The Use of Spanish Moss as a Biological Indicator to Examine Relationships Between Metal Air Pollution, Vegetation Cover, and Environmental Equity in Tampa, Florida

Studies of inequality in exposure to less common air pollutants, like metals, are often limited by the costs of high spatial resolution measurements. Spanish moss (Tillandsia usneoides) is a promising bioindicator for measuring air pollution due to its lower cost, enabling capture of time-average environmental concentrations at high spatial resolution. This study had three major aims. First, I aimed to use Spanish moss as a bioindicator to characterize ambient concentrations of selected metals (Ti, Cr, Mn, Co, Ni, Cd, Hg, Pb, As, and Sb) in Tampa, Florida. My second goal was to determine the impact of vegetation cover on metals air pollutants. Last, I aimed to determine the environmental equity implications of metals air pollution in the area. As primary data, metals air pollution concentrations were obtained from Spanish moss. A mixture of acids was used to extract the metals from Spanish moss. Inductively Coupled Plasma Mass Spectrometry (ICP-MS) was then used to analyze the metals from the final solution. Secondary data include population demographics from the U.S. Census from the American Community Survey 2014–2018, tree cover, and road data. For the first goal, samples were collected at three sites near air pollution sources and at three sites within conservation lands around Tampa to examine differences in concentrations of Spanish moss. Three samples were collected from each site. As a complementary analysis, the correlation of metals concentrations in Spanish moss with major roads and distance to pollution point sources was examined. The Mann-Whitney U test was performed to examine metals concentrations differences between potentially polluted and conservation land sites, while Spearman’s correlation was performed to examine the correlation between metals concentrations and major roads and distance to pollution point source. The findings showed that Spanish moss can distinguish between variations in metals pollutant concentrations in the air. Statistical analysis results indicate that there was a significant difference in metals concentrations (Ti, Cr, Co, Ni, Cd, Pd, As and Sb) between polluted and conservation land sites within high mean concentrations in Spanish moss at polluted sites and low mean concentrations at conservation land sites. Manganese showed an opposite association of other metals. This opposite association was explained by the enrichment of this type of pollutant in forest grounds since this metal is an essential nutrient found in plants (litterfall likely caused the forest ground enrichment of this metal). For the second goal which examines the effect of tree cover on metal air pollution concentrations, a total of 180 samples of Spanish moss were collected within the city of Tampa. Tree cover percentages were calculated in different buffer sizes around the sampling sites to examine their correlations with metals concentrations. Road density was also calculated and introduced as an explanatory variable since roads are one of the main air pollution sources. Spearman’s correlation, ordinary least square (OLS) and simultaneous autoregressive (SAR) regression models were performed to examine the correlation between tree cover percentage in different buffer sizes and metals concentrations. The findings showed that Cd, Pb and Sb showed negative significant associations with tree cover percentage in Spearman’s correlation only, while no association were found in regression models. Thus, the finding of the association between these metals and tree cover percentage is considered to be weak. It is commonly known that the examination of tree cover as an air filter is complex due to the many variables that could affect this correlation. Future work is needed to consider other factors that affect air pollution in the air and air pollution foliar deposition such as vegetation characteristics, meteorological data, and chemical/physical characteristics of pollutant types when examining the efficiency of vegetation in enhancing air quality. For the third goal of this dissertation, environmental inequity was examined by comparing demographic data at the block group level, with metals air pollution data measured by Spanish moss. Sixty block groups were randomly selected to represent my block group samples. The averaged metal concentration value of three samples was assigned to each block group. For statistical analysis, Spearman’s correlation and linear regression were performed (full model and backward step-wise model). Results for race/ethnicity variables showed that Blacks and Hispanics are disproportionately exposed to high level of metals concentrations (Ti, Cd, Pb, and Sb for African Americans, and Pb only for Hispanics). Elderly population were found to be negatively associated with Sb and Pb metals concentrations. There were no associations between poverty level and income and metals concentration. However, other wealth indicator variables such as renters, household size, and single-family units indicated that poor people were not associated with high levels of metals air pollution. Similar results were found in previous studies conducted in the area that modeled hazardous air pollution. Further research is needed to examine the latter association since it appears to be complex in the city of Tampa