China is the world's largest producer and consumer of pork. As the “chip” of the swine industry, the quality of breeding pigs directly determines business competitiveness and the industry's capacity for sustainable development. Over the past several decades, China has imported high-quality breeding stock from Europe and the United States. Pig breeding in China, guided by industry alliances and large companies, established pyramid breeding systems modeled on classic European and American systems, as well as breeding programs for commercial pig crossbreeding lines. These provided important genetic support for the industry's rapid expansion. However, closer consideration and research reveal persistent problems in the commercial implementation of breeding, including a disconnect between nucleus herd selection and market demand, a lack of feedback from multiplication and commercial production data, a focus on importing breeding stock over selection, and a mismatch between selection intensity and replacement efficiency. Market fluctuations, disease disruptions, and bottlenecks in the genetic supply chain make it difficult to break the vicious cycle of “import—decline—reimport,” seriously constraining the high-quality development of China's pig industry. This article offers some preliminary thoughts on applying reverse-pyramid thinking to the commercial implementation of pig breeding, in the hope of prompting further discussion.

 

 

1 International Breeding Concepts and Trends

 

“The core principle is that breeding must serve commercial production, with coordinated improvement of nucleus and production herds guided by regional markets, meat-quality requirements, and end-market returns.” Leading breeding companies in Europe and North America have already taken the lead in upgrading from the “traditional pyramid” to a model driven in reverse by the market, involving the whole herd in selection, and iteratively refining a closed-loop data system, thereby establishing mature commercial breeding systems. For example, Denmark's Danbred, through a nationwide joint breeding system, has maintained PSY at above 35 head, with the top 10% of nucleus farms reaching 42.8 head; the Dutch Topigs Norsvin (Topigs) has optimized specialized lines: its TN70 maternal line is known for exceptional lactation ability and uniformity in offspring, while the TN Tempo sire line stands out for robustness and profitability. France's NUCLEUS and AXIOM, leaders in French pig genetics, focus on selecting for maternal reproductive performance and rearing ability. They have selected for 8-9 pairs of teats, allocated 55% of breeding weight to reproductive indicators such as number born alive, birth weight, and uniformity, and measured the birth weights of 73000 piglets in 2014, with strict control of within-litter uniformity. PIC, the world's largest pig breeding company, applies customer-specific selection indices to optimize breeding objectives based on market demand. In addition to promoting technological innovation and genetic progress in nucleus herds, breeding companies pay greater attention to the actual performance and returns of commercial pig herds. For example, producing more piglets may require nurse sows or supplemental feeding machines; weaning capacity may be a selling point promoted by French genetics and TOPIGS; while maternal traits and the ratio of biological mother sows to piglets are a key economic feature of PIC. Reverse-pyramid breeding does not reject the hierarchical structure of pyramid breeding. Instead, it reverses the logic of selection, the direction of data flow, and the direction in which value is transmitted, changing “transmitting genetic progress downward from the nucleus herd” to “defining selection objectives in reverse, from the commercial end upward.” More broadly, the fierce competition among new energy vehicle manufacturers today also reflects a shift in thinking toward commercializing technology and making it tangible to customers. Denmark has had a unified national joint breeding system, DanAvl, for over 120 years. In 2018, the market structure changed: Danish pig breeding moved from a single monopoly to market competition, giving rise to two technical approaches, Danbred and Danish Genetics (some of which became PIC partners, while Danish Genetics merged with Hypor); and in early 6 of 2026, Denmark abolished its Ministry of Agriculture for the first time in 130 years. The advanced Danish breeding system is undergoing change, which helps explain some of the reasons behind it.

 

2 Breaking through breeding bottlenecks while realizing returns from pig farming

 

After nearly 30 years of change and upheaval, the pig farming industry is very different from its early days of intensive and industrialized lean-type pig production before 2000. While maintaining meat quality and body conformation, large-scale breeding pig enterprises in China seek to improve their core competitiveness through lines focused on growth and feed conversion efficiency, as well as carcasses with thin, non-exudative backfat. Looking back at early breeding methods, BLUP was initially hampered by insufficient computing power, while whole-genome selection was initially held back by the high costs of chips and testing. Applying both technologies in breeding is no longer a problem, but the genetic levels of breeding pigs and the quality of commercial pigs still vary among breeding companies and enterprises. Classic pyramid breeding is the traditional standard architecture of the global pig breeding industry. Its defining feature is a clear hierarchy with one-way transmission. It consists of three core tiers, each with distinct functions and responsibilities: The first tier is the Nucleus, the “source” of the entire breeding system. It focuses primarily on pure-line selection, bringing together individuals with the highest genetic levels. The population is relatively small, but selection intensity is highest. Its core task is to develop superior pure lines and accelerate genetic progress, determining the long-term potential of the entire breeding system. The second tier is the Multiplier, the “amplifier” of genetic advantages. Its main task is to multiply the purebred or parent-stock pigs developed by the nucleus and pass on the nucleus’s superior genes in large numbers to meet the production needs of the commercial tier. The third tier is the Commercial tier, the “end point” for realizing genetic value. It primarily uses parent-stock pigs supplied by the multiplier to produce commercial pigs, realizing the ultimate economic value; it is the “revenue-generating end” of the entire breeding system. To enable large-scale breeding and meet biosecurity requirements, the nucleus herd can also be divided into the Genetic Nucleus (GN) and the Production Nucleus (PN). Pig farming enterprises and producers consistently focus on the genetic advantages and progress of the parent-stock sows and terminal sires they receive. The scientific mechanisms of genetic transmission and the economic mechanisms of value together form the essential link between breeding and production.

 

Europe and North America have maintained leading positions in pig breeding over the long term not because of the pyramid structure itself, but because they rely on well-developed operating mechanisms, rigorous standards, and advanced technologies. Their efficient operations are reflected in four main areas, offering important lessons for China’s inverted-pyramid breeding approach:

 

1 First, efficient selection in the nucleus ensures the pace of genetic progress. For example, PIC has performance-testing records for more than 900 pigs and complete pedigrees spanning 27 generations, providing precise data to support breeding value evaluations. In 2014, French NUCLEUS recorded maternal rearing ability indicators for 8241 litters and leg-quality assessment data for 10000 pigs, achieving comprehensive coverage of selection indicators.

 

2, full-chain data interoperability makes genetic progress traceable. From the growth performance, reproductive performance, and carcass quality of individuals in the nucleus herd, to the reproductive efficiency of the multiplier herd, and then to the feed-to-weight ratio, lean meat percentage, survival rate, and meat quality indicators of the commercial herd, the entire process is traceable and analyzable.

 

3, specialized lines have clear positioning and focus on commercial value. European and American companies strictly distinguish between the breeding objectives for sire and dam lines. Sire lines focus on production-efficiency indicators such as growth rate, feed-to-weight ratio, and carcass quality, while dam lines focus on reproductive indicators such as total number born, number weaned, lactation capacity, and service life. This avoids the slow genetic progress caused by “balanced selection.” Breeding objectives are adjusted approximately 5 years, using genetic response prediction algorithms to adapt to new market demands. For example, French companies AXIOM and NUCLEUS raised the weight assigned to reproductive performance in dam lines to 55%, with a focus on optimizing the number of live-born piglets and within-litter uniformity. In recent years, they have also placed greater emphasis on high productivity in French Large White pigs through intensified selection.

 

4, genomic selection and the large-scale application of CT shorten the generation interval. In addition, the EU project Feed-a-Gene is continuously opening up new possibilities for selecting traits, such as traits related to maternal behavior in sows (e.g., reducing piglet crushing) and the trait of piglets not biting tails.

 

In summary, the commercialization of breeding must neither focus solely on realizing short-term returns from current benefits nor solely on making isolated breakthroughs in breeding technology bottlenecks; neither can be neglected.

 

3 Preventing the nucleus herd from becoming an “isolated island” and strengthening the commercial application of pig breeding economics

 

Genetic progress is not transmitted effectively: superior genes developed in the nucleus herd decline during multiplication, and some multiplier herds, in an effort to reduce costs, select unqualified breeding pigs indiscriminately. As a result, commercial herds cannot benefit from the genetic advantages of the nucleus herd. Another factor is that performance differences between lines must be evaluated at the same selection rate. Breeding must reflect production realities while taking regional adaptability and market demand into account. For example, in Yunnan and Guizhou, where large pigs are raised, conformation requirements are not high; the suitable lines or cross combinations may differ from those for South China or East China. Batch production means that foster sows are not convenient to use. Lines with 8 pairs of teats are more suitable and easier to manage than sow lines with around 7 pairs.

 

In short, the classic pyramid structure is itself a mature breeding model. However, for a long time, China has adopted only its “hierarchical structure” without mastering its “commercial breeding logic”—overlooking market demand, lacking a data feedback loop, and lacking coordination mechanisms. This has disconnected breeding work from commercial production, preventing genetic value from being converted into economic benefits and maximized. A typical example is that during market downturns, cash-strapped operations cannot purchase superior breeding pigs, while during periods of high prices, indiscriminate purchasing of breeding pigs can lead to a decline in genetic quality.

 

Reverse-pyramid breeding does not overturn the traditional pyramid’s hierarchical structure of “nucleus herd → multiplier herd → commercial herd.” Instead, it fundamentally reverses the breeding logic, direction of data flow, and direction of value transfer. Its core is “putting commercial value at the center, responding to market demand, and achieving coordinated breeding across the entire chain.” Its core definition can be summarized in four “reversals”:

 

1 is a reversal in breeding direction: from “top-down” to “bottom-up.” It breaks with the traditional logic that “the nucleus herd determines the breeding direction,” starting instead from market demand and farming pain points at the commercial end, then working backward to set breeding objectives for the multiplier and nucleus herds. This ensures that every trait selected for in the nucleus herd can translate into economic benefits at the commercial end.

 

2 is a reversal in data flow: from “internal circulation within the nucleus” to a closed feedback loop of “commercial end → multiplier end → nucleus end,” creating a virtuous cycle of “breeding—validation—iteration.”

 

3 is a reversal in value orientation: from “pursuing excellent phenotypes in the nucleus herd” to “pursuing benefits across the commercial end’s entire life cycle,” abandoning the traditional misconception of “emphasizing phenotype and body conformation while neglecting economic returns.” For example, the ranking of crossbred boars based on their own performance testing may reflect hybrid vigor rather than the value of economic traits within the range of heritability.

 

4 is an expansion of the breeding participants: from “breeding only in the nucleus herd” to “precision breeding in the nucleus herd + intensified selection in the multiplier herd + big-data validation at the commercial end.” The selection of international breeding companies for weaning weight and Robust traits is indeed reflected in tracking commercial herds. By focusing on the Guangdong and coastal markets in China, Wens’s selection for body conformation and feed conversion ratio has long earned market recognition in slaughter rate and uniformity.

 

In the traditional pyramid model, the multiplication herd serves only to increase numbers. It multiplies animals without breeding selection or scientifically guided retention, causing the superior genes of the nucleus herd to deteriorate during multiplication and creating a “bottleneck” in the transfer of genetic progress. One of the key improvements of the inverted pyramid model is to develop the multiplication herd into a “second selection tier,” responsible for reverse selection and data feedback, thereby opening up the channel for genetic transfer from the nucleus herd to the commercial herd. All purebred animals in France are tested and evaluated. In China, however, the sow retention mechanism is often overlooked and interrupted at the multiplication-herd level. Strengthening boar selection and using low-density insemination is a feasible strategy that can partly address this issue.

 

The core value of the traditional pyramid breeding system is that it addressed the “multiplication” of genetic advantages in pigs. The inverted pyramid breeding concept does not reject the traditional pyramid; rather, it retains its tiered structure while fundamentally optimizing its selection logic, data flows, and value orientation. At its core, it reinforces the principle of returning to the essence of breeding: serving commercial pig production, the interests of downstream pig enterprises, slaughter companies, and consumers. Both theory and practice show that pig production has moved beyond “single-technology improvement” (such as relying solely on breeding) and entered a stage of upgrading through a combination of factors across the entire chain under the “inverted pyramid” approach.