2007 ◽  
Vol 16 (3) ◽  
pp. 295 ◽  
Author(s):  
Owen F. Price ◽  
Andrew C. Edwards ◽  
Jeremy Russell-Smith

We investigated the efficacy of firebreaks in the prevention of wildfires in the Arnhem Land Plateau, a vast, rugged and sparsely populated region with high biodiversity value and frequent wildfires. A total of 623 events where a fire met a permanent firebreak (cliffs, stream order, tracks and roads) in different fire seasons were compiled. Cliffs were more effective than streams at stopping fires, which were more effective than roads. Larger streams were more effective than small ones. The largest streams stop 75% of early dry season fires, but there are no firebreak types with more than 50% likelihood of stopping a late dry season fire. Geographic Information System (GIS) surfaces of the relative density of the three firebreak features in the landscape were randomly sampled and compared with the total number of fires and late dry season fires using generalised linear modelling. Several of the density variables were weakly but significantly related to fire frequency, and it appears that late dry season fires are influenced by features at a larger scale (16-km radius) than total fires (4 km). The Aerial Prescribed Burning program for 2004 was studied to identify how effective it was at stopping subsequent wildfires by dividing ignition lines into 137 5-km sections. Only 20% of sections achieved a 100% burn and where gaps occurred, a subsequent fire was 88% likely to penetrate the line. Firebreaks are not certain instruments for fire management in this area.


1996 ◽  
Vol 2 (2) ◽  
pp. 449-477 ◽  
Author(s):  
A.E. Renshaw ◽  
S. Haberman ◽  
P. Hatzopoulos

ABSTRACTDeaths and exposures by individual calendar year and individual years of age for the U.K. male assured lives experience over the recent past are comprehensively modelled using generalised linear modelling techniques. Our principal objective is to develop a model which incorporates both the age variation in mortality and the underlying time trends in the mortality rates. The approach has considerable advantages over ad hoc methods of fitting parametric models to represent the age variation in mortality and then separately attempting to represent the time trends in the parameters of these models. The approach advocated can be seen as an extension to the conventional parametric graduation techniques used by the CMI Bureau to represent trends in mortality.


2009 ◽  
Vol 156 (9) ◽  
pp. 1941-1947 ◽  
Author(s):  
Bridget S. Green ◽  
Caleb Gardner ◽  
Robert B. Kennedy

2000 ◽  
Vol 57 (2) ◽  
pp. 429-440 ◽  
Author(s):  
D J Agnew ◽  
C P Nolan ◽  
J R Beddington ◽  
R Baranowski

Eleven rajid species are taken around the Falkland Islands, with four species, Bathyraja griseocauda, Bathyraja albomaculata, Bathyraja brachyurops, and Raja flavirostris dominating commercial catches and generally occurring together. Catch limits for individual species are not used in management because species are not separated in the catch or reported separately. The catch per unit effort for the mixed rajid assemblage was standardised using generalised linear modelling techniques, and two production models were used to estimate stock size and sustainable yield. Maximum likelihood methods were used to demonstrate that there are two distinct rajid communities, one to the north and one to the south of the Falkland Islands, which have different sustainable yields. Changes in species composition over the 10-year course of the fishery confirm theoretical expectations that the larger, later-maturing B. griseocauda is being replaced in catches by the smaller, earlier-maturing B. albomaculata and B. brachyurops. These changes in composition were evident after only 6 years of directed fishing. The current fishery to the north of the Falkland Islands appears to be stable at an annual catch of about 3000 t, which is between 6.5 and 7.6% of the estimated pre-exploitation biomass.


2004 ◽  
Vol 31 (2) ◽  
pp. 129 ◽  
Author(s):  
A. S. Kutt ◽  
N. Y. Thurgate ◽  
D. S. Hannah

Two systematic regional fauna surveys conducted in the central-north of the state, and an ancillary review of existing published data, identified many new localities and provided an opportunity to describe aspects of this species' ecology and habitat. As well as in typical Triodia grasslands, Pseudomys desertor was recorded in a wide range of sub-tropical savanna woodland, shrubland and grassland vegetation types, many within a zone with rainfall of 500–750 mm. In these higher-rainfall areas, average bodyweight was significantly larger than previously reported for arid-zone specimens. Generalised linear modelling was used to examine which environmental factors best explained the species' abundance, and ground cover was a consistent predictor across the two study areas. Despite occurring in sites mechanically cleared of trees, P. desertor was more abundant in those pastures with a high cover of native hummock grasses. Though superficially indicating a degree of tolerance of disturbance, the evidence suggests that P. desertor is a fire- and grazing-decreaser. Overall, P. desertor was most abundant in areas with the least historical grazing pressure and disturbance.


Sign in / Sign up

Export Citation Format

Share Document