Sources, distribution and variability of airborne trace metals in La Plata City area, Argentina C. Bilos, J.C. Colombo *, C.N. Skorupka, M.J. Rodriguez Presa Laboratorio de QuõÂmica Ambiental y BiogeoquõÂmica, Facultad de Ciencias Naturales y Museo, Universidad Nacional de La Plata, Av. Calchaquõ km 23,500, 1888 Florencio Varela, Buenos Aires, Argentina Received 14 July 1999; accepted 10 December 1999 ``Capsule'': Exhaust from vehicles appeared to be the major source of metal aerosols in the La Plata, Argentina region. Abstract Airborne particulate trace metals have been measured bimonthly during day and night hours in four permanent stations located in residential, industrial and commercial sectors of La Plata City region, to characterize the sources and variability of atmospheric inputs. Airborne trace metal regional averages (Pb 64�62, Cu 30�27, Mn 26�20, Zn 273�227, Fe 1183�838, Ca 5343�3614, Mg 1472�967, Cr 4.3�2.4, Ni 3.2�3.5 and Cd 0.41�0.42 ng/m3) are comparable to the values reported for not grossly polluted cities and below the general trend described for urban particulates. Two- and three-way analysis of variance and variance compo- nents tests (P<0.05) were performed to assess the contribution of the diurnal (day vs. night), spatial (inter-station) and temporal (inter-month) components of variability. Trace metal concentrations followed the behavior of total suspended particles with higher concentrations during the day and at the Downtown station and lower at the Residential site. In general, spatial and temporal variations prevailed over diurnal di�erences. Spatial di�erences were clearly most signi®cant for Pb, which presented higher values at the Downtown site re¯ecting the importance of motor exhaust inputs. In contrast, diurnal di�erences were more important for Mn due to increased dust resuspension during day hours. A seasonal trend with concentrations usually increasing during winter months and decreasing in spring±summer was also detected. Enrichment factors (EF) were calculated to evaluate anthropogenic versus natural element sources. High EF (21±376) were obtained for Pb, Zn, Cd and Cu re¯ecting the importance of anthropic inputs. The comparison with EF calculated for high-emitting vehicle particle emissions indicated that motor exhausts are the most important source of these elements in La Plata region. In contrast, the EF calculated for Mn, Cr, Ni, Ca and Mg were low (1.3±7.5) suggesting chie¯y natural sources, i.e. soil-derived dusts. # 2000 Elsevier Science Ltd. All rights reserved. Keywords: Air pollution; Heavy metals; Airborne particulate matter; Enrichment factors; Urban air quality 1. Introduction Man-induced mobilization of trace metals into the biosphere has become an important process in the glo- bal geochemical cycling of these elements (Nriagu and Pacyna, 1988). This e�ect is most evident in urban areas where several stationary and mobile sources (industrial activities, energy production, construction, urban waste treatment, vehicle exhausts) release large quantities of trace metals into the atmosphere, soils and aquatic eco- systems very often exceeding the natural emission rates. The atmosphere, in particular, is a key compartment in the global geochemical cycling of trace elements (Lantzy and Mackenzie, 1979). The main inputs of trace metals to the atmospheric cycle are strongly related to particle emission processes. For most of the toxic trace metals (e.g. Pb, Cd, Zn, Ni, Cu) anthropogenic inputs are more important than natural sources such as continental dust, volcanic dust and gas, sea spray, and biogenic particles (Bertine and Goldberg, 1971; Nriagu, 1979, 1989). From a toxicological perspective, the airborne parti- culate material has important health implications, basi- cally through the inhalation of small particles with diameters of 10 mm or less which can be absorbed in the alveolar region of the lung (Hileman, 1981). It is well established that these inhalable particles have higher concentrations of many trace elements, such as Pb, Cd, Zn, Cr, Ni, Mn and Cu (Natusch et al., 1974; Hlavay et al., 1992), and thus, are considered of major health concern. However, some sources have been reported to have emissions highly enriched in toxic metals without a 0269-7491/00/$ - see front matter # 2000 Elsevier Science Ltd. All rights reserved. PI I : S0269-7491(99 )00328-0 Environmental Pollution 111 (2001) 149±158 www.elsevier.com/locate/envpol * Corresponding author. Tel.:/fax: +54-11-4275-8266. E-mail address: colombo@museo.fcnym.unlp.edu.ar (J.C.Colombo). signi®cant increase in PM10 levels (Sweet et al., 1993). Thus, from a biogeochemical perspective, characteriza- tion of the total suspended fraction is relevant to iden- tify the sources and variability of the airborne material. In this study we determine the concentration of selec- ted trace elements in airborne particles from La Plata region in order to: (1) assess the heavy metal con- centration status of the region; (2) characterize the principal sources of atmospheric particulate trace metals; (3) characterize diurnal versus nocturnal trends; and (4) evaluate spatial and temporal variations. 2. Materials and methods 2.1. Study area and sampling The study area is centered around La Plata City, capital of the Buenos Aires state situated 15 km away from the Rõ o de la Plata Estuary coast. It includes the nearby cities of Berisso and Ensenada and comprises a total population of about 1,000,000 people and a heavy industrial sector oriented to petrochemical activities. In order to characterize the contamination status and the most important emission sources of airborne parti- culate metals, four permanent sampling stations were established in a 25-km-long NE±SW transect passing through the petrochemical zone and La Plata City cen- ter (Fig. 1). The stations were located in La Plata Port (Port), in the limit of the petrochemical area between Ensenada and La Plata cities (Petrochemical), in La Plata City center (Downtown), and in a less urbanized area 12 km away from La Plata (Residential). Total suspended particles (TSP) were collected at each station for seven to nine samplings carried out during the year 1993 (January, February, March, April, May, July, August, September, December), using two General Metals Works BM 2200X portable high volume sam- plers (HVS). For security reasons, the HVS were located �5 m above the ground. To establish day±night di�er- ences, the HVS were operated for 48 h in two dis- continuous periods of 12 h, during the day (from 09:00 to 21:00) and night (from 21:00 to 09:00). During nor- mal operation the HVS pumped 631±3418 m3 of air. Due to electricity failures and incorrect operation of the sampling equipment, the data set is incomplete for some months and stations (especially for July and the Resi- dential site). TSP were collected on Whatman EPM 2000 borosilicate glass micro®ber ®lters (99.999% retention for 0.6 mm NaCl particles). TSP masses were determined by gravimetry, drying and weighing the ®lters in glass tubes (40�C to constant weight). Dried ®lters were cut into two equal halves, re-weighed, and frozen (ÿ20�C ) until analysis of trace metals and hydrocarbons (Colombo et al., 1999). 2.2. Chemical analysis The ®lters for trace metals were extracted in glass tubes with 25 ml of concentrated hydrochloric acid± nitric acid±deionized water mixture (5:3:2) in a hot water bath (20 min at 70�C), followed by ultrasonic treatment (20 min at 30±40 kHz working frequency and 300 W ultrasonic power). This extraction scheme was repeated ®ve times plus an additional extraction with deionized water. All the extracts were pooled and con- centrated to 10 ml at 60�C in polyethylene ¯asks. The extracts were centrifuged (20 min at 4000 rpm) and solid potassium chloride was added as a modi®er. The extraction e�ciency was tested by re-extracting two previously treated ®lter-samples from each station. The recovery was null for Cd, Ni, Fe and Zn, averaged 1% for Pb and Cu, and 11% of the ®rst extraction for Cr. Extracts were brought to a ®nal volume of 25 ml with deionized water and stored at room temperature in polyethylene bottles until analysis. Trace metal analysis was performed with a Perkin Elmer 3110 Atomic Absorption Spectrometer and an air±acetylene ¯ame with a Deuterium Lamp back- ground correction to overcome molecular interferences. The instrument operating conditions and general set- tings for the di�erent elements (¯ame stoichiometry, slit Fig. 1. Study area and station location. Major roads and parks (open circles and rectangles) of La Plata City are indicated. Sampling sta- tions are shown as ®lled circles. 150 C. Bilos et al. / Environmental Pollution 111 (2001) 149±158 height and width, lamp current and wavelength) were de®ned according to the optimum recommended for the instrument. Determinations were done with Perkin- Elmer (Cu, Fe, Pb, Zn) and ISTC (Cd, Cr, Mn, Ni) hollow cathode lamps; Ca and Mg were determined by ¯ame emission analysis. Working standard solutions were prepared with high-purity Johnson Mathey PLC original standard materials. Multi-element standard solutions were made for Cd, Cu, Mn, Ni, Pb, and Zn; two-element standard solutions for Ca and Mg, and single-element standard solutions for Cr and Fe. Filter and reagent blanks were processed following the sample treatment. The average metal content of the blanks for Pb, Mn, Cu and Zn represented less than 5% of the samples average metal content. For the other metals the blank contents averaged 9% up to 41% for Cr. The analytical detection limits were 0.012 mg/ml for Cd, 0.020 mg/ml for Zn, 0.022 mg/ml for Cu, 0.041 mg/ml for Mn, 0.055 mg/ml for Cr, 0.065 mg/ml for Ni, 0.068 mg/ml for Fe, and 0.081 mg/ml for Pb. 3. Results and discussion 3.1. Trace metal concentrations in La Plata area Table 1 shows the concentrations of airborne trace metals at each station and the regional average for La Plata area compared with other world cities. In general, most airborne trace metals in La Plata area are lower than the average values reported for urban particulates (Lantzy and Mackenzie, 1979). La Plata Pb average (65�62 ng/m3) is 12 times lower than the urban par- ticulates average (790 ng/m3), similar to the values reported for Birmingham and Bondville, a rural site in Illinois. La Plata average concentration for Pb is well below the national regulation (2 mg/m3, 30 min average). Other cities in USA or Greece show much higher Pb levels (400 to >1000 ng/m3) whereas the background values reported for Antarctica (0.07±5.4 ng/m3) are sev- eral orders of magnitude lower comparable to the atmospheric continental USA baseline (8.0 ng/m3; Chow et al., 1972). The concentrations of Cu, Mn, Cr, Ni and Cd in La Plata area are four to 10 times lower than those repor- ted for urban particulates, comparable to the values registered in Birmingham and Sindos; Antarctica back- ground levels are two to four orders of magnitude lower. Fe follows broadly the same pattern; the con- centrations in La Plata air (1183�838 ng/m3) are below the general urban average and similar to the values reported for Chicago, St. Louis and Mallipo but are higher than the levels registered in Birmingham. The sole heavy metal which appears to be enriched in La Plata air is Zn, with levels (273�227 ng/m3) more comparable to the average urban particulates (359 ng/ m3). The values are similar to those of Mallipo, Bir- mingham and many USA cities, whereas Beijing and Grecian cities show higher Zn concentrations. Ca is another major element which appears to be enriched in La Plata area relative to other world cities; the average concentration (5343�3614 ng/m3) is �10 times higher than those reported for Chicago, St. Louis and Bir- mingham (171±1918 ng/m3). This high Ca abundance is probably related to the composition of the carbonac- eous soil-source material (loess and loess-like sediments with CaO contents of 2.8±3.1% of total oxides; Cami- lio n, 1993). In summary, the concentrations of airborne metals in La Plata area are relatively low, similar to the values reported for not grossly polluted cities and below the general trend described for urban particulates, probably re¯ecting the major administrative and trade-oriented character of the city. Zn and Ca are the sole elements showing some degree of enrichment relative to other world cities. 3.2. Diurnal, spatial and temporal variability of trace metal concentrations Table 2 presents trace metal and TSP concentrations during day and night for each station. The general mean concentrations of non-detected elements were calculated using one-half of the detection limit. The data evidence a considerable degree of variability including diurnal (day vs. night), spatial (inter-station) and temporal (inter- month) di�erences. Fig. 2 shows all these variability components. In general, as has been previously observed for airborne hydrocarbons (Colombo et al., 1999), trace metal concentrations tend to follow the behavior of TSP with concentrations usually higher during the day, par- ticularly at the Downtown site (squares in Fig. 2). Spatial di�erences are also evident in the graphs, with con- centrations usually higher at Downtown station and lower at the Residential site (triangles in Fig. 2). In addition to these diurnal and spatial variations, a seaso- nal trend can also be observed; the concentrations increase during April±May±August (autumn±winter) and decrease in September±December (spring±summer). In addition to these general trends associated with TSP behavior, some metals such as Cu and Zn display a dif- ferent pattern, e.g. no day±night di�erences or decreas- ing concentrations throughout the year. In order to more precisely assess the contribution of these de®ned sources of variation, two- and three-way analysis of variance (ANOVA) and variance components tests (P<0.05) were performed for the whole log-transformed data set (excluding July, due to incomplete sampling). 3.2.1. Diurnal versus spatial variability Fig. 3 shows the percentage of the total variability explained by spatial and diurnal di�erences considering C. Bilos et al. / Environmental Pollution 111 (2001) 149±158 151 T a b le 1 P a rt ic u la te el em en t co n ce n tr a ti o n s in a ir o f L a P la ta co m p a re d w it h o th er u rb a n a n d re m o te a re a s a ro u n d th e w o rl d C o n ce n tr a ti o n (n g /m 3 ) R ef er en ce s P b C u M n Z n F e C a M g C r N i C d L a P la ta a re a P o rt R a n g e 9 .2 ± 1 3 5 8 .4 ± 1 0 0 6 .8 ± 9 0 5 .1 ± 6 8 9 4 6 7 ± 2 3 1 9 2 9 2 3 ± 1 7 7 4 2 3 8 1 ± 4 9 5 4 2 .5 ± 8 .3 < 1 .0 ± 7 .2 < 0 .1 7 ± 1 .3 T h is st u d y P et ro ch em ic a l R a n g e 9 .5 ± 1 5 2 4 .5 ± 7 6 7 .4 ± 7 3 1 7 ± 6 9 5 3 6 9 ± 1 6 6 9 2 0 1 4 ± 1 2 3 0 6 4 3 7 ± 4 5 3 3 0 .8 ± 7 .2 0 .7 0 ± 1 6 < 0 .1 1 ± 1 .4 T h is st u d y D o w n to w n R a n g e 4 4 ± 2 6 8 8 .9 ± 7 3 8 .8 ± 9 2 2 0 ± 1 0 4 9 7 4 7 ± 5 9 6 7 3 4 9 4 ± 1 5 7 4 6 6 9 6 ± 3 1 5 8 3 .5 ± 1 2 < 1 .0 ± 1 5 < 0 .1 6 ± 2 .0 T h is st u d y R es id en ti a l R a n g e 2 .0 ± 1 0 1 7 .6 ± 1 6 3 4 .1 ± 3 1 3 4 ± 6 5 8 1 7 8 ± 1 4 9 5 3 9 4 ± 7 4 8 9 5 4 4 ± 2 6 7 4 0 .6 5 ± 7 .9 < 1 .1 ± 5 .2 0 .1 3 ± 1 .2 T h is st u d y G ra n d m ea n � S .D . 6 4 .5 � 6 1 .8 2 9 .5 � 2 7 .3 2 5 .5 � 1 9 .7 2 7 3 � 2 2 7 1 1 8 3 � 8 3 8 5 3 4 3 � 3 6 1 4 1 4 7 2 � 9 6 7 4 .3 2 � 2 .3 9 3 .1 5 � 3 .5 2 0 .4 1 � 0 .4 2 T h is st u d y U rb a n a re a s G en er a l a v er a g ea 7 9 0 1 1 0 1 4 9 3 5 9 3 7 1 0 ± ± 3 2 3 0 2 L a n tz y a n d M a ck en zi e (1 9 7 9 ) W a sh in g to n D .C . (U S A )b A v er a g e 1 4 2 0 ± ± 1 5 0 ± ± ± ± ± 3 .5 G re en b er g (1 9 9 0 ) N ew Y o rk C it y (U S A )b A v er a g e 1 2 2 0 ± ± 3 2 0 ± ± ± ± ± 7 G re en b er g (1 9 9 0 ) B o st o n (U S A )b A v er a g e 1 4 0 0 ± ± 3 4 0 ± ± ± ± ± 2 G re en b er g (1 9 9 0 ) C h ic a g o (U S A )b A v er a g e 1 5 0 0 ± ± 5 9 0 ± ± ± ± ± 6 G re en b er g (1 9 9 0 ) S t. L o u is (U S A )b A v er a g e 4 0 0 ± ± 2 4 0 ± ± ± ± ± 1 5 G re en b er g (1 9 9 0 ) N o rt h w es t In d ia n a (U S A )b A v er a g e 1 7 0 0 ± ± 2 7 0 ± ± ± ± ± 1 2 G re en b er g (1 9 9 0 ) B ei ji n g (C h in a )c R a n g e ± ± 7 3 ± 6 8 0 6 2 ± 1 7 0 0 4 3 0 0 ± 2 3 0 0 0 ± ± 1 0 ± 7 7 ± ± G a o et a l. (1 9 9 2 ) M a ll ip o (S o u th K o re a )c R a n g e ± ± 1 3 ± 1 2 0 6 .5 ± 3 5 0 4 3 0 ± 4 3 0 0 ± ± 1 .3 ± 1 4 ± ± G a o et a l. (1 9 9 2 ) Il li n o is (U S A )d B o n d v il le si te A v er a g e 2 2 .4 5 .4 1 0 .1 2 8 .5 2 3 3 4 4 2 ± 1 .9 1 .2 < 4 S w ee t et a l. (1 9 9 3 ) E a st S t. L o u is A v er a g e 2 0 9 1 3 8 2 4 .3 2 3 1 6 6 6 1 9 1 8 ± 5 .7 3 .9 2 5 S w ee t et a l. (1 9 9 3 ) S o u th ea st C h ic a g o A v er a g e 1 2 7 1 5 .4 8 6 1 4 8 1 1 8 5 1 0 8 8 ± 1 1 .2 4 .8 < 4 S w ee t et a l. (1 9 9 3 ) T h es sa lo n ik i (G re ec e) e Io n ia R a n g e 4 6 ± 4 2 0 9 .0 ± 2 1 0 1 9 ± 3 4 0 2 5 0 ± 3 2 0 0 ± ± ± 0 .0 1 ± 9 .9 ± 0 .8 6 ± 6 .6 V o u ts a a n d S a m a ra (1 9 9 6 ) S in d o s R a n g e 7 0 ± 1 0 2 0 3 1 ± 9 3 1 7 7 ± 4 2 0 0 6 1 0 ± 4 7 0 0 ± ± ± 3 .2 ± 2 4 ± 0 .7 6 ± 7 .8 V o u ts a a n d S a m a ra (1 9 9 6 ) B ir m in g h a m (U K )f R a n g e 6 9 ± 1 1 3 1 2 ± 6 6 1 0 ± 2 3 6 4 ± 6 4 1 2 5 4 ± 3 4 8 1 7 1 ± 2 4 5 ± 7 .1 ± 1 8 2 .2 ± 7 .4 ± H a rr is o n et a l. (1 9 9 6 ) R em o te a re a s A tl a n ti c O ce a n g R a n g e 0 .1 0 ± 6 4 0 .1 2 ± 5 6 0 .0 5 ± 9 .7 0 .3 0 ± 1 5 4 3 .4 ± 2 4 0 ± ± 0 .0 7 ± 1 .1 8 .0 ± 1 2 0 .0 0 3 ± 0 .6 2 D u ce et a l. (1 9 7 5 ) (3 0 � N ± 4 3 � N ) V er o n et a l. (1 9 9 2 ) A n ta rc ti ca h R a n g e 0 .0 7 1 ± 5 .4 1 0 .0 2 5 ± 1 .1 7 0 .0 0 4 ± 0 .9 9 0 .0 1 8 ± 2 4 .8 0 .2 2 ± 4 6 .8 0 .1 5 ± 1 0 0 3 0 .3 2 ± 2 .0 0 .0 0 2 5 ± 0 .1 0 0 .0 3 ± 0 .0 6 0 .0 0 5 ± 0 .5 Z o ll er et a l. (1 9 7 4 ) L o u re ir o et a l. (1 9 9 2 ) R a È d le in a n d H eu m a n n (1 9 9 2 ) K h a n d ek a r et a l. (1 9 9 2 ) a G en er a l a v er a g e v a lu es o b ta in ed fr o m a re v ie w o f p u b li sh ed d a ta fr o m U S A a n d E u ro p ea n ci ti es . b G en er a l a v er a g e v a lu es fr o m so m e U S A im p o rt a n t ci ti es . c R a n g e v a lu es fo r 7 2 d a il y sa m p le s co ll ec te d d u ri n g A p ri l a n d M a y 1 9 8 9 . d A v er a g e v a lu es fo r 1 2 - o r 2 4 -h sa m p le s o f P M -1 0 p a rt ic le s co ll ec te d o v er a 2 -y ea r p er io d in a ru ra l si te a n d in tw o u rb a n /i n d u st ri a l a re a s (B o n d v il le si te , E . S t. L o u is a n d S E C h ic a g o , re sp ec ti v el y ). e R a n g e v a lu es fo r 2 4 -h sa m p le s (n = 2 8 ) co ll ec te d d u ri n g th e p er io d su m m er 1 9 9 3 ± su m m er 1 9 9 4 in tw o re si d en ti a l a re a s. f R a n g e o f m ea n v a lu es fo r a to ta l 5 5 d a il y 2 4 -h sa m p le s co ll ec te d d u ri n g w in te r (J a n u a ry 2 ± F eb ru a ry 2 8 ) a n d su m m er (J u ly 2 7 ± A u g u st 2 3 ) 1 9 9 2 . g R a n g e v a lu es fo r a er o so l sa m p le s co ll ec te d o n -b o a rd sh ip d u ri n g cr u is es in th e N o rt h A tl a n ti c. h R a n g e v a lu es fo r a er o so l sa m p le s co ll ec te d d u ri n g d i� er en t sa m p li n g p er io d s a n d ca rr ie d o u t o n -b o a rd sh ip a n d in A n ta rc ti ca st a ti o n s. 152 C. Bilos et al. / Environmental Pollution 111 (2001) 149±158 T a b le 2 C o n ce n tr a ti o n o f p a rt ic u la te el em en ts a n d to ta l su sp en d ed p a rt ic le s in a ir sa m p le d d u ri n g d a y a n d n ig h t in L a P la ta re g io n a P b (n g /m 3 ) C u (n g /m 3 ) M n (n g /m 3 ) Z n (n g /m 3 ) F e (n g /m 3 ) C a (n g /m 3 ) M g (n g /m 3 ) C r (n g /m 3 ) N i (n g /m 3 ) C d (n g /m 3 ) T S P (m g /m 3 ) D N D N D N D N D N D N D N D N D N D N D N J a n u a ry P o rt 1 0 .9 9 .2 1 8 .5 2 8 .3 8 8 .2 5 6 .8 5 2 3 2 6 8 9 4 6 7 7 3 7 3 5 4 4 5 1 2 9 7 4 6 2 0 6 5 3 .0 0 4 .5 3 1 .3 8 1 .1 8 0 .1 7 0 .4 1 4 2 .1 2 4 .7 P et ro ch em ic a l 4 0 .5 3 4 .5 9 .1 4 1 2 .0 1 7 .0 9 .5 3 2 8 6 5 8 5 1 5 9 6 1 0 1 4 3 0 1 0 3 9 6 3 7 5 8 2 4 2 8 3 .9 5 3 .0 9 < 1 .1 1 < 1 .1 3 0 .4 9 < 0 .1 8 3 9 .8 3 1 .2 D o w n to w n 2 0 5 1 2 4 2 6 .3 2 5 .4 6 7 .7 3 3 .5 1 0 4 9 4 5 7 5 9 6 7 1 9 1 7 9 3 2 4 6 3 4 4 2 6 2 1 1 7 6 1 5 .1 0 3 .9 2 1 0 .0 4 .5 1 0 .2 3 0 .2 5 1 5 0 6 7 .3 R es id en ti a l 5 .0 3 1 .9 9 7 .6 4 1 2 .1 1 4 .7 1 0 .7 6 5 8 3 3 5 1 1 5 5 1 1 0 7 6 1 4 3 9 4 1 1 0 1 5 6 6 0 .7 4 1 .0 9 < 1 .2 4 < 1 .0 8 0 .1 9 0 .1 7 3 4 .6 2 4 .3 F eb ru a ry P o rt 2 2 .2 1 7 .6 1 0 .7 9 .3 0 2 0 .9 2 1 .0 2 2 5 5 0 2 8 0 2 5 2 5 3 6 0 7 4 5 8 4 3 8 1 1 4 6 9 6 .4 2 5 .1 2 < 1 .1 3 < 1 .2 2 0 .2 8 0 .2 3 6 5 .9 4 9 .3 P et ro ch em ic a l 9 .4 7 2 2 .8 4 .5 2 8 .9 7 1 0 .2 7 .4 3 2 9 3 4 4 3 5 5 6 8 4 5 2 3 7 8 2 6 4 8 8 3 5 1 1 3 2 4 .6 2 2 .3 2 3 .1 9 3 .2 3 0 .2 9 < 0 .1 8 3 6 .2 2 8 .2 D o w n to w n 1 8 1 1 1 9 2 3 .1 1 8 .9 5 2 .9 2 3 .3 3 7 2 4 2 4 1 8 7 4 1 2 7 4 1 1 5 8 9 3 4 9 4 1 3 9 8 1 1 2 5 5 .2 7 3 .5 1 2 .1 7 < 1 .1 4 0 .3 1 0 .2 4 1 0 7 6 8 .2 R es id en ti a l 2 .3 7 1 1 .7 8 .6 4 1 4 .4 1 6 .7 1 0 .3 1 5 9 2 9 7 1 4 1 9 1 4 9 5 1 5 8 2 1 2 5 2 5 6 7 9 6 1 0 .6 7 0 .6 5 < 1 .1 2 < 1 .0 9 0 .1 7 0 .1 7 4 7 .6 2 7 .1 M a rc h P o rt 7 0 .6 7 1 .2 2 8 .1 3 5 .0 1 6 .6 1 5 .9 3 4 7 4 0 2 8 3 6 1 1 3 0 3 1 8 8 2 9 2 3 4 5 7 1 0 3 0 4 .5 5 4 .4 5 1 .3 7 1 .7 6 0 .5 3 0 .4 2 7 9 .9 6 3 .8 P et ro ch em ic a l 4 9 .2 7 9 .2 1 9 .4 2 8 .5 3 3 .1 3 2 .3 6 9 5 2 2 1 1 1 0 7 1 5 1 4 4 8 7 0 4 8 7 7 1 3 7 3 5 5 7 5 .4 9 7 .1 5 3 .3 8 5 .5 0 0 .3 7 0 .4 8 4 6 .0 6 9 .9 D o w n to w n 1 3 2 7 4 .9 2 6 .4 1 7 .8 3 1 .5 1 6 .6 2 6 8 4 3 8 2 8 4 7 1 7 2 8 6 8 0 5 3 6 3 9 1 0 7 5 1 4 9 7 6 .3 6 4 .1 1 < 1 .1 7 < 1 .1 0 0 .7 7 0 .3 7 7 7 .7 4 6 .1 R es id en ti a l 4 .5 7 8 .0 5 2 0 .9 2 1 .7 1 6 .7 9 .8 1 2 9 5 4 7 1 4 6 5 3 9 8 1 5 8 7 1 9 3 2 1 4 5 0 2 2 9 9 0 .7 1 0 .7 0 < 1 .1 9 < 1 .1 7 0 .1 8 0 .1 8 3 9 .3 2 3 .7 A p ri l P o rt 2 6 .0 2 4 .9 1 4 .5 1 5 .7 1 4 .7 1 3 .2 2 8 4 2 9 9 6 0 2 6 1 0 4 0 2 6 4 1 0 4 1 1 8 6 6 8 8 3 .2 4 7 .9 5 6 .1 3 2 .2 4 0 .3 2 0 .2 6 5 2 .8 4 4 .2 P et ro ch em ic a l 4 1 .6 5 2 .6 1 1 .7 2 1 .7 1 5 .4 1 0 .8 5 1 9 6 6 8 9 2 5 3 1 8 3 1 9 7 3 1 2 0 1 1 3 7 1 3 3 3 4 .5 8 0 .8 0 1 .6 9 2 .4 5 0 .3 6 0 .1 8 5 3 .2 3 5 .5 D o w n to w n 2 3 1 7 9 .1 4 2 .5 2 2 .6 5 3 .9 1 6 .3 2 1 7 6 1 .4 2 8 4 4 1 3 0 1 1 3 2 0 2 5 5 5 8 2 5 4 9 6 9 6 7 .2 7 3 .5 1 7 .7 3 3 .4 9 0 .5 7 0 .1 7 1 4 7 5 3 .2 R es id en ti a l 4 5 .8 1 0 1 1 8 .7 3 5 .2 3 0 .6 1 9 .3 1 2 2 4 1 4 1 0 0 8 9 1 5 4 0 3 4 3 1 1 1 5 4 4 9 8 9 4 .2 8 4 .9 9 1 .2 3 5 .2 2 1 .2 0 < 0 .1 8 7 7 .2 6 7 .1 M a y P o rt 4 8 .0 9 0 .1 2 1 .5 4 2 .6 2 5 .9 1 7 .9 1 3 8 1 7 4 1 4 2 2 1 3 0 9 4 6 8 9 3 5 9 6 8 5 2 1 0 7 4 4 .3 9 5 .2 9 4 .2 0 7 .2 5 0 .2 7 0 .9 9 6 1 .5 7 4 .8 P et ro ch em ic a l 7 0 .0 6 3 .4 3 0 .2 1 6 .6 1 7 .0 1 0 .7 1 0 5 6 0 .8 5 3 2 4 5 3 3 1 9 5 2 0 1 4 6 8 2 4 3 7 0 .7 5 1 .8 5 2 .3 6 0 .9 1 0 .4 8 0 .3 0 5 4 .5 4 4 .7 D o w n to w n 1 8 1 6 8 .3 5 4 .5 9 .6 7 4 8 .3 8 .8 4 3 9 1 1 4 6 2 2 5 2 1 1 5 8 1 5 7 4 6 5 9 7 4 2 4 1 6 8 0 4 1 1 .8 3 .7 3 6 .2 1 < 1 .0 3 0 .3 4 < 0 .1 6 1 2 2 3 4 .3 R es id en ti a l 4 4 .8 3 3 .6 7 9 .4 1 6 3 2 1 .7 4 .0 5 4 6 1 6 4 2 6 0 1 3 8 8 6 7 8 6 2 8 9 7 1 7 5 0 1 4 2 8 7 .6 8 7 .9 0 < 3 .3 5 < 3 .8 0 0 .5 2 0 .5 9 7 9 .2 3 9 .9 J u ly P o rt 1 0 0 3 1 .8 3 3 .4 3 3 .0 2 0 .1 9 .9 9 7 8 .2 2 6 .8 7 5 0 7 4 1 9 2 0 2 4 1 1 5 3 9 6 4 1 9 4 4 4 .6 8 3 .4 4 2 .3 6 3 .7 4 0 .5 4 < 0 .1 7 7 2 .0 3 2 .0 P et ro ch em ic a l 2 6 .5 9 .5 1 1 1 .3 1 6 .6 3 6 9 5 6 6 7 2 9 0 4 2 .1 5 < 1 .1 6 0 .2 7 3 0 .4 D o w n to w n R es id en ti a l A u g u st P o rt 1 3 5 1 3 3 5 3 .2 9 9 .9 9 0 .3 6 1 .5 1 8 5 6 1 4 2 3 1 9 1 7 4 8 1 7 7 4 2 8 4 3 4 4 9 5 4 2 9 5 0 8 .3 4 5 .2 3 3 .8 5 6 .1 8 0 .9 6 1 .2 6 1 6 2 1 0 5 P et ro ch em ic a l 1 3 8 1 5 2 7 5 .8 6 4 .9 7 3 .1 3 7 .0 1 8 6 1 3 2 1 6 6 9 1 2 2 5 1 2 3 0 6 1 1 4 9 9 4 0 9 3 4 5 3 3 5 .5 5 4 .4 6 1 6 .3 7 .6 9 1 .4 2 1 .3 2 1 6 2 1 1 0 D o w n to w n 2 6 8 1 6 5 7 2 .8 5 7 .5 9 2 .0 3 9 .2 2 8 1 1 3 1 1 4 2 6 9 5 7 9 2 9 5 8 2 2 3 3 1 5 8 2 4 8 3 1 1 .6 7 .1 5 1 2 .5 1 5 .1 1 .9 8 1 .7 5 2 1 9 1 0 5 R es id en ti a l 2 4 .4 2 4 .1 6 9 .1 3 8 .6 3 1 .3 6 .2 7 5 2 .2 3 4 .3 1 0 3 3 1 7 8 7 4 8 9 3 4 7 6 2 6 7 4 1 5 0 3 3 .6 0 2 .1 3 1 .0 6 1 .2 4 < 0 .1 7 0 .1 3 8 1 .1 2 3 .3 ( T a b le co n ti n u ed o n n ex t p a g e) C. Bilos et al. / Environmental Pollution 111 (2001) 149±158 153 the whole data set. In general, spatial di�erences are more important than diurnal variations. Overall, spatial di�erences explain 24±67% (Cd±Pb) of the total var- iance, whereas diurnal di�erences account for 0.5±35% (Ni±Mn) of the total variability. Spatial di�erences are clearly most signi®cant for Pb. This trend is almost exclusively determined by the higher values of the Downtown site. This station pre- sents signi®cantly (S±N±K method, P<0.05) higher Pb concentrations (mean=136�64 ng/m3) compared to the rest of the stations (53.3�40, 47.2�42 and 23.7�27 ng/m3, for Petrochemical, Port and Residential, respec- tively). This di�erence, that follows the trend of TSP levels, re¯ects the well-established relation between atmospheric Pb concentrations and motor exhaust inputs due to the use of leaded gasoline (Chow et al., 1972; Eisenreich et al., 1986; Veron et al., 1992). Diurnal di�erences are more important for Mn re¯ecting the predominant natural soil source of this element and the increased dust resuspension during the most active day hours (see next section). 3.2.2. Diurnal versus temporal variability The evaluation of the diurnal and temporal vari- abilities by station evidence di�erent patterns (Fig. 4). Overall, at Residential, Petrochemical and Port sites the temporal variation is more important, whereas at Downtown, diurnal di�erences prevail. At the Residential site, the total variance explained by temporal di�erences ranges from 69 to 96% (Pb±Cr; P=0.005); for Mn, Fe, Mg and Cd the statistical results are not signi®cant. Diurnal variations are signi®cant only for Mn (54%, of the total variance explained; P<0.05). At Petrochemical station, temporal variations explain 58±89% (Cd±Ca; P<0.05); Pb, Fe and Cr tem- poral di�erences are not signi®cant (P<0.05). Diurnal di�erences at this station are only signi®cant for Cd and Mn (22±25%, of total variation, respectively; P=0.005). At Port site, temporal di�erences explain from 79 to 97% of the total variation for Ca and Pb, respectively (P<0.02). Mg and Cr temporal di�erences are not sig- ni®cant (P<0.05). At this site, all the elements showed non-signi®cant (P<0.05) diurnal variations. In contrast with the temporal-dominated pattern pre- viously described, at Downtown site, diurnal variations are more important, accounting for 21±72% of the total variation (Ni±Ca; P<0.005). For Cu, Zn and Mg, the di�erences are not signi®cant (P<0.05). Temporal dif- ferences prevail for Cd, Ni and especially for Zn (45± 68%; P<0.005). The higher incidence of the diurnal variability at the Downtown site was also observed for airborne hydro- carbons (Colombo et al., 1999) re¯ecting the dominant role of mobile sources, and the generally higher activity during the day (construction, material handling, etc.), which in association with higher temperatures and lowerT a b le 2 (c o n ti n u ed ) P b (n g /m 3 ) C u (n g /m 3 ) M n (n g /m 3 ) Z n (n g /m 3 ) F e (n g /m 3 ) C a (n g /m 3 ) M g (n g /m 3 ) C r (n g /m 3 ) N i (n g /m 3 ) C d (n g /m 3 ) T S P (m g /m 3 ) D N D N D N D N D N D N D N D N D N D N D N S ep te m b er P o rt 1 9 .7 1 1 .8 1 4 .1 1 1 .3 5 5 .3 1 3 .4 2 9 .1 3 5 .9 1 5 2 9 8 5 7 1 1 1 0 5 5 6 4 5 1 4 9 6 1 1 2 0 4 .4 6 3 .0 2 2 .3 9 1 .3 6 < 0 .1 7 0 .1 6 8 1 .5 2 9 .7 P et ro ch em ic a l 6 2 .0 2 5 .0 2 2 .5 1 2 .6 2 8 .7 1 3 .4 5 4 .0 1 9 .5 7 3 9 1 0 8 2 3 7 3 2 3 2 2 7 5 8 4 9 7 9 2 .3 6 4 .3 8 5 .2 3 7 .9 4 0 .6 6 0 .2 0 5 1 .3 5 4 .2 D o w n to w n 1 3 9 4 4 .0 2 4 .9 8 .9 1 3 3 .7 1 1 .9 7 8 .5 2 0 .0 1 2 6 0 7 4 7 9 9 1 2 3 8 4 4 1 4 5 7 1 4 5 7 5 .6 0 3 .6 9 3 .6 0 < 1 .0 4 0 .6 4 < 0 .1 6 9 4 .5 5 5 .0 R es id en ti a l D ec em b er P o rt 1 4 .3 1 3 .1 6 .2 5 1 6 .7 1 5 .5 1 0 .7 6 .9 7 5 .1 1 9 8 3 9 8 6 5 3 9 6 3 1 7 4 9 8 5 8 9 9 2 .9 6 2 .5 2 < 1 .0 9 < 1 .1 0 < 0 .1 7 0 .1 7 2 9 .2 2 7 .8 P et ro ch em ic a l 1 4 .8 2 4 .7 6 .9 2 2 1 .5 2 7 .2 1 2 .9 2 0 .8 2 9 .4 9 1 3 5 0 1 3 6 6 2 2 5 2 5 9 6 1 9 1 9 3 .7 6 2 .2 7 0 .7 0 0 .8 0 0 .1 1 < 0 .1 1 8 5 .4 3 7 .4 D o w n to w n 8 8 .9 7 8 .7 2 3 .6 3 7 .6 3 0 .6 2 9 .4 7 9 .9 4 4 .0 1 2 8 4 1 2 2 9 7 8 9 7 5 9 4 2 2 0 9 7 1 8 4 2 5 .8 6 4 .7 7 3 .3 2 3 .1 9 0 .2 7 0 .2 0 9 8 .6 1 0 1 R es id en ti a l 4 .3 5 2 0 .3 2 0 .6 6 2 .4 1 8 .0 1 7 .1 3 6 .3 3 4 .4 6 5 1 6 1 7 4 0 5 8 4 1 7 9 9 0 9 9 7 4 3 .3 8 3 .3 3 < 1 .0 7 < 1 .1 3 < 0 .1 7 0 .2 2 3 6 .2 3 7 .2 a T S P , to ta l su sp en d ed p a rt ic le s; D , d a y ; N , n ig h t. 154 C. Bilos et al. / Environmental Pollution 111 (2001) 149±158 ambient humidity favors dust resuspension. At this site, TSP concentrations during the day are signi®cantly higher (P=0.005) than night values (Fig. 2). The annual average increment of day versus night TSP, expressed as percentage of night values, is 92%. This clear TSP trend is also followed by Pb and Mn, which present sig- ni®cantly higher values during the day, e.g. 90 and 129% annual average day increment, respectively. Simultaneously to this TSP-related volumetric trend (concentrations in ng/m3), an opposite mass proportion pattern (concentrations in mg/g TSP mass) is also sug- gested. Although this trend has no statistical sig- ni®cance (P<0.05) for the whole data set, a trend of higher mass concentrations during the night is suggested for most of the metals. This is clearly evidenced by the annual average night increments expressed as percen- tage of day values, e.g. Pb 4±150%, Cu 23±175%, Zn 39±160%, Cd 9±70%, Ni 7±115%, Cr 15±61%, Fe 8± 33%, Ca 1±21%, Mg 36±77%. The only exception is Mn, which shows an opposite pattern, i.e. day mass Fig. 2. Concentrations of airborne trace metals and total suspended particles in the day and night air samples at the four stations and eight sampling months. Stations are identi®ed by diamonds (Port), circles (Petrochemical), squares (Downtown) and triangles (Residential). C. Bilos et al. / Environmental Pollution 111 (2001) 149±158 155 concentrations are higher than night values (7±32% annual average increment). Hydrocarbons showed a similar trend of night enrichment to most of the metals (Colombo et al., 1999). These di�erent patterns of the trace element mass concentrations (night vs. day enrichment) re¯ect the existence of two particle pools: a smaller size pool enriched in anthropogenic metals, chie¯y derived from gasoline combustion, which due to longer atmospheric residence time prevails during the night, and a larger-size, dust-derived rapidly sediment- ing particle pool more abundant during the day. Mn enrichment in the coarser day particles supports a pre- dominant soil-origin associated with mechanic resus- pension of street-dust (Sweet et al., 1993). 3.2.3. Overall diurnal, spatial and temporal variability Fig. 5 shows the results of a three-way ANOVA test performed to simultaneously evaluate all three sources of variation. Analyzing the percentage of the total variability explained by each source, four di�erent pat- terns can be identi®ed. TSP, Mn and Ca present a similar pattern, with temporal, spatial and diurnal dif- ferences of the same order of magnitude, explaining each 16±23% (P<0.001), for TSP, 13±31% (P<0.001), for Mn, and 8±34% (P<0.001), for Ca. Pb pattern is comparable to the previous group, but lacks signi®cant diurnal di�erences. Temporal and spa- tial di�erences dominate, explaining 16 and 52% (P<0.001), respectively, of the total variation. A third group is formed by Cu±Mg±Zn±Ni±Cd with a clear predominance of the temporal variability (26±69% Fig. 3. Diurnal versus spatial variance components test results for the whole data set. Fig. 4. Diurnal versus temporal variance components test results for the whole data set. 156 C. Bilos et al. / Environmental Pollution 111 (2001) 149±158 of the total variation explained, P<0.001). The spatial variability is signi®cant only for Cu and Ni (5.8±14%, P<0.05, respectively), whereas day±night di�erences are not signi®cant for all the metals (P<0.05). Fe and Cr are the last group, characterized by the prevalence of spatial di�erences, which explained 29% (P<0.001) of the total variability for both metals. Day± night variations are low, signi®cant only for Fe (8.5% of variation explained, P<0.05), and temporal variations are not statistically signi®cant (P<0.05). This descriptive classi®cation of the elements' behav- ior is di�cult to interpret on a source-base perspective, probably because of the large variability of the whole data set. In order to complement the interpretation based on variability components, in the next section enrichment factors (EFs) are calculated to more clearly identify the principal element sources. 3.3. Trace element EFs The calculation of trace metal EFs in airborne parti- cles, relative to soil or crustal abundances, has been used to evaluate anthropic versus natural sources (Zol- ler et al., 1974; Duce et al., 1975; RaÈ dlein and Heu- mann, 1992; Veron et al., 1992). In this study, within the elements which are commonly considered as reference for crustal material (e.g. Al, Sc, Ba, Fe), Fe was used to calculate the EFFes of the trace metals (Me), according to: EF � Me=Fe� �TSP Me=Fe� �crust ; where [Me/Fe]TSP and [Me/Fe]crust refer, respectively, to the TSP and mean crustal concentration ratios of the metal and Fe. The EFs were calculated on the basis of Earth's Crust mean abundances of the elements given by Taylor (1964). Trace element EFs include some degree of uncertainty related to the natural variations of the earth crustal composition. For this reason it is usually assumed that the EFs should be more than an order of magnitude higher than unity to suggest an anthropic origin. On the basis of the EFs calculated at the four stations, TSP collected at La Plata region appear as enriched in Pb, Zn, Cd and Cu, whereas Mn, Cr, Ni, Ca and Mg EFs are lower, compatible with prevailing natural sources. Pb shows the highest EFs, ranging from 8 to 860 with the highest values at Downtown (average=376) and lowest at Port and Residential stations (average=180). For Zn, the EFs oscillate between 4 and 1700, with highest values at Petrochemical and Residential sites (300±366, respectively) and lowest at Downtown station (113). In the case of Cd, the EFs range from 10 to 456 with rather homogeneous station averages (90±116). Cu EFs are 5±429 with highest values at the Residential site (84), perhaps re¯ecting the use of Cu as biocide in the numerous swimming pools of the area, and lower homogeneous values in the other sites (22±28). The high EFs calculated for Pb, Zn, Cd and Cu indi- cate that they are present in the airborne particles in concentrations too high to be explained in terms of normal crustal weathering processes. On a global basis, anthropogenic inputs of these elements predominate over natural sources accounting for 96, 66, 85 and 56% of the total emissions, respectively (Nriagu, 1989). Among these anthropogenic inputs, vehicle particle emissions are a relevant source. EFs calculated for par- ticles from high-emitting vehicles (Cadle et al., 1997) are 560, 550, 480 and 24, for Pb, Zn, Cd and Cu, respec- tively. These values are similar to those calculated for our TSP samples supporting the interpretation of vehi- cle emissions as the most important source of these ele- ments in La Plata region. The existence of two particle populations, a larger±low concentration diurnal one and a smaller±high mass concentration pool prevailing during the night, is also supported by the EFs which are higher in the night samplings (e.g. annual average night EFs increments expressed as percentage of day values=Pb 8±137%, Cu 8±178%, Zn 15±75%, Cd 15±38%). Pb, Zn, Cd and Cu are considered relatively volatile metals (Zoller et al., 1974; Duce et al., 1975) and, because they are mainly transported through the atmos- phere, they have been termed atmophile elements (Lantzy and Mackenzie, 1979). In contrast, Mn, Cr and Ni have been termed litophile elements because their masses are principally transported by streams (Lantzy and Mackenzie, 1979). As expected, the calculated EFs for these elements are not signi®cantly di�erent from unity (station mean range=1.3±3.1), indicating that their main source in airborne particles are soil- derived dusts. For these metals, natural emissions are very important and normally exceed anthropogenic Fig. 5. Overall diurnal, spatial and temporal variance components test results for the whole data set. C. Bilos et al. / Environmental Pollution 111 (2001) 149±158 157 sources, especially in the case of Mn and Cr where they account for 89 and 59%, respectively (Nriagu, 1989). Related to these elements are Ca and Mg, which present comparable EFs (station mean range=2.8±7.5) also suggesting a prevailing natural origin. 4. Conclusions Airborne particulate trace metals have been deter- mined bimonthly during day and night hours in four permanent stations located in residential, industrial and commercial sectors of La Plata City region. Concentra- tions of airborne metals in La Plata area were relatively low, similar to the values reported for not heavily pol- luted cities, probably re¯ecting the major administrative and trade-oriented character of the city. The data set show an important degree of variability including diurnal (day vs. night), spatial (inter-station) and temporal (inter-month) components which were evaluated by two- and three-way ANOVA and variance components tests (P<0.05). In general, trace metal con- centrations follow the behavior of TSP with higher concentrations during the day, particularly at Down- town site. Spatial di�erences were also evident with higher levels at Downtown station and lower at the Residential site. Overall, spatial di�erences were more important than diurnal variations; these di�erences were clearly most signi®cant for Pb, which presents higher values at the Downtown site re¯ecting the importance of motor exhaust inputs. In contrast, diur- nal di�erences are more important for Mn indicating increased dust resuspension during day hours. A seaso- nal trend of increasing concentrations during April± May±August and decreasing values in September± December was also detected. EFs relative to earth crust abundances were calcu- lated to evaluate anthropic versus natural sources. EFs were highest for Pb, Zn, Cd and Cu, indicating that anthropic inputs prevail over normal crustal weathering processes. These TSP EFs were similar to those calcu- lated for vehicle particle emissions suggesting that motor exhausts are the most important source of these elements in La Plata region. EFs calculated for Mn, Cr, Ni, Ca andMg are 20±60 times lower indicating that their main source in airborne particles are soil-derived dusts. References Bertine, K.K., Goldberg, E.D., 1971. Fossil fuel combustion and the major sedimentary cycle. Science 173, 233±235. Cadle, S.H., Mulawa, P.A., Ball, J., Donase, C., Weibel, A., Sagebiel, J.C., Knapp, K.T., Snow, R., 1997. Particulate emission rates from in-use high-emitting vehicles recruited in Orange Country, Cali- fornia. Environmental Science and Technology 31, 3405±3412. Camilio n, M.C., 1993. Clay mineral composition of Pampean Loess (Argentina). Quaternary Internacional 17, 27±31. Chow, T.J., Earl, J.L., Snyder, C.B., 1972. Lead aerosol baseline: concentration at White Mountain, California. Science 178, 401±402. Colombo, J.C., Landoni, P., Bilos, C., 1999. Sources, distribution and variability of airborne particles and hydrocarbons in La Plata Area, Argentina. Environmental Pollution 104, 305±314. Duce, R.A., Ho�man, G.L., Zoller, W.H., 1975. Atmospheric trace metals at remote northern and southern hemisphere sites: pollution or natural? Science 187, 59±61. Eisenreich, S.J., Metzer, N.A., Urban, N.R., Robbins, J., 1986. Response of atmospheric lead to decreased use of lead in gasoline. Environmental Science and Technology 20, 171±174. Gao, Y., Arimoto, R., Zhou, M.Y., Lee, D.S., Duce, R.A., 1992. Input of atmospheric trace-elements and mineral matter to the Yel- low Sea during the Spring of low dust year. Journal of Geophysical Research-Atmosphere 97 (4), 3767±3777. Greenberg, R.R., 1990. Analysis of atmospheric particulate samples via instrumental neutron activation analysis. In: Zielinski Jr., W.L., Dorko, W.D. (Eds.), Monitoring Methods for Toxics in the Atmosphere, ASTM STP 1052. American Society for Testing and Materials, Philadelphia, pp. 175±184. Harrison, R.M., Smith, D.J.T., Luhana, L., 1996. Source apportion- ment of atmospheric polycyclic aromatic hydrocarbons collected from an urban location in Birmingham, U.K. Environmental Sci- ence and Technology 30, 825±832. Hileman, B., 1981. Particulate matter: the inhalable variety. Environ- mental Science and Technology 15, 983±986. Hlavay, J., Polya k, K., Wesemann, G., 1992. Particle size distribution of minerals phases and metals in dusts collected at di�erent work- places. Fresenius J. Anal. Chem. 344, 319±321. Khandekar, R.N., Tripathi, R.M., Raghunath, R., Sathe, A.P., Nambi, K.S.V., 1992. Heavy metal concentrations in air, water and rock sam- ples at Antarctica during 1989±1990. Current Science 63, 201±204. Lantzy, R.J., Mackenzie, F.T., 1979. Atmospheric trace metals: global cycles and assessment of man's impact. Geochimica et Cosmochi- mica Acta 43, 511±525. Loureiro, A.L.M., Vasconcellos, M.B.A., Pereira, E.B., 1992. Trace elements determination in aerosols from the Antarctic Peninsula by neutron activation analysis. Journal of Radioanalytical and Nuclear Chemistry 159, 21±28. Natusch, D.F.S., Wallace, J.R., Evans Jr., C.A., 1974. Toxic trace elements: preferencial concentration in respirable particles. Science 183, 202±204. Nriagu, J.O., 1979. Global inventory of natural and anthropogenic emissions of trace metals to the atmosphere. Nature 279, 409±411. Nriagu, J.O., 1989. A global assessment of natural sources of atmos- pheric trace metals. Nature 338, 47±49. Nriagu, J.O., Pacyna, J.M., 1988. Quantitative assessment of world- wide contamination of air, water and soils by trace metals. Nature 333, 134±139. RaÈ dlein, N., Heumann, K.G., 1992. Trace analysis of heavy metals in aerosol over the Atlantic Ocean from Antarctica to Europe. Intern. J. Environ. Anal. Chem. 48, 127±150. Sweet, C.W., Vermette, S.J., Landsberg, S., 1993. Sources of toxic trace elements in urban air in Illinois. Environmental Science and Technology 27, 2502±2510. Taylor, S.R., 1964. Abundance of chemical elements in the Continental Crust: a new table. Geochimica et Cosmochimica Acta 28, 1273±1285. Veron, A., Church, T.M., Patterson, C.C., Erel, Y., Merril, J.T., 1992. Continental origin and industrial sources of trace metals in the Northwest Atlantic troposphere. Journal of Atmospheric Chemistry 14, 339±351. Voutsa, D., Samara, C., 1996. Trace elements in vegetables grown in an industrial area in relation to soil and air particulate matter. Environmental Pollution 94, 325±335. Zoller, W.H., Gladney, E.S., Duce, R.A., 1974. Atmospheric con- centrations and sources of trace metals at the South Pole. Science 183, 198±200. 158 C. Bilos et al. / Environmental Pollution 111 (2001) 149±158