4.3.5 Individual pesticide sorption
Triazine pesticides
A wide range o f ten pesticides from two broad groups (triazines and
organophosphates) were chosen to provide inform ation on pesticides representative o f contaminants found globally. They were also chosen to exhibit a diverse range o f solubility, atomic mass and structure. The triazine pesticides (simazine, atrazine, propazine, desmetryn, prometryn, terbutryn and cyanazine) share structural sim ilarities, in that all are based around a triazine ring (see Figures 4.1 and 4.2). The organophosphate compounds are distinct from this group. Not only do they contain phosphate, but they also lack the distinct structure shared by the triazine group (see Figure 4.3). Partly fo r these reasons, the first fo u r compounds (simazine, atrazine, propazine and to a certain extent, desmetryn) are often sorbed to a sim ilar degree. Not only are these four compounds structurally sim ilar, they are also o f a sim ilar weight class (see Figure 4.1) and are detected by the GC unit in very close succession (meaning that chem ically they are also very sim ilar, as the GC column separates compounds according to a variety o f criteria, including weight, structure, side ion groups, and im portantly, solubility). Figures 4.6 - 4.12 illustrate that these four compounds are sorbed to a sim ilar (comparatively low) degree. These four compounds are in fact the most reliably detected compounds o f the suite o f pesticides used and are by far the most clearly distinguished in a chromatograph, despite their close proxim ity o f emergence from the separating column. The fact that these compounds are sorbed to a lesser extent than others by the organic material tested here is more an indication o f the relative solubility o f these compounds in water. Generally, the order in which compounds are listed on each o f Figures 4.6 - 4.12 (the order in which they emerge from the separating column) is an indication o f their relative solubility or conversely, hydrophobicity, the most soluble being simazine, the most hydrophobic being parathion. (See also values listed in Table 4.1, for exact values, and exceptions to this rough guide). The fact that such compounds (simazine, atrazine, propazine and desmetryn) are sorbed to this degree is an indication o f the
high sorptive potential o f the material tested, and they should not be overlooked because the sorption o f other compounds may be higher. Other compounds w ith in this suite would be expected to be sorbed to a far greater degree. To begin w ith, the other pesticides are more hydrophobic (and are therefore more lik e ly to sorb onto anything rather than remain in solution), and they also have higher values, another indication o f their preference fo r material less polar than water. Cyanazine is a notable exception to this trend. It has a relatively low (see Table 4.1) due to a side chain o f a cyanide group which is hydrophilic. However, despite this relative a ffin ity to remain in aqueous solution, cyanazine is still sorbed to a strong degree by blackfly silk and snail mucus, indicating that such materials are themselves hydrophobic, relative to water.
Organophosphate pesticides
Fenitrothion tends to dominate each figure (Figures 4.6 - 4.12) in the series, as do each o f the other organophosphates (malathion and parathion). This is partly because o f the different structure and thereby, chemical nature o f these compounds, and also in part because o f the higher sensitivity o f the GC detector to fenitrothion in particular. For a demonstration o f the relevance o f these results, attention must be turned to Figure 4.13, which compares in diagrammatic form , the sorption o f various materials, as well as the predicted sorption o f materials, based on published, predictive equations. It is apparent that silk demonstrates a strong a ffin ity fo r pesticides well beyond any other material either measured or postulated. Equally, snail mucus, when measured according to the wet weight o f the live snails, is comparable to measured K j and values fo r sediments and soils. When measured using the correction factor described in section 4.3.3 where the amount o f mucus (dry weight) produced per unit snail (wet weight) is used to calculate the fo r mucus, the resulting values are comparable to silk, and higher than those o f any other substance measured.
4.3.6 Are certain pesticides sorbed to a greater extent than others?
Sorption o f different pesticides onto mucus and silk are statistically indistinct. Analysis of a random subset o f values (fo r all silk samples on CN membranes)
145
found that with the exception o f parathion, pesticides did not d iffe r statistically from each other in their sorption onto the same material (Anova Fg^g = 1.94, P = 0.065). Parathion is the most hydrophobic pesticide o f the suite examined, emerging last from the GC column, and is therefore expected to sorb on to materials to a greater degree. For a random subset o f values based on mucus samples (all mucus samples fo r P. jenkinsi) no differences were found between pesticides in terms o f sorption (Anova
Fg ,70 = 1.46, P = 0.166).
These findings im ply that despite the considerable variation in hydrophobicity exhibited by the diverse suite o f pesticides chosen, and despite the considerable range o f molecular weight, chemical reactivity, side chain groups etc, sorption onto mucus and silk is uniform ly high. This suggests that it is the material examined rather than the hydrophobic nature o f the pesticides that is causing sorption to this degree, and that both snail mucus and blackfly silk may be considered to be strongly adsorptive.
4.3.7 Experimental constraints
Experimental design was dictated by the quantities o f mucus and silk generated. Attempts were therefore made to maximise sorption differences by reducing the volume o f sand on which snails crawled and also by reducing the concentration o f pesticides exposed to the mucus and sand mixture. Sim ilarly, the low masses o f silk produced by larvae reared in laboratory conditions again necessitated low pesticide concentrations in order to detect appreciable differences before, and after, exposure due to sorption. A t 5 and 10 \ig l ‘, pesticide concentrations are close to the w orking lim it o f detection. M icrogram s per litre are equivalent to nanograms per ml, GC injection samples prepared from each treatment were just 0.3 ml. For this reason, certain pesticides were more reliably detected than others. This explains the rather large standard deviation on Figures 4.6 - 4.12. That small quantities o f organic material (mucus and silk) were produced in the laboratory does not indicate that such material is lim ited in the environment. Ladle et ai, (1972) measured a density o f up to 300,000 Simuliidae larvae per square metre o f vegetation in the area from which egg masses were collected. Potamopyrgus spp. can exist at very high densities: up to
28,000 individuals per m ' o f stream bed (Arm itage pers. comm.). Lymnaea spp. are also able to exist at high densities in rivers, up to 965 (Extence, 1981). The organisms that produce the materials examined w ith in this study can occur at high densities, producing considerable quantities o f mucus and silk. However, in order to produce mucus and silk in the laboratory in a useable, fresh form , certain lim itations are imposed upon production. For successful hatching o f blackfly eggs in a glass ja r incubator, most o f the space available is required fo r aeration blocks. S im ilarly, the snails were observed to remain im m obile in a dish o f sand, unless they have sufficient space in which to move around. The practical considerations associated w ith these experiments on snails and b lackfly larvae tend to lim it production o f mucus and silk.
4.3.8 Mucus and silk as water purifiers
The examination o f pesticide attachment to sand coated in snail mucus is analogous to the examination o f slow sand filte r beds as agents o f pesticide sorption. Slow sand filte r beds are used by the water industry as a purification stage. Woudneh et al., (1996) reported the removal o f up to 17% o f pesticides at optim um conditions using slow sand filte r beds. Although they did not define the means o f removal, they presented their results as a means o f bioremediation. By not defining the means o f removal it can only be assumed that the pesticides were either p h o tolytically degraded and their degradation products not detected, or that the pesticides were sorbed by schmutzdecke (the layer o f particles and organisms that forms on a filte r bed). Significantly, they found no removal o f atrazine at all. I f these results are compared with up to 54% removal by snail mucus and up to 98% removal by silk (see Table 4.10), the a bility o f the organic material (examined in this study) to sequester xenobiotics is apparent.
Table 4.10. Percentage sorption o f pesticides by blackfly silk and pedal mucus from two species o f freshwater snail.
Pesticide Simazine Atrazine Propazine Desmetryn Prometryn Terbutryn Fenitrothion Malathion Cyanazine Parathion
"B la c k fly silk 25 2 0 .6 15.3 2 1 .2 32.8 58.0 94.9 88.0 16.9 96.9
^ P .Jenkinsi pedal mucus 1.8 3.3 2 .6 54.1 50.3 0.1 2.7 — 0.1 9.3
' L p e r e g r a pedal mucus 1.0 1.9 1.3 14.9 15.9 - - 0.1 0.3 7.1
“ Silk samples numbered 29, over the range 0.1 - 4.9 mg, exposed to 25 ml o f pesticide suspension o f either 5 or 10 pg P. je n k in s i mucus samples were produced using masses o f snails weighing 2.4 - 25 g.
,1
Eighteen samples were exposed to 25 ml o f pesticide suspension o f either 5 or 10 pg 1 ' L p e re g ra mucus samples were produced using masses o f snails weighing 0.7 - 5.0 g. Sixteen samples were exposed to 25 ml o f pesticide suspension o f either 5 or 10 pg l '
4.3.9 Placing blackfly silk and snail mucus in the context o f other materials studied.
W orking w ith pyrethroids which have exceptionally high values due to their extremely hydrophobic nature, Zhou et ai, (1995 c) found values o f 120,000 to 770,000 on a variety o f organics coating clay minerals, most notably aromatic containing polyelectrolytes such as humic acids. Zhou et al., (1995 d) presented a table o f organic compounds most likely to interact w ith pyrethroids: humic acids, fu lv ic acids, hydrophilic macromolecular acids, and then natural coatings on estuarine suspended particles. It is unfortunate that they lim ite d themselves to pyrethroids which are expected to have exceptionally high values anyway, and that there is no overlap with the suite o f compounds used in this study, to place macro in vertebrate organics such as silk and mucus into such a table. T heir papers do illustrate however, the great potential o f organic materials to sorb pesticides.
Some reviews state that pesticides are harmless, and that examination o f sorption onto materials other than sediments is futile. H ill (1989), stated that pyrethroids, despite very high LC^q values, would have no effect on freshwater ecosystems, and that such pesticides were in fact transitory and harmless. The basis fo r this claim is that, as pyrethroids have such high K j values, they w ill sorb onto almost any material, and it is possible that they may not undergo desorption. However, soils and sediments are not stable, non-dynamic media. K e ilty et al., (1988) demonstrated that oligochaetes move pesticide contaminated sediments upwards, disturbing clean surface sediments and refuting the concept o f bioremediation o f contaminated water bodies by adding additional sediment to encase the pesticide (Isensee, 1983). The sorption o f almost all pesticide compounds is pH dependent, due to the attachment o f hetero-atoms (atoms other than carbon or hydrogen). The pH o f river water is unlike ly to present a sufficient pH range to have any effect on sorption or desorption. The pH o f some invertebrate guts however, does present a pH range sufficiently broad as to affect the sorption o f pesticides on D O M (W otton 1996).
149
by pyrethroid contamination. Furthermore, pesticides may be transferred to local predators via the food chain. M ost important o f all, is the consideration that organic material produced by macroinvertebrates (such as mucus and silk) which has a strong a ffin ity fo r pesticides (far more so than soils and sediments), is also a food source for other organisms. It would seem therefore that i f pesticides enter a freshwater system, they have the opportunities to attach and persist. The problem up to now has been in not identifying the correct component o f sediments or soils to sample. A problem remains in trying to sample the appropriate fraction o f sediments, the organic coatings around minerals that sorb pesticides.
4.3.10 The importance o f accounting fo r pesticide a ffin ity to biological
material
The last thirty years have seen many advances in pesticide technology, in their manufacture, detection and modelling o f their persistence. However, pesticide technology has failed in form ing cross-links between related disciplines. Clark et al., (1991), while examining pesticides in water from a chalk catchment area, pointed out that, historically, pesticides have been used fo r far longer than they have been monitored. In order to understand the behaviour o f such contaminants, far greater information about aquifers is required. For example. Gomme et al., (1991), investigating the same catchment, noticed that factors such as season are important. In spring-fed streams, summer flo w w ill be low and winter flo w w ill be fast, w hile aerial deposition and land ru n -o ff w ill be a series o f discrete events. A new research front (Spark and S w ift, 1994) is pesticide interactions w ith humic acids. However, examination o f materials other than sediments and soils or sub-lethal responses by individual species has been limited.
In measuring sorption onto sediment, K a ric k h o ff et ai, (1979), stated that values can generally be predicted from values using the predictive equations already available, since Kenanga and Goring (1980) had shown that there is essentially little difference between them. As far as sediments and soils are concerned they are largely correct. There seems to be little discernible difference between soil and sediment
types, with regard to sorption o f individual pesticides (see Table 4.11). However, that is only the case if the minerals are washed. I f the minerals are coated in organic material, then considerable variations are detected. Hence the importance o f isolating and defining sorption o f individual organic materials (Zhou et al., 1995 a,b,c,d). In attempting to predict pesticide sorption, desorption and persistence in rivers and streams, it is no longer sufficient to examine pesticide sorption onto soils or bed sediments in isolation. Biological material exists w ith in freshwater systems, which is capable o f sorbing xenobiotics to a high degree. Indeed, many biological materials such as mucopolysaccharide, silk and b io film communities, due to their chemical structure and physical design, are particularly suited to sorbing micro-organic compounds such as pesticides (Bishop et al., 1995). Furthermore, not only are these materials ubiquitous and often present in large quantities over almost every surface (mucus and b io film exist on sediment, stones, roots etc), but they are also an essential food source fo r macroinvertebrates, and ultim ately all consumers w ith in freshwater systems.
These results show that organic materials such as snail mucus and silk have a high a ffin ity for a range o f pesticides, and by sorbing pesticides from the water column they provide access fo r pesticides into the food chain.
51
T a b le 4.11 C o m p a ris o n o f d is trib u tio n c o e ffic ie n ts fo r b la c k fly la rv a l s ilk w ith o th e r d is trib u tio n c o e ffic ie n ts fo u n d in the lite ra tu re fo r the p esticides used in th is stu dy ( m l m g ')
R e f e r e n c e a b b c d e f g Pesticide K„ K.... K.; K . K,. K K,. Simazine 135 135 1.9x10" 130 Atrazine - - 550/3000 148 149 1.7x10" 100 Propazine - - - 158 160 1.6x10" 154 Desmetryn - - - - 1.7x10" - Prometryn - - - - 810 2.4x10" 400 Terbutryn - - 700 4.9x10" 2000 Fenitrothion - - - 13x10" 2000 Malathion 0 57 - 0.72 - - - 1778 24x10" 1800 Cyanazine - - 200 200 7.1x10" 190 Parathion 0 29-0.39 0.78 ■64.9 351 - 972 - 6457 4800 31x10" 5000
a Sujatha and Chacko (1992) on three unwashed estuarine sediments,
b Get S t I and M engelgrin (1984) on 11 unwashed soils and 4 unwashed lake sediments c Bottero ei u l (1994) on zeolite and organoclay respectively,
d K arickho ff (1981) on 17 soils and sediments from across the USA
e Kcnaga and G oring (1980) A general review o f all available predictive equations,
f This study, distribution coefficients for blackfly silk