A review of extended lactation in dairy cows managed in high-input and pasture-based farming systems

Introduction

As a result of ongoing genetic selection, improved nutrition and feeding practices, and overall management, the milk yield and lactation length of the modern dairy cow are well in excess of those necessary to raise its calf. Milk production is ultimately a function of the number of milk-producing cells in the mammary gland and the extent of their synthetic capacity. Their synthetic capacity is the main determinant of milk yield up to peak lactation, whereas the number of secretory cells is the major determinant during the post-peak declining phase of lactation (Capuco and Ellis 2005). The gradual post-peak decline of milk yield typically coincides with the cow being pregnant at this time and ceases when milking is stopped for the final 2 months of gestation, i.e. the dry period. The mammogenic and lactogenic stimuli as a result of pregnancy are critical to prevent complete involution of the secretory cells in the mammary gland and allow for replacement of damaged or quiescent secretory cells during the dry period, to ensure that, upon parturition, the lactation cycle can start over again (Capuco and Akers 1999; Capuco and Ellis 2005).

This is why conventional dairy-farming practices are based on a 12-month calving interval, allowing for 305 days of lactation, followed by a dry period of 60 days. However, this does require the cow to become pregnant again within approximately 2.5–3 months following parturition, which means that she needs to conceive at the same time when she also produces the most milk (i.e. during peak production) and is metabolically in a negative energy balance. These physiological states, conception and high milk production, compete for a limited pool of metabolites and, given that milk production, in a biological sense, ensures the immediate survival of its offspring and, therefore the species, it will take priority over the ability to conceive at that time. Indeed, selection for higher milk production negatively affects fertility (Pryce et al. 2004; Walsh et al. 2011). The implication of delayed conception and pregnancy in high-producing dairy cows is that it may no longer be practical to adhere to a strict 12-month calving interval.

Another challenge of managing high-producing dairy cows within a 12-month calving interval is that they can still have a high daily milk production (i.e. >25 kg/day) at the time of drying off, which may slow the required immunological changes at this time (Silanikove et al. 2013). Such immune changes are necessary to prevent mammary infections during this critical and high-risk period in the lactation cycle, when the mammary gland is very susceptible to existing subclinical infections becoming clinical or to developing new infections (Silanikove et al. 2013; Vilar and Rajala-Schultz 2020).

Voluntarily extending the calving interval beyond the conventional 12 months not only results in an extended lactation (EL), it allows cows to conceive and become pregnant later during lactation, once they have returned to a state of positive energy balance. It also allows them to be dried-off when their daily milk yield is at a much lower level, thereby reducing possible mammary health issues at that time. Additionally, there are other benefits of an extended calving interval (and EL), such as fewer ‘surplus’ calves and fewer health-risk periods over the lifetime of the cow (van Knegsel et al. 2022). Surplus calves are mostly male calves or those not required as herd replacements. Such calves are sometimes euthanised shortly after birth or more commonly culled after several days and used for veal production (Creutzinger et al. 2021). The issue of surplus calves as well as the related farm practice of separating mother and calf immediately after birth, not only raise animal welfare concerns but increasingly adversely influence public perception of dairy farming (Ritter et al. 2022; Sirovica et al. 2022).

As discussed, at the time of drying off, cows are at an increased risk of mammary infections, which may also carry over into the new subsequent lactation (Cameron et al. 2014; Vilar and Rajala-Schultz 2020). The periparturient period or transition period is another major health-risk period for the cow. This is a time when the immune competency of the cow is compromised and during this time and the ensuing period of negative energy balance, she is more prone to metabolic disorders, uterine diseases, mastitis and lameness (Mezzetti et al. 2021).

Since the 1990s, a number of controlled studies have been conducted with high-yielding dairy cows under intensive farming conditions to explore the concept of EL, considering lactation performance, reproduction and health implications (Bertilsson et al. 1997; van Amburgh et al. 1997; Ratnayake et al. 1998; Österman and Bertilsson 2003; Sorensen et al. 2008; Delany et al. 2010; Mellado et al. 2016; Niozas et al. 2019a, 2019b; Edvardsson Rasmussen et al. 2023a, 2023b). More recently, comprehensive EL research programs have been conducted in Denmark (Sehested et al. 2019) and The Netherlands (van Knegsel et al. 2022), not only considering the effects of EL on animal factors, but also considering environmental (e.g. greenhouse-gas emission), milk-quality and processing, and economic implications.

Although high milk yields prior to drying off within a 12-month calving interval are less of an issue in low(er)-input pasture-based dairy management systems, there has been an increasing interest in the concept of EL for these systems (Auldist et al. 2007; Kolver et al. 2007; Grainger et al. 2009; Butler et al. 2010; Marett et al. 2011). However, such a farming system, based around seasonal calving to optimise pasture utilisation, offers unique challenges for implementing an EL strategy in terms of feeding strategies and maintaining ongoing lactation.

Therefore, in this review, we will explore the impact of EL as a management tool on milk production, composition and processing characteristics, as well as on reproductive performance, animal health and welfare, and environmental and economic consequences under both intensive, high-input systems, and seasonal pasture-based farm management systems.

Extended calving interval versus extended lactation

For the purpose of the current review, the term extended calving interval is defined as any calving interval greater than 12 months. It is assumed that extending the calving interval does not extend the normal 60-day dry period at the end of lactation. Therefore, the term extended calving interval is assumed to be equal to the more accurate term ‘extended lactation’, which will be used in this review.

Effect of extended lactation on milk yield and composition

Extended lactation and dairy cow nutrition

Milk production comes at a high metabolic cost to the cow and requires a high input of digestible nutrients. Indeed, nutrition is one of the key drivers of milk yield. In high-input/TMR dairy systems, nutrition is aimed at maintaining high milk yields, by feeding a similar diet year-round. In contrast, pasture-based systems aim to maximise DMI and milk yield from pasture and are therefore much more affected by changes in seasonal conditions. Cows usually calve in spring when high-quality pasture is readily available to meet their high metabolic demands and are commonly dried-off in early winter when pasture growth slows. Such a system lends itself well to a 12-month calving interval.

Thus, whereas managing nutrition of lactating cows during EL in a high-input system is relatively straight forward, when similar diets can continue to be fed without restrictions, in a pasture-based system nutrition and maintaining a consistent feed supply becomes more challenging. It requires limited pasture availability to be supplemented with conserved forages and/or concentrate (including grains) in winter to prevent milk production dropping below the level where drying off becomes inevitable. When cows with EL enter the second spring, often an increase in milk yield is observed, although not to the same extent as during early lactation (Kolver et al. 2007; Williams et al. 2013; Jarman et al. 2020). This effect is likely to be due to a photoperiodic effect of increasing daylength (Dahl et al. 2000) and the increased abundance of high-quality pasture during that time.

The difference between these TMR and pasture-based farming systems is also reflected in the available amount of research on the impact of nutrition during an EL. Only three published studies conducted under high-input conditions were found, whereas the impact of nutrition on milk production during EL in pasture-based systems has been much more thoroughly investigated (Table 2).

Sorensen et al. (2008) showed that feeding cows, managed under high-input conditions, a more energy-dense diet post-peak lactation did not affect milk production, LW, BCS or profiles of metabolic hormones (growth hormone, insulin and IGF-I; Sorensen and Knight 2002) during EL. Gaillard et al. (2016a, 2016b) fed cows a more energy-dense diet within a high-input system during the first 7 weeks of lactation when the cows were in negative energy balance. They expected to see positive carry-over effects during subsequent EL; however, although the energy-dense diet improved the energy balance in early lactation and increased milk production, it did not improve production subsequently during EL. In fact, there was a negative effect on lactation persistency past 250 DIM. The energy-dense diet had no effect on plasma insulin or IGF-I but increased NEFA till Week 36 of lactation. Overall, increasing the energy density of the diet appeared to have little benefit for cows in a high-input system. Recently, Burgers et al. (2023) looked at the effect of EL in primi- and multiparous cows on metabolism (i.e. NEFA, β-hydroxybutyrate, glucose, insulin and IGF-I in plasma) and BCS, by using three groups of cows subjected to a VWP (i.e. the post-calving interval during which cows are deliberately not serviced) of 50 (i.e. control group), 125 or 200 days. In primiparous cows EL did not affect milk yield or metabolic parameters. However, in multiparous cows, EL (VWP of 200 days) resulted in less milk and increased BCS during pregnancy and a more negative energy balance during the beginning of the next lactation.

The effect of feeding supplements (i.e. silage and/or grain) on production during EL in cows grazing pasture has been investigated in a number of studies conducted in Australia and New Zealand (Table 2). In all but one of these studies, lactation was extended to between 20 and 22 months.

Rius et al. (2011) showed that short-term (i.e. 68 days, starting at 328 DIM) supplementation of a pasture-only diet with 6 kg DM/day per cow of concentrates during EL can transiently increase daily milk yield and leads to a sustained increase in LW and BCS. However, most studies were conducted with supplementation during the entire EL. Auldist et al. (2011) fed cows a pasture-only diet or pasture supplemented with different levels of grain and showed that grain feeding increased DMI and resulted in an increased feed conversion efficiency (FCE; i.e. units milk/DMI) and a marginal milk-solids response during the EL period, but only during the autumn when the supplements compensated for poorer pasture quality at that time. Williams et al. (2013) also showed that grain supplementation (5 kg/day per cow) to pasture-fed cows during EL can increase milk yield, this time during both spring and autumn.

Marett and co-workers conducted a number of studies in cows with EL, exploring the effect of supplementing a pasture diet with grain (6 kg/day per cow) on the metabolic responses in cows during EL, using a glucose-tolerance test (Marett et al. 2015), insulin-tolerance test (Marett et al. 2017) and epinephrine challenges (Marett et al. 2018). Whereas the observed metabolic changes are complex, collectively their results indicated that grain supplementation increased milk and MS yields and tended to increase LW. The latter is consistent with the suggestion that during EL, a higher proportion of dietary energy is partitioned towards body reserves. The response is likely to be dependent on the level of feed supply, because when cows fed pasture supplemented with grain were fed a restricted diet during only the EL period, milk yield was decreased, BCS slightly increased and metabolic changes suggested increased lipid mobilisation and more energy being partitioned towards milk production (Marett et al. 2011). In contrast, when cows normally fed a pasture-based diet were adjusted to a TMR diet, metabolic hormone changes were consistent with partitioning nutrients to LW gain rather than into milk production and a higher proportion of cows in the TMR group had to be dried-off before 22-months of EL because of excessive LW gain (Grainger et al. 2009; Delany et al. 2010).

In a more recent study, Marett et al. (2019) selected two groups of cows on the basis of either high or low milk yield at 450 DIM (∼15 months). Subsequently, all cows received a pasture diet supplemented with grain (6 kg/day per cow) between 450 DIM and 580 DIM. The high milk-yield cows maintained a higher milk yield throughout the experiment and exhibited a lower plasma insulin response, slower glucose clearance and increased epinephrine sensitivity, resulting in greater lipid mobilisation and more dietary energy being partitioned towards milk production. Given that both high- and low-producing cows were fed the same diet, these differences indicated that in addition to simply feeding more nutrients to maintain milk production during an EL, genetic differences among cows in how these nutrients are utilised are also important.

This is also apparent from the genetic-strain trial by Kolver et al. (2007), which compared NZF and HF cows on pasture diets supplemented with either 0, 3 or 6 kg DM/day per cow of concentrates during a 22-month EL. The HF had a higher FCE than did NZF when supplemented with 6 kg DM concentrate, whereas no differences between these genetic strains were found for cows supplemented with 0 and 3 kg DM during the EL period. Furthermore, Phyn et al. (2008) showed that FCE during the EL, compared with the normal 10-month lactation, was greatest for HF (i.e. 98% vs 88%). The BCS increased linearly with an increasing supplementation level in both NZF and HF cows, but the increases were much lower in the HF. Metabolic responses also appeared to differ between NZF and HF, indicating that HF cows favoured partitioning of nutrients towards milk production rather than to body tissues during EL (Kay et al. 2009).

Taken together, the data indicated that supplementing cows with ≤6 kg/day per cow of grain or concentrates, depending on the pasture quality, makes it possible to maintain a marginal milk response during a period of EL. Higher levels of supplementation may lead to increased partitioning of nutrient towards body gains at the expense of milk production. Moreover, genetics are also important in maintaining milk production during a period of EL and in determining the extent of nutrient partitioning towards either milk or body reserves.

Effect of extended lactation on milk processing and dairy products

As discussed earlier, milk composition is not constant during lactation, and during the post-peak declining phase of lactation significant changes in milk composition occur. Collectively, these changes may adversely affect the processing characteristics of milk and potentially the type of product that can be made from such milk (Lucey 1996; Truong et al. 2016).

During EL, a greater proportion of milk is produced during the post-peak declining phase of lactation, which may exacerbate the effect of undesirable changes in milk composition on dairy processing. A small number of studies on the effect of EL on milk-processing have been reported (Table 3).

Maciel et al. (2016) measured milk composition and quality, as well as milk sensory and cheesemaking properties (i.e. rennet coagulation time, curd firming rate, gel strength, wet and dry curd yields) in the same cows during what would normally be mid-lactation (140–175 DIM) and late-lactation (280–315 DIM), as well as during EL (385–420 DIM). Interestingly, all measures of cheesemaking properties showed an improvement as lactation progressed, with the biggest improvement during EL. In addition, no adverse sensory characteristics were detected by a trained sensory panel. In a follow-up study on a commercial dairy herd, no adverse effects of EL on cheesemaking and milk sensory characteristics were detected (Maciel et al. 2017). Consistent with these results, there were no adverse effects of EL on milk quality, including proteolytic activity in milk. The latter is in contrast to the results of a study conducted with late-lactation cows on an all-pasture diet (Nicholas et al. 2002). However, in that study the daily milk yield of the cows was much lower (i.e. <10 kg/day) than in the two studies of Maciel et al. (2016, 2017) (approximately 20 kg/day) with intensively managed cows.

Consistent with the observation by Maciel et al. (2016), milk sampled from pasture-fed cows during EL, compared with those with a normal-length lactation (i.e. 300 days) had improved cheesemaking properties (i.e. curd strength and curd firming rate) and cheese yield, regardless of feeding management (Table 3; Auldist et al. 2010).

Sorensen et al. (2008) assessed processing properties of milk indirectly by calculating the casein number (i.e. casein as a percentage of total milk protein) at various stages of lactation and predicted that the lowering of the casein number with increasing EL would lead to a lower yield of cheese and fermented dairy products. However, this was not what was observed in practice when actual cheese was made from EL milk (Auldist et al. 2010; Maciel et al. 2016, 2017). Therefore, the casein number may not be a good indicator for dairy manufacturing properties from EL milk.

Turner et al. (2008) looked at simulated dairy product yields based on milk yields and composition in cows having a normal lactation length (i.e. 300 days) or EL (Table 3). Milk from EL cows on an all-pasture diet resulted in lower predicted yields of whole milk powder, skim milk powder and whey powder, but increased predicted yields of cheese, butter, butter milk powder, casein and whey protein concentrate. However, in cows where pasture was supplemented with concentrates, the decline in predicted whole milk powder was smaller.

Collectively, the data from the limited number of studies investigating the impact of EL on milk-processing indicated that there may not be an adverse effect of EL on milk processing characteristics. It must be noted that very little research has been conducted in this area and that the focus has been mainly on cheese production. A further caveat is that milk yield during EL was higher in these studies than that achieved with a pasture-only diet. To maintain milk yield in cows on pasture during EL would require feed supplementation. Further research is necessary to determine the extent of feed supplementation that is necessary to maintain processing characteristics for a range of common dairy products, including various milk powders.

Effect of extended lactation on reproduction

From the cow’s perspective, maintaining milk production to support its calf takes, biologically, preference over getting pregnant again. Although improved management and genetic selection during the past 40–50 years have led to impressive annual increases in milk production in the modern dairy cow, this appears to have come at the expense of cow reproductive performance, in both high-input/TMR and pasture-based systems (Pryce et al. 2004; Walsh et al. 2011; Abdelsayed et al. 2015).

One of the inherent implications of extending the length of the lactation is that the interval between parturition and pregnancy is increased and re-breeding is delayed till after peak lactation, when the animal has returned to a state of positive energy balance. Such a practice may lead to improved, albeit intentionally delayed, reproductive performance, which is defined as ‘the inherent capacity of the cow to establish ovarian function postpartum, to show overt oestrus, and to conceive and maintain pregnancy when inseminated at the appropriate time’ (p. 1227) (Darwash et al. 1997).

A number of studies has been conducted to investigate the effect of EL on reproductive performance of dairy cows. Given the negative relationship between high milk production and reproductive performance, it is perhaps not surprising that most of the studies on reproductive performance in association with EL were conducted with cows in high-input/TMR systems (Table 4).

Van Amburgh et al. (1997) found that extending the lactation by 3 months had little effect on reproductive measures. Similar findings were reported when different EL scenarios were modelled (Gaillard et al. 2016c). However, perhaps this lack of an effect was due to cows being treated with somatotropin, which increases milk yield and energy partitioning, or in the latter case, insufficient sensitivity of the model. Other studies have shown a positive effect of EL on reproductive measures, including conception and pregnancy rates (Ratnayake et al. 1998; Larsson and Berglund 2000; Niozas et al. 2019b). Niozas et al. (2019b) showed a clear benefit of increasing the VWP from 40 to 120 or 180 days. Moreover, 74% fewer Ovsynch™ protocols were required in cows with an extended VWP. These results support those in the study by Ma et al. (2022a) where ovarian cycling improved with EL in groups of cows subjected to a VWP of 50, 125 or 200 days. Normal ovarian cycling during 100 days around the end of the VWP (i.e. −50 to 50 days) was 58.0%, 77.3% and 91.9% respectively and fewer cows had prolonged cycles (32.7%, 19.0% and 6.0% respectively). Days until pregnancy after the end of the VWP were fewest in cows with the longest VWP (35.5, 37.3, 19.4 days respectively). Similarly, Edvardsson Rasmussen et al. (2023a) showed that increasing the VWP from 25–95 to 145–215 days in primiparous cows increased oestrus intensity and conception at first-service rate and required fewer inseminations per conception. This will likely result in fewer cows being culled as a result of reproductive failures, as shown by Allore and Erb (2000) by using a simulation model in which the VWP was arbitrarily increased from 50 to 150 days.

A retrospective study of Burgers et al. (2021b), by using data from 13 commercial farms each with different strategies for delaying the calving to first-service interval, indicated no benefit of increasing the VWP (i.e. EL) on services per conception and conception rate. Although these results seem to be at odds with those of other studies, which have reported positive effects of EL on reproductive performance, the authors pointed out that results may not be comparable because of the different strategies for increasing calving interval employed on the different farms. Other studies in commercial herds have shown positive effects of EL. Mellado et al. (2016) and, more recently, Rodríguez-Godina et al. (2021) suggested that EL may be a useful management tool with high-producing cows that are exposed to heat-stress conditions and that fail to conceive in <230 DIM, because many of these cows maintain a high level of production and eventually will become pregnant again.

Most studies looking at the possible benefits of EL on reproductive performance have been conducted with high-yielding cows in high-input/TMR systems (Table 4). However, this does not mean that in a pasture-based system, where the production level of cows is lower than that in high-input/TMR-based systems, there is no adverse relationship between the level of production and reproductive performance (Walsh et al. 2011; Abdelsayed et al. 2015). Butler et al. (2010) showed that in a pasture-based system carrying cows that failed to get pregnant during a 12-month calving interval through an EL may offer an alternative to culling, as 85% became pregnant during the period of 10 months of EL. Although it was noted that the higher the milk production potential of the cows is, the better they are suited to EL. Moreover, in a farmlet comparison, cows having EL (i.e. 16 months) had better 3-week submission, 6-week in-calf and conception rates and lower not-in-calf rates (Jarman et al. 2020) than did those with a normal 10-month lactation length.

Overall, EL appears to improve reproductive performance and allows more time for cows to become pregnant again. Even so, with EL, the negative relationship between milk production and reproductive performance remains. Lehmann et al. (2017) showed that when cows during EL were stratified in low, medium and high milk-yield groups, the cows in the high group had the lowest reproductive performance. Further, when comparing NZF with HF cows in a pasture-based system, Kolver et al. (2007) showed that the high genetic merit for milk production cows (HF) had the worst reproductive performance during a normal 10-month lactation. And although during EL, the 3-week submission rate and the first-service conception rate were similar for NZF and HF, the latter still had a lower 6-week pregnancy rate and higher not-in calf rate.

Effect of extended lactation on animal health and welfare

Extended lactation and environmental consequences

Effect of extended lactation and economics

For EL to become an acceptable on-farm management tool, its implementation has to be at least cost-neutral and preferably increase farm profitability. A number of studies have been conducted to quantify the economic impact of EL, on the basis of actual farm data (van Amburgh et al. 1997; Arbel et al. 2001; Malcolm 2005; Butler et al. 2010; Browne et al. 2014; Dalcq et al. 2018; Burgers et al. 2022) or on simulated scenarios (Rotz et al. 2005; Inchaisri et al. 2011; Kok et al. 2019; Lehmann et al. 2019). Although most studies have indicated a neutral or positive impact of EL on profitability, it is difficult to draw firm conclusions, given the range of models, input variables and assumptions employed and different durations of EL (Table 5). Moreover, the outputs of the models vary from profitability per cow (van Amburgh et al. 1997; Kok et al. 2019) to that on a whole-farm basis (Butler et al. 2010; Browne et al. 2014; Lehmann et al. 2019). Given that EL affects animal factors and farm-wide factors, such as amount of feed grown or purchased, land and fertiliser use and labour requirements, a whole-farm model is preferrable however complex it may be. The model used in the study by Lehmann et al. (2019) is the most comprehensive in terms of the range of inputs.

Feed costs and the ability of cows to maintain milk production during EL (i.e. lactation persistency) are key drivers determining the profitability of EL (Table 5). However, to focus mainly on feed costs and the production level is too simplistic and future models may need to accommodate considerably more detail. For example, if there are fewer health-risk periods associated with EL, simply adjusting culling rate may not be sufficient. Cows that fail to get pregnant within the constraints of a 12-month calving interval may become pregnant again when EL allows for an extended VWP, and remain profitable, in both high- and low-input farm systems (Butler et al. 2010; Mellado et al. 2016). Moreover, fewer health-risk periods may result in fewer veterinary expenses and less discarded milk resulting from antibiotic treatments (Burgers et al. 2022). To maintain genetic progress, with fewer calves to select from, so called genetic lag, the use of sexed semen may have a positive effect on economic returns with EL (Clasen et al. 2019).

Increasingly, environmental impacts of farm management practices have to be taken into account when evaluating farm profitability. Extending lactation may decrease GHG emissions (Browne et al. 2014; Lehmann et al. 2019) and this may increase profitability depending on the carbon-offset cost (Browne et al. 2014).

Changes in consumer perception around animal welfare, euthanasia or treatment of dairy calves, and antibiotic use are increasingly affecting the social license to operate of the dairy sector. All have the potential to be improved by EL and increase the market value of the products derived from these systems. Models need to be able to accommodate such changes to be useful to the sector to guide system and management decisions.

Effect of extended lactation on the seasonal pasture-based farming system

Implementation of EL as a routine management practice in a low(er)-cost- and -input pasture-based system is more complex than in a high-input/TMR-based dairy system. This is because in the latter, a similar diet, based largely on conserved forages supplemented with concentrates and/or grains, is fed throughout the lactation, and extending the lactation will, by and large, not affect the daily management of the dairy operation. Whereas in a pasture-based system, cows usually calve in spring, when high-quality pasture is readily available to meet their high metabolic demands and are commonly dried-off during winter when pasture growth is at its lowest. Such a system lends itself well to a 12-month calving interval. Although grass growth can, as a result of climatic change, increasingly continue during the winter season (Jarman et al. 2020) and this can increase farm profitability when cows are milked during winter (Chikazhe et al. 2017), extending lactation means that (more) forage will have to be conserved on-farm and additional supplements (e.g. concentrates and/or grains; maize silage) will need to be brought in to supplement cows through at least one winter season, so as to maintain milk production longer.

Another complexity is that in a pasture-based system, the level of milk production per cow is usually lower than that from cows in high-input/TMR systems, where cows have been selected to maximise milk production per unit of feed, rather than maximise milk production from pasture (Kolver et al. 2007). This means that in pasture-fed cows maintaining milk yield during an EL at a profitable level may be challenging.

Whereas feed costs and maintaining milk production (i.e. lactation persistency) are key drivers determining profitability of EL (Table 5), there are other factors to consider. These include a spread in labour input (especially around seasonal calving, when all cows in the herd calve in a short period of time, and subsequent mating time) and income across the year (Borman et al. 2004) and, potentially, other factors such as animal health and welfare and environmental implications, as discussed earlier.

This means that implementing EL in a pasture-based farm system must be seen in the context of a full dairy-systems approach (Borman et al. 2004; Butler et al. 2010; Lehmann et al. 2014, 2019). The only study in which a systems approach was applied to assess the effect of EL is the New Zealand study by Jarman et al. (2020; Table 1). These researchers compared cows milked for 10 months with those milked for 16 months (i.e. an 18-month calving interval) in a farmlet-type trial (Table 1). Despite similar levels of total MS per cow between the two farmlets, more feed supplements were used in the EL farmlet. Reproductive performance and BCS were improved with EL. Importantly, the authors noted that because of such carry-over effects, the impact of EL must be measured over several lactation cycles. This study did not take into account other factors such as the need for fewer replacements, animal health and welfare, environmental impacts or labour requirements, nor was an economic analysis performed.

Finally, the seasonality of pasture-based farming means that all cows are dried off during the winter period. Traditionally, this ‘quiet time’ of approximately 2 months allows time to catch up on farm-maintenance activities and more time for the pursuit of off-farm activities. Implementing EL means that farmers no longer have the luxury of this ‘quiet time’ period.

Given the unique challenges faced when implementing EL into a pasture-based system, more research is needed to establish the short- and long-term effects of EL at a farm-system level, what such a system might look like (e.g. should all cows in the herd or only high-producing and/or primiparous cows be subjected to EL), and how it affects animal health and welfare, labour requirements and work environment, farm economics, environmental parameters, milk flow to processors, and milk quality and processability.

Conclusions

The concept of EL, where lactation is extended beyond 305-day lactation length, has been discussed in relation to high-input/TMR and low(er)-input pasture-based farming systems. High-producing cows and those that have high lactation persistency are most suited for EL, as late-lactation milk-yield losses are the smallest in such cows. Changes in milk composition mimic those normally occurring during late lactation and they appear to have little to no effect on milk sensory and processing characteristics; although mostly only cheese manufacturing has been considered at this stage. The greater interval between calving and mating associated with EL can improve reproductive performance. Animal health and welfare may also improve with EL because there are fewer health-risk days over the lifetime of the animal. Similarly, EL results in fewer pregnancies during the lifetime of the cow. This means that there are fewer cow–calf separation events and fewer ‘surplus’ calves being slaughtered. As such, EL may help address adverse consumer perceptions towards such practices. Compared with EL in an intensive farming system where feed supply and consistency remain relatively unchanged, implementing EL in a pasture-based system is more challenging in terms of securing feed supplements to manage effects of seasonal changes in pasture quantity and quality to maintain milk yield at a high-enough level to prevent cows drying themselves off. Moreover, limited (≤6 kg/day) supplementation of concentrates may help maintain production but genetically cows bred for pasture-based systems seem to be more prone to partitioning the extra energy towards increasing body condition than towards milk. In contrast, high-producing cows reared under intensive farming conditions appear to be genetically inclined to direct nutrients more towards milk production. More research to explore how best to implement an EL strategy in a pasture-based system is warranted, given the potential benefits of such a practice. This requires a systems approach to assess the impact on farm economics, GHG emissions, freshwater quality, animal health and welfare, labour requirements and work environment, milk flow to processors, and milk quality and processability.