Morphological and biochemical changes in Quercus humboldtii in response to warmer and polluted urban areas in a tropical Andean city

Abstract

Introduction

Urban areas cover ∼3% of the Earth's land surface (Liu et al. 2014) and are home to over 55% of the global population (Esperón-Rodríguez et al. 2022). However, projections indicate that by 2050, this percentage will increase to 68% (United Nations 2019). The rapid population growth in cities has led to significant changes in human settlement patterns, energy consumption, transportation, and industry (Ulpiani 2021). These changes have resulted in the formation of Urban Heat Islands (UHI), characterized by high temperatures and low relative humidity (McMichael et al. 2003; Ulpiani 2021). As changes in temperature and pollution are both strictly related to combustion processes from transport, industry and other anthropogenic activities (Ulpiani 2021), it is common for UHI also to exhibit high pollution levels (Jin et al. 2022). Both air pollution and elevated temperatures are widely recognized as major environmental risk factors for public health (Heaviside, Macintyre, and Vardoulakis 2017; Murray et al. 2020). Alarmingly, air pollution is responsible for ∼7 million deaths worldwide each year (Rodríguez-Villamizar et al. 2023). This issue is particularly concerning in Latin America, where cities are densely populated, and a significant portion of the population lacks access to basic healthcare and nutrition, making them highly vulnerable (Bell et al. 2006).

Trees have long been recognized as a nature-based solution due to their ability to regulate environmental temperatures, reduce air conditioning costs (Rosenfeld et al. 1998; Stone et al. 2014), filter particulate matter from the air, and decrease hospital admissions for respiratory diseases (Escobedo et al. 2008; Tiwary et al. 2009; Morani et al. 2011; Chen et al. 2017). However, trees in urban areas are negatively affected by the environmental conditions caused by UHI (Wolfenden and Mansfield 1990; Zvereva, Roitto, and Kozlov 2010; Martin, Simmons, and Ashton 2016; Hilbert et al. 2019), which poses a challenge for urban planning, management, and ultimately human health. Previous research has shown that smaller and thicker leaves may develop in response to pollution (Wen, Kuang, and Zhou 2004; Tiwari, Agrawal, and Marshall 2006; Liu and Ding 2008) as a strategy to reduce the entry of pollutants into the leaves (Pal et al. 2002). Additionally, plants may exhibit these leaf traits in response to high temperatures (Leigh et al. 2017; Slot et al. 2021; Tserej and Feeley, 2021) in order to maintain a stable internal leaf temperature and protect the photosynthetic apparatus from excessive heat (Leigh et al. 2017; Liu et al. 2020; Slot et al. 2021; Tserej and Feeley, 2021). Both heat and pollution have an impact on transpiration and photosynthetic rates. Particulate matter, such as PM_2.5 and PM₁₀ (particulates smaller than 2.5 µm and 10 µm), can adhere to the surface of leaves and alter the chemical composition of the cuticle (Wolfenden and Mansfield 1990). This can also affect the characteristics of stomata and parenchyma (Qin et al. 2014), resulting in a decrease in the rate of photosynthesis (Saunders and Godzik 1986). Additionally, when plants are exposed to high vapor pressure deficit, which is also caused by UHI, they may reduce their transpiration rates to minimize water loss, increasing leaf temperatures that may reach the thermal limits of photosynthetic damage (Fauset et al. 2018). Pollution also increases cell permeability (Keller and Lamprecht 1995), resulting in the loss of water and dissolved nutrients, and premature leaf senescence (Masuch et al. 1988). Additionally, pollution may promote the production of ascorbic acid, an antioxidant that protects cells from free radicals and other toxic molecules (Singh et al. 1991; Smirnoff 1996). Urban heat islands provide a unique opportunity to study the combined effect of several environmental conditions on plant responses to stress, which is crucial for supporting management decisions in cities facing rapid environmental changes.

In Bogotá, Colombia, the levels of atmospheric pollution are alarmingly high, with an average PM_2.5 concentration of 13.9 μg/m³ (Rodríguez-Villamizar et al. 2023), exceeding the recommended thresholds set by the World Health Organization (World Health Organization 2021). Additionally, the tropical Andean region has experienced a temperature increase of 0.34°C per decade over the past 25 years, with projected average temperature increases of 3 ± 1.5°C over the next 50 years (Anderson et al. 2010; Buytaert, Cuesta-Camacho, and Tobón 2011) with strong impacts on plants. Under this scenario, it is essential to study the morphological and biochemical responses of trees to contrasting climatic conditions and pollution levels, especially in the context of urban planning and the ecosystem services provided by trees. Unfortunately, research in this area is still limited in tropical cities. Therefore, our goal was to evaluate the impact of environmental conditions of UHI on the functional traits of Quercus humboldtii, a native and endangered tree species, by comparing trees growing in urban and forest environments in Bogotá, a major city in the Global South. We expect morphological and biochemical differences among the localities, with native forest trees exhibiting greater differentiation due to lower pollution and temperature compared to trees located in Bogotá.

Materials and methods

Study area and species

This study was conducted in Bogotá, Colombia, one of the largest cities in Latin America, which is home to over 7 million people (Rodríguez-Villamizar et al. 2023). Bogotá is located in a region that has experienced significant human impact (Etter and Villa 2000). The city is situated at an altitude of ∼2650 m.a.s.l. and has an average annual temperature of 14°C. It receives ∼1.000 mm of rainfall annually, and the average relative humidity is 73% (Salamanca-Fonseca et al. 2023).

We selected three localities with contrasting pollution levels and climatic environments (Fig. 1; Table 1). L1 is a native forest surrounded by grasslands in the suburban municipality of San Francisco (Cundinamarca). This forest is located away from main roads and experiences minimal vehicular traffic, making it the control site for our study. This forest is dominated by native species such as Critoniopsis bogotana, Palicourea angustifolia, Viburnum lasiophyllum, and Quercus humboldtii. L2 is located in the University City of the Universidad Nacional de Colombia, which is 80% covered by green areas such as gardens, sports areas, and outdoor wooded areas composed of both native and exotic species such as Fraxinus chilensis, Cupressus sempevirens, Tecoma stans and Lafoensia acuminata. Lastly, L3 is situated on a main street in Bogotá (Street 53), with constant vehicular traffic and close proximity to a heavily trafficked main avenue (Street 30). Precipitation and maximum temperatures varied among the three localities (Table 1). L1 and L2 showed the lowest precipitation, with significant differences compared to L3. The native forest (L1) exhibited the lower maximum temperature in comparison with L2 and L3. Pollution levels (PM_2.5 and PM₁₀) also varied among these localities (Table 1), with the highest values found in L3 for both parameters. We extracted the climate data for each study locality from the WorldClim global climate data set (http://www.worldclim.org/) between 2011 and 2020. We obtained pollution data from the meteorological monitoring stations of the Oficina de Gestión Ambiental of the Universidad Nacional de Colombia. For L2 (Nursery Faculty station), we used data from March 2022 to February 2023, while for L3 (Kinder Garden station), we only had information from August 2022 to March 2023.

Figure 1:

Localities studied. L1, native forest surrounded by grasslands in the suburban municipality of San Francisco (Cundinamarca). L2, located in the University City of the Universidad Nacional de Colombia, which is 80% covered by green areas. L3, situated on a main street in Bogotá (Street 53), with constant vehicular traffic and close proximity to a heavily trafficked main avenue (Street 30).

Open in new tabDownload slide

We chose the Andean oak (Quercus humboldtii, Fagaceae) because it is an endemic tree species of Colombia and Panama and dominates the Andean Forest ecosystems (Avella, Camacho, and Torres 2016). Over the past 20 years, this species has been planted as an urban tree in Bogotá due to its ability to tolerate the unique climatic and soil conditions of the city, including frost, poor drainage, and slightly acidic soils with moderate to low fertility (Mahecha et al. 2010). Q. humboldtii is the third most common native species in Bogotá, with nearly 19.000 individuals in the city (www.jbb.gov.co/sigau/).

Morphological traits

Biochemical traits

In the same individuals that were measured for morphological traits, we also measured biochemical traits on three leaves per individual: pH, ascorbic acid (AA, mg/g), and relative chlorophyll content (Chl, mg/g). The biochemical traits were measured on physiologically mature leaves, specifically the fourth and fifth leaves below the branch apex. For leaf pH, we followed the method described by Prasad and Rao (1982) and Singh (2021) with minor modifications. Briefly, we took 0.1 g of fresh leaves and macerated them with liquid nitrogen until obtaining a fine powder. Then, we added 10 ml of distilled water and continued the maceration process until achieving a homogeneous mixture. This mixture was then placed in a shaker for 10 min at 150 RPM. After that, the sample was centrifuged at 6000 RPM for 20 min at 4° C. The supernatant was transferred, and the pH was measured using a digital pH meter (Ohaus aquasearcher TM). To estimate the ascorbic acid (AA), we used the titration method with 2,6-dichlorophenolindophenol (DCPIP) as proposed by Rao and Deshpande (2006), with modifications by Dinesh et al. (2015), and following the equation proposed by Rao and Deshpande (2006). The measurement of relative chlorophyll content, was conducted in situ using an optical meter (SPAD-502; Konica Minolta Sensig, Inc., Sakai, Osaka, Japan). We measured three physiologically mature leaves from the same six individuals at each site between 10:00 and 12:00 noon (Parry, Blonquist, and Bugbee 2014). We performed six readings in the center of the leaf blade avoiding the main vein to obtain an average value. In order to convert the optical chlorophyll values obtained with the SPAD to total chlorophyll content (mg/g), we constructed a calibration curve for the species by extracting chlorophyll with acetone 80% (Lichtenthaler 1987; Melgarejo et al. 2010; Rodríguez, Melgarejo, and Blair 2019) from leaves of different colors (Parry, Blonquist, and Bugbee 2014), which were previously measured with the SPAD. This calibration curve relates the values of total chlorophyll content (mg/g) and optical chlorophyll (SPAD units), with both variables measured at the same point on each leaf.

Data analysis

To assess significant differences in morphological and biochemical traits among the three localities, we conducted a series of univariate ANOVAs followed by a post hoc Tukey test. Since the leaf dry matter content and ascorbic acid did not meet the normality assumption, we performed a non-parametric Kruskal-Wallis test followed by a Dunn test. To further explore the overall multivariate relationships and trait differences among the localities, a principal component analysis (PCA) was conducted. All statistical analyses were performed using R 4.2.1 (R Core Team 2022).

Results

Plant traits differed among localities, as revealed by PCA. The first PCA axis, which explained 30% of trait variation, clearly separated the colder and unpolluted native forest (L1) from the other two warmer and polluted localities (L2 and L3, Fig. 2). This axis was negatively correlated with high SLA and RWC and positively correlated with pH and LDMC. The second PCA component explained an additional 27.3% of the total variance, and it was dominated by the trade-off between Chl and Lth, with AA and APTI. This axis separated the trees located in the warmest and most polluted locality (L3) from the other localities. Similar results were detected when each trait was analyzed separately. Trees growing in the warmer and most polluted locality (L3) exhibited lower LA and SLA and higher LDMC than trees growing in colder and unpolluted native forests (L1), but they did not differ from the middle-polluted locality (L2) (Fig. 3). Regarding biochemical traits, trees from the warmer and most polluted locality (L3) exhibited higher pH values than colder and unpolluted native forests (L1, Fig. 4). Surprisingly, values of AA and APTI did not vary between the warmer and most polluted locality (L3) and colder and unpolluted native forest (L1), but both varied from the L2 (Fig. 4). The APTI values were between 1 and 16 indicating that Q. humboldtii is a sensitive species to contamination (Singh et al. 1991). We did not report significant differences in four traits: Lth, WD, RWC and Chl, among localities (Figs 3 and 4).

Figure 2:

Principal component analysis (PCA) of morphological and biochemical traits for individuals of Quercus humboldtii in three localities with contrasting pollution levels and climate conditions in Bogotá (Colombia). The first two principal components explained the 57% of the variance of plants among localities. Morphological traits: leaf area (LA), specific leaf area (SLA), leaf thickness (Lth), leaf dry matter content (LDMC), relative water content (RWC) and wood density (WD). Biochemical traits: leaf pH (pH), chlorophyll content (Chl), ascorbic acid (AA) and air pollution tolerance index (APTI).

Open in new tabDownload slide

Figure 3:

Boxplots of morphological traits of Quercus humboldtii in three localities with contrasting pollution levels and climate conditions in Bogotá (Colombia). L1, native forest of San Francisco (Cundinamarca-Colombia). L2, University City of the Universidad Nacional de Colombia, covered by green areas (80%). L3, situated on a main street in Bogotá (Street 53), with constant vehicular traffic and close proximity to a heavily trafficked main avenue (Street 30). Different letters indicate significant differences.

Open in new tabDownload slide

Figure 4:

Boxplots of biochemical traits of Quercus humboldtii in three localities with contrasting pollution levels and climate conditions in Bogotá (Colombia). L1, native forest of San Francisco (Cundinamarca-Colombia). L2, University City of the Universidad Nacional de Colombia, covered by green areas (80%). L3, situated on a main street in Bogotá (Street 53), with constant vehicular traffic and close proximity to a heavily trafficked main avenue (Street 30). Different letters indicate significant differences.

Open in new tabDownload slide

Discussion

Our results indicated that Q. humboldtii adjusted its morphological and biochemical traits in response to the environmental challenges posed by urban heat islands in Bogotá. The morphological and biochemical strategies varied among individuals, with thin and large leaves, as well as acid pH in the colder and unpolluted native forest (L1), to individuals with a conservative resource strategy and a pH close to neutral, in the warmer and most polluted locality (L3). Despite changes in these traits, the values of several traits indicate that Q. humboldtii is not a species tolerant to pollution, which is further supported by the APTI values. Therefore, it is recommended to plant this species in low-polluted areas in Bogota in order to increase its longevity and enhance the delivery of ecosystem services.

We found lower LA, SLA, and higher LDMC in trees growing in warmer and polluted localities (L2 and L3) compared to the colder and unpolluted native forests (L1). These morphological changes are consistent with a resource-conservative strategy, which is common in polluted environments (Liu and Ding 2008) and when resources are limited (Diaz et al. 2004). Strengthening cell walls and protecting leaves from mechanical damage enhance nutrient conservation (Warren and Adams 2004). Several studies have also reported a decrease in SLA with pollution (Wen, Kuang, and Zhou 2004; Zhu and Xu 2021), as reduced area and increased leaf density and thickness minimize pollution uptake (Wen, Kuang, and Zhou 2004; Tiwari, Agrawal, and Marshall 2006). Additionally, high LDMC and smaller leaves are advantageous in warmer and drier conditions. Increased LDMC reduces water loss by enhancing resistance to water diffusion from leaves to leaf surfaces (Zhu and Xu 2021). Thick and dense leaves offer protection to the photosynthetic apparatus against high air temperatures (Leigh et al. 2017; Liu et al. 2020; Slot et al. 2021; Tserej and Feeley, 2021), maintain internal leaf temperature stability (Liu et al. 2020), and mitigate the damage caused by high levels of ultraviolet radiation, particularly relevant at high altitudes (Ma et al. 2012) such as Bogotá (2600 m.a.s.l.). We did not find any differences in wood density among localities. However, previous studies have reported changes in xylem traits in tropical trees due to pollution (Gupta and Iqbal 2005; Rajput and Rao 2005; Iqbal et al. 2010; da Silva, de Vasconcellos, and Callado 2023). Considering that pollution, heat, and water deficit are often correlated in urban ecosystems (Ulpiani 2021), it is likely that the presence of shorter and narrower vessel elements in polluted areas could be linked to enhanced hydraulic safety, reducing the risk of cavitation under water deficit conditions (Hacke et al. 2001). The absence of differences in wood density among our localities is probably due to the relatively small gradient in precipitation and temperature values. Further research is necessary to fully understand the mechanisms underlying changes in wood anatomy under the environmental conditions of UHI.

Q. humboldtti exhibited a higher leaf pH in the warmer and polluted locality (L3), suggesting that it is a pollution-sensitive species. Previous research has shown that plants tolerant to SO₂ and NO₂ keep their stomata open when exposed to these pollutants. Open stomata allow a greater influx of contaminants into the leaves, resulting in a more acidic foliar pH compared to sensitive plants, which tend to have a higher pH due to early stomatal closure (Thambavani and Prathipa 2012; Patel and Kumar 2018). In addition, the increase in pH could also be due to specific pollutants such as cement dust, which is a component of PM₁₀ in urban areas (Zeng et al. 2010). Studies have shown that cement dust can enter leaves and dissolve into calcium hydroxide, raising foliar pH (Guderian 1986). In Bogotá, it has been reported that there is an increased contribution of enriched fugitive dust (resuspension of crustal material and soil dust) and secondary PM compared to vehicular emissions (Ramírez et al. 2019). Moreover, alkalization of pH can be a response mechanism to pollution stress, as it has been found to enhance the conversion efficiency from hexose carbohydrates to ascorbic acid, which is a natural antioxidant (Kamble et al. 2021). Additionally, other studies have reported an increase in leaf pH values (close to neutral) with increases in annual temperature (Liu et al. 2019), however, the specific mechanisms underlying these observations remain unclear.

It has been observed that plants tolerant to pollution generally have high RWC values (Kuddus, Kumari, and Ramteke 2011). Plants have the ability to assimilate, metabolize, and eliminate pollutants through deposition or translocation to leaves or other organs, but this process is highly dependent on the hydration level (Sharma, Sharma, and Bhardwaj 2017). The lack of significant variations in RWC among different localities suggests that Q. humboldtti is not tolerant to atmospheric pollution. Ascorbic acid is a natural antioxidant that helps maintain the stability of plant cell membranes and scavenges free radicals induced by different types of stress, such as pollution or drought (Dhindsa, Plumb-Dhindsa, and Thorpe 1981; Akram, Shafiq, and Ashraf 2017). Surprisingly, we did not find any differences in the levels of ascorbic acid between the localities with contrasting pollution, precipitation, and temperature conditions (L1 and L3). The lower levels of ascorbic acid in the colder and unpolluted native forest (L1) are likely due to the downregulation of genes involved in ascorbic acid production or the presence of basal levels (Li et al. 2013). Conversely, in the warmer and polluted locality, the reduction in ascorbic acid levels may be attributed to its consumption during the removal of free radicals generated in chain reactions following the penetration of air pollutants into the foliar tissue (Wolfenden and Mansfield 1990; Szymańska et al. 2017). The consumption of ascorbic acid suggests that Q. humboldtii is not a pollution-tolerant plant, as tolerant plants are able to maintain high levels of ascorbic acid even under high contamination levels (Keller and Schwager 1977; Pandey and Agrawal 1994; Rai 2016). The lack of changes in chlorophyll content among the different localities is likely due to the fact that ascorbic acid is primarily located in the chloroplasts (Keller and Schwager 1977), which helps protect their integrity from the action of free radicals, including those derived from atmospheric pollutants. Additionally, it assists in maintaining a stable pH, which is crucial for sustaining photosynthetic efficiency (Werdan, Heldt, and Milovancev 1975). The differences in the APTI were primarily centered in the most polluted localities (L2 and L3). This may be due to the notable differences in the concentration of ascorbic acid among individuals from both sites. Nevertheless, the decrease in APTI with increasing atmospheric pollution strengthens the notion that Q. humboldtii is susceptible to air pollution. It is important to emphasize that our study was conducted during a dry period, when there is an increased concentration of pollution in the environment and a reduced removal of particles from leaves and stems through wet scavenging (Wang et al. 2017; Vicente-Serrano et al. 2020). However, it is crucial to comprehend how trees respond to the temporal variation of pollution and climatic conditions.

Climatic changes expected to occur in tropical mountains in the coming decades will greatly impact human health, energy consumption, and nutrient cycling in cities. In light of these new climate scenarios, it is crucial to prioritize urban tree planning as a highly effective strategy for mitigating heat and pollution. Vegetation plays a critical role in cooling the environment and removing pollutants (Rosenfeld et al. 1998; Stone et al. 2014). Our study offers valuable insights into the functional traits that determine how species respond to high levels of pollution and warmer weather, which are caused by the urban heat island effect. Further research should investigate the responses of other tree species and consider a broader range of environmental variables. This step is necessary to fully understand the complex dynamics of species-rich urban areas.

Acknowledgements

We thank Claudia Durana, owner of El Silencio Reserve (San Francisco, Cundinamarca), who opened the doors to this project. Financial support was provided by Agencia Distrital para la Educación Superior, la Ciencia y la Tecnología (ATENEA) and Universidad Nacional de Colombia, who funded the project: “Los árboles como una solución basada en la naturaleza para regular el clima, reducir la contaminación y mejorar el bienestar de la población en la ciudad de Bogotá”, No. ATENEA-087–2023.

Author contributions

Miguel Angel Camargo (Conceptualization [supporting], Data curation [equal], Writing—original draft [equal], Writing—review & editing [equal]), Luz Marina Melgarejo (Conceptualization [equal], Investigation [equal], Methodology [equal], Supervision [equal], Writing—review & editing [equal]), Geisa Faerito (Methodology [equal], Writing—review & editing [equal]), Ingry Pérez (Methodology [equal], Writing—review & editing [equal]), and Beatriz Salgado-Negret (Conceptualization [lead], Data curation [lead], Formal analysis [lead], Funding acquisition [lead], Methodology [equal], Project administration [lead], Supervision [equal], Validation [equal], Writing—original draft [equal], Writing—review & editing [equal])

Conflict of interest statement

All authors declare that they have no conflicts of interest.

Data availability

The data used in this paper will be available in Dryad Digital Repository.

References