Anthropogenic Impact and Antimicrobial Resistance Occurrence in South American Wild Animals: A Systematic Review and Meta-Analysis

Abstract

Antimicrobial resistance (AMR) is a significant global challenge that affects environmental, animal, and human health, with reports of antimicrobial-resistant bacteria and antimicrobial resistance genes becoming increasingly common across these domains. This study aimed to systematically review and compare the occurrence of AMR in bacterial isolates from wild animals in South America, focusing on environments with varying levels of anthropogenic impact. Half of the countries in South America documented AMR in wild animals at least once. Most studies focused on specific animal classes, particularly Aves and Mammalia, with a notable emphasis on the orders Chiroptera and Rodentia, as well as the bacterial species Escherichia coli and Salmonella enterica. Subgroup meta-analyses revealed that, for most antimicrobials, the proportion of AMR was significantly higher in environments with a high anthropogenic impact compared to those with a low anthropogenic impact. However, there were no significant differences between the two types of environments for some antimicrobials. Interestingly, certain beta-lactams showed a higher proportion of AMR in environments with low anthropogenic impact. These findings raise important questions regarding the origins and spread of AMR in wild animals, underscoring the necessity for further research to understand the dynamics of AMR in areas with varying levels of human intervention.

1. Introduction

It has been reported that 60–75% of emerging infectious diseases are zoonotic, with approximately 72% of emerging zoonotic events occurring between 1940 and 2004 originating from wild animals [1,2]. However, human activities, such as land-use change and environmental pollution, facilitate the transmission of biological and chemical agents to wildlife, threatening their health and conservation status.

Antimicrobials are particularly concerning among the various chemical compounds due to their widespread use in human and animal health and agriculture. Approximately 73% of antimicrobials used globally are applied in animal production to maintain high yields [3]. This extensive use impacts ecosystems and promotes antimicrobial resistance (AMR), posing a threat to global public health [4,5].

By 2050, AMR is projected to become the leading cause of human mortality, with an estimated 10 million deaths annually, surpassing cancer [6].

Studies suggest that wild animals could serve as effective sentinels for AMR in the environment, as some species act as reservoirs and/or vectors for antimicrobial-resistant bacteria (ARB) and antimicrobial resistance genes (ARG) [7,8,9,10,11,12]. Wild animals living in contaminated environments or near human activity, such as cities, landfills, or intensive agricultural areas, are more likely to acquire ARB and ARG than those living in areas with limited human presence [13,14].

Various clinically relevant ARB have been identified in wild animals, including Enterobacteriacea producing Extended Spectrum Beta-Lactamases (ESBL) and AmpC Type Beta Lactamases, vancomycin-resistant enterococci, methicillin-resistant staphylococci, carbapenemase-producing bacteria, plasmid-mediated colistin resistance, bacteria resistant to fluoroquinolones, and non-typhoidal Salmonella resistant to antimicrobials, among others [15,16,17,18,19,20]. These bacteria have recently emerged at the environment–animal–human interface, with pathogenic and zoonotic potential and limited treatment options. Their presence has been detected globally in wild birds, deer, foxes, small rodents, wild boars, fish, and even insects associated with animal production environments [15,16,17,18,19,20].

A total of 5 out of 25 conservation hotspots in the world, areas with the highest biodiversity and species endemism, are in South America [21]. However, extensive land-use changes have led to significant biodiversity losses, with local species richness declining by up to 31% in highly impacted areas. These changes have affected ecosystem functionality and increased the risk of pathogen transmission from wildlife to humans [22,23,24]. Additionally, data from the region indicate an AMR crisis in South America, with approximately 60% of infections in Brazil and 51% in Bolivia and Peru caused by pathogenic ARB in human populations [25,26,27].

Given the presence of AMR in wild animals, especially in areas with high human activity, and the growing AMR challenge in South America, this systematic review and meta-analysis aim to update, categorize, and analyze the data on AMR in wild animals across the region.

2. Materials and Methods

We conducted this systematic review and meta-analysis following the guidelines of the “Preferred Reporting Items for Systematic Reviews and Meta-Analyses” (PRISMA) [28].

2.1. Search Strategy

A systematic search was conducted across PubMed, SciELO, Scopus, ProQuest, and Web of Science (WoS), widely used academic databases. The search included the terms “antibiotic”, “antimicrobial”, “resistance gen*”, “antimicrobial resistan*”, and “antibiotic resistan*” to identify studies on AMR. To focus on research conducted in South American countries, the search incorporated the terms “Argentina”, “Bolivia”, “Brasil”, “Brazil”, “Chile”, “Colombia”, “Ecuador”, “French Guiana”, “Guyana”, “Paraguay”, “Peru”, “Suriname”, “Venezuela”, “Uruguay”, and “South America”. Finally, the terms “wildlife”, “wild”, “mammal*”, “bird*”, “reptile*”, “amphibian*”, “fish*”, “invertebrate*”, and “mollusk*” were added to the target studies on AMR in wild animals, covering the major animal classes.

It is important to note that while the term “bacterial species” was not included in the search strategy, this was a deliberate choice to avoid excluding abstracts that referenced bacterial species using specific scientific names (e.g., Escherichia coli) without explicitly using broader terms like “bacteria” or “bacterial species”.

The search was conducted across the five electronic databases listed above to identify publications up to 31 December 2022, without setting a lower date limit. We used an Excel spreadsheet to manually match duplicate studies to ensure accuracy. Duplicates were defined as publications with identical titles, authors, and publication years across the mentioned databases. Removing duplicates was critical to avoid inflating the number of studies counted and ensure each study was represented only once in the final dataset. This process also allowed us to accurately analyze the proportions of AMR in the meta-analysis, helping to avoid underestimating and overestimating these proportions.

Finally, we screened the remaining articles according to the inclusion/exclusion criteria for the systematic review and meta-analysis.

2.2. Inclusion and Exclusion Criteria

The inclusion criteria for the systematic review were as follows: (1) publications in English, Spanish, or Portuguese; (2) AMR studies in free-living or captive wild animals in South America; (3) AMR in animal species native to South America. The exclusion criteria were as follows: (1) studies conducted outside South America; (2) studies on non-native animal species; (3) studies without specific data on AMR (e.g., reviews without original data). Three independent researchers (M.P.M., C.U., N.A.) reviewed the initial database and applied the inclusion and exclusion criteria, resulting in a unified database.

For the meta-analyses, we selected studies based on the following additional criteria: (1) the studies reported the number of bacterial isolates tested for AMR; (2) they reported the number of isolates resistant to one or more antimicrobials; (3) they provided the geographic coordinates of the sampling site; (4) they specified the animal species sampled; (5) the studies were conducted on free-living wild animals.

2.3. Data Extraction and Classification

For the systematic review, we organized AMR studies in wild animals from South America into an Excel sheet with the following classifications: publication details (title, authors, year of publication, and year of sampling), country, region, epidemiological approach (diagnostic for resistant bacteria in diseased animals or preventive for bacteria in healthy animals), study design (descriptive or analytical), animal species, bacterial species or samples studied, AMR diagnostic method, sampling method, sample type, and environment (free-living or captive wild animals).

For the meta-analysis, we organized the selected studies into reports in a separate Excel sheet. We defined a report as one or more isolates from specific bacterial species tested for AMR in each study. Consequently, a single study could include multiple reports if different bacterial species were tested or if isolates were obtained from different animal species (Supplementary Material, S1).

For each report, the classification variables included the title of the study, authors, year of publication, sampling year, country, sample size, animal taxonomy, bacterial taxonomy, number of bacterial isolates, antimicrobials and antimicrobial classes tested, coordinates of the sampling location, and the degree of anthropogenic impact at the sampling site (low or high). The classification of environments with low or high anthropogenic impact was determined using the coordinates of the sampling points and a world map with anthropogenic impact layers created by Jacobson and Riggio [29] in Quantum Geographic Information System (QGIS) [30].

Additionally, we recorded the response variable as the proportion of ARB for each specific antimicrobial tested. We created a map of South America, highlighting areas with low and high anthropogenic impact layers, including the sampled reports. This map was created in Excel [31] (Supplementary Material, S2). Moreover, we summarized in two graphics the most studied animal and bacterial species in captive and free-living wild animals. These graphics were created in Miro [32].

2.4. Statistical Analysis

We utilized a meta-analysis approach to quantitatively synthesize data from multiple studies, enabling us to assess trends in AMR across diverse locations.

We conducted two main subsets of analyses using the same dataset. The first subset focused on bacterial species, in general, while the second subset specifically examined the order Enterobacterales. In the first subset, we included all isolates reported in the selected studies. In the second subset, we limited our analysis to isolates from the Enterobacterales order. In both subsets, the response variable was the proportion of isolates exhibiting resistance to a specific antimicrobial, and the explanatory variable was the level of anthropogenic impact on the environment, categorized as either low or high.

For all the analyses, we calculated the proportion of estimated AMR and 95% confidence intervals using a random effects model with the metaprop command from the meta package in R version 4.4.0 [33]. We used Cochran’s Q test to determine whether differences existed between ARB proportions in the subgroups (low versus high anthropogenic impact), with a p-value < 0.05 indicating significant differences. Additionally, we conducted a heterogeneity analysis using the “inconsistency index” statistic (I²). We considered heterogeneity with (I²) > 40% medium or high [34,35].

In the meta-analysis, heterogeneity refers to the differences in the effects the studies aim to measure (in this case, the proportion of ARB). These differences may stem from clinical heterogeneity (e.g., variations in population) or methodological heterogeneity (e.g., differences in how the effect is measured or biases present in the studies) [35]. We did not conduct publication bias analysis, as funnel plots in proportion studies can lead to inaccurate results, such as indicating publication bias when it does not exist [36].

3. Results

3.1. Systematic Review

The systematic review selected 153 studies from 2948 results across PubMed, Scopus, Web of Science, ProQuest, and Scielo. Of these, 60 studies met the inclusion criteria for the meta-analysis, which included 456 reports of ARB isolates from free-living wild animal samples (Figure 1). All reports (isolates) and resistance to each antibiotic can be found in the Supplementary Material (Supplementary Material, S1).

3.1.1. Studies Classification

Brazil had the highest number of publications ($n=105$), followed by Chile ($n=25$), Ecuador ($n=7$), Argentina ($n=6$), Peru ($n=4$), Colombia ($n=3$), French Guiana ($n=2$), and Venezuela ($n=1$). The search yielded no results from Bolivia, Guyana, Paraguay, Suriname, or Uruguay. The publications spanned from 1972 to 2022, with a noticeable increase in studies starting in 2010 (Figure 2).

Of the 153 publications, 122 were descriptive studies (77 wild and 49 captive), while 31 were analytical studies (26 wild and 7 captive). Furthermore, 130 studies had a preventive approach, sampling healthy animals (91 wild and 45 captive), and 23 had a diagnostic approach, sampling sick or dead animals (12 wild and 11 captive). Most studies analyzed samples of cloacal ($n=59$) and rectal ($n=42$). The most used AMR diagnostic technique was the Kirby-Bauer disk diffusion test ($n=115$), followed by Minimum Inhibitory Concentration tests ($n=25$), Whole Genome Sequencing ($n=20$), PCR ($n=20$), and qPCR ($n=6$).

3.1.2. Type of Environment, Animals, and Bacteria Studied

The publications in South America primarily focused on the wild environment, with 103 publications compared to 56 studies on animals in captivity. Most reports on ARB isolated from free-living and captive wild animals mainly involved E. coli and S. enterica (

Table 1. AMR studies according to bacterial species isolated from South American free-living and captive wild animals.

Bacteria Species 1	Free-Living Wild Animals	Bacteria Species 1	Captive Wild Animals
Escherichia coli	38	Escherichia coli	28
Salmonella enterica	21	Salmonella enterica	13
Enterococcus faecium	6	Klebsiella pneumoniae	6
Enterobacter cloacae	5	Proteus mirabilis	5
Enterococcus faecalis	5	Enterobacter cloacae	4
Klebsiella pneumoniae	4	Pseudomonas aeruginosa	4
Pantoea agglomerans	4	Staphylococcus aureus	4
Enterococcus hirae	4	Enterococcus faecalis	3
Citrobacter freundii	4	Pantoea agglomerans	3
Enterococcus casseliflavus	4	Staphylococcus epidermidis	3

¹ This table lists the ten bacterial species with the highest number of studies, while the remaining species are provided in the Supplementary Material, S1.

).

In free-living wild animals, the most studied classes were Aves and Mammalia; the orders Charadriiformes and Chiroptera (detailed in Supplementary Material, S1); and the species Kelp gull (Larus dominicanus), Magellanic penguin (Spheniscus magellanicus), and common vampire bat (Desmodus rotundus). For captive wild animals, the most studied classes were Aves and Mammalia, the order Psittaciformes (detailed in Supplementary Material, S1), and the species Blue-fronted Amazon (Amazona aestiva) (

Table 2. AMR studies according to the animal species of South American free-living and captive wild animals.

Animal Species 1	Free-Living Wild Animals	Animal Species 1	Captive Wild Animals
Larus dominicanus- Kelp gull	8	Amazona aestiva- Blue-fronted amazon	10
Spheniscus magellanicus- Magellanic penguin	7	Amazona amazonica- Orange-winged amazon	4
Larus pipixcan- Franklin’s gull	6	Anodorhynchus hyacinthinus- Hyacinth macaw	4
Chelonia mydas- Green sea turtle	3	Eupsittula cactorum- Caatinga parakeet	4
Carollia perspicillata- Seba’s short-tailed bat	3	Sicalis flaveola- Saffron finch	3
Desmodus rotundus- Common vampire bat	3	Chrysocyon brachyurus- Brazilian maned-wolf	3
Didelphis marsupialis- Common opossum	3	Ara chloropterus- Red-and-green macaw	3
Fregata magnificens- Magnificent frigatebird	3	Ara macao- Scarlet macaw	3
Molossus molossus- Velvety free-tailed bat	3	Cerdocyon thous- Crab-eating fox	3
Nasua nasua- South American coati	2	Ara ararauna- Blue-and-yellow macaws	3

¹ This table lists the ten animal species with the highest number of studies, while the remaining species are provided in Supplementary Material, S1.

).

All 56 studies on captive wild animals specified the animal species from which bacteria or samples were analyzed for AMR. In contrast, 99 out of the 103 studies on free-living wild animals provided this information (Figure 3 and Figure 4).

3.1.3. Meta-Analysis

The meta-analyses included 456 reports from 60 publications on wild environments. Brazil contributed the most publications ($n=36$), followed by Chile ($n=12$), Argentina ($n=4$), Ecuador ($n=3$), Peru ($n=3$), and French Guiana ($n=2$). Most of these studies were conducted in areas with a high anthropogenic impact (46 studies; 288 reports across 78 sampling locations), with fewer studies from regions with a low anthropogenic impact (17 studies; 166 reports across 31 sampling locations) (Supplementary Material, S1). Bolivia, Colombia, Guyana, Paraguay, Suriname, Uruguay, and Venezuela did not contribute studies to the meta-analyses. Out of 456 reports, AMR to 76 different antimicrobials was analyzed, with studies examined between 2 and 30 antimicrobials.

The 76 antimicrobials studied belonged to 18 classes: tetracyclines, sulfonamides, aminoglycosides, macrolides, quinolones, beta-lactams, nitrofurans, amphenicols, beta-lactamase inhibitors, polypeptides, antimycobacterial agents, glycopeptides, streptogramins, fosfomycin, lipopeptides, oxazolidinones, pyridopyrimidines, and lincosamides. Among these, twelve classes (tetracyclines, sulfonamides, aminoglycosides, macrolides, quinolones, beta-lactams, nitrofurans, amphenicols, beta-lactamase inhibitors, glycopeptides, polypeptides, and lincosamides) provided sufficient data to compare ARB proportions between environments with low and high anthropogenic impact.

3.2. AMR Proportions in Bacteria Isolated from Free-Living Wild Animals

Among all antimicrobials analyzed, 20 showed a significantly higher proportion of AMR in environments with high anthropogenic impact compared to those with low anthropogenic impact. Conversely, four antimicrobials exhibited higher AMR in low anthropogenic impact environments. For the remaining 19 antimicrobials, there were no significant differences in AMR proportions between the two types of environments (

Table 3. Proportion of ARB in wild animals based on the anthropogenic impact (AI) of the environments where they were isolated. Proportions with significant differences (p < 0.05) are highlighted in bold. The number of isolates analyzed for each AMR proportion is shown in brackets. I² represents the heterogeneity measure.

Antimicrobial Class	Antimicrobial	Low AI	High AI	p-Value	I2
$\beta$-lactams	Amoxicillin	0.296 (160)	0.916 (190)	<0.0001	82.9%
	Ampicillin	0.552 (388)	0.235 (1565)	0.0001	89.4%
	Aztreonam	0.002 (95)	0.087 (207)	0.0136	51%
	Cefepime	0.076 (198)	0.119 (258)	0.4638	72.6%
	Cefotaxime	0.071 (273)	0.110 (391)	0.4446	77.8%
	Cefoxitin	0.012 (184)	0.026 (216)	0.4864	44.3%
	Ceftazidime	0.039 (338)	0.098 (215)	0.1467	63.2%
	Ceftiofur	0.385 (36)	0.005 (393)	0.0093	42.6%
	Ceftriaxone	0.114 (141)	0.060 (246)	0.3400	74.2%
	Cephalexin	0.669 (97)	0.431 (97)	0.2256	81.2%
	Cephalothin	0.075 (132)	0.675 (110)	<0.0001	77.6%
	Ertapenem	0.000 (165)	0.001 (225)	0.5360	0%
	Imipenem	0.007 (224)	0.004 (166)	0.7132	0%
	Meropenem	0.063 (57)	0.001 (323)	0.0346	0%
	Piperacillin	0.000 (89)	1 (24)	<0.0001	79.1%
	Ticarcillin	0.003 (89)	1 (24)	<0.0001	77.9%
Aminoglycosides	Amikacin	0.000 (356)	0.004 (486)	0.1330	0%
	Gentamicin	0.003 (515)	0.019 (1643)	0.0098	34.6%
	Kanamycin	0.001 (326)	0.043 (466)	0.0111	65.5%
	Netilmicin	0.000 (89)	0.116 (95)	0.0077	37.2%
	Streptomycin	0.003 (253)	0.074 (1433)	<0.0001	83.5%
	Tobramycin	0.000 (89)	0.060 (229)	0.0041	58.5%
Quinolones	Ciprofloxacin	0.015 (375)	0.049 (1188)	0.0566	68.8%
	Enrofloxacin	0.074 (26)	0.032 (299)	0.6245	59%
	Levofloxacin	0.000 (123)	0.170 (68)	0.0080	60.1%
	Nalidixic acid	0.000 (351)	0.193 (667)	<0.0001	80.7%
	Norfloxacin	0.000 (10)	0.125 (723)	0.0367	88.8%
	Ofloxacin	0.000 (89)	0.477 (39)	0.0006	69.5%
$\beta$-lactamase inhibitors	Amoxicillin–clavulanic acid	0.213 (270)	0.286 (361)	0.3563	73.5%
	Piperacillin plus tazobactam	0.000 (101)	0.007 (175)	0.1668	0%
	Ticarcillin–clavulanic acid	0.000 (89)	0.854 (24)	<0.0001	66.4%
Macrolides	Azithromycin	0.000 (34)	0.275 (188)	<0.0001	65.6%
	Erythromycin	0.136 (68)	0.170 (610)	0.7361	86.3%
	Fosfomycin	0.002 (94)	0.444 (23)	0.0008	36%
Sulfonamides	Sulphonamide	0.014 (204)	0.105 (161)	0.0119	41.5%
	Trimethoprim-sulfamethoxazole	0.044 (258)	0.043 (818)	0.9666	66.6%
Tetracyclines	Doxycycline	0.022 (79)	0.016 (29)	0.8757	41.2%
	Tetracyclines	0.055 (430)	0.095 (1873)	0.1892	82.9%
Amphenicols	Chloramphenicol	0.006 (349)	0.034 (1788)	0.0092	65%
Antimycobacterial	Rifampicin	0.775 (4)	0.538 (539)	0.6269	93.3%
Glycopeptides	Vancomycin	0.000 (34)	0.001 (603)	0.6663	69.1%
Nitrofurans	Nitrofurantoin	0.000 (172)	0.014 (737)	0.0246	20.5%
Polypeptides	Colistin	0.112 (135)	0.046 (62)	0.2091	58.3%

). All forest plots are available in the Supplementary Material (Supplementary Material, S3.1).

3.3. AMR Proportions in Enterobacterales Isolated from Free-Living Wild Animals

Within the Enterobacterales order, 13 antimicrobials were significantly associated with environments of high anthropogenic impact, while two were associated with low anthropogenic impact. No significant differences were observed for the remaining 16 antimicrobials (

Table 4. The proportion of Enterobacterales resistant to antimicrobials in wild animals across environments with low and high anthropogenic impact (AI). Proportions with significant differences (p < 0.05) are highlighted in bold. The number of isolates studied within each subgroup for the reported AMR proportions is shown in parentheses. I² represents heterogeneity.

Antimicrobial Class	Antimicrobial	Low AI	High AI	p-Value	I2
$\beta$-lactams	Amoxicillin	0.296 (160)	0.881 (154)	<0.0001	82.5%
	Ampicillin	0.630 ( 320)	0.355 (519)	0.0096	85.1%
	Aztreonam	0.002 (95)	0.113 (191)	0.0079	53.9%
	Cefepime	0.027 (165)	0.146 (229)	0.0340	73.5%
	Cefotaxime	0.011 ( 239)	0.116 (379)	0.0053	78.4%
	Cefoxitin	0.012 (184)	0.033 (178)	0.4223	50.7%
	Ceftazidime	0.002 (295)	0.175 (215)	<0.0001	62.5%
	Ceftiofur	0.443 (35)	0.005 (392)	0.0068	43.8%
	Ceftriaxone	0.039 (96)	0.065 (217)	0.5650	60.5%
	Cephalexin	0.669 (97)	0.325 (81)	0.1372	82%
	Cephalothin	0.018 (99)	0.675 (110)	<0.0001	80.4%
	Ertapenem	0.000 (165)	0.002 (209)	0.5005	0%
	Imipenem	0.000 (181)	0.006 (137)	0.1945	0%
	Meropenem	0.004 (24)	0.001 (258)	0.7756	0%
Aminoglycosides	Amikacin	0.000 (467)	0.002 (927)	0.1764	0%
	Gentamicin	0.001 (437)	0.027 (772)	0.0012	22.5%
	Kanamycin	0.001 (292)	0.001 (75)	0.9175	0%
	Streptomycin	0.003 (219)	0.077 (419)	0.0069	76.1%
	Tobramycin	0.000 (89)	0.058 (195)	0.0063	51.3%
Quinolones	Ciprofloxacin	0.010 (297)	0.054 (600)	0.0399	63.2%
	Enrofloxacin	0.091 (25)	0.025 (261)	0.4846	8.2%
	Nalidixic acid	0.001 (351)	0.216 (437)	<0.0001	81.9%
$\beta$-lactamase inhibitors	Amoxicillin–clavulanic acid	0.167 (236)	0.315 (310)	0.0635	71.2%
	Piperacillin plus tazobactam	0.000 (101)	0.012 (138)	0.0943	0%
Sulfonamides	Sulphonamide	0.014 (204)	0.105 (161)	0.0119	41.5%
	Trimethoprim-sulfamethoxazole	0.036 (183)	0.048 (702)	0.7239	74.7%
Tetracyclines	Doxycycline	0.022 (79)	0.016 (29)	0.8757	41.2%
	Tetracyclines	0.064 (362)	0.116 (772)	0.2118	78.8%
Amphenicols	Chloramphenicol	0.004(314)	0.042(715)	0.0050	52.4%
Nitrofurans	Nitrofurantoin	0.000 (138)	0.004 (150)	0.2907	0%

). All forest plots are available in the Supplementary Material (Supplementary Material, S3.2).

3.4. Heterogeneity

Most of the meta-analyses revealed significant differences between settings that exhibited high heterogeneity.

4. Discussion

Numerous studies have investigated the proportion of AMR in samples or isolates from wild animals, with findings broadly indicating a correlation between environments with a high anthropogenic impact and the prevalence of ARB [9,13,14,37,38,39,40,41]. However, some studies of wild animals have reported no such differences [42,43].

Systematic reviews on AMR in wild animals from a global perspective often include limited data from South America. For instance, Vittecoq et al. [39] reported 9 South American studies out of 256 studies worldwide, and the Norwegian Scientific Committee for Food Safety (VKM) [40] included 22 South American studies out of 230 total studies. The broader timeframe and use of region-specific keywords in the present work likely contributed to identifying more studies. Nevertheless, data on AMR in wild animals from South America remain sparse, with gaps in published data from countries like Bolivia, Guyana, Paraguay, Suriname, and Uruguay, as well as certain animal classes such as Amphibia and Insecta.

Encouragingly, regional and global research efforts have increased in recent years [18,20]. This progress is significant as the role of wild animals in AMR transmission dynamics at the environment–animal–human interface remains poorly understood [39,40,44].

Brazil leads South America in AMR studies in wild animals, accounting for nearly 70% of publications in this and previous reviews [40]. The predominant bacterial genera studied in South America include Escherichia, Salmonella, and Enterococcus; this finding aligns with global trends [39,40,45,46]. These genera are commonly reported in both free-living and captive wild animals, focusing on multi-host bacteria of public health relevance, particularly those with zoonotic potential [45,46,47,48]. This emphasis is further underscored by the surveillance-oriented nature of the studies, which primarily aim to determine the proportion of ARB in healthy wild animal populations, a trend also observed in global AMR research [39].

Most studies are conducted in environments with high anthropogenic impact, likely due to the interest in wildlife closely interacting with human populations. This research highlights the presence of ARB with potential transmission pathways to the environment, other animals (wild or domestic), and humans.

Birds, mammals, and reptiles were the animal classes that were analyzed the most. However, studies in South America revealed a greater diversity of orders and species (both animal and bacterial) compared to those conducted on other continents. This increased diversity may be attributed to the continent’s vast biodiversity [12,21,39,40].

For the meta-analysis, the outcome measured was the proportion of ARB isolates in free-living wild animals. Specifically, the numerator represented the number of ARB, while the denominator indicated the total number of isolates tested for AMR. This distinction is crucial, as some studies reported the proportion of AMR as the numerator and the total number of samples as the denominator, which introduces variability that should be considered when comparing different studies [49]. In this review, we opted to use the proportion of ARB, as this measurement was consistently reported across all selected studies, unlike the measurements of AMR based on the total number of samples or animals tested.

Wild animals showed different proportions of ARB based on the anthropogenic impact on their environment. The significantly higher proportions of ARB to aminoglycosides (gentamicin, kanamycin, netilmicin, streptomycin, tobramycin), beta-lactams (amoxicillin, aztreonam, cephalothin, piperacillin, ticarcillin), quinolones (levofloxacin, nalidixic acid, norfloxacin, ofloxacin), macrolides (azithromycin, fosfomycin), amphenicols (chloramphenicol), sulfonamides (sulfonamide), nitrofurans (nitrofurantoin), and beta-lactamase inhibitors (ticarcillin–clavulanic acid) in environments with high anthropogenic impact (Table 3 and Table 4), agrees with previous reports in the world [39]. However, the higher proportion of ARB to beta-lactams (ampicillin, ceftiofur, meropenem) in environments with a low anthropogenic impact had not been previously reported.

To minimize heterogeneity in the data analysis, we focused on AMR enterobacteria isolated from wild animals in low- and high anthropogenic impacted environments to evaluate differences. Similar to the previous analysis for all bacteria, the proportion of ARB was significantly higher in environments with high anthropogenic impact, except the beta-lactams (ampicillin, ceftiofur). This observation may be attributed to the fact that most reports of ampicillin resistance involved Serratia marcescens and Klebsiella oxytoca, while those related to ceftiofur were primarily linked to E. coli. These bacterial species are known to be intrinsically resistant to these antibiotics, suggesting that their overrepresentation in low anthropogenic impact environments could be a confounding factor [50,51].

Regarding meropenem, the higher proportion of ARB in low anthropogenic impacted areas can primarily be attributed to one study that analyzed flamingo feces from altiplano wetlands [52]. The specific origin of the samples in this study likely explains the elevated resistance observed for these antimicrobials in areas with low anthropogenic impact. Dib et al. [53] suggested a correlation between exposition to UV radiation in extreme environments at high altitudes and the presence of ARB.

One limitation of this study is that the higher number of studies conducted in Brazil could distort the results. This may either lead to an overestimation or underestimation of the proportion of ARB in environments with either low or high human impact in other South American countries or regions that have not been investigated.

Moreover, a general limitation of this type of study is that less than 1% of environmental bacteria are considered cultivable, meaning that the proportion of AMR could be underestimated in individual studies and, by extension, in reviews and meta-analyses. To improve this, the inclusion of more PCR, qPCR, whole genome sequencing, and metagenomic studies would be valuable, as these methods would allow for a more thorough investigation of AMR by capturing both culturable and non-culturable bacteria, thereby providing a more complete understanding of the bacterial resistance profiles in wildlife [54].

Finally, to compare multidrug-resistant isolates in studies involving cultivable bacteria, researchers must establish a standardized protocol that specifies a consistent number of antimicrobials to be tested across different research efforts. In this study, we encountered challenges because isolates from various studies were typically tested with different antimicrobials, which hindered our ability to compare the multidrug-resistant isolates effectively.

5. Conclusions

The study of AMR in wild animals is limited in South America compared to North America and Europe, despite the region being home to one of the highest levels of biodiversity and species richness worldwide.

Most AMR studies focus on birds and mammals, with fewer studies on reptiles and amphibians. E. coli and S. enterica were the most commonly studied bacteria. These trends were observed in both free-living and captive wild animals. Additionally, studies on AMR in wild animals focused on preventive measures or surveillance efforts (such as sampling healthy animals), with a noticeable trend of conducting research in areas near human settlements across South America.

Subgroup meta-analyses revealed significant differences in the proportion of ARB to certain antimicrobials between environments with low and high anthropogenic impact on bacteria isolated from wild animals in South America. Specifically, in bacteria of the order Enterobacterales, a higher proportion of AMR to aminoglycosides, beta-lactams, quinolones, amphenicols, sulfonamides, macrolides, nitrofurans, and beta-lactamase inhibitors was observed in environments with high anthropogenic impact, which aligns with existing literature. Conversely, AMR to certain beta-lactams and polypeptides was more prevalent in environments with low anthropogenic impact. Although these particular findings have not been commonly reported in comparative studies, individual studies have observed AMR to beta-lactams in remote areas with minimal human influence.

It is essential to promote the study of AMR in wild animals across environments with different levels of anthropogenic impact, including fragmented environments or those affected by land-use change. This would enhance our understanding of wildlife’s role in maintaining and disseminating ARB and ARG in the environment. Additionally, it would provide insights into how the growing proximity of human settlements influences wildlife populations.