Can Swine Movement Networks Drive Regional Patterns of PRRSV Genetic Diversity?

Introduction/Background

Characterizing PRRSV genetic diversity is important for understanding viral circulation patterns, identifying areas with greater co-circulation of diverse variants, and assessing the risk of new variant introduction. These conditions may promote viral recombination, increase disease pressure, and challenge vaccine effectiveness, ultimately supporting targeted disease surveillance and control strategies across swine-producing regions. Traditional administrative regions (e.g., counties, states) and animal movement-derived communities provide complementary approaches for describing viral diversity and epidemiological connectivity. Movement-derived communities may reveal indirect connections between geographically separated production systems but connected through other channels, e.g., animal movements, highlighting potential pathways for viral circulation and spread that are not captured by traditional regional boundaries. The objective of this study was to characterize PRRSV genetic diversity across counties and animal movement-derived communities within different swine production systems. By comparing diversity patterns across administrative and movement-based spatial frameworks, the study aimed to understand how swine movement connectivity shapes PRRSV diversity across swine-producing regions and identify regions with increased co-circulation of genetically diverse strains.

Methods

Animal movement data, ORF5 PRRSV genetic sequences, and premises metadata from the Porcine Regional Information Management Ecosystem (PRIME), an epidemiological cyberinfrastructure developed at Iowa State University, were analyzed for the 2023-2024 and 2024-2025 PRRSV seasons. PRRSV seasons were defined according to industry convention, running from July 1 through June 30 of the following year. Eligibility criteria included sites from systems with consistent movement data availability throughout the study period. Premise-level movement networks were first constructed using individual premises as nodes to evaluate direct animal movement patterns. County-level movement networks were then constructed by aggregating premises-level movements into county-to-county edges. Community detection was performed using the fast-greedy modularity optimization algorithm applied to undirected weighted networks. Genetic diversity within counties and movement-derived communities was characterized using sequence counts, genetic richness, and Simpson’s diversity index (Simpson, 1949). Genetic richness was defined as the number of distinct PRRSV sublineages identified within each geographic or movement-defined unit and, when variant classification data were available, the number of distinct PRRSV variants detected. Simpson’s diversity index was used to describe the diversity of PRRSV sublineages and variants circulating within each county or movement community. The index ranges from 0 to 1, where values closer to 0 indicate low diversity dominated by a single sublineage or variant, and values closer to 1 indicate higher diversity with multiple sublineages or variants circulating more evenly. The index can also be interpreted as the probability that two randomly selected sequences belong to different sublineages or variants.

Results

Genetic data included 2,032 distinct PRRSV ORF5 sequences from 652 sites. The most frequently identified PRRSV genetic group was L1C.5 (n = 507 sequences), followed by L1D (n = 339), L5A (n = 306), L1A (n = 286), and L1C (n = 235). Additional groups included L5 (n = 131), L1C.2 (n = 123), and L1H (n = 71), while the remaining groups each had fewer than 15 sequences. County-level diversity analysis showed that Southeast, Central, and Northwest Iowa, as well as Western Illinois, had higher Simpson’s diversity values, suggesting a more heterogeneous viral population and potentially multiple sources of introductions of the virus (Figure 1). This is consistent with high animal density, multiple production systems, and high animal movement connectivity. Clusters of higher diversity values were observed, gradually decreasing across neighboring counties.

When evaluated by movement-derived communities, 10 county-based movement communities were identified across the study region (Figure 2). The four communities in Iowa (1, 3, 4, and 8) exhibited high Simpson’s diversity values (>0.75), indicating relatively even circulation of multiple PRRSV variants within highly connected movement regions.

Community 4 showed the highest genetic richness (12), while Communities 1, 3, and 8 also demonstrated elevated diversity despite differences in sequence counts (Table 1). In contrast, Communities 5, 7, and 10 exhibited low diversity values, meaning dominance of a limited number of viral variants. Several counties with low or moderate individual diversity values were part of highly connected movement communities that collectively exhibited high genetic diversity. This suggests that animal movement networks capture epidemiologically relevant links between geographically separated counties that may share PRRSV circulation dynamics despite administrative boundaries.

The premises-level movement network included 1,452 unique premises, of which 88.4% were growing herds (n = 1,297), 8.2% were breeding herds (n = 120), and 2.4% were gilt development units and isolation sites (n = 35). Community detection using premises as nodes is shown in Figure 3, which depicts communities containing at least 5 premises, yielding 17 distinct movement communities. Colored polygons highlight four communities that included premises from multiple production systems, suggesting broader inter-system connectivity patterns (Communities 3, 5, 6, and 9). Community 3 included 201 premises distributed across four states and three production systems, with 335 sequences and a genetic richness of 12, and a Simpson’s diversity index of 0.80. Community 5 contained 181 premises across three states and two production systems, totaling 228 sequences, a genetic richness of eight, and a Simpson’s diversity index of 0.76. Community 6 included 58 premises distributed across three states and four production systems, with 21 sequences, a genetic richness of four, and a Simpson’s diversity index of 0.62. Community 9 contained 145 premises across four states and three production systems, totaling 285 sequences, a genetic richness of 10, and a Simpson’s diversity index of 0.76. The presence of multiple PRRSV sublineages and variants within highly connected multi-system communities suggests that animal movement connectivity may facilitate the introduction, co-circulation, and maintenance of genetically diverse viral populations across production systems and geographic regions.

Conclusions

These findings highlight the importance of evaluating PRRSV diversity beyond traditional regional boundaries and considering the role of animal movement connectivity in shaping viral circulation patterns. While county-level analyses identified important geographic differences in diversity, movement-derived communities revealed indirect epidemiological connections between counties that would otherwise appear unrelated. This shows that systems and regions are not epidemiologically isolated, and that movements occurring within one production system or county may influence viral circulation across broader connected networks. The integration of movement and genetic data through the PRIME infrastructure provides an important framework for identifying regional transmission patterns and supporting more coordinated regional disease surveillance and control strategies.