Prescribed Fire Effects on Hummingbird Taxonomic and Functional Diversity in Pine–Oak Forests in West-Central Mexico

Simple Summary

Pine–oak forests in Mexico experience frequent wildfires, making fire management a key conservation strategy. Prescribed fire is widely used to prevent high-intensity wildfires, regulate vegetation structure, and create habitats for wildlife adapted to post-fire conditions. This study evaluated the impact of low-severity prescribed fire on hummingbird diversity in the Sierra de Manantlán Biosphere Reserve. We conducted surveys in two recently low-severity prescribed fire sites and one site where wildfire had been suppressed for over 40 years. Our results revealed significant differences in the composition of abundant and very abundant hummingbird species at sites characterized by low-severity prescribed fire and concave summit topography. However, species richness did not align with functional richness, as the site with long-term fire suppression exhibited the highest functional diversity. Functional diversity refers to the range of ecological roles and traits that species have within an ecosystem, such as differences in feeding behavior or resource use. These findings suggest that fire history, topography, vegetation structure, and floral resource availability shape hummingbird communities. We recommend maintaining a heterogeneous forest matrix, including fire-suppressed patches, and areas with prescribed fires of varying severity. This approach can enhance taxonomic and functional biodiversity, ensuring the availability of diverse habitats for hummingbird species.

Abstract

Prescribed fires are a management strategy involving the controlled application of fire to achieve specific ecological objectives. In the pine–oak forests in west-central Mexico, we conducted an experimental low-severity prescribed fire to assess its effects on hummingbird diversity. We hypothesized that low-severity prescribed fire would enhance both taxonomic and functional diversity by modifying understory vegetation structure and increasing floral resource availability. To test this, we performed point count censuses in both low-severity prescribed fire and fire-suppressed sites where wildfire had been excluded for over 40 years. Taxonomic diversity was assessed using Hill numbers to estimate true diversity across different abundance weights, while functional diversity was evaluated through indices such as functional richness, functional evenness, and functional divergence. Our results indicated that low-severity prescribed fires did not affect overall hummingbird diversity as both low-severity prescribed fire sites and fire-suppressed sites exhibited comparable species richness. However, sites with low-severity prescribed fire and concave summits showed a significantly higher abundance of common and highly abundant species. Notably, species richness did not align with functional richness, as the fire-suppressed site exhibited the highest functional diversity. These findings suggest that hummingbird community structure is influenced by a combination of fire history, topography, vegetation structure, and floral resource availability. We recommend maintaining a heterogeneous forest matrix, incorporating patches with fire suppression, and areas subjected to prescribed fires of varying severity. This multifaceted approach enhances both taxonomic and functional biodiversity, promoting habitat heterogeneity and ensuring the persistence of diverse hummingbird assemblages in fire-prone ecosystems.

1. Introduction

Wildfires in temperate forests play a fundamental role in shaping avian communities by altering vegetation structure, floristic composition, and habitat availability [1,2,3]. In the boreal forest biome, several studies have reported strong impacts of fire on bird diversity and species composition, particularly due to changes in understory structure and successional dynamics [4,5]. High-severity fires tend to reduce specialist species while favoring generalists and early colonizers [6,7]. However, responses can be variable, and some fire-adapted ecosystems show resilience, maintaining stable avian diversity post-disturbance [8]. These findings highlight the need to understand fire effects across biomes to inform adaptive fire management strategies under increasing climate-driven fire regimes.

Pine–oak forests in Mexico experience frequent surface fires that occur every 5 to 35 years [9,10]. Occasionally, high-severity fires create openings by burning all biomass in a torch-like manner. However, most fires in these forests are of low to moderate severity, with plant mortality primarily affecting the understory [11]. Canopy reduction is generally less than 25%, though small gaps can form due to the death of clusters of contiguous trees [12]. The vegetation’s response to fire is predominantly positive, as these forests are fire-resistant and have a strong capacity for regeneration [11]. Fire incidence is higher in convex landforms and at elevated topographic positions [12].

Wildfires are becoming more frequent and severe compared to previous decades due to changes in land use, global temperature, and precipitation patterns. The compound effect of these factors is impacting ecosystems and biodiversity [13,14]. According to the National Forestry Commission, there were 50,317 wildfires in west-central Mexico between 1991 and March 2024 [15]. Additionally, towards the southern region of the country, forest fires have increased by 42% compared to the year 2000, resulting in a larger affected area [16]. Therefore, it is essential to implement management actions to mitigate the impact of forest fires effectively [17].

Prescribed fires are crucial for decreasing the intensity of wildfires. They accomplish this by lowering the amount of available fuel, which hinders the rapid spread and uncontrollability of wildfires [18,19]. Prescribed fire is a habitat management strategy that involves applying fire in a controlled manner to achieve specific goals. Their objectives include maintaining or restoring the historical fire regime, controlling the structure and composition of vegetation, preventing or reducing the environmental impacts of severe wildfires, and generating habitats for certain species [20].

Prescribed fires are part of the 2020–2024 Fire Management Program of the National Forestry Commission of Mexico, promoting their use in a fire-dependent ecosystem [21]. The fires should mimic the characteristics and patterns of natural fires, including intensity (amount of energy released), severity (impact on the ecosystem), frequency (fire return interval), seasonality (in which seasons it occurs if the disturbance is recurrent), and the size area affected [22].

Prescribed fires affect plant and animal communities differently because they modify habitats, enhancing conditions for some species while degrading them for others [23,24]. For instance, in west-central Mexico, hummingbirds (Trochilidae family) exhibit varied responses to fires in the temperate forest (pine, pine–oak, and cloud forest). Some hummingbird species suffer negative impacts after a fire, resulting in decreased abundance. Others remain unaffected, while some thrive in post-fire areas, exhibiting increased abundance [25,26,27]. Investigating the impact of fire on hummingbird populations is crucial for their conservation [20]. Hummingbirds play a significant role in ecosystems as pollinators, promoting the growth of numerous plant species [28,29,30]. They also act as predators of invertebrates and serve as prey for other terrestrial vertebrates [31,32]. Changes in hummingbird diversity can affect the ecological functions they perform.

Examining the different aspects of biodiversity, including taxonomic and functional diversity, in terms of alpha (local diversity) and beta (turnover diversity) components, helps us understand the complexity of local communities. This analysis also allows us to assess how organisms respond to spatial gradients and their patterns at the regional level [33,34]. The traditional measures of alpha and beta diversity assume all species have equal importance [35,36] and, thus, fail to consider the function of the different species [37,38]. Information about functional diversity will help to understand the mechanisms that shape communities and which functional groups influence ecosystem functioning [29,39]. However, most ecological studies tend to focus on taxonomic diversity, ignoring functional diversity, even though the latter influences ecosystem services [39,40]. Anthropogenic disturbances may negatively impact diversity because species maintain homogenized traits to adapt to the disturbances [41,42,43].

We evaluated hummingbirds’ taxonomic and functional diversity in two perturbed pine–oak forests with prescribed fire and one pine–oak forest with fire suppression in the Sierra de Manantlán Biosphere Reserve (SMBR). Low-severity prescribed fire might be similar to a low-severity wildfire. After this last type of fire, the vegetation structure in terms of vegetation strata is expected to remain the same after a low severity fire because the ecosystem is fire resistant, but we expected an increase in herbaceous cover and the availability of flowering plants used by hummingbirds. The SMBR pine–oak forests have a high-frequency, low-severity fire regime [44], which makes this ecosystem fire-resistant. Still, there will be an increase in herbaceous cover and the availability of plants with flowers used by hummingbirds. We hypothesized that low-severity prescribed fire would increase taxonomic and functional diversity of hummingbirds because it would change the vegetation structure at the understory level, providing a higher resource availability. This study will provide management recommendations to aid in the conservation of hummingbirds.

2. Materials and Methods

2.1. Study Area

The SMBR is located in west-central Mexico, with 90% of its area in Jalisco state and the remaining 10% in Colima state. It lies in the northern portion of the Sierra Madre del Sur, a mountain range in southern Mexico. SMBR borders the Trans-Mexican Volcanic Belt—a volcanic physiographic province—to the east and the Sierra Madre Occidental—a vast mountain range—to the north. The reserve covers a total area of 139,577 hectares, featuring complex and rugged terrain that ranges from 400 to 2860 m above sea level [45]. The climate is warm and temperate subhumid, with an average of 16 to 22 °C and annual precipitation of 600–1700 mm [46]. There is a marked rainy season from June to September with occasional rainfall in winter and a dry season from February to May.

We selected three similar sites regarding the type of vegetation (pine–oak forest), the size (45–55 ha approximately each one), and elevation between 1700 and 2200 masl. They were located in the northwestern part of the SMBR (Figure 1A–C). The first one, San Campus (SC-sf), where the vegetation is mature pine–oak forest, was dominated in the arboreal strata by douglas pine (Pinus douglasiana), herrera pine (Pinus herrerae), white oak (Quercus scytophylla), yellow oak (Quercus glaucescens), powder puff (Calliandra longipedicellata), texas madrone (Arbutus xalapensis), cucharo (Clethra fragans), and Mexican yellow pine (Pinus oocarpa) [46]. The terrain consists of low, convex slopes with moderate to steep inclines [47]. The dominant herbaceous species are the lady’s eardrops (Fuchsia cylindracea), and purple sage (Salvia iodantha). The last recorded wildfire was in 1983, with a high-severity effect that burned the treetops in a torch-like manner, lasting 2–3 days (Jardel, pers. communication). Since then, the area has been under a fire-suppression regime (Figure 1D).

The second site, Cuartones (CT-pf), is a mature pine forest. The dominant species are douglas pine, flagpole oak (Quercus calophylla), and Mexican white oak (Quercus obtusata). The dominant herbaceous species are lavender-leaf sage (Salvia lavanduloides) and purple sage. The terrain features low, convex slopes with moderate to steep gradients [45]. The last wildfire was in 1994, with a 2-day duration, a low-severity effect that did not create clearings, and a low severity prescribed fire, which was in 2020, with a 2-day duration (Figure 1D).

The last site, Mantequillas (MT-pf), is a mature pine forest vegetation. The dominant species are douglas pine, Mexican yellow pine, flagpole oak, and Mexican white oak. The dominant herbaceous species is Mexican sage (Salvia mexicana). The terrain is characterized by concave summits and moderate to steep slopes [47]. The last recorded wildfire was in 2008, with 2-day duration, a high-severity effect mainly in canyons, and a low-severity prescribed fire, which was carried out in 2020, with a 2-day duration (Figure 1D).

2.2. Data Collection Methodology

From October 2020 to March 2021, we conducted monthly surveys to study migratory and resident hummingbirds, as this period coincides with peak blooming in the area. We established eight-point counts (circular) at the fire-suppressed site (SC-sf), and ten-point counts at each experimental low-severity prescribed fire site (CT-pf and MT-pf). These point counts were established at each site following the recommendations by Hutto et al. [50]. The point count survey order varied each month to reduce potential biases. Each circular plot had a 25 m radius, established with a distance-meter (BOBLOV 6 × 42). We followed standard bird census protocols [51,52], to ensure habitat representativeness and independence between samples. A single trained observer conducted all surveys, standing at the center of each plot for ten minutes and recording all the hummingbirds seen or heard within the fixed radius. To reduce potential biases, the order of survey points varied each month, and observations were conducted from 30 min after sunrise until 11:00 a.m.

Vegetation structure was assessed using the Relevé method [53] within circular plots of a 25 m radius in each stratum corresponding to the bird sampling plots. Within these plots, we identified three distinct vegetation strata: the tree stratum (vegetation above 5 m), the shrub stratum (vegetation between 1 and 5 m), and the herbaceous stratum (vegetation below 1 m). For each stratum, we estimated the cover of dominant species to characterize vegetation composition and structure. Cover estimation followed the standard Relevé procedure [53], based on visual assessment of the vertical projection of vegetation onto the ground surface. A single trained observer conducted all estimates from the center of each plot to ensure consistency and minimize observer bias.

To evaluate habitat conditions, we recorded several environmental variables, including average temperature and average relative humidity, using a handheld Kestrel 3000 weather meter. The variables measurements were taken at each sampling plot at ~1.5 m height, in shaded conditions, at the start of each bird survey to avoid bias from direct sunlight. To assess floral resource availability, we quantified the total number of flowers and recorded the abundance of key plant species, specifically Mexican sage, lavender-leaf sage, purple sage, Salvia thyrsiflora (no known common name), and lady’s eardrops (Supplementary Material, Table S1).

Flower abundance was evaluated using Gentry transects [54], consisting of 20 m × 5 m plots. Flowering plants were counted within a 2.5 m strip on each side of the 20 m transect. We identified the species present and counted their total number of flowers to assess the floral resources available for hummingbirds.

2.3. Statistical Analysis

2.3.1. Taxonomic Diversity

Hummingbird taxonomic and functional diversity was estimated by site (spatial approach) with a one-way mensurative design. The site was a fixed factor with three levels (SC-sf, CT-pf, and MT-pf). We considered each point count per month per site as an independent replicate. We calculated the Hill numbers from the pool (sum) of all the point counts per site using iNEXT Online 4 steps [55,56]. The diversity of q orders (Hill numbers) represents the effective number of species. We considered q = 0 to be the species richness; q = 1, all species included with a weight proportional to their abundance in the community, i.e., the abundant species; and q = 2, the very abundant species [55,57,58]. The approach based on the mentioned framework is a procedure that consists of four steps: (1) The evaluation of the profile of sample completeness, which assesses what percentage of the community species have been recorded. (2) An analysis of rarefaction and extrapolation based on the size and profile of asymptotic diversity for 0 ≤ q ≤ 2. This step helps to identify how many species in the community have yet to be recorded. (3) An extrapolation and rarefaction analysis based on non-asymptotic coverage for the orders q = 0, 1, and 2. This allows us to compare the confidence intervals of each sample (based on sites) to determine if there are any significant differences in q0, q1, or q2. And, finally, (4) the evenness profile. This analysis was performed with 1000 iterations and a 95% confidence interval.

2.3.2. Functional Diversity

We selected six functional traits relevant to understanding the contribution of hummingbirds to ecosystem functioning (Supplementary Material, Table S2). The functional traits selected were bill length and curvature, wing and tail length, weight, and migratory status. The information on functional traits data (except bill curvature) was provided by Contreras-Martínez (2010–2023) from the hummingbird database of the Department of Ecology and Natural Resources-IMECBIO, Universidad de Guadalajara, Centro Universitario de la Costa Sur. To obtain data on bill curvature, we conducted field surveys within the SMBR from October 2022 to March 2023. During these surveys, we took photographs of the bills of hummingbirds and used the ImageJ Software v. 1.54h [59] to measure the angle between the exposed culmen and the tip of the bill [29].

To estimate functional diversity, we first calculated the three components proposed by Villéger [60]. The functional richness (FRic index) measures the amount of functional space occupied by the species in a community, regardless of their abundance. The functional evenness (FEve index) measures the homogeneity of the distribution of species abundances in a community’s functional space. Finally, the functional divergence (FDiv index) is a measure of functional similarity between the dominant species of a community [61]. The 95% confidence intervals were used to evaluate the differences in functional indices by analyzing the interval overlap. This last measurement was estimated using null models. These null models were created by randomizing the community matrix 1000 times using a swapping algorithm to generate null distributions for each functional diversity (FD) index. This method uses the observed matrix as the starting point, preserving the total species richness at each site and the number of sites where each species can be found [62]. We estimated the functional indices using species composition, abundances, and functional traits using the FD package v. 1.0-12.3 [61,63] and null models using Picante v. 1.8.2 [64] in R-project, version 4.2.2.

To identify functional groups of hummingbirds, we used the functional traits matrix to perform cluster analysis using the Gower distance [65,66]. The clustering analysis was performed using the group average linkage method and Whittaker’s coefficient of association. Additionally, we did a Principal Coordinates Analysis (PCO) to visualize the ordination of hummingbird functional groups and functional traits. The PCO was built with the same cluster analysis resemblance coefficient, complemented with a multiple correlation analysis that shows the functional traits as vectors.

Finally, considering their relative abundance, a shade diagram was created to visualize the species’ contribution to each site. The functional groups cluster were coupled in this diagram. The clustering analysis, PCO, and a shade diagram were conducted in the program PRIMER v7 with PERMANOVA+ software from PRIMER-e [67].

2.3.3. Hummingbird Diversity and Site Characteristics

We performed a PCO to visualize the relationship between hummingbirds and habitats. We used the species-abundance per site matrix and the Bray and Curtis resemblance matrix. After, we performed a multiple correlation analysis, showing the species associated with the habitats as vectors.

For the relationship of the sites with the environmental variables, we performed a PCA. We considered a matrix with 16 environmental variables relevant to hummingbirds (Supplementary Material, Table S1). Before performing the ordination, we did a multicollinearity test using Pearson correlation, eliminating variables with r ≥ 0.9 and reducing the matrix to four variables: time since last fire, temperature, abundance of lavender-leaved sage, and abundance of lady’s eardrops.

3. Results

3.1. Taxonomic Diversity

We registered five genera and eight species: Rivoli’s Hummingbird (Eugenes fulgens), Amethyst-throated Mountain-gem (Lampornis amethystinus), Calliope Hummingbird (Selasphorus calliope), Rufous Hummingbird (Selasphorus. rufus), Broad-tailed Hummingbird (Selasphorus. platycercus), Bumblebee Hummingbird (Selasphorus heloisa), White-eared Hummingbird (Basilinna leucotis), and Berylline Hummingbird (Saucerottia beryllina). Five species are migratory, two are residents, and one has both resident and migratory populations. All species have some endemic status; one species is considered near threatened (Supplementary Material, Table S3). The estimated sample completeness for all the sites was between 67% and 100%. The sample coverage of the abundant and very abundant species detected in the sites was from 96% to 100% and 100%, respectively (Supplementary Material, Table S4). The undetected species richness within the sites varied from less than one (CT-pf, 0.25 and MT-pf, 0 species, respectively) to three (SC-sf-, 2.96 species). Fewer than one species remained undetected across all sites for abundant species (q = 1) and very abundant species (q = 2) (Supplementary Material, Table S4). Non-asymptotic analysis for the diversity of order q = 0 showed no significant difference among sites (Figure 2A). Regarding the diversity of order q = 1 and the diversity of q = 2 analysis, CT-pf and SC-sf had fewer species compared to MT (Figure 2B,C). We used a coverage value of 98.6% to compare the sites. Regarding species evenness, SC-sf showed the lower Pielou’s evenness (0.57), and MT-pf the highest value (0.86). The same pattern was shown for the evenness of abundant (q1) and very abundant species (q2) (Supplementary Material, Table S4).

3.2. Functional Diversity

Functional Diversity Indices showed that SC-sf had the highest FRic value (1.35), an intermedium value of FEve (0.387), and a low value of FDiv (0.599) compared with the other sites. In contrast, CT-pf showed low values of FRic (0.40) and FEve (0.326) but a higher value of FDiv (0.882). Moreover, MT-pf had intermedium values of FRic (1.14), the highest FEve value (0.741) and intermedium FDiv values (0.611) (Figure 2D–F, Supplementary Material, Table S4). The PCO ordination, coupled with cluster analysis, explained 96.3% of the total variation in the first two axes, generating four functional groups (Figure 3A). Group A (blue square) was associated with big size and resident hummingbirds (Rivoli’s Hummingbird, and Amethyst-throated Mountain-gem), group B (orange diamond) with small migratory hummingbirds (Rufous Hummingbird, Broad-tailed Hummingbird, and Calliope Hummingbird), group C (grey triangle) with very small resident species (Bumblebee Hummingbird), and group D (blue triangle) with small-medium resident hummingbirds (White-eared Hummingbird, Berylline Hummingbird; Figure 3A).

All functional groups were present in all sites except for group D with Bumblebee Hummingbird, which was only present in SC-sf. The groups “c” (one species), “a”, and “d” (two species, respectively) had the lowest number of species. In contrast, group “b” had the highest number of species, which implies greater functional redundancy, compared to other groups (Figure 4).

The shade plot showed that the Broad-tailed Hummingbird and Amethyst-throated Mountain-gem were the most abundant species regardless of the site. The dominant species in the SC-sf was the Broad-tailed Hummingbird. The dominant species in SC-sf was the same species but they were reverse in their abundances (Amethyst-throated Mountain-gem was more abundant). In the MT-pf, the Berylline Hummingbird, Amethyst-throated Mountain-gem, and Broad-tailed Hummingbird were the most abundant species. Concerning the hummingbird species’ rarity, Bumblebee Hummingbird (1 ind.) was only present in SC-sf, and Calliope Hummingbird (2 ind.) in SC-sf and MT-pf (Figure 4).

3.3. Hummingbird Diversity and Site Characteristics

Habitat analysis showed that hummingbird diversity was mainly explained by the time since the last fire, temperature, and lady’s eardrops, which explained 98% of the total variation in the first two axes (Figure 3C). The ordination indicated that lady’s eardrops was the strongest predictor for SC-sf, with the Broad-tailed Hummingbird shown to be associated with it (Figure 3B,C). In contrast, MT-pf was related to higher temperature and the Berylline Hummingbird. Nevertheless, CT-pf was the only site associated with lavender-leaf sage and the Amethyst-throated Mountain-gem (Figure 3B,C).

4. Discussion

We found eight of the ten hummingbird species reported in pine–oak at SMBR [25]. Two species were absent from our survey. The Allen’s Hummingbird (Selasphorus sasin) is an extremely rare species that has not been recorded in our long-term hummingbird monitoring project with mist nets over the past ten years. The Mexican Violetear (Colibri thalassinus) has only been recorded in SMBR at locations affected by high-severity wildfires [25].

4.1. Taxonomic Diversity

We found no significant difference in the species richness (q0) among the sites. However, the abundant species (q1) and very abundant species are significantly more numerous in MT-pf. It has been reported that sites with concave topography and wildfire present higher floral resources, which is the case of MT-pf [25,27,47,68,69]. Therefore, the higher availability of floral resources supports a higher abundance of hummingbirds. Lara [70] reported that availability and quality of resources can explain the temporal and spatial composition of the hummingbird community on a local scale.

We registered similar values for hummingbird species richness (eight species) but different abundances (237 individuals) to other temperate forests. For example, Lara [70] studied the community of hummingbirds in the pine, oak, and fir forest in the National Park “La Malinche”, in the Tlaxcala state. They observed eight species of hummingbirds but 765 individuals. The abundant species were the Ruby-throated Hummingbird (Archilochus colubris) and White-eared Hummingbird, and the very abundant Green Violetear. This difference in abundance might be attributed to the fact that this author surveyed more than one temperate type (oak and fir forest). Contreras-Martínez [25] cites 10 hummingbird species in pine forests in western Mexico, with seasonal variations in their abundances. The most abundant species throughout the year were the Amethyst-throated Hummingbird and the White-eared Hummingbird. On the other hand, species richness varies depending on the availability and quality of resources, as these factors directly influence habitat suitability and species interactions, shaping plant–hummingbird networks across elevations and seasons. Greater floral resource availability can support a higher number of species, while resource quality influences species composition and functional diversity [71].

In this study, we hypothesized that low-severity prescribed fire would increase taxonomic and functional diversity of hummingbirds in the sites CT-pf and MT-pf. We expected this because the fire changed the vegetation structure at the understory level, providing higher resource availability.

Although, in this study, there was not a match between taxonomic and functional diversity, given there was no difference in the species richness among sites, but higher functional richness was recorded in SC-sf. This is important for the conservation of these birds. It was very important to maintain the landscape forest patches for a long time since the last wildfire (35 years), as well as the patches with low-intensity prescribed fire. This is because hummingbird identity and functional groups were different.

4.2. Functional Diversity

Functional diversity indices indicate that SC-sf has the highest functional richness, while CT-pf has the lowest. This means that SC-sf has the greatest variety of functional traits, occupying a larger functional space and supporting a more diverse set of functional roles within the ecosystem, potentially increasing its resilience to environmental change [60,72]. This can be explained by the presence of the Bumblebee Hummingbird at this site (Figure 4), as its small body weight represents an extreme trait that expands the functional space of SC-sf. This may indicate greater diversity in ecological strategies. The absence of the Calliope Hummingbird in CT-pf (Figure 4) and its low functional richness may indicate the importance of migratory status and body weight in the functional matrix, a finding confirmed by the functional groups generated in this study.

Regarding functional evenness, MT-pf had the highest value, and SC-sf intermediate values and CT-pf the lowest value. Considering that functional evenness reflects the homogeneity in the distribution of species abundances across functional space [60,72], this means that the abundance of species is similar across different functional groups. This is evident in the relatively similar abundance distribution across functional groups at this site. On the other hand, in SC-sf and CT-pf, there is a disparity in abundance across different functional groups. This suggests that some resources in SC-sf and CT-pf are underutilized [72].

Functional divergence, which is the functional similarity measure between the dominant species, was the highest and most significant in CT-pf. Taking into consideration that CT-pf also had the lowest functional richness, this can mean that the species in CT-pf occupy a narrower functional space but there is a higher degree of functional specialization among the dominant species. In SC-sf and MT-pf, similar functional divergence indicates that hummingbirds behave as generalists, with low resource competition [73,74]. This last observation aligns with the clade composition of the hummingbird community we observed, which includes emeralds, bees, and mountain gems, all of which are known to be generalists [75]. We found four functional groups (Figure 3A): (1) large and resident hummingbirds, (2) small migratory hummingbirds, (3) the smallest resident hummingbird, and (4) medium resident hummingbirds. The main traits that separated functional groups were weight and migratory status, which are known to be important traits for hummingbirds [30,74,76]. Bill morphology is also known to be important, but in this case, our hummingbird community consists of species with medium-large straight bills with low variation between species [77,78,79]. Functional groups were composed of hummingbird species with different responses to fire (Supplementary Material, Table S3). We recommend maintaining a heterogeneous landscape with patches that vary in time since the last wildfire, as well as areas subjected to low- and high-intensity prescribed fires.

4.3. Hummingbird Diversity and Site Characteristics

Prescribed fires have been shown to increase the abundance of Salvia flowers [68], a key food resource for hummingbirds in the SMBR [80]. Our results suggest that low-severity prescribed fire enhances the abundance and very abundant hummingbird species; however, its effects vary depending on topography, which explains the differences observed between MT-pf (concave) and CT-pf (convex). This variation is likely driven by environmental factors such as humidity, light availability, and temperature, which influence the establishment of Salvia species and other ornithophilous plants [25,80].

4.3.1. Immediate vs. Delayed Response to Fire

Some studies suggest that bird populations may exhibit a time lag in their response to habitat changes [75,81]. However, there is evidence that hummingbirds can respond to fire within a short period (less than two months post-fire) [82]. Additionally, Contreras-Martínez [25] observed that hummingbirds in SMBR exhibit an immediate response to fire due to the increased abundance of flowers of the genus Salvia flowers. Therefore, a delayed response may not be the primary explanation. Furthermore, prescribed fires are typically of lower intensity than wildfires and may not significantly open the forest canopy, potentially influencing the observed patterns [83].

4.3.2. The Role of Open Habitats in Hummingbird Preferences

Hummingbirds might prefer more open areas for foraging and maneuvering [71]. Saab [81] found the same for other groups of birds that benefit from burned and open areas. This author suggested that logging could be a way to create such a habitat. In this scenario, the beneficial effects of fire for hummingbird species can likely be detected only with the combination of the generation of open areas and the increase in flowers that result from more severe wildfires.

The history of wildfires may be influencing these results. The time since the last wildfire has been a crucial factor because it triggers a series of ecological succession processes in pine–oak forests. These processes vary depending on the fire’s severity, frequency, and extent. Differences in bird community composition are more related to the time and severity of the last wildfire than to the presence or absence of prescribed fire. This finding aligns with previous research on bird community responses to fires in North America and Australia [18,75,77], as well as in the SMBR, where areas with the most recent wildfires have higher hummingbird diversity [25]. These results suggest that long-term fire effects on habitat structure play a crucial role in shaping bird community composition [4,23,27].

An advanced successional state is reflected in vegetation structure, which favors species that prefer mature pine–oak forests of SMBR, such as the Amethyst-throated Mountain-gem and Broad-tailed Hummingbird, both of which experience population declines after a wildfire [25]. The Amethyst-throated Mountain-gem reaches its highest abundance in CT-pf, preferring habitats with high arboreal cover [25]. This species also tends to defend monospecific flower patches [83]. In this context, it primarily feeds on and defends the high abundance of lavender-leaf sage, a small herb that proliferates following disturbances such as fire [84]. However, the Broad-tailed Hummingbird abundance was higher in SC-sf, where the abundance of lady’s eardrops was also greater.

Our findings contrast with those of [80,84], which reported that the Broad-tailed Hummingbird feeds more on species of Salvia and less on lady’s eardrops, while the Amethyst-throated Mountain-gem feeds primarily on Fuchsia rather than Salvia [84]. However, our results align with a pollen–hummingbird interaction study in SMBR, which found that the Amethyst-throated Mountain-gem frequently visits the Salvia species [80].

This apparent discrepancy may be explained by differences in floral availability at each study site. In our sites, lady’s eardrops were highly abundant and widely available, providing a key nectar resource, particularly for the Broad-tailed Hummingbird. Plant species exhibit phenological variations influenced by local site conditions, meaning that hummingbirds do not necessarily prefer lady’s eardrops over Salvia species, but rather feed on the most available floral resource at a given time. In this case, the dominance of lady’s eardrops in our sites is likely shaped by the observed feeding patterns. These findings highlight the importance of considering habitat-specific resource availability when interpreting hummingbird foraging behavior and plant–hummingbird interactions.

4.4. Study Limitations

Although this study provides valuable insights into the effects of low-severity prescribed fire on hummingbird diversity, it also presents some limitations. First, the spatial scale of our study was limited to three sites within a single biosphere reserve, which may constrain the generalizability of our findings to other pine–oak forests with different fire histories or floristic compositions. Second, hummingbird observations were conducted during specific morning hours and across a single flowering season. This fact may not fully capture daily or seasonal variations in species behavior and floral resource use. Despite these limitations, the study contributes to the growing understanding of how fire management influences hummingbird communities.

5. Management Implications

5.1. Integrating Fire Management and Hummingbird Conservation

To conserve hummingbird populations within a fire regime management framework, particularly in the context of climate change, it is necessary to implement adaptive strategies that integrate prescribed fire application, habitat restoration, and ongoing ecological monitoring. The presence of all functional groups in both low-severity prescribed fire and fire suppression sites underscores the significance of sustaining a fire regime that fosters habitat heterogeneity while preserving essential resources.

5.2. Key Environmental Factors Influencing Hummingbird Diversity

Hummingbird diversity is significantly affected by the time elapsed since the last fire, temperature, and the availability of lavender-leaf sage and lady’s eardrops. Therefore, fire management strategies should aim to maintain floral resources year-round, particularly considering how climate change is altering precipitation patterns and plant phenology. Rising temperatures and shifting climatic conditions may also cause shifts in hummingbird populations, necessitating conservation approaches that enhance habitat connectivity and ensure resource availability across the landscape. Larger resident species, such as the Amethyst-throated Mountain-gem, may be more vulnerable to habitat degradation caused by changes in fire regimes and resource availability.

5.3. Site-Specific Fire Management Recommendations

Site-specific fire management strategies should be implemented to mitigate these threats. In San Campus (SC-sf), where the Broad-tailed Hummingbird resides, a rotational prescribed fire strategy is recommended to sustain a diverse flowering plant community and facilitate sufficient post-fire recovery. In Cuartones (CT-pf), where the Amethyst-throated Mountain-gem is most prevalent, low-intensity prescribed burns should be implemented to ensure the persistence of lavender-leaf sage while preserving fire-free refugia for nesting and foraging. In Mantequillas (MT-pf), where the Berylline Hummingbird is predominant, fire management should prioritize the preservation of nectar-rich understory vegetation and the maintenance of unburned forest areas for shelter and nesting purposes.

5.4. Long-Term Conservation Strategies

To mitigate the impacts of fire and climate change on hummingbird populations, conservation efforts should focus on (1) post-fire restoration of essential nectar plants, (2) enhancing habitat connectivity to facilitate species movement, (3) implementing long-term monitoring programs to assess population changes in response to fire and climate alterations, and finally (4) incorporating climate projections into fire management planning to anticipate environmental changes and adapt conservation strategies proactively.