Explore the evolution of selection indices in dairy farming, their impact on genetic diversity, and how they cater to unique farming needs. Are there too many indices? Find out here.
If you’ve ever wondered how selection indices evolved over time, then you’re in for a treat. Once regarded as just another ticker tape in the realms of dairy farming, selection indices have morphed into a more nuanced system underpinned by advances in data analytics. It’s a constellation of traits, each bearing its own weight, culminating in a nuanced system we see today. But before we delve deeper, let’s start at the beginning.
“The selection index journey began with the USDA’s Predicted Difference Dollars index which was based only on milk and fat production. The shift over time has been influenced by emerging knowledge on the biology of the cow, innovation in data collection and the ever-evolving dairy economics.”
Now, we see indices that bear the weight of multiple physical and economic traits. Intriguingly, with every leap in scientific understanding and data analytics, the focus expanded from just production to fitness and conformation traits as well. Ready to know how it evolved and transformed over the years? Let’s dive in!
What Traits are in the Index?
So, what really goes into these selection indices? Let’s lay it all out. You’ll notice that over time, the level of emphasis on each trait has seen dynamic shifts with each index revision. What’s interesting is, there’s been a notable quickening in the rate of new traits making their way into the index. This can be attributed to shifts in dairy economics, a deeper grasp of bovine biology, and the enhanced ease of data gathering and transmission.
Travel back in time to 1971 when the USDA released the Predicted Difference Dollars index. This was the first of its kind and primarily revolved around milk and fat production (Norman and Dickinson, 1971). While other traits were considered economically significant back then, milk and fat were the stars of the show due to the ample phenotypic data available.
Fast forward to 1976, and things begin to spice up. Protein yield was incorporated into the Predicted Difference Dollars index, birthing the Milk-Fat-Protein Dollars index (Norman et al., 1979). Later on, in 1984, an exciting new index was introduced focusing on cheese yield (Norman, 1986).
It all fundamentally changed in 1994 when productive life and Somatic Cell Score (SCS) found a place alongside yield traits, marking the first rendition of the Lifetime Net Merit index (VanRaden and Wiggans, 1995). Here’s where it gets ultra-interesting. The amalgamation of fitness, conformation, and production traits made NM$ stand out from its contemporaries.
Meanwhile, Scandinavian countries had already begun recorded health and fertility data in the 1960s, computing genetic evaluations for these traits in the 1970s (Philipsson and Lindhé, 2003). They discovered that selection objectives encompassing traits with low heritabilities could lead to significant improvements in cow health and fertility. Leitch (1994) examined 19 modern selection indices and found that merely two – Danish S-Index and US NM$ included mastitis resistance, and only one incorporated fertility (Danish S-Index) and productive life (US NM$).
In a revealing review based on an independent survey, Philipsson et al., 1994 identified several other countries’ indices (Finland, Norway, Slovenia, and Sweden) that included fitness traits as well. This trend took off, and a decade later, each of the 17 indices reviewed encompassed at least one fitness trait (Miglior et al., 2005).
Today, indices are being progressively packed with more fitness traits (Cole and VanRaden, 2018), to the point that it’s considered unusual if an index fails to include such traits. With this shift in focus and the continued development of these indices, one thing is clear – the understanding and evaluation of overall dairy cow merit is moving towards a more holistic paradigm.
There is No Universal Standard
While it may be tempting to define a single, universal total merit index, the reality is that this is not attainable. The reason being is that every farmer operates in a unique economic and environmental context from their neighboring farms. This concept was first proposed by Gjedrem in 1972, who theorized that every farm should actually be using its own customized selection index that is tailored to its specific financial situation and business objectives.
In practice, farms with overlapping operating and financial characteristics can potentially use the same index with minimal efficiency loss. However, there are challenges in assigning direct economic values to some traits, most notably conformation traits. Breeders’ goals vary considerably which directly impacts their breeding programs. For instance, a commercial dairy that primarily earns its income from the sale of milk solids will have differing income streams and expenses compared to a seedstock breeder who also sells embryos and premium germplasm. Hence, using different indices may be beneficial.
Specifically, the Lifetime Net Merit (LNM) was explicitly developed for use by commercial dairy farmers (VanRaden, 2004), whereas the Holstein Association USA’s Total Performance Index is aimed at registered cattle breeders who often sell both genetics and milk. The need for more than one index stems from the fact that farmers sell their products to varying markets (VanRaden, 2000), and they have personal preferences (Martin-Collado et al., 2015), as well as different strategies for maximizing profits (Berry et al., 2019).
Recognizing these variations, the CDCB currently publishes four separate indices (Lifetime Net Merit, Fluid Merit, Cheese Merit, and Grazing Merit) to offer farmers options that best suit their needs. This approach of providing multiple indices to farmers isn’t unique to the United States. For instance, when the Australian Dairy Herd Improvement Scheme (now DataGene) revised the Australian Profit Ranking index in 2016, it was replaced with three new indices – Balanced Performance Index, Health Weighted Index, and Type Weighted Index (Byrne et al., 2016). These indices offer their farmers the ability to focus on trait groups that align with their priorities and are sound from a technical standpoint.
Are There Too Many Indices Already?
The recent years have witnessed the emergence of numerous new selection indices that are being aggressively marketed to commercial dairy farmers. This is different from the norm observed with the Net Merit Dollars (NM$) and indices released by noteworthy organizations such as the Purebred Dairy Cattle Association (PDCA). Many of these novel indices are being promoted by breeding companies as a strategy to differentiate their products.
Table 2 provides a list of a few selection indices currently available to American dairy farmers. However, this list is not exhaustive, as some organizations prefer to keep their indices confidential. These indices have been developed by different agencies such as the United States Department of Agriculture (USDA), PDCA – particularly by the American Jersey Cattle Association, and commercial establishments like Zoetis.
Despite the variations, there’s a remarkable similarity among most indices, with a bidirectional focus on productivity (a key source of income for most farms) and fitness traits (often directly linked to costs). However, making direct comparisons is a challenge due to availability restrictions, as some indices are only accessible for bulls promoted by the index publisher.
Differences between indices are typically attributed to the inclusion of unique sets of traits, or the differential priority given to these traits in the index. Some companies even opt for proprietary evaluations to differentiate their offerings from their competitors. Correlations among these indices are generally very strong, resulting in minimal reranking of bulls when switching from one index to another. Nevertheless, many farmers may struggle to clearly describe the differences between each index, thereby creating room for confusion. Furthermore, there are concerns that marketers might exaggerate the significance of differences between the indices. Table 2: Some selection indices currently offered to US dairy farmers
Note: The correspondence between the indices is often quite remarkable, revealed by the work of T. J. Lawlor Jr., from the Holstein Association USA (personal communication), and this minimal reranking of bulls when transitioning from one index to another.
| Trait2 | BS PPR (2017) | AY CPI (2019) | GU PTI (2020) | JE JPI (2020) | HO ICC$ (2020) | JE ICC$ (2020) | HO TPI (2020) | USDA NM$ (2018) |
|---|---|---|---|---|---|---|---|---|
| Milk | — | — | — | — | 5 | — | — | −1 |
| Fat | 28 | 25 | 25 | 19 | 14 | 22 | 19 | 27 |
| Protein | 34 | 35 | 25 | 27 | 12 | 22 | 19 | 17 |
| PL | 6 | — | 6 | 5 | 6 | 12 | 5 | 12 |
| SCS | — | −4 | — | −4.5 | −4 | −4 | −4 | −4 |
| UC | 10 | — | 10 | — | 7 | 5 | 11 | 7 |
| FLC | — | — | 10 | — | — | — | 6 | 3 |
| BWC | — | — | — | — | — | — | — | −5 |
| DPR | 12 | 6 | 15 | 9 | 8 | 12 | 9.1 | 7 |
| SCE | — | — | — | — | −2 | — | — | — |
| DCE | — | — | — | — | −1 | — | −0.5 | — |
| SSB | — | — | — | — | −1 | — | — | |
| DSB | — | — | — | — | −1 | — | −1.5 | |
| CA$ | — | — | — | — | — | — | — | 5 |
| HCR | — | — | — | 2 | 5 | 6 | 1.3 | 1 |
| CCR | — | — | — | 3.5 | — | 3 | 1.3 | 2 |
| LIV | 4 | — | 3 | 3 | 2 | 5 | 3 | 7 |
| HLTH | — | — | — | 4.6 | 4 | — | 2 | 2 |
| MO | 6 | — | — | — | — | — | — | |
| TYPE | — | 25 | — | 19.4 | — | — | 8 | |
| UDEP | — | 5 | — | — | — | — | — | |
| STR | — | — | 3 | — | — | — | — | |
| STAT | — | — | 3 | — | — | — | — | |
| DENS | — | — | — | −3 | — | — | — | |
| FEED | — | — | — | — | 16 | — | 8 | |
| POLL | — | — | — | — | 1 | — | — | |
| HAPL | — | — | — | — | <1 | 1 | — | |
| LOCO | — | — | — | — | 6 | — | — | |
| HOOF | — | — | — | — | 1 | — | — | |
| BCS | — | — | — | — | 1 | — | — | |
| MAST | — | — | — | — | 1 | 4 | — | |
| SPD | — | — | — | — | 1 | — | — | |
| TEMP | — | — | — | — | <1 | — | — | |
| CALF | — | — | — | — | — | 4 | — | |
| EFC | — | — | — | — | — | 1 | 1.3 |
1 Due to rounding, columns will sometimes sum to a value slightly smaller or larger than 100. BS PPR = Brown Swiss Progressive Performance Ranking (Brown Swiss Association, 2017); AY CPI = Cow Performance Index (U.S. Ayrshire Breeders’ Association, 2020); GU PTI = Performance and Type Index (American Guernsey Association, 2020); JE JPI = Jersey Performance Index (Tauchen, 2020); HO ICC$ = Ideal Commercial Cows for Holsteins (Genex, 2020a, Genex, 2020b); JE ICC$ = Ideal Commercial Cows for Jerseys (Genex, 2020a,b); HO TPI = Total Performance Index (Holstein Association USA, 2020); USDA NM$ = Net Merit Dollars (VanRaden et al., 2018).2 PL = productive life; UC = udder composite (varies by breed and index); FLC = feet and legs composite; BWC = body weight composite; DPR = daughter pregnancy rate; SCE = sire (direct) calving ease; DCE = daughter (maternal) calving ease; CA$ = calving ability dollars; HCR = heifer conception rate; CCR = cow conception rate; LIV = cow livability; HLTH = health traits (varies by breed and index); MO = mobility (Brown Swiss); TYPE = type (conformation) composite (varies by breed); UDEP = udder depth; STR = strength; STAT = stature; DENS = milk density; FEED = feed intake/feed cost (varies by breed and index); SSB = sire (direct) stillbirth; DSB = daughter (maternal) stillbirth; POLL = polled status; HAPL = haplotypes affecting fertility; LOCO = locomotion; HOOF = hoof health; MAST = clinical mastitis; SPD = milking speed; TEMP = milking temperament; CALF = calf survivability; EFC = early first calving (age at first calving).
Are Selection Indices Responsible for Reducing Diversity in Some Breeds?
At first glance, you’d be forgiven for suggesting that the continued decrease in genetic diversity, particularly in US Holsteins (e.g., Maltecca et al., 2020) could be attributed to breeders doggedly pursuing high-index animals. However, the reality is not so straightforward. The escalation in inbreeding rates are, in fact, more likely sparked by improvements in selection intensity, largely thanks to advances in genomic technology (García-Ruiz et al., 2016).
The rapid cycling of generations, paired with significant gains achieved in each, has led seedstock producers to place heavy focus on the lines that have consistently produced successful bull families. With limited resources available for the identification of elite animals, the threat of losing market share to rivals is a far more potent concern now than in the days of traditional progeny testing programs. This is because of the rapidly accruing genetic gains. As such, the expected decline in the rate of inbreeding under genomic selection, as anticipated by Daetwyler et al., 2007, has not come to pass. Simply put, no major AI company is prepared to risk sourcing largely from outcross families.
Should there be a market for outcross bulls, the collected phenotypes would primarily come from daughters of popular families, leading to a drop in prediction accuracies for the outcross animals. However, looking at the long-term picture, the benefits of diversifying the genetic base could well justify a bit of short-term inaccuracy. You could compare this situation to the balancing act in optimal contribution theory, where alterations in inbreeding are offset by rates of genetic improvement (Clark et al., 2013).
It’s also feasible that the surging number of indices could encourage the emerging development of more distinct Holstein strains. This would increase inbreeding within individual strains but would enhance diversity overall when the strains are crossed. Strategies like this echo those proposed for nucleus herd programs (e.g., Meuwissen, 1998), common features in the swine and poultry sectors. In fact, certain breeding companies offer mating schemes predicated on assigning young sires to genetic lines within the breed (e.g., Select Sires Inc, 2020). The specifics of how bulls are assigned to lines, however, remain undisclosed.
The Bottom Line
In the final analysis, we’ve seen how selection indices have evolved over time, progressively expanding to encompass a wider array of traits reflecting economic, health, and fitness factors. This holistically reflects the varying needs and goals of individual farmers, set in their unique farm environments and economic situations. While there’s been an exponential increase in selection indices, each serves a distinct purpose and is aimed at providing the best possible outcomes for different farming models. Even though multiple indices could induce a level of complexity and confusion among farmers, their fundamental similarity lies in striking a balance between productivity and fitness traits.
One common criticism regarding indices, increased inbreeding resulting in reduced genetic diversity, is not solely tied to the selection indices. Advanced or genomic technologies have accelerated this trend more than the indices themselves. It is paramount that the value of broadening the genetic base is considered, possibly at the expense of some short-term gains. The potential for diversity may also lie in the development of various strains within breeds due to the multiple indices available.
The selection index is a potent tool that empowers farmers to make informed decisions that align with their individual operational context, ultimately working towards the shared goal of maximizing productivity and profitability. Considering the nuances of the indices and striving towards an understanding that serves their unique needs can make a crucial difference to their farming success. As developments and research continue within this field, the hope is for the creation and application of even more comprehensive and farmer-oriented indices in the future.
