3.3.1 Sampling method

Regarding the sampling method, samplers can be divided into active and passive. Active, or volumetric samplers, are devices that operate by drawing in or aspirating air through a controlled airflow. In these, the sample can be collected in different ways: (1) through multi-stage impactors which are designed to draw air in and capture particles in one or multiple stages, separating the particles according to their size for later analysis (e.g., the Andersen six-stage sampler); (2) filtered through a membrane (e.g., the high volume total suspended particulate air sampler); or (3) passed through a liquid phase (e.g., the bubble flask with culture medium or saline solution). In the analyzed publications, the airflow rate through the active samplers ranged from 10 L min^-1 (for example, using a Hirst spore trap) (HTS) to 600 L min^-1 (using a high throughput jet sampler), and the filtering time also had a wide range of variation (from minutes to days), so the total volume of air sampled was very different from study to study.

On the other hand, passive or gravimetric methods are devices used to capture airborne particles, contaminants, and microorganisms, without the need for active suction. These samplers are based on the natural diffusion of pollutants towards a capture surface. Examples are Erlenmeyer flasks and Petri dishes, both with general or specific culture media (solid or liquid), and Durham spore traps. Unlike active samplers, passive samplers do not require an external power source to draw in air, making them simpler and more cost-effective to use as they do not require a pump or controlled airflow.

As can be seen in Tables II and III, in both bacterial and fungal studies, active samplers have been used much more frequently than passive samplers.

In a study by ^{Serrano-Silva and Calderón-Esquerro (2018)}, three samplers were compared: (i) a Durham-type spore trap (passive), (ii) a Hirst-type spore trap (active), and (iii) a high-throughput ‘jet’ sampler (active), and each of these three samplers was found to preferentially collect certain groups of bacteria. The passive Durham-type spore trap preferentially collected cyanobacteria, the jet sampler collected mainly proteobacteria and firmicutes, and the Hirst-type spore trap collected mainly actinobacteria, proteobacteria, and firmicutes, with the latter collecting the greatest diversity of microorganisms.

Once the collection is complete, the sample may consist of a Petri dish with a culture medium, a liquid culture medium or saline solution with microorganisms in suspension, a Melinex tape with microorganisms attached, a membrane or paper filter, or even bulk particles.

3.3.2 Laboratory approach

The second key step in the study of airborne microorganisms is the laboratory approach, which refers to the main technology used in the laboratory to analyze the microorganisms contained in the samples, and depends on the purpose of the study, the microorganisms of interest, and the facilities and financial resources available. In the 35 studies reviewed, three different laboratory approaches were identified: (i) those based on culture techniques, (ii) those using microscopy, and (iii) those using metagenomic technology (Fig. 3).

In culture-based studies, the microorganisms in the sample were grown and isolated either using general culture media, such as trypticase soy agar (TSA) for bacteria, malt extract-agar for fungi, or specific culture media such as xylose lysine deoxycholate-agar for enteric pathogens, MacConkey-agar for enterobacteria, mannitol salt-agar for staphylococci, or sodium azide esculin bile-agar for enterocococci, among others.

In turn, microscopy-based studies involved the observation of a sample of bulk particles under a microscope, which can be optical or a scanning electron microscope. However, in both cases, this technique only visualizes particles and microorganisms within specific size ranges.

As for the third approach, metagenomic studies involve the extraction of total DNA from the sample by using an extraction kit or laboratory-made solutions. This DNA is then sent for either amplicon or shotgun sequencing. In the papers we analyzed using this approach, sequencing was performed by amplicons, using as target the hypervariable regions of the rRNA16S gene for bacteria (e.g., ^{Calderón-Ezquerro et al., 2020}, ²⁰²¹; ^{Rodríguez-Gómez et al., 2020}), and the ITSs regions for fungi (e.g., ^{Calderón-Ezquerro et al., 2020}). The main sequencing platforms used were Ion Torrent and Illumina-MiSeq, both considered next-generation sequencing (NGS) platforms. So far, no study has used shotgun sequencing to investigate airborne microbiota in Mexico.

It is important to note that each of the different laboratory approaches has some bias, and it is not possible to directly compare results from different studies without taking this into account. For example, metagenomic studies allow the detection of a much greater richness of microorganisms but often do not allow their reliable identification down to low taxonomic levels (such as species or strain). In contrast, culture-based techniques do not allow detection of the vast majority of microorganisms present in a sample but do allow very precise classification. That is why the results presented in Tables II and III are at the genus level. It is essential to bear this in mind when comparing results from different studies. In general, it can be said that it is possible (and important) to compare which microorganisms were present in different studies, but it is not possible to do so at low taxonomic levels, and it does not make sense to compare absences of micro-organisms between studies that used different approaches. It would be desirable to develop a protocol that would allow comparison of results between studies.

3.3.3 Microorganisms identification technique

The third key step is the identification of the microorganisms. Depending on the methodological approach followed, there is a wide diversity of identification methods. In the case of studies of culturable organisms, identification can range from morphological identification using taxonomic keys and biochemical galleries, to the identification of specific organisms of interest, such as Escherichia coli, by PCR or DNA sequencing of colonies growing on a specific culture media.

In studies following a metagenomic approach, once DNA reads are available, they are analyzed using bioinformatics procedures. This usually involves cleaning, filtering and assembling the sequences to eliminate sequencing errors and chimeras, and finally, taxonomic classification is carried out by comparing them against existing databases. The databases that have been used for taxonomic classification in studies in Mexico have been the Ribosomal Database Project (RDP), Greengenes for bacteria, and UNITE + Findley for fungi.

Finally, studies using microscopy have been based on the counting of spores collected in the environmental sample, followed by morphological identification using taxonomic keys.

Despite their differences, advantages, or limitations, the insights provided by different methodological approaches can help to gradually piece together the jigsaw of airborne microbial ecology. Of the 35 studies conducted in Mexico, 27 (77%) have been carried out using culture techniques, four (11.5%) have been carried out using microscopy, and the remaining four (11.5%) have been carried out using metagenomics. Figure 4 shows a trend where studies using culture techniques dominated until 2017, and from 2018 onwards, studies with a metagenomic approach have become increasingly frequent. Although the number of published studies using metagenomics is still quite low (totaling four, all of them published in the last six years) this is expected to increase, as the information they provide is broader. However, this change is likely to be gradual, as metagenomic studies are more expensive and the number of researchers in Mexico who have mastered this methodology is still limited. Additionally, it is worth mentioning that so far almost all metagenomic studies have been done by amplicon sequencing (16S, 18S, ITS) and have reported almost exclusively bacteria and fungi. Shotgun studies are needed to identify a wider range of microorganisms in the samples.

Fig. 4 Number of studies by approach. 

3.4.1 Limitations of comparing results from different studies

Differences in sampling methods, laboratory approaches, and microorganism identification techniques make it difficult to compare results between studies. First, there is an inherent bias in any sampling method that may result in certain microorganisms being over or under-represented. Second, regarding laboratory approaches, there is a gap in the spectrum of microorganisms identified using culture techniques. It has been believed for many years that only 1% of microorganisms can be cultured using conventional culture media (^{Torsvik and Øvreås, 2002}); however, authors such as ^{Martiny (2019)} argue that there have been significant advances in cultivation techniques in recent years and currently the percentage of culturable organisms is much higher. However, virtually all researchers agree that there is a huge gap in the microorganisms that can be identified by culture-based techniques, and that this can be minimized by using metagenomic techniques, as these can identify a broader spectrum of microorganisms in a community. Third, methodologies for the identification of microorganisms also have an important bias, as different results can be produced depending on whether morphological, biochemical, or molecular techniques are used.

The most studied group of airborne microorganisms in Mexico has been the fungal kingdom, accounting for 57% of the publications, and it is the only group that has been studied in Mexico using the three approaches discussed above. The second most studied group in terms of number of publications has been bacteria, which have been studied with both culture techniques and metagenomics. Finally, studies on archaea, microalgae, amoebae, yeasts, and protozoa are limited to four publications, and these have only been carried out by analysis of culture techniques and/or microscopy (Fig. 5).

Fig. 5 Number of publications on airborne microorganisms in Mexico according to the different approaches for their study. Note: The sum of the studies does not match the total number of studies conducted (35) because some studies investigate two or more types of organisms. 

Due to the great importance of both bacteria and fungi for people and ecosystems, it is to be expected that the trend in the future will be towards an equal increase in the number of studies on these two groups of microorganisms. However, this is an assumption based on what has happened internationally and not on what has happened so far in Mexico, where fungal studies have received the most attention.

3.4.2 Locations where studies have been conducted

Airborne microbiota studies in Mexico have only been conducted in 14 of the 32 states in the country: Aguascalientes, Baja California, Chihuahua, Mexico City, State of Mexico, Michoacán, Morelos, Nuevo León, Puebla, San Luis Potosí, Sinaloa, Sonora, Veracruz, and Yucatán. Furthermore, only one study has been conducted in 10 of these states, and it was performed exclusively with culture techniques (Fig. 6).

Fig. 6 Map of Mexico Map indicating the states where studies of airborne microbial communities have been reported. Note: The number of studies conducted in different states of Mexico represented on the map adds up to 37, as one study was conducted in both Mexico City and Veracruz (^{Rosas et al., 1989}), and another study was conducted in both Mexico City and the State of Mexico (^{García-Mena et al., 2016}). 

The state where most studies on airborne microbial communities have been carried out is Mexico City, with 46% of the publications. Furthermore, it is also the only state in which studies have been carried out with a metagenomic approach, and one of only two states in which, in addition to culture techniques, studies have been carried out using microscopy (the other being Sonora).

It is worth noting that of the 35 studies conducted to date, 31 have been carried out in densely populated cities, and only four have been conducted in places uninhabited by humans (one in an avocado plantation, one in a dairy cattle production system, one in a landfill, and one in the Chihuahua desert, along the US-Mexico border). This is important because only by studying unpopulated and relatively pristine places will it be possible to have a contrast to serve as a control against which to compare the impact of cities on airborne microorganisms. So, it is to be expected that studies in other types of locations will gradually appear in the future.

3.4.3. Main bacterial genera identified

According to the findings presented in Table II, a total of 14 scientific papers have been published regarding airborne bacterial microbiota in Mexico. The majority of these studies have utilized active samplers as their primary sampling method. Nine of these investigations have employed culture techniques, while four have utilized metagenomics. It is noteworthy that approximately 40% of these studies were conducted specifically in Mexico City. Consequently, a significant knowledge gap exists regarding airborne bacterial communities in Mexico, arising from two main factors: first, the employment of culture techniques, which only allow for the identification of a small fraction of the genera present in a given sample, and second, the numerous unexplored and intriguing locations and environments within the country.

Most of these studies have focused on the identification of pathogenic bacteria. Consequently, it is not surprising that the vast majority of the genera found include species pathogenic to plants (e.g., Citrobacter, Erwinia, Pseudomonas) or to animals and humans (e.g., Acinetobacter, Bacillus, Brucella). In fact, many of the genera found belong to the order Enterobacterales (e.g., Enterobacter, Escherichia, Salmonella), which is not only associated with a wide range of clinical syndromes but is also a major causative agent of foodborne enteritis and zoonotic infections, including sporadic to pandemic outbreaks, although it is worth mentioning that is also widely dispersed in nature in many natural ecosystems (^{Janda and Abbot, 2021}). On the one hand, this is important since all studies have been conducted in densely populated cities, and therefore imply that there is a large number of people in direct contact with airborne pathogens, but on the other hand it does not provide knowledge and understanding on the diversity, behavior, and ecological characteristics of airborne bacterial communities.

It is worth mentioning that metagenomic studies have identified a much larger number of bacterial genera than culture and microscopy studies. For example, a metagenomic study by ^{Serrano-Silva and Calderón-Ezquerro (2018)} reported 637 genera in the air of Mexico City, far from any of the studies using cultivation techniques. In addition, these methods have allowed the detection of low-abundance genera (those with < 1% relative abundance) such as Gordonia, Ruminococcus, Comamonas, Arcobacter, and Fusobacterium (^{Calderón-Ezquerro et al., 2020}).

Most studies on airborne bacteria in Mexico have been carried out in temperate (57%) or arid (36%) regions, with tropical areas being the least studied (7%). As for the seasonality, most of them (64%) have focused on the description and general characterization of the airborne microbiota regardless of its longitudinal changes. That is, they have not compared microbial communities from the same site at different times of the year (Tables II and III). The few longitudinal studies carried out in Mexico have mainly compared the dry, wet and cold front seasons, and only a few have considered the conventional seasons (spring, summer, autumn and winter).

It is also important to note that the existing studies on airborne bacterial microbiota in Mexico have not yet generated sufficient data to discern notable trends or patterns. For instance, only very few studies have explored seasonal variations, and there is a notable dearth of information regarding altitude-based differentiations, comparisons between tropical and northern regions, or distinctions between polluted and pristine areas, among other aspects. Furthermore, limited attention has been given to investigating potential correlations between airborne bacterial communities and the physical and chemical properties of airborne particulate matter. Similarly, the influence of climatic characteristics, including humidity, temperature, and UV radiation on these communities remains significantly underexplored. Only few studies have considered the integration of environmental variables as metadata for understanding the ecology of airborne bacteria.

3.4.4. Main fungal genera identified

Fungi have been the most extensively studied group of airborne microorganisms in Mexico, with a total of 15 studies, conducted in nine of Mexico’s 32 states (Table III). As in the case of bacteria, most of the fungal studies have been conducted in state capitals or densely populated cities. In fact, only two of them fall outside this pattern, one conducted in the coastal city of Sisal, Yucatan, and the other in Ocuituco, Morelos. The latter was carried out with the aim of identifying avocado pathogens, which highlights the importance of studying airborne fungi also for their role as pathogens of economically important crops.

As in the case of bacteria, active samplers were the most commonly used to study airborne fungi, and culture techniques were the most commonly used laboratory approach (65% of studies), followed by microscopy (two studies) and metagenomic (two studies). It is interesting to note that, in addition to the studies included in this review where airborne microorganisms were collected directly from the air, fungi have been also collected indirectly from the atmosphere via spores in rainwater, as in the study by ^{Rosas et al. (1986)}.

Regarding identification, the most common method was the use of taxonomic keys. It should be noted that metagenomic approaches to the study of airborne fungi have only been used in Mexico City, indicating that in most parts of the country there is still much to be discovered about the composition of airborne fungal communities and their ecology.

The genera Cladosporium and Alternaria stand out by far as the most commonly found genera in the studies, having been identified in 70 and 60% of the publications, respectively. This is interesting, as both have human and plant pathogenic species (e.g., ^{Thomma, 2003}; ^{Rivas and Thomas, 2005}; ^{Mamgain et al., 2013}). Another fungal genus of importance is Fusarium, which encompasses a wide variety of species}, known not only as plant pathogens, but also for their ability to produce mycotoxins and to be opportunistic pathogens for humans (^{Ma et al., 2013}). Additionally, other less represented genera that are also potentially plant pathogens were identified, such as Acremonium, Bipolaris, Colletotrichum, Epicoccum, and Helminthosporium, among others.

In addition to their importance as potential human and plant pathogens, some of these genera stand out for their importance as allergens. Among the main genera identified and reported as potential allergens are Alternaria, Aspergillus, Aureobasidium, Cladosporium, Epicoccum, Fusarium, Helminthosporium, Mucor, Nigrospora, Penicillium, and Rhizopus (^{Burge, 1985}; ^{Horner et al., 1995}; ^{Kurup, 2003}). This suggests a wide distribution of potentially allergenic fungi in Mexico and that they are taxonomically diverse. Nevertheless, there are very few studies on the relationship between the presence of airborne allergens and hospital admissions for allergies (in particular asthma), so it is important to highlight the work by ^{Rosas et al. (1998)}, in which pollen and fungal spores were identified as potential allergens.

Finally, molds were also found, of which several species of genera widely distributed in the analyzed literature, such as Alternaria, Aspergillus, Fusarium, Mucor, Phoma, Penicillium and Rhizopus, are known to cause food spoilage (^{Shapaval et al., 2013}). The genus Ulucladium is known to spoil in particular nuts, fruits and cereals (^{Andersen and Hollensted, 2008}). Furthermore, the genera Acremonium, Aspergillus, Penicillium, and Ulocladium have been reported as the most common fungi found in moisture-damaged buildings, which could have implications for human health due to their potential allergenicity, besides their role in the degradation processes of building materials (^{Andersen et al., 2011}).

3.4.5. Other eukaryotes identified

Interestingly, in addition to studies focusing on bacteria and fungi, seven publications were also found on other types of airborne organisms, such as algae, amoebae, protozoa, and yeasts (Table IV). All of these investigations exclusively utilized active samplers and culture-based techniques, and their scope has been limited to four states of the Mexican Republic, most of them carried out in Mexico City (50%). It should be noted that several of these studies found genera that contained both free-living species and some that are hazardous to human health (e.g., Acanthamoeba, Bodo, Candida).

It is worth mentioning that two new species of algae, Brotryokoryne simplex and Chloroglouea microcystoides, were discovered in Mexico City in one of these works (^{Rosas et al., 1987}).

Another interesting finding is that only one of the 35 publications studied airborne pollen. This is probably because pollen is usually considered a bioparticle and is not a microorganism. In the study by ^{Ortega et al. (2019)}, in addition to investigating airborne fungi, airborne pollen was also examined. They found that the plant taxa that contributed the most pollen to the air were Poaceae, Nyctaginaceae, Ambrosia, and Urticaceae. It is worth mentioning that in addition to this study, there are some other publications on airborne pollen in Mexico (e.g., ^{Tarragó, 1996}; ^{Calderón-Ezquerro et al., 2016}; ^{Ríos et al., 2016}).

The study of other eukaryotes such as algae, protozoa, and amoebae has focused mainly on temperate climate regions, and only few attempts have been made to determine correlations with environmental variables such as solar radiation, relative humidity, temperature, and wind speed. In the case of algae, their abundance has been found to be related to wind speed, vapor pressure, and temperature (^{Rosas et al., 1989}), while the higher abundance of protozoa has been linked to temperature and wind speed and direction (^{Rivera et al., 1992}). Finally, studies on amoebae indicate higher abundance during the dry season and their association with wind speed and direction, and low relative humidity (^{Rodríguez-Zaragoza et al., 1993}).