Enrichment, isolation and characterisation of ruminal bacteria that degrade non-protein amino acids from the tropical legume Acacia angustissima

Acacia angustissima has been proposed as a protein supplement in countries where low quality forages predominate. A number of non-protein amino acids have been identified in the leaves of A. angustissima and these have been linked to toxicity in ruminants. The non-protein amino acid 4-n-acetyl-2,4-diaminobutyric acid (ADAB) has been shown to be the major amino acid in the leaves of A. angustissima. The current study aimed to identify micro-organisms from the rumen environment capable of degrading ADAB by using a defined rumen-simulating media with an amino acid extract from A. angustissima. A mixed enrichment culture was obtained that exhibited substantial ADAB-degrading ability. Attempts to isolate an ADAB-degrading micro-organism were carried out, however no isolates were able to degrade ADAB in pure culture. This enrichment culture was also able to degrade the non-protein amino acids diaminobutyric acid (DABA) and diaminopropionic acid (DAPA) which have structural similarities to ADAB. Two isolates were obtained which could degrade DAPA. One isolate is a novel Grain-positive rod (strain LPLR3) which belongs to the Firmicutes and is not closely related to any previously isolated bacterium. The other isolate is strain LPSR1 which belongs to the Gammaproteobacteria and is closely related (99.93% similar) to Klebsiella pneumoniae subsp. ozaenae. The studies demonstrate that the rumen is a potential rich source of undiscovered micro-organisms which have novel capacities to degrade plant secondary compounds. (c) 2005 Elsevier B.V. All rights reserved.

Chemical deterrence of a marine cyanobacterium against sympatric and non-sympatric consumers

This study investigates the influence of mesograzer prior exposure to toxic metabolites on palatability of the marine cyanobacterium, Lyngbya majuscula. We examined the palatability of L. majuscula crude extract obtained from a bloom in Moreton Bay, South East Queensland, Australia, containing lyngbyatoxin-a (LTA) and debromoaplysiatoxin (DAT), to two groups: (1) mesograzers of L. majuscula from Guam where LTA and DAT production is rare; and (2) macro- and mesograzers found feeding on L. majuscula blooms in Moreton Bay where LTA and DAT are often prevalent secondary metabolites. Pair-wise feeding assays using artificial diets consisting of Ulva clathrata suspended in agar (control) or coated with Moreton Bay L. majuscula crude extracts (treatment) were used to determine palatability to a variety of consumers. In Guam, the amphipods, Parhyale hawaiensis and Cymadusa imbroglio; the majid crab Menaethius monoceros; and the urchin Echinometra mathaei were significantly deterred by the Moreton Bay crude extract. The sea hares, Stylocheilus striatus, from Guam were stimulated to feed by treatment food whereas S. striatus collected from Moreton Bay showed no discrimination between food types. In Moreton Bay, the cephalaspidean Diniatys dentifer and wild caught rabbitfish Siganus fuscescens were significantly deterred by the crude extract. However, captive-bred S. fuscescens with no known experience with L. majuscula did not clearly discriminate between food choices. Lyngbya majuscula crude extract deters feeding by most mesograzers regardless of prior contact or association with blooms.

Customer information system data pre-processing with feature selection techniques for non-technical losses prediction in an electricity market

Non-technical losses (NTL) identification and prediction are important tasks for many utilities. Data from customer information system (CIS) can be used for NTL analysis. However, in order to accurately and efficiently perform NTL analysis, the original data from CIS need to be pre-processed before any detailed NTL analysis can be carried out. In this paper, we propose a feature selection based method for CIS data pre-processing in order to extract the most relevant information for further analysis such as clustering and classifications. By removing irrelevant and redundant features, feature selection is an essential step in data mining process in finding optimal subset of features to improve the quality of result by giving faster time processing, higher accuracy and simpler results with fewer features. Detailed feature selection analysis is presented in the paper. Both time-domain and load shape data are compared based on the accuracy, consistency and statistical dependencies between features.