Reproductive Issues Arising from Different Management Systems in the Dairy Industry


Author’s address (for correspondence): JF Mee, Teagasc, Animal and Bioscience Research Department, Animal & Grassland Research & Innovation Centre, Moorepark, Fermoy, Co. Cork, Ireland. E-mail:


The objective of this review is to address the reproductive issues arising from different dairy management systems by describing divergent systems and comparing their reproductive outcomes. The increasing global demand for dairy products has led to the majority of the world’s milk being produced in intensive management systems. This intensification has occurred in both zero-grazed (ZG) and in pasture-based (PB) systems, and it has contributed directly to the current decline in dairy cow fertility globally. Given the heterogeneous nature of both ZG and PB systems, comparisons between them in dairy cow reproductive performance need to be treated with caution. In general, cows in ZG systems have higher milk production and better energy balance but more of some animal health problems, lower ovarian activity post-partum, reduced oestrous expression, reduced conception success, and higher culling and mortality rates, than cows in PB systems. Key environmental descriptors affecting reproductive performance within management systems include the type and duration of housing and the pre- and post-partum diet composition. Genetic by environment (GxE) interactions for dairy cow fertility have been detected for some, but not for other, management systems. Given the concerns of some consumers within the EU about the health, fertility and welfare of dairy cows in modern dairy herd management systems, there is a need to address these concerns with large-scale experimental and epidemiological studies.


As the world’s population grows, global demand for milk for human consumption is predicted to rise by more than 50% by 2050 (FAO, 2011). This along with concerns about ‘food security’ has led to demand for higher milk output from dairy farmers nationally. The FAO states that this demand can only be met by large scale, intensive livestock operations. Dairy farmers have responded by increasing milk production per cow and by increasing herd size. World milk production increased by 27% between 1997 and 2007 (IDF, 2008). The 27 states in the European Union (EU-27) are now the biggest milk producing block in the world. This trend toward herd expansion will intensify following EU milk quota deregulation in 2015 (O’Donnell et al. 2011) but then land area will be the new key factor limiting production indicating more producers may adopt confinement systems (O’Brien et al. 2012). However, rising input costs, fluctuating milk commodity prices, the purchasing power of oligopolies and recent broader macroeconomic difficulties have led to rationalization of dairy production industries internationally with only the most efficient farmers surviving. Notwithstanding existing EU milk quota regulation, this has resulted in intensification of milk production systems worldwide whether they are pasture-based (PB) [e.g. intensive rotational grazing, extended grazing season, high stocking rates (SR)] or zero-grazed (ZG) [e.g. total mixed ration (TMR) feeding, genetic selection for increased milk yield, automated milking]. In an attempt to address these trends, some milk processors (e.g. Friesland Campina in the Netherlands) pay a bonus to farmers who commit to grazing for specified periods.

As this intensification has gradually occurred, dairy cow fertility has gradually declined internationally to such an extent that it has been predicted that high-performance production systems will become unsustainable by 2020 (Maas et al. 2009). Thus, poor dairy herd fertility performance has become a ‘wicked problem’. A wicked problem is a problem that lacks a consensus about the optimal solution and an inability of experts and science to resolve the conflicts (Gerloff 2011). While many hypotheses have been posited as to the multifactorial aetiology of this problem, effects of changes in dairy herd management systems have not featured prominently. Only a limited number of recent studies have highlighted the association between herd-level management factors and reproductive performance (Garcia-Ispierto et al. 2007; Bach et al. 2008; Schefers et al. 2010). This blind spot may reflect the reductionist nature of much research on dairy cow reproductive physiology and the paucity of systems thinking in this scientific milieu, a fact alluded to in a recent editorial (Hudson 2011). This systematic review focuses on the effects of management systems on reproductive performance in dairy herds.

Dairy Herd Management Systems

Settled agricultural farming systems have recently been classified by the European Food Safety Authority (EFSA) as mixed (crop) livestock farming, intensive production systems and extensive production systems, with the latter two predominating on dairy farms in Europe (EFSA, 2009a). There are numerous descriptors used to characterize dairy herd management systems internationally. The term management system is used here as a synonym for the farm environment. The hazards within the farm environment which may impact reproductive performance include the housing, nutrition, health, genetic selection and management. This review focuses on those systems most common in Europe, intensive, often ZG systems and PB systems.

The defining characteristics of ZG systems are housing all year round, outdoor access only for loafing, non-seasonal calving, TMR feeding and high milk yield per cow. The defining characteristics of PB systems are grazing for the provision of forage for at least 6 months of the year with housing for the remainder and seasonal calving. Easy-feeding is now more common than self-feeding in both PB and ZG systems. In some countries, management systems have regulatory definitions. For example, in Ireland, an ‘intensive unit’ is a premise where farm animals are kept which relies on automatic equipment to such an extent that if it failed it would cause the animals unnecessary suffering (DAFM, 2011). In the USA, any dairy with more than 700 cows is defined as a ‘concentrated animal feeding operation’. These typically have all-year-round total confinement in freestall barns with the milking parlour operating around the clock and are often described as ‘factory farms’ (Cook 2011). Organic management systems are value-based variants of PB systems with grazing and adherence to organic codes of practice mandatory. Within the EU, organic farming accounts for a few per cent of dairy production, with an upward trend in some countries.

According to a recent report from the EFSA, intensive systems account for 85% of all milk production in the EU (EFSA, 2009a). While only approximately 10% of world milk production is from grazing systems, pasture grazing is the most common system for managing dairy cows worldwide (Boken et al. 2005). As herd size increases within the EU, there has been an increase in the number of farmers adopting a ZG management system (EFSA, 2009a, DEFRA, 2011). Following a review of the relevant literature, a synthesis of the main characteristics of these two contrasting management systems is shown in Table 1.

Table 1.   Characteristics of zero-grazing and pasture-based dairy herd management systems
  1. TMR, total mixed ration; PMR, partial mixed ration; AMS, automated milking systems.

SynonymsBarn housed, confinement, factory farming, feedlot, high-input/high-output, high intensity indoor, industrial, intensive, super dairy, total confinementExtensive, grazing, high forage, low cost, low input, outdoor, pastoral, seasonal
Duration of housingAll year round, continuousNone or seasonal in winter
Pasture usageLoafing/exercise lot used (freestall and drylot) or not used (total confinement)Grazing ≥6 months/year
Calving patternAll year roundSeasonal (winter/spring, autumn/winter) or bi-seasonal (split)
Accommodation typeIndoor-loose: cubicles (freestalls) with various bedding types on a concrete or sand base with slatted or solid passageways (rubber/no rubber) or bedded pens (deep litter, bedded pack, straw yard); Indoor-tied: Tie-stalls (stanchions, tethered). Outdoor – loose: open lot (earthen corral, dry lot)Indoor: as for zero-grazing systems. Outdoors: pasture [grasses (ryegrass, alfalfa, lucerne, bermuda) +/− legumes (clover), rotational or set-stocked] or out-wintering pads/stand-off pads.
Complimentary feed typeTMRConcentrate or TMR or PMR
Conserved roughageMaize silage predominantlyGrass silage predominantly
Milk productionHigh milk volume output/cowHigh milk solids output/hectare
Herd sizeLargeSmall-large
Milking frequency×2, ×3, AMS (robots)×1, ×2

While these systems operate outside the EU also, they may differ in some characteristics compared to within the EU. For example, herd size in both systems tends to be larger in North America (ZG) and in the Antipodes (PB), pasture herbage varieties and use of irrigation differ, labour costs tend to be lower and pharmaceutical intervention in breeding programmes tends to be greater.

Internationally, ZG is more common in some countries (e.g. Germany, Israel, Mediterranean countries, North America, the Netherlands,), while PB systems are more common in other countries (e.g. Argentina, Australia, Azores, France, Great Britain, Ireland, New Zealand). However, most countries have a mixture of management systems with some types predominating. For example, in the UK, 99% of farms are PB with the remainder housed throughout the year (Haskell et al. 2007). Whereas in Chile, 40% of farms have dry lots, 40% cubicles and the remainder grazing and mixed systems (Melendez and Pinedo 2007). In Germany, in one region 2% of herds are ZG while in another region 99% are ZG (EFSA, 2009a). While in some Scandinavian countries, for example, Finland and Sweden, regulations mandate access to pasture (EFSA, 2009a).

Management systems are heterogeneous within system, and sometimes between years, and they continue to evolve. For example, the type of housing provided in both ZG and PB systems is highly variable. To deal with this complexity, herd cluster models have been developed to group herds with similar production environment descriptors using environmental gradients along with intensity and production indices (Fikse et al. 2003, Strandberg et al. 2009, Windig et al. 2011). Across-country grouping of herds into similar production environments is particularly important for borderless genetic evaluations, for example, for fertility traits (Zwald et al. 2003).

Which Dairy System Do Cows Choose?

In recent choice experiments in the UK, cows, although naturally grazing animals, displayed a partial preference to be indoors where a TMR was fed, particularly for high yielders and when it rained (Charlton et al. 2011). Conversely, in conditions of high ambient temperature and humidity in Canada, cows chose to spend more time indoors during the day and at pasture during the night (Legrand et al. 2009). It is concluded that cow preference for management system is dependent upon the prevailing climatic conditions, diet offered and level of milk production.

Which Dairy System Do Consumers Prefer?

In the UK, the majority of consumers surveyed (n = 363) thought that keeping dairy cows permanently indoors (95%) or tethered (95%) was unacceptable and a minority (32%) thought it unacceptable to keep dairy cows outdoors all year round. A majority (73%) thought it acceptable to keep dairy cows outdoors in summer and indoors in winter with the remainder (27%) viewing this practice as maybe or not acceptable (Ellis et al. 2009). In Canada, the majority (73%) of participants (n = 178) in an agricultural university on-line survey voted that dairy cows should be provided with access to pasture (Weary et al. 2012). When US consumers were given the choice of milk from TMR-fed compared with pasture-fed cows, they could not differentiate between these systems even though distinct flavour and compositional differences are present (Croissant et al. 2007; Mohammed et al. 2009).

Internationally, and especially within the EU with the imminent abolition of milk quota (2015), there is a trend toward dairy farm intensification with fewer, larger herds as scale is a key driver of profitability. Farmers with larger herds adopt more mechanization and automation technologies (Khanal et al. 2010), and as herd size increases, more farmers are likely to move to ZG systems (Burow et al. 2011). However, changes in husbandry systems toward more intensive management of dairy cows may conflict with consumer welfare opinions and with policy-makers views on sustainability. For example, the environmental impacts of ZG systems are greater than those of PB systems (O’Brien et al. 2012). While in some countries, ZG systems are traditional and accepted as normal, in others ZG, large scale (‘super dairies’) and intensive farms are controversial (Huxley and Green 2010). To address these concerns in the UK, a large-scale experimental and survey design study was initiated in 2008 to examine the management, health, fertility and welfare of continuously housed cows (ZG) when compared with cows in summer grazing systems (PB). The results of this significant study are due to be released in 2012 (DEFRA, 2011).

Farmer Demographics and Dairy Management System

Within each production system, management practices in animal health, nutrition, genetics, stockpersonship and reproduction can vary significantly. Hence, the characteristics of the farmer such as age, education and full/part-time working on the farm, owner of the farm/manager of the farm are a key component of either system. However, interactions between management system and farmer demographics have been detected in the US where PB farmers were older and less likely to hold college degrees (Khanal et al. 2010) and in the UK where the quality of stockpersonship (positive interactions shown by stockpersons toward cows corroborated by the inquisitive nature of the cows’ interaction with an unfamiliar human) was above average on ZG farms (DEFRA, 2003). In addition, veterinary dairy herd fertility service provision differs markedly between seasonal and non-seasonal management systems (Mee 2010).

Housing vs Pasture

Despite the enormous corpus of literature on factors affecting reproductive performance in dairy herds, there are very few recent peer-review papers comparing management systems. This may reflect the increasingly reductionist approach to modern scientific research. Most studies involving the effect of management system on reproductive performance vary one or more factor within the system but do not compare between management systems. However, in comparing whole farm management systems inevitably, there is confounding of effects with, for example, higher milk production per cow in ZG compared with PB systems; thus, reproductive outcomes are due not just to management system per se but to the effects it has on cow health and lactational and reproductive physiology. The same, standardized fertility management was implemented for PB and ZG treatments within designed experiments (e.g. Palmer et al. 2012, Pollott and Coffey 2008, Washburn et al. 2002). In observational studies comparing these systems, the fertility management was not specified as the same or different between management systems (e.g. Bruun et al. 2002; Windig et al. 2005; Burow et al. 2011); though, the same oestrous detection method (HeatWatch) was used in all herds in the observational study of Dransfield et al. (1998).

Replacement heifers

In the limited number of studies carried out comparing ZG and PB systems for rearing replacement heifers, differences in age at first calving (Hagjihacek et al. 1992), reproductive performance (Troccon 1993), health (Troccon 1993) and milk production (Hagjihacek et al. 1992) were non-significant. While a Swedish study found a shorter productive life in heifers grazed before first calving (Hultgren and Svensson 2009), a French study did not (Troccon 1993). It is concluded that there are insufficient studies to be definitive about the effects of divergent management systems on replacement heifer performance but existing data indicate minimal effects.

Dairy cows

It is generally accepted that milk production is higher in ZG than in PB systems (by up to 20%), (White et al. 2002; Fontaneli et al. 2005; Davis et al. 2006). This may be explained by greater dry matter and energy intake and lactation persistency and length in ZG systems (Kolver and Muller 1998; Boken et al. 2005; Davis et al. 2006). Negative energy balance [plasma glucose, insulin, non-esterified fatty acids, body condition score (BCS) loss and body weight loss] is higher in PB compared with ZG systems (Kolver and Muller 1998; Washburn et al. 2002; Boken et al. 2005; Fontaneli et al. 2005).

Dairy cow health

The incidence of some animal health problems tends to be greater in ZG than in PB systems; clinical mastitis and high somatic cell count (Bendixen et al. 1986; Beri et al. 1995a; Fontaneli et al. 2005), clinical lameness and leg injuries (Haskell et al. 2006; Olmos et al. 2009a), parturient problems (paresis, dystocia and retained foetal membranes) (Bendixen et al. 1986) and ‘loser’ cows (Thomsen et al. 2007). However, some other animal health problems can be more common in PB systems (grass tetany, exposure to inclement weather, internal and external parasitism, low rumen fill, metabolic stress, ketosis, sub-acute ruminal acidosis and phytotoxicities), (EFSA, 2009a; Olmos et al. 2009b, Cook 2011). Hence, it should not be assumed that animal health or welfare is automatically better in either system.

Given these differences between management systems in milk production, energy balance and animal health status, it would be expected that differences might exist in reproductive physiology and performance.

Dairy cow reproductive performance

This hypothesis is supported by the findings of a factorial design experiment in the UK. Contemporaneous comparison of a ZG and PB system using 363 lactations from 229 cows on the same farm revealed that while the pattern of luteal activity post-partum was ‘better’ (earlier onset, longer luteal phases, shorter inter-ovulatory intervals, and less frequent delayed ovulation) in the PB system, reproductive performance (number of luteal cycles needed to become pregnant, conception characteristics, gestation length) was poorer (Pollott and Coffey 2008). The differences in luteal activity were entirely explainable by the differences in energy balance characteristics of the two systems. A recent Irish study using 46 spring-calved cows has shown that the oestrous behaviour of cows in a ZG system differs significantly from that in a PB system. There was a higher frequency of standing to be mounted, mounting other cows and ano-genital sniffing in PB compared with ZG cows (Palmer et al. 2012). As a consequence of these differences, efficiency of oestrous detection was significantly higher in PB cows, irrespective of the method of oestrous detection employed (visual observation, tail paint, radiotelemetry), although accuracy of oestrous detection did not differ between treatment groups. In addition, the intervals between calving and oestrus were longer in the ZG treatment (Palmer et al. 2010). These findings are supported by the results of a North American study with 36 cows which showed that PB cows had a significantly higher peak plasma progesterone concentration (in the first post-partum cycle), had a significantly higher oestrous intensity (number of mounts received), and had a significantly higher pregnancy rate than ZG cows (Boken et al. 2005). However, management system had no effect on calving to onset of luteal activity or first oestrus, duration of first oestrus or number of services per conception. In agreement with these peak progesterone data, Bilby et al. (1998) found that progesterone clearance was lower in PB cows in New Zealand than in ZG TMR-fed cows in North America. A Danish study in 2144 herds found significantly higher odds of metritis in ZG compared with PB herds (Bruun et al. 2002).

In a Dutch epidemiological study in 3904 herds relating herd environment to herd fertility using principal component analysis, high intensity herds had a shorter interval from calving to first service but a lower first service conception rate (Windig et al. 2005). Supporting this result, in a recent Croatian study with 860 cows, early pregnancy loss rates (day 32–86 post-insemination) were significantly higher in a ZG (11%) compared with a PB system (5%) and were associated with increased BCS loss and higher milk yield (Zobel et al. 2011). These results are in agreement with an earlier study in two Hungarian herds where PB cows had a higher fertility index and shorter calving interval than ZG cows (Beri et al. 1995b). Recent Danish research in 391 herds comparing ZG and PB systems concluded that cow mortality was lower in PB herds (Burow et al. 2011). In addition, the risk of mortality decreased with increasing number of hours on pasture. PB systems also had lower culling rates than ZG systems (Washburn et al. 2002; White et al. 2002).

In contrast to these results, a 4-year study on a university farm in the USA did not find any significant differences in reproductive performance between a ZG and a PB system although 10% less Holstein-Friesians in the ZG survived until a subsequent lactation compared to in the PB system (Washburn et al. 2002). In addition, a North American study of 17 herds showed no relationship between conception rate and housing type (PB and ZG), (Dransfield et al. 1998). In support of these results, a semi-quantitative risk assessment carried out by the European Panel on Animal Health and Animal Welfare concluded that the risk of suffering reproductive or metabolic disorders was independent of the housing system (cubicle, tie-stalls, straw yards or pasture) (EFSA, 2009b). No conclusions were drawn from the scientific report of the panel which reviewed these systems. It is concluded that luteal and oestrous activity is higher and metritis, early embryonic mortality and cow culling and mortality rates are lower in PB compared with ZG systems. However, there are conflicting data on effects of management system on conception metrics.

Genotype-by-farming environment interactions for reproductive performance

A genotype-by-environment interaction (GxE) is one where the difference between genotypes is not constant from one environment to another. Thus, a given environmental difference has more effect on some genotypes than on others. While some studies have demonstrated a GxE interaction for dairy herd management systems and reproductive performance, others have not. This reflects the variety of herd management systems, the variation in metrics used to describe reproductive performance and the limitations of some study designs. Studies in Australia (Fulkerson et al. 2008), Brazil and Columbia (Ceron-Munoz et al. 2004), the Netherlands (Calus et al. 2005) and in the UK (Haskell et al. 2007; Strandberg et al. 2009) found GxE interactions for reproductive performance in different farm environments as defined by intensity and production indices and concentrate feeding levels. However, other studies in Australia (Haile-Mariam et al. 2008), Canada (Boettcher et al. 2003), Ireland (Buckley et al. 2003a), the UK (Pryce et al. 1999) and in the USA (Kearney et al. 2004) did not find GxE interactions for reproductive performance in different farm environments as defined by production intensity or concentrate feeding level or when comparing ZG with PB systems. While GxE interactions were detected between conventional and organic management systems for reproductive performance in one Swedish study (Sundberg et al. 2010), they were not in another Swedish study (Ahlman et al. 2011).

Where significant GxE interactions exist sires vary in the sensitivity of their daughters to different farm environments. Hence, sires should be matched to particular farm management systems and farmers would benefit from selecting bull’s progeny tested in a similar environment as their own. Depending on the relative importance of fertility traits in national selection indices, where a significant GxE exists for reproductive performance, environment-specific breeding values for fertility traits should be estimated for use in genetic indices. However, where the effects of GxE are small between management systems such customization of genetic indices is unwarranted. It is concluded that given the variation in the results of studies attempting to detect a GxE interaction for dairy management systems internationally studies need to be carried out in each dairy industry to reflect the relevant genetics and indigenous management systems.

Variations in Accommodation within Housing Systems

In both PB and ZG systems, there is variation in the type of accommodation provided when cows are housed. The studies reviewed here describe the effects of these variations during the housing period.

Cubicle vs tie-stall housing systems

Prior to the development of milking parlours, stanchion houses were the most common type of dairy housing. Now in Europe, their use has diminished but with wide variation between countries, for example, from 10% in the Netherlands to 60% in Sweden (EFSA, 2009a). The incidence of some clinical animal health problems (mastitis, teat injuries and ketosis) is lower, while the incidence of others (poor claw health, SCC) is higher in cubicle compared with tie-stall systems (Simensen et al. 2010). Many, (Eriksson et al. 2006; Lof et al. 2007; Simensen et al. 2010), although not all (Coleman et al. 1985; Kinsel and Etherington 1998, Bruun et al. 2002), studies show poorer reproductive performance in tie stall systems compared with cubicle house systems. It is concluded that good reproductive performance can be achieved in either tie-stall or cubicle management systems.

Cubicle vs loose housing systems

Open cubicles are the most common housing in European dairy farms; loose housing is the least common system in Europe (EFSA, 2009a). It is generally accepted that the incidence of clinical lameness and of leg injuries is higher in cubicle compared with loose housing, for example, in straw courts or compost-bedded pack barns (DEFRA, 2003, Lobeck et al. 2011). Attempts to improve reproductive performance in loose housing by separating cows for 1 month after calving failed but milk production did increase in primiparous cows (Ostergaard et al. 2010). A recent study was carried out in Ireland on low cost winter housing systems comparing uncovered or covered loose out-wintering pads with cubicle housing. Type of accommodation did not significantly affect reproductive performance (O’Driscoll et al. 2007). It is concluded that there are insufficient studies comparing cubicle and loose housing to determine their effects on reproductive performance.

Tie-stall vs loose housing systems

Loose housing has a more positive influence on post-partum ovarian activity than tie-stalls with a greater incidence of atypical progesterone profiles and longer interval to first ovulation in the latter (Claus et al. 1983; Petersson et al. 2006). In addition, higher milk yields and greater longevity have been found in cows housed loose compared with those in tethered systems (Bader 1987), with a transitory drop in yield when cows move from tethered to loose housing (Soch et al. 1997). However, Coleman et al. (1985) reported a higher incidence of retained foetal membranes and of culling for low production in open lot compared with stanchion systems. It is concluded that luteal activity may be better in loose housing compared with tie-stalls but there are insufficient data comparing reproductive performance.

Floor surface in cubicle housing

Wet and slippery flooring is associated with poorer reproductive performance (Przewozny 2011); hence, alternative floor surfaces have been tested to improve cow health, welfare and reproduction. A recent study by Kremer et al. (2010) found that cows on rubber flooring had an earlier onset of oestrus post-partum and superior reproductive performance to those on concrete flooring. In contrast, an earlier study by Boyle et al. (2007) found a negligible beneficial effect of rubber flooring on heel erosion and differences in behaviour but no effect on other claw lesions or on oestrous expression or on reproductive performance. The authors emphasized the importance of floor friction coefficient in expression of standing oestrous behaviour, a point made by other authors also (EFSA, 2009a). While soft flooring materials within the cubicle have been associated with increased milk yield, and fewer incidences of clinical mastitis, teat lesions and removal of cows, data on reproductive performance are lacking (Ruud et al. 2010). It is concluded that while floor surface affects reproductive performance, there are conflicting results on the benefits of rubber passageway flooring.

Variations in Feeding Practices within Management Systems: Reproductive Performance

Unlike variations in accommodation systems, there is a substantial and growing corpus of literature on the effects of various feeding regimes on reproductive performance of dairy cows in different management systems.

Stocking rate at pasture

A central tenant of PB systems is to maximize milk production (per hectare) from pasture which in intensive grazing systems is achieved through higher SR. However, the effects of SR or pasture intake on reproductive performance have not been widely investigated. Two studies have shown the benefit of high pasture intakes (unrestricted access or 11.9 kg pasture DM per cow per day) pre-partum on post-partum reproductive performance (shorter post-partum anovulatory interval) in heifers and in cows (Chagas et al. 2006, Burke and Roche 2007).

Three recent studies have, in general, shown no effect of post-partum SR at pasture on reproductive performance across the range from 1.6 to 4.3 cows/hectare (Baudracco et al. 2011; Macdonald et al. 2011; McCarthy et al. 2011). However, Ryan and Mee (1994) reported a longer calving to conception interval for cows on tight compared with lax grazing (4 vs 8 cm post-grazing height) and McCarthy et al. (2011) showed higher late embryonic mortality at low SR (2.5 cows per ha). Recently, Macdonald et al. (2011) commented that cows/ha may not be the most robust measure of SR and that comparative SR (kg of cow body weight at a standard BCS/tonne of feed DM) may be more meaningful. It is concluded that these studies show that for all-year-round grazing systems, maximizing pasture intake pre-calving benefits reproductive performance but that unless pasture intake is severely restricted, post-partum SR has negligible effect on reproductive performance.

Concentrate supplementation in pasture-based systems

Where concentrate supplementation has been practised, pre-calving in PB systems low-fed cows tended to have better reproductive performance (onset of ovarian activity, ovulation detection rate, embryo mortality) than high-fed cows (Fahy et al. 2005; Cavestany et al. 2009; Cutullic et al. 2011); though, this is BCS-dependent (Adrien et al. 2011). In a review of the literature on dry period feeding in PB systems, Mee (2008) concluded that grass or grass silage alone resulted in an earlier onset of ovarian cyclicity and better reproductive performance than forage plus concentrate supplement.

A series of experiments have been carried out on the effects of concentrate supplementation after calving at pasture on reproductive performance. In general, these studies have failed to show a benefit of such supplementation, irrespective of cow genotypic merit for milk production, on reproductive performance (Logue et al. 1999; Kennedy et al. 2003; Horan et al. 2004, 2005; Kolver et al. 2005; Pedernera et al. 2008; Coleman et al. 2009). Supplementation levels varied from 350 to 2000 kg of concentrate/cow/lactation. However, Fulkerson et al. (2001) found that while cows fed higher levels of concentrates (0.84 and 1.7 tonnes/cow/lactation) had an earlier onset of ovarian activity post-partum and higher submission rates, they also had lower first service conception rates than those fed lower levels of concentrates (0.34 tonnes). This was supported by Jonsson et al. (1999) who showed that the probability of ovulation post-partum was higher in high-fed pluriparous cows (2000 kg concentrates/cow/lactation) but not in primiparous cows. It is concluded that these studies show that post-partum luteal activity is better in cows on low-feed regimes pre-partum and on high-feed regimes post-partum and that conception success is generally not affected by concentrate feed level post-partum where pasture allowance is adequate.

Concentrate supplementation in zero-grazing systems

Similar to the results from PB systems, data from ZG systems show that increasing the energy density of the pre-calving diet (grass silage supplemented with concentrates or straw or a TMR diet offered during the dry period) is not beneficial for reproductive performance and can be detrimental (longer calving to onset of luteal activity, greater number of services/conception) (Keady et al. 2001; Holtenius et al. 2003, McNamara et al. 2003a). Such outcomes have led to the theory of limit feeding during the entire dry period and avoidance of luxury feeding to prevent over-conditioning at calving by feeding a low energy high fibre TMR pre-calving (Beever 2006). There is some evidence to support the effect of this controlled-energy diet feeding regime on reproductive performance using straw-based (Mee 2008) or grass silage only (Fahy et al. 2005) diets pre-calving.

Earlier studies showed no benefit of high concentrate supplementation in complete diets offered post-calving for the complete lactation (2.5 tonnes) on reproductive performance (McGowan et al. 1996; Pryce et al. 1999). Reduced reproductive performance on high concentrate diets was entirely because of increased milk production. More recently, the theory of feeding an insulogenic diet (‘cycling diet’; e.g. high starch) in early lactation and a lipogenic diet (‘mating diet’; e.g. low starch, high fat) pre-breeding has been posited to reduce the calving to onset of luteal activity interval and to improve embryo development and conception success, respectively. While Garnsworthy et al. (2008) showed detrimental effects (delayed post-ovulatory progesterone rise) of a very high starch diet (230 g/kg DM), lower levels of dietary starch (160–180 g/kg DM) and fat (<44 g/kg DM) had a positive effect on ovarian function. These results were not replicated recently by Gilmore et al. (2011); hence, there is a need for larger scale confirmatory studies.

It is concluded that these results suggest that high feed levels during the dry period or post-calving in ZG systems are not beneficial to reproductive performance but that limit feeding pre-calving and altering post-calving dietary energy source may have a positive effect.

While there are numerous studies showing effects of specific components of the diet (e.g. fat, protein, microelement and vitamin supplements, (e.g. McNamara et al. 2003b; Mee 2011), BCS (e.g. Buckley et al. 2003b), and the aberrant metabolism of modern dairy cows’ somatotropic and gonadotropic axes (e.g. Roche 2006) on dairy cow reproduction, these are not management system-specific and so have not been reviewed here.


Cows in intensive management systems produce the majority of milk worldwide, and the trend toward intensification in both ZG and PB systems is increasing. This intensification has contributed directly to the current decline in dairy cow fertility globally. Given the heterogeneous nature of both ZG and PB systems, comparisons between them in dairy cow reproductive performance need to be treated with caution. However, in general, cows in ZG systems have higher milk production and better energy balance but more animal health problems, lower ovarian activity post-partum, reduced oestrous expression, reduced conception success, and higher culling and mortality rates, than cows in PB systems. Key environmental descriptors affecting reproductive performance within management systems include the type and duration of housing and the pre- and post-partum diet composition. GxE interactions for dairy cow fertility have been detected for some, but not for other, management systems. Given the concerns of some consumers within the EU about the health, fertility and welfare of dairy cows in modern herd management systems, there is a need to address these concerns with large-scale experimental and epidemiological studies.

Conflict of interest

The author has no conflict of interest to declare.