Avian Urine: Its Potential as a Non-Invasive Biomonitor of Environmental Metal Exposure in Birds
- 253 Downloads
Current non-invasive biomonitoring techniques to measure heavy metal exposure in free ranging birds using eggs, feathers and guano are problematic because essential metals copper (Cu) and zinc (Zn) deposited in eggs and feathers are under physiological control, feathers accumulate metals from surface contamination and guano may contain faecal metals of mixed bioavailability. This paper reports a new technique of measuring lead (Pb), Cu and Zn in avian urate spheres (AUS), the solid component of avian urine. These metal levels in AUS (theoretically representing the level of metal taken into the bloodstream, i.e. bioavailable to birds) were compared with levels in eggs (yolk and shell), feathers and whole guano from chickens (Gallus gallus domesticus) exposed to a heavy metal-contaminated soil (an allotment soil containing Pb 555 mg kg−1 dry mass (dm), Cu 273 mg kg−1 dm and Zn 827 mg kg−1 dm). The median metal levels (n = 2) in AUS from chickens exposed to this contaminated soil were Pb 208 μg g−1 uric acid, Cu 66 μg g−1 uric acid and Zn: 526 μg g−1 uric acid. Lead concentrations in egg yolk and shell samples (n = 3) were below the limit of detection (<2 mg kg−1), while Cu and Zn were only consistently detected in the yolk, with median values of 3 and 70 mg kg−1 (dm), respectively, restricting the usefulness of eggs as a biomonitor. Feathers (n = 4) had median Pb, Cu and Zn levels respectively of 15, 10 and 140 mg kg−1 (dm), while whole guano samples (n = 6) were 140, 70 and 230 mg kg−1 (dm). Control samples were collected from another chicken flock; however, because they had no access to soil and their diet was significantly higher in Cu and Zn, no meaningful comparison was possible. Six months after site remediation, by top soil replacement, the exposed chickens had median Pb, Cu and Zn levels respectively in whole guano (n = 6) of 30, 20 and 103 mg kg−1 (dm) and in AUS (n = 4) of 147, 16 and 85 μg g−1 uric acid. We suggest the persistent high Pb level in AUS was a consequence of bone mobilised for egg production, releasing chronically sequestered Pb deposits into the bloodstream. In contrast, AUS levels of Cu and Zn (metals under homeostatic control and sparingly stored) had declined, reflecting the lower current exposure. However because pre- and post-remediation samples were measured using different methods carried out at different laboratories, such comparisons should be guarded. The present study showed that metals can be measured in AUS, but no assessment could be made of availability or uptake to the birds because tissue and blood samples were not concomitantly analysed. A major short coming of the study was the inappropriate control group, having no access to uncontaminated soil and being fed a different diet to the exposed birds. Furthermore guano and urine analysis should have been carried out on samples from individual birds, so biological (rather than just technical) variation of metal levels could have been determined. Future studies into using AUS for biomonitoring environmental heavy metals must resolve such experimental design issues.
KeywordsAvian urine Biomonitoring Heavy metals Chickens Allotment soil
Metals are natural substances, but their increased availability through mining and smelting, with subsequent release into the environment from numerous anthropogenic sources, makes them a serious threat as persistent toxic pollutants (Walker et al. 2001). Metals like copper (Cu) and zinc (Zn) are essential trace elements but are toxic at higher concentrations, others, e.g. lead (Pb) and cadmium (Cd) have no known biological function and can be toxic at low levels. High levels of Cu and Zn are commonly found with other metals in mine waste, and their toxic impact on fauna including wild birds is illustrated by the 1998 mine waste spill in Spain’s Donana National Park (Gomez et al. 2004). Lead poisoning of free ranging birds following incidental uptake from spent fishing weights and gunshot is also well recognised (Scheuhammer and Norris 1996). This widespread use of metals in the world makes it necessary to monitor their levels in the environment to safeguard biological systems. Typically, metals accumulate in soil and sediment where they can exist in many states, which determine their bioavailability (Ruby 2004); for this reason a measure of their biological uptake through biomonitoring techniques is more relevant than total environmental levels (Peakall and Burger 2003). Following certain guidelines (Hollamby et al. 2006), birds are commonly used for biomonitoring environmental heavy metals (Furness 1993), with most tests involving invasive or destructive sampling techniques (Swaileh and Sansur 2006). The advantages of using non-invasive strategies in biomonitoring programmes (Fossi 1994) has led to the increasing use of eggs (Burger and Gochfeld 1993), feathers (Burger et al. 1992) and guano (Fitzner et al. 1995), to biomonitor avian exposure to environmental heavy metals. In order for such materials to be valid metal biomonitors, they need to correlate with blood levels (Hollamby et al. 2006), which represent the birds’ current bioavailable intake (Furness 1993). Although the production of eggs from wild birds is restricted to the laying season and feathers to the time of moult, whole guano has the advantage of being continuously produced and easy to collect but does not necessarily relate to blood levels.
Guano is a mixture of faecal and urinary excretions resulting from the digestive and urinary systems sharing a common outlet in birds called the cloaca. The faecal component of guano contains a variable mixture of bile excreted (representing bioavailable) and unabsorbed heavy metals simply transiting the digestive system (Mohanna and Nys 1998). The white, urine component is derived from the blood passing through the kidneys and so its heavy metal content is linked to the birds’ current ‘bioavailable’ intake (Furness 1993), i.e. the metal has been taken up into the blood stream from the environment.
2 Materials and Methods
To enable us to measure urine heavy metal levels, it was necessary to devise a method to separate the urine component from whole guano (see Section 2.3). To determine the potential of the avian urine-based technique for measuring metal bioavailability in contaminated soil, we were able to use a local city allotment (just prior to its remediation by topsoil replacement), on which laying female chickens were kept and had a soil known to be contaminated with Cu, Pb and Zn from incinerator bottom ash (Pless-Mulloli et al. 2004). These chickens provided us with an ideal sentinel species (Peakall and Burger 2003) for biomonitoring the soil heavy metal contamination before and 6 months after site remediation. As the chickens had access to the contaminated soil and ingested soil may typically constitute 10 % of the diet in such birds (Stephens et al. 1995; Beyer et al. 1994), we expected them to have elevated body metal levels before remediation.
Metal levels in urine from laying female chickens on a similar control allotment site where no soil exposure was allowed were compared with urine from birds on the contaminated allotment. We also compared the proposed urine-based method with currently used non-invasive biomonitoring techniques (egg, feather and whole guano analysis) for heavy metal exposure in birds.
2.2 Sample Collection
2.2.1 Site Metal Levels and Soil Sampling
Soil samples were collected for metal analysis in order to confirm the allotment contamination status previously reported as having geometric mean soil levels for Cu, Pb and Zn of 166, 674 and 823 mg kg−1 (dry mass), respectively (Hartley et al. 2004). Six surface horizon (0–10 cm) soil samples were taken from a chicken pen on the contaminated allotment to determine total soil heavy metal levels available to the chickens. The soil samples were kept refrigerated (0–4°C) in sealed plastic bags until processed (see Section 2.5). It was not necessary to take soil samples from the control allotment because the birds here were kept entirely on wood shavings with no access to soil. Six months after site remediation, soil samples were taken from the rebuilt chicken pen in which the birds previously exposed to contaminated soil were kept.
2.2.2 Biomonitor Sampling
On the contaminated allotment, prior to remediation, the chickens were kept in a group of 20 birds confined to a caged enclosure (approximately 6 m2) with a floor of exposed contaminated soil, where they were fed. The birds had an adjoining separate night roost consisting of a wooden-floored shed with perches and nest boxes. The diet of these birds was a mixture of a commercial brand feed for laying chickens and whole grains fed ad lib, domestic supply tap water in buckets and occasional kitchen scraps (vegetable peelings and stale bread).
On the control allotment, a single group of 35 chickens were housed in a purpose-built shed having a concrete floor covered with regularly replenished wood shavings. The birds were fed exclusively on a commercial laying ration ad lib, and water was provided from a domestic supply. Birds were not exposed to soil on this allotment. Six representative samples of the rations as fed (excluding kitchen scraps) were collected from each allotment for metal analysis; the exact same ration was fed before and after remediation to the chickens on the contaminated allotment.
Sample Collection for Biomonitoring
From the night roost and nesting boxes on both the contaminated and control allotments, several kilograms of guano, a few freshly laid eggs (three contaminated and six controls) and numerous feathers were collected; following remediation of the contaminated site just guano was collected for analysis.
2.2.3 Sample Preparation
The eggs were thoroughly washed using warm tap water and a nylon brush, rinsed with 18 MΩ deionised water and then dried prior to being stored frozen (−20°C). To prepare the eggs for metal analysis, the shell was peeled from the frozen eggs and internally adherent albumin washed off with18 MΩ deionised water. The outer albumin was discarded, this being facilitated by it thawing quicker than the yolk, which was retained for analysis. The albumin was not analysed because of its reported low affinity for Pb and Cu in chickens exposed to elevated intake levels (Flores and Martins 1997; Skrivan et al. 2006).
Clean, intact wing primary feathers were collected and grouped by their colour into four samples corresponding to the chicken breeds on the contaminated allotment (Rhode Island Red, Maran, White Leghorn and Black Rock). The feathers from the control allotment had only three different colours so provided just three sample groups for analysis. Feathers were first washed with warm tap water, then thoroughly rinsed with 18 MΩ deionised water and dried before storage at room temperature in sealed plastic bags. This simple cleaning method was chosen because despite repeated acetone and water washes, favoured by many authors (Burger et al. 1992), feathers still appear to retain the heavy metals accumulated from external contamination (Dauwe et al. 2003). Furthermore washing feathers with distilled water was recommended over using acetone in a study correlating feather with muscle levels of organohalogen pollutants (Jaspers et al. 2011).
Only fresh, whole guano pellets were selected with any adherent feathers or bedding material removed prior to storing frozen (−20°C) in sealed plastic bags. Individual pellets were analysed for each whole guano metal determination. But urine analysis required more guano and typically used between 20 and 30 pellets per metal determination.
2.3 Urine Extraction
Avian urine is composed of a colloidal suspension of discrete spherical urates (Fig. 1) measuring 0.5 to 10 μm (Casotti and Braun 2004); these spheres are insoluble in ethanol or acetone but are disrupted in aqueous solutions (Drees and Manu 1996). Our extraction technique depended on the principle of differential sedimentation; the small size of the urate spheres allowing them to remain suspended in the ethanol for longer than larger or denser particulates including soil from the faecal component.
To extract the urate spheres from whole guano samples, an approximately 200 g representative sample of frozen guano was defrosted at room temperature with 300 mL absolute ethanol (general purpose grade) in a glass beaker. On breaking the guano up with a glass rod, the white urinary component readily formed a suspension, any gross floating faecal contaminants being removed at this time. To avoid the transfer of denser faecal material including soil particulates, only the upper portion of the supernatant was decanted into a 50-mL glass test tube. The supernatant was allowed to settle for 5 min, after which the top 20–30 mL portion (representing the extracted urine sample) was decanted into a 50-mL polypropylene centrifuge tube (Fisherbrand®). The residue from the glass test tube was returned to the beaker and the process repeated four to five times until no further white urine could be extracted. The extracted urine sample was centrifuged at 2,000 × g for 2 min; the discoloured ethanol discarded and the solid urine washed twice with approximately 40 mL of fresh ethanol by vortexing and again centrifuging (2,000 × g for 2 min).
Qualitative purity of the solid urine extract was determined by examining a small representative fraction under a medium power light microscope; an adequately pure sample consisting almost entirely of characteristic urate spheres. From the finding of Braun and Pacelli (1991), later confirmed by the more extensive work of Casotti and Braun (2004) that urate spheres are consistently 65 % uric acid by dry mass, we determined the sample uric acid content representing a quantitative measure of the extracted urine purity.
2.4 Uric acid Analysis of Extracted Avian Urine
Uric acid analysis was carried out using a modified method (Van Handel 1975; Adeola and Rogler 1994). In brief, an accurately weighed 30–40 mg representative fraction of the dry extracted urine sample was digested in 10 mL of 0.5 % (w/v) LiCO3 solution in a boiling water bath for 10 min. The resulting solution was made up to 100 mL with 18 MΩ deionised water, of which a well-mixed representative 5 mL sample was filtered using a 0.45-μm pore size syringe filter (Whatman® PuradiscTM 25 AS). One hundred microlitres of this filtrate was diluted to 1 mL, added directly to a cuvette and mixed with an equal volume (1 mL) of the freshly made up chromogen reagent (Van Handel 1975). Although the original method used a 5-min end point, we found that 10 min was required, at which time the absorbance (450 nm) of the yellow product was measured using a Biochrom Libra S12 UV/Visible spectrophotometer (Biochrom Ltd, Cambridge, UK). Calibration was achieved using water blanks and uric acid standard solutions made up in 0.5 % (w/v) LiCO3. Specificity to uric acid was determined by overnight incubation at room temperature of sample duplicates with approx 0.5 U/mL uricase (Sigma), which resulted in equivalent to blank readings.
2.5 Sample Analysis
All soil, chicken feed, egg, feather, whole guano and extracted urine samples were oven dried to constant mass at 65°C, then ground and sieved to homogenise prior to taking a representative sample for heavy metal analysis. Pb, Cu and Zn levels for all samples other than chicken feed and those collected following remediation were determined by a UKAS (United Kingdom Accreditation Service) accredited laboratory using inductively coupled plasma mass spectrometry after a 2-h acid reflux pre-digestion. Levels of detection were poorer for some of the smaller samples but generally were between 2 and 8 mg kg−1 for Cu and Pb although >0.8 mg kg−1 for Zn on a dry mass basis. The high reactivity on acid pre-digestion of the extracted urine resulted in two samples from the chickens kept on the contaminated soil being lost. Metal analysis of chicken feed (from both allotments), whole guano, extracted urine and soil (from the remediated allotment) was performed using a Unicam 701 ICP-OES (Unicam Instruments, Cambridge, England) after acid reflux using a standard method (EPA 1991). Calibration standards (Sigma) were freshly made up for each metal, having limits of detection in a complicated matrix solution of 0.1 mg L−1 for Pb and 0.01 mg L−1 for Cu and Zn. The metal levels in the two acids were determined by digesting acid blanks in the same manner as the samples; this value being subtracted from the sample results before calculating metal values. The resulting limits of detection in dry samples for Pb, Cu and Zn were 5, 0.5 and 0.5 mg kg−1, respectively. All metal concentrations are expressed as milligrams per kilogram on a dry mass basis, with urine levels as micrograms per gram uric acid, allowing for sample comparison to compensate for possible variations in extraction purity.
Representative samples of the dried, ground and sieved (2 mm) soil; were analysed for pH and soil organic matter (SOM) using a modified method (Clark et al. 2006) because these properties have a strong influence on the bioavailability of soil heavy metals. The pHw was determined after reacting soil for 1 h 1:1 with 18 MΩ deionised water (5 g/5 mL), centrifuged (2,000 × g for 2 min) and the pHw of the supernatant measured using a glass electrode (Denver Instruments) pre-calibrated with pH 4, 7 and 10 standards. The SOM content was determined on accurately weighed 5 g soil samples by mass reduction from overnight ignition in a 450°C furnace.
2.6 EDAX Analysis of Extracted Urate Spheres
The opportunity arose to use the facilities of the environmental electron microscope in the Advanced Chemicals & Materials Analysis Service at Newcastle University. This unit can analyse small particulates similar to the urate spheres for their elemental constituents using energy dispersive X-ray micro-analysis (EDAX) (Vesk and Byrne 1999). Furthermore this method of analysis has previously been used to identify the elemental composition of urate spheres from birds, including domestic chickens (Casotti and Braun 1997, 2004).
2.7 Data Analysis
Statistical analyses were not performed on the whole guano and avian urate spheres (AUS) data because bulked samples had been used, giving effectively n = 1; consequently sampling would reflect technical rather than biological variability. Data from the small number of individual egg and feather samples were analysed using the non-parametric Mann–Whitney U test. For these, sample results are reported as median and range values. However the larger number of soil and feed samples (being normally distributed) allowed parametric analysis using the Student’s t test, with significant differences (p < 0.05) set at a 95 % confidence interval. Where samples were below the limit of detection (LOD), a value of half the LOD was used in data analysis (Nicholson et al. 1999; Dauwe et al. 2005).
3.1 Soil Analysis
Top soil pH, percentage soil organic matter (% SOM), lead (Pb), copper (Cu) and zinc (Zn) from the chicken pen on the contaminated allotment before and after remediation by soil replacement, compared with previous data for the allotment soil and mean background values for soils in England and Wales reported in the literature
Chicken pen soil
7.30 ± 0.30
33.5 ± 10.2
555 ± 301
273 ± 59
827 ± 241
7.31 ± 0.03
12.3 ± 0.5
58.5 ± 10.8
15.1 ± 1.1
57.7 ± 2.3
Soil concentrations previous studya (n = 12)
7.3 ± 0.2
674 ± 286
166 ± 76
823 ± 194
Mean background soil concentrationsb
Both the contaminated and replacement soils had neutral pH values (7.3 ± 0.3 and 7.31 ± 0.03 respectively) confirming earlier reported values (Hartley et al. 2004). The higher SOM value of the contaminated soil (33.5 ± 10.2 %) in comparison to the typical range of 7–13 % reported for allotment soils in the literature (Clark et al. 2006) accounted for its dark colour and visible residues of chicken manure. The replacement soil was lighter coloured, appearing more clay-like and had a lower SOM value (12.3 ± 0.5 %) despite visible feed and guano content.
3.2 Chicken Feed Analysis
Lead (Pb), copper (Cu) and zinc (Zn) concentrations (in milligrams per kilogram dm median and range) in chicken feed as fed on the contaminated (pre- and post-remediation) and control allotments, with comparison for the reported range of values in home mixed feed fed to laying chickens in England and Wales
Reported range of valuesa
3.3 Metal Levels in Eggs
Pb values in all egg samples were below the LOD of 2 mg kg−1 dry mass. Cu and Zn levels in the egg yolks were not significantly different (p > 0.05) between the contaminated allotment and control site. Cu levels in egg shell samples were below the LOD of 2 mg kg−1 dry mass in all control eggs and two of the three contaminated samples making comparison impossible. The Zn level in one shell sample from the contaminated site chickens was 7 mg kg−1 with the others below the detection limit of 4 mg kg−1, suggesting elevation over the control site egg shell values that were all below the lower detection limit of 2 mg kg−1.
3.4 Metal Levels in Feathers
Lead (Pb), copper (Cu) and zinc (Zn) concentrations in milligrams per kilogram as median (and range) in feathers from chickens on the contaminated and control allotment
Control site (n = 3)
Contaminated site (n = 4)
3.5 Metal Levels in Whole Guano
3.6 Avian Urine Analysis
3.6.1 Purity of Sample
Representative samples of extracted AUS were examined by light microscopy and all appeared to be free of contamination from soil or faecal material. The chemical analysis of each sample of extracted AUS ranged from 54 to 60 % uric acid by dry mass, representing a purity of 83 to 92 %, respectively.
3.6.2 Avian Urine Metal Levels
4.1 Metal Levels and Properties of Contaminated Soil
The lead concentration in the pre-remediation chicken pen soil (Table 1) was above the soil guideline values of 450 mg Pb kg−1 dry mass, confirming this allotment’s contaminated land status (DEFRA 2000). Also Cu and Zn exceeded the now withdrawn Inter-departmental Committee on the Redevelopment of Contaminated Land intervention concentrations of 130 and 300 mg kg−1 dry mass, respectively. These elevated concentrations of Pb, Cu and Zn in the soil samples, although characteristic of the incinerator bottom ash added to the site (reported as 760, 870 and 1,100 mg kg−1 dry mass, respectively, Pless-Mulloli et al. 2004), may also have originated from other sources. These include the use of agro-chemicals on the gardens (Rimmer et al. 2006) and burning rubbish especially plastics (Meharg and French 1995).
Chicken manure is likely to be responsible for the chicken pen SOM values being higher than the 7–13 % reported for allotment garden soils (Clark et al. 2006). This was backed up by the later analysis of chicken guano using the same technique giving an organic matter content of 76 %. Although the soil metal load may also be derived from chicken manure input, due to high Cu and Zn inclusion in commercial diets (Mohanna and Nys 1998), we showed on the analysis that the feed of these birds was low in these metals, discounting this as a major source of soil metal contamination. From my findings, the combination of neutral pH and high SOM of these samples would be expected to reduce metal bioavailability in the contaminated chicken pen soil (Clark et al. 2006).
4.2 Metal Levels in Chicken Feed
Chickens on the control allotment were fed rations high in added Cu and Zn with concentrations being comparable to literature values (Nicholson et al. 1999). However birds on the contaminated site before and after remediation were fed a ration significantly lower in Cu and Zn. As the control birds had no access to soil, the feed was the major source of metal uptake, in contrast to the birds on the contaminated site in which soil was the major metal exposure route. The Pb concentrations in both rations were low and would not be expected to significantly contribute to the uptake of this metal by chickens. A consequence of the higher concentrations of Cu and Zn in the feed of control site chickens compared to the contaminated soil exposed birds prevented a meaningful comparison of these metals in biomonitor samples. However, post-remediation when the soil contribution to uptake of these metals was drastically reduced in the birds on the contaminated site, the different feed concentrations (Table 2) were reflected in the guano samples (Fig. 2).
4.3 Biomonitor Samples
4.3.1 Metal Levels in Eggs
In the current study (bearing in mind the LOD of 2 mg kg−1), egg yolk and shell were unsuitable materials to biomonitor increased Pb exposure from the contaminated soil, confirming that lead has a low transfer to eggs (Walsh 1990). However the sensitivity of the analysis method used here may have compromised my results and should not preclude eggs from being a valid material for biomonitoring. Pb concentrations in egg yolks and shells from Pb-exposed chickens (from ingested Pb-based paint chips in their environment) were 0.4 and 0.45 mg kg−1, respectively, and significantly above concentrations in eggs from the control birds (Trampel et al. 2003). In addition Mazliah et al. (1989) reported eggshells from Pb-dosed hens had 6–12 times the Pb concentration of eggshells from controls, while the Pb content of the egg yolks from dosed hens was significantly higher than the controls. In a study on environmental uptake of heavy metals by house sparrows (Passer domesticus), there was significant correlation between Pb, Cu and Zn in egg shell and egg content (Swaileh and Sansur 2006), making the shell a valuable biomonitoring matrix for these metals. However Grand et al. (2002) cast doubt on the value of eggs for biomonitoring Pb exposure in birds, reporting no correlation between blood and egg Pb concentrations in two species of wild duck.
Skrivan et al. (2006) raised the dietary Cu intake of laying chickens increasing Cu concentrations in egg yolk and shell from 3.5 to 5.0 mg kg−1 and 2.0 to 2.5 mg kg−1 (dry mass), respectively. However these small increases required more than a tenfold elevation in dietary Cu intake, further emphasising that eggs are an insensitive monitor for this metal. In a study on concentrations of heavy metals in laying great tits (Parus major) and their eggs, Dauwe et al. (2005) hypothesised that egg content or shell were unsuitable as a measure of exposure because Cu and Zn are under physiological control. These limitations on eggs as biomonitors for Cu and Zn exposure are reflected in egg yolks in this study but the egg shell values are too few to draw any conclusion. In the light of these findings, no eggs were collected from the chickens after remediation for metal analysis.
As the eggs from the contaminated allotment were used for human consumption, an estimate of dietary Pb exposure was calculated in a manner similar to Trampel et al. (2003). By assuming Pb concentration in the egg yolks, typically 8–9 g dry mass had reached the LOD (2 mg kg−1 dry mass), a 60-kg person would require a daily intake of greater than 12 eggs to exceed the provisional tolerable weekly intake of 0.025 mg kg−1 body mass (JFWEC 1999).
4.3.2 Metal Levels in Feathers
While feathers reflected the elevated Pb exposure on the contaminated allotment compared with the control site, Zn concentrations were no different and Cu concentrations were significantly lower in feathers collected from the contaminated site compared to those from the control. Again the results obtained in relation to essential metals Cu and Zn could be explained by their homeostatic regulation in birds (Dauwe et al. 2003) and it is known that the internal deposition of heavy metals in feathers is only a fraction of the total body burden, with the exception of mercury (Jaspers et al. 2004) and organo-tin (Kannan and Falandysz 1997). The surface affinity of feathers to bind heavy metals is shown by their use in wastewater cleanup (Al-Asheh et al. 2003) and how cleaning techniques can add to their metal content (Hogstad et al. 2003). Pb is recognised as principally a surface contaminant in feathers (Nam et al. 2004), so the higher concentration of Pb we report here may be from contaminated soil or guano accumulation on their surface. However similar differences from surface contamination with soil-derived Cu and Zn could be masked by feed or guano contamination from their typically high concentrations in commercial chicken feed (Mohanna and Nys 1998). As the inter-moult period dictates how long the feathers have to accumulate surface metals (Jaspers et al. 2004) and this may not be the same on the two allotments, this factor could also adversely influence the results.
The reported lack of correlation between Pb concentrations in different feather groups or between feather and blood Pb concentrations in blackbirds (Turdus merula) from Pb-polluted urban areas (Scheifler et al. 2006) further confirms the limitations of this technique for measuring bioavailable Pb exposure. Feathers were not collected following remediation as it was considered likely that they would still reflect pre-remediation contamination. This was because typically adult domestic chickens moult only once a year in autumn (King and McLelland 1975) and as the soil remediation took place in early November, any feathers collected 6 months later were likely to have been shed prior to remediation, so their metal content would be derived from contaminated soil exposure.
4.3.3 Metal Levels in Whole Guano
Pb concentrations in whole guano from birds on the contaminated allotment were elevated over control samples (Fig. 2) and normal background concentrations reported in the literature (Nicholson et al. 1999). This indicates that oral uptake of contaminated soil was responsible for the high Pb guano concentrations and suggests whole guano could be a suitable biomonitor for Pb exposure. By assuming that the majority of ingested Pb was from the contaminated soil and dietary Pb is concentrated 3.25 times in chicken guano as reported for Cu (Kunkle et al. 1981), the calculated percentage of soil uptake on a dry matter basis was 8 %. This is in agreement with estimates of soil intake by chickens reported in the literature (Beyer et al. 1994).
Following remediation of the contaminated allotment with clean soil (having a mean value of 59 mg Pb kg−1dry mass), the whole guano lead concentration dropped to a median of 31 mg Pb kg−1. But if the same soil intake value of 8 % as determined above is assumed, on calculation, this should have resulted in a guano concentration of 15.6 mg Pb kg−1, which is nearer to the value of 8.3 mg Pb kg−1 obtained from the guano of chickens on the control allotment (Fig. 2). This elevated concentration of Pb in guano after site remediation is likely to have resulted from the urine component of the guano (see “Metal Levels” below).
Cu and Zn concentrations in the guano of chickens on contaminated soil were not apparently different from guano metal concentrations sampled from the control site (Fig. 2), while guano samples post-remediation appeared to have lower metal concentrations than those found in guano from control and pre-remediation birds. This result shows how the typically high dietary inclusion rates of these metals (Mohanna and Nys 1998) in the ration fed to the control site birds masked the elevated uptake from the soil in birds on the contaminated allotment. The similar metal concentrations found in the guano samples would suggest birds from the control and contaminated sites are being equally exposed to Cu and Zn. This is not the case because the metals come from different sources (either food or contaminated soil) and so may be in different chemical forms, which can affect their relative bioavailability (Ruby 2004). Cu and Zn in guano from the control allotment birds is entirely from the feed, and these metals are reported to be poorly absorbed in chickens with less than 6 % being retained in the body from commercial diets (Mohanna and Nys 1998). As concentrations of Cu and Zn are low in the feed (5.7 and 20 mg kg−1) but high in the soil (273 and 800 mg kg−1) on the contaminated allotment, it can be calculated that most of the guano derived metal originates from the soil intake of these chickens. Assuming again a soil intake of 8 % (dry mass basis), for each kilogram of dry diet consumed, the Zn intake from soil would be 8 % of 827, i.e. 66.2 mg Zn, while from feed 92 % of 20 results in an intake of 18.4 mg Zn; similarly the Cu intake from soil is 8 % of 273, i.e. 22 mg Cu compared with the lower intake from the feed being 92 % of 5.7, i.e. 5 mg Cu. The calculated intake of Cu and Zn in the control chickens (being entirely from feed) of 10.6 mg Cu kg−1 and 60.5 mg Zn kg−1 was lower than the intake of chickens on the contaminated allotment, 27 mg Cu kg−1 and 84.6 mg Zn kg−1, respectively. As this difference between contaminated and control chickens is not shown in the whole guano analysis (Fig. 2), it may suggest that the soil-derived metals are more readily absorbed from the digestive system as a consequence of being more bioavailable. In a separate study, we determined the Cu and Zn bioavailability in the soil from the contaminated allotment using an in vitro method (Rieuwerts et al. 2000) and found them to be high (75–84 %). This is consistent with these metals’ likely origin from added anthropogenic products of combustion (incinerator bottom ash) in contrast to geological sources found in background soils (Rieuwerts et al. 2000). This highlights a potential problem of using whole guano in metal exposure studies because it does not take into account variations in metal bioavailability (Ruby 2004).
4.3.4 Avian Urine
The post-remediation extracted AUS (Fig. 3) appear to show urine to be a route for Pb excretion. Additionally, the persistent high AUS concentration, despite low feed and soil values, confirms that this metal’s presence in extracted urate spheres is not simply from faecal contamination.
In contrast to other methods (eggs, feathers or whole guano), AUS samples from chickens on the contaminated soil, when compared with controls, appeared to give a better representation of the birds’ elevated exposure to all three heavy metals (Fig. 3). This may be a consequence of the AUS content consisting of heavy metals entirely derived from the bloodstream following digestive absorption, hence representing the fraction of metals from environmental sources that are bioavailable to the birds. In comparison, AUS samples from the chickens 6 months after site remediation reflected the reduced Cu and Zn exposure from the clean replacement soil. Interestingly Pb concentrations in the AUS remained high after site remediation. This continued elevated excretion may be a consequence of bone mobilised for egg production (Dacke 2000; King and McLelland 1975), releasing chronically sequestered Pb deposits into the bloodstream. Bone Pb concentrations in birds account for approximately 90 % of the body burden, with egg laying females accumulating more than males (Scheuhammer et al. 1999). It would be expected following remediation that urine Pb concentrations should decline as the bone Pb is excreted over time. In humans, this decline may be quite prolonged (decades) and varies with bone type, metabolic state and subject age (Hu et al. 1998). Similarly, whether or not a chicken was laying eggs would be expected to affect the rate of bone mobilisation and therefore Pb urinary excretion. Pain et al. (1997) reported that blood Pb concentrations remained elevated for longer time (several months) following higher exposure from Pb shot ingestion in marsh harriers (Circus aeruginosus). Persistent excretion of this quantity of Pb 6 months after reducing the birds’ Pb intake may suggest substantial bone deposits of Pb from their previous prolonged exposure.
Cu and Zn do not substantially accumulate in the body like Pb (Walsh 1990), being essential metals under metabolic control. Therefore they did not show a prolonged excretion in the AUS following remediation. AUS concentrations of Cu and Zn are excess to the bird’s requirement excreted under homeostatic control, while Pb AUS concentrations reflect unregulated blood concentrations. In the light of blood concentrations of essential metals being kept within a normal range, AUS sampling may be a better measure of excessive exposure than blood, casting doubt on blood being the ‘gold standard’ for monitoring purposes (Furness 1993). For the non-essential metal, Pb, AUS concentrations could be expected to reflect blood concentrations. However the present study showed AUS concentrations may not directly relate to the birds’ current intake due to Pb accumulation in bone and its subsequent release due to bone remobilisation.
The EDAX analysis of AUS has been reported previously by Casotti and Braun (1997 and 2004), where they determined the ionic composition of individual urate spheres. Chicken urate spheres were reported to contain Mg using EDAX analysis (Casotti and Braun 1997). However in a later paper, the authors suspected it had been from background analysis of the stub (Casotti and Braun 2004).
In the present study, the EDAX analysis of individual AUS was unable to detect any of the three metal ions identified with the contaminated soil. This was because the sensitivity of the EDAX analysis is restricted to 0.1 % (dry mass) of a sample, equivalent to 1 g/kg and several times higher than the concentrations detected in the AUS by ICP-OES. The elemental analysis by EDAX of individual AUS (Fig. 4b) showed that K and Ca were the predominant cations, in agreement with Casotti and Braun (2004).
4.3.5 Health Implications for Chickens Ingesting Heavy Metal Contaminated Soil
The toxic and sub-lethal effects of ingested Pb, Cu and Zn on birds varies widely between species, age, sex and the chemical form of each metal (Eisler 2000 and references within). Experimental poisoning of captive birds with Pb showed wide species variation in susceptibility (Beyer et al. 1988), with a similar finding reported for Cu and Zn (Eisler 2000 and references within). Among avian species, domestic chickens are comparatively resistant to Pb toxicosis, with a diet containing 1.85 g kg−1 as Pb acetate given over 4 weeks to domestic cockerels being non-lethal (Franson and Custer 1982). Diets in domestic chickens with concentrations above 500 and 2,000 mg kg−1 (dry mass) Cu and Zn respectively are reported to be toxic (Eisler 2000 and references within). Such a relative insensitivity of the domestic chicken to metal toxicosis compared to other avian species (Eisler 2000) combined with its habit of ingesting soil makes it a suitable sentinel species to biomonitor heavy metal-contaminated soils.
Dietary intake of lead (Pb), copper (Cu) and zinc (Zn) from the combined soil and feed components, in chickens on the contaminated allotment
Total dietary concentration this study
Toxic concentration reporteda
In the context of heavy metal pollution, avian biomonitoring attempts to determine a bird’s internal exposure to bioavailable metals from the environment classically represented by circulating blood concentrations (Furness 1993). It is evident that current non-destructive biomonitoring methods using eggs, feathers or guano may not adequately reflect this. Egg and feather production draw upon both current intake and sequestered body reserves, so may not reflect current heavy metal body uptake from environmental exposure. Also, the homeostatic control of essential metals (e.g. Cu and Zn) in blood restricts their deposition in eggs and feathers to within a normal range (Walsh 1990). Feathers may gain variable amounts of surface accumulated heavy metals, which on analysis are indistinguishable from bioavailable internal deposits (Scheifler et al. 2006). Analysis of guano is complicated by it being a mixture of faecal and urinary excretions. The faecal heavy metals may have varied bioavailability, with non- bioavailable metals simply transiting the digestive system.
The present study has shown that metals can be measured in AUS but not that concentrations reflect biological availability because there was no assessment of availability or uptake. This could have been achieved by measuring metal residues in tissues and/or blood from the birds. The shortcomings of the study include the lack of a proper control group: the control birds in this study were at a different site, were not kept on uncontaminated soil and their diet was different. Taking representative samples of bulked guano and AUS meant that statistical analysis was not possible for these samples, giving only a measure of technical rather than biological variation. Guano should have been collected and analysed from individual birds, and concurrently residues of metals determined in their blood and a range of tissues, to assess uptake. Another problem with the study was that metal concentrations in pre- and post-remediation samples were measured using different methods carried out at different laboratories. This seriously affects any comparison between the two measurements and it adds to the problem of using bulked samples.
We are grateful for the assistance of P. Hartley of Newcastle City Council in the collection of samples from the two allotments and the generous cooperation of the allotment gardeners. We thank Dave Dunbar for metal analysis and Grant Staines for SEM imaging and laboratory assistance from Fiona McLachlan, all from Newcastle University, UK.
- Ademuyiwa, O., Arowolo, T., Ojo, D. A., Odukoya, O. O., Yusuf, A. A., & Akinhanmi, T. F. (2002). Lead levels in blood and urine of some residents of Abeokuta, Nigeria. Trace Elements and Electrolytes, 19(2), 63–69.Google Scholar
- Adeola, O., & Rogler, J. C. (1994). Comparative extraction methods for spectrophotometric analysis of uric acid in avian excreta. Archives of Animal Nutrition-Archiv Fur Tierernahrung, 47(1), 1–10.Google Scholar
- Beyer, W. N., Spann, J. W., Sileo, L., & Franson, J. C. (1988). Lead poisoning in six captive avian species. Archives of Environmental Contamination and Toxicology, 17, 121–130.Google Scholar
- Braun, E. J., & Pacelli, M. M. (1991). The packaging of uric acid in avian urine. The FASEB Journal, A1408.Google Scholar
- Casotti, G., & Braun, E. J. (1997). Ionic composition of urate-containing spheres in the urine of domestic fowl. Comparative Biochemistry and Physiology, Part A 118(3), 585–588.Google Scholar
- CDC (2009). Fourth national report on human exposure to environmental chemicals. Atlanta, GA: Centers for Disease Control and Prevention. Available: http://www.cdc.gov/exposurereport/. Accessed 10 April 2012.
- Dacke, G. C. (2000). In G. C. Whittow (Ed.), Sturkie’s avian physiology (5th ed., Vol. Chapter 18, pp. 472–485). London: Academic.Google Scholar
- Davies, K. J. A., Sevanian, A., Muakkassahkelly, S. F., & Hochstein, P. (1986). Uric-acid iron-ion complexes—a new aspect of the antioxidant functions of uric-acid. Biochemical Journal, 235(3), 747–754.Google Scholar
- DEFRA (2000). Statutory guidance. UK DETR circular 2/2000.Google Scholar
- EPA publication number 600491010 (1991). Methods for the determination of metals in environmental samples. Office of Research and Development, Washington DC 20460. Available at http://www.epa.gov/nscep/index.html. Accessed 10 April 2012.
- Fitzner, R. E., Gray, R. H., & Hinds, W. T. (1995). Heavy-metal concentrations in great blue heron fecal castings in Washington State—a technique for monitoring regional and global trends in environmental contaminants. Bulletin of Environmental Contamination and Toxicology, 55(3), 398–403.CrossRefGoogle Scholar
- Franson, J. C., & Custer, T. W. (1982). Toxicity of dietary lead in young cockerels. Veterinary and Human Toxicology, 24, 421–423.Google Scholar
- Furness, R. W. (1993). Birds as monitors of pollutants. In R. W. Furness & J. J. D. Greenwood (Eds.), Birds as monitors of environmental change (pp. 86–143). London: Chapman & Hall.Google Scholar
- Gomez, G., Baos, R., Gomara, B., Jimenez, B., Benito, V., Montoro, R., Hiraldo, F., & Gonzalez, M. J. (2004). Influence of a mine tailing accident near Donana National Park (Spain) on heavy metals and arsenic accumulation in 14 species of waterfowl (1998 to 2000). Archives of Environmental Contamination and Toxicology, 47(4), 521–529.CrossRefGoogle Scholar
- Grand, J. B., Franson, J. C., Flint, P. L., & Petersen, M. R. (2002). Concentrations of trace elements in eggs and blood of spectacled and common eiders on the Yukon-Kuskokwim Delta, Alaska, USA. Environmental Toxicology and Chemistry, 21(8), 1673–1678.Google Scholar
- Hartley, C.P., Vizard, K., & Air, V. (2004). Branxton A and Branxton B allotment sites. Desk top study and site investigation. Newcastle City Council. Public Health and Environmental Protection. Civic Centre, Newcastle upon Tyne, NE1 8PB, UK.Google Scholar
- Jaspers, V. L. B., Rodriguez, F. S., Boertmann, D., Sonne, C., Dietz, R., Rasmussen, L. M., Eens, M., & Covaci, A. (2011). Body feathers as a potential new biomonitoring tool in raptors: a study on organohalogenated contaminants in different feather types and preen oil of West Greenland white-tailed eagles (Haliaeetus albicilla). Environment International, 37, 1349–1356.CrossRefGoogle Scholar
- JFWEC (1999) Joint FAO/WHO Expert Committee on food additives, Report TRS 896-JECFA 53/81, Monograph FAS 44-JECFA 53/273.Google Scholar
- King, A. S., & McLelland, J. (1975). Outlines of avian anatomy (pp. 25–26). London: Bailliere Tindall.Google Scholar
- Kunkle, W. E., Carr, L. E., Carter, T. A., & Bossard, E. H. (1981). Effect of flock and floor type on the levels of nutrients and heavy-metals in broiler litter. Poultry Science, 60(6), 1160–1164.Google Scholar
- McGrath, S. P., & Loveland, P. (1992). The soil geochemical atlas of England and Wales. London: Blackie Academic and Professional.Google Scholar
- McNabb, R. A., & McNabb, F. M. A. (1980). Physiological chemistry of uric acid: solubility, colloid and ion binding properties. Comparative Biochemistry and Physiology, Part A 67, 27–34.Google Scholar
- Scheifler, R., Coeurdassier, M., Morilhat, C., Bernard, N., Faivre, B., Flicoteaux, P., Giraudoux, P., Noel, M., Piotte, P., Rieffel, D., de Vaufleurs, A., & Badot, P. M. (2006). Lead concentrations in feathers and blood of common blackbirds (Turdus merula) and in earthworms inhabiting unpolluted and moderately polluted urban areas. Science of the Total Environment, 371(1–3), 197–205.CrossRefGoogle Scholar
- Taylor, M. G., & Simkiss, K. (1989). Structural and analytical studies on metal ion-containing granules. In S. Mann, J. Webb, & R. J. P. Williams (Eds.), Biomineralisation: chemical and biochemical perspectives (pp. 427–460). Weinheim: VCH.Google Scholar
- Walker, C. H., Hopkins, S. P., Sibly, R. M., & Peakall, D. B. (2001). Principles of ecotoxicology (2nd ed., pp. 3–6). New York: Taylor and Francis.Google Scholar
- Walsh, P. M. (1990). Use of seabirds as monitors of heavy metals in the marine environment. In R. W. Furness & P. S. Rainbow (Eds.), Heavy metals in the marine environment (pp. 183–204). Boca Raton: CRC.Google Scholar