scrumby
Well for 1 the problem is not that it is getting into the medium it is offgassing off the tube into the air were it is taken in by the stoma. Then once inside it disrupts the chlorophyll prodution. And 2nd DBP and DIBP are the most unstable and most prone to offgas than any other phthalate. Not to mention the EU has banned the 2 chemicals in all horticultural supplies because of this same reason. So do a little research here I helped you out read the following paragraphs.
Journal of Experimental Botany, Vol. 37, No. 179, pp. 883S97, June 1986
Phytotoxicity of Phthalate Plasticisers
1. DIAGNOSIS AND COMMERCIAL IMPLICATIONS
J. W. HANNAY1 AND D. J. MILLAR2
Department of Pure and Applied Biology, Imperial College, Prince Consort Road,
London SW7 2BB, U.K.
Received 16 October 1985
ABSTRACT
Hannay, J. W. and Millar, D. J. 1986. Phytotoxicity of phthalate plasticisers. 1. Diagnosis and
commercial implications??J. exp. Bot. 37: 883-897.
The toxicity caused by a volatile constituent from certain samples of flexible polyvinyl chloride (PVC)
was due to dibutyl or diisobutyl phthalate (DBP or DIBP) plasticisers. It has caused serious financial
losses in the horticultural industry. The two phthalate esters have low volatilities, so any toxicity lasts
for many years. Radish (Raphanus sativus L. cv. Cherry Belle) seedlings, exposed to an air stream
containing 160-180 ng dm~3 of butyl phthalates developed chlorotic leaves within 3-4 d and died
within 12 d. Neither dioctyl nor diisodecyl phthalate (DOP nor DIDP) produced damage in the test
plants. Measurements of photosynthetic and respiratory gas exchange in intact shoots of affected
radishes showed that photosynthesis was severely inhibited whilst respiration was virtually
unaffected. Electron micrographs of sections from young leaves showed disruption of thylakoid
formation and granal stacking. In mature leaves, thylakoids and grana were well formed but
chloroplasts were swollen and the thylakoids were pushed towards the vacuolar side of the
chloroplast. Sensitivity to toxic phthalates varies between species; all members of the Crucifcrae tested
were susceptible, tomato less so, and lettuce and ryegrass were resistant. Toxicity of DIBP, from PVC
glazing strip, caused a reduction in crop value of £20000 per acre per year in commercially grown,
monocrop tomatoes.
Key words??Phthalates, plasticised PVC, radish bioassay, glasshouse, tomato, toxicity.
Correspondence to: Brunei Institute for Bioengineenng, Brunei University, Uxbridge, Middlesex
UP8 3PH, U.K.
INTRODUCTION
A tomato crop failure during the winter of 1978-79 was traced to the introduction of some
flexible polyvinyl chloride (PVC), used in a novel plant supporting system. This PVC gave off
a volatile component which was toxic to tomato but even more toxic to radish. A note was
published in the Grower (Hannay, 1980) warning nurserymen of the potential hazard from
flexible PVC especially in enclosed spaces such as glasshouses during wintertime. On seeing
this publication, a correspondent drew attention to an advertising leaflet issued by BASF,
probably around 1976 ('Plasticised PVC in Horticulture'). The leaflet describes experiments
in which several crop plants were grown in each of four temporary greenhouses; three
covered with PVC film each plasticised with a different phthalate to be compared with one
covered with polythene film as a control. The greenhouse containing DBP as a plasticiser was
toxic to several crop species, whereas those greenhouses having PVC plasticised with either
1 Present address: 18 Goddington Chase, Orpington, Kent BR6 9EA, U.K.
2 Present address and to whom correspondence should be sent: Brunei Institute for Bioengineering, Brunei
University, Uxbridge, Middlesex UP8 3PH, U.K.
© Oxford University Press 1986
Downloaded from
Oxford Journals | Life Sciences | Journal of Experimental Botany by Brandon Eckel on July 9, 2010
884 Hannay and Millar??Phytotoxicity of Phthalate Plasticisers
diethylhexyl phthalate (DEHP) or diisodecyl phthalate (DIDP) were not toxic. Apparently
the experimental work was carried out in 1970-71 in response to litigation over some flexible
PVC sheet which had proved to be similarly toxic in commercial use: no other details of this
work have been published (private communication). Subsequent enquiries revealed several
papers in Japanese dating from the early 1970's of which the comprehensive paper by Inden
and Tachibana (1975) is the only one which has been translated.
Before embarking on the experimental work it may be helpful to give a brief explanation
of the nature of plasticisers. A plasticiser is used to impart flexibility to a compound and it is
common to have 30-40% by weight of plasticiser in the flexible PVC used in tubing and
sheeting. Several kinds of plasticiser can be used but the most common are the dialkyl
phthalates. Industrially these are made from phthalic anhydride and an alkyl alcohol in the
presence of p-toluene sulphonic acid:
-COOR
-COOR
R may be straight chain or branched. The common R groups are:
(a) Butyl or isobutyl??to give DBP or diisobutyl phthalate (DIBP).
(b) Ethylhexyl??to give DEHP, often referred to in the literature as dioctyl phthalate
(DOP) since the straight chain n-octyl is hardly used. DEHP is the most commonly
used phthalate plasticiser accounting for around 25% of world production of
phthalates.
(c) Isooctyl??diisooctyl phthalate (DIOP) a mixture of octyl alcohols.
(d) Isodecyl??DIDP??together with DIOP this accounts for a further 25% of production.
(Sears and Darby, 1982.)
DBP and DIBP have excellent plasticising properties but have lost favour with the plastics
industry because of their higher volatility compared to phthalates of higher molecular
weight. Their toxicity to plants is a new factor but their effects on animals have been a
concern since the early 1970's (Autian, 1973).
MATERIALS AND METHODS
Plastics
Plastics were prepared in standard, clear rod format to specified formulations (as detailed in the text) by
ICI Plastics and Petrochemicals Division. Commercial glasshouse glazing strip was removed from
glasshouses under the auspices of officers of the Agricultural Development and Advisory Service
(ADAS) and was forwarded to us with appropriate documentation by the officer. The Technical
Services Department at ICI carried out the chemical analyses of the plastics.
Plants
Radish seeds??Raphanus sativus L. cv. Cherry Belle??were sown in a mixture of equal parts of
Levington Potting Compost and John Innes No. 2 Compost. After 7 d they were transplanted into
small seedtrays (220 mm x 165 mm x 50 mm) so that each tray contained nine plants in rows of three.
The plants were grown in a greenhouse at approximately 23 °C under natural lighting: in winter this
was supplemented with 400 W mercury fluorescent lamps. Prior to use in the bioassay young plants
were transferred for 1 d to the controlled environment room at 22 ± 1 °C and with a photon fluence
rate 185 /imol m'2 s"1 {PAR) from nine, 8 ft 125 W warm white fluorescent tubes plus four 60 W
tungsten lamps for a photoperiod of 18 h d~'.
886 Hannay and Millar??Phytotoxicity of Phthalate Plasticisers
Light measurements
Light measurements were made with a Li-Cor Quantum Meter Model LI 185.
Gas liquid chromatography
A Pye Series 104 Model 4 gas chromatograph was used with a flame ionisation detector. The glass
column (500 mm x 2 mm i.d.) was packed with 3% Dexsil 300 on 80-100 mesh Chromosorb G, and the
oven temperature was normally 230 °C. The carrier gas was oxygen-free nitrogen at 40 cm3 min"1;
hydrogen flow rate was also 40 cm3 min"1.
Glass sampling tubes (90 mm x 4 mm i.d.) were thoroughly cleaned with Spectrograde cyclohexane
(Fisons) and then heated to 300 °C for 4 h. One end of each tube was clearly marked with a diamond
marker. Tenax GC 60-80 mesh (Phase Sep) was used to pack the cleaned tubes. The Tenax had also
been preconditioned by heating to 300°C in a stream (40 cm3 min" ') of oxygen-free nitrogen for 24 h.
The filled tubes were closed at each end with caps fashioned from polythene tubing, when not in use.
Immediately prior to use they were purged with nitrogen at 100 cm3 min""1 at 300°C for 10 min.
Sampling of growth chamber atmosphere
Volatile phthalate esters were collected by drawing a known volume of air from the growth chamber
through the glass sample tube packed with Tenax. Normally the sampling was for a period of 48 h. The
flow rate was regulated to 180-200 cm3 min"1 and the total volume of air passing through the Tenax
tube was measured with a dry-type gas meter. Phthalate collected by the Tenax was desorbed by adding
0-5 cm3 cyclohexane into the unmarked end of the tube and allowing it to percolate through. The eluate
was collected in a graduated micro-vial; when the first 0-5 cm3 had percolated through, a second and
then a third 0-5 cm3 was added. The total volume of eluate collected was 1-0 cm3 and the vial was
immediately stoppered. Aliquots (50 mm3) were used for injection into the gas chromatograph.
Calibration curves were prepared by making known additions of individual phthalates into Tenax
tubes and eluting as outlined above.
Care was taken to minimize contamination from extraneous phthalates which are common
pollutants in the environment (Crosby and Singmaster, 1973; Giam, Chan, and Neff, 1975a, b; Gross
and Colony, 1973). All glassware was washed in Spectrograde cyclohexane and heated to 230 °C for
several hours immediately prior to use; polythene end caps were stored at 60 °C until used.
Infrared gas analysis
CO2 exchange was measured using an Infrared Gas Analyser (IRGA) type GC 225.2A (Analytical
Development Company Ltd., England). The IRGA was connected to the test chambers through a
three-way valve??Fig. 2(a). Air from a standard air cylinder (British Oxygen Co.) flowed through each
of the three chambers at 200 cm3 min"1. Only the shoots of the plants were exposed inside the plastic
bag; the roots and pot were excluded by placing a perspex base around the hypocotyl and sealing with
lanolin??Fig. 2(b). The volume enclosed in the chamber was about 600 cm3. The three plants were
illuminated by four, 5 ft 80 W warm white fluorescent tubes plus two 25 W tungsten bulbs??which gave
a photon fluence rate 135 /jmol m~2 s"1 (PAR) at the mid-height of the chambers.
Using a multi-channel switching unit, air could be passed through a valve to the IRGA, or exhausted
to the outside. Air from each chamber flowed through the IRGA for 10 min before switching to the next
chamber.
Electron microscopy
Small pieces of leaf (approximately 2-3 mm2) were excised from a region halfway along the lamina
between the midrib and the leaf margin of treated and untreated leaves. These were placed immediately
in a solution of 2-5% glutaraldehyde (EM grade, EM-scope) fixative in a 100 mol m " 3 cacodylate buffer.
Each excised piece of leaf was then cut into smaller pieces under the glutaraldehyde (approximately 0-25
mm2 in area). The samples were then fixed for 24 h at room temperature, post fixed in osmium tetroxide
and then routinely processed (Glauert, 1980). Cut sections were 60 nm thick and were stained in uranyl
acetate and lead citrate.
RESULTS
Certain plastics are toxic
Using the bioassay system illustrated in Fig. 1 it was possible to identify those combinations
of PVC resin, plasticiser and stabiliser which were toxic and thus to identify the constituent
TABLE 1. Formulations of plastics supplied by ICI
Numbers are parts by weight of each constituent.
Formulation
PVC
DBP
DEHP
ESBO"
Stearic acid
Ba/Cd stabiliser
Ca/Zn stabiliser
Chlorosis and death
in bioassay
Plastic
A
100
42
^_
??
m2??
+
B
100
42
0-2
2
??
_
C
100
42
0-2
2
??
D
100
42
??
??
2-5
+
E
100
.. ..
42
,??,
? _
??
2-5
F
100
..
.
42??
??.
2-5
" Epoxydised soya bean oil.
which caused toxicity. The various formulations were made up into approximately 70 mm
diameter clear rod by ICI.
The plastic rod was cut into 10 cm lengths for ease of packing and to ensure a fairly
turbulent flow over the surfaces. The glass tubes were packed with 170 g of plastic for
bioassay. The same sample could be used repeatedly, since the toxin continues to volatilize at
room temperature for several years.
Formulations A and D caused obvious chlorosis within a few days and the plants were
dead within about 12 d. Formulations B, C, E and F caused no obvious differences from the
controls. The only thing in common between samples A and D were the PVC resin and the
plasticiser??which was DBP.
FIG. 3. The effect of continuing exposure to the vapour from various samples of PVC on increase in
weight of radish seedlings. (Each point is the mean of three individual weights.)
Chemical analysis of, the original clear plastic tubing which had started these investigations
had showed that it contained DBP as plasticiser. The black plastic from the
Humberside tomato nursery had a mixture of diisobutyl phthalate (DIBP) and diisooctyl
phthalate (DIOP) as its-plasticiser. DIBP might account for it's toxicity, whereas DIOP has
not been shown to be toxic. A further batch of plastics were prepared by ICI, on a similar
basis to those already used, containing DIBP, DIOP, DNP and DIDP. Only DIBP was
toxic. In fact no plastic so far tested in our bioassay has been found to be toxic unless it
contained either DBP or DIBP.
When the plastics were prepared by ICI a sample of each constituent was taken from the
identical batch so that each could be tested separately. Preliminary tests with the pure
plasticiser did not confirm the toxicity of the butyl phthalates. Then it was realised that the
surface area of the liquid plasticiser being presented was considerably less than the surface
area of the pieces of plastic rod normally used. When a filter paper 'sail' was erected in the
long, narrow, glass 'boat' in which the plasticiser was contained, so that the total surface area
was equal to that of the plastic, then chlorosis occurred quite quickly and the radishes died
within about 14 d. If the surface area of the plasticiser was reduced to about one third that of
the plastic then the toxicity was much less than from the plastic. This suggests that the
plasticiser must form a film on the surface of the plastic, since it only constitutes about one
third of the total mass.
With equal weights and similar surface areas of three plastics the results shown in Fig. 3
were obtained.
Only the plastic containing DBP produced an obvious difference in weight from the
control. The DBP-treated plants after 10 d had produced three leaves whereas the controls
had four. Moreover, the leaves of the DBP-treated plants were by now withered and some
cotyledons had also started to collapse. Cotyledons which were still turgid remained green.
In contrast to this the leaves of the control plants were healthy and green??as were their
cotyledons. Chlorosis was beginning to show in DBP-treated plants after 3 d together with
curling of some leaves. At 6 d chlorosis was developing in the veinal regions of older leaves.
The youngest leaf on these plants did not develop any chlorophyll. This is typical of the effects
of both DBP and DIBP, either when incorporated into flexible PVC or as pure substances.
Glazing strip in some glasshouses causes toxicity
In February 1983, we first became aware of a disorder which was afflicting tomato plants in
a few commercial glasshouses. It's cause was unknown but Mr N. Starkey, the officer in
charge of tomato cultivation trials at the Efford Experimental Station of ADAS was
convinced that the toxin was in the atmosphere but had eliminated most of the obvious
possibilities such as ethylene, propylene and sulphur dioxide. After further investigation he
concluded that the most likely source of toxin was the PVC glazing strip. This is the 'cushion'
of thin PVC tubing on which the sheets of glass are bedded to separate them from the
aluminium frame and to give an air-tight seal.
A sample of this glazing strip was tested in our bioassay and was found to be very toxic. A
similar sample from a commercial tomato nursery in which similar toxicity symptoms were
appearing, was also toxic. However, glazing strip from an adjacent glasshouse, on the same
site and planted with the same cultivar, was found to be non-toxic in the bioassay. This was
anticipated since it was not causing any disorder in the glasshouse tomatoes. Chemical
analysis of the two toxic glazing strips revealed that both contained approximately equal
amounts of DIBP and DEHP whereas the non-toxic strip contained only DEHP. The
non-toxic strip came from an older glasshouse whereas the toxic strip came from recently
glazed structures. The manufacturers confirmed that a change in formulation of the glazing
strip had occurred in 1981. Of the several cases now known, in which similar disorders have
been found in monocrop tomatoes, all were in recently erected or reglazed glasshouses and
new glazing strip had been used. At NVRS a new glasshouse was being used to raise Brassica
species and these plants were abnormal even in early autumn, when ventilation was
occurring. Discussions with the glasshouse manufacturers and their suppliers of glazing strip,
together with collaborators from ICI, led to the recommendation that the toxic glazing strip
should be replaced by strip containing only DIDP as plasticiser. We tested the new
formulation in our bioassay and found it to be non-toxic. An additional change was
incorporated into this new strip in that the original aluminium powder which gave a silvery
appearance to the glazing strip was replaced with carbon black. This was likely to improve
the stability of the material and it made the new glazing strip conspicuously different from the
original. It should be emphasized that only the recently manufactured batches of the silvery
glazing strip might be toxic.
Comparison of the three types of glazing strip in a radish bioassay is shown in Fig. 4. It is
obvious that the strips containing DEHP or DIDP were innocuous but the strip which
included DIBP was very toxic. The remaining three plants after 10 d treatment are shown in
Plate 1.
In addition to the favourable bioassay for the new, black glazing strip it was also necessary
to have long term tests under commercial conditions. One such test was at NVRS in which
two of the five sections of the experimental glasshouse had the old glazing strip replaced
whereas the other three did not. The two renewed sections have now shown no disorders for
more than a year whereas in the sections containing the original strip, brassicas still produced
the standard chlorotic syndrome (Hardwick, Cole, and Fyfield, 1984).
Value of losses in monocrop tomatoes
Investigation of the commercial implications was made possible by the kind cooperation
of the tomato nursery in which we first diagnosed the toxic glazing strip in the new
glasshouse. This nursery also had similar glasshouses with non-toxic glazing strip. The
nurserymen kept records of yields in the two comparable houses from the winter of 1981
when the new, toxic house first came into use. In the autumn of 1983 the toxic glazing strip
was replaced with the non-toxic black strip. The yield comparisons for 1983 and 1984 seasons
are shown in Table 2.
Hannay and Millar??Phytotoxicity of Phthalate Plasticisers 891
TABLE 2. Calculations for loss in gross crop value due to toxicity from glazing strip
Year and month
1983 March
April
May
June
July
August
September
October
Annual totals
1984 March
April
May
June
July
August
September
October
Annual totals
Average
value"
(£ ton"1)
1199
1180
886
697
683
503
332
427
1069
1107
1075
629
726
623
298
498
Toxic House
Yield
(ton acre"1)
1-3
9-8
19-9
26-8
318
23-7
17-5
121
142 9
9-5
16-8
21-4
28-8
283
22 1
14-0
18-9
159-9
Crop value
(f month"1)
(A)
1559
11564
17631
18680
21719
11921
5810
5 167
£94051
10156
18 598
23005
18115
20546
13768
4 172
9412
£117772
Control
Yield
(ton acre"1)
41
17-6
24-7
29-4
31-8
23-8
19-9
172
168-5
9-2
191
24 5
287
25-3
20-2
12 4
16-1
155 7
Crop value
(£ month"1)
(B)
4916
20768
21884
20492
21719
11971
6607
7344
£115 701
9835
21 144
26338
18052
18368
12 585
3 695
8018
£118035
Monthly loss
(gain) in crop
value £
(B-A)
3 357
9204
4 253
1812
_
50
797
2177
£21650
(321)
2546
3 333
(63)
(2178)
(1 183)
(477)
(1394)
£263
* Data for mean monthly crop values from ADAS, South Coast Glasshouse and Mushroom Advisory UniL,
Chichesten baled on gross returns for Class 1 tomatoes.
The glazing strip in the Control contained only Di-octyl phthalate as plasticiser and was assumed to be non-toxic.
The glazing strip in the TOXJC house contained Di-octyl phthalate plus Di-isobutyl phthalate during 1983 but in
November 1983 this was replaced by glazing strip containing only Di-isodecyl phthalale as plasticiser.
Each house was about one third of an acre but they were not identical in size. All the plants
were grown by nutrient film technique.
The main points to notice are:
(a) The loss in yield in 1983 occurred mainly in the first three months but also in
September and October.
(b) Because the price of tomatoes varies with the time of season, the financial loss was
relatively greater than the loss in yield.
(c) In 1984 there was virtually no difference in yield, or crop value, between the original
glasshouse and that in which the toxic glazing strip had been replaced.
Circumstances affecting the overall picture were:
(i) In 1983 all the young plants were kept for 8 weeks in the 'toxic' house, from
mid-October to mid-December, until they were planted out into their final positions.
Thus even the 'controls' would be partially retarded and, therefore, will have caused an
underestimate of the relative damage.
(ii) In 1984 the reglazed house had a fairly serious root infection from Phytophthera in
March and April which probably depressed yields temporarily, whilst in July the
'control' house was infected with whitefly which would depress yields till the end of the
season.
In spite of these vagaries, the figures provide a good approximation to the sort of losses
that can occur. In 1982 the loss was around £22000 per acre in the toxic house but some of
this was due to differences in planting dates since the new house was not ready until
FIG. 5. The effects of dibutyl phthalate, either alone or incorporated into PVC, on photosynthesis and
respiration in radish shoots. Each point above the axis (photosynthesis) is the mean value during the 18 h
photoperiod and those below the axis (respiration) are the mean values for the 6 h dark period.
December 1981. In 1985, up to the end of August, the 'loss' was about £2000 on the reglazed
house but the monthly yields are variable and the 'loss' may well be within the normal limits
of variation between houses. The problem may be confined to tomato crops in the winter and
early spring when the glasshouse is almost sealed to retain heat and to facilitate carbon
dioxide enrichment of the atmosphere. Once free ventilation occurs the toxin level falls and
the tomatoes grow well.
Toxicity due to inhibition of photosynthesis
In an attempt to discover the reason for the toxicity of the phthalates, whole shoots of
radish plants were enclosed in small individual growth chambers, so that the change in
carbon dioxide concentration of the air passing over each shoot could be monitored.
Photosynthesis in treated plants was inhibited but there was virtually no effect on respiration
(Fig. 5).
The data are from only three individual plants but the trend shown is typical of several
experiments. The plants were of different sizes at the end of 10 d but this was compensated
for by using unit leaf area for comparison. This does not allow for the presence of
chlorotic patches on some leaves of the treated plants, so a decrease in photosynthesis
was not surprising; a more useful comparison may have been on a unit chlorophyll
basis. Nevertheless, under these circumstances the respiratory process was not seriously
affected; it appears much less sensitive to the toxic phthalates than the photosynthetic
Hannay and Millar??Phytotoxicity of Phthalate Plasticisers 893
Measurements were also made of the stomatal resistance of each plant at the beginning
and end of the 10 d enclosure in the IRGA system. Table 3 shows the results. In all treatments
the stomatal resistance was lower at the end of the experiment than at the beginning but there
was no obvious difference between the three treatments. The decrease in photosynthesis does
not appear to be due to stomatal closure effected by DBP.
TABLE 3. Total leaf areas and stomatal resistances at the start and end of CO2 exchange
measurements
Mean of three leaves ±s.e.
Treatment
1. control
start 2. DBP plastic
3. DBP plasticiser
1. control
end 2. DBP plastic
3. DBP plasticiser
Leaf area
(cm2)
39-5
450
43-75
73-6
50-00
55-10
Stomatal
resistance
(±s.e.)(s m"1)
7-3 ±0-84
6 9± 1 16
8-2±l-75
2-4 ±0-01
21 ±0-36
2-9 ±0-45
Chloroplast development disrupted
Investigations of the fine structure of treated leaves, in 1979-80, by third year
undergraduates revealed that chloroplast development in the youngest chlorotic leaves,
which had developed during exposure to phthalates, was severely disrupted. There were few
grana. The treated leaves also had increased numbers of plastoglobuli but starch grains were
absent.
These early, unreplicated observations have been confirmed and extended to an
examination of the effects of butyl phthalates on the more mature leaves, which were already
green and fully expanded when the exposure to DBP was commenced. These leaves normally
remain green during a 10 d exposure in the bioassay even though the younger leaves are
chlorotic. Plate 2 shows four photomicrographs to illustrate these observations. The upper
two photographs (A and B) show the comparison between a control and a treated young leaf
and the lower photographs (C and D) compare control and treated mature leaves. The
immature leaves show the typical lack of development of grana, as found previously, but the
mature leaves have well developed grana, though they have never been found to include any
starch grains when sampled after 5 d exposure to DBP. The obvious effect in the mature
chloroplast is the swelling which causes the normal ovoid shape to round out and the
thylakoid system to be pushed towards the vacuole. Plastoglobuli also appear to be more
numerous but there is no obvious breakdown of the thylakoids.
From these observations it is not surprising that photosynthesis is inhibited in
DBP-treated leaves but it is not clear whether the inhibition is equally severe in the young
versus the mature leaves. Mitochondria appeared normal in sections of DBP-treated leaves.
This is consistent with the absence of inhibition of respiration. Virgin, Hoist, and Morner
(1981) also remarked that mitochondria in young leaves appeared to be unaffected by an
exposure to DBP which completely disrupted the chloroplasts.
PLATE 2. Photomicrographs of sections through radish leaves to show chloroplasts from: (A) a young
radish leaf ( x 18 000); (B) a young radish leaf exposed to DBP vapour for 3 d ( x 18 000); (C) a mature
radish leaf (x 26000); (D) a mature leaf exposed to DBP vapour for 5 d (x 26000). G??granum;
Th??thylakoid; PG??plastoglobuli; SG??starch grain; M??mitochondrion; NA??nucleic acid (DNA);
S??stroma; CE??chloroplast envelope; V??vacuole.
DISCUSSION
In September 1981, a computer search of the literature revealed nothing about PVC toxicity
to plants. At that time it was not known that phthalates were involved and this keyword was
not used. Since then many important references have turned up (see Millar, 1985 for full
bibliography). The most relevant came from personal enquiries to Japanese and German
scientists. In the early 1970's important investigations into phthalate toxicity in PVC film had
clearly established that the butyl phthalates were the cause. Unfortunately records of these
investigations were relatively inaccessible in bulletins of research stations, or even in
advertising literature. In view of the importance of PVC film in horticulture in Japan, which
uses over a quarter of a million tons of plasticised PVC film annually (Dubois, 1978), it is
surprising that the manufacturers in the U.K. had not sought the reason why it was not being
used here. There is no reason to castigate the manufacturer who included DIBP in the glazing
Hannay and Millar??Phytotoxicity of Phthalate Plasticisers 895
strip formulation. Butyl phthalates are still listed as satisfactory for food-grade film. Any
enquiries made in this country in 1980-81 would have been unlikely to have raised any
objection to the use of DIBP. Similar errors have been made previously, in Germany and
Japan. There is no reference to PVC toxicity to plants in the book 'Plastics in Agriculture'
(Dubois, 1978). To alert the horticultural industry to the cause and cures of the glazing strip
problems, a note was published anonymously in the Grower (Anon, 1983).
The commercial data on tomato crops shows that the toxic glazing strip can cause serious
losses of around £20 000 per acre per year. The cure is expensive because it is a skilled and
time consuming job to replace the glazing strip. Apart from glazing strip other glasshouse
equipment such as hose-pipes and trickle irrigation systems could be a hazard though only a
few cases are known. Most of the glazing strip and trickle irrigation problems are likely to be
confined to PVC made during mid-1981 to mid-1983. Anything purchased since then should
contain no butyl phthalates because all plastics manufacturers were informed of the problem
in summer 1983. However, hose-pipes and paints might still be problematical. Different
species show varying sensitivity to the toxic phthalates. Brassicas are particularly sensitive
and are useful for bioassays. Other crucifers such as alyssum (Lobularia maritima L.) and
ten-week stock {Mathiola incana L.) are also sensitive. Tomato is less sensitive than the
brassicas but lettuce will grow well when tomato and brassicas are badly damaged; rye grass
(Lolium perenne L.) is also very resistant. It is probable that most plants are fairly resistant but
Virgin et al. (1981) reported that some house plants were affected by butyl phthalates
volatilizing from paint in newly decorated rooms. Hardwick et al. (1984) reports significant
differences in sensitivity between different cultivars of cabbage. More work needs to be done
to screen glasshouse plants for sensitivity. There may be other species among bedding plants
and house plants which are susceptible, but this may not be diagnosed correctly if most other
species growing alongside them appear to be healthy.
It is still unknown why some plants are susceptible when others are resistant but it is
important to discover the reason. Then it may be possible to combat toxicity in susceptible
plants, or even to turn the differences in sensitivity to advantage as a basis for a herbicide.
Only the butyl phthalates are phytotoxic under commercial horticultural conditions and
in our bioassay. The bioassay produces the toxic syndrome within a few days and young
seedlings die within 2 weeks of exposure to atmospheric concentrations in the region of
160-180 ng dm " 3 of DBP or DIBP with an air flow of 30 dm3 min " ' (this gave about one air
change every 5 min in the growth chamber). Hardwick et al. (1984) reported a maximum
concentration of butyl phthalates in the affected glasshouse at NVRS of approximately 2-0 ng
dm ~3 (average 1 -2 ng dm "3) which is about a hundred times lower than that measured in the
bioassay. At that concentration many brassicas could not be grown satisfactorily. The
concentration dropped to around 10 ng dm"3 during ventilation in late summer; brassicas
did not grow satisfactorily even at this reduced level. No symptoms were visible when the
detected concentration was lower than 014 ng dm"3. Presumably the concentrations of
phthalate in the tomato nursery were higher than 2-0 ng dm"3 DIBP during the
non-ventilated winter period when the plants were affected. No measurements were made at
this time because the analytical technique was not developed until 1984.
At the concentrations present in glasshouses during wintertime, young tomato plants need
about 3 weeks exposure before a trained eye can spot the first signs of disorder (N. Starkey,
personal communication). With a prolonged exposure of three months (December to
February) symptoms are severe. The young leaves have a yellow-green interveinal mottling
but not an obvious chlorosis; the old leaves have quite large blotches of interveinal necrosis
and at worst the necrotic patch goes completely white or transparent and may be 2-3 cm2 in
area. Only about half to two thirds of the leaf is green; the rest is papery and with occasional
896 Hannay and Millar??Phytotoxicity of Phthalate Plasticisers
transparent patches. Nevertheless, the first truss sets but the fruits are smaller and fewer in
number. This also happens in the next few trusses and even when ventilation commences it
may take another 2 months or so before the plants produce a near normal crop. They do not
make up the early losses and towards the end of the season, when ventilation is reduced, they
again show a small decrease in yield.
The cytological changes of chloroplast structure seen in the electron micrographs are
consistent with those found by Virgin et al. (1981) so far as the young leaves are concerned.
Virgin et al. (1981) showed no pictures of mature leaves following treatment but implied that
such leaves also became chlorotic. This was not so during the fairly short exposure time given
in our bioassay, although it produced severe chlorosis in the young leaves. The mature leaves
showed no chlorosis and no malformation of thylakoids or grana. However, the whole granal
system was displaced towards the vacuole. This visual evidence suggested that the
chloroplasts could be capable of carrying out photosynthesis in the mature leaves but not in
the young leaves. Virgin et al. (1981) also found a marked decrease in carotenes in young
treated leaves and an accumulation of a carotene precursor??probably phytoene. They
suggested that the mechanism of action of DBP could be interference with carotene
metabolism. Thus, free radicals generated by chlorophylls during photosynthesis would
not be quenched and bleaching would result in chlorosis. If this hypothesis is correct, yet
mature green leaves do not become chlorotic during treatment, then carotene turnover in
mature leaves must be negligible. The carotene formed in the young leaf, prior to exposure
to the toxin, must continue to act as a quenching agent. After several weeks exposure of older
plants we occasionally do find mature green leaves which start to turn chlorotic. This may
be due to enhanced senescence rather than a direct effect of the phthalates on carotene
synthesis.
ACKNOWLEDGEMENTS
The authors' wish to thank the SERC and ICI for a CASE studentship (DJ.M.) and the latter
for chemical analysis of plastics; undergraduates for assistance with EM work (D. Moss, C.
Mallory and A. Waite); and to V.M.C. Baileys and partners for providing data on their
tomato yields.
LITERATURE CITED
ANON (Undated). Plasticised PVC in Horticulture. Report on BASF plasticisers. BASF, Ludwigshafen,
Germany.
ANON, 1983. New glazing strip damages glass crops. Grower, 23 June, pp. 5.
AUTTAN, J., 1973. Toxicity and health threats of phthalate esters: review of the literature. Environmental
Health Perspectives, 4, 3-26.
CROSBY, D. G., and SINGMASTER III, J. A., 1973. Phthalate esters as artifacts in pesticide research.
Abstracts of Papers??American Chemical Society 166th Meeting??Pesticide Chemistry, p. 59.
DUBOIS, P., 1978. Plastics in agriculture translated from the French edition (1973) and revised by C. A.
Brighton. Applied Science Publishers, London.
GIAM, C. S., CHAN, H. S., and NEFF, G. S., 1975a Sensitive method for determination of phthalate
plasticisers in open ocean biota samples. Analytical Chemistry, 47, 2225-9.
19756. Rapid and inexpensive method for detection of polychlorinated biphenyls and
phthalates in air. Ibid. 47, 2319-20.
GLAUERT, A. M., 1980. Fixation, dehydration and embedding of biological specimens. In Practical
methods in electron microscopy. Volume 3. Ed. A. M. Glauert. North-Holland, Amsterdam.
Pp. 1-201.
GROSS, F. C, and COLONY, J. A., 1973. The ubiquitous nature and objectionable characteristics of
phthalate esters in aerospace technology. Environmental Health Perspectives, 3, 37-47.
Hannay and Millar??Phytotoxicity of Phthalate Plasticisers 897
HANNAY, J. W., 1980. Fumes from flexible PVC damage young plants. Grower, 21 August,
Pp. 28-9.
HARDWICK, R. C, COLE, R. A., and FYFIELD, T. P., 1984. Injury to and death of cabbage (Brassica
oleracea) seedlings caused by vapours of dibutyl phthalate emitted from certain plastics. Annals of
Applied Biology, 105, 97-105.
INDEN, T., and TACHIBANA, S., 1975. Damage of crops by gases from the plastic materials under
covering conditions. Bulletin of Mie University, Faculty of Agriculture, 50(6), 1-10.
MILLAR, D. J., 1985. Phytotoxicity of phthalate plasticisers. Ph.D. Thesis, University of London.
SEARS, J. K., and DARBY, J. R., 1982. The technology of plasticizers. Wiley-Interscience, New York.
VIRGIN, H. I., HOLST, A. M., and MORNER, J., 1981. Effect of di-n-butylphthalate on the carotenoid
synthesis in green plants. Physiologia plantarum, 53, 158-63.
