Air Pollution Distribution Patterns in the San Bernardino Mountains of Southern California: a 40-Year Perspective

Since the mid-1950s, native pines in the San Bernardino Mountains (SBM) in southern California have shown symptoms of decline. Initial studies in 1963 showed that ozone (O3) generated in the upwind Los Angeles Basin was responsible for the injury and decline of sensitive trees. Ambient O3 decreased significantly by the mid-1990s, resulting in decreased O3 injury and improved tree growth. Increased growth of trees may also be attributed to elevated atmospheric nitrogen (N) deposition. Since most of the N deposition to mixed conifer forest stands in the SBM results from dry deposition of nitric acid vapor (HNO3) and ammonia (NH3), characterization of spatial and temporal distribution of these two pollutants has become essential. Although maximum daytime O3 concentrations over last 40 years have significantly decreased (~3-fold), seasonal means have been reduced much less (~1.5-fold), with 2-week long means occasionally exceeding 100 ppb in the western part of the range. In the same area, significantly elevated concentrations of HNO3 and NH3, up to 17.5 and 18.5 μg/m3 as 2-week averages, respectively, have been determined. Elevated levels of O3 and increased N deposition together with long-term drought predispose the SBM forests to massive bark beetle attacks making them susceptible to catastrophic fires.


INTRODUCTION
Since the mid-1950s, mid-elevation forests in the San Bernardino Mountains (SBM) have shown symptoms of significant decline, which was first shown on its most sensitive species, ponderosa pine (Pinus ponderosa). Initially, the condition was termed "x-disease" because its cause was unknown [1]. However, as a result of field and controlled experiments, in which seedlings, saplings, and branches of trees were exposed to charcoal-filtered clean air and ambient smog, it was revealed that ozone (O 3 ), an important component of smog, was the main cause of the decline [2]. Photochemical smog generated in the heavily populated Los Angeles Basin (LA Basin) is trapped by thermal inversion (up to about 2,000 m above the sea level). In a typical summer season, the polluted LA Basin air masses move inland with easterly winds [3] and are intercepted by the SBM range.
About 70% of air pollution in the LA Basin, where presently ~14 million of people live, results from combustion of gasoline and diesel fuel by ~12 million motor vehicles [4]. Population growth in southern California is the highest in the entire nation. While the U.S. population has nearly doubled since 1950, in the entire state of California it tripled, and it grew even more in southern California [5]. Historical and projected population changes for the six counties of the Southern California Association of Governments (SCAG) region indicate that growth was the highest in those counties from which emissions directly affect air pollution status of the SBM. The highest population growth between 1960 and 2000 was recorded for Riverside County (5.05-fold increase), Orange County (4.04-fold increase), and San Bernardino County (3.4-fold increase), compared with 1.58-fold growth in Los Angeles County [6].
During this rapid expansion of the southern California population, several air pollution control initiatives were initiated by the California Air Resources Board (CARB) and implemented by South Coast Air Quality Management District (SCAQMD) [7]. These measures resulted in a significant reduction in pollutant emissions as well as a large reduction of ambient O 3 concentrations in the LA Basin (Fig. 1). Cleaner burning fuel also delayed onset of O 3 formation, resulting in smaller O 3 decreases in more easterly mountain and desert areas than near urban areas. Recent studies [8] indicate the spatial distribution of air pollution may also have changed, but a lack of monitoring data has prevented characterization of spatial trends. Since 1990, Forest Service scientists have studied interactive effects of O 3 , wet and dry deposited nitrogenous (N) compounds, climatic variability, water availability, and outbreaks of insects on physiological responses of trees, their growth and mortality, and changes in hydrological and soil processes in forests. The following first description of spatial trends of major air pollutants in the SBM is critical to understanding the ongoing impacts of air pollution on the highly stressed forest areas of southern California.

Monitoring Network
In 2002, 18 sites were established in the SBM for air pollution and forest health monitoring (Fig. 2). These sites included 11 original sites of the 1968-1974 study [2] as well as the newly established sites [9]. The sites were selected following ground inspection and available information on distribution of photochemical smog, soil condition, forest condition, and stand/tree health [9]. Monitoring sites were selected in open-terrain locations, such as forest clearings, burnt areas, forest nurseries, etc. Sites were located on a western aspect, at least 100 m from a local road and 200 m from main roads. Free air movement from all directions was required; however, sites exposed to continuously strong winds were avoided. Sampler stands were placed at a distance at least two times the height of the tallest tree from forest edges, however, sparsely dispersed smaller trees or shrubs that did not directly obstruct the samplers were allowed. Passive samplers were exposed for 2-week periods during the summer and fall seasons of 2002-2005. Monitoring was performed between May and October/November; beginning and end of the monitoring campaigns depended on weather conditions. Samplers in protective caps were hung on a wooden stand about 2 m above ground level.

Equipment Used
Ogawa samplers (Pompano Beach, FL) were used to measure O 3 concentrations [10]. In each sampler, two replicate cellulose filters saturated with nitrite (NO 2 -) were used. Nitrite on the filters is oxidized by ambient O 3 to nitrate (NO 3 -). Nitrate concentration in filter extracts was determined by ion chromatography (Dionex, Model 4000i) and the rate of NO 3 formation served as a measure of 2-week average ambient O 3 concentration at every site. Rates of NO 3 formation in the passive samplers were compared to real-time O 3 concentration measurements by UV absorption (2 B Technologies) and the empirically derived coefficients were used to calculate O 3 concentrations.
Passive samplers developed by the USDA Forest Service [11] were used for monitoring nitric acid (HNO 3 ) concentrations. In that sampler, ambient air passes through Teflon membrane and gaseous HNO 3 is absorbed on Nylasorb nylon filter as NO 3 -. Nitrate concentrations in sample extracts were analyzed by ion chromatography (Dionex, Model 4000i) and concentrations of HNO 3 were calculated using the calibration curves [11]. Three replicate HNO 3 samplers were used at each monitoring site.
Ammonia (NH 3 ) concentrations were determined with Ogawa passive samplers. Ammonia is absorbed on two replicate cellulose pads coated with citric acid forming ammonium citrate. Ammonium concentrations in filter extracts were determined colorimetrically on a Technicon Autoanalyzer. Concentrations of NH 3 from passive samplers were calculated based on a comparison with colocated anular denuder systems [12].

Geostatistical Analysis
Pollutant distribution maps were developed with the Geostatistical Analyst, extension of ArcGIS software (ESRI, Redlands, CA) software. The Geostatistical Analyst uses values measured at sample points at different locations in the landscape and interpolates them into a continuous surface. Using a set of pollutant concentration measurements for the study area, a spatial model of O 3 concentration was developed using the inversed distance weighted (IDW) method [13].

RESULTS
• Ozone -Seasonal average concentrations of the pollutants in 4 years of the study based on consecutive 2-week long collection periods were used in the analysis ( Table 1) 3 were found at the HV site north of Big Bear Lake (Fig. 4A-D).

DISCUSSION
Most available historical information regarding air pollution in the SBM is on O 3 . Lee et al. [14] reconstructed hourly O 3 data for the Crestline site, which has been near the CP site of our network since  [17]. High mean O 3 values in the mountain locations may be partially explained by the elevated levels of the pollutant continuing throughout the night hours due to a lack of nitric oxide (NO) that effectively scavenges O 3 in the urban source areas [18]. In general, our study shows that although the maximum O 3 concentrations have significantly decreased over 40 years of studies, high mean values occurring in the western part of the range have a strong potential for physiological and biochemical effects [19], damage of sensitive species and individuals, and predisposition of trees to drought stress and bark beetle attacks [20]. Much less is known on distribution of HNO 3 and NH 3 in the studied area. First measurements of these pollutants were performed in 1992 and 1993 at Barton Flats in the middle of the SBM air pollution gradient. Summer season mean values calculated for June-September 1991-1993 were 1.4 μg/m 3 for HNO 3 [22] while those for NH 3 should be less than 1 µg/m 3 [23]. The observed increase in NH 3 concentrations may reflect effects of the reformulation of gasoline that promotes higher production of reduced N compounds from combustion engines equipped with catalytic converters [24]. Reasons for the relatively high increase in HNO 3 levels since 1991-1993 are not clear and will have to be further investigated. Elevated levels of HNO 3 and NH 3 , especially in the western part of the range, explain why these areas experience high levels of N deposition. Both HNO 3 and NH 3 have very high deposition velocities to foliar surfaces [23], and can also be effectively taken up by stomata [25]. Elevated levels of N deposition increase drought stress of trees, lead to build up of dead biomass in forest stands, and contribute to contamination of stream and surface water with nitrate [26]. These effects and those caused by elevated O 3 levels may cause deterioration of health of trees, change forest species composition, and predispose the SBM forests to widespread outbreaks of bark beetle and catastrophic fires, such as those which took place in the fall of 2003 [27].   [14,15].