JEQ Journal of Natural Resources and Life Sciences Education
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF) Free
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Related articles in JEQ
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via ISI Web of Science (3)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Casey, J. W.
Right arrow Articles by Holden, N. M.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Casey, J. W.
Right arrow Articles by Holden, N. M.
Agricola
Right arrow Articles by Casey, J. W.
Right arrow Articles by Holden, N. M.
Related Collections
Right arrow Sustainable Agriculture
Right arrow Global Change
Right arrow Ecosystem Management
Right arrow Agricultural Systems
Right arrow Air Pollution
Published in J. Environ. Qual. 34:429-436 (2005).
© ASA, CSSA, SSSA
677 S. Segoe Rd., Madison, WI 53711 USA

TECHNICAL REPORTS

Atmospheric Pollutants and Trace Gases

The Relationship between Greenhouse Gas Emissions and the Intensity of Milk Production in Ireland

J. W. Casey* and N. M. Holden

Department of Biosystems Engineering (Bioresources Modelling Group), Univ. College Dublin, Earlsfort Terrace, Dublin 2, Ireland

* Corresponding author (john.casey{at}ucd.ie)

Received for publication June 8, 2004.

   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 CONCLUSIONS
 REFERENCES
 
European Union agri-environmental schemes aim to reduce the environmental impact of agricultural production, but were developed before consideration of greenhouse gas emissions from agriculture. Life cycle assessment methodology provided a framework for comparing emissions as kg CO2 equivalent per kg of energy corrected milk (ECM) (kg CO2 kg–1 ECM yr–1) and per hectare (kg CO2 ha–1 yr–1) for farms both within and outside the Irish agri-environmental scheme. The agri-environmental scheme farms operate extensive systems from 40 to 120 cows producing between 3032 and 5946 kg ECM cow–1 lactation–1. The cows are fed on grass, conserved silage, and concentrates. Supplementation ranged between 250 and 620 kg cow–1 yr–1. The conventional farms had between 30 and 77 milking cows producing 4736 to 6944 kg ECM cow–1 lactation–1. Supplementation ranged from 400 to 1000 kg cow–1 yr–1. The emissions from each unit were estimated using published emissions factors and possible error was evaluated by using ranges for each factor. Calculated emissions ranged from 0.92 to 1.51 kg CO2 kg–1 ECM yr–1 and 5924 to 8323 kg CO2 ha–1. On average, total emissions from conventional farms were around 18% (p = 0.01) greater than the agri-environmental scheme farms and emissions per hectare (total area required) were 17% greater (p = 0.02) but there was no significant difference (p = 0.335) in terms of emission per unit milk produced. To evaluate greenhouse gas emissions for each farm in terms of the system intensity it was necessary to define a measure of intensification and area per liter of milk produced that was best.

Abbreviations: AES, agri-environmental scheme • CAP, common agricultural policy • ECM, energy corrected milk • EF, emissions factor • EU, European Union • FU, functional unit • GHG, greenhouse gas • GWP, global warming potential index • HA, unit area • ISO, international standards organization • LCA, life cycle assessment • LU, livestock unit • REPS, rural environment protection scheme • TGE, total greenhouse gas emissions


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 CONCLUSIONS
 REFERENCES
 
THE TOTAL greenhouse gas (GHG) emissions associated with milk production (excluding postfarm processing) include sources from enteric fermentation, energy, fertilizer, cereal production for concentrate feed, transport of products to the farm, and manure management on the farm (Cederberg and Mattsson, 2000). Ireland is an unusual country with respect to GHG emissions because it has a small human population but a large per capita emission (United Nations Framework Convention on Climate Change, 2002), 16% of which was estimated to be contributed by farm animals (Environmental Protection Agency, 1998). Any attempt to modify the emissions from the major sources on dairy farms should be assessed in terms of whether the emissions burden is reduced or simply transferred to another part of the production system (Casey and Holden, 2005). In addition to considering emissions management, the overall emissions from the Irish dairy sector will be influenced by trends in intensification. European Union (EU) and national policy, social pressure, and the economic context of dairying will dictate whether Irish dairying becomes more intensive (i.e., more product per unit area and greater input use) or more extensive (i.e., less inputs and a greater area per unit output). This paper describes an analysis of a number of dairy units in Ireland to identify the role of intensification in contributing to GHG emissions per kilogram of ECM produced and per hectare utilized to produce the milk.

Life cycle assessment (LCA) is a holistic tool used to assess the environmental impacts of a system or product from "cradle to grave" (Udo de Haes, 1996). An International Standard (ISO, 1997, 1998) defines LCA methodologies, and to carry out a full LCA it is normal to use a number of impact categories ranging from acidification to energy consumption. Life cycle assessment methodology is appropriate for agricultural systems because it ensures that consideration has been given to the boundaries of the system under investigation (Audsley et al., 1997; Ceuterick, 1996, 1998; Haas et al., 2001; Cederberg and Mattsson, 2000). When used with only one impact category, the LCA methodology facilitates understanding of a specific aspect of a system (Casey and Holden, 2005; Kramer et al., 1999; Flessa et al., 2002), and is a useful tool for examining the role of farm intensification on GHG emissions.

The Common Agricultural Policy (CAP) is used to regulate the production, trade, and processing of agricultural products in the EU. Agri-Environmental Schemes (AES) became accompanying measures to the CAP in 1992 with the intention of promoting farming methods that are compatible with the protection of the environment (European Community, 1999). In Ireland, AES was implemented as the Rural Environmental Protection Scheme (REPS), which was designed to reward farmers for carrying out activities in an environmentally friendly manner. Strict guidelines are set out for nutrient management and habitat conservation (Government of Ireland, 2000). Adopting the REPS scheme is generally taken to result in extensification (lower yield per unit area from less inputs, therefore more area per unit product) because of the limitations placed on inputs and stocking rates. The scheme was developed before serious consideration of GHG emissions from agriculture. The question being addressed by this work is whether extensive production in the dairy sector will result in a greater GHG emissions per kilogram (ECM) produced or per hectare used to produce the milk, and thus work in a manner contrary to the intent of REPS. The consideration of emissions is not limited to the land area of the farm or the geo-political boundary of Ireland. It encompasses all the estimable emissions associated with the system wherever they occur.

The objective of the work presented in this paper was to assess the GHG emissions from case study dairy units including both REPS and conventional production to assess the relationship between production intensity and GHG emissions. The LCA methodology was adopted because it provided a reproducible, objective method of delimiting the production system, defining the GHG emissions and of quantifying the impact in terms of scalable outputs (the functional units: kg CO2 kg–1 ECM yr–1 and kg CO2 ha–1 yr–1).


   MATERIALS AND METHODS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS
 DISCUSSION
 CONCLUSIONS
 REFERENCES
 
Life cycle assessment methodology requires the definition of a functional unit (FU) that is an attribute of the product or system, and is used as a quantitative scalar for comparison purposes. There can be more than one FU because a system may have a number of possible functions but the FU must be definable and measurable (ISO, 1997). In the case of dairy systems the main product is milk; therefore, one FU defined for this study was: the production of 1 kg of ECM during a time frame of 1 yr. Energy corrected milk (ECM) (kg) was determined as:

[1]
where M is the mass of the milk (kg), F is the fat (kg), and P is the protein (kg) (Sjaunja et al., 1990). Milk produced in Ireland has an average density of 1.03 kg L–1. The time frame encompasses a lactation period in Irish dairy production, which is a cycle of 1 yr. To facilitate comparison with area based inventory studies a second FU encompassing the total land area including that used for the production of externally sourced concentrate feed was also used. The total land area to operate the dairy unit was recorded for the farms studied. The additional land area required to supply concentrate feed was calculated from yields of crops included and scaled to the required amount in area to supply the feed.

The global warming potential index (GWP) was used to determine contribution to the greenhouse effect. The index is defined as the cumulative radiative forcing effect between the present moment and a selected time in the future caused by a unit mass of gas emitted in the present. The emissions are measured in terms of a reference gas, CO2 (IPCC, 1996a). The GWP of 1 kg CO2 (with a time span of 100 yr) is 1, 1 kg CH4 is 21, and 1 kg N2O is 310 (Audsley et al., 1997). The total greenhouse gas emissions (TGE) are determined as:

[2]
where mi is the mass (kg) of the emitted gas. The result is expressed in terms of kg CO2 equivalents (kg CO2) (Heijungs et al., 1992). The total impact is expressed as TGE/ECM (kg CO2 kg–1 ECM yr–1) (Casey and Holden, 2005) relative to product output and TGE/HA (kg CO2 ha–1 yr–1) relative to land area used.

The system boundary was defined by the GHG emissions associated with milk production up to the point of transportation away from the dairy unit. The system included the physical limits of the dairy unit and its activities (excluding other nonrelated activities on the farm) and the production of externally sourced inputs that permit the unit to operate (such as fertilizer, concentrate feed, and fuel).

The dairy units investigated did not all run identical production systems so a number of assumptions were made to facilitate comparison. It was assumed all stock other than milking cows and cow replacements were sold at 10 d. The replacement rate was set at 16%. This was achieved using the recorded stocking rate per dairy unit and reducing the land area according to the number of dairy stock. The stock carrying capacity of the REPS units is determined by environmental sensitivities, livestock housing and waste facilities, and a number of other minor parameters (Government of Ireland, 2000). Conventional units can operate at any stocking rate the manager wishes. All the dairy units followed the same basic method of spring calving and utilize a combination of grass, concentrates, and silage for the diet. The number of grazing days available per dairy unit was a function of geographical location but ranged from March to October. Other management practices such as slurry spreading and slurry storage were assumed to be similar for each dairy unit. Slurry was assumed to be stored under slats with 75% spread as liquid in the first week of June and the remaining 25% in the first week of July. A constant pelleted ration formulation was also assumed that consisted of a mixture of crops, molasses, vegetable oil, minerals, and vitamins (Table 1) in descending order of mass contribution. The REPS farmers are legally limited to 260 kg N ha–1 that is compatible with the maximum stocking rate permissible.


View this table:
[in this window]
[in a new window]
  		Table 1. Emissions (and data source) associated with the supply of 1 Mg of concentrate feed to dairy units in Ireland (percentage contribution to total in brackets). The inclusion rate of individual ingredients is commercially sensitive and therefore not detailed at the request of the feed suppliers consulted (Casey and Holden, 2005).

 
Detailed questionnaires were circulated to 52 farms with dairy units in the southern half of Ireland, 13 replies were received, 10 of which provided suitable data. Site visits to the 10 farms, A to J, were performed to ensure the high quality of the data supplied. Exact location details are not presented because the farmers were guaranteed anonymity. Properties of each farm are presented in Table 2.


View this table:
[in this window]
[in a new window]
  		Table 2. Summary of the 10 case study dairy units investigated.

 
The emissions from each dairy unit were estimated using emissions factors (EF) collated by Casey and Holden (2005). To account for uncertainty regarding the EF values for the different production situations a range of values was assumed (Table 3). For methane emissions from enteric fermentation (whole year, not just lactation period) the following rules were used: low output cows (3000–4000 kg head–1 yr–1), 85 to 95 kg CH4 head; low–medium output cows (4000–5000 kg head–1 yr–1), 95 to 105 kg CH4 head; medium–high output cows (5000–6000 kg head–1 yr–1), 105 to 115 kg CH4 head; and high output cows (6000–7000 kg head–1 yr–1), 115 to 125 kg CH4 head. These were reasonable assumptions given the IPCC default values for lactating cows in Western Europe are 100 kg CH4 yr–1 for a cow producing 4200 and 118 kg head–1 yr–1 for a cow producing 6700 kg head–1 yr–1(IPCC, 1996b). The selection of EF range values for other stock was based on both IPCC default values and field data (Lovett and O'Mara, 2003). Range values for CH4 from slurry storage were applied from error ranges in Husted (1994), which were selected as best suited to dairy units with different winter periods and negligible slurry storage over the summer period. Ranges of values for N2O emissions from slurry stored under slats were adapted from the IPCC guidelines (IPCC, 1996b). The uncertainty associated with these values was large. A range of N2O emissions have been suggested (Oenema et al., 1997; Anger et al., 2003; Yamulki et al., 1998) for manure excreted at pasture of between 0.5 and 3.0% of total excreted N. Emissions of N2O are variable because those that are livestock-induced in pasture are related to the stage of growth of the grass and the time of year. The IPCC (1996b) default value of 2% was used because it was expected to be the most reliable. Methane emissions from dung at pasture (derived from Jarvis et al., 1995) were selected because they were most suited to the Irish situation. Range off values for southwest England for N2O and CH4 from slurry application were taken as ±50%. Chadwick et al. (1999) presented values for emissions from slurry application. These United Kingdom data were considered to be most relevant to the Irish situation (Casey and Holden, 2005). Range values for N2O emissions from N application were applied from IPCC guidelines (IPCC, 1996b). A range of ±50% was applied to CH4 emissions from pasture. Collecting yard emissions are very small (Casey and Holden, 2005) and no reliable data for silage storage were obtained (and are thought by the authors to be small) so both are omitted in the calculations. The emissions associated with fertilizer production, shipping, trucking, electricity, and diesel combustion were selected from the limited data available and a ±5% range was used to account for uncertainty. The emissions associated with concentrate feed were adapted from Casey and Holden (2005) and a ±5% range was also used.


View this table:
[in this window]
[in a new window]
  		Table 3. Emission factor (EF) ranges (and data sources) used to calculate greenhouse gas (GHG) emissions from the case study dairy units.

 
The possible error associated with the calculation of TGE/FU (for both FUs) was assessed by taking 2000 random combinations of EF values, from within EF ranges assumed for each dairy unit, to obtain an estimated distribution. Thus, using all minimum or maximum EF values would give an estimate of the extremes of emissions that might occur. Sampling within this space provided an estimate of the possible error associated with the EF uncertainty. Differences between dairy units were assessed using a paired t test (H0, there is no significant difference between the average TGE/FU for the two dairy units) to examine whether trends in average TGE/FU reflected a response that accounted for the uncertainty regarding the EF values.

A number of intensity measures (Table 2) were derived for each farm and the trend in TGE/FU was assessed with respect to the various measures. The significance of linear regression parameter values (t statistic) was used to identify the most appropriate measure of system intensity based on relationship with TGE/ECM and TGE/HA. The percentage contribution for each component of the system was calculated relative to TGE, and simple predictive models for TGE/ECM and TGE/HA were derived using step-wise linear regression with dairy unit properties. The significance of the parameter values was assessed using the t statistic.


   RESULTS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
DISCUSSION
 CONCLUSIONS
 REFERENCES
 
Each of the dairy units used for case study (Table 2) operated a variation of a rotational grazing system employed for dairy production in Ireland. The result of the paired t tests indicated that the 10 dairy units were all significantly different from each other (p < 0.05) in terms of TGE/ECM and TGE/HA (HA, per unit area) (Table 4) with the exception of Farms A and I with respect to TGE/ECM. The contribution of the various emissions sources from the dairy units as a percentage of the total are also presented in Table 4. Overall TGE from conventional farms were around 18% (p < 0.05) greater than the REPS farms, TGE/HA were 17% greater (p < 0.05) for conventional farms, but there was no significant difference (p > 0.05) in terms of TGE/ECM.


View this table:
[in this window]
[in a new window]
  		Table 4. Summary of estimated range of CO2 equivalent percentage contribution of the main emissions components TGE/ECM and TGE/HA for each dairy unit.{dagger}

 
Of the REPS farms, Farm C has 31% greater milk output per cow than Farm D at the cost of more N input but less concentrate feed. The land area and number of cows required to operate the system on Farm C are significantly greater than any other farm in the survey, but the TGE/ECM was the lowest of all farms surveyed. Farm A used a relatively large amount of concentrate feed to achieve a balanced production system. The relatively low output of milk per cow means that the TGE/ECM is relatively high. Farm B operates with a very low output of milk per cow, the lowest of all farms surveyed, giving the highest TGE/ECM and lowest TGE/HA of all the farms surveyed. By rank order (lowest to highest) the four REPS dairy units tended to have greater TGE/ECM than the conventional dairy units (ranked 1, 6, 7, and 10) but lower TGE/HA (ranked 1, 2, 4, and 5).

Of the conventional farms, Farm F is using a very high level of concentrate feed per cow for poorly performing animals (compare with Farmer G). Farms E, G, and J are operating systems with TGE/ECM < 1.0. Farm E had the highest TGE/HA because of a high stocking rate that produces a lot of product from a small area. Farms C (REPS), E, and J are situated in the best agroclimatic location for dairying with the longest outdoor grazing periods. As a result, the concentrates fed and dung stored are relatively small and the systems can support quite high yielding cows. The farm management method and geographic location of Farm C allows relatively low TGE/ECM (ranked 1) and TGE/HA (ranked 5) (Table 4) relative to all the farms investigated. Farm C is producing a large amount of milk from a large area, which is beneficial from both perspectives—the emissions are spread over the large milk output and the large area farmed. Farm G is operating an intensive system, but the high concentrate feed input coupled with high N input per hectare does not yield the lowest TGE/ECM (ranked 4) and causes a high TGE/HA (ranked 7). Farm H is using moderate performing cows with moderate inputs, and perhaps typifies the system that the 1990s average Irish dairy unit is going to move toward; however, the intensity of the system leads to the second highest TGE/HA (ranked 9). Farm I is similar to Farm H, except output per cow is lower yielding similar TGE/HA. Farm J is operating a system very similar to Farm E but with greater output, giving a lower TGE/ECM. However Farm E is operating a more intensive system with a higher stocking rate (Table 2), giving rise to a higher TGE/HA (ranked 10).

To evaluate the GHG emissions for each dairy unit in terms of the system intensity, it is necessary to define a suitable measure of intensification. Measures relative to unit product (ECM) and per unit area (HA) were considered, which were N fertilizer rate (Fig. 1) , concentrates fed (Fig. 2) , grazing intensity (Fig. 3) , and area per unit output (Fig. 4) . The area per unit output intensity measure showed the best relationship and had the advantage of relating the product output to the area used, thus linking the two FU measures.



View larger version (23K):
[in this window]
[in a new window]
  		Fig. 1. The relationship between (a) TGE/ECM* and (b) TGE/HA* and rate of N fertilizer applied. [REPS farms = triangles (Farms A–D). Conventional farms = squares (Farms E–J)]. *Gradient parameter significant at p < 0.05.

 


View larger version (23K):
[in this window]
[in a new window]
  		Fig. 2. The relationship between (a) TGE/ECM* and (b) TGE/HA* and concentrates fed. [REPS farms = triangles (Farms A–D). Conventional farms = squares (Farms E–J)]. *Gradient parameter not significant at p > 0.05.

 


View larger version (21K):
[in this window]
[in a new window]
  		Fig. 3. The relationship between (a) TGE/ECM* and (b) TGE/HA{dagger}, and stocking density. [REPS farms = triangles (Farms A–D). Conventional farms = squares (Farms E–J)]. *Gradient parameter significant at p < 0.05; {dagger}Gradient parameter not significant at p > 0.05.

 


View larger version (20K):
[in this window]
[in a new window]
  		Fig. 4. The relationship between (a) TGE/ECM* and (b) TGE/HA* and area per kg of ECM produced. [REPS farms = pale gray (Farms A–D). Conventional farms = black (Farms E–J)]. *Gradient parameter significant at p < 0.05.

 
The data that had the most significant linear correlation with TGE/ECM (r = 0.88, p < 0.05) was area per kilogram (ECM) produced (Fig. 4a); as the area per kilogram increased so did TGE/ECM. As fertilizer input per unit area was increased TGE/ECM decreased (r = –0.64, p < 0.05) but no relationship was found with stocking rate (r = –0.21, p > 0.05) or concentrate fed (r = –0.15, p > 0.05). In general, the trends suggest that as intensity increases the TGE/ECM decreases, but an examination of the relative positions of the REPS dairy units (indicated by triangles on Fig. 1a, 4a, and 5a) shows that it is possible to operate an extensive type system with similar TGE/ECM as a conventional dairy unit, but being in REPS does not necessarily result in lower TGE/ECM. Expressing TGE/ECM as a function of output per cow (Fig. 5a) reveals a significant linear relationship.



View larger version (23K):
[in this window]
[in a new window]
  		Fig. 5. The relationship between (a) TGE/ECM* and (b) TGE/HA{dagger} and output per cow. [REPS farms = triangles (Farms A–D). Conventional farms = squares (Farms E–J)]. *Gradient parameter significant at p < 0.05; {dagger}Gradient parameter not significant at p > 0.05.

 
The data that had the most significant linear correlation with TGE/HA (r = 0.86, p < 0.05) was stocking rate LU ha–1 (Fig. 3b); as stocking rate increased so did TGE/HA. Positive correlations were also found with fertilizer input (r = 0.80, p < 0.05) (Fig. 1b) and output per cow (r = 0.65, p < 0.05) (Fig. 5b). As area per kilogram (ECM) reduced there was an increase in TGE/HA (r = –0.76, p < 0.05) (Fig. 4b) and there was no relationship with concentrates fed per kg cow (r = 0.07, p > 0.05). The area assessment confirmed the view derived from the ECM assessment that it is possible to operate an extensive system with similar or lower TGE/HA as a conventional dairy unit (Fig. 1b, 3b, 4b, and 5b).


   DISCUSSION
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
CONCLUSIONS
 REFERENCES
 
The two FUs used arose from two approaches to analyzing GHG emissions. From a milk production perspective it is necessary to choose a FU that a strict measure of the function that the system delivers, but from an IPCC inventory reporting perspective, choosing a FU coupled with land area is necessary. The lack of a significant relationship (r = –0.52, p > 0.05) between the FUs arises because the milk production system is not simple; farmers can use many low or high output cows on a large or small area and can compensate with N and concentrates. To find the best production system it is necessary to examine both measures, and as can be seem from Farm C, it is possible to achieve low values for both measures by using high output cows at low stocking rates.

Previous work that estimated the TGE/ECM of the 1990s average dairy unit in Ireland (Casey and Holden, 2005) derived a figure of 1.50 kg (CO2) kg–1 (ECM) yr–1. This figure was not subject to the standardising assumptions used here to compare the various dairy units. When recalculated using the same assumptions the value is 1.25 kg (CO2) kg–1 (ECM) yr–1 and 7421 kg (CO2) ha–1. Legislative drivers (Common Agricultural Policy and EU directives) are pushing farmers to be more efficient than the 1990s average. The 10 dairy units examined yielded a mean TGE/FU of 1.1 kg (CO2) kg–1 (ECM) yr–1 and 6868 kg (CO2) ha–1.

The impact of management on emissions can be seen by examining some of the individual farms. Farm B was operating a dairy unit with very low inputs and outputs resulting in the largest TGE/ECM but the lowest TGE/HA. If this farm increased output per cow without greatly increasing inputs (which should be physically and biologically possible) then it would operate more like Farm C, which has low TGE/ECM and TGE/HA as a result of high output cows at low stocking rate. Farm E is using high output cows, a high stocking rate, and relatively little concentrate feed, which means that the TGE/ECM is amongst the lowest calculated, but the TGE/HA is the highest because the land area used is small relative to the product output. Farms H and I fall between the extremes of Farms B and E. Both use quite high stocking rates that result in relatively high TGE/HA but use moderate output cows so TGE/ECM is in the middle of the range assessed. Farms A, D, and F are all using moderate stocking rates but relatively low output cows so have low TGE/HA values but high TGE/ECM. By comparison Farms G and J are using high output cows at relatively high stocking rates so have low TGE/ECM but quite high TGE/HA. These farmers do not achieve combined estimates of TGE/FU as good as Farm C because of the high N and concentrate inputs needed to compensate for the higher stocking rates.

The results may be compared with emissions of 1.3 kg (CO2) kg–1 (milk) and 9400 kg CO2 ha–1 for intensive dairy units in the Allgau region of Germany (Haas et al., 2001). The higher emissions estimated in the German study may be attributed to different management practices at farm level. The result is quite similar to the 1.03 kg CO2 L–1 (milk) estimated by Howden and Reyenga (1999) for a dairy unit in South East Queensland Australia. However, these data were not collated or processed using the LCA methodology and therefore it is difficult to discern what emissions are included or excluded. Martin and Seeland (1999) observed that a more intensive dairy unit would decrease the methane production per liter of milk produced and the results reported here support this idea. However, as indicated by Phetteplace et al. (2001), the whole system must be considered when both estimating emissions and testing mitigation options, and the LCA approach allows a more comprehensive analysis than simple area-based accounting or just examining a single component within the system.

Limiting N input and stocking rate, as required in the REPS scheme, generally increases the area needed to produce 1 kg ECM (Fig. 4). However, a REPS dairy unit can achieve a similar TGE/ECM as a more intensive dairy unit (e.g., Farm C) but with the advantage of a lower TGE/HA. The important factor in achieving similar emissions to a conventional farm is to have animals with a high annual output of milk but utilizing grass from a relatively large area. The trend in Fig. 5a shows that increasing animal output will have a significant impact on reducing TGE/ECM but in general with current practice in Ireland this seems to be associated with an increase in TGE/HA. Animal numbers in each dairy unit do not directly relate to milk output (r = 0.30, p > 0.05) because of the interaction between feed management (concentrates) and fertilizer for grass production.

Stepwise linear regression using the data in Table 2 indicated that TGE/ECM could best be predicted using a combination of kg ECM yr–1 per cow (M) and kg concentrate fed (C):

[3]
where p < 0.07 for all regression parameters. The simplicity of this relationship can be readily explained by considering the function of the system. The kg ECM yr–1 per cow will dictate the feed demand (MAFF, 1984), which will determine how much grass is needed, and will have to be balanced to the area available, N input, and stocking rate used. Any misjudgment of the system has to be balanced by feeding concentrates. A significant negative correlation (r = –0.83, p < 0.05) between the percentage contribution from enteric fermentation and the percentage contribution of concentrates fed indicates that the relative importance of enteric fermentation decreases as more concentrates are used. The prediction of TGE/HA can be achieved quite well using just kg ECM yr–1 per cow (M) and stocking rate (LU ha–1) (S) (R2 = 0.947) but is greatly improved by including nitrogen (N) and concentrates (C):

[4]
The main strategic management options available to the farmer can be used to estimate the GHG emissions associated with the system on both a product and area basis.

A required reduction in cow numbers in line with Irish national policy (Dep. of Environ. and Local Gov., 2000) will have to be offset by increased output per cow to fulfill national milk quota. To ensure that the overall goal of reducing GHG emissions is achieved, various factors will have to be considered. Dairy units located in areas with longer growing seasons will require less imported feed, can achieve more grass yield with less fertilizer, and have shorter dung storage periods. In such locations greater output per cow will mean fewer cows are needed, thereby decreasing enteric fermentation. However, the potential for a shift of emissions within the system as a result of greater cow output also has to be considered.

Feed demand of higher output cows will mean that more grass (more fertilizer input) and more concentrate feed are required if stocking rate remains the same (as evidenced by the significant relationship between ECM per cow and N application rates (r = 0.763, p < 0.05)). Nitrous oxide emissions, which occur mainly in late autumn/winter/early spring from urine spots and following effluent or fertilizer applications (Flessa et al., 2002) will have to be controlled, perhaps by reducing the amount of N excreted in urine and by good animal and N management. Minimizing N fertilizer by using properly managed effluent, applied only in the amount required and during rapid N uptake by actively growing pasture, should have a positive effect on GHG emissions. The use of low N supplements such as maize (Zea mays L.) silage should permit both increased output per cow and less N in the system (O'Mara et al., 1998), and sourcing locally grown inputs should reduce emissions associated with transport.

These points of discussion raise the issue of how the CO2 equivalent emissions are expressed for a dairy unit. If the emissions are expressed on a land area basis (e.g., IPCC level 1 type assessment; IPCC, 1996a), then the emissions beyond the boundary of the dairy unit have to be considered to reflect the true cost of the system. For national reporting purposes, however, the nonnational emissions can be omitted. From a scientific (as opposed to policy) perspective, an LCA-type approach, considering the whole system and using a FU that is an output of the system is probably more satisfactory for evaluating management options of dairy units. This approach is more likely to yield net reductions rather than transfers to other areas.


   CONCLUSIONS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 CONCLUSIONS
REFERENCES
 
The work presented in this paper is the first case study data GHG emissions from dairy units in Ireland. The application of LCA system inventory methods has permitted the comparison of farms in an agri-environmental scheme with those that were not. The trend found indicates that a move toward fewer cows producing more milk at lower stocking rates is required. Such a move would represent extensification in terms of area but intensification in terms of animal husbandry. The efficiency of the farm is the most important factor in terms of the balance of output per cow and feed supply. Compensating for imbalanced grass supply to feed demand by over feeding concentrates not only erodes profit but also leads to greater CO2 equivalent emissions. The quality of management per se is more important than the intensity of production, but the role of geographical location requires further research, as do the interactions of production viability, economic viability, and GHG emissions.


   ACKNOWLEDGMENTS
 
This work was supported by the Environmental Protection Agency (Ireland) under the Environmental Research Technological Development and Innovation programme and was funded by the National Development Plan (2000–2006). The advice and constructive criticism from the anonymous reviewers was very much appreciated.


   REFERENCES
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 CONCLUSIONS
 REFERENCES
 


Related articles in JEQ:

This Issue in Journal of Environmental Quality

JEQ 2005 34: 403-407. [Full Text]  



This article has been cited by other articles:


Home page
 J. Environ. Qual.Home page
J. W. Casey and N. M. Holden
Greenhouse Gas Emissions from Conventional, Agri-Environmental Scheme, and Organic Irish Suckler-Beef Units
J. Environ. Qual., January 5, 2006; 35(1): 231 - 239.
[Abstract] [Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF) Free
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Related articles in JEQ
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via ISI Web of Science (3)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Casey, J. W.
Right arrow Articles by Holden, N. M.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Casey, J. W.
Right arrow Articles by Holden, N. M.
Agricola
Right arrow Articles by Casey, J. W.
Right arrow Articles by Holden, N. M.
Related Collections
Right arrow Sustainable Agriculture
Right arrow Global Change
Right arrow Ecosystem Management
Right arrow Agricultural Systems
Right arrow Air Pollution

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
The SCI Journals Agronomy Journal Crop Science
Vadose Zone Journal Journal of Plant Registrations
Journal of Natural Resources
and Life Sciences Education		 Soil Science Society of America Journal