The treatment and desalination of inland water via reverse osmosis (RO) technology is gaining momentum for upgrading brackish groundwater and developing supplemental fresh water for various regions. In brackish RO plants, high water recovery is critical in order to minimize the volume of residual RO concentrate (brine), given the economic and environmental challenges of concentrate management. However, high recovery may be limited by mineral salt scaling resulting from supersaturation of sparingly soluble minerals (e.g., CaSO4, BaSO4, CaCO3, SiO2). Mineral scaling results in membrane surface blockage, reduction of permeate flux and shortening of membrane lifetime.

In order to control or prevent mineral scaling, effective mitigation methods must be developed and tested (i.e., the relationship between RO operating conditions and mineral scaling). To accomplish the above, early detection of mineral scaling is essential. To meet the above challenges, a novel high-pressure RO membrane monitoring system was developed to allow direct membrane surface imaging. The membrane monitoring system (MMS) was interfaced with a slipstream from an RO plant, whereby captured membrane surface images were analyzed online. The present monitoring approach demonstrated, for the first time, early detection of silica scale formation and growth kinetics. Detailed silica scaling studies revealed that scaling consisted of discrete silica particles embedded in a silica “gel” layer. The membrane monitoring approach also served to evaluate the impact of a membrane biofilm on concentration polarization (CP) and mineral scaling kinetics. It was shown that biofilms enhanced CP within the biofilm and significantly exacerbated mineral scaling.

The membrane monitoring approach was subsequently deployed in a field study of desalination of brackish agricultural drainage (AD) water. In these studies, the membrane monitor was integrated with a mobile RO pilot plant, developed at UCLA, for real-time field optimization of RO operating conditions for averting mineral scaling. The above approach demonstrated that effective feed filtration for removal of suspended particles is critical for mitigating mineral scaling and reducing nucleation triggered by surface deposited particles. Moreover, antiscalant selection and dosage optimization was feasible under field conditions which also enabled determination of the maximum feasible water recovery level.

Species Distribution Models (SDMs) are the primary tool used to predict the future niche spaces for species. These models, however, do not usually incorporate species interactions that may restrict available niche space. Analysis of the widely studied coevolving network of Greya moths and Lithophragma plants shows that incorporating species interactions into SDMs can produce results that differ greatly from models without species interaction effects. This effect is highly variable, however, even among very closely related species. Moreover, choosing different phylogenetic levels to run SDMs can alter the predictions greatly, but this effect is also highly variable. The differences among projections that incorporate species interactions and phylogenetics, and those that include neither effect, may result from the relative ranges of these species and populations, and their relationships to each other. We also show here how these models predict changes in the particular combinations of plant and insect species that interact in each region, which could, in turn, lead to yet more cascading effects on the future of the geographic distributions of these coevolving lineages of interacting species.

Floral volatiles are often used by plants to attract the subset of floral visitors that are mostly likely to contribute to pollination. Consequently, the composition and complexity of floral scents differ greatly among plant species. Recent studies of woodland stars, Lithophragma (Saxifragaceae), have found that species in this genus often have scents composed of an unusually large number of compounds, and even populations of the same species can differ extremely in the composition of scents. Here, I assess how the composition and complexity of floral scents change when closely related species of woodland stars hybridize. I evaluated floral scent in experimentally produced hybrids between two sister taxa, L. parviflorum and L. affine. Most F1 and F2 hybrids produced scents that were intermediate mixtures of the two parental species, but the scents of hybrids were often novel in the relative proportions of compounds they produced. Moreover, some hybrids produced scents with either fewer or more compounds than found in any parental individual. Alteration of the mix and complexity of monoterpenes contributed greatly to the variation among hybrid individuals, but most individuals produced scents composed of compounds arising from at least several biochemical pathways. Hence, the chemical profiles of these hybrids suggest the possibility of transgressive segregation of this complex floral trait. These results suggest that natural hybridization in woodland stars would produce novel scent combination that could alter the interactions with their coevolved Greya moth pollinators and other visiting insects. Because the distributions of many plant species are rapidly changing, altered interactions with pollinators through novel scent combinations in hybrids could become common in many species-rich taxa capable of hybridization.

Background and aims

Methods

Key results

Conclusions