Resilience of the zooplankton community in the northeast Gulf of Mexico during and after the Deepwater Horizon oil spill.

We evaluated the resilience of the zooplankton community to the Deepwater Horizon oil spill in the northeast Gulf of Mexico, by assessing abundance, biomass, spatial distribution, species composition, and diversity indices during spring, summer, and winter, May 2010 to August 2014. SEAMAP samples collected between spring and summer 2005-2009 were analyzed as a baseline. Our results did not indicate that there was a long-term impact from the oil spill, but did demonstrate that environmental variability and riverine processes strongly governed zooplankton community dynamics. Zooplankton abundances during the oil spill (spring 2010) were not significantly different from abundances during spring 2011 and 2012. Summer 2010 abundances were the highest observed for the 2005 to 2014 period, due to high river discharge, high chlorophyll, and aggregation in eddies. High densities of the dinoflagellate, Noctiluca, during the oil spill, and the copepod, Centropages velificatus, and larvaceans in all years, suggest that these taxa warrant further investigation. Ecosystem connectivity (zooplankton transport by currents into the oil spill region), high fecundity, relatively short generation times, and refugia in deeper depths are key factors in zooplankton resilience to major perturbations. This study serves as a baseline for assessment of future impacts to this system.

Transformation and release of nanoparticle additives & byproducts from commercially available surface coatings on pressure treated lumber via dermal contact.

Production and marketing of "nano-enabled" products for consumer purchase has continued to expand. However, many questions remain about the potential release and transformation of these nanoparticle (NP) additives from products throughout their lifecycle. In this work, two surface coating products advertised as containing ZnO NPs as active ingredients, were applied to micronized copper azol (MCA) and aqueous copper azol (ACA) pressure treated lumber. Coated lumber was weathered outdoors for a period of six months and the surface was sampled using a method developed by the Consumer Product Safety Commission (CPSC) to track potential human exposure to ZnO NPs and byproducts through simulated dermal contact. Using this method, the total amount of zinc extracted during a single sampling event was <1 mg/m2 and no evidence of free ZnO NPs was found. Approximately 0.5% of applied zinc was removed via simulated dermal contact over 6-months, with increased weathering periods resulting in increased zinc release. XAFS analysis found that only 27% of the zinc in the as received coating could be described as crystalline ZnO and highlights the transformation of these mineral phases to organically bound zinc complexes during the six-month weathering period. Additionally, SEM images collected after sampling found no evidence of free NP ZnO release during simulated dermal contact. Both simulated dermal contact experiments, and separate leaching studies demonstrate the application of surface coating solutions to either MCA and ACA lumber will reduce the release of copper from the pressure treated lumber. This work provides clear evidence of the transformation of NP additives in consumer products during their use stage.

Release and transformation of ZnO nanoparticles used in outdoor surface coatings for UV protection.

A major area of growth for "nano-enabled" products has been the addition of nanoparticles (NPs) to surface coatings including paints, stains and sealants. Zinc oxide (ZnO) NPs, long used in sunscreens and sunblocks, have found growing use in surface coating formulations to increase their UV resistance, especially on outdoor products. In this work, ZnO NPs, marketed as an additive to paints and stains, were dispersed in Milli-Q water and a commercial deck stain. Resulting solutions were applied to either Micronized-Copper Azole (MCA) pressure treated lumber or a commercially available composite decking. A portion of coated surfaces were placed outdoors to undergo environmental weathering, while the remaining samples were stored indoors to function as experimental controls. Weathered and control treatments were subsequently sampled periodically for 6 months using a simulated dermal contact method developed by the Consumer Product Safety Commission (CPSC). The release of ZnO NPs, and their associated degradation products, was determined through sequential filtration, atomic spectroscopy, X-Ray Absorption Fine Structure Spectroscopy, and electron microscopy. Across all treatments, the percentage of applied zinc released through simulated dermal contact did not exceed 4%, although transformation and release of zinc was highly dependent on dispersion medium. For MCA samples weathered outdoors, water-based applications released significantly more zinc than stain-based, 180 ±â€¯28, and 65 ±â€¯9 mg/m2 respectively. Moreover, results indicate that the number of contact events drives material release.

Dermal transfer and environmental release of CeO2 nanoparticles used as UV inhibitors on outdoor surfaces: Implications for human and environmental health.

A major area of growth for "nano-enabled" consumer products have been surface coatings, including paints stains and sealants. Ceria (CeO2) nanoparticles (NPs) are of interest as they have been used as additives in these these products to increase UV resistance. Currently, there is a lack of detailed information on the potential release, and speciation (i.e., ion vs. particle) of CeO2 NPs used in consumer-available surface coatings during intended use scenarios. In this study, both Micronized-Copper Azole pressure-treated lumber (MCA), and a commercially available composite decking were coated with CeO2 NPs dispersed in Milli-Q water or wood stain. Coated surfaces were divided into two groups. The first was placed outdoors to undergo environmental weathering, while the second was placed indoors to act as experimental controls. Both weathered surfaces and controls were sampled over a period of 6months via simulated dermal contact using methods developed by the Consumer Product Safety Commission (CPSC). The size and speciation of material released was determined through sequential filtration, total metals analysis, X-Ray Absorption Fine Structure Spectroscopy, and electron microscopy. The total ceria release from MCA coated surfaces was found to be dependent on dispersion matrix with aqueous applications releasing greater quantities of CeO2 than stain based applications, 66±12mg/m2 and 36±7mg/m2, respectively. Additionally, a substantial quantity of CeO2 was reduced to Ce(III), present as Ce(III)-organic complexes, over the 6-month experimental period in aqueous based applications.

Fast support vector machines for continuous data.

Support vector machines (SVMs) can be trained to be very accurate classifiers and have been used in many applications. However, the training time and, to a lesser extent, prediction time of SVMs on very large data sets can be very long. This paper presents a fast compression method to scale up SVMs to large data sets. A simple bit-reduction method is applied to reduce the cardinality of the data by weighting representative examples. We then develop SVMs trained on the weighted data. Experiments indicate that bit-reduction SVM produces a significant reduction in the time required for both training and prediction with minimum loss in accuracy. It is also shown to typically be more accurate than random sampling when the data are not overcompressed.

Recognizing plankton images from the shadow image particle profiling evaluation recorder.

We present a system to recognize underwater plankton images from the shadow image particle profiling evaluation recorder (SIPPER). The challenge of the SIPPER image set is that many images do not have clear contours. To address that, shape features that do not heavily depend on contour information were developed. A soft margin support vector machine (SVM) was used as the classifier. We developed a way to assign probability after multiclass SVM classification. Our approach achieved approximately 90% accuracy on a collection of plankton images. On another larger image set containing manually unidentifiable particles, it also provided 75.6% overall accuracy. The proposed approach was statistically significantly more accurate on the two data sets than a C4.5 decision tree and a cascade correlation neural network. The single SVM significantly outperformed ensembles of decision trees created by bagging and random forests on the smaller data set and was slightly better on the other data set. The 15-feature subset produced by our feature selection approach provided slightly better accuracy than using all 29 features. Our probability model gave us a reasonable rejection curve on the larger data set.