Introduction Keeping backyard, or hobby, poultry (i.e., non-commercial, privately owned poultry including chickens, ducks, geese, guinea fowl, and turkeys, maintained in a household backyard setting) has become very popular in the United States [ 1 ], where some estimates put the number of households with chickens at approximately 10.6 million, up about six percent (6%) in the last two years alone [ 2 ]. Concerns about rising grocery prices and a growing interest in homesteading have driven a surge in poultry-related sales, including a staggering 569% increase in incubator sales in January 2023 compared to the previous year, according to Manna Pro, a major poultry product manufacturer [ 3 ]. Backyard hobby poultry flocks generally range in size from three to five birds in urban settings to 35 birds in mixed urban and rural settings [ 4 , 5 ].
In 2022, North Carolina (NC) was ranked #1 in the U.S. for the commercial production of poultry and eggs, with poultry and egg sales accounting for 51.9% of the state’s total agricultural sales [ 6 ]. Despite the state’s prominence in commercial poultry production, data on backyard hobby poultry flocks remains scarce. Approximately 32,000 NC families with large backyard flocks of 50–500 birds produced an estimated half a million eggs daily for local communities in 2023 [ 7 ].
However, even with the presence of major national and state poultry organizations, such as the National Poultry Improvement Plan (NPIP) and the NC Extension Poultry Extension, there is still a significant gap in the collection of data on backyard poultry, including hobby flocks. In the spring of 2024, 470 cases of Salmonellosis were reported across 48 states, resulting in 125 hospitalizations and one death due to contact with backyard poultry [ 8 ]. Zoonotic infection from backyard poultry could be related to other bacteria commonly associated with poultry, as well; Escherichia coli ( E. coli ) and Campylobacter spp. [ 9 ].
In the U.S., the National Antimicrobial Resistance Monitoring System (NARMS), a joint initiative of the Centers for Disease Control and Prevention (CDC), U.S. Department of Agriculture, and Food and Drug Administration, monitor antibiotic resistance in major foodborne pathogens, including Salmonella spp., E. coli , and Campylobacter spp. [ 10 ]. E. coli producing extended-spectrum beta-lactamase (ESBL-Ec) pose a significant threat to public health and are the recommended target of the World Health Organization’s (WHO’s) global antimicrobial surveillance protocol for low-resource settings, referred to as the “tricycle protocol”, alluding to the integrated nature of “human health, food chain (animals), and the environment” [ 11 ].
As outbreaks of zoonotic diseases continue to rise, collaboration between healthcare providers, veterinarians, and public health officials is critical to educating animal owners, reducing disease transmission risks, and promoting preventative approaches to zoonotic infections, in other words, embracing the One Health approach which incorporates human, animal, and environmental practice [ 12 ]. Nontraditional pets have become more common [ 13 ], including backyard poultry treated as pets, and understanding the nuances of pathogen transmission across this interaction has become increasingly important. Limited information is available regarding the prevalence of antimicrobial resistant bacteria in backyard poultry and their environments [ 14 ], with data on the most common bacteria (e.g., Campylobacter spp.) primarily originating from large scale commercial flocks or retail meat samples [ 14 – 16 ].
It is expected that the microbial ecology of commercial poultry operations is very different from what is present in backyard poultry settings, including hobby poultry [ 17 ]. Commercial and backyard poultry systems differ in several interrelated ways that shape their resident microbiota and resistomes, including housing density and ventilation, diet and feed delivery, antimicrobial use patterns, degree of environmental containment, genetic uniformity of the flock, and the standardization of biosecurity practices [ 17 – 19 ]. Commercial operations typically maintain birds at high stocking densities within enclosed, climate-controlled houses, use standardized formulated feed, follow federal and state biosecurity standards such as those of the National Poultry Improvement Plan, and raise genetically uniform production lines.
In contrast, backyard hobby flocks are smaller, are often housed with outdoor access, are fed varied diets that may include forage and household scraps, contain heritage and mixed breeds, and operate without enforced biosecurity requirements [ 18 ]. Additionally, backyard poultry have exposures unique to their setting, such as proximity to wild birds and other animal species living in the area [ 20 ]. Additional research on the spread of bacteria and antimicrobial resistance in backyard poultry environments is needed [ 15 , 21 – 24 ].
We conducted a two-part project to: 1) to characterize the practices, management, and resource preferences of backyard hobby poultry owners in central NC using a qualitative study design, and 2) describe the occurrence and antimicrobial susceptibility profiles of Salmonella spp., E. coli , and Campylobacter spp. isolated from human and environmental samples obtained at participating backyard hobby flock enclosures. Here, we report on the second part of the project, the characterization of bacterial species, and their antimicrobial susceptibility profiles, isolated from human and environmental samples collected at backyard hobby poultry settings in central NC. Materials and methods Ethics statement This study was approved by the Campbell University Institutional Review Board under protocol number 845 and the North Carolina State University’s Institutional Review Board reliance agreement protocol number 26333.
Study enrollment and sample collection From January 1, 2024, to June 30, 2024, human and environmental samples were collected at 15 backyard hobby poultry settings located across the five NC counties of Iredell, Wake, Johnston, Chatham, and Richmond. Using recruitment methods, eligibility requirements, and written informed consent as previously described, nasopharyngeal swabs (NP) and questionnaire data from backyard hobby poultry owners aged 18 years or older were collected, along with environmental samples from their properties [ 25 ]. The human (i.e., NP swabs) and environmental samples were collected in phosphate-buffered saline, using standard techniques [ 26 ], placed on ice for no more than 2 hours, and transported to the laboratory at Campbell University where they were stored at -80°C until further processing could be completed.
No permits were required for the collection of samples in this study as all samples were collected with permission on the property of backyard hobby chicken owners. Laboratory analysis Samples were processed in accordance with the WHO Tricycle Protocol recommendations [ 11 ]. Sample types included the human NP swabs and convenience samples collected from the backyard poultry settings of enrolled participants, namely poultry fecal matter swabs, poultry water bowl swabs, soil samples, poultry bedding samples, and other environmental swabs.
NP swabs were collected using techniques developed in other studies examining enteric bacteria spread among households exposed to animals [ 27 , 28 ]. Poultry fecal matter swabs were collected from areas with high concentrations of fecal matter, such as below nesting areas. Water bowl samples were collected by swabbing poultry water bowls or water containers.
For each soil sample, a small amount of soil from the chicken coop was collected. Bedding samples were collected from used poultry bedding areas. Other environmental swabs were collected from a variety of areas around the farm, including roosting areas and farm buildings.
Specifically, samples were thawed from frozen and initially processed using a double-culture enrichment procedure using selective broth culture media (Tryptic Soy Broth [TSB] for E. coli , Rappaport Vassiliadis [RV] Broth for Salmonella , and Bolton Broth for Campylobacter ), followed by selective agar plate media [ 29 , 30 ]. After the initial culture of bacteria on non-antibiotic agar plates, up to five bacteria per positive sample were randomly selected for further antibiotic resistance testing and regrown in broth culture. Antibiotic resistance testing E. coli isolates were confirmed as ESBL by Kirby-Bauer susceptibility testing using the criteria defined in the WHO Tricycle Protocol for cefotaxime (CTX), ceftazidime (CAZ), CTX + clavulanic acid (CLA), and CAZ + CLA as paper discs [ 11 ].
All presumptively positive E. coli isolates were also examined for the molecular presence of the two most common ESBL antibiotic resistance genes, TEM and CTX, using previously described methods [ 31 ]. Briefly, isolates were grown in liquid culture medium overnight, and 1 mL of the culture was pelletized at 14,000 rpm for three (3) minutes, and the pellet was resuspended in 300 µL sterile TE Buffer and heated to 98°C for 10 minutes. This suspension was cooled on ice and then centrifuged at 14,000 rpm for two (2) minutes and the supernatant was used directly for PCR.
Real-time PCR [ 31 ] was performed in 20 µL reactions as using 10 µL Absolute qPCR Mix (ThermoFisher Scientific), along with 5 pmol of each primer, 10 pmol of the probe, and RNase-free water. PCR reaction conditions included an initial heating step of 95°C for 20 seconds, and 40 cycles of 95°C for one (1) second and 60°C for 20 seconds. Salmonella and Campylobacter isolates were examined by Kirby-Bauer susceptibility testing using criteria defined in the NARMS and modified by Luangtongkum et al.
(2007) for Campylobacter [ 32 , 33 ]. Isolates were examined for resistance to erythromycin, ciprofloxacin, nalidixic acid, gentamicin, ampicillin, tetracycline, and norfloxacin. Results Our study team collected 17 nasopharyngeal swabs, 15 fecal matter swabs, 12 water bowl swabs, 14 soil samples, 10 bedding samples, and 31 other environmental swabs for a total of 99 samples across the 15 participating farms ( Table 1 ).
10.1371/journal.pwat.0000565.t001 Table 1 Distribution of positive E. coli , Campylobacter , and Salmonella by sample type (n = 99). Sample Type Isolate type present* (N (%)) Human Nasopharyngeal Swab (n = 17)* Fecal Matter (n = 15) Other Environmental Swab (n = 31) Water Bowl Swab (n = 12) Soil Sample (n = 14) Bedding Sample (n = 10) E. coli 0 (0%) 12 (80%) 8 (26%) 5 (42%) 11 (79%) 4 (40%) Campylobacter 0 (0%) 11 (73%) 12 (39%) 6 (50%) 12 (86%) 5 (50%) Salmonella 0 (0%) 0 (0%) 0 (0%) 0 (0%) 1 (7%) 0 (0%) *>=1 isolate of this genus or genus/species identified in the plated sample. E. coli or Campylobacter were found in more than 70% of poultry fecal and of soil samples.
Across the 15 farms examined, 14 had a sample positive for E. coli and 13 had positive samples for ESBL E. coli . For Campylobacter, 14 farms also had positive samples ( S1 Table ). The distribution of each bacterium in each type of environmental sample is presented in Table 1 .
Forty percent (40/99) of the total number of samples presented with E. coli . Eighty percent (12/15) of fecal samples, 26% (8/31) of environmental samples, 42% (5/12) of water bowl samples, 79% (11/14) of soil samples, and 40% (4/10) of bedding samples contained E. coli . Similarly, 46% (46/99) of the total number of samples presented with Campylobacter .
Seventy three percent (11/15) of fecal samples, 39% (12/31) of environmental samples, 50% (6/12) of water bowl samples, 86% (12/14) of soil samples, and 50% (5/10) of bedding samples contained Campylobacter . In contrast, Salmonella was isolated from only one sample, a soil sample; this represents 7% of the soil samples and only 1% of the total number of samples collected. None of the bacteria targeted by our study was detected in the nasopharyngeal swabs collected from participants.
Table 2 presents the results of the Kirby-Bauer susceptibility testing amongst colonies of Campylobacter cultured from the samples collected [ 32 ]. 4.3% (2/46) of Campylobacter isolates presented resistance to ampicillin and 95.7% (44/46) were susceptible. Fifteen of the forty-six (32.6%) of the Campylobacter isolates were resistant to tetracycline, while 60.9% (28/46) were susceptible.
Six and one-half percent (3/46) of samples presented with intermediate resistance to tetracycline. Sixteen of the forty-six (34.8%) of the Campylobacter isolates were resistant to gentamicin, with 54.4% (25/46) expressing intermediate resistance and 10.9% (5/46) fully susceptible. Erythromycin resistance was detected in 91.3% (42/46) of Campylobacter samples.
Intermediate resistance to erythromycin was detected in the remaining 8.7% (4/46) of samples. Four of the forty-six (8.7%) of the Campylobacter isolates were resistant to ciprofloxacin, with 15.2% (7/46) presenting intermediate resistance, and 76% (35/46) of the samples were susceptible to this antibiotic. None of the samples presented full resistance to norfloxacin, although 8.7% (4/46) measured intermediate resistance, and the rest, 91.3% (42/46), were susceptible.
Lastly, 8.7% of Campylobacter isolates were resistant to nalidixic acid; the majority (91.3%; 42/46) were susceptible. The single Salmonella isolate detected was a soil sample and was resistant to erythromycin and tetracycline. 10.1371/journal.pwat.0000565.t002 Table 2 Distribution of antibiotic resistance in Campylobacter isolates (n = 46).
Antimicrobial Agent Susceptible Intermediate Resistant % Resistant Ampicillin 44 0 2 4.3% Tetracycline 28 3 15 32.6% Gentamicin 5 25 16 34.8% Erythromycin 0 4 42 91.3% Ciprofloxacin 35 7 4 8.7% Norfloxacin 42 4 0 0.0% Nalidixic Acid 42 0 4 8.7% All presumptive E. coli positive samples (n = 40) for each sample type were confirmed as ESBL by Kirby-Bauer susceptibility testing ( Table 3 ). In the confirmation of antibiotic resistance genes, all 40 E. coli isolates were tested for the presence of TEM and CTX genes. None of the isolates were positive for CTX; however, 13 (32.5%) were positive for TEM.
10.1371/journal.pwat.0000565.t003 Table 3 ESBL E. coli confirmations by resistance and sample type (n = 40)*. Sample Type Confirmed** ESBL TEM Gene Positive (n = presumptive positive) E. coli E. coli Fecal Matter (n = 12) 9 (75%) 4 (33%) Environmental Swab (n = 8) 3 (38%) 3 (38%) Water Bowl Swab (n = 5) 3 (60%) 1 (20%) Soil Sample (n = 11) 7 (64%) 2 (18%) Bedding Sample (n = 4) 3 (75%) 2 (50%) Total (n = 40) 25 (61%) 13 (32%) *No samples were positive for CTX genes (n = 40). ** Kirby-Bauer susceptibility testing using the criteria defined in the World Health Organization Tricycle Protocol [ 11 ]. Discussion The results of this study indicate that E. coli , Campylobacter , and Salmonella were present among environmental samples collected from 15 unique backyard hobby poultry settings in central NC.
In addition, some of these isolates were resistant to antimicrobials used in human medicine. Human nasal swabs collected in this study did not show presence of E. coli , Campylobacter , or Salmonella isolates. Considering the limited data available regarding presence of these bacteria, or their antimicrobial resistance, in backyard hobby poultry settings - and our finding of antibiotic resistance in isolates from variety of environmental samples - further study to characterize the presence and spread of AMR in this setting is warranted.
Such study could include whole genome sequencing of the detected bacteria or potentially soil profiles of areas with backyard farms. E. coli was the most common bacterium detected in the poultry fecal and soil samples we collected. This suggests that E. coli is a common microorganism found in backyard (hobby) poultry environments.
These results are similar to previous work investigating chickens exclusively raised for meat consumption, where 12.95% (101/780) backyard farm samples comprising “fecal, soil, litter or compost, and feeder and waterer swabs” presented with ESBL E. coli [ 34 ]. Although our study examined backyard poultry primarily produced for egg production or raised as pets, we also saw that 61% of E. coli isolates in this study were ESBL E. coli . The CDC has recently expressed concern about the CTX-M enzyme and the ESBL-Ec ST131 strain, which are spreading resistance to antimicrobials, leading healthcare providers to use more aggressive treatments, which in turn increases the risk of additional resistance developing [ 35 ].
Our study did not detect CTX genes, but 31.7% of the E. coli isolates did present TEM genes. TEM genes, a family of ESBL enzymes, have become less prevalent in recent years, as CTX-M enzymes have risen in prevalence since the early 2000s [ 36 ]. However, Parzygnat et al. [ 14 , 34 ] found that all E. coli isolates from both large backyard (50–500 chickens) and commercial poultry settings contained plasmid-mediated blaCTX-M genes.
Similarly, studies in rural Western Nepal and Northern Thailand have identified variants of blaCTX-M genes in ESBL E. coli isolates from backyard poultry samples [ 37 , 38 ]. Despite screening for AMR E. coli in some settings, there is still the lack of data on backyard poultry and a gap in understanding how AMR detected in these settings can impact human and animal health. When evaluating Campylobacter isolates from samples collected for our study, 8.7% (4/46) were resistant to ciprofloxacin and 8.7% (4/46) were resistant to nalidixic acid.
Both drugs are members of the fluoroquinolone family, a class of bactericidal agents that inhibit bacterial DNA synthesis, and are commonly used for respiratory and urinary tract infections, respectively. Resistance to these antibiotics can complicate patient care in both acute and chronic healthcare settings. In addition, based on the antimicrobial resistance testing performed, more than 90% of the Campylobacter isolates in this study were resistant to a macrolide antibiotic, erythromycin, which inhibits bacterial protein synthesis [ 39 ].
Currently, azithromycin and clarithromycin, derivatives of erythromycin, are more commonly used to treat respiratory tract and mycobacterial infections [ 40 ]. Testing for resistance to these two macrolide antibiotics could provide further insight into the specificity of antimicrobial resistance in the Campylobacter isolates. In our study, 32.6% (15/46) of Campylobacter samples were resistant to tetracyclines.
These patterns of resistance are also reflected in a study from Ontario, Canada, that investigated AMR in two Campylobacter species, C. jejuni and C. coli , from small poultry flocks [ 41 ]. Fecal samples they collected contained C. jejuni with resistance to tetracycline in 77% of chicken samples, 100% in turkey samples, and C. coli resistance in 50% of samples from turkey and 40% of samples from game birds. The researchers in this study also distinguished the different rates of resistance to tetracyclines in the two different species of Campylobacter , and investigated resistance to quinolones and macrolides, which were not as common.
This is consistent with our findings on quinolone resistance and contrasts with our macrolide resistance sampling, as discussed above. Another study conducted in NC found a higher prevalence of Campylobacter on backyard farms (21.9%; 171/780), in flocks defined as 50–500 birds, when compared to than in commercial production (12.1%; 95/780) [ 14 ]. The majority of isolates were C. jejuni , and the remainder were C. coli .
Based on their AMR susceptibility testing from the backyard poultry samples only, isolates presented resistance to ciprofloxacin (27.3% 47/171) and tetracycline (22.7% 39/171). The proportion of isolates with resistance to ciprofloxacin in that study is higher than was observed in our study (8.7%), and lower for tetracycline resistance compared to our study (32.6%). However, our study’s sample size was much smaller: 46 Campylobacter isolates compared with their 171 isolates, and their collection methodology differed from ours.
Only one sample collected in this study, a soil sample, tested positive for Salmonella . Sample processing for Salmonella is notoriously difficult for environmental samples [ 42 ], particularly for previously frozen samples, and it may be that in our samples the processing was a limitation to detection. With only one sample positive for Salmonella in this study, it is difficult to comment on the concern for increased Salmonella resistance in chicken sources and humans, as reported by the NARMS report [ 39 ].
In a similar study conducted in the Greater Boston, MA area, only one backyard poultry flock (2%) of the 53 flocks sampled was positive for Salmonella and possessed resistance to tetracycline and streptomycin [ 43 ]. Two studies from the West Coast presented similar findings regarding the low frequency of Salmonella in backyard flocks. Of the sampled 34 backyard poultry flocks in Washington State, only one flock (3%) was positive for Salmonella [ 44 ].
From the California Animal Health and Food Safety Laboratory system, 1.7% (44/2627) of samples representing 2,347 backyard flocks were positive for Salmonella [ 45 ]. Through PCR-based DNA analysis, the researchers identified serotypes associated with foodborne diseases. Future studies could investigate the relationship between serovars detected in backyard flocks and those detected in clinical cases amongst human and animal illnesses.
In addition, continued monitoring is recommended to identify and mitigate further development and spread of AMR in the Salmonella species. The limitations of our study include environmental factors affecting sample collection, the processing time, the relatively small number of samples taken, and environmental nature of the samples themselves. Recovery of bacteria across the environmental sample types collected in this study likely reflects differences in the conditions that support bacterial persistence.
Soil and accumulated fecal matter retain moisture and are not routinely removed, providing reservoirs that favor bacterial survival; Campylobacter jejuni , in particular, has been shown to survive substantially longer in poultry feces than in used broiler litter, where lower moisture and aeration shorten persistence [ 46 ]. In contrast, poultry bedding is typically drier and, when adequately aerated, less hospitable to desiccation-sensitive enteric organisms [ 46 , 47 ]. Water bowls are generally refreshed frequently and may be treated with chlorine or other sanitizers to limit microbial growth, which further reduces recovery of viable bacteria from this sample type [ 47 ].
Additionally, it would have been beneficial to have other types of human samples included in the study to examine further colonization. The environmental samples were collected from the areas inside poultry habitats; however, future studies that collect fresh fecal samples or cloacal swabs from individual chickens may show a more accurate representation of microbial colonization in the poultry flocks. Similarly, exposure to the elements of weather, i.e., UV radiation, drying, washing, heat, could have influenced the concentration of bacterial colonies, in contrast with fresh fecal, bedding, water, etc. samples.
It is also possible that environmental swabs were procured from areas where only portions of the flock reside. This may provide an inaccurate portrayal of the entire flock. In addition, there could be other bacterial reservoirs on the properties that were not sampled.
Additionally, we did not include a control sample from an area without backyard poultry in this pilot study. Given that NC is such a dense poultry state, it would be useful to document and compare genotypes of bacteria cultured from commercial industrial poultry facilities in proximity to backyard hobby flocks. Additionally, future research could consider collecting additional human samples focused on bacterial colonization of caretakers.
Conclusion The presence of zoonotic pathogens, including E. coli, Campylobacter and Salmonella , in the backyard environment could inform how backyard poultry caretakers consider biosecurity and farming practices in a residential environment. In this study, we did not detect E. coli, Campylobacter , or Salmonella in human NP samples; however, the interaction between backyard poultry owners and their flocks remains an understudied and potential area for zoonotic pathogen transfer, including spread of antibiotic-resistant bacterial species. As a pilot, the results of this study help lay the groundwork for future research investigating the prevalence of resistant bacteria in backyard hobby poultry settings in NC.
Such studies would inform our understanding of the role that backyard hobby flocks play in both human health and agriculture. Supporting information S1 Table Distribution of sample positives across farms for E. coli , Campylobacter , and Salmonella. (DOCX)