Pathogen-Induced Stress in Wild House Finches (Haemorhous mexicanus): Leukocyte Dynamics as Health Indicators Open Access

Understanding stress in wildlife is essential for evaluating the impacts of environmental and biological challenges, particularly by pathogens and parasites, which can have significant effects on individual health and population dynamics. Stress might serve as an indicator of overall host health, including how animals handle stressors, such as infection. Measuring stress in wildlife is therefore important in assessing the physiological consequences of disease exposure, and it can provide insights into the broader ecological impacts of pathogens on wildlife populations. Wild organisms are exposed to multiple stressors, among which pathogens play a major role. Cases of co-infection (when an individual hosts multiple pathogens simultaneously), may compound these effects, altering stress responses and complicating the interpretation of physiological data. A common approach to measure stress in wildlife is through the assessment of adrenal glucocorticoids such as corticosterone, which are “stress” hormones released in response to different challenges, in plasma (Wikelski and Cooke 2006). This technique has the disadvantage that corticosterone levels rise quickly immediately after capture (Romero and Reed 2005), complicating the determination of baseline data. This is especially problematic in wild animals, where capturing the individuals can induce immediate stress that affects hormone levels (Vleck et al. 2000).

Gross and Siegel (1983) observed that when chickens are stressed and their corticosterone levels increase, the proportion of heterophils in the white blood cell (WBC) population increases while the proportion of lymphocytes decreases. They concluded not only that an increase in the H:L ratio is a reliable indicator of physiological stress, but that it should be a better measure of long-term changes in the environment, whereas the concentrations of corticosteroids should be a better measure of short-term changes. The technique has since been extended to explore responses to various stress factors in wild birds (Davis et al. 2008; Cīrule et al. 2012; Carbó‐Ramírez and Zuria 2017; Skwarska 2019).

This differential response of heterophils and lymphocytes is believed to ensure appropriate WBC types are concentrated where they are required for an effective stress response (Dhabhar et al. 1996). Under stress, the release of heterophils into the blood is stimulated, thereby increasing their proportion in the circulation, whereas lymphocytes are redistributed from the blood to organs, resulting in a decrease in their circulating number (Dhabhar 2002). The ratio between these two leukocytes (hereafter H:L ratio) thus offers useful information about an individual’s health and stress status (Lobato et al. 2005; Davis and Maney 2018).

In many ornithological studies, pathogens have been chosen as the stressor of interest to test how the presence or infection intensity of pathogens affects hematological parameters. Some have not found any significant impact when birds were infected (e.g., Krams et al. 2010; Cornelius et al. 2014; Granthon and Williams 2017); one reported a reduced H:L ratio in wild birds dealing with haemoparasites (Dunn et al. 2013), but in general, elevated H:L ratios have been observed in infected wild birds when compared to those not infected (Ots and Hõrak 1998; Davis et al. 2004; Lobato et al. 2005; Wojczulanis-Jakubas et al. 2012; Bale et al. 2020).

Around 1994, a novel strain of the poultry parasitic bacterium Mycoplasma gallisepticum emerged in wild birds and rapidly spread across North America (Dhondt et al. 1998), causing conjunctivitis especially in House Finches (Haemorhous mexicanus). This House Finch–Mycoplasma gallisepticum system has become an excellent model that has provided valuable insights into complex host-pathogen interactions, coevolution, and the dynamics of multiple pathogens in co-infected hosts (e.g., Faustino et al. 2004; Bonneaud et al. 2012, 2020; Grodio et al. 2012; Love et al. 2016; Vinkler et al. 2018; Reinoso-Pérez et al. 2020). Significant changes in leukocyte levels have been reported in both captive and wild House Finches, with experimentally infected birds showing elevated WBC counts compared to noninfected individuals (Davis et al. 2004; Fratto et al. 2014; Bale et al. 2020; Reinoso-Pérez et al. 2020). The increase in WBC counts correlated with pathogen virulence; House Finches can be naturally infected by either poultry M. gallisepticum strains, which are less virulent in House Finches, and in which the M. gallisepticum load in the conjunctiva is not higher than in the choana, or by House Finch M. gallisepticum strains, which are more virulent in House Finches, and have a higher M. gallisepticum load in the conjunctiva (see Reinoso-Pérez et al. 2022). House Finches infected with more virulent M. gallisepticum strains exhibited higher H:L ratios compared to those infected with less virulent strains (Bale et al. 2020). According to Davis et al. (2004) the primary cause for the elevated H:L ratio was an increase in heterophils, while other studies highlighted that a reduction of lymphocytes also contributed to elevated H:L ratios in infected House Finches (Fratto et al. 2014; Bale et al. 2020).

Although most previous studies evaluated the impact of a single pathogen on total WBC counts or H:L ratios, we found in a previous experimental study that in House Finches infected with Haemosporidia (intracellular blood parasites), WBC counts increased following inoculation with M. gallisepticum compared to birds that were not co-infected, indicating an increased immune response to concurrent pathogens. Additionally, in co-infected birds, both haemosporidian parasitemia and M. gallisepticum load increased compared to individuals without co-infection, indicating an increase in stress resulting from co-infection (Reinoso-Pérez et al. 2020). Following up on these findings, this study had the following objectives: 1) Measure WBC count and H:L ratio as a stress indicator in four groups of wild Mexican House Finches that were tested for the presence of M. gallisepticum and Haemosporidia: (a) pathogen-free (i.e., not infected with either M. gallisepticum or haemosporidia), (b), (c) single infected birds with either M. gallisepticum or haemosporidia, and (d) birds co-infected with both pathogens. We hypothesized that physiological stress would increase with infection, reflected in significant variation in leukocyte parameters and H:L ratio among groups. We expected that stress would be more severe in co-infected birds, leading to the highest H:L ratios, whereas pathogen-free birds would exhibit the lowest values, and birds with a single infection would show intermediate values. We further predicted that birds with higher M. gallisepticum loads in the conjunctiva compared to the choana would exhibit larger H:L ratios. 2) Validate in wild birds the findings from captive House Finches, specifically that co-infected birds have higher M. gallisepticum loads and haemosporidian parasitemia compared to those with a single infection (Dhondt et al. 2017b; Reinoso-Pérez et al. 2020).

Between July 2019 and February 2020, 246 House Finches were mist-netted at 20 sites across 12 Mexican states (Baja California, Chihuahua, Durango, Estado de Mexico, Hidalgo, Jalisco, Michoacan, Nuevo Leon, Puebla, Queretaro, San Luis Potosi, and Sonora; Fig. 1). The sampling sites represented a range of habitats, including xerophilous scrublands, grasslands, areas near livestock and poultry farms, and urban zones. Feeders were not used to attract the birds. Instead, before setting up mist nets, each location was visited 1–2 d in advance to observe the behavior of the birds and identify optimal sites for net placement. This strategy was particularly important given the tendency of House Finches to form flocks. Sampling was conducted immediately after capture to minimize handling time and ensure accurate assessment of leukocyte parameters. All applicable institutional and national guidelines for the care and use of animals were followed. Samples were collected under the permit SEMARNAT No. SGPA/DGVS/07141/19.

The following data were collected: date and location of capture, age, sex, weight, wing length, and eye lesion score, following Sydenstricker et al. (2006); Dhondt et al. 2007b. Birds were manually restrained and blood samples were taken from the brachial vein using a 27-gauge needle and a heparinized capillary tube. Additionally, two thin blood smears were prepared for each bird following standard techniques for haemosporidian studies (Valkiūnas 2005), and a drop of blood was collected on a filter paper for subsequent DNA extraction. In the field, the blood smears were air-dried and fixed with 100% methanol. Using 1.25-mm cotton applicators (Puritan Medical Products, Guilford, Maine, USA), swabs from the conjunctiva and choana were collected separately, placed separately in tryptose phosphate broth, and stored at −20 C until DNA extraction in the laboratory.

Blood capillary tubes were centrifuged to record the packed cell volume (PCV%) for each bird. Using the plasma, a rapid plate agglutination (RPA) assay (Sydenstricker et al. 2006) using M. gallisepticum antigen (Charles River Lab Inc., Wilmington, MA, USA) was conducted to detect the presence of antibodies to M. gallisepticum, which would indicate prior exposure to the pathogen (Kleven 2008).

Blood smears were stained in the laboratory with Giemsa stain as described (Petithory et al. 2005). Slides were examined using a Olympus BH2 microscope (Center Valley, Pennsylvania, USA) under 100× oil immersion objective with 1.40 numerical aperture (NA). Examination was conducted by two independent observers, blind to the birds’ infection status, and their counts averaged (mean). Parasitemia was quantified by counting the number of infected erythrocytes in 100 randomly selected fields (Godfrey et al. 1987), with approximately 250 erythrocytes each, for a total of approximately 25,000 examined cells and expressed as parasitized cells per 25,000 erythrocytes.

We determined both total and differential WBC counts on each blood smear. Total WBC counts were conducted using a microscope (Olympus BH2) under 40× objective with 0.95 NA. All WBC were counted in 10 randomly selected fields where cells were evenly distributed, excluding any damaged WBC. To estimate the total WBC count per milliliter of blood, the sum of WBC counted in 10 fields was multiplied by 200 (Fudge 2000). Differential counts were conducted similarly, but using a 100× oil immersion objective. A total of 100 WBCs were identified and categorized, based on their morphological and staining characteristics, into heterophils, lymphocytes, monocytes, eosinophils, and basophils (Clark et al. 2009). The heterophil-to-lymphocyte (H:L) ratio was then calculated.

We extracted DNA from blood samples using the DNeasy Blood & Tissue Isolation Kit (QIAGEN, Germantown, Maryland, USA) and from conjunctival and choanal swabs, separately from each other for each bird, using the DNeasy 96 Blood & Tissue Kit (QIAGEN), following the manufacturer’s instructions for each procedure. The integrity of all extracted DNA was verified by visualization on 1% agarose gels with ethidium bromide staining. We measured DNA concentration and purity (A260/A280 ratio) using a NanoDrop Lite Spectrophotometer (Thermo Fisher Scientific Inc., Waltham, Massachusetts, USA).

We targeted the mitochondrial cytochrome b gene of Haemosporidia using a nested PCR protocol as described (Bensch et al. 2009) and the primers described in Bensch et al. (2000), Hellgren et al. (2004), and Waldenström et al. (2004). Separated PCR reactions were performed to detect Plasmodium-Haemoproteus spp. or Leucocytozoon spp. For Plasmodium-Haemoproteus spp., two pairs of specific primers were used: HAEMNF CATATATTAAGAGAATTATGGAG-HAEMNR2 AGAGGTGTAGCATATCTATCTAC and HAEMF ATGGTGCTTTCGATATATGCATG-HAEMR2 GCATTATCTGGATGTGATAATGGT. For Leucocytozoon spp., the primers HAEMNFI CATATATTAAGAGAAITATGGAG-HAEMNR3 ATAGAAAGATAAGAAATACCATTC and HAEMFL ATGGTGTTTTAGATACTTACATT-HAEMR2L CATTATCTGGATGAGATAATGGIGC were used. Each sample was tested by three independent PCR reactions with positive and negative controls included. The primary standard PCR included 1 μL of total DNA in a total of 10 μL volume with the outer correspondent primers (HAEMNF-HAEMNR2 or HAEMNFI-HAEMNR3). This was followed by the nested PCR with 1 μL of the primary PCR product in a 25-μL reaction using the specific primers (HAEMF-HAEMR2 or HAEMFL-HAEMR2L). Roche Taq DNA polymerase (Roche Holding AG, Basel, Switzerland) was used for all PCR reactions under cycling conditions: initial denaturation at 94 C (3 min); 20 cycles of 94 C (30 s), 50 C (30 s), and 72 C (45 s); with a final extension at 72 C (10 min). Nested PCR was conducted with the same conditions but for 35 cycles. All PCR products were confirmed by 2% agarose gel electrophoresis with ethidium bromide staining and documented using a Labnet Enduro GDS UV transilluminator (Labnet International Inc., Edison, New Jersey, USA) and its associated software (ENDURO-GDS 2013). The intensity of the bands was scored visually on a scale from 0 (no band) to 2 (dark band) as per Reinoso-Pérez et al. (2020), with the cumulative score from three replicates serving as an indicator of infection intensity. Sequencing was performed on both strands of all PCR products. Consensus amplicons were constructed by aligning the forward and reverse sequences using BioEdit 7.2 (Hall 1999) and MEGA6 (Tamura et al. 2013). Sequence identity was confirmed through a Basic Local Alignment Service Tool (BLAST) search on the MalAvi database (Bensch et al. 2009). All haemosporidian sequences obtained and the information from the avian host were deposited at Mendeley Data, V1, DOI: 10.17632/8rv4ng834j.1.

The presence of M. gallisepticum was determined using DNA extracted from swab samples. To ensure reliability, each sample was independently processed in triplicate following a PCR protocol with primers specific to the bacterial gene mgc2: mgc2-2F CGCAATTTGGTCCTAATCCCCAACA and mgc2-2R TAAACCCACCTCCAGCTTTATTTCC (García et al. 2005). Amplification products were confirmed by 2% agarose gel electrophoresis as described above. Samples testing positive for M. gallisepticum were further analyzed to quantify the M. gallisepticum load following a qPCR protocol specifically designed for this purpose (Grodio et al. 2008).

Samples were categorized into four groups based on their infection status determined by the diagnostic tests applied. The first group tested positive only for M. gallisepticum (the “M. gallisepticum” group). Based on the higher M. gallisepticum load detected in the tissue sampled, this group was subdivided into “conjunctiva” and “choana” subgroups. The second group tested positive only for Haemosporidia. Birds that tested positive for both pathogens formed the third, co-infection group. The last group was composed of birds in which neither M. gallisepticum nor Haemosporidia were detected, the pathogen-free group (although they may have been infected with other pathogens). A fifth group was considered for Haemosporidian co-infection, which would include cases of Plasmodium-Haemoproteus and Leucocytozoon; however, there were not enough cases to provide statistical support.

Statistical analyses were conducted using RStudio (RStudio Team 2022) and STATISTIX v10.0 (STATISTIX 2014). To test if groups differed in measured variables such as WBC counts, M. gallisepticum load or Haemosporidia parasitemia, an analysis of variance (ANOVA) followed by a Tukey honestly significant difference (HSD) test (for normally distributed data), a Kruskal-Wallis ANOVA (H) followed by a Dunn’s test (for nonnormally distributed data), and a Wilcoxon rank-sum test (W; when comparing two groups) were used.

Of the 246 Mexican House Finches that we sampled, 181 (73.6%) were infected with M. gallisepticum and 106 (43.1%) with Haemosporidia. Sample sizes of the four groups based on their infection status were M. gallisepticum only (n=101), Haemosporidia only (n=26), co-infected (n=80), and pathogen free (n=39),

Haemosporidian parasites were detected by PCR in 43% (106/246) Mexican wild House Finches. The majority (56%, 59/106) were each infected with one Haemoproteus lineage, including HAEMEX01, PYERY01, PYERY02, SIAMEX05, and SISKIN1; 16% (17/106) were each parasitized with one Plasmodium lineage: GRW04, MOLATE01, POOHIS04, SEIAUR01, or TROAED24. Another 16% (17/106) were infected with one Leucocytozoon lineage: CARFLA04, CB1, CNEORN01, DENCOR06, DUMCAR01, JUNPHA06, QUIQUI02, SETAUD25, SETAUD28, SETCOR12, SPIPAS07, TUMIG12, and ZOLEU02. The remaining infected birds (12%; 13/106) tested positive for both Plasmodium-Haemoproteus and Leucocytozoon spp.

Only 2/246 of the trapped House Finches presented with very mild clinical signs of conjunctivitis. Infection with M. gallisepticum was detected in 8.5% (21/246) of the birds by serological test, in 24% (59/246) of the conjunctival samples, and 72% (177/246) of the choanal samples by qPCR. In the 57 birds that tested positive both in conjunctival and choanal samples, the mean bacterial load was significantly higher in choanal samples (mean±standard error [SE]=1.53±0.13) compared to conjunctival samples (0.64±0.03; Fig. 2; W=558, P<0.01). Therefore, M. gallisepticum results presented further are based exclusively on the analysis of choanal samples, except when comparing the H:L ratio between tissue subgroups.

Haemosporidian parasitemia was significantly higher in M. gallisepticum–co-infected birds (80/246, 62±16.08 parasites per 25,000 erythrocytes) compared to those without M. gallisepticum co-infection (26/236, 21±7.43; W = 736.5, P<0.05; Fig. 3a). The M. gallisepticum load was also significantly higher in co-infected birds (80/246, 1.20±0.08 mgc2 copies) than in the M. gallisepticum group (101/246,0.99±0.07; W=2724, P<0.01; Fig.3b).

The PCV did not vary significantly across the four groups categorized by infection status (H=3.81, df=3, P=0.28; Fig.4a). Conversely, the total WBC count significantly varied between groups (ANOVA, F=3.924, df=3, P<0.01; Fig. 4b), the co-infection group having significantly higher WBC count than the pathogen-free group (Table 1). Birds with a single infection did not differ significantly from each other, nor from either of the other groups. Heterophil proportions varied significantly between groups (ANOVA: F=22.41, df=3, 242, P<0.01). The proportion in the pathogen-free group was significantly lower than in any other group, and the single-infection groups did not differ from each other. The co-infection group had a significantly higher heterophils proportion than the M. gallisepticum group, but not significantly higher than the Haemosporidia group (Fig. 5a, Table 1). Lymphocyte proportions differed significantly between the four groups (ANOVA: F=10.69, df=3, 242, P<0.01). The co-infection group had a significantly lower proportion than the pathogen-free and M. gallisepticum groups, but the proportion was not significantly lower compared to the Haemosporidia group (Fig. 5b, Table 1). The H:L ratio differed significantly among the four groups (H=23.22, df=3, P<0001). The pathogen-free group had a significantly lower H:L ratio than any other group, whereas the single-infection groups did not differ from each other. In the co-infection group, the H:L ratio was significantly higher than the M. gallisepticum group, but it did not significantly differ from the Haemosporidia group (Fig. 5c). Monocytes (H=7.24, df=3, P=0.06), eosinophils (H=2.23, df=3, P=0.52), and basophils (H=1.61, df=3, P=0.65) showed consistent levels across all four groups (Table 2).

Among the single-infected M. gallisepticum group, we subdivided the 74 House Finches that tested positive in both sampled tissues into groups based on whether the M. gallisepticum load was higher in the conjunctiva (11/73) or the choana (63/74). Birds from the conjunctiva subgroup showed a significantly higher H:L ratio than birds in the choanal subgroup (W=174, P<0.01; Fig. 6).

This study evaluated the prevalence of Haemosporidia and M. gallisepticum in wild House Finches across 20 locations in Mexico and measured the bird’s immune response using the H:L ratio. Among 246 trapped House Finches, 43% were infected by Haemosporidia. This high prevalence and the lineages detected in House Finches in Mexico are consistent with previous haemosporidian studies in Mexican avian populations (Reinoso-Pérez et al. 2016; Ham-Dueñas et al. 2017; Tinajero et al. 2019; Hernández-Lara et al. 2020; Villalva-Pasillas et al. 2020; Penha et al. 2023), and are similar to those found in the western range of House Finches in the US (Davis et al. 2013). Interestingly, some of the lineages detected in this study, especially from the genus Leucocytozoon, have previously been reported from House Finch populations in New York, US (Reinoso-Pérez et al. 2024). Conversely, the most common Plasmodium lineage infecting House Finches in New York, US (PADOM11) was not detected in the Mexican birds. The diverse range of Plasmodium, Haemoproteus, and Leucocytozoon lineages identified emphasizes the diversity of Haemosporidia infecting avian hosts in the wild. Although House Finches are not migratory, it is possible that migratory bird species could introduce haemosporidian lineages into new areas.

Although M. gallisepticum has been reported in Mexican poultry (Petrone-Garcia et al. 2020), it has not been found in Mexican wildlife. Our observation that among the 246 House Finches we trapped across Mexico 73.6% tested positive for M. gallisepticum DNA by qPCR from the choanal and/or the conjunctival swabs, a very high proportion of current infections, was therefore surprising. Our finding that the proportion of birds in which we detected M. gallisepticum DNA was much higher than the proportion of seropositive birds seems to be typical for M. gallisepticum infections in wild birds, especially in House Finches (Dhondt et al. 2014); thus serology is not a reliable method to detect M. gallisepticum infection in a population.

Among our birds, just two birds showed very mild clinical signs of conjunctivitis. The absence of clinical signs in infected birds might have been caused by other factors than an M. gallisepticum infection, although it could also be the result of resistance to reinfection with the same or a less virulent M. gallisepticum strain, as shown experimentally (Sydenstricker et al. 2005; Dhondt et al. 2017a; Fleming-Davies et al. 2018), which is probable, given the high prevalence of M. gallisepticum in Mexican House Finches. This resistance may reflect the long-term evolutionary history of the Mexican population, the original population of House Finches in North America, which may have developed adaptive immune responses over time. It could also be the result of the route of infection: Dhondt et al. (2007) showed that although all birds infected in the conjunctiva developed bilateral conjunctivitis, only 20% of birds inoculated by the oral route developed mild unilateral conjunctivitis although all seroconverted. Our finding that in 11 birds the M. gallisepticum load was higher in conjunctival than in choanal swabs suggests that some of the House Finch-adapted strains that spread across the US and Canada have also reached Mexico (see Reinoso-Pérez et al. 2022).

The high rate of Mexican House Finches co-infected with Haemosporidia and M. gallisepticum also made it possible to test both the hypothesis that birds co-infected with multiple pathogens are more stressed than birds infected with a single pathogen, and that co-infections are beneficial for the pathogens. Our finding of significantly higher levels of haemosporidian parasitemia and M. gallisepticum loads in co-infected birds, compared to those with a single infection is a similar pattern to previous observations from experimental studies on co-infection in captive House Finches and suggests a mutualistic relationship between these pathogens in House Finches, leading to increased infection intensity of both (Dhondt and Dobson 2017; Reinoso-Pérez et al. 2020). Alternatively, this pattern could reflect a competitive interaction, where M. gallisepticum and haemosporidian parasites both grow to exclude one another, resulting in elevated immune responses that track the high loads of both parasite types.

Both PCV and total WBC counts in this study were comparable to those previously reported in experimental studies of House Finches under controlled and captive conditions (Reinoso-Pérez et al. 2020), but unlike the result in earlier experiments, the PCV in this study did not differ significantly between the groups. However, the total WBC count did differ significantly between the groups, although differences were only between the pathogen-free and the co-infected birds. The observed elevated WBC count compared to their specific-pathogen-free counterparts aligns with previous studies comparing this parameter in a similar way (Davis et al. 2004; Fratto et al. 2014; Bale et al. 2020; Reinoso-Pérez et al. 2020).

In contrast with the previous (few) studies, which have not found an impact of Haemosporidia infection on leukocyte profiles (e.g., Krams et al. 2010; Cornelius et al. 2014; Granthon and Williams 2017), the differential WBC count analysis in this study revealed significant changes in specific leukocytes as we had predicted. The observed significant variation in heterophil proportions across different infection status groups suggests a robust immune response from the host. As a critical component of the innate immune system, heterophils are primarily involved in phagocytosis. They are rapidly mobilized to the infection or inflammation site, especially during bacterial infections, where they play an important role in pathogen digestion (Maxwell 1993; Harmon 1998; Campbell 2005). Our observation strongly supports the hypothesis of increased proportions of heterophils in infected birds. This also corresponds with similar heterophil increases observed in both captive (Bale et al. 2020) and wild-caught House Finches (Davis et al. 2004) infected with M. gallisepticum.

Similarly, lymphocytes have a variety of immunological functions, and they are an important component of the adaptive immune system. They not only proliferate in response to infections but also play a crucial role in producing the immunoglobulins and modulating immune defense (Campbell 2005; Davis et al. 2008). Although a decrease in lymphocyte proportion was not observed in free-living House Finches infected with M. gallisepticum by Davis et al. (2004), our observation of lymphopenia is consistent with reports made by Fratto et al. (2014) and Bale et al. (2020) in their respective studies. Previous research in poultry indicates that lymphopenia can result from an increase in stress hormones such as corticosterone (Gross and Siegel 1983), a phenomenon also shown in House Finches infected with M. gallisepticum (Love et al. 2016). Thus, the observed decreased lymphocyte proportions we observed might be explained as a glucocorticoid-induced redistribution of these leukocytes (Dhabhar 2002).

The significant differences in the H:L ratio observed across infection status groups support our initial prediction. Birds with either single or double infection showed a higher H:L ratio compared to birds free of both infections, reflecting the leukocyte changes observed across different groups. This observation is consistent with the elevated ratios previously reported in infected House Finches (Davis et al. 2004; Bale et al. 2020). In all cases, the H:L ratio increased in infected birds. Contrary to observations reported by Dunn et al. (2013) studying birds infected by the Haemosporidia genus Haemoproteus, we did not observe a reduced H:L in infected birds, even when comparing birds infected only with Haemosporidia to pathogen-free birds. Although there was no significant difference in the response to single infections, it is worth noting that co-infected birds showed significantly higher H:L ratios compared to M. gallisepticum group but not to the Haemosporidia group. This suggests a stronger immune response in birds dealing with multiple pathogens, supporting our hypothesis.

When comparing H:L ratios of M. gallisepticum–infected birds on tissue-specific bacterial load, birds in the subgroup conjunctiva had a higher H:L ratio than in the choana subgroup. This supports the hypothesis of a differential response by tissue based on the M. gallisepticum variant infecting a bird. Although both variants of M. gallisepticum commonly found in wildlife, the poultry-adapted and the House Finch-adapted strains are pathogenic, the latter has evolved to increase virulence for the House Finches, with infected birds showing a higher M. gallisepticum load in the conjunctiva than in the choana (Ley et al. 1996; Dhondt et al. 2007a; Hawley et al. 2013; Reinoso-Pérez et al. 2022). Based on previous reports (Bale et al. 2020; Reinoso-Pérez et al. 2022), our observations suggest that birds with higher M. gallisepticum load in the conjunctiva, and higher H:L ratio, are probably infected with a House Finch–adapted strain. In contrast, birds with a higher load in the choana may be infected with a less virulent poultry strain. Further research is needed to determine the specific variant of M. gallisepticum circulating among Mexican House Finches conclusively.

Our study not only confirms previous experimental findings but also significantly extends our understanding by providing field-based evidence of the hematological impacts of both single and co-infections in wild House Finches. It highlights the importance of considering multiple pathogens and their interactions in wildlife when assessing wildlife health, thereby offering valuable perspectives for future research on avian disease ecology. Furthermore, this represents the first published report of M. gallisepticum infecting House Finches in Mexico, following the initial outbreak in the US approximately 30 yr ago.

All applicable institutional and national guidelines for the care and use of animals were followed. We express our sincere gratitude to Carlos Morales Mariñelarena, Celina González Gallegos, Rosa María Martínez García, Héctor Gerardo Pérez González, Diego Gerardo Pérez Martínez, Eduardo Rafael Pérez Martínez, José Antonio Reinoso Pérez, Rodrigo Reinoso Pérez, Roberto Gerardo Pérez Delgado, Angélica Tello Losa, Catalina María Vélez Argumedo, Andrés Felipe Arango Gómez, Gabriela Pérez Florencio, Alfredo Ramírez Bouchan, Jessica Gretel Loza León, Osvel Hinojosa Huerta; from Pronatura Noreste A.C: Alejandra Calvo Fonseca, Juan Butrón Méndez, José Juan Butrón Méndez, Juan Ángel Butrón, Benito Brambila, Carlos Medina, Stefany S. Villagómez Palma; from Club de Conservadores y observadores de aves de Durango: Jesús Favela Mesta, Olivia Rojas Flores; from Bosques Urbanos de Guadalajara: Cristian Alejandro de la Torre Gutiérrez, Romario Palacios Aguilar; from Universidad Autónoma del Estado de Hidalgo: Mariel de Jesús Rufino Flores, Guillermo Vargas Noguez; from Benemérita Universidad Autónoma De Puebla: Salvador Juan Loranca Bravo, José Antonio González Oreja. We thank Instituto de Investigación de Zonas Desérticas (IIZD, UASLP), and Coordinación para la Innovación y Aplicación de la Ciencia y la Tecnología (CIACYT, UASLP) for their invaluable assistance in the sample collection in Mexico. Their hospitality, support, expertise, and dedication were crucial to the successful completion of this project. M.T.R.P. thanks CONACYT for the grant scholarship that enabled the start of this project. We also gratefully acknowledge funding from the Athena Fund of the Cornell Lab of Ornithology, which provided essential support to complete this research. The authors declare that they have no conflict of interest. Data for this manuscript are freely available at Mendeley Data, V1, DOI: 10.17632/8rv4ng834j.1.