Exposure of feral swine to Coxiella burnetii overlaps with human Q fever incidence in California

Coxiella burnetii is a zoonotic pathogen that causes Q fever in humans. There are many known reservoirs of C. burnetii, including cattle, sheep, and goats with an expanding list of potential reservoirs including birds, reptiles, ticks and additional mammalian species, such as swine. Feral swine are a highly invasive species in the United States with significant populations and a broad geographic distribution. The role of feral swine in the transmission and spread of C. burnetii is poorly understood, although a recent report identified overlap between feral swine seroprevalence and human Q fever incidence in Texas. California accounts for a large proportion of human Q fever cases in the United States and in this study we characterized the seroprevalence of C. burnetii in feral swine populations in the state. Feral swine showed seropositivity rates up to 1.64% indicating some level of exposure and the possibility that they may serve as a reservoir for disease transmission and spread. Overlap with human Q fever incidence was identified in the central region of California. Although this study does not directly link feral swine to human infection, it identified spatial overlap between feral swine seroprevalence and human Q fever incidence in the state of California, possibly due to the presence of ruminants as the principal reservoirs of C. burnetii. The environmental stability and low infectious dose of C. burnetii, coupled with the geographic overlap between feral swine seroprevalence and human Q fever incidence suggests that feral swine may contribute to zoonotic disease transmission and spread.

Introduction

Feral swine are an invasive species in North America. In the early 1500s, explorers introduced swine as a source of food, subsequently leading to the establishment of feral swine populations within the modern day United States (1). Eurasian and/or Russian wild boar were introduced for sport hunting in the 1900s and domesticated pigs have also contributed to current feral swine populations in the United States. Current populations of feral swine in the United States represent a combination of Eurasian and Russian wild boar, escaped domesticated pigs, and their hybrids (2). Feral swine display a broad variety of phenotypic traits including variations in coat color, bristle length, length of nasal passages, and zygomatic breadth, yet are the same species as domesticated pigs, Sus scrofa (1). Common traits of S. scrofa, particularly in feral swine, help them survive in many environments leading to rapid population expansion. A flexible and opportunistic omnivorous diet including grasses, legumes, herbs, seeds, and other plant material increases the geographic range in which feral swine can survive (1). Populations of feral swine can increase quickly due to rapid reproduction rates and high fecundity with each mature female able to produce up to ten offspring each year (3). Decreases in predatory species in focal regions in the United States, combined with dietary flexibility and rapid reproduction rates, has led to significant geographic spread of feral swine in contemporary times (1, 2).

Feral swine are a known source of economic loss due to damage of crops and livestock throughout the world. In the United States, feral swine have contributed to major economic losses in crop producing states like Texas and California (4, 5). Predation of livestock or damage to livestock infrastructure by feral swine is another source of economic loss in the United States and other parts of the world (1, 6, 7). Overall, it has been estimated that damage and control efforts related to feral swine in the United States cost between $0.8 and $1.5 billion annually (8, 9). These estimates are likely conservative as it is difficult to characterize the economic impacts of many aspects of feral swine damage, including the costs associated with transmission of disease from feral swine to livestock, wildlife, or humans. However, a major impact that invasive feral swine have is the transmission of zoonotic diseases, like brucellosis, trichinellosis, tuberculosis, leptospirosis, and classical swine fever, to humans and/or livestock (10). Feral swine in the state of Hawaiʻi have recently been shown to have serological indications of past or current infections with Brucella species indicating that they may contribute to transmission of brucellosis to livestock and potentially humans (11). Increases in seroprevalence towards Trichinella species and Toxoplasma gondii, have been observed in feral swine populations throughout the United States with positivity rates rising to 12.4% and 40.8%, respectively (12, 13). Antibodies against pathogenic Leptospira species have been identified in feral swine of the United States with positivity rates up to 33.8%, suggesting a role in transmission of leptospirosis (14, 15). In Texas, feral swine have shown high seroprevalence rates against Bacillus anthracis indicating potential exposure to the causative agent of anthrax disease (16). Recently, feral swine in the state of Texas and Hawaiʻi were shown to have low but detectable levels of anti-Coxiella burnetii antibodies, indicating potential exposure to the pathogen that causes human Q fever (17).

Q fever is a significant disease in humans caused by C. burnetii that can result in severe complications including chronic hepatitis, osteoarticular infections, endocarditis, or pseudo-tumors (18–21). Coxiella burnetii is an obligate intracellular coccobacillus that grows in a biphasic pattern alternating between a large cell replicating form and an environmentally stable small cell, non-replicating form (22–25). Coxiella burnetii infects mononuclear phagocytes, replicating in a phagolysosome-like structure, and trophoblast cells of the placenta causing preterm birth and fetal mortality in animals and humans (26–30). Infected tissues and fluids from animals contain high concentrations of C. burnetii and have caused outbreaks of Q fever in humans (29, 31–37). Birth products, urine, feces, and milk have been shown to harbor C. burnetii in an increasing list of animal reservoirs including marine mammals, domestic mammals, birds, reptiles, and ticks, potentially leading to disease transmission to animals and humans (18, 28, 38–46). The infectious dose of C. burnetii has been reported to be a single bacterium, increasing the chances of transmission and spread from contaminated sources (47–49). Q fever is an important zoonosis with epidemic potential due to its environmental stability, low infectious dose, and numerous known/potential animal reservoirs.

Whether swine contribute to transmission or spread of C. burnetii to humans is currently unknown, although it is clear from serosurveys that they are susceptible to infection. Domesticated swine in South Korea, China, and Italy have shown low levels of seropositivity toward C. burnetii (50–52). Feral swine have also shown low levels of seropositivity to C. burnetii in broad geographic regions including Portugal, Spain, South Korea, and the United States (17, 53–56). In the United States, it was previously reported that feral swine in the state of California were positive for antibodies against C. burnetii via microagglutination test, with county level rates ranging from 24% to 68% (56). More recently, seropositive feral swine were identified via indirect ELISA in the states of Hawaiʻi and Texas, with positivity rates up to 0.19% and 6.03%, respectively (17). Hawaiʻi and Texas have contrasting rates of human Q fever that was also reflected in the seroprevalence rates of feral swine identified in each state. Overlap between human Q fever incidence and feral swine seropositivity was identified in Texas, indicating that feral swine may contribute to transmission and spread, or have a similar source of exposure to C. burnetii as humans (i.e., domestic ruminants) (17). Like Texas, California has a large population of feral swine and accounts for a significant proportion of human Q fever cases in the United States, accounting for ∼18% of all cases in 2019 (57). In the current investigation, we sought to survey contemporary feral swine populations in the state of California to determine seroprevalence rates and identify any overlap with human Q fever incidence that may exist.

Results

Characteristics of feral swine samples from California

Feral swine serum samples tested in this study were collected during routine population control efforts and archived by Wildlife Services, part of the United States Department of Agriculture, Animal and Plant Health Inspection Service. All sera from California collected between October 2019 and July 2024 were tested in this study for the presence of antibodies specific to C. burnetii. In total, there were 1,297 serum samples from feral swine collected from 27 counties in the state. Sample sizes ranged from one to 206 feral swine per county. Calaveras, Colusa, Sacramento, San Joaquin, Shasta, Tehama, Trinity, Tulare, and Yuba were the least sampled counties with under 10 samples/county. The most heavily sampled counties were Napa (120; 9.3%), San Luis Obispo (206; 15.9%), Santa Barbara (176; 13.6%), and Santa Clara (113; 8.7%). Swine serum samples were classified into three different groups based on age including adult (996; 76.8%), sub-adult (298; 23.0%), and unknown (3; 0.23%). There were 685 (52.8%) female pigs, 605 (46.7%) male pigs, and 7 (0.54%) pigs with no reported sex.

Seroprevalence of C. burnetii in feral swine

Feral swine have been reported to exist in 56 of 58 counties in the state according to the California Department of Fish and Wildlife. California also accounts for a large number (∼18%) of reported Q fever cases in the United States (57). Feral swine may serve as biosentinels to evaluate their potential contribution to the zoonotic spread of C. burnetii for this disease in the state. To test if feral swine in California have been exposed, serum samples were tested for antibodies specific to C. burnetii. Swine samples showed absorbance (Au) values ranging from 0.035 to 0.676 (μ = 0.120, σ = 0.061) with a non-normal distribution determined via the Shapiro–Wilk test for normality (P < 0.0001, Figure 1a). In the 1,297 samples tested, seven (7/1,297; 0.54%) were positive with an Au value above 0.5 indicating the presence of antibodies against C. burnetii (Figure 1a). The positive samples were identified in feral swine sampled in 2020, 2022, 2023, and 2024 (Figure 1b). Four of the positive samples were from male pigs and three from female pigs (Figure 1c). No significant differences were identified when Au values of gender were compared using the Mann–Whitney U test (P = 0.3619). In addition, when positivity and sex were compared using Fisher's exact test, no significant difference was identified (P = 0.7122). Five positive samples were from adult pigs and two positive samples were from sub-adult pigs and no significant difference between these two groups was identified when comparing Au values via the Mann–Whitney U test (P = 0.9260, Figure 1d). Positivity and age class showed no significant difference when compared using Fisher's exact test (P = 0.6644). Positive samples were identified in the months of February, May, June, August, and September (Figure 1e). These results indicate that feral swine have a low but detectable rate of seropositivity in the state of California.

Figure 1

Figure 1. Feral swine in California show low but detectable levels of anti-C. burnetii antibodies. (a) Histogram of 1,297 feral swine sera samples from the state of California tested by indirect ELISA targeting C. burnetii antigens. All samples were normalized to a positive control and any absorbance (Au) value above 0.5 is considered positive. The red line shows frequency distribution of all samples following a non-normal distribution determined by Shapiro–Wilk test for normality. (b) Distribution of Au values for samples separated by collection year showing positive samples identified in 2020, 2022, 2023, and 2024. No significant (ns) difference was identified between Au values based on gender (c) or age class (d) determined by both Mann–Whitney U test and Fisher's exact test. e) Distribution of Au values for samples separated by collection month showing positive samples identified in February, May, June, August, and September. The positivity cutoff on graphs is represented by a dotted line at Au = 0.5. Box plots represent the interquartile range (IQR) and whiskers represent the 5th and 95th percentiles. Dots in box plots are individual samples that fall outside of the whiskers.

Feral swine serum samples were collected from 27 counties in the state of California with a mean sample size of 46.25 (σ = 54.19) during the study period (Figure 2a). Positive samples were identified in Alameda County (Au = 0.64), Butte County (Au = 0.54), Lake County (Au = 0.51), Madera County (Au = 0.55), Santa Barbara County (Au = 0.55), Santa Clara County (Au = 0.68), and Yolo County (Au = 0.55). To better understand how positive samples were distributed throughout the sampled counties in state of California, we analyzed the seroprevalence rate for each county. Raw seropositivity rates ranged from 0% to 5.88% (μ = 0.56, σ = 1.30). Spatial empirical Bayesian smoothing (SBS) was employed to account for bias in sample sizes across counties. Smoothed seropositivity rates ranged from 0% to 1.64% (μ = 0.58, σ = 0.50). Spatial analysis identified regions with increased seropositivity rates in the central California counties of Madera, Merced, and Tuolumne (Figure 2B). SBS seropositivity rates in these counties ranged from 1.04% to 1.51% (Figure 2b). The highest SBS seropositivity rate was identified in the northern county of Tehama at 1.64% (Figure 2b). We analyzed the smoothed seroprevalence data to identify any local indicators of spatial association (LISA) using the Local Moran's I statistic. Through this analysis a significant high-high cluster centralized around Madera County was identified (Figure 2c). LISA analysis of C. burnetii seropositivity in feral swine populations of California show areas of increased prevalence.

Figure 2

Figure 2. Spatial distribution pattern of C. burnetii seroprevalence in California. (a) Sample distribution across the state of California. (b) Spatial empirical Bayesian smoothed (SBS) seroprevalence rates for feral swine in the state of California identified the highest rates in north central Tehama County and central Tuolumne and Madera counties. (c) Spatial clusters characterized by LISA analysis identified a high-high cluster centered around Madera County in the central region of California.

Incidence of human Q fever in California, 2014–2022

Q fever is a reportable disease in the state of California and any suspected cases must be reported to the local health department within one working day of diagnosis. We used publicly available Q fever count data published by California Department of Public Health to calculate rates. Between 2014 and 2022, there were 417 cases of Q fever reported in California accounting for a large proportion of all cases in the United States. Each year between 29 and 66 cases of Q fever were reported across the state (Figure 3a). Q fever was reported in 72.4% of counties with raw incidence rates as high as 8.78 cases per 100,000 population in Mariposa County. The mean raw incidence rate during this time frame was 1.35 cases per 100,000 with a standard deviation of 1.65 (Figure 3b). To account for population variance in each county, we employed SBS and identified rates as high as 3.62 cases per 100,000, with a mean rate across all counties of 1.36 with a standard deviation of 0.83 (Figure 3b). During this time period, Stanislaus County had the highest case rate in the state with 3.62 cases per 100,000 (Figure 3c). In this same region in central California, Mariposa, Tuolumne, Mono, Madera, and Calaveras counties had case rates of 3.36, 2.94, 2.67, 2.64, and 2.54 cases per 100,000, respectively (Figure 3b). Riverside County in the southern region of the state had a case rate of 2.89 cases per 100,000 and Del Norte County in the north had a case rate of 2.58 cases per 100,000 (Figure 3c). We analyzed the smoothed case rates for any clustering using local spatial association analysis via Local Moran's I. A high-high cluster was identified in central California around Calaveras, Madera, Mariposa, Mono, Stanislaus, and Tuolumne counties with two low-high spatial outliers identified in Merced and Alpine counties (Figure 3d). LISA analysis identified Riverside County as a high-low spatial outlier in the southern region of California (Figure 3d). A low-low cluster was identified around Butte, Lassen, Plumas, Shasta, Sierra, Tehama, and Yuba counties in the northeastern region of the state (Figure 3d). San Francisco and Ventura counties along the western coast were also identified as low-low spatial clusters (Figure 3d).

Figure 3

Figure 3. Human Q fever cases in California, 2014–2022. (a) Human Q fever case counts reported by the California Department of Health from 2014 to 2022. (b) Box plot showing the RAW and SBS smoothed case rates of human Q fever per 100,000 population in each county in California from 2014 to 2022. Box plots show the IQR with whiskers representing the 5th and 95th percentiles and dots are individual samples that fall outside of those ranges. (c) SBS case rates of human Q fever across the state of California from 2014 to 2022. (d) Spatial clusters characterized by LISA analysis identified a high-high cluster centered around Calaveras, Madera, Mariposa, Mono, Stanislaus, and Tuolumne counties in the central region of California. Low-high spatial outliers were identified in Merced and Alpine counties in central California. A high-low spatial outlier was identified in the southern region of the state around Riverside County. A low-low cluster was identified in the north eastern region of the state around Butte, Lassen, Plumas, Shasta, Sierra, Tehama, and Yuba counties. San Francisco and Ventura counties along with their spatial neighbors on the western coast were also identified as low-low spatial clusters by LISA analysis.

Discussion

California has large populations of feral swine and accounts for the largest proportion of human Q fever cases in the United States. Texas has similar characteristics to California in that there are large populations of feral swine and a high number of human Q fever cases. The goal of the current investigation was to determine if there was geographic overlap between feral swine seropositivity and human Q fever in California. In addition, previous reports of C. burnetii seroprevalence in California feral swine populations showed county level rates ten times those observed in Texas (17, 56). This work, conducted over 40 years ago, used microagglutination assays to characterize current or previous exposure to C. burnetii, which may account for the large discrepency (56). While this technique has lower sensitivity and specificity than the techniques employed in the current study, it did identify that feral swine may be a potential reservoir for C. burnetii transmission and spread, a question that is still not completely understood to this day.

We found that a small percentage of feral swine in California showed serological signs of current or previous infection with C. burnetii. Male and female feral swine, including adult and subadult age classes, were found to be positive. Positive samples were identified in the months of February, May, June, August, and September with 71.4% of the positive samples occurring between May and August. This trend shows similarity to national observations of human Q fever temporal dynamics and previous reports of feral swine seropositivity (17, 58). While the overall positivity rate for feral swine in California was low, at 0.54%, smoothed county level rates of seropositivity as high as 1.64% were identified, consistent with reports of seroprevalence in domestic and feral swine ranging from 0.8% to 14.6% (17, 50–52). The geographically central counties of Madera and Tuolumne showed high rates of seropositivity in feral swine with the highest reported rate identified in Tehama County in the north central region of the state. A high-high cluster was identified around Madera County in central California, indicating significantly higher rates of seropositivity in this region when compared to the mean of the state-wide rate. This analysis identified regions of higher seropositivity in feral swine populations across the state of California indicating potential areas where they may contribute to transmission and geographic spread of C. burnetii.

While the majority of Q fever cases in humans are asymptomatic, Q fever remains a significant disease that is nationally reportable (22). Between 2014 and 2022 there were 417 cases of human Q fever reported in California with annualized raw incidence rates ranging from 0.07 to 0.17 cases per 100,000. Geographic dissection of human Q fever cases at the county level identified raw rates are as high as 8.78 cases per 100,000 between 2014 and 2022. Smoothing on a spatial scale was used to account for variations in population between counties, identifying smoothed rates as high as 3.62 cases per 100,000 over the nine-year period. The highest smoothed human Q fever incidence rates were identified in the central region of the state including Tuolumne, Mariposa, and Stanislaus counties, with 2.94, 3.36, and 3.62 cases per 100,000, respectively. Analysis of this central region also identified a high-high incidence cluster around Calaveras, Madera, Mariposa, Mono, Stanislaus, and Tuolumne counties and their neighbors, indicating that human rates of Q fever in this region were significantly higher than the overall mean for the state. Merced and Alpine counties neighbor this high-high cluster, but represent low-high spatial outliers due to low rates of human Q fever. In the southern region of the state Riverside County was identified as a high-low spatial outlier with smoothed case rates of 2.89 cases per 100,000. Other reports have highlighted numerous cases of human Q fever from Riverside and the neighboring county of San Bernadino over the last several decades (59, 60). Investigation of Q fever cases identified at the Loma Linda Veterans Affairs hospital in San Bernadino County showed that 70% of all cases reported contact with animals, and 15% reported contact with swine (60). This highlights the importance of understanding the role of animal reservoirs in the spread and transmission of this disease.

While this work does not directly link feral swine to human infection, it identified spatial overlap between feral swine seroprevalence and human Q fever incidence in the state of California. Madera County in the central region of the state showed a high rate of feral swine seropositivity towards C. burnetii and a high rate of human Q fever. In Madera County and its spatial neighbors, feral swine and humans are likely exposed to C. burnetii via the same source: contaminated birth products, urine, feces, or milk from other infected animals. All of these counties contain domesticated goats, sheep, and cattle that are known and accepted reservoirs of C. burnetii (18, 61). This overlap between the presence of known reservoirs, feral swine seropositivity, and human Q fever infections suggests that C. burnetii circulating among domestic ruminants is a likely source of exposure. Additional research is required to confirm the hypothesis that there is a shared source of infection between feral swine and humans. Other regions in the state showed a contradictory pattern. In the north central region of the state high feral swine seroprevalence rates were identified in the county of Tehama. This county and its spatial neighbors have significantly lower rates of human Q fever, the low-low cluster identified around Butte, Lassen, Plumas, Shasta, Sierra, Tehama, and Yuba counties. A similar observation was made in Texas, where feral swine in Jefferson County on the eastern coast showed high rates of feral swine C. burnetii seropositivity but low rates of human Q fever (17). Future work will be focused on identifying the ecological mechanisms that drive this observation where feral swine have high seropositivity and there are low rates of human Q fever.

This work identifies trends of the zoonotic pathogen C. burnetii relating to coincidence in feral swine and humans from 2019 to 2024 but is not without limitations. We do not directly link feral swine to human Q fever incidence and therefore can make no definitive conclusions on the role of feral swine in transmission. Feral swine were sampled from 27 out of 56 (46.6%) counties in California reported to have populations of feral swine (62). We did not survey feral swine from the southern counties of Riverside and San Bernadino, where there are high rates of human Q fever associated with animal exposures (59, 60). Targeted expansion of feral swine serosurveillance for C. burnetii in this region could help expand our understanding of the ecologic mechanisms supporting maintenance of Q fever in animals and its transmission to humans. We used a strict cutoff for positivity that could lead to an underestimate of seroprevalence because this approach reduces false positives. Using a lower cutoff value as other groups have done for serosurveillance would have increased the overall raw positivity rate from 0.54% to 1.54% (63). The feral swine that did not meet our strict cutoff for a positive could have waning anti-C. burnetii antibody titers based on their time of exposure relative to when the samples were collected. Cross-reactions to species like Bartonella, Chlamydia, Rickettsiae, and Legionella are a possibility that could lead to a false positive, however, off-target antibody titers are typically low, making misdiagnosis less likely (40). The commercial kit used in the current study has been extensively used to test swine and other animals for antibodies against C. burnetii, lending confidence to the results (50–52, 55, 64–66).

At present, we cannot exclude the possibility that feral swine contribute to the direct transmission of C. burnetii to humans, but do not provide evidence to suggest that is occurring. Here, we show that feral swine in the state of California are able to generate an immune response against C. burnetii infection and therefore could serve as a biosentinel, or support geographic spread and transmission to other animal reservoirs. This is consistent with other reports identifying feral swine with serological indications of C. burnetii infection in the states of Hawaiʻi and Texas (17). The immune responses mounted by feral swine could be due to predation of domesticated animals, as has been observed in Australia, or the opportunistic feeding of aborted birth products from infected cattle, goats, or sheep (1, 67). Contact of feral swine with contaminated urine, feces, or milk could also be a contributor to seroprevalence rates. Disease spread over time is likely facilitated by multiple factors associated with the pathogen itself, including the low infectious dose of C. burnetii, environmental stability of the small cell variant, and the environmental conditions that support its growth. These factors are additionally augmented by the complex interactions between wild and domesticated animals. Increased geographic range and rising populations of feral swine could potentially exacerbate risks associated with Q fever, highlighting the need for continuing population surveillance and a better understanding of the contribution feral swine have on the spread and transmission of C. burnetii.

Methods and materials

Sample acquisition

Feral swine sera was collected through a program of the United States Department of Agriculture, Animal and Plant Health Inspection Service (USDA-APHIS). The goal of the National Feral Swine Damage Management Program (NFSP) of USDA-APHIS is to help protect agricultural and natural resources, public and private property, and human health and safety by management of invasive feral swine populations in the United States. The NFSP routinely removes feral swine from environments in efforts towards this goal. During routine removal efforts, field personnel collect serum samples with associated metadata allowing for retrospective national disease surveillance and risk analysis (11, 16, 17). Serum samples studied in the present manuscript were obtained through a partnership with the NFSP. Samples were collected between October 2019 and July 2024 in the state of California (n = 1,297). Serum was extracted within 12 h of clotting from postmortem blood collections as previously described (11, 17). The age of male and female feral swine were classified into two groups, adult (≥1 year of age) or subadult (<1 year of age). Serum samples were shipped and stored at −80 °C at the National Wildlife Research Center in Fort Collins, CO, USA until testing.

Serological testing and analysis

An indirect ELISA was used to determine the presence of antibodies against C. burnetii as previously described (17, 50–52, 55). Briefly, all samples were tested using ID Screen® Q Fever Indirect Multi-species ELISA kit (Innovative Diagnostics, France) following manufacturer's instructions in duplicate. This indirect ELISA tests for antibodies against phase I and phase II C. burnetii antigens and has been used to test swine serum previously (17, 50–52, 55). The specificity for this commercial ELISA kit in swine is 100% (CI_95%: 95.58%–100%, n = 83) as determined by the manufacturer. The positive cutoff published by the manufacturer is 50% of the positive control and is the cutoff used in this study. Briefly, each sample's raw absorbance value at 450 nm (Au) was normalized to a positive control on the same plate (Au = 1.0) and any sample with an Au above 0.5 was considered positive, aligning with the manufacturer's cutoff. Of the 1,297 samples tested, 44.5% were tested in an additional independent experiment to ensure accuracy of the results. Normalized Au values were visually represented and analyzed in GraphPad Prism version 9.5.1 (San Diego, California, USA). Statistical analysis of Au values were carried out as described in the text using Prism.

Human Q fever incidence in California from 2014 to 2022

The state of California publishes county level counts of human Q fever cases in yearly summaries of selected communicable diseases. This information is freely available on the California Department of Public Health webpage in the surveillance and statistics section of the Infectious Diseases Branch (https://www.cdph.ca.gov/). County level case counts were extracted from the yearly summary for 2014–2022 to analyze the spatial distribution of human Q fever in the state of California. Populations for each county were extracted from the WorldPop database (https://hub.worldpop.org/) using constrained data at a 100 m resolution for the United States in 2020 (68, 69). Case rates (raw and SBS) for human Q fever were calculated using GeoDa version 1.22 as previously described (11, 17).

Mapping seroprevalence and incidence rates

Feral swine seroprevalence and human Q fever incidence rates were mapped to the county level in the state of California. County level GIS data was accessed from the California State Geoportal (https://gis.data.ca.gov/) and downloaded on May 9, 2024. All maps were generated using the California County Boundaries shapefile in QGIS version 3.34.3-Prizen. Raw feral swine seroprevalence rates were calculated using the number of positive samples and the number of total samples from each county. Human Q fever raw incidence rates were calculated using the number of cases and the population in each county. To account for variations in sample size or population in each county, spatial empirical Bayesian smoothing (SBS) was carried out in GeoDa version 1.22 as previously described (11, 17).

Spatial cluster analysis of seroprevalence and Q fever incidence

Local indicators of spatial association (LISA) analysis was carried out using the local Moran's I method as previously described (70). This method identifies any high-high clusters (hot spots), low-low clusters (cold spots), or high-low/low-high clusters (spatial outliers) using a given data set. Briefly, SBS seroprevalence rates or human Q fever incidence rates were used to identify significant spatial clusters using the local Moran's I statistic in GeoDa version 1.22 as previously empolyed (11, 17, 63). The local Moran's I statistic is designed to reject the null hypotheses of spatial randomness and we used 99,999 computational permutations to determine a pseudo-P-value. We used a statistical cutoff value of P < 0.05 to determine significant spatial clusters based on a first order queen contiguity weight matrix.

Data availability statement

The original contributions presented in the study are included in the article/Supplementary Material, further inquiries can be directed to the corresponding authors.

Ethics statement

Ethical approval was not required for the study involving humans in accordance with the local legislation and institutional requirements. Written informed consent to participate in this study was not required from the participants or the participants' legal guardians/next of kin in accordance with the national legislation and the institutional requirements. The animal study was approved by the United States Department of Agriculture. The study was conducted in accordance with the local legislation and institutional requirements.

Author contributions

IM: Writing – original draft, Writing – review & editing, Funding acquisition, Validation, Investigation, Data curation, Supervision, Formal analysis, Conceptualization, Methodology, Visualization, Project administration. SG: Validation, Methodology, Data curation, Investigation, Writing – review & editing, Formal analysis. MN: Methodology, Formal analysis, Writing – review & editing, Writing – original draft. GF: Data curation, Investigation, Writing – review & editing. JG: Investigation, Writing – review & editing, Data curation. RB: Writing – review & editing, Resources, Formal analysis. VB: Data curation, Resources, Conceptualization, Project administration, Writing – review & editing. BB: Supervision, Conceptualization, Validation, Writing – review & editing, Formal analysis, Funding acquisition, Project administration, Resources, Methodology.

Funding

The author(s) declared that financial support was received for this work and/or its publication. The funding for this project was a Cooperative agreement from the NFSP of USDA-APHIS. The award number is: AP23WSHQ0000C011.

Acknowledgments

This project was supported by a grant from the NFSP of USDA-APHIS. Thank you to the Wildlife Services personnel in California for collecting the feral swine samples used in this study and to the Wildlife Services technicians with the NFSP for organizing the samples and corresponding metadata for this project.

Conflict of interest

The author(s) declared that this work was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Generative AI statement

The author(s) declared that generative AI was not used in the creation of this manuscript.

Any alternative text (alt text) provided alongside figures in this article has been generated by Frontiers with the support of artificial intelligence and reasonable efforts have been made to ensure accuracy, including review by the authors wherever possible. If you identify any issues, please contact us.

Publisher's note

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.

References

1. Bevins SN, Pedersen K, Lutman MW, Gidlewski T, Deliberto TJ. Consequences associated with the recent range expansion of nonnative feral swine. BioScience. (2014) 64:291–9. doi: 10.1093/biosci/biu015

Crossref Full Text | Google Scholar

2. USDA APHIS. History of feral swine in the Americas. Available online at: https://www.aphis.usda.gov/aphis/ourfocus/wildlifedamage/operational-activities/feral-swine/sa-fs-history (Accessed July 07, 2025).

Google Scholar

3. Choquenot D, McIlr J, Korn T. Managing vertebrate pests: feral pigs.

Google Scholar

4. Adams CE, Higginbotham BJ, Rollins D, Taylor RB, Skiles R, Mapston M, et al. Regional perspectives and opportunities for feral hog management in Texas. Wildl Soc Bull. (2005) 33:1312–20. doi: 10.2193/0091-7648(2005)33[1312:RPAOFF]2.0.CO;2

Crossref Full Text | Google Scholar

5. Frederick J. Overview of wild pig damage in California. Vertebr. Pest Conf. Proc. 18th 1998 (1998).

Google Scholar

6. Choquenot D, Lukins B, Curran G. Assessing lamb predation by feral pigs in Australia’s semi-arid rangelands. J Appl Ecol. (1997) 34:1445–54. doi: 10.2307/2405260

Crossref Full Text | Google Scholar

7. Anderson A, Slootmaker C, Harper E, Miller RS, Shwiff SA. Predation and disease-related economic impacts of wild pigs on livestock producers in 13 states. Crop Prot. (2019) 121:121–6. doi: 10.1016/j.cropro.2019.03.007

Crossref Full Text | Google Scholar

8. Pimentel D, Zuniga R, Morrison D. Update on the environmental and economic costs associated with alien-invasive species in the United States. Ecol Econ. (2005) 52:273–88. doi: 10.1016/j.ecolecon.2004.10.002

Crossref Full Text | Google Scholar

9. Pimental D. Environmental and economic costs of vertebrate Species invasions into the United States. In Manag. Vertebr. Invasive Species Proc. Int. Symp. Witmer GW, Pitt WC, Fagerstone K editors USDAAPHISWS Natl. Wildl. Res. Cent. Fort Collins Co, 2007 (2007).

Google Scholar

10. Meng XJ, Lindsay DS, Sriranganathan N. Wild boars as sources for infectious diseases in livestock and humans. Philos Trans R Soc B Biol Sci. (2009) 364:2697–707. doi: 10.1098/rstb.2009.0086

PubMed Abstract | Crossref Full Text | Google Scholar

11. La THA, McMillan IA, Dahal P, Burger AH, Belcaid M, Phelps DM, et al. Tracking sero-molecular trends of swine brucellosis in hawai‘i and the central Pacific. Front Public Health. (2024) 12:1440933. doi: 10.3389/fpubh.2024.1440933

PubMed Abstract | Crossref Full Text | Google Scholar

12. Hill DE, Dubey JP, Baroch JA, Swafford SR, Fournet VF, Hawkins-Cooper D, et al. Surveillance of feral swine for Trichinella spp. And Toxoplasma gondii in the USA and host-related factors associated with infection. Vet Parasitol. (2014) 205:653–65. doi: 10.1016/j.vetpar.2014.07.026

PubMed Abstract | Crossref Full Text | Google Scholar

13. Cleveland CA, Haynes E, Callaghan KC, Fojtik A, Coker S, Doub E, et al. Distribution and prevalence of antibodies to Trichinella spp. And Toxoplasma gondii in wild pigs (Sus scrofa) in the United States. Vet Parasitol. (2024) 325:110090. doi: 10.1016/j.vetpar.2023.110090

PubMed Abstract | Crossref Full Text | Google Scholar

14. Buchholz AE, Katz AR, Galloway R, Stoddard RA, Goldstein SM. Feral swine Leptospira seroprevalence survey in Hawaii, USA, 2007–2009. Zoonoses Public Health. (2016) 63:584–7. doi: 10.1111/zph.12266

PubMed Abstract | Crossref Full Text | Google Scholar

15. Pedersen K, Pabilonia KL, Anderson TD, Bevins SN, Hicks CR, Kloft JM, et al. Widespread detection of antibodies to Leptospira in feral swine in the United States. Epidemiol Infect. (2015) 143:2131–6. doi: 10.1017/S0950268814003148

PubMed Abstract | Crossref Full Text | Google Scholar

16. Maison RM, Pierce CF, Ragan IK, Brown VR, Bodenchuk MJ, Bowen RA, et al. Potential use for serosurveillance of feral swine to map risk for anthrax exposure, Texas, USA. Emerg Infect Dis. (2021) 27:3103–10. doi: 10.3201/eid2712.211482

PubMed Abstract | Crossref Full Text | Google Scholar

17. McMillan IA, Norris MH, Golon SJ, Franckowiak GA, Grinolds JM, Goldstein SM, et al. Serosurveillance of Coxiella burnetii in feral swine populations of Hawaiʻi and Texas identifies overlap with human Q fever incidence. J Clin Microbiol. (2024) 62:e0078024. doi: 10.1128/jcm.00780-24

PubMed Abstract | Crossref Full Text | Google Scholar

18. Angelakis E, Raoult D. Q fever. Vet Microbiol. (2010) 140:297–309. doi: 10.1016/j.vetmic.2009.07.016

PubMed Abstract | Crossref Full Text | Google Scholar

19. Raoult D, Tissot-Dupont H, Foucault C, Gouvernet J, Fournier PE, Bernit E, et al. Q fever 1985–1998. Clinical and epidemiologic features of 1,383 infections. Medicine (Baltimore). (2000) 79:109–23. doi: 10.1097/00005792-200003000-00005

PubMed Abstract | Crossref Full Text | Google Scholar

20. Maurin M, Raoult D. Q fever. Clin Microbiol Rev. (1999) 12:518–53. doi: 10.1128/CMR.12.4.518

PubMed Abstract | Crossref Full Text | Google Scholar

21. Lipton JH, Fong TC, Gill MJ, Burgess K, Elliott PD. Q fever inflammatory pseudotumor of the lung. Chest. (1987) 92:756–7. doi: 10.1378/chest.92.4.756

PubMed Abstract | Crossref Full Text | Google Scholar

22. Eldin C, Mélenotte C, Mediannikov O, Ghigo E, Million M, Edouard S, et al. From Q fever to Coxiella burnetii infection: a paradigm change. Clin Microbiol Rev. (2017) 30:115–90. doi: 10.1128/CMR.00045-16

PubMed Abstract | Crossref Full Text | Google Scholar

23. Coleman SA, Fischer ER, Howe D, Mead DJ, Heinzen RA. Temporal analysis of Coxiella burnetii morphological differentiation. J Bacteriol. (2004) 186:7344–52. doi: 10.1128/JB.186.21.7344-7352.2004

PubMed Abstract | Crossref Full Text | Google Scholar

24. Sandoz KM, Sturdevant DE, Hansen B, Heinzen RA. Developmental transitions of Coxiella burnetii grown in axenic media. J Microbiol Methods. (2014) 96:104–10. doi: 10.1016/j.mimet.2013.11.010

PubMed Abstract | Crossref Full Text | Google Scholar

25. Sandoz KM, Popham DL, Beare PA, Sturdevant DE, Hansen B, Nair V, et al. Transcriptional profiling of Coxiella burnetii reveals extensive cell wall remodeling in the small cell variant developmental form. PLoS One. (2016) 11:e0149957. doi: 10.1371/journal.pone.0149957

PubMed Abstract | Crossref Full Text | Google Scholar

26. Howe D, Shannon JG, Winfree S, Dorward DW, Heinzen RA. Coxiella burnetii phase I and II variants replicate with similar kinetics in degradative phagolysosome-like compartments of human macrophages. Infect Immun. (2010) 78:3465–74. doi: 10.1128/IAI.00406-10

PubMed Abstract | Crossref Full Text | Google Scholar

27. Khavkin T, Tabibzadeh SS. Histologic, immunofluorescence, and electron microscopic study of infectious process in mouse lung after intranasal challenge with Coxiella burnetii. Infect Immun. (1988) 56:1792–9. doi: 10.1128/iai.56.7.1792-1799.1988

PubMed Abstract | Crossref Full Text | Google Scholar

28. Stein A, Louveau C, Lepidi H, Ricci F, Baylac P, Davoust B, et al. Q fever pneumonia: virulence of Coxiella burnetii pathovars in a murine model of aerosol infection. Infect Immun. (2005) 73:2469–77. doi: 10.1128/IAI.73.4.2469-2477.2005

PubMed Abstract | Crossref Full Text | Google Scholar

29. Sánchez J, Souriau A, Buendía AJ, Arricau-Bouvery N, Martínez CM, Salinas J, et al. Experimental Coxiella burnetii infection in pregnant goats: a histopathological and immunohistochemical study. J Comp Pathol. (2006) 135:108–15. doi: 10.1016/j.jcpa.2006.06.003

Crossref Full Text | Google Scholar

30. Million M, Roblot F, Carles D, D'Amato F, Protopopescu C, Carrieri MP, et al. Reevaluation of the risk of fetal death and malformation after Q fever. Clin Infect Dis Off Publ Infect Dis Soc Am. (2014) 59:256–60. doi: 10.1093/cid/ciu259

PubMed Abstract | Crossref Full Text | Google Scholar

31. Delsing CE, Kullberg BJ, Bleeker-Rovers CP. Q fever in The Netherlands from 2007 to 2010. Neth J Med. (2010) 68:382–7. PMID: 2120946321209463

PubMed Abstract | Google Scholar

32. Lennette EH, Welsh HH. Q fever in California. X. Recovery of Coxiella burneti from the air of premises harboring infected goats. Am J Hyg. (1951) 54:44–9.14846789

PubMed Abstract | Google Scholar

33. Tissot-Dupont H, Amadei M-A, Nezri M, Raoult D. Wind in November, Q fever in December. Emerg Infect Dis. (2004) 10:1264–9. doi: 10.3201/eid1007.030724

PubMed Abstract | Crossref Full Text | Google Scholar

34. Porten K, Rissland J, Tigges A, Broll S, Hopp W, Lunemann M, et al. A super-spreading ewe infects hundreds with Q fever at a farmers' market in Germany. BMC Infect Dis. (2006) 6:147. doi: 10.1186/1471-2334-6-147

PubMed Abstract | Crossref Full Text | Google Scholar

35. Hackert VH, van der Hoek W, Dukers-Muijrers N, de Bruin A, Al Dahouk S, Neubauer H, et al. Q fever: single-point source outbreak with high attack rates and massive numbers of undetected infections across an entire region. Clin Infect Dis Off Publ Infect Dis Soc Am. (2012) 55:1591–9. doi: 10.1093/cid/cis734

PubMed Abstract | Crossref Full Text | Google Scholar

36. van der Hoek W, Hunink J, Vellema P, Droogers P. Q fever in The Netherlands: the role of local environmental conditions. Int J Environ Health Res. (2011) 21:441–51. doi: 10.1080/09603123.2011.574270

PubMed Abstract | Crossref Full Text | Google Scholar

37. Welsh HH, Lennette EH, Abinanti FR, Winn JF. Q fever in California. IV. Occurrence of Coxiella burnetii in the placenta of naturally infected sheep. Public Health Rep Wash DC 1896. (1951) 66:1473–7. doi: 10.2307/4587909

Crossref Full Text | Google Scholar

38. Guatteo R, Beaudeau F, Joly A, Seegers H. Coxiella burnetii shedding by dairy cows. Vet Res. (2007) 38:849–60. doi: 10.1051/vetres:2007038

PubMed Abstract | Crossref Full Text | Google Scholar

39. Rodolakis A, Berri M, Héchard C, Caudron C, Souriau A, Bodier CC, et al. Comparison of Coxiella burnetii shedding in milk of dairy bovine, caprine, and ovine herds. J Dairy Sci. (2007) 90:5352–60. doi: 10.3168/jds.2006-815

PubMed Abstract | Crossref Full Text | Google Scholar

40. Anderson A, Bijlmer H, Fournier PE, Graves S, Hartzell J, Kersh GJ, et al. Diagnosis and management of Q fever–United States, 2013: recommendations from CDC and the Q fever working group. MMWR Recomm Rep Morb Mortal Wkly Rep Recomm Rep. (2013) 62:1–30.

Google Scholar

41. Eklund CM, Parker RR, Lackman DB. A case of Q fever probably contracted by exposure to ticks in nature. Public Health Rep Wash DC 1896. (1947) 62:1413–6. doi: 10.2307/4586287

Crossref Full Text | Google Scholar

42. Beaman MH, Hung J. Pericarditis associated with tick-borne Q fever. Aust N Z J Med. (1989) 19:254–6. doi: 10.1111/j.1445-5994.1989.tb00258.x

PubMed Abstract | Crossref Full Text | Google Scholar

43. Pinsky RL, Fishbein DB, Greene CR, Gensheimer KF. An outbreak of cat-associated Q fever in the United States. J Infect Dis. (1991) 164:202–4. doi: 10.1093/infdis/164.1.202

PubMed Abstract | Crossref Full Text | Google Scholar

44. Langley JM, Marrie TJ, Covert A, Waag DM, Williams JC. Poker players' Pneumonia. An urban outbreak of Q fever following exposure to a parturient cat. N Engl J Med. (1988) 319:354–6. doi: 10.1056/NEJM198808113190607

PubMed Abstract | Crossref Full Text | Google Scholar

45. Marrie TJ, Schlech WF, Williams JC, Yates L. Q fever pneumonia associated with exposure to wild rabbits. Lancet Lond Engl. (1986) 1:427–9. doi: 10.1016/S0140-6736(86)92380-9

PubMed Abstract | Crossref Full Text | Google Scholar

46. Buhariwalla F, Cann B, Marrie TJ. A dog-related outbreak of Q fever. Clin Infect Dis Off Publ Infect Dis Soc Am. (1996) 23:753–5. doi: 10.1093/clinids/23.4.753

PubMed Abstract | Crossref Full Text | Google Scholar

47. Tigertt WD, Benenson AS, Gochenour WS. Airborne Q Fever. Bacteriol Rev. (1961) 25:285–93. doi: 10.1128/br.25.3.285-293.1961

PubMed Abstract | Crossref Full Text | Google Scholar

48. Jones RM, Nicas M, Hubbard AE, Reingold AL. The infectious dose of Coxiella Burnetii (Q fever). Appl Biosaf. (2006) 11:32–41. doi: 10.1177/153567600601100106

Crossref Full Text | Google Scholar

49. Brooke RJ, Kretzschmar ME, Mutters NT, Teunis PF. Human dose response relation for airborne exposure to Coxiella burnetii. BMC Infect Dis. (2013) 13:488. doi: 10.1186/1471-2334-13-488

PubMed Abstract | Crossref Full Text | Google Scholar

50. Seo M-G, Ouh I-O, Lee S-H, Kwak D. Detection and genotyping of Coxiella burnetii in pigs, South Korea, 2014–2015. Emerg Infect Dis. (2016) 22:2192–5. doi: 10.3201/eid2212.161236

PubMed Abstract | Crossref Full Text | Google Scholar

51. El-Mahallawy HS, Kelly P, Zhang J, Yang Y, Wei L, Tian L, et al. Serological and molecular evidence of Coxiella burnetii in samples from humans and animals in China. Ann Agric Environ Med AAEM. (2016) 23:87–91. doi: 10.5604/12321966.1196859

PubMed Abstract | Crossref Full Text | Google Scholar

52. Ferrara G, Pagnini U, Improda E, Iovane G, Montagnaro S. Pigs in southern Italy are exposed to three ruminant pathogens: an analysis of seroprevalence and risk factors analysis study. BMC Vet Res. (2024) 20:183. doi: 10.1186/s12917-024-04037-4

PubMed Abstract | Crossref Full Text | Google Scholar

53. González-Barrio D, Martín-Hernando MP, Ruiz-Fons F. Shedding patterns of endemic Eurasian wild boar (Sus scrofa) pathogens. Res Vet Sci. (2015) 102:206–11. doi: 10.1016/j.rvsc.2015.08.014

Crossref Full Text | Google Scholar

54. Pires H, Cardoso L, Lopes A, Fontes M, Matos M, Pintado C, et al. Seropositivity for Coxiella burnetii in wild boar (Sus scrofa) and red deer (Cervus elaphus) in Portugal. Pathogens. (2023) 12:421. doi: 10.3390/pathogens12030421

PubMed Abstract | Crossref Full Text | Google Scholar

55. Choi Y, Yeh J-Y, Lee JK, Michelow IC, Park S. Risk factors associated with Coxiella burnetii in wild boars: a study in South Korea. Prev Vet Med. (2024) 225:106157. doi: 10.1016/j.prevetmed.2024.106157

PubMed Abstract | Crossref Full Text | Google Scholar

56. Clark RK, Jessup DA, Hird DW, Ruppanner R, Meyer ME. Serologic survey of California wild hogs for antibodies against selected zoonotic disease agents. J Am Vet Med Assoc. (1983) 183:1248–51. doi: 10.2460/javma.1983.183.11.1248

PubMed Abstract | Crossref Full Text | Google Scholar

57. CDC. Epidemiology and Statistics. Q Fever. (2025). Available online at: https://www.cdc.gov/q-fever/data-research/index.html (Accessed on July 11, 2025).

Google Scholar

58. Cherry CC, Nichols Heitman K, Bestul NC, Kersh GJ. Acute and chronic Q fever national surveillance—united States, 2008–2017. Zoonoses Public Health. (2022) 69:73–82. doi: 10.1111/zph.12896

PubMed Abstract | Crossref Full Text | Google Scholar

59. Johnson AK, Patel FR, Persichino JG. Q fever in southern California: 11 patient cases from a community-based organization. Am J Trop Med Hyg. (2019) 101:1185–6. doi: 10.4269/ajtmh.19-0557

PubMed Abstract | Crossref Full Text | Google Scholar

60. Akamine CM, Perez ML, Lee JH, Ing MB. Q fever in southern California: a case series of 20 patients from a VA medical center. Am J Trop Med Hyg. (2019) 101:33–9. doi: 10.4269/ajtmh.18-0283

PubMed Abstract | Crossref Full Text | Google Scholar

61. Gilbert M, Nicolas G, Cinardi G, Van Boeckel TP, Vanwambeke SO, Wint GRW, et al. Global distribution data for cattle, buffaloes, horses, sheep, goats, pigs, chickens and ducks in 2010. Sci Data. (2018) 5:180227. doi: 10.1038/sdata.2018.227

PubMed Abstract | Crossref Full Text | Google Scholar

62. Wild Pig Management. Published by the California Department of Fish and Wildlife. Available online at: https://wildlife.ca.gov/Conservation/Mammals/Wild-Pig (Accessed on June 24, 2025).

Google Scholar

63. Ekakoro JE, Lubega A, Kayaga EB, Ndoboli D, Bluhm AP, Wampande EM, et al. An investigation of Burkholderia pseudomallei seroprevalence in market pigs slaughtered at selected pig abattoirs in Uganda. Pathogens. (2022) 11:1363. doi: 10.3390/pathogens11111363

PubMed Abstract | Crossref Full Text | Google Scholar

64. Meredith AL, Cleaveland SC, Denwood MJ, Brown JK, Shaw DJ. Coxiella burnetii (Q-fever) seroprevalence in prey and predators in the United Kingdom: evaluation of infection in wild rodents, foxes and domestic cats using a modified ELISA. Transbound Emerg Dis. (2015) 62:639–49. doi: 10.1111/tbed.12211

PubMed Abstract | Crossref Full Text | Google Scholar

65. Joulié A, Rousset E, Gasqui P, Lepetitcolin E, Leblond A, Sidi-Boumedine K, et al. Coxiella burnetii circulation in a naturally infected flock of sheep: individual follow-up of antibodies in Serum and milk. Appl Environ Microbiol. (2017) 83:e00222-17. doi: 10.1128/AEM.00222-17

Crossref Full Text | Google Scholar

66. Seo M-G, Lee S-H, VanBik D, Ouh I-O, Yun S-H, Choi E, et al. Detection and genotyping of Coxiella burnetii and Coxiella-like Bacteria in horses in South Korea. PLoS One. (2016) 11:e0156710. doi: 10.1371/journal.pone.0156710

PubMed Abstract | Crossref Full Text | Google Scholar

67. Pavlov PM, Hone J. The behaviour of feral pigs, Sus scrofa, in flocks of lambing ewes. Wildl Res. (1982) 9:101–9. doi: 10.1071/WR9820101

Crossref Full Text | Google Scholar

68. Bondarenko M, Kerr D, Sorichetta A, Tatem A. Census/projection-disaggregated Gridded Population Datasets for 189 Countries in 2020 Using Built-Settlement Growth Model (BSGM) Outputs. University of Southampton (2020). doi: 10.5258/SOTON/WP00684

Crossref Full Text | Google Scholar

69. Tatem AJ. Worldpop, open data for spatial demography. Sci Data. (2017) 4:170004. doi: 10.1038/sdata.2017.4

PubMed Abstract | Crossref Full Text | Google Scholar

70. Anselin L. Local indicators of spatial association—LISA. Geogr Anal. (1995) 27:93–115. doi: 10.1111/j.1538-4632.1995.tb00338.x

Crossref Full Text | Google Scholar