Urbanization, Electromagnetic Radiation and Sparrows: A Case Study from Guwahati, India

The persisting growth of wireless telecommunication technology causes increased electrosmog. Exposure to wide-ranging radiofrequency electromagnetic elds is thought to be a concern for all the living species across the globe. Studies have shown possible effects of electromagnetic radiation on various organisms including electromagnetic radiation risk on birds, but the outcomes were inconsistent. Here, we investigated if there is any impact of electromagnetic radiation on the abundance and habitat use of sympatric House Sparrow and Tree Sparrow in Guwahati City, India. In addition, we did a comparative analysis of previous work that had been conducted on possible effects of EMR on wild birds in eld condition. We collected sparrow abundance in selected urbanization gradients temporally over a period of two years in 45 locations and spatially covering 168 locations in the eastern part of Guwahati City. Point counts were carried out, and successively we measured the electromagnetic radiation along with other ecological covariates. It was observed that ecological variables contributed signicant variation to explain the habitat use of sparrows in Guwahati City compared to electromagnetic radiation. We observed that House Sparrow had quadratic relationship and Tree sparrow had negative association with increasing urbanization. Studies carried out in the past on the impacts of electromagnetic radiation on birds outside the lab condition had not considered other ecological covariates, which could also inuence the life history needs of the bird species. Our study emphasized that the ecological covariates should be taken into consideration while studying the effect of electromagnetic radiation on wild organisms.


Introduction
On a global scale, humans are shifting the composition of habitats, with ever-increasing human population. In recent years, the importance of incorporating the impact of human activities into the study of ecology is becoming more widely recognized (Bürgi et al. 2017). The incremental growth of wireless telecommunication technologies has causes increased electrosmog (Cucurachi et al. 2013). Exposure to wide-ranging radiofrequency electromagnetic elds (RF-EMF) shown as a concern for variety of species and groups across the globe. Studies have recognized possible effects on organisms (e.g., bees, fruit ies, frogs, birds, bats, and humans) with the increase of RF-EMF in the environment ( Rejt et al. 2007). As "birds are candidates for being good biological indicators for low-intensity electromagnetic radiation: they have thin skulls, their feathers can act as dielectric receptors of microwave radiation, many species use magnetic navigation, they are very mobile and possible psychosomatic effects are absent" (Balmori 2005). Balmori (2004) also stated that the urban birds that are subjected to electromagnetic contamination are more vulnerable to the negative impacts of RF-EMF.
Six species of sparrows Passer spp. have been reported from the Indian subcontinent (Ali et al. 1987). Out of them, the House Sparrow Passer domesticus and the Eurasian Tree Sparrow Passer montanus are chie y associated with man-made habitats and human activities (Summers-Smith 1988). In India, the distribution of the House Sparrow is widespread; however, the Eurasian Tree Sparrow, generally a northern species, has a small resident population in the Eastern Ghats too (Grimmett et al. 1999). House Sparrow and Tree Sparrow ll similar ecological niches (Summers-Smith 1988). House sparrows usually live in urban environment, where electromagnetic contamination is comparatively higher; for this reason, sparrows may be a useful biological indicator for detecting the effects of this radiation. Successively, the recent decline of House Sparrow worldwide thought to have the possible link with electromagnetic radiation (Balmori and Hallberg 2007). Some published report anticipated that sparrows tend to avoid places with high levels of electromagnetic signals (Balmori 2003). Moreover, small organisms are especially vulnerable as they possess thinner skulls, which may facilitate radiation penetration into the brain (Hyland, 2000). However, it was also questioned, "if the electromagnetic radiations are deleterious to birds; they should be so to a number of other animals that shares urban landscape with humans. Why must the house sparrow be vulnerable when crows, pigeons, owls, mynas, bats and geckos have not succumbed to electromagnetic radiation that is attributed telecommunication?" (Daniels 2008).
Henceforth, in the present study we seek to understand is there any impact of electromagnetic radiation on the abundance and habitat use of sympatric House Sparrow and Tree Sparrow in the Guwahati City, India. In India, the composition of telephone subscribers using wireless form of communication is 63.27% in urban area and 33.20% in rural area (TRAI 2012).We hypothesized that if House Sparrow abundance and distribution (as hypothesized by Balmori 2004) was in uenced by electromagnetic radiation that would have more or less similar consequences on Tree Sparrow too, since Tree Sparrow is ecologically similar to House Sparrow concerning body size and habitat use with highly overlapped zones within the limit of their distribution (Ali et al. 1987). Thus, we collected data on the abundance of Tree Sparrow along with House Sparrow and other associated urban bird species such as the House Crow Corvus splendens, Common Myna Acridotheres tristis, Feral Pigeon Columba livia domestica, Spotted Dove Spilopelia chinensis, Jungle Myna Acridotheres fuscus and Asian Pied Starling Gracupica contra. Balmori and Hallberg (2007) found that when the mean eld strength of electromagnetic radiation was greater than 3 volt/meter, sparrows declined drastically. Besides, they also mentioned that an apparently strong relationship between bird density and electromagnetic eld strength could be conducted in a more controlled study to test the hypothesis. However, studies that have been carried out on the possible effect of electromagnetic radiation on sparrow abundance/density had not considered other ecological factors that might also in uence the abundance/distribution of sparrows. Hence our speci c objectives were-a) to investigate is there any impact of electromagnetic radiation on the abundance of sparrows in the presence of other ecological variables; b) to nd out the association of House Crow, Common Myna, Feral Pigeon, Spotted Dove and Jungle Myna in relation to EMR; and c) to probe on the previous studies carried out on wild birds and possible effects of EMR outside laboratory condition.

Study Area
The study was carried out in Guwahati, which is the major and growing city in the north-eastern part of India (Alam, 2011). The human population of Guwahati Municipal Corporation Area in 2011 was 963,429 (Guwahati City Census 2011). The city is located on the southern bank of the Brahmaputra River, adjacent to undulating plain foothills of Meghalaya plateau, between 26°50' to 26°150' N and 91°350' to 91°550' E, and covers an area of 264 km 2 (Fig. 1a). Guwahati falls under the tropical monsoon climate and receives about 1,600 mm annual rainfall. The major habitat types of Guwahati comprised of forest patches (mostly hillocks), scrublands, plantations, agriculture and human settlements and Deepor beel, a Ramsar site. There are several commercial and residential nuclei (e.g., Paltan Bazar, Ganeshguri, Beltola, Sixmile, Khanapara, Maligaon) in the city. Paltan Bazar surrounded by Pan Bazar and Fancy Bazar is the central part of the city. This is the transportation hub and includes the Guwahati Railway Station, bus stations, stoppage of numerous private bus services, o ces, numerous hotels and restaurants, making it one of the crowded and congested areas of the city. On the other hand, Beltola area in the eastern part of the city sharing geographic boundary with the hills of Meghalaya, has been developing in a rapid manner since 1980s and now extends up to National Highway 37 (NH-37) in the extreme south of the city. The Beltola Bazar in the central part of Beltola is a bi-weekly fruits and vegetable market, which has been place of congregation of people during the market-days since ages. It is an important traditional trade point between the people from Khasi Hills (Meghalaya) and the local people of Guwahati. Beltola area forms a gradient of urbanization scale starting from highly urbanized area such as congested market places at the road sides to residential complexes to fellow land. Additionally, the old railway quarters in Maligaon and the veterinary staff quarters in Khanapara occupy a very small area of the Guwahati city, but have Assam type traditional housing structures, with predominantly tin roofs.

Methods
Initially, we collected sparrow abundance in selected habitat gradients (highly commercial/semi-commercial city centers, residential apartments/buildings, and Assam-type houses) over a period of time (temporal) to monitor the trend. Starling (APS). Point counts were carried out for ve minutes within 30 m radius circle in the morning between 0600-0900 hrs. Simultaneously, we measured the electromagnetic radiation for each location using EMF Detector (Model-AI195, Three-Axis RF Strength meter; frequency range 50 MHz ~ 3.5 GHz). We used the radiation measurements of the maximum average for ve minutes in the V/m unit. Subsequently, we also collected microhabitat information of 12 ecologically important (see Nath et al. 2019) variables on each sample location: distance to closest green patch-delineated the boundary of nearest green patch using Google Earth and measured the closest distance of the boundary of the patch from the center of the 30 m radius circle; (ii) green cover-digitized green cover (tree + grass cover) using Google Earth; (iii) grass cover-we used handheld laser range nder and Google Earth to estimate the percent grass cover; (iv) Plant diversity-counted the number of individuals of plant species (trees and sapling) and calculated Shannon (H´) diversity index (v) distance to closest marketplace-generated the center point of the open market places (open daily markets, congested areas with hotel restaurants, shopping malls, crowded places nearby busstops and railway station etc.) and measured the distance from the center point of 30 m radius circle; (vi) Food shops-number of hotels, restaurants, grocery shops within the 30 m radius. We also counted the (vii) number of rolling shutters people used for the front gate of shops and garage. Moreover, counted the number of (viii) single storied and (ix) multistoried house along with the roof type such as (x) tin & (xi) concrete within the 30 m radius. In addition, we also measured the (xii) area opening of drains using laser range nder, and then converted the area into percentage by dividing the area of 30 m radius circle.
Later, from November 2014 to March 2015, we also collected data on the spatial scale in the eastern part of Guwahati City, where the overlap of both sparrow species was higher compared to other parts of the city (Nath et al. 2019). A total of 168 locations were surveyed over the urban gradients and the number of House Sparrow and (avoid using ampersand in a running text, use word 'and') Tree Sparrow was counted, and the value of electromagnetic radiation was also measured using EMF Detector as mentioned above. Here, we classi ed the LISS-IV image (5 m resolution) of Guwahati city using ERDAS 9.1 to get the built-up cover, green cover (tree cover + grass cover), and the proportion of each class within the 100 m radius of the sampling location using ArcGIS 10.3.
In addition, we presented a systematic comparative review of the published scienti c studies on the potential ecological effects of electromagnetic elds on birds at the natural setup (i.e., outside laboratory condition). We excluded all studies on birds that had not been conducted in eld condition. Subsequently, we summarized the following: study design adopted, ndings and conclusive remarks these studies delivered.

Data analysis
Pearson correlation test was performed to nd out the association between EMR and abundance birds. For the temporal dataset, we conducted PCA for ecological covariates and components representing similar ecological gradients (degree of urbanization PC1: distance to nearest marketplace, number of food shops, number of rolling shutters; greenness PC2: overall green-cover, grass-cover percentage, distance to nearest green patch and plant diversity; house-type PC3: no. of single & multistoried house, no. of tin and concrete roof) (for details see Nath et al. 2019). Before performing PCA analysis, we transformed all the variables into Z-score. We used exploratory analysis such as graphs (scatter and box plots) and univariate regressions to choose the best among several indices corresponding to a single apriori hypothesis and settle on the appropriate function (linear or quadratic) before formal modeling. We built candidate regression models (Generalized linear model) using EMR and its additive and interactions with the PCA components − degree of urbanization, greenness, and house-type as an independent variable, and sparrow abundance as a dependent variable.
For the spatial dataset, we built candidate regression models with covariates -percent built-up cover and EMR, and interactions of built-up × EMR on sparrow abundance. Following Burnham and Anderson (2002), we computed Akaike-weight (Wi) of select candidate models to provide model-averaged regression coe cients, unconditional standard errors, and importance (summed Akaike-weights) of each predictor using MuMIN (Bartoń 2015) in program R v 3.2.5. Finally, we used R package "effects" (Fox and Hong 2009) to plot the interaction model.

Results
For the temporal habitat use dataset, both House Sparrow (r = -0.022, p = 0.88) and Tree Sparrow (r = -0.07, p = 0.64) found to have no association with EMR measures. Out of eight species of urban birds studied, signi cantly Spotted Dove (r = -0.31, p = 0.03) and Asian Pied Starling (r = -0.48, p = 0.001) showed negative association with increasing electromagnetic radiation (Figs. 2 & 3). The House Sparrow abundance was found to be in uenced by the degree of urbanization (Table 1; S1 Table).
Electromagnetic radiation solely was not found to have any effect on the House Sparrow abundance. EMR was found to have an additive effect on House sparrow numbers when modeled with the quadratic form of degree of urbanization (not signi cant; see Table 2; S1 Table). A similar trend was also observed for Tree Sparrow as EMR did not in uence the Tree Sparrow numbers.
Interaction of greenness and house type had a positive in uence, whereas the degree of urbanization had negative (see Tables 3   & 4). Among all other covariates, degree of urbanization was found to play a critical role in determining the number of sparrows in Guwahati City. House Sparrow had a quadratic relation, Tree Sparrow had a negative, whereas EMR was positively associated with the increasing urbanization (Fig. 4).  Key factors in uencing the habitat use using spatial dataset of sparrows are given in Tables 5 and 6 Table 7). The studies claimed the possible signi cant effect of RF-EMF on birds' occupancy, abundance, breeding density, reproduction, and environmental variables that can also in uence the habitat use of these studied birds (see Table 7). We have provided a detailed on the methodology adopted, ndings and conclusion drawn by the authors. Besides, we also commented and argued on the outcomes provided in relation to impact of EMR with published research works, and cite literature those who mentioned other ecological variables that could also in uence the population and habitat use parameters of the studied bird species. Signi cantly Blue tits occupied more nest boxes in high exposed areas near Radar station compared to Great Tits (p < 0.05).
No signi cant difference was noticed in the number of eggs, and number of nestling per box observed between tits occupying exposed and control nest boxes.
However, they concluded that radar station generally doesn't lead to decrease of number of tits, but may cause shifts in tit species living The study highlights that signi cantly Blue Tits occupied more nest boxes in high exposed areas near Radar station compared to Great Tits (p < 0.05) and mentioned that no signi cant difference was noticed in the number of eggs, and number of nestling per box observed between tits occupying exposed and control nest boxes. They also mentioned that sites where radiation level was higher were inhibited mainly by Blue Tits.

Discussion
The present study showed that electromagnetic radiation did not show any direct effect on the abundance and habitat use of sparrows in Guwahati city. Besides, the sparrows were found to be in uenced by other complex ecological variables (e.g., distance to the nearest market place, number of food shops, number of rolling shutter in shops, distance to nearest green patch, percent green cover and plant diversity, percent built-up area and percent green cover). However, the magnitudes and impacts of microhabitat variables mentioned above may vary among the speci c needs of the life history traits of both the sparrow species (for details see Nath et al. 2019). It was observed that on spatial and temporal scales, ecological variables contributed signi cant variation to explain the habitat use of sparrows in the study area.
In general, increasing urbanization has positive association with electromagnetic radiation. The present scenario in most of the cities around the world is clear-cut─ the area with higher developmental activity is also high in electromagnetic radiation because of the growing demand for telecommunication and mobile phone services. Henceforth, the model output of our spatial data set showed a complex interaction of percent built-up and EMR since the percent built-up area partially characterizes the degree of urbanization. The interaction model (House Sparrow ~ Built-up * EMR) showed a positive association, which implies that the sampling locations with more built-up and area with high EMR measures have a large number of House Sparrow sightings (Fig. 5). Balmori and Hallberg (2007) showed that the mean density of sparrows corresponds to the lowest electromagnetic eld intensity areas. Despite that, they also come up with the discordant statement that no sparrows would be expected to be found in an area with eld strength > 4V/m. However, we recorded House Sparrow in the places where EMR reached up to 7.81 V/m, and even we recorded 1-17 individuals of sparrows (5.0 Mean ± 0.87 S.E ) within 4.00-7.81 V/m range. On the other hand, Tree Sparrow preferred habitat with minimum urbanization (Zhang and Zheng 2010), and its number was signi cantly in uenced by greenness (% green cover, plant diversity) and type of housing structure (primarily residential apartment/building). It was found that EMR had no signi cant role in governing the habitat use of the Tree Sparrow; even at the different levels of EMR.
In our study, EMR was positively correlated with the degree of urbanization (r = 0.53, p = 0.0001) and negatively with greenness (r = -0.51, p = 0.0002). Subsequently, the negative association of Asian Pied Starling and Spotted Dove could be directly associated with the degree of urbanization. Both the species observed to occupy areas with low level of urbanization in the study area. Sivakumar et al. (2006) also found that the density of Asian Pied starling was more in village edge forest compared to other habitat types in Buxa Tiger Reserve, West Bengal, India. Therefore, solely based on the EMR measures the cause and effect relationship on avifaunal abundance may not be studied in the eld condition. Subsequently, our analysis showed that EMR has no direct impact on the distribution and habitat use of urban birds, which collectively depending on the microhabitat needs of the species.
Previously, several studies have examined electromagnetic radiation risk on birds, but outcomes were inconsistent, and contributing factors were incoherent (Cucurachi et al., 2013). Most of the studies carried out on the effect of electromagnetic radiation on birds were laboratory investigation-on chicken (Gallus domesticus) and Japanese quail (Coturnix coturnix subsp. japonica). Cucurachi  The study design and methods used in each of these above mentioned studies have been discussed in details in Table 7. None of these studies, considered possible interaction of electromagnetic radiation with other life history needs (ecological variables) of the studied species. Henceforth, nding out the reliable conclusion is di cult in studies without the other controlling factors (i.e. covariates). For urban birds, the sampling approach, and the intensity of urbanization and its complex association with EMR, could play a major role in the statistically signi cant cause and effect relationship due to sampling artifact, which has no relevance biologically. Apart from that, most of the studies were carried out in the laboratory showed signi cant negative impact of EMR. The duration of exposure of EMR was twice as high on average studies, which showed impact of the radiation than that of studies which did not show any effect (Cucurachi et al. 2013).
Moreover, studies that were carried out in the laboratory condition showed the signi cant negative impact of EMR. The duration of exposure of EMR was twice as high on average in studies, which showed the impact of the radiation than that of studies that did not show any effect (Cucurachi et al. 2013). They have also mentioned, the laboratory studies carried out in the past exposed targets to the high level of MW-EMF, which probably determined the uncontrolled rise in temperature affecting the exposed systems. The information and results on the effects of radiation gathered in laboratory studies may need to be guardedly handled due to the steep nature of the laboratory solutions adopted. The conditions applied in the laboratory studies, in fact, do not always re ect real conditions of exposure, and at times it is essential to carefully evaluate the plausibility that biological systems exposed to RF-EMF could likely translate into ecologically relevant effects.
The WHO con rms that to date the accepted health effects due to high dosage ascribable to RF-EMF are caused by temperature rise (van Deventer et al., 2011). On the contrary, in the present study in Guwahati city, and the other studies which have been carried out in the eld underexposure of such a low level of EMR (< 9 V/m) were unable to generate the thermal effect. The measured electric eld strength values were far below the required to produce heat as low as 0.5°C (i.e., 10 mW/cm² or ca. 194 V/m; Bernhardt, 1992). However, few studies on the non-thermal effects of electromagnetic radiation provide evidence about the hypothesis of non-thermal effects of electromagnetic radiation on the brain physiology (Cucurachi et al. 2013) Subsequently, they also found that about two-thirds of the reviewed studies on the ecological effects of EMR were reported at high as well as at low dosages. The very low dosages are compatible with actual eld situations and could be found under environmental conditions. However, the non-thermal effects on biological tissues still under investigation, although calcium e ux and free radical production are among the candidates of the possible mechanism responsible for the non-thermal effects of EMR. Then again, results are still not conclusive, and there is still some uncertainty about the low dosages and non-thermal effects applied in some studies which did nd an effect (Cucurachi et al. 2013).
In conclusion, the studies carried out on the impact of EMR on birds in eld condition, is limited. Moreover, the negative correlation strength reported by the previous studies between sparrow density and EMR may simply attribute to sampling artifacts rather than factual causation. In addition, large scale survey through questionnaire across India with the help of citizens did not able to make conclusion on any correlation in the decline of house sparrows with cellphone tower radiation (Rahmani et al. 2013). The present study also emphasized that the complexity of ecological processes involved in the cause and effect relationship should be understood in multidimensional means. Therefore, site covariates need to de ne and quantify along with EMR measures before concluding remarks on the impact of radiation emitted by cell phone towers.

Declarations Acknowledgment
We are thankful to Aaranyak for providing the necessary facilities during the eld surveys. We would like to thank Qamar Qureshi, Sutirtha Dutta, Sonali Ghosh and Ranjana Pal of Wildlife Institute of India in overall preparation of the manuscript. We are grateful to Minarul Haque, Samrat Sengupta, Nilutpal Mahnta for their help during the eld work.

CONFLICT OF INTEREST STATEMENT
Manuscript title: Urbanization, Electromagnetic Radiation and Sparrows: a case study from Guwahati, India. The authors whose names are listed immediately below certify that they have NO a liations with or involvement in any organization or entity with any nancial interest (such as honoraria; educational grants; participation in speakers' bureaus; membership, employment, consultancies, stock ownership, or other equity interest; and expert testimony or patent-licensing arrangements), or non nancial interest (such as personal or professional relationships, a liations, knowledge or beliefs) in the subject matter or materials discussed in this manuscript. Author names: Anukul Nath, Hilloljyoti Singha, Bibhuti P Lahkar.
FUNDING: The rst author received fund from Aaranyk-rufford seed grant to carry out the eld activity.    Relationship among sparrows and EMR with increasing degree of urbanization