3. Livestock - Environment Interactions (LEI)

3.1. General description and concepts
3.2. Relevance and magnitude of interactions

3.1. General description and concepts

LLR systems have direct and indirect effects on the environment. The livestock - environment interactions (LEI) can be negative, but also positive interactions occur. The LEI are related to the stage of the production process:

(a) the production and delivery of inputs to LLR systems;
(b) directly to the production process itself; and
(c) the processing and marketing of the products from LLR systems.

a. LEI related to inputs

The main external inputs of the LLR production system are: (1) feed, mainly in the form of concentrates; and (2) livestock for further fattening (or milking).

(1) Feed (concentrate) production.

The effect on the environment of the demand for concentrates has been described extensively by Hendy et al. (1995). The main aspects are:

-The requirement of land for production of concentrates. The environmental impact includes competition with humans for food, changes in land-use, increased land pressure (use of rangeland or forest), effects on soil, water and the atmosphere (use of fertilizers and herbicides/pesticides), and effects on soil erosion as a result of intensified cropping.

-The utilization of crop-residues and by-products of the agro-processing industry. Large quantities of these types of products are consumed in LLR systems and converted into useful high quality food. If not used, the residues and by-products would form a gigantic environmental problem.

-The production of forage for LLR systems. The intensive forage production on a small land area can absorb part of the manure produced by the animals and reduces the land area required for animal production making more land available for wildlife, forest, rangeland etc..

-Energy requirement for transport of feed to feedmills, the feed milling and mixing process, and the transport to the farm. Transport and processing requires fossil energy resulting in emissions to the air/atmosphere of CO₂.

Cattle and lambs are required for fattening in feedlots, and dairy cows and she- buffalos are required for the urban dairies. The main environmental aspects are:

- Cattle and lambs for fattening. Increased numbers of livestock required for fattening put extra pressure on rangeland and/or results in expanded rangeland areas at the expense of wildlife and nature conservation. On the other hand, the lamb fattening system in the Middle East has been developed and is being subsidized to reduce pressure on the rangeland and prevent rangeland degradation.

- Dairy cows. For the urban dairy systems dairy cows come from the rural mixed farming areas. This trade itself does not have a direct environmental effect on the mixed farming system but because the best animals are selected it forms a drain on the genetic resource.

LEI are related to (1) the production of manure; (2) the use of drugs, growth stimulators etc.; and (3) the use of fossil energy.

(1) Manure production.

The main environmental effects are caused by emissions from manure in the stables, during storage, after application to the land or when manure is simply dumped. Emissions are in the form of nitrogen, phosphorus, methane, organic matter and possibly heavy metals. Manure is also a useful fertilizer and an input for crop and pasture production. Biogas production is another form of manure utilization. Brandjes et al. (1995) deal extensively with most of these aspects. The main issues are:

- Manure as fertilizer. The use of manure can improve soil fertility. In certain areas (mainly developing countries) manure is a valuable commodity and sold especially for vegetable production. The small areas of land required for intensive roughage production can absorb a part of the manure from the LLR system.

- Manure for biogas. Manure can be used for the production of energy (Methane) and the remaining liquid slurry can be used as fertilizer. The manure from intensive systems (low roughage consumption resulting in low DM content of the manure) is not very suitable for use as fuel in a dried form.

- Nitrogen (N) emissions are in the form of (i) volatilization of NH₃; (ii) volatilization of N₂ and N_xO_x; and (iii) run-off and leaching of N-compounds.

(i) Volatilization of NH₃ during storage and application on the field. NH₃ emissions cause acid rain and eutrophication of the ecosystem.

(ii) Volatilization of N₂ and N_xO_x occurs in anaerobic situations (in lagoons and after application to the soil as (by)-products of nitrification and denitrification processes. N₂O is especially harmful as it contributes to global warming and breakdown of the ozone layer.

(iii) Run-off and leaching of N-compounds (nitrate etc.) during storage and after application to the land. These compounds can reach the ground water and make the water unsuitable for drinking water, eventually contribute to eutrophication of the surface water.

- Main P emissions are through run-off and causes eutrophication of the surface water. However, if P fertilization is in excess of crop P requirements for longer periods, P saturization of the soil occurs leading to P leaching to the ground water.

- Odours: animal manure contains a number of volatile organic compounds with an obnoxious odour The compounds do not have a direct negative impact on the environment except that they are a nuisance to the people of the surrounding area;

- Methane (CH₄) emission is from two sources in LLR systems (1) direct from the digestive process in the rumen; and (2) from the anaerobic decomposition process of organic matter in the manure during storage. Methane causes breakdown of the ozone layer. The impact domain study Methane will deal with this aspect of environmental impact;

- Heavy metals in the faeces can form a problem where high levels of manure are used as fertilizer, particularly if removal of heavy metals from the land is low, e.g. in case of predominant livestock farms.

The following categories can be distinguished:

- drug residues: residues in animal products following preventive or curative treatment of diseases;

- pesticides residues: pesticides in products following spraying or dipping for controlling external parasites;

- growth stimulators: effects of the use of hormones etc. to stimulate and regulate growth;

- pesticides/herbicides residues: herbicides which enter the system through concentrates and crop-residues and appear as residues in the animal products or in the manure.

Fossil energy is used as energy source for the operation of equipment and transport. Emissions are in the form of CO₂, SO₂ and N_xO_x contributing to global warming and acid rain problems.

(c) LEI in relation to processing and marketing of animal products.

Environmental effects are in the form of waste production from slaughterhouses, tanneries and dairy processing plants and for the use of fossil energy for transport and conservation (e.g. chilling) of products. These aspects are described in the impact domain study 'Environmental effects of animal product processing' by Verheyen et al (1995).

(1) Environmental effects of slaughterhouses; meat and by-product processing.

Emissions are in the form of:

- Solid waste: manure, paunch, hooves, horns and solid slaughter offal and by-products. Most of the solid waste can be composted and used as fertilizer, but may pose a threat to human health and surface water if not treated well.

- Waste water: water from cleaning can contain slaughter offal and by-products (e.g. blood). The waste water mainly contains organic matter. The quality is measured in Biological oxygen demand (BOD) i.e. the quantity of oxygen required to break down the organic matter.

- Volatile compounds: emissions of volatile compounds is mainly from the use of fossil energy, further from the singeing of pig skins for hair removal, and for the further processing of meat (smoking).

For the tanning of 1 ton of raw hides around 300 kg of salts and minerals are required. Chromium is the main tanning agent and at the same time a major polluting factor as it is highly toxic. Emissions are in the form of:

- Solid waste: in the form of scrapings of the raw hide (meat offal etc.), and chromium containing scrapings and cuttings of semi- or fully tanned hides. Discarded leather products also contain ± 3 % chromium.

- Waste water: from cleaning, soaking etc contains organic material, salts and chromium.

- Volatile compounds: emitted to the air mainly from the use of fossil energy and the use of dyes for finishing leather products.

Emissions are in the form of:

- Waste water: containing residues of milk and milk products; whey from cheese production is a major water pollutant in some cases.

- Volatile compounds: emitted to the air mainly the result of use of fossil energy, and limited dust emissions in milk powder manufacturing.

- Whey and other dairy by-products: which can be, and are increasingly, used for feeding calves and pigs.

3.2. Relevance and magnitude of interactions

3.2.1. Introduction
3.2.2. Nutrient excretion and manure management
3.2.3. Methane production
3.2.4. Concentrate demand
3.2.5. Rangelands
3.2.6. Animal genetic resources
3.2.7. Contaminations of LLR-products and food safety
3.2.8. Demand for fossil energy
3.2.9. Waste from processing of animal products
3.2.10. Biodiversity

3.2.1. Introduction

In the description of the magnitude and relevance of LEI the following aspects are important (modified from Willeke-Wetstein et al. 1994):

- the flow of nutrients (input - output, losses, recycling);
- eco-toxic effects in the production system;
- effects on limited resources, e.g. soil conservation, water, genetic resources and others; and
- energy requirements.

3.2.2. Nutrient excretion and manure management

3.2.2.1. Nutrient excretion.
3.2.2.2. Manure management
3.2.2.3. Heavy metals

3.2.2.1. Nutrient excretion.

In Table 5 estimates are given of the total manure production of the different sub-systems within LLR systems. More information on assumptions and sources are given in Annex 1-4. The estimations of the manure production from feedlots in the USA and from veal production in the EU are based on reasonably reliable data on animal numbers, feed rations, etc.. Estimates for sheep fattening in WANA are less reliable as only little information is available on feed rations and animal numbers. Lack of in formation on feed composition (see also Section 3.2.4) and, more important, animal numbers in the EE and the CIS and in the urban dairy system (see Section 2.3.4 and 5) precludes a reliable assessment of the manure production.

Table 5 Estimated annual nutrient excretion in the different LLR-systems5 Estimated annual nutrient excretion in the different LLR-systems (in 10³ ton).

	Total N	Mineral N	Total P
Feedlot - beef fattening	497	348	133
Veal production	29.4	-	3.6
Sheep fattening	18.7	12.7	4.0
LL in EE and CIS	n.a.	n.a.	n.a.
Urban dairies	n.a.	n.a.	n.a.

3.2.2.2. Manure management

Losses or emissions in the form of volatilization, leakage, run-off and dumping occur during storage and application of the manure. Manure storage systems are described by Safley et al. (1992) and Brandjes et al. (1995). In the latter report also an assessment is made for the losses from the various manure management systems.

In the USA solid storage is the most prominent manure storage system. As most feedlots are unpaved, leaching losses may be considerable, depending on the type of soil. Run-off water, containing considerable amounts of dissolved manure particles and nutrients, is increasingly being collected and treated in lagoons, before being discharged to surface waters. Nevertheless, imperfect lining of the feedlots and overloading still results in part of the run-off being discharged without treatment. This treatment in lagoons does not affect the height of the nutrient losses, but reduces the negative effects of run-off on surface water. Slurry systems are beginning to become more important.

Manure surpluses hardly occur at regional level: even on a county basis the manure production/ land-base ratio is low to medium (Fedkiw, 1992; Sweeten, 1994). However, due to the size of the large-scale feedlots where well over 50,000 head is marketed annually, major manure disposal problems do occur. Such feedlots produce about 1000 tons N and 266 tons P, while crop requirement of e.g. high productive maize silage is less than 29 kg P; thus such feedlots require more than 9,000 ha (over which the manure is distributed evenly!) to maintain P equilibrium fertilization. The problems of the enormous land requirements are aggravated by:

- Underestimation of actual fertilization rates: it is not uncommon for some producers to apply 2-5 times more manure than estimated (Wiese, 1992).

- Incorrect habit to base manure application rates on N requirements of crops and N availability in manure (e.g. Clanton, 1992), thus risking P over-fertilization and often not considering N residual effects (Brandjes et al., 1995).

- Lack of manure storage: most crops do not need manure application for a major part of the year, while manure has to be removed after each cycle of fattening (Foster, 1992).

Manure from veal production is exclusively stored in slurry systems. As veal manure has a very low DM content of about 2% and a low nutrient content compared to other types of cattle manure, the fertilizing value of veal manure is very low. The high water content of veal manure also results in high transportation costs. Consequently, veal manure is increasingly treated as sewage water, also facilitated by the centralized production of the veal.

Feedlots for sheep fattening are unpaved and only the solid manure is collected. Losses are through volatilization, leaching and runoff, the latter particularly in places with high and heavy rainfall. Most of the remaining solid manure is well used, often sold to farmers in the area. Only in a few cases are feedlots more densely concentrated, particularly near cities, thus posing some manure disposal problems.

Manure management in EE and the CIS is rather unclear. Firstly, several types of storage systems are important, but the proportional importance is far from clear, particularly for something as vague as the LLR systems. Secondly, manure management is changing. Until recently, manure was wasted on a large scale, often applied mainly on the fields nearby the livestock farm buildings, but also directly discharged to surface waters or dumped on wasteland. Reasons given were the enormous scale of livestock production units, inadequate manure application equipment, cheap fertilizer and/or lack of interest. Underlying reasons are less clear, but partly related to planned economy (van der Graaf et al., 1991). However, in various countries this situation is changing considerably. In some countries, large livestock units are being abolished in the process of privatization, while manure utilization is improving as artificial fertilizer becomes more expensive or unavailable.

In most countries with urban dairies, manure is a highly valuable product, used as fuel or fertilizer. However, because of the high concentrate rations, the DM content of manure is fairly low, rendering it less suitable for fuel while transportation to vegetable production fields, for instance, is also problematic. Where surpluses are common, manure is often discharged directly into the public sewage system, open surface waters or nearby land.

Table 6 presents the different manure storage systems in use in the LRR systems, together with estimates of the nutrient losses.

3.2.2.3. Heavy metals

Heavy metals in LLR systems originate from mineral supplements for P and from fertilizers for crop production. An extensive study in the Netherlands on the presence of heavy metals in animal products and manure indicated that with current manure application rates, Cu, Cd and Zn are the main problem (Heidemij, undated). It is unknown to what extent heavy metals are problematic in LLR systems. It seems reasonable to assume that Cd and Zn concentration in soils will rise to high levels when these soils are overdosed with P. As indicated already by Brandjes et al. (1995), heavy metals are unlikely to pose major problems if fertilization is based on P equilibrium.

3.2.3. Methane production

3.2.3.1. Methane emission from the digestive process of ruminants
3.2.3.2. Methane emission from manure
3.2.3.3. Effect of landless ruminants on global warming

The atmospheric concentration of methane (CH₄), currently about 1.7 ppmv (parts per million by volume) is increasing at a rate of about 1% per year and has more than doubled over the past two centuries. Prior to this doubling, the atmospheric concentration of methane remained fairly constant, at least as far back as 160,000 years.

The increased abundance of methane will have an important impact on global climatic change, tropospheric (ground-based) ozone, and the stratospheric ozone layer. Estimates are that methane contributes to about 20% of the expected global warming from the greenhouse effect, second only to carbon dioxide which contributes about 50% (Safley et al., 1992). Other related to livestock production estimates, however, show a large variation (Khalil et al., 1994). There are two main sources of methane emission related to livestock production:

Production system *)	Deep litter	Manure solids	Liquid/ slurry	Lagoon system
- USA beef	6	88	6	-
- EU veal	-	-	100	-
- EE/CIS **)	5	45	40	10
- WANA mutton	-	100	-	-
- Urban dairy	-	100	-	+
Losses (%):
- N urine	15	100	20	70
- N manure	0	0	0	10
- P manure	0	25	0	10
CH4 emission (%) ***)	5	10	20	90

Sources: this table is based on authors' interpretation of data of various sources, including: Schulte (1993), Brandjes et al. (1995), NRC (1989), Safley et al. (1992), Haskoning (1994).

*) % in each manure system.

**) the assessment of actual situation in EE and the CIS is highly unreliable; the inventory by Safley et al. (1992) is inconsistent, but also inadequate for this study as it does not distinguish in different type of livestock systems.

***) Methane emission is discussed in Section 3.2.3. Data are based on Safley et al. (1992). The realized methane emission is expressed as a percentage of the potential methane emission.

(1) emission from the digestive processes of ruminants; and
(2) emission from the decomposition of animal manure.

3.2.3.1. Methane emission from the digestive process of ruminants

Methane emissions from the digestive process of ruminants depend largely on the crude fibre percentage of the ration: the higher the crude fibre content, the higher the methane emission percentage of the gross energy intake. However, as lower crude fibre contents are nearly always combined with higher total energy intake, the effect per head is often small (see also Table 7).

Different estimates exist for methane emission from beef cattle in the USA. Byers (1994a) estimates that in the intensive system around 80 kg of methane is produced during the production of an 500 kg animal of 18 months old and the total methane emission from the beef industry in the USA is 2.9 Tg (= million metric tons) per year. Johnson et al. (1994) on the other hand, estimate the total methane emission from the USA beef industry as being 3.8 Tg per year based on methane production of different classes of cattle (Table 7). The emission for feedlot cattle in this table is around 39.1 kg per feedlot place, which is much lower than the 65 Kg given by Crutzen et al. (1986), while Khalil et al. (1994) even assume a production of 103 kg methane per feedlot place.

Table 7 Methane production of different classes of beef cattle 7 Methane production of different classes of beef cattle (Johnson et al., 1994).

Class of livestock	numbers million	Days fed per annum	% of diet 1)	Methane production
				Ltr * hd-1 * d-1	Tg per year 2)
beef cows	33.7	365	6.2	262	2.3
calves	38.6	210	6.0 3)	53	0.3
stockers	37.9	150	6.5	202	0.8
feedlot	26.6	140	3.5	153	0.4

1) methane emission estimated as percentage of gross energy intake
2) Tg = million metric tons per year
3) % energy from dry feed
1 ltr CH4 is appr. 0.7 grams.

Little data is available for the methane emission of sheep fattening; all refer to Crutzen et al. (1986) who give an annual methane production for sheep in developed countries of 8 kg and 5 kg for sheep in developing countries and Australia. Considering that the sheep in the feedlot systems in WANA are raised on low to medium roughage rations, an estimate of 3 - 5 kg with an average of 4 kg per annum is probably fair. The estimate results in a total methane emission of 10,960 mt from sheep fattened in WANA.

The methane emission from LLR systems in EE and the CIS cannot be assessed as information is lacking on the type of ration and average live weight.

3.2.3.2. Methane emission from manure

Methane emission from manure depends on the composition of the manure, the storage, and the distribution system. Potential methane emission is closely related to diet composition, higher digestible rations producing a higher potential methane emission. Under similar conditions, the manure of cattle fed on a high-energy corn-based diet will produce about twice as much methane as the manure of the cattle fed on a roughage diet (Safley et al., 1992). Thus the reduction of methane emission from the rumen due to higher digestibility is partially countered by the increased methane emission from the manure. Which part of this potential methane emission is actually emitted depends, among other things, on the type of manure storage (see Table 6).

Safley et al., 1992 estimate that the methane emission from beef cattle waste is 1.4 Tg per year of which 0.26 Tg is from beef cattle in feedlots. This assessment, however, was based on 11.2 million feedlot places, thus methane emission from manure of ca. 10 million feedlot places is approximately 0.23 Tg, or 23 kg per feedlot place.

3.2.3.3. Effect of landless ruminants on global warming

Though information on the contribution of LLR systems to methane is far from complete and estimates vary widely, partly due to large differences in estimated methane emissions from ruminants, some general conclusions may be drawn. Hopefully the "Methane study" will soon produce more accurate information.

First, methane production per head in LLR systems is higher than from ruminants in land-based systems, partly because of the higher methane emission from the manure and because of higher feed intake. However, methane production per kg output is much lower because of the higher animal production levels in LLR systems. The only exception might be when all manure is treated in lagoons from which process the methane emission from manure is 18 times higher than emitted from range animals. Second, though LLR-systems are characterized by high fossil energy utilization, this effect is negligible compared to the effect of the lower methane emission per kg output (see section 3.2.8). Third, manure treatment in lagoons has a large impact on global warming as both methane and N₂O emissions are high, moreover 1 g of N₂O has an effect on global warming equivalent to 15 g of CH₄.

3.2.4. Concentrate demand

Hendy et al. (1995) estimated the total concentrate consumption in LLR systems at 139,443 MT, which is almost 13% of the total world consumption of concentrate. This includes an estimate for consumption of in EE and the CIS of more than 75% of the concentrate consumed in LLR systems. Though we are not in the position to check the values in EE and the CIS, they do not seem abnormal as it is known that high levels of concentrate are (or at least were) common in these countries. Their estimates of concentrate consumption in other countries are higher than expected. This can be due to over estimates of live weights of beef cattle in OECD countries and of fattening periods of sheep in WANA..

Also no distinction was made between basic feeds, specially cultivated for animal production, and by-products, which become available from food processing, industrial use, etc. When waste or by-products, such as inedible grain, soybeancake, molasses and slaughter offal are consumed by livestock, then the burden on the environment from these products will be reduced as part of it will be converted into edible products.

Table 8 gives an estimate of the basic feed and by-product consumption of LLR systems, based on assumptions explained in Annex 2. These estimates are based on average standard feed rations, despite the large differences within each system, often related to differences in feed ingredient prices. Feed requirement estimates for EE and the CIS, and urban dairies are not given due to lack of data. For comparison, Hendy et al. (1995) estimated the concentrate utilization in LLR systems in WANA and OECD countries to be 1.6 and 33 million MT respectively.

As already mentioned in Section 2.3, high concentrate rations common to all LLR systems are a logical reaction to low concentrate prices. Alternative rations in which less basic products are used and with comparable livestock productivity, are indeed available (e.g. Algeo, 1994; Chenost and Preston, 1992; Harb, 1986) but implemented only in few situations mainly due to economic reasons. As shown by Algeo (1994), replacement of grain by roughage would reduce profit per head drastically (US$ 76 per head in his example of beef fattening).

Table 8: Concentrate requirements in LLR systems Concentrate requirements in LLR systems (million tons)

	Feedlot beef	Intensive sheep	Veal
basic feed	14.9	0.63	1.1
by-products	5.0	0.12	0.5
Total "concentrate"	19.9	0.75	1.6

3.2.5. Rangelands

An often claimed positive effect of landless ruminant systems is a reduction of rangeland degradation through lower animal densities and higher off-take rates. The assumed side-effects have been important reasons for subsidizing barley in many WANA countries and for stimulating feedlot systems in many African countries, most of which failed (Simpson, 1988). However, no convincing evidence has been found for this assumed positive effect, probably because total meat production levels are not fixed. Thus, even though an initial reduction in animal numbers may appear, animal numbers on rangelands can also increase as a result of an increase in demand for sheep for the feedlot system (Treacher, 1993).

A more fundamental and theoretical question is whether LLR systems require more or less land, compared to a situation where the same amount of meat is produced in land-based systems, using less feed grain. Considering that in most regions grass productivity (in kg DM per hectare) is around 1.5 - 2 times grain productivity under comparable conditions, differences are likely to be small but probably advantageous in situations where less grain is fed. The exact outcome of such comparison cannot be assessed in general as it is highly dependent on local conditions, e.g. type, yields and quality of alternative feeds, feed ration composition. Ward (1994) made a comparison between the present feedlot system in the USA and a situation where animals were purely range fed. Not surprisingly more feed energy (about 20%) is required per unit beef if animals are purely range fed. However, this comparison ignores the possibility of using by-products and inedible grain as possible supplements which would reduce the estimated difference.

A major exception of the general picture is veal production, as this is based on milk powder rations, being a highly inefficient way to produce meat.

The relevance of the above considerations is limited. Firstly, because the quality of meat from animals fed grain-based rations and veal is often considered to be much higher than the quality of meat from animals mainly fed grass; many LLR systems exist mainly because consumers are prepared to pay the high premiums. Secondly, a reduction in production from these systems in a specific country is not likely to cause a linear reduction in arable land, considering the surpluses of grain which already exist in the EU and North America, while the reduced demand for grain is likely to result in lower yields per hectare and more rotational fallow land.

3.2.6. Animal genetic resources

There are no specific dangers to animal genetic resources from LLR systems, except for the demand for good quality dairy cattle for the urban dairies. These animals are purchased from the rural areas and are normally slaughtered at the end of the lactation, resulting in a negative selection in the dairy herds, especially the dairy buffalo herds, in the rural areas.

3.2.7. Contaminations of LLR-products and food safety

There is much concern about contamination of livestock products, as a result of the use of feed additives and drugs, particularly in intensive systems such as LLR systems, and the negative direct and indirect effect on human health (e.g. Rifkin, 1992; Harrington, 1991). Different types of contamination may be relevant, e.g. antibiotics, antibiotic resistant bacteria, heavy metals, chemicals, hormones and mycotoxins, each with its own problems.

The extensive use of non-therapeutic use of antibiotics to increase productivity to improve the immune status of large herds (in particular after transportation (Walton, 1986)), has come under criticism, because of increasing antibiotic resistance in bacteria which becomes a danger to human health. Most antibiotics are excreted in a relatively short time, and no residues will appear in animal products if recommended withdrawal times are complied with (Hapke and Grahwit, 1987). However, there is evidence that bacteria pathogenic to both animals and man can acquire multiple antibiotic resistance in the gut of farm animals and can be transmitted to man via food or direct contact by farm workers (Willinger, 1987). Moreover, non-pathogenic E.coli can act as a source of resistance to pathogenic E.coli and salmonella, though in practice transfers are rare (Strauch and Ballarini, 1993).

The extent to which antibiotic resistant pathogens are the cause of extensive non-therapeutic use of antibiotics is hardly known, because of difficulties in tracing where a resistant strain originates from. Resistance can also be due to therapeutic use, both in animals and in man. Only few cases have proved that antibiotic resistant bacteria in humans did originate from (non)-therapeutic antibiotic use in animal husbandry (Willinger, 1987).

Because of the potential problems with antibiotic resistant pathogens, only some antibiotics are allowed for non-therapeutic use in animal feeds, but thus far it has hardly resulted in a reduction in the prevalence of antibiotic resistant pathogens. This is probably because no measures have been taken simultaneously in respect to human use, thus still providing a major input to the pool of bacterial resistance (Walton, 1987)

The main public concern about chemicals is related to organo-chlorines such as DDT and lindane, as these compounds are highly persistent. The presence of organo-chlorines in animal fat tissues (see Table 9), where they are mainly deposited, is not only the result of their use against ectoparasites in animal husbandry, but also by their presence in animal feed. In the latter case animals are victims of pesticide use in grain production. As with heavy metals, concentration of organochlorines in livestock products may be higher than that in the feed ration, because several kilograms of feed are used to produce one kilogram of meat. However, this 'biomagnification' is partly neutralized by detoxification, metabolization and excretion of chemicals by livestock, as by all living species (Byers, 1994b).

In most developed countries, levels are well below security levels as organo-chlorines are replaced by other less problematic ectoparasitic drugs, and also as they cease to be used in grain production. In other countries they are still used both in livestock and crop production so high levels of organo-chlorines may occur in livestock products. In India, for example, DDT levels up to 7.2 ppm in animal fat have been measured, compared with levels up to 175 ppm in pulses (Gupta, 1993).

Table 9. Levels of organo-chlorines in animal fat tissues in Central Europe. Levels of organo-chlorines in animal fat tissues in Central Europe (mg/kg fat).

DDT	0.02 - 0.25
DDE 1	0.02 - 0.1
DDD 1	0.01 - 0.06
Lindane	0.01 - 1.0
HCB	0.01 - 3.3
Dieldrin	0.005 - 0.02
PCB 2	0.01 - 0.41

¹: Metabolites of DDT.
²: Polycyclic biphenyls are not pesticides but have a toxicological relevance identical to organo-chlorines

Source: Hapke and Grahwit, 1987

- anabolic steroids (e.g. testosterone, progresteron, zeranol),
- beta-agonists (e.g. clenbutarol, salbutamol),
- corticoids (e.g. cortisone), and
- somatotropines (e.g. BST).

Heavy metal contamination of livestock products, mainly in the kidneys and liver, are rare (Craigmill, 1995; Livesey, 1994). If incidents do occur (mainly with lead and arsenium), they are usually related to high soil intake, which is uncommon in landless systems, and contamination of feed during transport (Livesey, 1994). Regular additions to feed of Zn and Cd (originating from P additions) do not add to these problems.

Mycotoxic problems, such as aflatoxin, are not of major public concern, in contrast to what scientists believe (Craigmill, 1995). The biological effects of mycotoxins include liver damage, and nephrotoxic, neurotoxic, mutagenic, carcinogenic and teratogenic effects (El-Darawany and Marai, 1994; Gupta, 1993). Their impact on livestock is of prime importance; humans can also be affected but it is highly unlikely that livestock products contribute to this. Mycotoxic problems are not considered as problematic in developed countries (Strauch and Ballarini, 1993), mainly due to rigorous screening of feed samples by feed mills, though incidents with high animal losses do sometimes occur (Gupta, 1993). In developing countries mycotoxic problems are likely to be much higher as storage conditions are much more problematic (mould), and where monitoring of feed quality scarcely exists (Gupta, 1993).

Note, data mentioned are not specifically related to LLR systems, as no such division is utilized in monitoring programs. Moreover, most aspects are not specific to landless systems or even livestock systems, e.g.:

- improper use of antibiotic is still a major problem in situations where educational standards are low, veterinary services are scarce and antibiotic products freely available in the markets;

- contaminations with heavy metals are more common in wild animals and fish than in domestic animals (Hapke and Grahwit, 1987), probably related to higher total feed quantities consumed by these animals as they generally become older.

3.2.8. Demand for fossil energy

Intensive livestock production systems are characterized by a high fossil energy utilization. Nearly all research on fossil energy dates back to the seventies s and early eighties, but since then attention has declined with the reduction in fossil energy prices. Now, awareness is growing mainly because of the importance of CO₂ in global warming: the increase in CO₂ is expected to cause about 50% of the global warming occurring in the next half century (Johnson et al., 1994).

Most comprehensive estimates on fossil energy utilization for beef production amounts to 43.9 MJ (10.5 Mcal) per kg of edible beef, i.e. about 18.5 MJ per kg LW (Fox, 1994; Keener and Roller, 1994). Most of the fossil energy requirements are related to fattening in feedlots, accounting for 63% of the total requirements (Fox, 1994), mainly for feed production (ca. 86% of all fossil energy allocated to feedlot fattening). As these estimates are mainly based on data from the seventies, energy efficiency may be improved somewhat by now. These improvements, however, are likely to be small as the improved FCR is already accounted for, while fossil energy utilization in crop production has probably not decreased since a major part is related to fertilizer use and irrigation.

Although no estimates are available for the other LLR systems on fossil energy utilization per kg product, a few general remarks can be made.

Fossil energy requirements for veal production are likely to be much higher than for beef. Firstly, direct fossil energy use is higher due to the more sophisticated housing systems used, often including mechanical ventilation. Secondly, the fossil energy use for feed production is very high, e.g. drying wet substances to obtain milk powder. Brand and Melman (1992) calculated the fossil energy requirements for the Netherlands for 1 kg milk substitute as being 27.83 MJ. With a feed conversion of 1.5 the fossil energy requirements for 1 kg LWG would be 41.7 MJ!

Fossil energy requirements for sheep fattening are likely to be lower, mainly because of the better FCR and, to a lesser extent, also because less fuel and electricity is used in a direct form for farm operations. The fossil energy requirements for the production of one unit feed for sheep is not likely to be much different. The largest proportion is imported from developed countries and the part which is produced locally requires less fossil energy because of lower fertilizer use. However, this low fertilizer use could result in soil mining.

A major issue is how fossil energy requirements will change if livestock production becomes more land-based. The answer is highly dependent on local situations, e.g. alternative feed rations and transport distances. For beef fattening it is indicated that fossil energy requirements will reduce by roughly 16% if the fattening period in feedlots is reduced from 280 to 120 days per animal (Keener and Roller, 1994), while it is reduced by 24% if beef animals are all range fed (Ward, 1994).

These reductions will, however, only have a small effect on global warming. The reduction of 16% for instance, will result in an effect on global warming equal to about 10 g methane per kg LW, compared to a typical methane emission from rumen digestion only of about 150 - 250 g per kg LW².

²A reduction of ca. 3000 kJ is estimated by Keener and Roller (1994), equivalent to 0.075 l gasoline, resulting in an reduced CO₂-emission of 0.31 kg. The relative absorption of CO₂ is ca. 1/30 of methane.

3.2.9. Waste from processing of animal products

3.2.9.1. Waste from the slaughter process
3.2.9.2. Waste from tanneries
3.2.9.3. Waste from dairy plants

It is difficult to assess the contribution of processing to the LEI because:

(1) animals within LLR systems originally come from other systems and it is difficult to know which proportion of the LEI should be allocated to LLR systems and which to other systems;

(2) the emission factors from the different animal processing systems are highly variable and are largely dependent on the 'house keeping practices' of the processing plant; and

(3) the data regarding quantities produced and quantities processed is unreliable or unavailable..

3.2.9.1. Waste from the slaughter process

Beef production in the USA: Total LWK (live weight kill), attributed to LLR systems, is estimated at 3.55 million tons to be processed in slaughterhouses. Average emission factors per ton LWK for the slaughter process are estimated at (based on Verheyen et al., 1995): 5 kg BOD, 0.3 N-Kj and 0.1 kg P in waste water. Emissions into the air cannot be estimated, as these depend too much on the energy sources used.

The total emission from slaughtering amounts to around 17.8 million kg BOD, 1.07 million kg N-Kj and 0.36 million kg P in waste water. Most of this waste water may cause major problems as slaughterhouses have a large capacity, but it is unknown to what extent waste water treatment is applied. Probably primary and secondary treatment is widely done, which mainly removes a fair amount of BOD, while tertiary treatment to remove nutrients is still to be implemented in most waste water treatment plants.

Veal production in the EU: Total LWK, attributed to LLR systems, is estimated at 1.08 million tons (disregarding the initial live weight of the calf, which is to be attributed to the land-based system). Based on similar estimates as above, the total emission from the slaughter process of veal in OECD countries, mainly the EU is: 5.4 million kg BOD, 0.32 million kg N-Kj and 0.11 million kg P in waste water.

Intensive mutton production in WANA: Total LWK, attributed to LLR-systems, is estimated at 0.22 million tons. Emission factors of the slaughter process in WANA should be estimated somewhat higher than for the USA and the EU, even though less meat processing is apparent. Per ton LWK estimates are assumed to amount to: 10 kg BOD, 1.0 kg N-Kj and 0.3 kg P in waste water. Total emission amounts to 2.1 million kg BOD, 0.22 million kg N-Kj and 0.06 million kg P. These figures are based on slaughtering in slaughterhouses. However, it can be assumed that a fair proportion of the sheep is slaughtered at home or in small butcheries. In that case waste production is difficult to estimate, and the figures are probably higher. Concentration of the waste water production may be lower in this case though the LLR-sheep-system mainly supplies cities, thus also this waste water production is likely to pose environmental and sanitary problems.

3.2.9.2. Waste from tanneries

Beef production in the USA: The quantity of raw hides is around 8% of LW, i.e. 0.28 million tons. Verheyen et al. (1995) estimate the emission factors per ton raw hides from the tanning process at 550 kg solid waste (of which 40% contains Cr), 100 kg BOD and 5 kg Chromium. Emissions to the air depend on the energy sources used during the tanning process. Further emissions to the air of volatile organic compounds are related to the finishing of the products and the paints used; these emissions depend on the end products.

Total emissions from the tanning of 284,000 tons of raw hides are around 156,000 tons solid waste (of which 62,480 tons contains Cr), and waste water containing ca.28,400 tons BOD and 1,420 tons Cr. As tanneries are known for their waste water production it is likely that major part of this water is treated, particularly in OECD countries. In most cases, however, considerable part of the Cr is still discharged to surface waters.

No emission factors are known related to the tanning of sheepskins, but many problems around tanneries are common, sanitary as well as environmental..

3.2.9.3. Waste from dairy plants

Milk is produced within the LLR systems in the urban dairy system. Most of the milk is traded directly from here without further processing to consumers and shop keepers, and only a small portion is processed in factories. The waste in dairy plants in this system is negligible.

3.2.10. Biodiversity

LLR systems mainly affect biodiversity indirectly. Direct negative effects occur locally and are related to emissions of nutrients and organic matter to surface water. Indirect effects include those as a result of the production of animals and feed for the LLR systems. The effects on biodiversity of rearing young stock and breeding animals, mainly on pastures, will be described in the impact studies on grassland based livestock systems in temperate zones and (semi-)arid tropics and subtropics. Effects of producing feed are mainly related to the conversion of potential nature area into land for forage and grain production, which has an enormous negative impact on biodiversity (de Wit et al., 1995). One of the principal questions is to what extent are total land requirements increased by livestock production in LLR systems and compared with what. As noted already in Section 3.2.5., land requirements of the present situation are probably only slightly higher compared to a system in which less grain is fed. However, the present requirements are substantionally higher when compared to a situation where ruminants mainly utilize "waste" biomass, as the positive effect of livestock production on human carrying capacity is only revealed at relatively low levels of livestock production (Kaasschieter et al., 1992).

The relevance of such comparison, again, is limited as it would imply a major change in the economy and a drastic reduction of the consumption rate of livestock products. Moreover, reduced demand of LLR systems for feed concentrate do not necessarily result in more nature areas. Given the already existing grain surpluses in EU and North America, for instance, and because lower demand for grain will partly result in lower grain yields per hectare and more rotational fallow land, the result will be much less valuable from a biodiversity point of view.