Intercomparison of Indoor and Outdoor Pollen Concentrations in Rural and Suburban Research Workplaces

: Pollen exposure in occupational settings involves different categories of workers. In this paper the effects of diurnal pollen variations have been evaluated in two sites characterized by different vegetation and urbanization: the suburban site of Tor Vergata (TV) and the rural site of Monte Porzio Catone (MPC). Aerobiological and meteorological monitoring was performed in the two sites during the winter of 2017. The data analysis focuses on the comparison between pollen concentrations observed in relation to meteorological variables. In general, it can be stated that the indoor and outdoor dynamics for MPC and TV are different, with the outdoor concentration of pollen for MPC always higher than for TV, in accordance with signiﬁcant presence of vegetation. The high nocturnal peaks detected in MPC and completely absent in TV could be caused by the presence of particular conditions of stagnation combined with greater emissions from the pollen sources. Furthermore the higher I/O ratio observed during the working hours in TV compared to MPC could be ascribed to the workers’ behavior. Exposure to pollen can be responsible for several health effects and the knowledge of its level can be useful to improve the evaluation and management of this biological risk. profoundly exhibit similar values If we consider the daily mean, values remain very close (I/O TV 24h and and the slight difference could be to the different habits and behaviors


Introduction
Pollen is one of the main components of bioaerosol. As the microgametophyte of spermatophytes, its role in the life cycle of seed plants (both wild and cultivated) is crucial, and its presence in the air is ubiquitous. From a biological, ecological and agricultural point of view, pollen is indispensable. However, from a human health perspective, pollen can be seen as an aerobiological pollutant mainly responsible for respiratory and allergic diseases [1-6] and has been recently associated with different health effects. Lung function, seasonality of flu-like illnesses, SARS-CoV-2 infection rates, and occurrence of several cancers have been put related to airborne pollen variations [7][8][9][10][11].
Environmental exposure is a critical aspect and several studies evidenced that different outdoor, indoor, rural, and urban environments may contribute to determine or exacerbate the health effects [12][13][14]. In this regard a key role should be attributed to occupational settings that may increase pre-existing sensitization and/or respiratory pathologies due to other biological, chemical and physical agents mainly present in workplaces. With respect The aim of the study is to evaluate the effects of diurnal pollen variation in two occupational settings with different vegetation and urbanization characteristics in relation to meteorological variables, working days (WDs), non-working days (NWDs), working hours (WHs), and non-working hours (WHs), considering the indoor and outdoor ratios.

Study Sites and Vegetation
The study was conducted inside and outside the Research Centre building of the National Workers' Compensation Authority (INAIL) in Monte Porzio Catone (MPC) as a rural area, and inside and outside the University of Tor Vergata (TV) in Rome as a suburban area. Monte Porzio Catone is a hilly area located in southeast Rome, in the Castelli Romani regional park at~300 m a.s.l. (41 • 49 19.5 N, 12 • 42 24.1 E); Tor Vergata is located in a south-easterly neighborhood of Rome at about 80 m a.s.l. (41 • 51 13.5 N, 12 • 36 14.2 E); the two sites are located in the Mediterranean biogeographical region. The climate of Rome is broadly sub-Mediterranean, with mild winters and hot summers (three/four months of aridity). The average annual temperature is 15 • C and average annual rainfall is 839 mm [69,70]. The vegetation is complex and varied, characterized by flora of high biodiversity and many taxa of high conservation value. The dominant types are sub-Mediterranean deciduous and evergreen mixed forests with prevalence of oak woods with Quercus cerris, Q. frainetto, Quercus ilex, Q. robur, Q. dalechampii and Q. pubescens, other common woody species such as A. monspessulanum, Acer campestre, Fraxinus ornus, Corylus avellana, Crataegus monogyna, Viburnum tinus, Pistacia lentiscus and trees belonging to the Robinia, Ulmus, and Ailanthus genera [69][70][71]. In the rural area (MPC) cultivates areas are most common including species such as Olea europaea, Vitis vinifera, Corylus avellana and chestnut woods (Castanea sativa Miller) fields, continuous prairies, hedges, many forests with oak trees, coniferous trees, vineyards, olive groves, and vegetable gardens. The area is surrounded by Cupressaceae, Pinaceae and Oleaceae. Herbaceous plants such as Urticaceae, Plantaginaceae and Poaceae are also present (http://websit.cittametropolitanaroma.it/DescriviMappa.aspx?i=7 accessed on 12 January 2021). The greenery of the suburban area (TV) is composed of pastures, abandoned fields, uncultivated land, urbanized and degraded areas, artificial surfaces, numerous native species due to human impact that are well adapted to human presence, and non-native species of uncertain origin [71].
Data regarding the vegetation maps around the sampling area were retrieved from the website http://websit.cittametropolitanaroma.it/DescriviMappa.aspx?i=7 (accessed on 12 January 2021) while data on buildings were obtained from https://dati.lazio.it/it (accessed on 22 April 2021). The retrieved data were processed using QGIS 3.16.1, obtaining measures of the areas covered by vegetation, artificial surfaces and buildings. The area measures were used to calculate the plan-area density (λp) of buildings and the Vegetation Index (Vi). The λp indicator (i.e., the fraction of area), defined as the ratio between the built area and the total area, is used to identify the presence of buildings whereas, regarding the green areas, the Vegetation index (Vi), defined as the ratio between the overall vegetated area and the total area, is considered.

Aerobiological Monitoring
Aerobiological monitoring was performed during winter 2017 (from 2 to 21 February 2017), collecting aerobiological particles in accordance with the UNI 11108/2004 and following UNI CEN/TS 16868:2015, using 7-day volumetric samplers, Lanzoni VPPS 2000 (Bologna, Italy), Hirst-type [72]. In the rural area (MPC) one sampler was placed outside the Research Center of INAIL, Monte Porzio Catone, at 1.10 m above ground level, another one inside the building at 0.6 m above ground level.
In the suburban area (TV) one sampler was located on the roof of the Biology Department building of the University Tor Vergata, Rome, at about 12 m above the ground level, and another one inside a room of the same building at 1 m above ground level.
The sampler consists of different components: a single-stage impactor of particles, a suction pump, an intake orifice, a rotating drum, and a directional wing. The air is captured by a vacuum pump to assure a constant flow of 10 L/min corresponding to the human mean breath rate, through an orifice of known dimension (2 × 14 mm) oriented continually against the wind due to a directional wing and positioned at a height of at least 1 m. The air flow is directed onto a surface consisting of a transparent plastic Melinex ® tape properly equipped, coated with a 2% silicon solution as trapping surface. The trapping surface moves at a speed of 2 mm/h on a cylindrical rotating drum that makes a complete round in a week [19,73,74].
At the end of the exposure the sampling surface (sticky tape) is cut into segments of 48 mm representing daily samplings and prepared to be mounted on microscope slides. The tape is treated with glycerin jelly in order to adhere perfectly to the slides and stained with basic fuchsine solution which colors only pollen grains. Lastly the slides are observed via a light microscope with a magnification of 400×.
According to the Italian standard methodology the minimum number of horizontal sweeps must correspond to an area close to 20% of the impacted surface and the sweeps spaced 2 mm from each other, as well as from the edge of the sampling surface [73]. The pollen counts are expressed as number of particles/m 3 of air [19,73,75], pollen concentrations are provided with a 30-min time resolution and, in the analysis, 2-h and 6-h averages are calculated and considered.

Meteorological Monitoring
Meteorological data were provided by the University of Rome "Tor Vergata" and collected simultaneously with the aerobiological measurements, during winter 2017 (from 2 to 21 February 2017), by a meteorological station located on the roof of the Experimental Ecology and Aquaculture Laboratory (LESA), 2.18 Km from the TV pollen trap. The monitoring was performed using a CAE SPM 20 station (Bologna, Italy) connected to sensors for air temperature, humidity, wind speed and direction (at 10 m above ground level) and rain (at 2 m above ground level). Data for air temperature ( • C), relative humidity (%), wind speed and direction (m/s and deg), precipitation (mm) and solar radiation (W/m 2 ) were recorded with a 30-min time resolution and, regarding the pollen analysis, the 2-and 6-h averages were subsequently considered. Preliminarily, the daily evolution of the main meteorological variables, also in relation to the vegetation maps, allow better characterization of the possible pollen transport/dilution conditions during the study period. The arithmetical means and the standard deviations of the meteorological variables over the entire period were calculated to complete the site description. Daily evolution of the meteorological variables is also presented and scatter plot between wind speed and direction highlights the main sectors of provenience related to the intensity of wind. It must be kept in mind that, for the MPC site, meteorological measurements are not available but, considering the short distance (7.12 km) and the objectives of the analysis presented in this study, we can consider the meteorological condition as homogeneous between the two sites.

Data Analysis
Data analysis focuses on the comparison between pollen concentrations observed in two sites, relatively close to each other but characterized by profoundly different land use (see Section 2.1); the analysis aims also to highlight the differences that can occur during the working days (WDs) and the non-working days (NWDs), working hours (WHs) and the non-working hours (NWHs), in each monitoring site in order to observe possible and different behavior patterns in relation to specific conditions, regarding also the exposure of workers [19].
In brief, three topics are the main components of the data analysis and the results will be presented as follows: 1.
outdoor and indoor pollen characterization in WDs and NWDs, WHs and NWHs in the two sites; 2.
the peak analysis, where the maximum of the outdoor pollen concentration for each day of measure is selected and related to the meteorological conditions; 3.
finally, the indoor and outdoor Ratios (I/O) are calculated and compared for the two sites.
As well as described above for the meteorological data, pollen concentrations are provided with a 30-min time resolution and, in the analysis, 2-h and 6-h averages are calculated and considered.

Average Pollen Values for TV and MPC
Starting from the observations collected every 30-min, average values for indoor and outdoor pollen concentrations were calculated for the considered study period. Mean values for the two sites and their standard deviations are reported in Table 1. As shown in Table 1, the two sites exhibit different pollen concentrations; regarding the outdoor component, MPC is characterized by higher values of about 46% more compared to TV.
The observed values of the outdoor pollen concentration compared to the indoor are higher in both sites (about 59 times for MPC and 22 times for TV).
In general, it can be stated that the indoor and outdoor dynamics for MPC and TV are different.

Land Use Characterization of TV and MPC
In order to interpret the observed pollen behavior, the buildings and vegetation distribution of an area of about 12 km 2 around the two sites are taken into account and shown in Figures 1 and 2.
The area of interest has been classified into two categories (green areas and artificial surfaces), grouping the different types of vegetation into the green area category and nonvegetated areas into artificial surfaces. Green areas are the main potential sources of pollen; however, no consideration has been made regarding the type of plants or pollen species.
The pollen impact measured at the two sites depends, firstly, on transport related to the wind direction but also on the distance between the site and the possible source of pollen. For this reason, data have been considered as a function of different classes of distance, reported in Table 2.   In this way, it is possible to identify, at different distances (up to a maximum of 2 km), the characteristics of the terrain in the vicinity of the two examined sites. Eight sectors (octants) are considered for the wind direction. Figure 1 shows the built-up area for TV and MPC, whereas Figure 2 reports the green areas. In both figures the angular sectors representing the wind direction, as in Table 2, are highlighted.
The average values of λp (calculated from Table 3, as function of the distance and the sector of provenance) for TV and MPC are shown in Figure 3.  Both sites are characterized by unbuilt areas (except for TV in the first 100 m, where λp = 0.47), but TV is characterized by a higher presence of buildings compared to MPC (λp = 0.19 and λp = 0.06, respectively). Table 4 reports the Vi as function of the distance and the wind direction for TV and MPC. However, the vegetation index is small in the areas close to the two sites (first 100 m) but grows with the distance (Figure 4a  Considering the wind direction, the vegetation index, for MPC, does not vary considerably whereas, for TV, Vi remains low, showing values comparable to MPC when the wind comes from the east (Figure 4a,b).

Study of Meteorological Variables and Pollen Concentrations
TV and MPC terrains are profoundly different and this aspect influences the possible source of pollen with respect to the wind direction.
To better evaluate this aspect, considering the meteorological condition as homogeneous between the two sites as indicated in Section 3.3, meteorological variables measured on TV were related to the outdoor pollen concentration observed in the two sites. Table 5 shows the mean values of the meteorological variables observed during the measurement campaign. Wind speed values are on average higher for the sector between North-South (WS = 1.5 m/s) while lower values are observed from West (WS = 0.8 m/s). Data for solar radiation show higher values for directions from SW-W-NW. During the night (not shown in the table), the mean temperature and wind speed are lower (Tnight = 6.8 • and WS = 1.02 m/s) than those observed during the day (Tday = 11.6 • and WS = 1.33 m/s), corresponding to a mean relative humidity of 92.5% and 72.5%, respectively.
Considering the pollen at TV and MPC, Figure 5 shows the observed values as function of the wind direction measured at TV. It should be emphasized that while for TV, meteorological variables and pollen concentration are measured on the same site, for MPC we assume that the wind regime is similar to that observed in TV; our assumption is supported by the evidence that the terrain between the two sites is generally free from obstacles, as shown in Table 3. Figure 5 shows that, for TV, the minimum pollen concentration corresponds to the South direction (P (TV) = 1.24 particles/m 3 ), while the maximum of transport occurs for winds from NW (P (TV) = 6.57 particles/m 3 ). For MPC, the observed values are, on average, higher and less dependent on the wind direction in accordance with the vegetation index (Table 4), which shows higher values for MPC, independently of the wind direction (see Figure 4).

Analysis of Typical Days NWDS and WDS for Indoor and Outdoor Pollen Concentrations
Aiming to evaluate occupational exposure, the hourly average values are considered in the analysis of a typical day, taking into account separately the working days and the non-working days. Figure 6a,b shows the outdoor concentrations for the MPC and TV. The outdoor concentrations for MPC are always higher than for TV, about 79% and 27% more during the working days and the non-working days, respectively; the difference between the pollen concentration in the two sites is probably due to the influence of the Vi that characterizes MPC and TV, but also to the high agricultural vocation of the MPC area, where the average pollen values reach the maximum during the holidays (P NWDs (MPC) = 6.5 particles/m 3 ), definitely higher than that observed at TV (1.1 particles/m 3 ) for the same period.  A first consideration that can be made regards the habits of workers (their presence and behaviors) that influence the indoor pollen concentration during the day. This aspect will be discussed in Section 3.6.

Two Hour Resolution Analysis
The purpose of the peak analysis is to identify the outdoor events characterized by the maximum concentration of pollen, also considering when during the day these events occur. In general, peaks can be studied at the maximum resolution (in our case 30 min); however, this temporal resolution is too high, giving rise to multiple maxima of equal intensity; thus, the 2-h and 6-h averages were considered. In the first case, this permits evaluation of the impact of pollen following the evolution during the day, whereas, in the second case, it studies which time slot of the day (morning, afternoon, evening or night) is characterized by the presence of the maximum.   Considering the time of the day when peaks occur, during working activities the most critical time range is from 1:00 to 5:00 p.m. for both MPC and TV (Figure 8).
This result shows how the two sites respond differently to the solicitation related to the emission sources and to the presence of turbulence.
As reported by other papers, the peaks can occur at all hours of day and their variations depend on a complex effect of meteorological factors and distance from the pollen source [50,56]. Analysis of two hours peak resolution have shown differences in relation to areas of sampling due to pollen transport dynamics, with the highest pollen concentrations during daylight hours and maximum intradiurnal values recorded usually between 12 and 14 a.m. and early afternoon [25,26,48,51,53]. Furthermore, several authors have shown that pollen concentrations are higher in rural than in urban areas [36].

Six Hours Resolution Analysis
Pollen concentration maxima are also considered for the following time ranges of the day (6-h averaged data).  As stated before, in the literature, the daily pattern distribution of pollen has been investigated and several authors have highlighted the presence of high values of pollen in the morning (06:00-12:00 a.m.) and in the afternoon (12:00-18:00 p.m.) [40,47,48].

Evaluation of I/O Ratios for TV and MPC
The I/O ratio indicates the presence of the indoor pollen we can expect in relation to the outdoor pollution. This relationship is based on the assumption of an instantaneous correspondence between external and internal pollution, but this is, especially for a short time range, not always true; thus, to obtain more reliable results, data averages (2-h, 6-h and 24-h averages) are considered.
In Figure 10, the evolution during the day of the mean I/O ratio for the two sites is shown. During the night and in the early morning (from 7:00 p.m. to 9:00 a.m.), TV and MPC exhibit similar values, slightly higher for TV (0.06 and 0.02, respectively). Instead, during the working hours (from 11:00 a.m. to 5:00 p.m.), values observed for TV and MPC are very different. TV is characterized by higher values of the ratio (mean I/O TV = 0.40) compared to MPC (mean I/O MPC = 0.03). As reported in the previous section, it should be kept in mind that, at MPC, the outdoor pollen concentrations are higher than those observed at TV.  [78].
To evaluate the differences observed in the I/O ratios for the two sites, worker habits, as highlighted in Table 6; have been taken into account; data on these habits are obtained by means of a survey form [19,20] filled in by workers, reporting any changes in occupant numbers present in the room and the indication of open windows or doors (in minutes). Data are then averaged. showing that ventilation conditions are fundamental for the determination of the I/O ratio and that differences between the two sites are principally ascribable to habits rather than to the outdoor pollen concentrations.
In Figure 11

Conclusions
Bioaerosol exposure derived by natural and anthropogenic sources has assumed greater importance in recent years [5]. Exposure to pollen biocomponents can be responsible for several health effects and knowledge of their levels, interactions with other components, modification in relationship with the presence and absence of occupants, as well as their diffusion in several environments can be useful in improving the evaluation and management of this biological risk.
The concern in urban environments with respect to rural ones is peculiar causing different health consequences, also due to interactions with pollutants of different origins, and with urban greening that can be associated with increased exposure to pesticides and allergenic pollen, although this evidence is weaker [79,80]. On the other hand, health benefits due to the presence of urban green spaces and infrastructures have been reported, including the attenuation of indoor temperatures and heat islands, the improving of air quality, the reduction of temperature and air pollution [81] and a lack of vegetation associated with a higher risk of heat-related mortality [82]. In addition, Sustainable Development Goal 3 of the 2030 Agenda for Sustainable Development is to "ensure healthy lives and promoting well-being for all at all ages". Moreover, another target is to reduce the number of deaths and illnesses from hazardous chemicals and pollution (https://sdgs.un.org/topics/health-and-population accessed on 3 June 2021), and goal 11 aims to "Make cities and human settlements inclusive, safe, resilient and sustainable" (https://sdgs.un.org/goals/goal11 accessed on 3 June 2021).
It is essential to improve pollen monitoring techniques and to implement automatic methods, favoring the use of personal pollen samplers and the application of molecular methodologies [83], and taking into account the role of occupants as a potential cause of indoor pollen level variation, as highlighted in previous studies [19,66,67] and in the current study, conducted in two different occupational settings. The important role of human behavior in pollen diffusion and/or retention has also been evidenced by experimental studies, where a mean of 0.93% of the initial pollen load was retained after a single wash of hands and traces of several species were found after numerous hand-wash cycles [84]. Moreover, the knowledge of pollen hourly peaks creates the opportunity to plan appropriate preventive measures and recommendations, addressed mainly to the sensitization of workers [85].
We suggest that different items should be considered in order to evaluate health effects derived by pollen exposure, such as urbanization and, conversely, vegetation areas [37,38,86], implementing predictive models in order to reduce pollen exposure and understand the complex interactions between allergens and other particles, in order to investigate the genesis of allergy and develop new preventive and therapeutic strategies [6,87], in relationship also to death due to respiratory and cardiovascular diseases [88] and to health costs [59].
The promotion of synergic and integrated approaches is needed [22], coupling data derived by individual experimental and epidemiological studies in order to optimize preventive strategies considering the most peculiar variables belonging to several occupational settings.
The knowledge of the spatial and temporal dynamics of pollen intradiurnal variations is indispensable to plan in advance and implement effective prevention strategies [26,27,35,49,53]. In this respect it is fundamental to increase the awareness of workers regarding the risk derived from exposure to allergens and to define action mechanisms, control and corrective measures such as limitation of outdoor activities, reducing the time of exposure to allergenic pollen in the peak periods, and the taking of biological vaccines and drugs before the pollen season begins. Furthermore it is important to promote the use of new techniques (i.e., GPS, remote sensing technology) to identify areas free of allergens and progressively remove allergenic taxa of ornamental plants to prevent high pollen concentrations in urban environments [27,36,38,61].