CorkCounty Council Comhairle Contae Chorcaí Environmental Protection Agency

An Ghníomhaireacht um Chaomhnú Comhshaoil

Animal Health Surveillance of Dairy Herds in the vicinity of a large chemical industrial complex in the

Cork Harbour Region 2011 - 2016 in the vicinity of a large chemical industrial complex in the Anim

al Health Surveillance of D

airy Herds - Cork H

arbour Region 2011 - 2016 Cork County CouncilCorkCounty Council Comhairle Contae Chorcaí

Environmental Protection AgencyAn Ghníomhaireacht um Chaomhnú ComhshaoilVeterinary Department, Environment Directorate

Veterinary Department, Environment Directorate

Ringaskiddy

Carrigaline

Rochestown

Mahon

Fota Island

Douglas

Little Island

Glanmire To Dublin

To Cork City

To Rosslare

Whitegate Shanbally

Crosshaven

Roches Point Light House

Glenbrook

Monkstown

Carrigtwohill Glounthaune Midleton

Ballynacorra

Passage West

Camden Fort Meagher

Spike Island

Cobh

Roches Point

AghadaHaulbowline/ Rocky Island

Fort Carlisle (Davis)

R612

R612

R612

R613

R613

R611

R611

N28

N28

N28

R624

R624

R623

N25

N25 N25

R630

R630

R630

M8

N40

N40 R610

R610

Cork Harbour

Prepared for the Environmental Protection Agency

by Veterinary Department, Environment Directorate, Cork County Council

Animal Health Surveillance of Dairy Herds in the vicinity of a large chemical industrial complex in the

Cork Harbour Region 2011 - 2016

Environmental Protection

AgencyPharmaceutical

& Petrochemical

Industry

Food Safety Authority of

Ireland

Dairy Farmers

Private Veterinary

Practitioners

DAFM –Agricultural

House

DAFM –Regional

Veterinary Laboratory

Cork Clinical

Pathology Laboratory -

UCD

Dr. Jim O’Donovan

Agricola Processing

Fermoy

Environment SPC

Dr. Kevin O’Farrell

DAFM -Veterinary Inspectors

Health Service

Executive

School of Public

Health -UCD

Disclaimer

Although every effort has been made to ensure the accuracy of the material contained in this publication,

complete accuracy cannot be guaranteed. Neither the Environmental Protection Agency, Cork County

Council nor the authors accept any responsibility whatsoever for loss or damage occasioned or claimed

to have been occasioned, in part or in full, as a consequence of any person acting or refraining from

acting, as a result of a matter contained in this publication. All or part of the publication may be

reproduced without further permission, provided the source is acknowledged.

Acknowledgements

Graphics - Catherine O’Callaghan, Architects Department

Photographs courtesy of

Martin Walsh – Photographer

Denis Horgan – Photographer

Donal O Callaghan

Table of Contents

Executive Summary 1

1.0 Introduction 9

1.1 Background and current drivers to the Study 9

1.2 Objectives 10

1.3 Governance and ongoing review of the scheme 11

1.4 Study Design & Operation 12

1.5 Perception v Reality 12

1.6 Future Options 14

1.7 Current Status of AHSS 18

2.0 Markers of Performance: Herd Performance Parameters 19

2.1 Materials & Methods 19

2.1.1 Sample Population 19

2.1.2 Parameters: Dairy Herd Performance 20

2.1.3 Farm Performance 22

2.2 Results & Discussion 22

2.2.1 Farm Performance Parameters 22

2.2.2 Calving-related Biological Parameters 28

2.2.3 Fertility-related Biological Parameters 35

3.0 Markers of Effects: Clinical Pathology 39

3.1 Introduction 39

3.2 Materials & Methods 40

3.2.1 Analytical Methodology 42

3.2.2 Haematology Results 43

3.2.3 Clinical Biochemistry Parameters 49

3.3 Discussion 60

4.0 Markers of Exposure: Persistent Organic Pollutants (Dioxins, Furans and Dioxin-

Like Polychlorinated Biphenyls (PCBs) and Marker PCBs) in Bovine Milk 61

4.1 Background 61

4.1.1 Biological and Ecological Significance of Dioxins and Dioxin-like PCBs 62

4.1.2 Dioxins in Cows’ Milk 62

4.2 Objectives 63

4.3 Materials and Methods 63

4.3.1 Milk sampling procedure 63

4.3.2 Laboratory Testing 64

4.3.3 Recent Re-evaluation of “Toxic Equivalency Factors” for Assessment of Levels of Dioxins,

Furans and Dioxin-like PCBs 65

4.4 Results 66

4.4.1 Dioxins, Dioxin-like PCBs and Marker PCBs 66

4.4.2 FSAI / Cork Co Co Biannual Study on Dioxins and PCBs 2008 to 2015 – Evidence of

Seasonal Variation. 68

4.5 Discussion 72

5.0 Conclusion 75

6.0 Appendices and References 79

Appendix 1: Glossary/Definitions/Abbreviations 79

Appendix 2: Dr. Riona Sayers, Herd Health Senior Research Officer, Teagasc, Moorepark 85

Appendix 3: Target and Control Herdowner Survey 87

Appendix 4: Dr. Patrick Wall, Professor of Public Health, UCD 89

Appendix 5: Mr. Tim Lucey, Chief Executive, Cork County Council 91

References 93

Authors 98

Executive Summary

1

Executive Summary

A glossary and list of definitions relevant to this study is included in Appendix 1.

The study presented in this report (2011-2016) is a continuation of a surveillance programme

initiated in the early 1990’s. The study monitors the health and productivity of dairy herds in the

Cork Harbour basin and surrounding districts as a surrogate or “proxy” of human health and

overall environmental quality.

County Cork has a diverse socio-economic infrastructure consisting of agriculture, industry and

tourism. County Cork accounts for 26% of Ireland’s National Dairy output. The Cork area

experienced a rapid expansion of the pharmaceutical industry in the late 1980’s /early 1990’s

with a large number of international companies establishing manufacturing plants in the Cork

Harbour region and surrounding regions. These developments have resulted in substantial

benefit to the local economy and are major employers in the region.

There is currently an underlying growing perception of environmental risk in the Cork harbour

region. The Cork Harbour Alliance for Safe Environment (CHASE) and other community groups

state that more than 23,000 people signed a petition opposing the Indaver application for the

development of a hazardous waste incinerator in Ringaskiddy (CHASE, 2018).

In common with other industrialised countries around the world, there has been ongoing local

concern in Ireland regarding the potential for pollutants emitted from major industrial clusters

having detrimental effects on the health of animals and people in these regions. Previous

allegations of increased animal mortality/ morbidity and reduced productivity in Ireland and

elsewhere allegedly attributable to environmental contaminants have resulted in complex,

protracted, expensive investigations, with associated public health anxieties and significant

regional economic loss (ILRM 629, 1988; Anon, 1995; Anon, 1995a; DAFF, 2010).

Participation in this surveillance programme was made a condition of the air emission licence

granted to one large multinational company in the area in 1991. The scope of the study was

subsequently expanded to encompass the other major industrial facilities in the region. Since

1993, the programme has been coordinated by the Veterinary Department of Cork County

2

Council (VDCCC) on behalf of the Environmental Protection Agency and is funded by way of

contributions from industrial operators under the terms of their respective Integrated Pollution

Prevention & Control (IPPC) licences.

Dairy cows were chosen as the subject of this study as they are considered to be particularly

appropriate indicators of the quality of the ambient environment, i.e. they live in the area for

their entire lives inhaling the ambient air and consume, in the main, local forages and water and

as such are continually exposed to environmental factors. In addition, they are expected to

maintain a productive life and their inputs and outputs can be readily monitored. They are also

significant contributors to the human food chain both through milk and meat supplied.

The results of the study are reviewed on an ongoing basis by a multidisciplinary expert

committee, including consultants in animal health, production, fertility and clinical pathology,

and the recommendations for modification to the design of the scheme are implemented

accordingly. Previous reviews have been carried out in 1996, 2005, 2006 and 2010. This

continual review is considered a key feature of the programme.

The adverse effects of toxins on animal health have been well described (Covello and

Merkhofer, 1993) there may be a wide variety of target organs and consequently a wide variety

of clinical syndromes; the changes may be subtle and the latency period after exposure to a toxin

may be lengthy (years / decades).

Biological markers

Biological markers can make an important contribution to toxicological and epidemiological

investigations and risk assessments, particularly in relation to diseases or conditions of obscure

aetiology, where direct or indirect exposure to environmental toxicants is suspected (Schulte &

Mazukelli, 1991). Investigations of health problems in populations exposed to suspected

industrial pollution are controversial (Lloyd et al., 1988) because it is difficult to demonstrate a

link with environmental contamination unequivocally (Lloyd et al., 1991).

3

The baseline data on herd health, productivity and tissue residues assimilated and interrogated

during this study period were applied as Biological Markers which can be classified as follows:

• “Markers of Performance”, i.e. variations in production outputs, presence of inter-current

disease and background metabolic disease levels;

• “Markers of Effects”, i.e. signals of tissue dysfunction, e.g. liver enzyme activity,

morbidity, mortality, haematological, physiological and clinical findings;

• “Markers of Exposure”, i.e. dioxin/ Polychorinated Biphenyls (PCB) levels in bovine

milk.

Markers of Performance - Results

Over the period of this report (2011-2016), there were a total of five Control and six Target herds

monitored. Statistical comparison of parameters for Target and Control herds was not considered

relevant due to the small sample size as discussed under ‘Future Outlook’ and Section 1.6.

However, comparisons are made between means generated for Control and Target herds and

where relevant with other nationally published data.

The farm performance parameters and, in particular, the data generated for mean stocking rate,

meals fed per cow and milk yield indicate that the study herds were more intensively managed

(2.57 livestock units/ ha) than the top one third of herds involved in the Teagasc National Farm

Survey(TNFS) (2.37 livestock units/ ha) and milk solids output (kg/cow) were also higher for the

study herds (452 kg/cow) than National Farm Survey (NFS) herds (372 kg/cow) for the same

period.

Still-births, birth defects and fertility rates may be used as indicators of natural or man-induced

environmental alterations, and animal surveillance for these parameters may be of supplemental

value for investigations of perceived or real environmental contamination (Marianfeld, 1979).

Abortion, congenital defects and increased calf mortality may be caused by exposure to

4

environmental contaminants (Lloyd et al., 1991); however, congenital malformations in

ruminants are well documented and may be due to a variety of causes (Davies et al., 2012).

Over the period, for Control and Target herds, the proportion of multiple births (4.0%), live

births (96%) and perinatal mortality (4.2%) recorded were similar to previous years and in line

with national values. Both Control and Target herds in this study recorded a similar incidence of

still-births (3%) compared to the ICBF national figures (2%). The proportions of male (52.2%) to

female calves born were similar for both groups and similar to national figures.

The average calving to calving interval for the individual Control and Target herds and the

comparison with Irish Cattle Breeding Federation (ICBF) national herd averages shows that the

Control and Target herds are similar and like the national dairy ICBF herds with a trend line

showing a reduction in the calving interval from 403 days in 2011 to 395 days in 2016. The

results for Control, Target and ICBF herds were similar over the years.

Markers of Effects - Clinical Pathology

Haematology and clinical biochemistry tests provide information on selected trace elements and

the nutritional and mineral status of animals, as well as providing information on the

haematopoietic activity of the bone marrow, i.e. the formation of blood cells. These tests

monitor the physiological and pathological responses of animals, as well as their responses to

stress and inflammatory disease. In this regard they can help to identify the extent to which

environmental factors contribute to disease, morbidity, mortality or reduced productivity. Over

the six-year period of this report, a total of 32,764 haematological and 35,121 biochemical tests

were performed, analysed and compared for Control and Target herds.

While none of the blood tests performed were specific markers for individual environmental

contaminants, parameters that could be altered by a variety of disease states and farm

management practises have been identified. Variations in metabolic profiles in dairy herds have

been associated with age, season, stocking rate, nitrogen usage and genotype (O’Farrell et al.,

1986; Olmos et al., 2009). The effect of sampling time and season were significant on

5

practically all blood parameters in this study. Overall there was no evidence, on the basis of

comparison of Control and Target herd results that there was any adverse effect of location on

the clinical pathology parameters examined.

Markers of Exposure

The results for dioxins, furans and dl - PCBs in Target and Control milk samples taken over the

period 1991-2015 and included in the present study were within the range recorded from other

sites in Ireland (EPA, 2012). In addition, they were significantly less than the applicable limits

set by the EU in Council Regulation (EC) No 1881/ 2006 and also the recently enacted

Commission Regulation (EC) No 1259/ 2011. The reducing trends in total dioxins observed

over the study period are largely accounted for by the reduction in PCB contamination, while

dioxins and furan levels have remained generally stable at values considered as low background

levels in European terms. Over the period since 1995, total dioxin WHO-TEQ in both Target

and Control milk have exhibited strong downward trends; the decreases have been of the order of

40% between the late 1990s and the early 2000s and have fluctuated moderately around this new

lower level since then. This downward trend was strongly correlated with the concentration of

the dl- PCB component of total dioxin, which decreased markedly over the period of the study.

Between 1995 and 2009, Control milk levels were on average 20% lower than the comparable

Target values though during the period 2010 to 2015 the total dioxin WHO-TEQ concentrations

for Target and Control milk samples were, broadly speaking, very similar.

The minor differences in observed concentrations between Control and Target milk are

consistent with the relatively greater degree of urbanisation of the Target farms, while the

differences between spring and autumn levels are considered to reflect season variability in the

ingestion of soil by grazing animals.

6

Conclusion

The overall findings of this study indicated that for ‘Markers of Performance’ Control herds were

more intensively managed and had higher milk yields. Variations in management practices,

stocking rates, calving patterns, breed profiles, etc have been found to show differences in herd

performance. Surveillance for ‘Markers of Effects’ indicated that there was no evidence, on the

basis of comparison of Control and Target herd results, that there was any adverse effect of

location on clinical pathology parameters. In addition, for ‘Markers of Exposure’, the reduction

in total dioxins in milk produced in the Cork Harbour catchment and adjacent areas is similar to

the findings of other studies in the UK and Ireland for the same period.

7

Future Outlook

The Animal Health Surveillance Scheme (AHSS) has evolved to its present form over 25 years.

Initially the surveillance scheme involved 10 dairy herds located in the Cork harbour region and

two herds located at a distance of 15 to 20 miles from the harbour region who had agreed to

participate. Over time the scheme moved to a more ‘longitudinal type’ study design. Most

longitudinal studies examine associations between exposure to known or suspected causes of

disease and subsequent morbidity or mortality. In the simplest design a sample or cohort of

subjects exposed to a risk factor is identified along with a sample of unexposed controls. The two

groups are then followed up prospectively, and the incidence of disease in each is measured. By

comparing the incidence rates, attributable and relative risks can be estimated. Reviewers have

suggested that the objectives and design of the scheme should be re-examined as to whether it is

now ‘fit for purpose’. As a first step, it was decided to review the study design.

Comparisons of target herds with control/normal population are fraught with statistical

uncertainty within the existing project design. Thus, statistically significant differences that

might be found may not necessarily be due to a location effect and may not be biologically

relevant due to small sample size, calving pattern, management and breed effects. Some

measures are subject to this confounding ‘noise’, whereby ‘false positive’ (Type 1 error) and

‘false negative’ (Type 2 error) results occur. In addition, the geographical definition of the zone

from which the Target herds is based on proximity to industrial sites and not based on wind

dispersion studies.

Option 1.

Following on a review involving the School of Mathematical Sciences, University College Cork,

it was agreed that in order to provide statistical clarity, the objectives of the study might be better

served by having a Non-Inferiority Study design.

The Non-Inferiority design would require 22 Target herds compared with 88 Control herds.

These samples sizes were computed based on a number of assumptions, such as the measure of

8

interest, the clinical difference, and assumed standard deviation. Only one measure was used to

derive the sample size, and it is common that sample size would be computed for each primary

measure of interest and not based on a single measure.

Careful matching of herds is essential to minimising the ‘noise’ which confuses the picture and

makes interpretation of results more difficult. To achieve this level of statistical power will

require a considerable increase in number of herds, staff, financial support and input from other

organisations which have expertise in this area.

Option 2.

Modifications to existing study

Several decades of work have amassed a highly valuable amount of baseline data for herds in the

Target and Control area. This is an immense and almost unique resource. Over time the scheme

moved to a more ‘longitudinal type’ study design where ‘balance’ between the number of

Control and Target herds was sought. The sample size is important to establishing the ‘power’ of

the study. The established nature of the project with its underlying infrastructure and resources

has made it very difficult to obtain sufficient herds to give a statistical power of 80% which

would be considered desirable for this type of study. To achieve this level of statistical power

will also require increased resources similar to those outlined in option 1.

The benefit from the baseline data collected by this study would be greatly enhanced if it was

linked to the statistical power outlined in either option 1or 2.

11. 0 Introduction

1. 1 Background and current drivers to the Study

1. 2 Objectives

1. 3 Governance and ongoing review of the scheme

1. 4 Study Design and Operation

1. 5 Perception v Reality

1. 6 Future Options

1. 7 Current Status of AHSS

9

1.0 Introduction

1.1 Background and current drivers to the Study

This report is the 4th major periodic assessment of the outcome of the study and covers the period

2011 to 2016.

- Earlier reports in this series were issued in 1998, 2005, 2010;

- In addition, peer reviewed articles based on data generated in the study were published by

(Buckley et al., 2007) and (O’Donovan et al., 2010).

- In Ireland and many other countries, industries function in close proximity to pasture

land, and industrial effluents are discharged to the environment as waste streams or as

gases or particulate materials. Air, water, and vegetation in the vicinity of industrial

complexes may act as a source of toxins to animals, and people may be at risk directly

from the same source, or through consuming animal food products. Direct evidence of

adverse health effects due to industrial pollution are sometimes tragically indisputable as

in Bhopal, India (Kumar, 1993) and Seveso, Italy (Bertazzi, 1991) and Chernobyl (RPII,

1987). In other circumstances, the effects of pollutants may be more subtle and long term

and as there are few syndromes due to poisonous substance that may not also be

produced by some other causative agent (Clarke et al., 1981), (Buckley and Larkin,

1998).

- In parallel to this programme, investigations into instances of increased animal disease

morbidity/mortality due to various pollutants have been reported in Ireland. Case studies

such as Castlecomer investigations (Anon, 2010), accidental on farm lead poisoning in

adult cattle in Cork in 2014/15, as well as the detection of polybrominated by-phenyl

ethers (fire retardant chemicals) in bovine milk Co Cork (EPA, 2010) serve to heighten

concerns regarding direct threats to animal and human health and the safety of the food

chain.

10

- County Cork is the largest county in Ireland: it covers an area of 7,500 km sq. The Cork

Harbour Region has a thriving, and diverse, socio economic infrastructure including

agriculture, tourism, clusters of pharma industry and large pockets of urbanisation with

associated heavy traffic movement. Fourteen major multinational chemical companies

were located in the region of Cork harbour in 1991, some of whom contribute to EPA to

fund this programme in 2011 - 2017. In addition, County Cork is regarded as the dairy

capital of Ireland with a dairy cow population of approximately 368,000 dairy cows

accounting for 26% of the national dairy herd (ICBF, 2017).

- Ireland is already one of the world’s most efficient food producers, in terms of carbon

foot print per unit of output. “Sustainability” is one of the key drivers to the “FOOD

WISE STRATEGY” (2025) which states that “environmental protection and economic

competitiveness are equal and complimentary – one cannot be achieved at the expense of

the other” (DAFM, 2015).

1.2 Objectives

From the outset the specific objectives of the scheme were:

• To generate baseline data for the Cork harbour region using dairy herds located in the

vicinity of major industry as biomonitors.

• To compare the health, performance of target herds with control herds in non-industrial

areas and nationally.

• To examine temporal trends in herd health and productivity.

• To build a bank of tissues and milk (kidney, muscle and liver) from animals in these

herds which were archived (Buckley and Larkin, 1998).

• To facilitate analysis of milk and animal tissues for micro-organic pollutants such as

dioxins, PCBs, PCDDs, PCDFs and heavy metals in order to assess potential

bioaccumulation in dairy cattle and the food chain.

This baseline data provides a reference for the investigation of suspected incidents of animal ill

health, increased morbidity / mortality and reduced productivity that might be attributed to real

11

or perceived environmental pressures from industry or other local sources. The data may also be

interpreted to provide early warning for regulators and key stakeholders in the event of major

incidents (fires, explosions, leakages, discharges) with the potential for major environmental

contamination.

1.3 Governance and ongoing review of the scheme

The results of the study are reviewed on an ongoing basis by a multidisciplinary expert

committee, including consultants in animal health, production, fertility and clinical pathology,

and agreed recommendations for modification to the design of the scheme are implemented

accordingly. Previous reviews have been carried out in 1996, 2005, 2006 and 2010. This

continual review process is considered a key feature of the programme.

Initially the surveillance scheme involved 10 dairy herds located in the Cork harbour region and

two herds located at a distance of 15 to 20 miles from the harbour region who had agreed to

participate. A broad database of animal disease incidence, productivity and blood composition

was established. Some of the results were the subject of a publication in a peer reviewed

scientific journal (Buckley and Larkin, 1998).

A subsequent report for the period 2001 – 2004 was based on 10 harbour herds which were

matched with 10 control herds located in non-industrial areas (Buckley et al, 2007).

In 2013, a further report (2005- 2010) on the study herds was published in which the numbers of

Target and Control herds varied over the period of assessment. Two Control and up to 14 Target

herds were involved in this period. These changes were the result of ongoing reviews of

logistics, costs and participation rates within the overall scheme.

The current report (2011- 2016) includes 5 Control and 6 Target herds involving over 8,000

dairy cows. The Control herds were located in non-industrial areas of County Cork.

12

1.4 Study Design & Operation

Dairy cows are expected to maintain a productive life for approximately seven years and produce

a calf each year. Herd fertility and production are well understood and have been documented as

sensitive indicators of environmental quality (Lloyd et al., 1988; Buckley and Larkin, 1998).

Dairy cows are a significant contributor to the human food chain in providing raw material for

both the dairy and beef industries and as such constitute both direct and indirect exposure

pathways for the human population. This is considered to be particularly important in relation to

the potential for long-term bioaccumulations of lipophilic pollutants.

The ready availability of systematically collected herd productivity and fertility data through

national databases facilitate detailed comparisons between the study herds in Cork and national

data. (Irish Cattle Breeding Federation; Teagasc National Farm Survey Results).

Participation in the scheme was made a condition of the air emission licence granted to one large

multinational chemical company locating in the area in 1991. The scope of the study was

subsequently expanded to encompass the other major industrial facilities in the area. The study

was originally operated by the Veterinary Department Cork County Council (VDCCC) since

1991 and has been co-ordinated by VDCCC on behalf of the Environmental Protection Agency

(EPA) since 1993. The programme is funded by way of contributions from industrial operators

under the terms of their Integrated Pollution Prevention Control (IPPC) licences granted by the

EPA.

1.5. Perception v Reality

In the 1980’s and 1990’s there was a perception that the chemical pharmaceutical industry

located in the vicinity of the Cork harbour region had a poor environmental record. The aim of

this study was to develop a database on the health and productivity of dairy herds in the Cork

harbour catchment area and to assess the value of a multidisciplinary approach for environmental

health monitoring.

13

As part of its output, the scheme has investigated many incidents involving animal ill health that

could be linked to environmental pollution. In all cases the causes of ill health effects were

attributable to agents other than environmental contamination.

The World Health Organisation (WHO) points out that Global One Health security relies on all

countries/regions having the capacity to rapidly detect and control public health threats at their

source, and that mass and social media can play a major role in disseminating incorrect

information which may generate anxiety and can influence health seeking behaviour (Winters et

al., 2018). In the UK after a recent government vote on sentience, more than 360,000 people

signed a petition online that wrongly claimed the UK government had removed animal’s status

as sentient beings (Winters et al., 2018) (Mance, 2018).

It is envisaged that the information provided in this report will be of assistance to key

stakeholders in the region and provide a practical example of the global “One Health” initiative

which emphasises interdisciplinary collaboration for protecting and preserving human and

animal health and environmental quality.

There is currently an underlying and growing perception of environmental risk in the Cork

harbour region. CHASE and other community groups state that more than 23,000 people signed

a petition opposing the Indaver application for the development of a hazardous waste incinerator

in Ringaskiddy(CHASE, 2018).

More recently and currently (2017), local community groups in Co. Cork (Cork Harbour

Alliance for a Safe Environment - CHASE) and the East Cork Safety and Environment Group

have been campaigning against the proposed location of hazardous waste incinerators at

Ringaskiddy, Co Cork. This has resulted in heightened public perception of environmental risk

in the area. In 2013, eight Public Representatives (Councillors) from towns and communities

within the catchment area made a submission to the EPA to oppose the granting of a “Waste

License” for the proposed hazardous waste incinerator at Ringaskiddy. This submission states

that: “Sufficient research has not been carried out nor monitoring programmes put in place to

assess the dangers to health and the environment from such a proposal”

14

The “Ecological Risk Assessment for PCDD/F for Indaver Ringaskiddy Resource Recycling

Centre” Air, Water & Noise Consulting (AWN Consulting, 2015) concluded that “baseline

dioxin concentration in the eggs of fish-eating birds and in otters were considered to be low and

well within limit values for the eggs of fish-eating birds”. In addition, during the lifetime of the

proposed facility the report concluded that “the predicted change in dioxin concentration were

considered to be insignificant for both fish-eating bird’s eggs and otters based on exposure to

foraged fish”.

Cork Area Strategic Plan (CASP) is a pioneering initiative jointly sponsored by Cork County

Council and Cork City Council providing a vision and strategy up to 2020.

The CASP plan envisages that by 2020 the population of the CASP area, including Cork

harbour, will have grown to 440,000, with approximately 210,000 jobs provided in the region

and the number of households to be in excess of 166,000 (Cork County Council & Cork City

Council, 2001).

The Cork harbour area, as part of that plan is now acknowledged internationally as a prime

centre for tourism, leisure activities, heritage and historical sites, agriculture and wildlife all co-

existing with clusters of pharmaceutical and high-tech industries.

1.6 Future Options

Comparisons of target herds with control/normal population are fraught with statistical

uncertainty within the existing project design. Thus, statistically significant differences that

might be found may not necessarily be due to a location effect and may not be biologically

relevant due to small sample size, calving pattern, management and breed effects. Some

measures are subject to this confounding ‘noise’, whereby ‘false positive’ (Type 1 error) and

‘false negative’ (Type 2 error) results occur. In addition, the geographical definition of the zone

from which the Target herds are drawn was based on proximity to industry rather than on wind

dispersion studies.

15

Statistical significance is a measure of how likely an observed result could have occurred, on the

basis of a set of assumptions (Reese, 2004). Comparison of two populations can have different

outcomes to treatments which are statistically different but not biologically different. A

biologically relevant effect can be defined as an effect considered by expert judgement as

important and meaningful for human, animal, plant or environmental health. It therefore implies

a change that may alter how decisions for a specific problem are taken (EFSA, 2011). The larger

the sample size (power) of the study, the more likely it can detect the biologically defined

relevant effect as statistically significant.

Following on a review involving the School of Mathematical Sciences, University College Cork,

it was agreed that in order to provide statistical clarity, the objectives of the study should be able

to address the following questions:

1. Is there a difference in the mean performance between the target and control herds? – this

is a parallel treatment- control study of equality

2. Is the mean performance of the target and control herds equivalent, that is no better and

no worse? – this is an equivalence study

3. Is the mean performance of the target herds no worse than the control herds? – this is an

inferiority study

Although the questions may appear very similar, they are subtly different but this difference

constitutes a completely different experimental study design.

Using Haemoglobin (a blood parameter) as an example, the following assumptions are made on

the advice of the clinical study experts:

• A clinically acceptable margin of 3 g/dl.

• An anticipated difference between the true means of 0 g/dl.

• A standard deviation of 5 g/dl.

• Type I error rate of 0.05.

• A power of 80%.

16

Based on the above, the approximate number of herds required for the Target group for ratios

of Controls of 1:1, 2:1, 3:1 and 4:1 are given in Table 1.

Table 1. For the three types of study design the approximate number of Target herds

required to achieve a power of 80% with different ratios of Control herds for the blood

parameter Haemoglobin.

Ratio of Control

to Target herds

Study

design 1.

Equality

Study

Study design 2.

Equivalence

Study

Study design 3.

Non-Inferiority

Study

1:1 44 48 35

2:1 33 36 26

3:1 30 32 23

4:1 28 30 22

The table of sample sizes fixed power to 80% and we did not examine the impact on power of

different n’s but more what the n required would be for the Target group based on different ratio

of Controls to Target.

Power calculations based on binary traits (yes/no) would require a larger sample size than for

Haemoglobin especially where the incidence is low.

Given the number of herds outlined in the Target region of Cork County and their likely

participation rate it is thought that the maximum number of Target herds that could be enrolled is

40. However, it should be noted that greater statistical power could still be achieved if the

proportion of Control to Target herds was >1. The marginal benefit of an additional multiple of

17

control herds was minimal after 4 Control herds per Target herd had been reached

(https://onlinelibrary.wiley.com/doi/book/10.1002/9780470696750 Chapter 8).

1.6.1 Option 1.

Based on a Non-Inferiority Study which requires the smallest number (n=22) of Target herds,

88 Control herds would be required at a ratio of 4:1. Possible factors for stratifying herds include

herd size, breed composition, average herd milk yield, geographical location, management and

calving pattern. Careful matching of herds for these criteria is essential to minimising the ‘noise’

which confuses the picture and makes interpretation of results more difficult. To achieve this

level of statistical power will require a considerable increase in number of herds, staff and

financial support.

1.6.2 Option 2.

Modifications to existing study

Several decades of work have amassed a highly valuable amount of baseline data for herds in the

Target and Control area. This is an immense and almost unique resource. Over time the scheme

moved to a more ‘longitudinal type’ study design where ‘balance’ between the number of

Control and Target herds was sought. The sample size is important to establishing the ‘power’ of

the study. The established nature of the project with its underlying infrastructure and resources

has made it very difficult to obtain sufficient herds to give a statistical power of 80% which

would be considered desirable for this type of study. To achieve this level of statistical power

will also require increased resources similar to those outlined in option 1.

The benefit from the baseline data collected by his study would be greatly enhanced if it was

linked to the statistical power outlined in either option 1or 2.

18

1.7 Current Status of AHSS

The AHSS has ceased operation since funding was withdrawn in 2018. Data is no longer being

collected from participating herds and herdowners have expressed their concerns to the Council.

The detailed information provided by the AHSS includes stocking rate, fertiliser use, milk yield,

milk solids yield, calving data, including birth defects, stillbirths, calf sex ratio, multiple births,

abortions, neonatal deaths, calving intervals, calving rates and metabolic profile data on over 11

haematological and over 14 biochemical parameters in blood. This level of detail required for the

scheme is not available from any other source in the country. In addition, the multidisciplinary

format includes the human, technical and knowledge infrastructure which has been developed

over the years to give the scheme the professional and expert capacity it now has. The capability

now exists to provide rapid response and the capture of high quality, relevant data on dairy cattle

health and productivity in the area in the event of a major industrial incident or emergency. To

quote Dr. Patrick Wall, Professor of Public Health, UCD, ‘Ireland has a vibrant agri-food

industry and a growing tourist industry, both dependent on a pristine environment, so in addition

to protecting the health of the public, early monitoring for adverse effects on the environment is

essential if we are to maintain our point of differentiation from other jurisdictions’.

22. 0 Markers of Performance: Herd Performance Parameters

2. 1 Materials and Methods

2. 1.1 Sample Population

2. 1.2 Parameters: Dairy Herd Performance

2. 1.3 Farm Performance

2. 2 Results and Discussion

2. 2.1 Farm Performance Parameters

2. 2.2 Calving-related Biological Parameters

2. 2.3 Fertility-related Biological Parameters

19

2.0 Markers of Performance: Herd Performance Parameters

2.1 Materials & Methods

2.1.1 Sample Population

The sample population consists of a set of Target herds, located in the vicinity of industrial

facilities in County Cork, and a set of Control herds located in adjacent non-industrial regions.

The number of Control herds that participated in the study varied as detailed in Table 2.1. These

changes resulted from ongoing reviews of logistics, costs and options taken by some of the herd

owners during the six-year period of this report.

Table 2.1: Description of Target and Control herds for the years 2011-2016 based on the

number of calving events per herd per year (excluding abortions).

Year

GROUP FARM 2011 2012 2013 2014 2015 2016 Grand Total

CONTROL A 155 151 170 126 196 187 985

B 113 * * * * * 113

C 86 72 * * * * 158

D 76 86 86 94 105 101 548

E 91 127 126 107 90

541

F * 125 114 122 136 146 643

G * * 61 70 76 80 287

H * * * * * 88 88

Total births 521 561 557 519 603 602 3363 Total herds

5 5 5 5 5 5

TARGET I 253 240 269 251 294 299 1606

J 118 123 137 140 149 165 832

K 146 107 117 114 129 133 746

L 67 61 66 65 65 72 396

M 138 119 147 172 168 184 928

N 53 58 52 62 55 59 339

Total births

775 708 788 804 860 912 4847 Total herds

6 6 6 6 6 6

Grand Total

1296 1269 1345 1323 1463 1514 8210 * Not participating in scheme at time

20

2.1.2 Parameters: Dairy Herd Performance

The performance of a dairy herd is assessed by analysis of a range of measures, which are of

three main types:

• Farm Performance Parameters (production parameters, stocking rates, meals fed, farm

nutrient profiles and disease incidence);

• Calving-related biological parameters (sex ratio, multiple births, calving difficulty

(dystocia), perinatal mortality, etc.);

• Fertility-related biological parameters (calving interval, calving to first service interval,

six-week calving rate, females not calved in period)

Comparison of these measures with those of other herds can allow the identification of potential

underlying problems or concurrent disease, many of which may reflect production, metabolic or

environmental stresses. In this regard, comparative information representing the national dairy

herd population was obtained from ICBF website (www.icbf.com/wp/) and from Teagasc

National Farm Survey Results (www.teagasc.ie).

Performance data collected by the field officer during monthly farm visits were inputted and

analysed through the Teagasc Dairy Mis 5 Programme (Crosse, 1986).

Field Observations

Herd performance data for each herd in the study were collected on a monthly basis by field

officers, based on interviews with individual herd owners during monthly visits to each

participating farm. This included information on farm management, milk yield and fertility data

for cows and calf production (Table 2.2). In addition, each herd owner provided details of any

unusual events or incidents on or off the farm as recorded in their Farm Incident Diary.

For each Target and Control herd the following data were recorded (where available) during each

monthly visit to participating farms (Table 2.2).

21

Table 2.2: Production parameters recorded for the cows and calves in the participating herds

Herd Performance Fertility Calving Data

Stocking Rate

(LU/hectare)

Fertiliser Use (N, P, K

kg/hectare)

Milk Yield per cow

(litres)

Butterfat and protein per

cent

Concentrates fed per cow

(kg)

Cause of mortalities

Calving to calving interval

(days)

Calving to 1st service interval

(days)

Six-week calving rate %

Proportion of females not

calved in each period

Birth events

Calves alive at birth

Sex ratio

Multiple births

Abortions

Stillbirths

Neonatal deaths

In addition, the monthly Farm Incident Diary for the month preceding the visit was signed off by

the participating herd owner or his/her representatives and collected by the field officers. The

Farm Incident Diary is a key component of the study, in that such records could provide a link

between events that may have occurred and anomalies in the herd performance, health or

environmental factors, thus allowing the causes of such anomalies to be investigated.

22

2.1.3 Farm Performance

In comparative farm performance studies of this type, the herd rather than the individual animal

is the experimental unit. For the study period, there were five Control herds which changed over

the report period and six Target herds, and statistical comparison of farm performance

parameters for individual Target and Control herds was not considered relevant due to the small

sample size (see Chapter 1.6). Comparisons with National Farm Survey Results (NFS) are also

presented for the period 2011 – 2016.

2.2 Results & Discussion

2.2.1 Farm Performance Parameters

Table 2.3: Mean stocking rate (livestock units per ha), fertiliser use(kg) per hectare, meals (kg) fed per cow, milk yield per cow (Litres), butterfat and protein per cent and milk solids per kg/cow, for Target and Control herds and for Teagasc NFS Dairy Enterprise herds for

the years 2011-2016

Group SR Fertiliser Kg Meals fed Yield per Butterfat Protein Milk

Per ha. Solids LU/ha N P K Per

Cow(L) % % Kg/cow Control 2.51 82 4.9 11.7 896 6,006 4.01 3.51 465 Target 2.64 96 7.1 17.4 873 5,582 4.09 3.55 439 NFS 2.37* n/a n/a n/a 974 5,190 n/a n/a 372

*Top one third of herds n/a = not available

Stocking Rate / Fertiliser Usage

The mean stocking rate in livestock units (LU) per hectare overall for the Control herds was 2.51

compared to 2.64 for the Target group in the period 2011 to 2016 (Table 2.3 and Figure 2.1).

Over the period of the report both Control and Target groups stocking rates were highest in the

Control herds in 2011 and in 2011and 2016 for the Target herds (Figure 2.1). Inclement weather

conditions and feed shortages in 2012 and 2013 probably led to a reduction in stocking rates for

both groups.

23

Figure 2.1: Average stocking rate (LU/Ha) for Control (C) and Target (T) herds 2011-2016.

There was variation from farm to farm and year to year in the amount (kg/ha) of Nitrogen (N),

Phosphorus (P) and Potassium (K) used (Figure 2.2). The mean amounts of N, P and K used over

the six years were 82, 4.9 and 11.7 kg/ha for Control herds, and 96, 7.1 and 17.4 for Target

herds.

Figure 2.2: Average Fertiliser inputs for Control and Target herds for the years 2011-2016.

24

Concentrate (Meals) Fed

The Teagasc Road Map for dairy production has set performance indicators for farms for 2025.

For concentrates fed the target is <750kg per cow and for milk yield it is >5,420 litres per cow.

The mean amount of meals fed per cow for the Control group was 896 kg compared to 873 kg

for the Target group. The mean concentrate use for herds in the Teagasc National Farm Survey

for the 2011-2016 period was 974 kg per cow. The average meals fed per litre of milk produced

is shown in Figure 2.3. The amount varied from 0.12 to 0.18 kg per litre over the study period

and compares favourably to the 0.17 to 0.20 kg recorded for NFS herds for the same period.

Figure 2.3: Average meals fed(kg) per litre of milk produced for Control(C) and Target (T)

herds 2011-2016.

Milk Yield

The mean milk yield for the Control herds over the six-year period was 6,006 litres per cow per

annum, compared to 5,582 litres per cow per annum for Target herds. The milk yield per cow for

Control and Target herds in the 2001-2004 report was 5,820 litres and 5,664 litres, and for the

2005-2010 report was 6,767 litres and 5,525 litres, respectively. In comparison, the average milk

yield for Teagasc NFS herds for 2011-2016 was 5,190 litres per cow and milk solids of 379 kg

per cow. The milk solids content of the Control and Target herds was much higher at 462 and

435 kg per cow respectively (Table 2.3 and Figure 2.4).

25

Figure 2.4: Average concentrates fed (kg) and milk yield (Litres) per cow for Control (C)

and Target (T) herds 2011-2016.

Butterfat and Protein Content

The average percentage (%) butterfat and protein for the Control and Target herds for the six-

year study period were 4.01 and 3.51 and 4.09 and 3.54 respectively. Both were similar across

the years for each study group (Figure 2.5).

26

Figure 2.5. Mean butterfat and protein per cent for Control and Target herds 2011-2016

Mortalities

The total number of mortalities recorded for the six-year period was 162(2.0%) of which Control

and Target groups accounted for 80 (2.4% of calving cows) and 82 (1.7% of calving cows)

respectively (Figure 2.6). There were a wide range of causes of deaths over the period with

‘accidental’, ‘calving difficulty’ and ‘unknown’ recorded in 15% and 16% and 16% of cases,

respectively, for the two groups. This was similar to that found in earlier AHSS reports.

27

Figure 2.6: Total mortalities per year for Control (C) and Target (T) herds 2011-2016.

Culling

One of the performance parameters requested from the herd owners referred to the number of

animals culled from the herd in the month preceding the visit and the reason for culling. Such

culling records have been discussed in previous reports in the current series, and many similar

studies give detailed accounts as to the primary reasons for culling by herd owners in Ireland

(Crosse & O’Donovan, 1998), in the UK, (Esslemont, 1993) and in Australia (Stevenson &

Lean, 1998).

The main culling reasons from the Control and Target herds are summarised in Figure 2.7 for the

current study period. On average 12% and 15% of animals were culled from Control and Target

herds, respectively. The main reasons, by order of importance, were infertility, mastitis, old age,

low production, lameness, systemic infection and udder conformation. This is similar to previous

AHSS reports and corresponds with culling studies in other countries. Infertility was the most

common cause of culling in both groups and varied from 5% to 9% over the years for Control

herds and from 4% to 14% for Target herds. From year to year there were minor variations in

culling reasons except for one year where a single herd had an outbreak of Tuberculosis and 29%

of the herd was culled.

28

Figure 2.7: Main culling reasons and average culling rate% (expressed as a percentage of

calved cows) for Control (C) and Target (T) herds for 2011-2016.

While it is evident that data on culling could potentially contain useful information in relation to

environmental quality, experience has shown that decisions on culling by herd owners are

subject to a variety of considerations, including commercial ones, e.g. quota restrictions.

Consequently, data on reasons reported for culling are frequently lacking in objectivity. For

example, decisions to cull animals from herds could be influenced by secondary reasons and

factors such as age, breed and temperament (Bascom and Young, 1988).

2.2.2 Calving-related Biological Parameters

Data on 8,210 calving events from the five Control and six Target herds over the six-year period

are summarised in Table 2.4. Overall, 96% of calves were born alive, 3.3% stillborn and a

further 0.9% died within 3 weeks of birth. The incidence of twinning, still birth, calving

assistance, difficult calving (dystocia) and the sex ratio of male to female calves were also

determined.

29

Table 2.4: Number of calving records (includes multiple births as multiple records) per year across the Control and Target herds, 2011-2016

Year GROUP Born alive

(%) Still

Birth(%) Dead in 24 hrs

Dead in 48 hrs

Dead 3 days-3 weeks Abortion

Grand Total

2011 C 498 (95) 22 (4) 1

1 522

T 738 (95) 32 (4) 2

3 3 778

2012 C 538 (96) 19 (3) 2 1 1

561

T 671 (95) 32 (5) 1

4 1 709

2013 C 532 (95) 22 (4) 2 1

4 561

T 759 (96) 20 (3) 4 4 1 4 792

2014 C 498 (96) 12 (2) 2 1 6

519

T 778 (97) 21 (3) 3

2

804

2015 C 577 (96) 18 (3) 4 1 3

603

T 827 (96) 24 (3) 4

5 3 863

2016 C 582 (97) 15 (2) 1 1 3

602

T 865 (95) 36 (4) 7 2 2 2 914

Total C (%) 3225 (96) 108(3.2) 12 5 13 5(0.15) 3368 Total T (%) 4638 (95) 165(3.4) 21 6 17 13(0.27) 4860

Grand Total (%) 7863 (96) 273(3.3) 33 (0.4) 11(0.1) 30 (0.4) 18(0.22) 8228

Sex Ratio

Across the entire dataset, the proportion of males was 52.2%. In the Control and Target groups,

the male sex ratio was 51.3% and 52.8%, respectively.

Fig. 2.8 shows the frequency distribution of calf sex across both groups for the six-year period

and is similar to figures reported in the 2005-2010 report and for figures reported for dairy herds

nationally in Ireland (ICBF) for 2005 to 2010 where 51% of calves were male.

30

Figure 2.8: Proportion of male and female calves for Control(C) and Target (T) herds for 2011-2016.

Multiple Births / Live Births / Perinatal Mortality

The incidence of multiple births across the entire population was 4.06%; the incidence was 3.8%

(range 2.2% to 5.6%) and 4.3% (range 2.7% to 5.7%) in the Control and Target herds,

respectively (Figure 2.9). The mean proportion of multiple births appears to have increased

relative to the 2.3% reported in previous studies. The upward trend is more noticeable in 2016

but the rise was similar for both groups.

31

Figure 2.9: Proportion of multiple births for Control(C) and Target (T) herds 2011-2016.

Retained Placenta

There were no cases of retained placenta reported for the Control herds and only nine (0.1%) in

the Target herds over the six-year period.

Stillbirths

The mean incidence of stillbirth in the population was 3.3%. In the Control and Target groups

the incidence was 3.2% and 3.4%, respectively (Table 2.4 and Figure 2.10). This is similar to the

mortality rate of 2.0% for ICBF dairy herds for the 2011-2016 period and lower than the 5.2%

reported in the previous AHSS report. ICBF records for 2005 to 2010 show a 2.1% stillbirth rate

for over 5 million calvings. Many factors can affect the incidence of stillbirths, such as multiple

births, parity of dam, breed of dam, calf breed and sex and, perhaps the most important of all,

calving management where protracted calving and difficult calving have an adverse effect (Mee,

1988).

32

Figure 2.10: Proportion of live births and stillbirths for Control(C) and Target (T) herds 2011-2016.

The incidence of abortion in the present study was 0.15% and 0.22% for Control and Target

herds respectively, and is similar to that found in previous AHSS reports. The declining rate of

abortions reflected in current surveys is presumed to be due to the success of brucellosis

eradication, increased vaccination and improved bio-security and management practices in recent

years.

33

Dystocia (difficult calving)

The prevalence of calving difficulty is presented in Table 2.5 and Fig. 2.11.

Table 2.5: Prevalence (as numbers and proportions) of the different categories of calving difficulty for Control, Target and ICBF herds 2011-2016.

Unobserved Observed

and no assistance

Slight difficulty

Serious difficulty

Very serious

difficulty Caesarean

N % N % N % N % N % N % Control 156 4.6 3070 91.2 125

3.7 10 0.3 4 0.1 2 0.06 Target 541 7.9 4025 82.8 321 6.6 74 1.5 4 0.08 7 0.01 ICBF 13.2 68.7 14.4 2.3 1.4

Figure 2.11: Proportion of Observed Calving events that were Unassisted and Assisted for Control (C) and Target (T) herds 2011-2016.

34

The mean incidence of dystocia for Control and Target herds was 0.46% and 1.6% respectively

(Table 2.5 and Fig 2.11). Dystocia includes ‘serious difficulty’, ‘very serious difficulty’ and

‘surgical intervention’. ICBF records for dairy herds nationally found that up to 68.7% of

calvings were unassisted and dystocia was reported as 3.7% compared to 1.1% and 1.8% found

in the present study and previous AHSS reports.

In Control herds, 4.6% of calvings were “unobserved” compared to 7.9% in the Target herds

(Table 2.6). The “unobserved” calvings could also be classified as “no assistance”, which would

make a total of 95.8% of Control and 90.7% of Target calvings unassisted.

35

2.2.3 Fertility-related Biological Parameters

In seasonally calving (Spring) dairy herds in Ireland, the key breeding objective is to achieve the

highest pregnancy rate in the shortest period after the start of the breeding season to achieve a

concentrated calving pattern in the following year. An underperforming herd can be easily

detected through examination of fertility records available on most dairy farms in Ireland. An

overall comparison with national figures about fertility-related parameters shows that both

Target and Control herds in this study are performing satisfactorily, although improvements in

overall herd performance are achievable based on current Irish dairy research (www.teagasc.ie).

Due to the differing calving patterns (Spring versus Spring/Autumn) across herd groups, many

fertility variables are subject to changing farm management decisions. For example, a farmer

could decide to begin calving earlier next year and thus this year’s calving to service interval

(CSI) would have to be shortened. Interpretation of fertility results should therefore be treated

with caution.

Table 2.6 shows the average calving to calving interval for the individual Control and Target

herds and the comparison with ICBF national herd average. Figure 2.12 illustrates this more

clearly and shows that the Control and Target herds are like the ICBF herds and the trend line

shows a reduction in the calving interval from 2011 to 2016.

36

Table 2.6: Average calving intervals for individual Control and Target herds and

comparison with ICBF herds for the period 2011-2016.

CALVING INTERVAL AVERAGE YEAR 2011 2012 2013 2014 2015 2016 Average

Control Herds A 412 415 406 406 400 387 404 B 452 452 C 416 417 402 390 396 382 401 D 375 384 378 395 391 369 382 E 393 409 395 407 403 391 400 F 401 375 377 387 375 371 381 G 369 372 367 375 369 374 371

Control Year Average 403 395 387 393 389 379 391 Target Herds

I 384 377 361 369 373 366 371 J 392 385 395 379 366 374 382 K 450 413 423 432 395 397 382 L 364 368 355 373 362 366 365

M 377 373 377 372 373 371 374 N 406 398 387 423 364 387 394

Target Year Average 395 386 383 392 372 377 384 ICBF Average 403 397 394 396 392 389 395

Figure 2.12: Average calving interval for Control, Target and ICBF herds 2011-2016.

37

The mean calving to first service interval (CSI) for Control and Target herds for the report period

was 75 and 72 days for Control and Target herds respectively. In 2011, the CSI for Target herds

(80 days) was longer than for Control herds (76 days) but this was mainly due to one spring

/autumn calving herd which had a CSI of 120 days. S/A calving herds would be expected to have

a longer CSI than spring calving herds and removing this herd from the dataset would have

reduced the CSI to 69 days for that year.

Another useful measure of reproductive performance for spring calving dairy herds is the six-

week calving rate (Figure 2.13).

Figure 2.13: Six-week calving rate (%) for Control (C), Target (T) and ICBF herds 2011-

2016.

The six-week calving rate is a measure of the compactness of calving in herds and is affected by

submission rate and conception rate. The results for Control, Target and ICBF herds are similar

over the years.

The proportion of females that did not calve for each year for Control, Target and ICBF herds are

shown in Figure 2.14. The proportions for each group are similar to that for ICBF herds.

38

Figure 2.14: Percent females not calved in each year for Control (C), Target (T) and

ICBF herds 2011-2016.

3. 0 Markers of Effects: Clinical Pathology 33.1 Introduction

3.2 Materials and Methods

3.2.1 Analytical Methodology

3.2.2 Haematology Results

3.2.3 Clinical Biochemistry Parameters

3.3 Discussion

39

3.0 Markers of Effects: Clinical Pathology

3.1 Introduction

The adverse effects of toxins on animal health may be classified under the general categories of

neoplasia, reproductive/developmental defects and clinical and subclinical effects. A wide

variety of target organs may be affected and consequently a wide variety of clinical syndromes

may present. The changes may be subtle and the latency period after exposure to a toxin may be

lengthy, i.e. years or decades (Covello and Merkhofer, 1993).

Cattle have been used as indicator species in several instances of poisoning due to natural or

synthetic toxins (Lloyd et al., 1988; Lloyd et al., 1991; Buckley & Larkin, 1998; Singh &

Srivastava, 2010; Beeby, 2001; Rubes et al., 1997; Burger et al., 2001). The propensity of cattle

to accumulate pollutants such as chlorinated hydrocarbons and heavy metals means that

evaluations of animal blood, milk and tissues can provide more information regarding risk to

human health than the analysis of air and/or soils alone. While none of the blood tests performed

are specific markers of environmental pollutants, blood testing was used in this programme

because it can yield useful information regarding the nutritional and health status of cattle. The

parameters examined included several that may be altered by a variety of disease states and

environmental factors (Sgorbini et al., 2003; Gummow et al., 2006; Perillo et al., 2009;

Batterman et al., 2009).

Haematology and clinical biochemistry tests provide information on selected trace elements and

the nutritional and mineral status of animals, as well as providing information on the

haematopoietic activity of the bone marrow, i.e. the formation of blood cells. These tests

monitor the physiological and pathological responses of animals, as well as their responses to

stress and inflammatory disease. In this regard they can help to identify the extent to which

environmental factors contribute to disease, morbidity, mortality or reduced productivity.

40

3.2 Materials & Methods

For the purposes of this part of the programme, the herd is the unit of investigation, rather than

individual animals within herds. In other words, comparisons of the blood test results from

Target and Control herds were made at the level of the herd as a whole, based on the average of a

representative subset of animals in each herd.

Nine cows were sampled from each participating herd once every quarter year (referred to as

Seasons 1 to 4) between 2011 and 2016 inclusive by the Private Veterinary Practitioner (PVP).

There was no sample collection in the first quarter of 2012 due to suspension of service at the

testing laboratory during refurbishment works. Cows were selected to enter the sampling group

so that equal numbers of young, middle and older cattle were represented. The same cows

generally remained in the group, being blood sampled until removed from the herd. Their

replacements were recruited from the equivalent lactating cows on the same farm.

Venous blood was collected from the jugular or tail vein using the vacutainer system and multi-

sample needle (VacutainerTM; Unitech Ltd, Dublin, Ireland). For each cow a new needle was

used and vacutainer tubes containing the following anticoagulants were filled and mixed

according to the standard operating procedure (SOP); EDTA, lithium heparin, fluoride oxalate

and plain. The animal’s identification mark was written on each tube and the Animal Monitoring

Chart was completed and signed by the PVP collecting the samples. The Veterinary Clinical

Assessment Diary was also completed at this time.

Samples were delivered directly by the PVPs to the Regional Veterinary Laboratory (RVL),

Cork and tested on the same day. On each sample a full haemogram and a range of biochemical

analyses (Table 3.1) were performed.

41

Table 3.1 Total Number of tests carried out by year for each Haematological and

Biochemical parameter.

Year

Parameter 2011 2012 2013 2014 2015 2016 Total Fibrinogen 377 296 395 394 374 391 2227

Haemoglobin 404 296 395 394 375 391 2255 MCH 404 296 395 394 375 391 2255

MCHC 404 296 395 394 375 391 2255 MCV 404 296 395 394 375 391 2255 PCV 396 296 395 396 378 395 2256 RBC 404 296 395 394 375 391 2255

Platelets 360 296 395 394 375 391 2211 WBC 404 296 395 393 375 391 2254

Lymphocytes 404 296 395 394 375 391 2255 Mature Neutrophils 377 296 395 394 375 391 2228

Monocytes 404 296 395 394 375 391 2255 Basophils 44

116 394 375 391 1320

Eosinophils 404 296 395 394 375 391 2255 Band Neutrophils 377 296 395 394 375 391 2228

Total 5567 4144 5646 5911 5627 5869 32764 Total Protein 405 296 395 396 378 395 2265

Albumin 405 296 396 396 378 395 2266 Globulin 405 296 396 396 378 395 2266 A:G Ratio 405 296 396 396 378 395 2266

Urea 405 296 396 396 378 395 2266 Glucose 360 296 396 396 378 395 2221

AST 405 296 396 396 378 395 2266 βHB 405 296 396 396 378 395 2266 CK

18 387 396 378 395 1574

GGT 405 296 396 396 378 395 2266 GLDH 54 287 396 396 378 395 1906

Calcium 405 296 396 396 378 395 2266 Copper (Serum) 405 287 396 396 378 395 2257 GSH-PX Bovine (Units/ml PCV) 378 296 395 396 378 395 2238

Inorganic Phosphorus 405 296 396 396 378 395 2266 Magnesium

(Colorimetric) 405 296 396 396 378 395 2266 Total 5652 4440 6325 6336 6048 6320 35121

42

3.2.1 Analytical Methodology

At the laboratory routine haemograms were determined with the aid of an automated cell counter

(Cell-Dyn 3700®, Abbott Laboratories) and by using conventional techniques. Haematological

examination was carried out using the EDTA sample and, in cases where the EDTA sample was

clotted, the heparinised blood sample from the same animal was tested; if the results were

deemed to be accurate, they were included in the final analysis. A blood film was prepared from

each sample and stained manually using the Rapid Romanowsky Stain kit (TCS Biosciences Ltd,

Botolph Claydon, UK) to allow differential white blood cell counts and examination of the

smears by microscope.

The biochemical analyses were made using an autoanalyser (RX Daytona; Randox Laboratories)

and reagent kits (Randox Laboratories; Sigma-Aldrich). Copper was estimated on serum

samples by atomic absorption spectroscopy using an AA 240 (Varian Inc.). Biochemical analysis

of samples was performed using heparinised whole blood (glutathione peroxidase), heparinised

plasma (urea, AST, GGT, calcium, magnesium and beta-hydroxy-butyrate), fluoride oxalate

plasma (inorganic phosphate and glucose), plain serum (protein, albumin, copper and NEFA) or

EDTA plasma (fibrinogen).

All the blood results were transferred electronically from the analysers to the scheme database

(except for results calculated by the laboratory technician (GPx) which were entered manually.

Laboratory reports were issued to Cork County Council usually within two to three days and

then forwarded from there to the herdowner’s PVP. Results having immediate consequences for

cow survival (Mg) were communicated as soon as they became available to the Veterinary

Office at Cork County Council and then to the farmer.

In addition, blood results were examined and commented on by a clinical pathologist for

individual animals and herds. The reports of these assessments form part of the database of the

study.

43

3.2.2 Haematology Results

The average (± SD) haematology parameters for Control and Target herds are shown in Table

3.2. The average of the haematological parameters fell within the RVL ranges for each year for

Control and Target herds. Again, as in previous AHSS reports, there were effects of season, year

and age to be seen across Control and Target groups. These effects can be seen in Figures 3.1,

3.1a and 3.1b.

Table 3.2: Average (± SD) haematology parameters for Control and Target herds 2011-

2016 and Normal Range used by Department of Agriculture, Food and Marine, Regional

Veterinary Laboratories.

Haematology Parameter Control Target Normal Range

Average ±SD Average ±SD Units

Fibrinogen 4.8 1.3 4.7 1.3 3 - 7 g/l Haemoglobin 10.2 1.2 10.2 1.3 9.7 - 13.7 g/dl

MCH 17.7 1.4 17.5 1.5 13.9 - 21.9 pg MCHC 34.8 2.5 35.0 2.3 27 - 33 g/dl MCV 50.8 3.3 50.1 3.7 41 - 61 fl PCV 0.3 0.0 0.3 0.0 0.23 - 0.42 L/L RBC 5.8 0.6 5.8 0.7 5.3 - 7.9 x 10˄12/L

Platelets 300.0 113.4 329.2 316.2 100 - 800 x10˄9/L WBC 6.6 1.8 6.5 1.6 5.24 - 9.84 x 10˄9/L

Lymphocytes 2.5 0.8 2.5 0.9 2.08 - 5.53 x 10˄9/L Mature Neutrophils 2.8 1.2 2.7 0.9 0.75 - 4.07 x 10˄9/L

Monocytes 0.7 0.2 0.6 0.3 0.02 - 0.69 x 10˄9/L Basophils 0.0 0.0 0.0 0.0 0 - 0.1 x 10˄9/L

Eosinophils 0.7 0.6 0.6 0.5 0.11 - 1.79 x 10˄9/L Band Neutrophils 0.0 0.0 0.0 0.1 0 - 1 x 10˄9/L

The mean Red Blood Cell (RBC) counts remained largely within normal limits when measured

quarterly. Values on and below the minimum limits were associated with age and early lactation

(Figures 3.1a, 3.1b,). In a study on Holstein bulls those in older age groups had lower erythrocyte

numbers than younger or intermediate groups (Monke, et al., 1998). Haemoglobin readings

paralleled the RBC values and were also affected by season, age and lactation stage. These

effects can be seen in Figures 3.1c, 3.1d and 3.1e.

44

Figure 3.1: Average Red Blood Cell (RBC) count by year, sampling season and lower and upper normal range indicators.

Figure 3.1a: Mean Red Blood Cell (RBC) count for Control and Target herds by lactation number.

45

Figure 3.1b: Mean Red Blood Cell (RBC) count for Control and Target herds by week of lactation.

Figure 3.1c: Average Haemoglobin count by year, sampling season and lower and upper normal range indicators.

46

Figure 3.1d: Haemoglobin for Control and Target herds by lactation number.

Figure 3.1e: Haemoglobin for Control and Target herds by lactation number.

47

Figure 3.2: Mean White Blood Cell (WBC) count by Year, sampling season and lower and upper normal range indicators.

Figure 3.2a: Mean White Blood Cell (WBC) count by week of lactation and lower and upper normal range indicators.

48

Figure 3.2b: Mean White Blood Cell (WBC) count by lactation number and lower and upper normal range indicators.

The average White Blood Cell count (WBC) was also within defined limits throughout the

season. WBC showed gravitation to the lower end of the range with parity, and also in early

lactation (Figures 3.2, 3.2a, 3.2b). In a study by Monke, et al., (1998) WBC in yearling bulls was

also found to be markedly greater than for adult bulls. Similarly, Olmos, et al., (2009), in a study

from 3 weeks pre-calving to 7 weeks post-calving found that platelets, WBC and neutrophils

were at their highest values 31 – 39 days post calving in Holstein – Friesian dairy cows. In the

present study, WBC increased to a peak by 16 weeks post-calving and began to drop by week 34

(Figure 3.2a).

49

3.2.3 Clinical Biochemistry Parameters

Mean biochemistry results for the Control and Target herds for 2011-2016 are presented in Table

3.3. As in previous reports, average values were within the normal range over the study period.

Again, there were effects of season, year and age as in previous reports.

Table 3.3: Average (±SD) for Control and Target herds 2011-2016 and Normal Range used

by Department of Agriculture, Food and Marine, Regional Veterinary Laboratories, Cork.

Biochemistry Parameter Control Target Normal Range

Average ±SD Average ±SD Units

Total Protein 73.7 5.4 73.5 5.7 57 - 83 g/l Albumin 34.4 3.0 34.6 2.7 23 - 37 g/l Globulin 39.2 6.2 39.0 5.9 31 - 51 g/l

A/G Ratio 0.9 0.2 0.9 0.2 0.5 - 1.5 Urea 4.9 1.8 4.6 1.9 2.65 - 6.89 mmol/l

Glucose 3.3 0.4 3.4 0.4 2.0 - 3.5 mmol/l AST 111.3 49.2 100.0 34.1 38 - 120 iu/l βHB 0.5 0.2 0.5 0.3 0 - 0.9 mmol/l CK 206.6 494.9 184.3 463.3 50 - 130 iu/l

GGT 27.4 15.2 25.3 25.2 18 - 55 iu/l GLDH 46.4 47.7 33.6 36.9 0 - 25 iu/l

Calcium 2.4 0.2 2.4 0.2 2.1 - 3.1 mmol/l Copper (Serum) 11.8 2.2 12.2 2.1 9.4 - 24 µmol/l

GSH-PX Bovine (Units/ml PCV) 137.9 48.0 140.1 49.4

18.46 - 500 units/ml

Inorganic Phosphorus 1.9 0.4 1.9 0.4 1.4 - 2.5 mmol/l Magnesium (Colorimetric) 0.9 0.1 0.9 0.1 0.8 - 1.32 mmol/l

50

AST is often regarded as a hepatocellular leakage enzyme, but is found in a wide variety of

tissues other than liver, e.g. skeletal muscle. Muscle damage, e.g. due to recumbency in cattle

(downer cows) may result in marked increases in AST. In addition, it does not appear to be a

very sensitive test for lipidosis in dairy cows and should be supported with a specific liver

enzyme test. GGT is a liver specific enzyme and is used to diagnose and monitor hepatobiliary

disease. It is currently the most sensitive enzymatic indicator of liver disease.

Figure 3.3: AST and GGT levels by year, sampling season, lactation number and week of

lactation for Control and Target herds.

51

52

53

AST levels were above the upper range on several occasions for 2014, 2015 and 2016 without

having corresponding increases in GGT where values were more stable and within limits.

Similarly, for week of lactation, AST levels from week 22 to week 36 rose above the upper limit

without a corresponding increase in GGT. The raised AST levels were mainly associated with

the Control herds in both cases and are difficult to explain but perhaps due to changes in dietary

management.

54

Figure 3.4: Inorganic Phosphorous(P) and Magnesium (Mg) levels by year and season of

sampling.

55

Figure 3.5: Copper (Cu) and GSH-PX (Selenium) by year and season of sampling.

56

Most biochemistry parameters were within the reference range with considerable overlapping of

average-by-season results for Control and Target herds. Copper, on the other hand, was close to

the lower reference range for both groups at various times over the six-year period, a finding

similar to the 2001-2004 and 2005 - 2010 reports. Marginally low copper levels have been

found in other studies of dairy herds in the south of Ireland (Mee et. al., 1994).

The one exceptional biochemical parameter that remained above the reference range was

Glutamate dehydrogenase (GLDH). This finding was common to both Control and Target herds.

GLDH is a mitochrondrial enzyme and is used primarily to reflect leakage from damaged or

necrotic hepatocytes. Since it is quite a large mitochondrial enzyme, injury needs to be

sufficiently severe to damage mitochondria. Since the half-life in cattle is approximately 14

hours, persistently elevated serum concentrations reflect chronic hepatic stress and damage.

57

Figure 3.6: Glutamate dehydrogenase (GLDH) by year, season of sampling and week of

lactation.

58

59

The greatest seasonal increases were associated with the Spring and Summer sampling dates.

Further investigation of the data revealed that the average plasma concentration decreased with

parity in both Control and Target herds. This could be ascribed to metabolic stress being greater

in younger animals, though no herd reported any clinical illnesses associated with liver damage

in any age group. It is notable that the control herds had higher concentrations in all parities bar

9th and 10th lactation animals. There were not adequate numbers of animals to examine the data

combining parity with week of lactation or time of sampling. Parallel testing of bulk milk

samples by way of enzyme-linked immune sorbent assay (ELISA) was carried out on a quarterly

basis for fascioliasis. Results were within the normal range. It is generally accepted that assay

results reflect exposure to Fasciola hepatica rather than the presence of active infection. This

point serves to highlight the importance of adopting an overall herd health approach with

attention being paid to the cows in the context of clinical and subclinical disease as well as to

other diagnostic tests including coprological examination.

The significance of this finding lies in a reassessment of what constitutes a “normal” value for

this biomarker in the modern dairy cow and what other factors impact on plasma concentrations.

Had these findings been discovered in the Target herds in the absence of similar results in the

Control herds, it would have been difficult to exclude that possibility that the elevated liver

enzymes were due to chronic exposure to an undetected pollutant. The normal readings for other

(especially hepatic) biomarkers in the herds in question also allowed an appraisal of the GLDH

readings over time within the broad health and production screening that underpins the

programme. It may well be that the reference range for this enzyme will have to be revised for

the modern high yielding dairy cow lest such incorrect inferences be made in future. Such a

process is outside the scope of this study, despite its ability to validate new biomarkers proposed

to screen animal health.

60

3.3 Discussion

The reference or normal range is vital to the clinical interpretation of test results. To be truly

valid, the reference range should be established by the analytical laboratory using the equipment

and techniques that will be used on the test population. Published reference ranges are of limited

use because they are for cattle in general, not for different classes of cattle such as dairy cows,

beef bulls or calves. Test results should not be interpreted in isolation but rather in the context of

other diagnostic findings, including history and the results of physical examination. A clinical

diagnosis or treatment decision should not be based on a laboratory value that is inconsistent

with herd history clinical findings and other laboratory results. The normal ranges used in the

present study were those employed by the Department of Agriculture, Food and Marine

(DAFM), Regional Veterinary Laboratories (RVL).

In the last project report (covering the period 2006 to 2010) we demonstrated that there were

significant effects of season number, year and interactions. Olmos et al., (2009), also found that

sampling time relative to calving had a significant effect on all haematological parameters with

the exception of lymphocyte counts. There is a clear pattern of effect of sampling time (season)

in practically all blood parameters. Other studies in Irish dairy herds have also found similar

effects (O’Farrell et al., 1986; Olmos et al., 2009). Factors which affected results included

season, stocking rate, age, fertilizer nitrogen usage and genotype. Genetic group by time

interactions were found for most haematological parameters (Olmos et al., 2009). Interaction

effects are frequently seen where milk yield is affected by season, age, level of nutrition and

somatic cell count. However, with few exceptions, mean results were within the reference

ranges provided by the testing laboratory.

There was no evidence, on the basis of comparison of Control and Target herd results in this or

previous studies, that there was any adverse effect of location on clinical-pathology parameters.

Pesticides

l Aldrinl Chlordanel DDTl Dieldrinl Endrinl HCBl Heptachlorl Mirexl Toxaphene

Industrial Chemicals

l Hexachlorobenzene (HCB) l Polychlorinated Biphenyis (PCBs)

Unintentionally Produced by Products

l Dioxins l Furans l HCB l PCBs

The DirtyDozen

Persistent Organic Pollutants

44. 1 Background

4. 1.1 Biological and Ecological Significance

of Dioxins and Dioxin-like PCBs

4. 1.2 Dioxins in Cows’ Milk

4. 2 Objectives

4. 3 Materials and Methods

4. 3.1 Milk Sampling Procedure

4. 3.2 Laboratory Testing

4. 3.3 Recent Re-evaluation of “Toxic Equivalency Factors”

for Assessment of Levels of Dioxins, Furans & Dioxin-like PCBs

4. 4 Results

4. 4.1 Dioxins, Dioxin-like PCBs and Marker PCBs

4. 4.2 FSAI / Cork Co Co Biannual Study on Dioxins

and PCBs 2008 to 2015

4. 5 Discussion

4. 0 Markers of Exposure: Persistent Organic Pollutants in Bovine Milk

61

4.0 Markers of Exposure: Persistent Organic Pollutants (Dioxins, Furans and Dioxin-Like Polychlorinated Biphenyls (PCBs) and Marker PCBs) in Bovine Milk

4.1 Background

Persistent organic pollutants (POPs) are a group of toxic chemicals that persist in the

environment, bioaccumulate in the food chain and can be transported long distances mainly by

air and water. They are alternatively termed ‘micropollutants’ as they may exhibit toxic effects at

very low concentrations.

The most significant POPs in terms of potential for unintentional releases are dioxins and PCBs.

“Dioxins”, as referred to in Regulation (EC) No. 1881/2006, cover a group of 75 polychlorinated

dibenzo-p-dioxin (PCDD) congeners (closely related chemical substances) and 135

polychlorinated dibenzofurans (PCDFs) congeners, of which 17 are of toxicological concern.

Polychlorinated biphenyls (PCBs) are a group of 209 congeners based on the biphenyl molecule,

which is composed of two benzene rings, of which 12 are similar to dioxins in their chemical

forms and toxicological properties and are termed “dioxin-like PCBs”. The other PCBs do not

exhibit dioxin-like toxicity, but have a different toxicological profile of generally less concern.

Dioxins are not produced intentionally and have no known use except for research and analytical

purposes; the main potential sources for these pollutants in Ireland are accidental burning of

vehicles, buildings, traffic emissions, backyard burning of domestic waste and emissions from

domestic heating, industry and power generation (EPA, 2012). Polychlorinated biphenyls have

been replaced with non-PCB alternatives over the last two decades, PCB containing applications

included e.g. certain sealants used in construction, ballast fuses in fluorescent light and as the

“oil” in electrical transformers.

Public concern about the possibility of these substances being produced and/or emitted by

industrial facilities in the Cork area led to the inclusion of sampling and monitoring for any

potential accumulation of these substances within the present study. The dataset, which now

spans 25 years, addresses these concerns directly.

62

4.1.1 Biological and Ecological Significance of Dioxins and Dioxin-like PCBs

Animal and in vitro studies have identified toxic effects such as endocrine dysfunction,

immunotoxicity and carcinogenicity which may be attributed to long term exposure to dioxins,

furans and dioxin-like PCBs (E.U. Scientific Committee on Food, 2000). Because of production

and use of these substances in previous decades, PCBs and dioxins are now ubiquitous and are

detectable in most environments worldwide (EPA, 2012; EU Scientific Committee on Food,

2000). Studies in various countries have found trace quantities of dioxins in the environment,

including the atmosphere, soil, plants, wild and domestic animals and humans.

Ninety percent of the non-occupational human intake of PCDDs/PCDFs and of dioxin-like PCBs

is ingested in the diet, with <10% of intake attributed to inhalation and other routes. Dairy

products along with fish and seafood can be significant sources of dioxins in the human diet,

contributing, in the case of dairy products, to up to 40% of daily intake. Dairy products, eggs

and meats have been significant sources of dioxins and PCBs for humans in animal feedstuff

contamination incidents in the past (O’Donovan et al., 2010).

4.1.2 Dioxins in Cows’ Milk

The concentration of these compounds in bovine milk is dependent on their concentration in

pasture or other feed consumed by lactating animals. Dairy cattle have an average lifespan of

nine years in the study region. They are subject to local environmental pollution through

ingestion and inhalation, with the potential for bioaccumulation of lipophilic chemicals,

including dioxins, in their tissues. Dairy cattle occupy an important position in the human food

chain in terms of both milk and meats, and given that the primary mechanism for dioxins

entering the food chain is through atmospheric deposition, bovine milk is considered to be a

particularly suitable matrix for assessing their presence in the environment and the risk of human

exposure.

63

4.2 Objectives

The objectives of this part of the study were to determine the concentration of dioxins and

dioxin-like PCBs in pooled bovine milk samples taken from Target and Control herds in the

South Cork region for the period of the current study (2010 to 2015), to identify any temporal

trends in these concentrations over a period of years by comparing the current results to those

obtained for the period 1991-2009, and to compare observed levels with data from other similar

studies.

4.3 Materials and Methods

4.3.1 Milk sampling procedure

1991 - 2015

Between 1991 and 2015, samples were collected periodically from the on-farm bulk milk tanks

of participating Target Herds (up to ten herds in any given year) adjacent to chemical industrial

plants in the greater Cork Harbour area and four Control Herds (two to five herds in any given

year) situated in rural, non-industrial locations also in County Cork.

An equal volume of milk for each herd contributed to a single annual pooled sample for the

Target and for the Control Herds. Target herd pooled milk samples were prepared for each of

the years 1991 to 1993, 1995, 1997 to 2015. Control herd samples were taken in the same years

commencing in 1995. Once prepared, the pooled samples were frozen at -70°C and then shipped

overnight to the testing laboratories. In several years two or more pooled samples were prepared

from each category of herd (2008 to 2015 inclusive).

64

4.3.2 Laboratory Testing

Milk samples collected between the years 1991 and 2001 inclusive were tested in 2001 by

ERGO Forschungesgesellschaft mbH, Hamburg, Germany. Samples for the years 2003-2005,

inclusive, were tested in 2005 by Eurofins/GfA mbH, Gesellschaft fur Arbeitsplatz and

Umweltanalytik mbH, Munster, Germany. Both laboratories that provided analytical services for

the period 1991-2005 were accredited according to DIN EN ISO/IEC 17025:2000, had

accredited quality management systems in place and participated in national (German) and

international quality assurance schemes for dioxin and PCB testing. The samples were tested in

accordance with current EU regulations.

Samples for 2006-2007 and 2008-2010 were tested at Food and Environment Agency, UK, in

line with Commission Regulation (EC) No. 1181/2006 and Commission Regulation EC No.

1883/2006.

Samples for 2010 to 2015 were tested at the State Laboratory, Ireland in line with Commission

Regulations (EC) No. 1181/2006 and No.1883/2006.

Following thawing and mixing of the samples, milk fat was extracted and the concentration of

dioxins, furans and PCBs were measured using isotope dilution methods, high performance gas

chromatography (HPGC) and high resolution mass spectrometry (HRMS).

Samples were tested for dioxins (PCDD & PCDF) & dioxin-like PCBs (dl-PCBs) and were

reported based on the sum of toxicological equivalents (Total TEQ [Sum of TEQ PCDD &

PCDF plus TEQ dl-PCBs])1, expressed as pg per g fat and for Marker PCBs (PCBs 28, 52, 101,

138, 153 and 180) expressed as ng/g fat. All results are reported on a fat weight basis and as

“upper-bound”, which sets the concentration of chemicals present in the sample at concentrations

below the analytical limit of detection (LOD) as equal to the LOD. 1 Because real samples containing dioxins are made up of complex mixtures, various systems of Toxic Equivalents have been developed in order to address the problem of reporting of differing toxicities and environmental behaviour of these substances. These procedures use schemes of weighting factors which express the toxicity of each individual PCDD and PCDF in terms of an equivalent amount of the congener 2,3,7,8-TCDD. This weighting factor, called a toxic equivalent factor (TEF), is multiplied by the concentration of the individual compounds in a mixture to give a 2, 3, 7, 8-TCDD toxic equivalent, (TEQ) which is the sum of the concentrations of the individual congeners multiplied by their TEFs. Refer to EPA (2012) for further information on this topic.

65

4.3.3 Recent Re-evaluation of “Toxic Equivalency Factors” for Assessment of

Levels of Dioxins, Furans and Dioxin-like PCBs

Previous reports in this series have employed the WHO-TEF weighting system, which was

developed by the World Health Organization (WHO) in 1998 for the assessment of human risk

arising from exposure to dioxins and other “Persistent Organic Pollutant” substances. This

system was incorporated into EU Regulations (Regulation (EC) No 466/2001 setting maximum

levels for certain contaminants in foodstuffs as regards dioxins and dioxin-like PCBs and

successive amendments thereto) and has also been the basis of reporting for the ongoing EPA

national survey of dioxins in milk.

The WHO carried out a re-evaluation of Human and Mammalian Toxic Equivalency Factors for

Dioxins and Dioxin-like Compounds in 2005. Commission Regulation (EC) No 1259/2011,

which came into force on 01 January 2012, revised limits in line with the WHO re-evaluation by

amending EC 1881/2006. This Regulation has had the effect of reducing the maximum residue

limits for the concentration of dioxins and furans (PCDD and PCDF) in milk and milk products

from the previous value of 3.0 pg WHO-TEQ/g fat to 2.5 pg WHO-TEQ/g fat and the maximum

tolerable concentration concentrations of dioxins, furans and dl-PCB in milk and milk products

from 6.0 pg WHO-TEQ/g fat to 5.5 pg WHO-TEQ/g fat.

These amended TEF values are applied to data from all years to derive adjusted TEQs for the

data in this report.

On foot of EC1259/2011 a limit of 40ng/g fat for the sum of concentrations of PCBs 28, 52, 101,

138, 153 and 180 (ΣPCB) was set for raw milk and dairy products including butter fat.

66

4.4 Results

4.4.1 Dioxins, Dioxin-like PCBs and Marker PCBs

The annual average total dioxin and dioxin-like PCBs results in respect of milk samples from

Target and Control herds for the years 1991 to 2015 are presented in Figure 4.1.

The total dioxin content in milk from Target herds was highest between 1991 and 1992, peaking

at a WHO-TEQ of 1.67 pg TEQ/g milk fat. Total dioxin then decreased to around 0.75 pg TEQ/

g milk fat towards the end of the 1990s and further decreased through the 2000s to a generally

stable level at or below 0.6 pg TEQ/g milk fat, which is 10% of the limit specified in the

applicable EU legislation2. These “annualised” data represent the average concentration where

more than one sample was collected in a given year.

The annual total dioxin concentration in Control milk samples exhibited a similar marked

decrease from around 0.68 pg TEQ / g milk fat between 1995 and 1999, to levels fluctuating

moderately around a yearly average of 0.4 pg TEQ/ g fat between 2003 and 2015.

Over the period since 1995, therefore, total dioxin WHO-TEQ in both Target and Control milk

have exhibited strong downward trends; the decreases have been of the order of 40% between

the late 1990s and the early 2000s and have fluctuated moderately around this new lower level

since then. This downward trend was strongly correlated with the concentration of the dioxin-

like PCB component of total dioxin (Figure 4.2), which decreased markedly over the period of

the study.

Between 1995 and 2009, Control milk levels were on average 20% lower than the comparable

Target values though during the period 2010 to 2015 the total dioxin WHO-TEQ concentrations

for target and control milk samples were, broadly speaking, very similar.

2 Commission Regulation (EC) No 1881/2006 of 19 December 2006 setting maximum levels for certain contaminants in foodstuffs. As mentioned in the text, Commission Regulation (EC) No 1259/2011, which came into force in December 2011, provides revised limits in line with the WHO re-evaluation of Human and Mammalian Toxic Equivalency Factors for Dioxins and Dioxin-like Compounds in 2005; these will be incorporated into future reports in this series.

67

Figure 4.1 Total WHOTEQ in bovine milk samples collected between 1991 and 2015.

Total WHOTEQ values represent the sum of toxic equivalent values for dioxins, furans and

dioxin like PCBs in each sample. For the years 2008 to 2015 when two samples were tested per

year the average concentration is reported in this graph. TEQ values for all samples are based on

the 2005 revision of TEFs by the WHO.

0

0.2

0.4

0.6

0.8

1

1.2

1.4

1.6

1.8

1991

1992

1993

1995

1997

1998

1999

2000

2001

2003

2004

2005

2006

2007

2008

2009

2010

2011

2012

2013

2014

2015

pg T

EQ/g

fat

Total WHOTEQ

Control

Target

0.00

0.20

0.40

0.60

0.80

1.00

1.20

1.40

1.60

1.80

1991

1992

1993

1995

1997

1998

1999

2000

2001

2003

2004

2005

2006

2007

2008

2009

2010

2011

2012

2013

2014

2015

pg T

EQ/g

fat

Target Milk

DL-PCBs

PCDDF

68

Figure 4.2 Contributions of dioxins/furans & dioxin–like PCBs to total WHOTEQ in Target and Control milk samples.

For the years 2008 to 2015 when two samples were tested per year the average concentration is

reported in this graph. TEQ values for all samples are based on the 2005 revision of TEFs by the

WHO.

4.4.2 FSAI / Cork Co Co Biannual Study on Dioxins and PCBs 2008 to 2015 –

Evidence of Seasonal Variation.

Table 4.1 provides data from the Food Safety Authority of Ireland / Cork County Council

Biannual study, from 2008 to 2015 and illustrates the results for Total TEQ, PCDD/F TEQ,

dioxin-like PCBs and Sum of 6 Marker PCBs for each sample taken. Overall, all samples

collected between 2008 and 2015 for dioxin and dioxin-like PCBs were less than 15% of the

maximum tolerable limit of 5.5 pg/g fat established by EU Regulation (EC) No 1881/2006 (as

amended). Generally, for samples taken bi-annually, results for the spring sampling were higher

than results for samples taken in autumn of the same year. There is considerable overlap between

the results for Target and Control samples during the period 2008 to 2015. Between 2010 to

2015 dioxin-like PCBs contributed on average 44.5% and 39.5% of the Total TEQ in Target and

Control milk, respectively.

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1991

1992

1993

1995

1997

1998

1999

2000

2001

2003

2004

2005

2006

2007

2008

2009

2010

2011

2012

2013

2014

2015

pg T

EQ/g

fat

Control Milk

DL-PCBs

PCDDF

69

The concentrations of total dioxins, dioxin like PCBs and Indicator PCBS in the 2008 to 2015

period are presented in Table 4.1. and Figures 4.3, 4.4 and 4.5.

Figure 4.3. Total dioxin concentration in milk samples collected over the period 2008 to

2015. The year suffix on the horizontal axis indicates the season the sample was collected; S=

Spring and A= Autumn.

Table 4.1 Biannual sampling summary results 2008 to 2015 (Total TEQ, PCDD/F-TEQ, DL-PCB TEQ, in pg TEQ/g fat weight; Sum 6 Marker PCBs in ng/g fat weight)

Origin Season Year Total

TEQ PCDD/F DL-

PCB

Indicator PCBs

Target Spring 2008 0.61 0.33 0.28 0.60 Target Autumn 2008 0.36 0.19 0.17 0.59 Target Spring 2009 0.61 0.35 0.27 0.42 Target Autumn 2009 0.35 0.20 0.15 0.39 Target Spring 2010 0.78 0.49 0.28 0.61

Target Autumn 2010 0.49 0.24 0.25 0.59

00.10.20.30.40.50.60.70.80.9

2008

S20

08A

2009

S20

09A

2010

S20

10A

2011

S20

11A

2012

S20

12A

2013

S20

13A

2014

S20

14A

2015

S20

15A

pg T

EQ/g

fat

Total WHO TEQ

Target

Control

70

Origin Season Year Total TEQ

PCDD/F DL-PCB

Indicator PCBs

Target Spring 2011 0.46 0.24 0.22 0.60

Target Autumn 2011 0.31 0.17 0.14 0.41

Target Spring 2012 0.50 0.31 0.18 0.58

Target Autumn 2012 0.35 0.22 0.12 0.43

Target Spring 2013 0.69 0.43 0.26 0.73

Target Autumn 2013 0.33 0.23 0.10 0.35

Target Spring 2014 0.40 0.23 0.16 0.52

Target Autumn 2014 0.27 0.17 0.09 0.38

Target Spring 2015 0.55 0.31 0.24 0.62

Target Autumn 2015 0.29 0.18 0.11 0.38

Control Spring 2008 0.35 0.19 0.16 0.47 Control Autumn 2008 0.24 0.13 0.11 0.33 Control Spring 2009 0.35 0.22 0.14 0.47 Control Autumn 2009 0.32 0.18 0.14 0.29 Control Spring 2010 0.47 0.28 0.19 0.51

Control Autumn 2010 0.37 0.18 0.18 0.42

Control Spring 2011 0.51 0.29 0.22 0.56

Control Autumn 2011 0.46 0.27 0.19 0.48

Control Spring 2012 0.39 0.23 0.16 0.43

Control Autumn 2012 0.38 0.21 0.17 0.43

Control Spring 2013 0.71 0.42 0.29 0.73

Control Autumn 2013 0.32 0.16 0.15 0.33

Control Spring 2014 0.44 0.20 0.24 0.53

Control Autumn 2014 0.33 0.17 0.15 0.40

Control Spring 2015 0.44 0.23 0.21 0.50

Control Autumn 2015 0.43 0.24 0.19 0.50

Table 4.1 Continued. See previous page for legend.

71

Figure 4.4. The dioxin like PCB concentration in milk samples collected over the period 2008 to 2015. The year suffix on the horizontal axis indicates the season the sample was

collected; S= Spring and A= Autumn.

Figure 4.5. The Indicator PCB concentration in milk samples collected over the period

2008 to 2015. The year suffix on the horizontal axis indicates the season the sample was

collected; S= Spring and A= Autumn.

0

0.05

0.1

0.15

0.2

0.25

0.3

0.35

2008

S20

08A

2009

S20

09A

2010

S20

10A

2011

S20

11A

2012

S20

12A

2013

S20

13A

2014

S20

14A

2015

S20

15A

pg T

EQ/g

fat

Dioxin-like PCBs

Target

Control

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

2008

S20

08A

2009

S20

09A

2010

S20

10A

2011

S20

11A

2012

S20

12A

2013

S20

13A

2014

S20

14A

2015

S20

15A

ng/g

fat

Indicator PCBs

Target

Control

72

4.5 Discussion

Over the last twenty-five years milk samples from Target and Control herds have shown

generally similar total dioxin levels for the years where data was available for both, and have

decreased substantially since the commencement of the sampling programme.

O’Donovan et al. (2010) analysed the milk sample data from this study over the period 1991-

2005. They observed that the decreases in total dioxins were strongly correlated with the

decreasing concentrations of dioxin-like PCBs and Marker PCBs. They also found that the

dioxin and furan components of the total dioxins, that is, excluding the dioxin-like PCBs, were

similar in both Target and Control milk samples, implying that the PCDD/PCDF content of the

diets of Target herds were not significantly higher than that consumed by the Control herds in the

study. They observed that PCDD/PCDF levels in both Target and Control samples have been

consistently in the low background range and have remained generally stable over the period of

the study, and concluded that the decrease in total dioxin is mainly accounted for in the general

reduction in environmental contamination by PCBs that has occurred over the last two decades.

The reduction in total dioxins in milk produced in the Cork Harbour catchment and adjacent

areas is similar to the trend that has been observed in the UK and globally (EU SCF, 2000) and

also in Irish cows’ milk; a mean 33% decrease in dioxin levels nationally was noted between

1995 and 2004 in Irish Environment Protection Agency (EPA) surveys (Concannon, 2005).

Average values for total dioxins measured by Concannon since 2000 were 0.45 and 0.35 WHO-

TEQ ng/kg milk fat for Cork Harbour and adjacent rural areas respectively; these values compare

closely with the data for Target and Control milk from the present study.

The national EPA surveys in Ireland indicate that levels of dioxins & furans in milk

representative of both the main Cork Harbour area and nearby rural areas have remained

generally stable since 2000, while concentrations of dioxin like PCBs have exhibited consistent,

though slight, decreases in line with the results of the present study.

Reasons for the observed fall in the concentrations of PCDD/PCDFs and, in particular, PCBs in

the milk of both Target and Control herds between 1991 and 2001 are likely to include the

introduction of regulations banning burning of slack (introduced in 1993) and bituminous coal

73

(introduced in 1995) in the designated area of Cork, the gradual switch over to unleaded petrol

for vehicles and the general increase in engine efficiency that has been achieved over the last two

decades. Also coincident with the marked fall in total dioxin levels was the introduction of the

Integrated Pollution Control (IPC) licensing system by the EPA in 1993 (Berry et al., 2005) and

the tightening of environmental controls that accompanied that development.

A recent Air Emissions Inventory, prepared by the EPA for the UN/ECE Convention (EPA,

2012), showed that dioxin and furan emission levels in 2010 were 41 per cent lower compared

with a similar inventory for 1990 (Hayes and Marnane, 2002). The main contributor to the dioxin

and furan inventory was the “Other Waste sector”, which includes the residential burning of

waste, accounting for 47 per cent of the total estimated emissions to air in 2010 (EPA, 2012).

Some categories of industrial emissions were noted to have fallen substantially over the period,

particularly in respect of the closure of a number of hospital incinerators and the former Irish

Steel site on Haulbowline Island, Cork Harbour, while other sources, including combustion

emissions associated with energy use in the residential, industrial and power generation sectors,

have remained consistent in terms of dioxin releases (EPA, 2012).

The decrease in total dioxin levels observed in the period between the two inventories is also

reflected in the results of present study, and was strongly linked to a decrease in PCB levels;

while the changes already mentioned over the past two decades have contributed to this,

probably the main factor in relation to that reduction was the banning of the manufacture and use

of PCBs for industrial applications and the introduction of European legislation3 requiring the

cessation of use of PCBs in industrial and power distribution infrastructure.

The additional perspective provided by biannual sampling shows clearly that there are regular

and repetitive fluctuations in the concentrations of these chemicals between seasons. This could

reflect changes in the concentration of dioxins and dioxin like substances in the diet between

seasons, possibly due to different atmospheric deposition rates associated with weather

conditions. A more detailed consideration of housing/grazing arrangements, diet and dates of

sample collection would be required to explore this phenomenon further. Another potential

3 notably Council Directive 96/59/EC of 16 September 1996 on the disposal of polychlorinated biphenyls and polychlorinated terphenyls (PCB/PCT) and Regulation (EC) No 850/2004 on Persistent Organic Pollutants implementing the Stockholm Convention on Persistent Organic Pollutants.

74

explanatory factor is mobilisation of the body burden of dioxins accumulated by cows and stored

in body fat. Cows in the early weeks of each lactation increase milk yield significantly, reaching

a peak in milk yield approximately six weeks post calving. The increase in yield places

significant energy demands on the cow during that period and some cows will enter negative

balance and will utilise stored body fat to provide the energy and fat required. Mobilisation of

body fat also mobilizes the lipophilic dioxins and PCBs present in the fat. As the majority of our

herds have Spring or Spring-Autumn calving patterns, the seasonal Spring peak in dioxin and

PCB concentrations could therefore be related to the larger proportion of early lactation animals

contributing to the bulk tank at that time of year. It is import context to note that the Spring peaks

in concentrations seen are still well below the maximum tolerable limits (MTL) permitted for

both total dioxins (<15% of the MTL) and PCBs (<2.5% of the MTL) in marketable milk.

55.0 Conclusion

75

5.0 Conclusion

The study described in this report was designed to assimilate baseline data and to assess, using

biological markers, whether any insidious or acute adverse effects were evident in Target dairy

herds, located close to industrial plants in the South Cork region.

The baseline data on herd health, productivity and tissue residues assimilated and interrogated

during this study period were applied as Biological Markers which can be classified as follows:

• “Markers of Performance”, i.e. variations in production outputs, presence of inter-current

disease and background metabolic disease levels;

• “Markers of Effects”, i.e. signals of tissue dysfunction, e.g. liver enzyme activity,

morbidity, mortality, haematological, physiological and clinical findings;

• “Markers of Exposure”, i.e. dioxin/ Polychorinated Biphenyls (PCB) levels in bovine

milk.

Markers of Performance

In this study the farm performance parameters and, in particular, the data generated for mean

stocking rate, meals fed per cow and milk yield indicate that the Control and Target herds were

intensively managed and had outputs and performance in keeping with similar herds in the

National Farm Survey. Cows in the Control herds yielded 6,006 litres compared to 5,582 litres

for the Target herds. Milk solids output (kg fat and protein/cow) were similar for Target and

Control herds in this study period and 73kg greater than the average for NFS herds for the same

period.

76

Markers of Effects

The proportion of multiple births and sex ratio, which can be influenced by environmental

contamination, and perinatal mortality were similar to previous years and in line with national

values.

Both Control and Target herds in this study recorded a marginally higher incidence of still-births

than the NFS figures, with the incidence being similar for Control and Target herds.

While none of the blood tests performed were specific markers for possible environmental

contaminants, parameters that could be altered by a variety of disease states and farm

management practises have been identified. A number of authors (O’Farrell et al., 1986; Olmos

et al., 2009) have noted variations in the interpretation of metabolic profiles in dairy herds

associated with age, season, stocking rate, nitrogen usage and genotype. The effect of sampling

time and season affected practically all blood parameters in this study.

Overall there was no evidence, on the basis of comparison of Control and Target herd results that

there was any adverse effect of location on the clinical pathology parameters examined.

Markers of Exposure

The results for dioxins, furans and dioxin-like PCBs in Target and Control milk samples taken

over the period 2006 - 2015 and included in the present study were within the range recorded

from other sites in Ireland (Concannon, 2012). In addition, they were significantly less than the

applicable limits set by the EU in Commission Regulation (EC) No 1881/ 2006 and also the

recently enacted Commission Regulation (EC) No 1259/ 2011. The reducing trends in total

dioxins observed over the study period are largely accounted for by the reduction in PCB

contamination, while dioxins and furan levels have remained generally stable at values

considered as low background levels in European terms. The minor differences in observed

concentrations between Target and Control milk are consistent with the relatively greater degree

of urbanisation of the Target farms.

77

Future Options

The Cork Lower Harbour Animal Health Surveillance Programme began in 1991 and continued

until the withdrawal of funding for the programme in December 2017. The programme has

delivered an almost uninterrupted stream of data from farms in proximity to the chemical plants

in the Cork Harbour area. It has compared this data on an ongoing basis to similar data from

selected herds outside of this area and also to data from other national production monitoring

systems.

While the continuation of the programme is outside the remit of its original proposers, designers,

governors and participants, it remains to be seen if the value they placed in it will resonate at

national level and lead to its restoration. It is difficult to envisage how the loss of the information

the scheme provides will lead to anything other than the public uncertainty that led to its origin

in the first place.

A questionnaire of herdowners was undertaken whereby scores of 1(very poor) to 5 (very good)

were taken to elicit their views of the scheme. In relation to the ‘provision of reassurance

regarding your animal’s ongoing health’ and ‘reassurance regarding you and your family’s health

and residents in the area’, both scored 3.9 out of 5. In relation to farm management issues,

herdowners felt that the scheme ‘provided additional useful information on herd health planning’

(Score 3.7), ‘provided access to production monitoring systems’ (Score 3.8) and ‘provided early

warning of metabolic (Score 3.2) and parasitic diseases’ (Score 3.7). Additional comments

included ‘that the AHSS had a key ‘watch dog’ effect’ and that ‘the Regional Veterinary

Laboratory gave excellent feedback’. Other comments related to the issue that the Scheme

‘provides reassurance to the local community and to food processors in the area’. Herdowners

also felt that there was an overwhelming sense of opportunity cost if the AHSS was to cease in

its entirety.

Letters of support for the scheme continuing have been received from several eminent

individuals. Dr. Patrick Wall, Professor of Public Health, UCD, stated that ‘that it is now well

acknowledged that the health of the environment, the health of animals and the health of people

are inextricably linked’. Dr. Riona Sayers, Herd Health Senior Research Officer, Teagasc,

Moorepark, stated that ‘the Cork harbour monitoring scheme provides a method of future

78

proofing the industrialised Cork harbour region and provides continuous reassurance to both the

locality and to industry’. Tim Lucey Chief Executive of Cork County Council said “We see the

Cork Harbour Monitoring Scheme as an important element in our armoury, demonstrating that

we are all doing all in our power to monitor the impact of industrial developments and have an

additional early warning system in place that could alert to possible adverse human or

environmental health effects associated with new and existing facilities in the region.

The problem of managing environmental protection in tandem with sustainable development

presents a continuing challenge to planning authorities. The integration of environmental

protection within EU policy linked to local community socio economic activities has been

described as” the main modality for ensuring sustainable development”. (Brandon et al., 2005)

A 2016 EPA report found that Ireland’s levels of ammonia and nitrogen oxides are now in

breach of EU emission levels and are 90% attributed to fertiliser usage and animal manures. The

EPA report concluded that these levels would “cause damage to air quality and health”(EPA

2016).

Clusters of urbanisation, population growth, increased traffic density, tourism footfall and

associated activities are well documented as environmental pressures. The Indavar planning

proposal for a hazardous waste thermal treatment facility at Ringaskiddy has proven to be

controversial and there is currently a perception of increased environmental risk in the area.

Much of the data provided in this report is directly relevant to the “One Health Concept” which

states that “human health and animal health are interdependent and are bound to the ecosystems

in which they exist” (World Health Organisation, 2017).

Appendix 1: Glossary/Definitions/Abbreviations

Appendix 2: Dr. Riona Sayers, Herd Health Senior Research Officer,

Teagasc, Moorpark

Appendix 3: Target and Control Herdowner Survey

Appendix 4: Dr. Patrick Wall, Professor of Public Health, UCD

Appendix 5: Mr. Tim Lucey, Chief Executive, Cork County Council

References

Authors

6. 0 Appendices and References 6

79

6.0 Appendices and References

Appendix 1

Glossary / Definitions / Abbreviations

Bioaccumulation: biological pathways by which organisms accumulate elements or compounds

(Markert, 2007).

Bioconcentration: direct uptake of contaminants from the physical environment through tissues

and organs (Markert, 2007).

Bioindicator: an organism or part of an organism or a community of organisms that yields

information on the quality of the environment by their absence or abundance (McGeoch and

Chown, 1998; Beeby, 2001; Markert, 2007).

Biomagnification: absorption of contaminants from nutrients via the epithelia of the

gastrointestinal system (Markert, 2007).

Biomarker: is a measurable biological change occurring in an organism as a result of exposure

to a contaminant, and usually relates to a biological change at the cellular, biochemical,

molecular or physiolocial level. Biomarkers tend to be measured in cells, body fluids, tissues or

organs (Markert, 2007).

Biomonitor: an organism or part of an organism or a community of organisms that yield

information on the quantitative aspects of the quality of the environment by examination of

impairment of their function or performance (Beeby, 2001; Markert, 2007).

Biosensor: a measuring device consisting of a biological entity (e.g. a bacterial cell, an enzyme)

coupled to a physical transmission device which produces a proportional signal in response to

contact with a contaminant (Markert, 2007).

Dose: the amount of a contaminant deposited or absorbed in the body over a given time period

(Hatch and Thomas, 1993)

Emission: a substance discharged into the air, especially by an internal combustion engine.

80

Exposure: concentration of a contaminant at the boundary between an individual and the

environment as well as the duration between the two (Hatch and Thomas, 1993).

Sentinel : a biological monitor species that accumulates a pollutant in their tissue without

significant adverse effects. Used primarily to measure the amount of a pollutant that is

biologically available. A sentinel is an animal that is used to measure pollution exposure (or

effect) in a particular species as a measure of the ambient levels of pollutants in an area (Beeby,

2001).

Sustainability: development that meets the needs of the current population without

compromising the ability of future generations to meet their needs (Vatalis, 2010).

Waste : any substance which constitutes scrap material, an effluent or other unwanted surplus

arising from the application of any process or any substance or article which requires to be

disposed of which has been broken, worn out, contaminated or otherwise spoiled. Such

substances will remain waste until it has been fully recovered and no longer poses a potential

threat to the environment or to human health (European Directive 75/442/EC).

Animal Related Definitions & Calculations

Abortion: the expulsion of a non-viable foetus following a gestation period of less than 260 days.

Anaemia: the presence of a below normal red blood cell (RBC) count, haemoglobin

concentration (Hb), and/or packed cell volume (PCV).

Calving assistance: the amount of farmer/veterinary assistance administered during a calving.

Degrees of calving assistance are scored as (1) unassisted, (2) slight assistance, (3) serious

difficulty, (4) very serious difficulty, (5) caesarean section, (6) other reasons/foetotomy, and (7)

unobserved. (Category 6 removed from this analysis).

Calving interval (CI): the average time interval between successive calvings for the same cow

Calving to first service interval (CSI): the average interval from calving to when the animal

received her first service by artificial insemination.

Dystocia: a term used to describe varying degrees of birthing difficulty.

81

Fertiliser input: calculated annually per hectare based on fertiliser purchase information supplied

by the farmer. Used as an indicator of soil quality on a farming terms of its ability to grow

pasture.

Malpresentation: when a foetus is not presented in the correct position for delivery i.e. the

normal ‘head and two legs’ presentation is not present.

Meals fed: calculated based on the total amount of concentrates fed annually divided by the

average herd size. Is essential information in judging dairy farm management, pasture quality

and herd performance.

Milk yield: calculated based on all milk sold from the farm including the amount estimated to be

fed to calves divided by the mean number of cows in the herd for the year. Valuable indicator of

herd performance.

Multiple births: occurrence of twins, triplets.

Perinatal mortality: calves that were stillborn or died within 48 hours of calving, following a

gestation period of at least 260 days.

Retained foetal membranes: when foetal membranes were not completely voided within 12

hours of calving.

Services per cow (NS): the number of times an animal was served by artificial insemination

during a single breeding season in order to conceive.

Stillbirth: a calf that was born dead, or died within a few minutes of calving, following a

gestation period of at least 260 days.

Stocking rate: calculated based on the farm area divided by all stock on the farm converted into

livestock units (LU). Is essential information in judging dairy farm management and

determining whether herd performance parameters may be affected by either under- or over-

stocking.

General Abbreviations (listed alphabetically)

Alb: Albumin

AHSS: Animal Health Surveillance Scheme

82

AST: Aspartate Aminotransferase

BHB: β-hydroxybutyrate

BSE: Bovine Spongiform Encephalopathy

BVD: Bovine Viral Diarrhoea

C: Control

Ca: Calcium

CPL: Clinical Pathology Laboratory

CSI: Calving to first Service Interval

Cu: Copper

CVI: Calving to Calving Interval

DAFM: Department of Agriculture, Food & the Marine

dl-PCB: Dioxin-like Polychorinated Biphenyls

EC: European Community

Eos: Absolute Eosinophil Count

EPA: Environmental Protection Agency

EU: European Union

FIB: Fibrinogen

FSAI: Food Safety Authority of Ireland

GEE: Generalised Estimating Equations

GGT: Gamma-glutamyl transferase

Glob: Globulin

Gluc: Glucose

GPx: Glutathione Peroxidase

Hb: Haemoglobin

83

HPGC: High Performance Gas Chromatography

HRMS: High Resolution Mass Spectrometry

ICBF: Irish Cattle Breeding Federation

IT: Information Technology

IPC: Integrated Pollution & Control

IPPC: Integrated Pollution Prevention & Control

K: Potassium

LOD: Limit of Detection

LU: Livestock Units

Lym: Absolute Lymphocyte Count

MCHC: Mean Cell Haemoglobin Concentration

MCV: Mean Cell Volume

Mg: Magnesium

N: Nitrogen

NFS: National Farm Survey

P: Phosphorus

PCB: Polychorinated Biphenyls

PCDD: Polychorinated Dibenzo-p-dioxins

PCDF: Polychorinated Dibenzofurans

PCV: Packed Cell Volume

PLA: Platelet Count

POP: Persistent Organic Pollutants

Prot: Total Protein

PVP: Private Veterinary Practitioner

84

RBC: Red Blood Cell Count

ROI: Republic of Ireland

RVL: Regional Veterinary Laboratory

S: Spring

SA: Spring/Autumn

T: Target

TEF: Toxic Equivalency Factors

UCD: University College Dublin

VDCCC: Veterinary Department of Cork County Council

WBC: White Blood Cell Count

WHO: World Health Organisation

WHO – TEQ: World Health Organisation Toxic Equivalency

85

Appendix 2

ANIMAL & GRASSLAND RESEARCH &

INNOVATION CENTRE

MOOREPARK, FERMOY, CO. CORK, IRELAND

Tel: +353 25 42222 Fax: +353 25 42340

e-mail: mk_reception.mfrc@teagasc.ie

web: www.agresearch.teagasc.ie/moorepark

3rd November 2017.

Re: Cork Harbour Monitoring Scheme.

I have been asked to furnish my opinion regarding continuation of the Cork Harbour Animal Health Surveillance Scheme. I am doing so in my role as an independent research biochemist, veterinarian, and epidemiologist with Teagasc who has previous experience of the Cork harbour monitoring scheme.

Cork harbour remains one of the most industrialised areas in the Republic of Ireland and encompasses a dense collection of pharmaceutical manufacturing plants. In this regard, it is incumbent on both industry and regulatory authorities to conduct on-going and routine surveillance to monitor environment quality and potential public health risks. Data generated from such surveillance acts to identify longitudinal trends in environmental quality and provide assurance to local residents of the safe operation of industries in the region. Benefits to industry also accrue, as positive actions taken by companies to ensure environmental quality and safety to public health are highlighted and supported on an on-going basis.

Environmental monitoring is not a short term endeavour and baseline data must be continuously collected. Reasons for this include the monitoring of natural changes to the environment independent of industry impact, changes to industrial work practices which may impact the environment and local population either positively or negatively, and post-incident/accident investigations. The type of environmental monitoring method chosen to achieve effective surveillance is very much dependent on locality, the type of industry present, and what the scheme hopes to achieve. Different regions will chose different means of surveillance and include different biomarkers for investigation. As Ireland is first and foremost a food producing nation, a scheme which monitors food-producing animals provides an appropriate means of surveillance. Such a scheme not only monitors general animal health but livestock

86

act as an early warning system for environmental change that may impact human health. Additionally, the scheme provides assurances of safe food for human consumption. In this regard, it is noteworthy that a short survey of farmers, currently involved in the Cork harbour surveillance scheme, clearly outlines the assurance that the scheme provides.

As with any long-term longitudinal study, the Cork harbour scheme should be reviewed regularly to ensure that it remains ‘fit for purpose’ for industry, local residents, and regulatory authorities. A core set of analyses, such as clinical pathology and infectious disease testing, should remain part of the scheme on an on-going basis, however. Regardless of its design, the Cork harbour monitoring scheme provides a method of futureproofing the industrialised Cork harbour region and provides continuous reassurance to both the locality and to industry. It is my opinion, therefore, that discontinuation of the Cork harbour surveillance scheme would be short-sighted in the extreme.

Please do not hesitate to contact me should you require any further clarification of the points I have raised.

Yours sincerely,

_______________________

DR. RIONA SAYERS

BSc, MAnSc, MVB, PhD, Dip PM, CEDIII

Herd Health Senior Research Officer

AGRIC, Teagasc, Moorepark,

Fermoy, Co. Cork, Ireland.

Tel: +353 (0)25 42215

e-mail: riona.sayers@teagasc.ie

Web: www.teagasc.ie

87

Appendix 3

Herdowner Survey

1 How would you score the Animal Health Monitoring scheme in terms of protecting animal and human health?

( 1 = very poor 2 = poor, 3 fair, 4 =good, 5= very good)

Target & Control herds. JM RG NC AB DJ TC MB KD IA Mean Provides reassurance regarding your animals’ ongoing health

4 5 4 5 4 3 4 4 2 3.9

Provides reassurance regarding you and your family’s health as residents of the area

5 5 4 4 4 4 5 3 1 3.9

Provides reassurance to the local community that there is an early warning system to detect possible industrial threats directly to human health

3 3 4 5 2 4 3 4 1 3.2

Provides reassurance to the local community that there is an early warning system to detect possible industrial threats to the food supply

3 2.5 4 4 2 3 5 3 2 3.2

Provides reassurance to food processors regarding their compliance with legislative and commercial obligations

4 3 3 3 4 4 3 4 2 3.3

Provides the State with a method of demonstrating the safe co-existence of pastoral agri-food production and industrial development

4 1 2 3 2 4 3 5 1 2.8

Provides the industries in the region with third party confirmation that their environmental protection measures are effective

4 N/A 2 N/A N/A 3 2 4 N/A 3.0

88

How would you score the Animal Health Monitoring scheme in terms of farm management issues?

( 1 = disagree strongly 2 = disagree, 3 neutral, 4 =agree, 5= agree strongly)

Provides useful additional information for your own herd health planning (e.g. IBR, BVD)

5 5 3 5 1 4 3 4 3 3.7

Provides free access to production monitoring systems (e.g. Teagasc, ICBF, Private Vet)

4 3 4 5 2 4 4 5 3 3.8

Has provided early warning of parasitic diseases

3 3 3 4 5 5 3 3 4 3.7

Has provided early warning of metabolic diseases

4 2 2 3 4 3 4 3 4 3.2

Being a member adds considerably to your workload on bleeding dates

3 1 1 2 2 3 2 4 3 2.3

Being a member adds considerably to your workload organising post mortems

1 1 1 1 1 3 1 3 4 1.8

Being a member adds considerably to your workload regarding slaughter sampling

1 1 1 2 N/A 1 1 3 5 1.9

Being a member adds considerably to your workload regarding adverse incident reporting

1 1 1 1 2 1 1 4 3 1.7

Suggested improvements/Comments 1. Herbage sampling should be included 2. Key is the “watch dog” effect. 3. General lack of trust of Pharma/Chemical industry 4. Liaise with GE Global group 5. Communication with PVP’s need to be improved, eexcellent feedback from Regional Veterinary Laboratory. 6. Flagging system for cull cows sold for further feeding 7. Must plan to minimise duplication of record keeping 8. Should insist in signing up for Herd Plus and Herd Performance report and Profit monitor 9. Sustainable Quality Assurance Scheme. 10. Iodine testing should be included 11. Farm Nutrient Plan should be utilised as data source. 12. Should use Anonymised herds from ICBF as Controls.

89

Appendix 4

90

91

Appendix 5

92

93

References

Anon. (1995). Environmental Protection Agency 2nd Interim Report (October 1995 – December 1966) Investigations of Animal Health Problems at Askeaton, Co. Limerick. Published by: Environmental Protection Agency, Ardcavan, Wexford, Ireland.

Anon. (1995a) Waste Management No. 107, record breaking civil hearing ends in relief for Rechem, ENDS Report 245 June.

AWN Consulting (2015), Ecological Risk Assessment for PCDD/F for Indaver Ringaskiddy Resource Recovery Centre. Appendix 15 of the Natura Impact Statement. http://www.ringaskiddyrrc.ie/pdfs/Natura_Impact_Statement/NIS_Appendix_15_-_Ecological_Risk_Assessement_for_PCDD_F_Issue_1.pdf (accessed 10th of September, 2018) Brandon, P. and Lombardi, P. (2005). Evaluating Sustainable Development: in the Built Environment. https://books.google.ie/books?hl=en&lr=&id=ZOVqPbT0hiIC&oi=fnd&pg=PR6&dq=main+modality+for+ensuring+sustainable+development&ots=_5yim0-Pj_&sig=soaszqiWulQ0TwSd7T1-yvFoLKQ&redir_esc=y#v=onepage&q=main%20modality%20for%20ensuring%20sustainable%20development&f=false

Bascom, S.S and Young, A.J. (1988). A summary of the reasons why farmers cull cows. Journal of Dairy Science 81:2299-2305.

Batterman, S., Chernyak, S., Gouden, Y., Hayes, J., Robins, T., Chetty, S. (2009). PCBs in air, soil and milk in industrialized and urban area of KwaZulu-Natal, South Africa. Environmental Pollution. 157, 654-663.

Beeby, A. (2001). What do sentinels stand for? Environmental Pollution, 112, 285-298.

Berry, D., Buckley, J., O’Donovan, J., O’Mahoney, P., O’Farrell, K.J. (2005). Animal Health Surveillance of dairy herds in the vicinity of a large chemical complex in the Cork harbour region 2001-2004. Cork County Council Report to the Environmental Protection Agency. http://www.corkcoco.ie/co/pdf/129612327.pdf

Bertrazzi, P.A. (1991). Long Term Effects of Chemical Disasters. Lessons and Results from Seveso. Science of the Total Environment, 106, 5-20. Buckley, J., Larkin, H. (1998). Health surveillance of cattle in the vicinity of a chemical industrial complex. Veterinary Record, 143, 323-326.

94

Buckley, J., O’Donovan, J.V., Berry, D.P., O’Mahony,P., O’Farrell,K.J. (2007). Comparison of production and calving data for 10 Irish dairy herds in the vicinity of an industrial chemical complex and 10 dairy herds in rural, non-industrialised areas. Veterinary Record, 161, 841-845. Burger, J., Gochfeld, M. (2001). On developing bioindicators for human and ecological health. Environmental Monitoring and Assessment, 66, 23-46.

CHASE (2018) http://chasecorkharbour.com/about/ (accessed 21/8/2018)

Clarke, M.L., Harvey, D.G., Humpreys, D.J. (1981) Veterinary Toxicology, Second Edition. Balliere and Tindall, London pp 10,13, & 14-17.

Cork County Council, Cork City Council (2001). Cork Area Strategic Plan (CASP). https://www.localenterprise.ie/SouthCork/Enterprise-Development/Cork-Area-Strategic-Plan-2001-2020/CASP-2001-2020.pdf (accessed 10th of September 2018).

Covello, V.T. and Merkkhofer, MW (1993) “In” Risk Assessment Methods; approaches for assessing health and environmental issues. 3, 1.2.2., 4.1.1., 4.1.4 & 4.1.2

Crosse, J.G. (1986). The Development and Implementation of Computerised Management Information Systems for Irish Dairy Farmers. PhD.Thesis, National University of Ireland.

Crosse, S. and O’Donovan, S. (1998). Dairy disposal rates from commercial dairy farms participating in the DAIRYMIS II computerised management information system in Ireland. Irish Veterinary Journal 42: 75-78

Davies Ian, Vellema Piet and Roger Paul (2012). Schmallenberg virus – an emerging novel pathogen. In practice Nov/Dec, Vol. 34, 598-604.

Department of Agriculture, Fisheries and Food (DAFF), Teagasc, Environmental Protection Agency, Health Service Executive, Kilkenny County Council. (2010). Investigation of animal health problems on a farm near Castlecomer, Co. Kilkenny. An inter-agency report. http://www.agriculture.gov.ie/media/migration/publications/2010/CComerIAGFinRpt231009.pdf

DAFM (2015) Food Wise 2025, https://www.agriculture.gov.ie/foodwise2025 (Accessed on the 27th of August 2018)

Daly, L. & Bourke, M.A. (2008) Chapter 8: Sample Size Determination. In: Interpretation and uses of Medical Statistics, Fifth Edition pp 269-295.

EFSA (2011). Scientific Opinion: Statistical significance and biological relevance. EFSA Journal, 9:2372. https://efsa.onlinelibrary.wiley.com/doi/pdf/10.2903/j.efsa.2011.2372 (Accessed 10th of September 2018).

95

EPA (2012). Dioxin levels in the Irish environment. Ninth Assessment (Summer 2011) based on levels in cows’ milk. Environmental Protection Agency, Wexford, Ireland. http://www.epa.ie/pubs/reports/other/dioxinresults/dioxinlevelsintheirishenvironment2011.html (Accessed 21st of August, 2018)

EPA (2016). Ireland’s Environment 2016 – An Assessment. http://www.epa.ie/pubs/reports/indicators/SoE_Report_2016.pdf (Accessed 10th of September 2018)

Esslemont, R.J. and Kossaibati, M.A. (1997). Culling in 50 dairy herds in England. Veterinary Record 140: 36-39.

European Union Scientific Committee on Food (EU, SCF). (2000). Opinion of the SCF on the risk assessment of dioxin and dioxin-like PCBs in food. http://europa.eu.int/comm/foods/fs/sc/scf/index_en.html

Gummow, B., Kirsten, W.F.A., Gummow, R.J., Heesterbeek, J.A.P. (2006). A stochastic exposure assessment model to estimate vanadium intake by beef cattle used as sentinels for the South African vanadium mining industry. Preventive Veterinary Medicine, 76, 167-184.

Hayes, F., Marnane, I. (2002). Inventory of dioxin and furan emissions to air, land and water in Ireland for 2000 and 2010. EPA, Wexford, Ireland.

ILRM 629 (1988). Hanrahan versus Merck, Sharpe and Dohme (Ireland Ltd.), Supreme Court Decision, July 5th. www.Bailii.org/ie/cases/IESC/1988/1.html.

Kumar, S. (1993) India: the second Bhopal tragedy. The Lancet, 341, 1205-1206.

Lloyd, O.L., Lloyd, M.M., Williams, F.L.R., Lawson, A. (1988). Twinning in human populations and in cattle exposed to air pollution from incinerators. British Journal of Industrial Medicine, 45, 556-560.

Lloyd, O.L., Lloyd, M.M., Williams, F.L.R., McKenzie, A., Hay, A. (1991) Toxicity from ragwort and fat cow syndrome or from industrial chemicals: The value of epidemiological analysis for interpreting clinic-pathological findings. The Science of the Total Environment, 106: 83-96. Elsevier Science Publishes B.V. Amsterdam.

Mance, H. (2018) Animal sentience, politics and why I’m a vegetarian. Financial Times 21st February 2018. https://www.ft.com/content/2fcfcabc-15ca-11e8-9e9c-25c814761640 (accessed 10th of September 2018).

Marienfeld, Carl J. (1979) Detecting Teratogenic Substances by Watching Animal Populations. Contr. Epeidum. Biostatist., Vol. 1, ppp.57-70 (Karger, Basel)

96

Mee, John F. (1988). Calf Mortality in Irish Dairy Herds, Irish Grassland Animal Production Association J. 22, 106-110.

Mee, J. F., O’Farrell, K. J., Rogers, P.A.M. (1994). Base-line survey of blood trace element status of 50 dairy herds in the south of Ireland in the spring and autumn of 1991.Irish Veterinary Journal, 47, 115-122.

Monke, D.R., Kociba, G.J., DeJarnette, M., Anderson, D.E. and Ayars W.H. Jr. (1998) Reference values for selected hematologic and biochemical variables in Holstein bulls of various ages. American Journal of Veterinary Research, 59, 1386-1391

O’Donovan, J.V., O’Farrell, K.J., O’Mahoney, P., Buckley, J.F. (2010). Temporal trends in dioxin, furan and polychlorinated biphenyl concentrations in bovine milk from farms adjacent to industrial and chemical installations over a 15 year period. The Veterinary Journal, 190, 117-121. O’Farrell, K.J., MacCarthy, D.D., Crinion, R.A.P. and Sherington,J. (1986). Metabolic Profiles of Dairy Cows. 1. The effects of season, nitrogen level and stocking rate. Irish Veterinary Journal, 40, 42-52.

Olmos, G., Boyle, L., Horan, B., Berry, D., Sayers, R., Hanlon, A., Mee, J.F. (2009). Effect of Holstein-Friesian genetic group on peripartum and early lactation haematological and acute phase proteins profiles, health and fertility. Animal , 3, 1013-1024.

Perillo, A., Paciello, O., Tinelli, A., Morelli, A., Losacco, C., Troncone, A. (2009). Lesions associated with mineral depositin in the lymph nodes and lungs of cattle: a case-control study of environmental health hazard. Folia Histochemica Et Cytobiologica, 47, 633-638.

Radiological Protection Institute of Ireland (RPII). (1987). Chernobyl; its effect on Ireland. http://www.rpii.ie/Site/Publications.aspx?search=chernobyl&cat=653&page=3 (Accessed 30th January 2012).

Reese R.A. (2004) Does Significance Matter? Significance, 1, 39-40. https://doi.org/10.1111/j.1740-9713.2004.00009.x (accessed 10th of September 2018).

Rubeš, J., Pokorná, Z., Borkovec, L., Urbanová, Strnadová, V. (1997). Dairy cattle as a bioindicator of exposure to genotoxic substances in a heavily polluted area in Northern Bohemia. Mutation Research, 391, 57-70.

Schulte Paul and Mazzuckelli Lawerence F. (1991) Validation of Biological Markers for Quantitative Risk Assessment. Environmental Health Perspectives. Vol.90.pp.239-246.

Singh, N.N. Srivastava, A.K. (2010). Haematological parameters as bioindicators of insecticide exposure in teleosts. Exotoxicology, 19, 838-854.

97

Sgorbini, M., Tognetti, R., Bizzeti, M., (2003). Blood Cd-concentration in dairy cattle living in different area of Cd-pollution. Toxicology Letters, 144, Supplement 1, s137.

Stevenson, M.A. and Lean, I.J. (1998). Descriptive epidemiological study on culling and deaths in eight dairy herds. Australian Veterinary Journal 76:482-488

Winters, M., Jalloh, M. F., Sengeh, P., Jalloh, M. B., Conteh, L., Bunnell, R....Nordenstedt, H. (2018). Risk Communication and Ebola-Specific Knowledge and Behavior during 2014–2015 Outbreak, Sierra Leone. Emerging Infectious Diseases, 24(2), 336-344. https://dx.doi.org/10.3201/eid2402.171028. World Health Organisation (2017). One Health. http://www.who.int/features/qa/one-health/en/ (Accessed 10th of September, 2018).

98

Authors

Cork County Council

Dan Crowley, MVB, BA in Public Administration, MVPH, MRCVS

Chief Veterinary Officer, Veterinary Department, Cork County Council

Jim Buckley, MVB, MVM, MRCVS

Chief Veterinary Officer (Retired February 2012), Cork County Council

Catherine Keohane, Senior Staff Officer, Cork County Council

Programme Co-ordinator

Kevin J. O’Farrell, MVB, PhD, Dip. ECBHM, MRCVS

Consultant in Dairy Herd Health

Bill Cashman, MVB, MRCVS, MVM

Donagh Berry, BAgrSc, PhD

Animal & Grassland Research and Innovation Centre, Teagasc, Moorepark, Fermoy, Co. Cork.

Adjunct Professor to School of Biochemistry and Cell Biology, UCC.

Christina Tlustos, MNS, PhD of Philosophy in Nutritional Science

Technical Executive in Nutritional Sciences at FSAI

Professor Kathleen O’Sullivan, School of Mathematics, UCC.

Roisin Kiely, MVB. NUI

Veterinary Inspector, Cork County Council

Alan Nugent

Clerical Officer, Cork County Council

CorkCounty Council Comhairle Contae Chorcaí Environmental Protection Agency

An Ghníomhaireacht um Chaomhnú Comhshaoil

Animal Health Surveillance of Dairy Herds in the vicinity of a large chemical industrial complex in the

Cork Harbour Region 2011 - 2016 in the vicinity of a large chemical industrial complex in the Anim

al Health Surveillance of D

airy Herds - Cork H

arbour Region 2011 - 2016 Cork County CouncilCorkCounty Council Comhairle Contae Chorcaí

Environmental Protection AgencyAn Ghníomhaireacht um Chaomhnú ComhshaoilVeterinary Department, Environment Directorate

Veterinary Department, Environment Directorate