grain.
Discussion
The level of Zn contamination in the single super-
phosphate supplied adequate Zn for the three lupins
used in this study for up to 14 years after application
so that relative to current Zn applied as Zn oxide in
1997, it had a good residual value. This result is
consistent with results of previous research for cereal
crops (Takkar and Walker 1993; Brennan 1996).
Although the application of Zn to the soils of this
21
region is low (0.6 21.0 kg Zn kg ), previous re-
search with cereal crops has shown that the use of
21
superphosphate at 150 kg kg or more has supplied
21
sufficient amounts of Zn (.90 g Zn kg per year) to
meet the requirements of the current crop, maintain-
ing adequate Zn levels in the soil despite any decline
in the effectiveness of the original Zn application
(Brennan 1996; 1998). The residual value of Zn
fertilizer is due to undissolved element still present in
Figure
2.
The relationship between the effectiveness of Zn applied the fertilizer, the Zn that has dissolved from the
in the current year or each of the previous years relative to the fertilizer and which is either retained by the soil or
effectiveness of Zn in the current year (relative effectiveness or taken up by the plants and the Zn taken up by plants is
RE ) and length of time that the Zn fertilizer has been applied to
Zn
either returned to the soil as organic matter or re-
the soil for three lupin species (a) L
.
angustifolius cv. Gungurru, (b)
moved in grain as has been described for phosphorus
L
.
luteus cv. Motiv, and (c) L
.
luteus cv. Teo
by Barrow (1980). The amount of Zn removed in
product is typically low relative to the amount of Zn
21
fertilizer Zn is applied, in the current or previous applied. In the present study, where 6 kg Znha had
years, so less Cd accumulates in the grain. There was been applied to the soil in 1983, about 5% of the total
no relationship between the concentration of Cd in the Zn was removed in the grain from the preceding
seed and concentration of Zn in the seed of L
.
angus- crops. The pasture was not grazed or defoliated so no
tifolius
.
For cv. Teo, the fitted equation was: y[Cd Zn was removed. Losses of Zn from the soil system
212
mg kg ] 50.191 20.002 *Zn [mg / kg], r 50.62 through leaching and erosion are negligible (Brennan
(p 50.001). The fitted model for the Motiv cultivar 1998). Therefore, most of the residual Zn is still
2121
was: y[Cd mg kg ] 50.295 20.005 *Zn [mg kg ], present in the soil.
2
r50.58 (p 50.001). Note that, the concentration of However, the residual effects of Zn application are
Cd was measured in grain grown in 1997, so the Cd not as long-lasting for other soil types and plant
was derived from indigenous Cd in the soil and Cd species (Takkar and Walker 1993). Ten years after

216
21
Anon 1961. Residual value of trace elements. J. Dept. Agric. West
application at 18 kg Znha , Weir and Holland (1980)
Aust. (4 th Series) 2: 222– 223.
found that for maize production there was a decline in Barrow N.J. 1975. The response to phosphate of two annual
the effectiveness of Zn applied to a black earth species. 1. Effect of the soil’s ability to adsorb phosphate on
(higher pH and clay content than the soil used in this comparative phosphate requirements. Aust. J. Agric. Res. 26:
study). Takkar et al. (1975) suggested that an applica- 137–143.
21
tion of 22 kg Zn kg to a loamy sand would be Barrow N.J. 1980. Evaluation and utilization of residual phosphor-
us in soils. In: Khasawneh F.E., Sample E.C. and Kamprath E.J.
sufficient for at least seven crops in a wheat-ground-
(eds), The Role of Phosphorus in Agriculture. Am Soc Agron,
nut rotation, a conclusion based on observing the
Madison, WI, USA Ch 13., pp. 333– 359.
decline in soil extractable Zn over three crops. The Barrow N.J. 1985. Comparing the effectiveness of fertilizers. Fert.
residual effectiveness for maximum grain production Res. 8: 85–90.
has often been less than 10 years in alkaline and sodic Barrow N.J. 1987. Reactions with Variable Charge Soils. Nijhoff,
soils (Martens and Westermann 1991; Takkar and Dordrecht, The Netherlands.
Walker 1993). Barrow N.J. and Campbell N.A. 1972. Methods of measuring the
residual value of fertilizers. Aust. J. Expt. Agric. Anim. Husb.
Zinc concentration in grain can be used as a post-
12: 502 –510.
mortem method of determining Zn deficiency of
21
Barrow N.J. and Mendoza R.E. 1990. Equations for describing
lupins; concentrations of 21 mg Znkg in lupin grain sigmodial yield responses and their application to some phos-
may suggest future applications of Zn fertilizer are phate responses by lupins and subterranean clover. Fert. Res. 22:
required. Riley et al. (1992) found no further grain 181–188.
yield increase of L
.
angustifolius resulting from appli- Bolland M.D.A. 1998. Phosphorus. In: Moore G (ed.), Soilguide: A
Handbook for Understanding and Managing Agricultural Soils.
cations of Zn fertilizer where grain Zn concentration
21
Bulletin 4343. Agriculture Western Australia, South Perth 168–
was 19 mg kg . This study found that the critical 175.
level of Zn for L
.
angustifolius and L
.
luteus was Brennan R.F. 1990. Reactions of zinc with soil affecting its
21
similar, being 21–22 mg Zn kg which is in agree- availability to subterranean clover. II. Effect of soil properties on
ment with Riley et al. (1992) for L
.
angustifolious;the relative effectiveness of applied zinc. Aust. J. Soil Res. 28:
Riley et al. (1992) did not include L
.
luteus in their 303–310.
Brennan R.F. 1996. Availability of previous and current applica-
study.
tions of zinc fertilizer using single superphosphate for the grain
Compared with L
.
angustifolius cv. Gungurru, L
.
production of wheat on soils of south western Australia. J. Plant
luteus cv. Motiv required more current Zn to produce Nut. 19: 1099–1115.
the same relative yield, but it required less previous Brennan R.F. 1998. Zinc. In: Moore G (ed.), Soilguide: A Hand-
Zn, and vice versa for L
.
luteus cv. Teo. No explana- book for Understanding and Managing Agricultural Soils. Bul-
letin 4343. Agriculture Western Australia, South Perth 189– 192.
tion for these results can be provided. Luteus lupins
Brennan R.F. and Gartrell J.W. 1990. Reactions of zinc with soil
do have higher concentration of Cd in the seed than L
.
affecting its availability to subterranean clover I. The relation-
angustifolius. It may be possible to produce new ship between critical concentrations of extractable zinc and
cultivars of L
.
luteus that do not take up much Cd or properties of Australian soils responsive to applied zinc. Aust. J.
exclude Cd from the seed. Soil. Res. 28: 293–302.
Colwell J.D. 1963. The estimation of phosphorus fertilizer require-
ments of wheat in southern New South Wales by soil analysis.
Aust. J. Exp. Agric. Ani. Husb. 3: 190–197.
Acknowledgements Day P.R. 1965. Particle fractionation and particle size analysis. In:
Black C.A (ed.), Methods of soil analysis, Part 1. Agronomy
We thank the farmer (Jim Lay) for use of his land, monograph No. 9. Am Soil Sci Soc, Madison, WI, USA, pp.
545–545.
staff of Agriculture Western Australia who provided
Delane R.J., Nelson P. and French R.J. 1989. Role of grain legumes
technical assistance, chemists from the Chemistry in sustainable dry land cropping systems Proceedings of the 5th
Centre of Western Australia, for analyses of soil and Australian Agronomy Conference, Perth. Australian Society of
plant samples, and referees for comments on an Agronomy, Parkville, Victoria, pp. 181– 196.
earlier manuscript. The work was funded by the Pulse Dolling P.J. 1995. The effect of lupins and location on soil acidifi-
cation rates. Aust. J. Exp. Agric. 35: 753– 763.
and Oil-Seed programme of Agriculture Western Aus-
Dolling P.J., Porter W.M. and Rowland I.C. 1994. Acidification
tralia. rates in the central wheatbelt of Western Australia. 2. On a sandy
duplex soil. Aust. J. Exp. Agric. 34: 1165– 1175.
Foy C.D. 1997. Tolerance of lupin species and genotypes to acid
References soil and coal mine spoil. J. Plant. Nut. 20: 1095–1118.
Gartrell J.W. and Glencross R.N. 1968. Copper, zinc and
Allen J.E. 1961. The determination of zinc in agricultural materials molybdenum fertilizers for new land crops and pastures – 1969.
by atomic absorption spectrophotometry. Analyst 86: 531– 534. J. Dept. Agric. West Aust. (4 th series) 9: 517–21.

217
method of measuring exchange characteristics of soils. Aust. J. 1992. Zinc deficiency in wheat and lupins in Western Australia is
Soil Res. 24: 61 –66. affected by the source of phosphate fertilizer. Aust. J. Exp. Agric.
Gladstones J.S. 1990. The narrow-leafed lupin in Western Aus- 32: 455–463.
tralia. Bulletin 3990. Western Australian Department of Agricul- Rowland I.C., Mason M.G. and Hamblin J. 1988. Effects of lupins
ture, South Perth. and wheat on the yield of subsequent wheat crops grown at
Hamblin J. 1987. Grain legumes in Australia. In: Reeves T.G (ed.), several rates of applied nitrogen. Aust. J. Exp. Agric. 28: 91– 97.
Proceedings of the 4th Australian Agronomy conference, Mel- Soil Survey Staff 1975. Soil Taxonomy: A Basic System of Soil
bourne. Australian Society of Agronomy, Parkville,Victoria, pp. Classification for Making and Interpreting Soil Surveys US
65–82. Department of Agriculture Handbook No. 436. Government
Johnson C.M. and Ulrich A. 1959. Analytical methods for use in Printing Office, Washington, DC, USA.
plant analysis Bull Calif. Agric. Exp. Stn. No. 766. . Sweetingham M.W., French R.J., Dracup M., Macleod W.J., Carr
Lindsay W.L. and Norvell W.A. 1978. Development of a DTPA soil S.J., Davies C. et al. 1996. Yellow lupins: a new grain crop for
test for zinc, iron, manganese and copper. Proc. Soil Sci. Soc. acid soils. Aust. Grain. 6: 6–8.
Am. 42: 421–428. Takkar P.N. and Walker C.D. 1993. The distribution and correction
Martens D.C. and Westermann D.T. 1991. Fertilizer applications of zinc deficiency. In: Robson A.D (ed.), Zinc in Soils and Plants.
for correcting micronutrient deficiencies. In: Mortvedt J.J., Cox Kluwer Academic Publishers, Dordrecht, The Netherlands, pp.
F.R., Shuman L.M. and Welch R.M. (eds), Micronutrients in 151–165.
Agriculture. 2nd edn. Soil Sci Soc Am, Madison, WI, USA, pp. Takkar P.N., Mann M.S. and Randhawa N.S. 1975. Effect of direct
549–592. and residual availability zinc on yield, zinc concentration and its
McQuaker N.R., Brown D.F. and Klucker P.D. 1979. Digestion of uptake by wheat and ground crops. Indian J. Soil Sci. 23: 91–95.
environmental materials for analysis by inductively coupled Walkley A. and Black I.A. 1934. An examination of the Degtjareff
plasma-atomic emission spectrometry. Anal. Chem. 51: 1082 – method for determining soil organic matter and a proposed
1084. modification of the chromic acid titration method. Soil Sci. 37:
Nelder J.A. and Mead R. 1965. A simplex method for function 29–38.
minimisation. Comp. J. 7: 308–313. Weir R.G. and Holland J. 1980. The residual effects of fertilizer
Nelson P. and Delane R.J. 1990. Producing lupins in Western zinc on a black earth from north western New South Wales
Australia. Bulletin 4179. Western Australian Department of Australian Agronomy Conference, Queensland. Australian Soci-
Agriculture, South Perth. ety of Agronomy, Parkville, Victoria, pp. 151 –165.
Northcote K.H. 1979. A factual Key for the Recognition of Aus- Williams C.H. 1974. Heavy metal and other elements in fertilizer
tralian Soils. Rellim Technical Publications, Glenside, South environmental considerations. In: Leece D.R (ed.), Fertilizers and
Australia. the Environment. Aust Inst Agric Sci, Uni. Sydney, Sydney,
Ozanne P.G., Shaw T.C. and Kirton D.J. 1965. Pastures responses Australia, pp. 123 –130.
to traces of zinc in phosphate fertilizers. Aust. J. Exp. Agric. Ani. Yuen S.H. and Pollard A.G. 1954. Determination of nitrogen in
Husb. 5: 28 –33. agricultural materials by the Nessler reagent II. Micro-determi-
Ratkowsky D.A. 1990. Handbook of Nonlinear Regression. Marcel nation of plant tissue and soil extracts. J. Sci. Food Agric. 5: 364
Dekker, New York. –369.
Gillman G.P. and Sumpter E.A.1986.Modificat ionof the compulsive Riley M.M., Gartrell J.W., Brennan R.F., Hamblin J. and Coates P.