primarily led from a human health perspective with a need to assess the risks from
consuming edible food crops contaminated with pharmaceutical residues [57].
Some of the earliest published research, from mid-2000 onwards, started by
assessing the uptake and accumulation of pharmaceuticals in plants, from soils
spiked directly with chemical residues. Boxall et al. [57] provided some of the first
experimental evidence that veterinary medicines persist in soils and can accumulate
in carrot (Daucus carota) roots (tubers) and lettuce leaves. Results suggested that a
combination of chemical properties and the crop species influenced the degree of
uptake, with only two of the ten target analytes detected both in the carrot and the
lettuce leaves (florfenicol and trimethoprim) with levamisole also observed in lettuce
leaves and diazinon and enrofloxacin in the carrot roots. The lack of uptake of some
pharmaceuticals can be well explained by their degradation in soils, with >90% of
dissipation of amoxicillin, sulfadiazine and tylosin observed by the time the lettuce
plants were harvested.
Building on the results from the Boxall et al. [57] study, Carter et al. [5] carried
out a series experiments to elucidate relationships between the fate of pharmaceuticals in soils and potential uptake by plants. The consistently high carbamazepine
uptake into both radish (Raphanus sativus) and ryegrass (Lolium perenne) (<52 μg/
g, dry weight) was suggested to result from a combination of persistence in the soil
(DT50 > 40 days), a high degree of bioavailability in soil pore water and being a
moderately hydrophobic (log K ow 2.25) and unionised compound. Meanwhile, when
the pharmaceutical demonstrated fast dissipation in the soil (e.g. sulfamethazine),
this results in diminishing concentrations in the soil matrix and thus smaller fractions
are available for uptake. Carbamazepine is a neutral compound with a K ow value
similar to where maximum uptake of neutral organics is observed according to the
Gaussian distribution proposed by Briggs et al. [58]. Similar results were recently
published by Li et al. [28] who also observed that out of a suite of 15 pharmaceuticals
spiked into a sandy loam soil, carbamazepine accumulated to the greatest extent in
radish leaves and roots, which was up to 738 times greater than the accumulation of
the least accumulated compound estrone in the roots. Carbamazepine was weakly
sorbed to soil, as well as being highly persistent, making it highly favourable for
plant uptake.
Comparatively, Carter et al. [5] observed more hydrophobic pharmaceuticals
(e.g. fluoxetine and diclofenac) accumulated to a greater extent in the roots with
low translocation capacity to aerial plant organs. As these chemicals were extensively ionised at test soil pH, relationships between log D ow and accumulation in the
plants revealed a general increase in log D ow corresponded to an increase in plant
uptake factors. Meanwhile, [28] demonstrated, with a strong positive correlation (R
2
0.94), that log D ow was a good predictor of plant bioconcentration from pore water
(BCF pore water ) for non-ionised pharmaceuticals, with similar relationships not evident for ionised pharmaceuticals. Interestingly, the relationship with log D ow for
non-ionised pharmaceuticals could not be replicated for bioconcentration factors
(BCFs) based on bulk soil concentrations (BCF soil ) implying that pharmaceuticals
present in soil pore water represent the major bioavailable fractions for plant uptake.
Li et al. [28] concluded that BCFs calculated on the basis of pharmaceutical
188
L. J. Carter et al.
consuming edible food crops contaminated with pharmaceutical residues [57].
Some of the earliest published research, from mid-2000 onwards, started by
assessing the uptake and accumulation of pharmaceuticals in plants, from soils
spiked directly with chemical residues. Boxall et al. [57] provided some of the first
experimental evidence that veterinary medicines persist in soils and can accumulate
in carrot (Daucus carota) roots (tubers) and lettuce leaves. Results suggested that a
combination of chemical properties and the crop species influenced the degree of
uptake, with only two of the ten target analytes detected both in the carrot and the
lettuce leaves (florfenicol and trimethoprim) with levamisole also observed in lettuce
leaves and diazinon and enrofloxacin in the carrot roots. The lack of uptake of some
pharmaceuticals can be well explained by their degradation in soils, with >90% of
dissipation of amoxicillin, sulfadiazine and tylosin observed by the time the lettuce
plants were harvested.
Building on the results from the Boxall et al. [57] study, Carter et al. [5] carried
out a series experiments to elucidate relationships between the fate of pharmaceuticals in soils and potential uptake by plants. The consistently high carbamazepine
uptake into both radish (Raphanus sativus) and ryegrass (Lolium perenne) (<52 μg/
g, dry weight) was suggested to result from a combination of persistence in the soil
(DT50 > 40 days), a high degree of bioavailability in soil pore water and being a
moderately hydrophobic (log K ow 2.25) and unionised compound. Meanwhile, when
the pharmaceutical demonstrated fast dissipation in the soil (e.g. sulfamethazine),
this results in diminishing concentrations in the soil matrix and thus smaller fractions
are available for uptake. Carbamazepine is a neutral compound with a K ow value
similar to where maximum uptake of neutral organics is observed according to the
Gaussian distribution proposed by Briggs et al. [58]. Similar results were recently
published by Li et al. [28] who also observed that out of a suite of 15 pharmaceuticals
spiked into a sandy loam soil, carbamazepine accumulated to the greatest extent in
radish leaves and roots, which was up to 738 times greater than the accumulation of
the least accumulated compound estrone in the roots. Carbamazepine was weakly
sorbed to soil, as well as being highly persistent, making it highly favourable for
plant uptake.
Comparatively, Carter et al. [5] observed more hydrophobic pharmaceuticals
(e.g. fluoxetine and diclofenac) accumulated to a greater extent in the roots with
low translocation capacity to aerial plant organs. As these chemicals were extensively ionised at test soil pH, relationships between log D ow and accumulation in the
plants revealed a general increase in log D ow corresponded to an increase in plant
uptake factors. Meanwhile, [28] demonstrated, with a strong positive correlation (R
2
0.94), that log D ow was a good predictor of plant bioconcentration from pore water
(BCF pore water ) for non-ionised pharmaceuticals, with similar relationships not evident for ionised pharmaceuticals. Interestingly, the relationship with log D ow for
non-ionised pharmaceuticals could not be replicated for bioconcentration factors
(BCFs) based on bulk soil concentrations (BCF soil ) implying that pharmaceuticals
present in soil pore water represent the major bioavailable fractions for plant uptake.
Li et al. [28] concluded that BCFs calculated on the basis of pharmaceutical
188
L. J. Carter et al.