[This corrects the article DOI: 10.1371/journal.pone.0203694.].

The interrelationship between public interest in endangered species and the attention they receive from the conservation community is the 'flywheel' driving much effort to abate global extinction rates. Yet big international conservation non-governmental organisations have typically focused on the plight of a handful of appealing endangered species, while the public remains largely unaware of the majority. We quantified the existence of bias in popular interest towards species, by analysing global internet search interest in 36,873 vertebrate taxa. Web search interest was higher for mammals and birds at greater risk of extinction, but this was not so for fish, reptiles and amphibians. Our analysis reveals a global bias in popular interest towards vertebrates that is undermining incentives to invest financial capital in thousands of species threatened with extinction. Raising the popular profile of these lesser known endangered and critically endangered species will generate clearer political and financial incentives for their protection.

The melting of mountain glaciers and ice caps is expected to contribute significantly to sea-level rise in the twenty-first century although the magnitude of this contribution is not fully constrained. Glaciers in the Patagonian Icefields of South America are thought to have contributed about 10% of the total sea-level rise attributable to mountain glaciers in the past 50 years. However, it is unclear whether recent rates of glacier recession in Patagonia are unusual relative to the past few centuries. Here we reconstruct the recession of these glaciers using remote sensing and field determinations of trimline and terminal moraine location. We estimate that the North Patagonian Icefield has lost 103±20.7 km3 of ice since its late Holocene peak extent in AD 1870 and that the South Patagonian Icefield has lost 503±101.1 km3 since its peak in AD 1650. This equates to a sea-level contribution of 0.0018±0.0004 mm yr−1 since 1870 from the north and 0.0034±0.0007 mm yr−1 since 1650 from the south. The centennial rates of sea-level contribution we derive are one order of magnitude lower than estimates of melting over the past 50 years3, even when we account for possible thinning above the trimline. We conclude that the melt rate and sea-level contribution of the Patagonian Icefields increased markedly in the twentieth century.

The flow stress-elongation curves for a strain rate of 3 × 10 -3 s-1 have been obtained for plain C-Mn and Nb containing steels over the temperature range 600-1000°C after solution treating at 1330°C and cooling at 60 K min-1 to the test temperature. The type of curve depended on whether the part of the trough at the low temperature end of the hot ductility curve was wide or narrow. of the two phases, austenite and ferrite, ferrite independent of whether it is deformation induced or not, has been shown to have very much the better ductility. For plain C-Mn steels giving narrow trough behaviour, deformation induced ferrite (DIF) forms in equilibrium or near equilibrium amounts from the austenite, so that large amounts of DIF can form just below the Ae3. Ductility at the low temperature end of the trough is then dependent almost entirely on the amount of ferrite present. The improvement in ductility that occurs as the temperature is decreased, can therefore be related to the Ae 3 temperature, because a large reduction in flow stress occurs close to the Ae3 in the curve of peak stress against temperature. This DIF is believed to form almost immediately on yielding in both steels with narrow and wide trough behaviour. However, in steels showing wide trough behaviour, although DIF forms readily at close to the Ae3, it only forms as a thin film surrounding the austenite grains. This film grows very slowly in thickness as the temperature is reduced. In this case, it is necessary for the test temperature to be below the Ar3, so there is sufficient ferrite present prior to deformation for ductility to improve and this temperature can be recognised by a discontinuity in the curve of strain to the peak stress against temperature as well as in the curve of peak stress against the temperature. It is therefore possible from analysis of the flow curves, by plotting the peak stress against temperature and the strain to the peak stress against temperature to obtain the Ae3 for narrow troughs and the Ar3 for wide troughs. In contrast, to the very low strain required to produce DIF, the strain for the onset of dynamic recrystallisation (DRX) in these coarse grained steels is in excess of 20%. The temperature at which the strain to reach the peak stress for ferrite changed to that needed for DRX could also be used to determine the temperature for the onset of DRX. © 2006 Institute of Materials, Minerals and Mining.

Low (0.3%) and high manganese (1.4%) plain C - Mn steels with varying sulphur levels have had their hot ductility determined over the temperature range 700-1000°C, both after 'solution treatment' at 1330°C and directly after casting. It has been established that the width, depth and position of the hot ductility curves after solution treatment is more related to the transformation behaviour than either the sulphur in solution or the sulphide volume fraction or distribution. The growth of deformation induced ferrite at the austenite boundaries seems to be mainly diffusion controlled, and the higher is the transformation temperature for the γ - α phase change, the faster is the growth. Large amounts of ferrite can then form, giving good ductility. Thus, high transformation temperatures Ae3 or Ar 3 are required to produce narrow ductility troughs. It is believed that any detrimental influence of the sulphides on these 'solution treated' steels is swamped by the rapid increase in ferrite volume fraction. For the as cast state, as more sulphides are able to precipitate at the interdendritic boundaries and austeuite grain boundaries than in the solution treated condition, increasing the sulphur level causes a small deterioration in ductility at the high temperature end of the trough. In the present work, only narrow troughs have been found. This is in contrast to previous work on as cast C-Mn-Nb-Al steels, which exhibited wide troughs in the ductility curves, where it was shown that higher total sulphur levels lead to considerably worse ductility and that sulphur can be as detrimental to the ductility as niobium. It is recommended that, to avoid transverse cracking during continuous casting, in addition to keeping the sulphur level low, the carbon and manganese should also be as low as possible. © 2004 Institute of Materials, Minerals and Mining. Published by Maney on behalf of the Institute.

Hot ductility curves for high carbon Nb and Nb free steels have been determined immediately after casting at two P levels, ∼0.01% and ∼0.045%. High strain rates of 0.1-0.55 s-1 were generally used but some limited low strain rate testing at 7 × 10-3 s-1 was carried out on Nb containing steels. Nb containing steels showed, as expected, worse ductility than the Nb free steel but high P level was detrimental to ductility for both steels and ductility in general was very poor. Failure was intergranular with the presence of films of a P rich phase at the boundaries in the case of the Nb free steels and in addition to this, in Nb containing steel there was a Nb rich phase. The films were thicker and more continuous in the higher P steels. It is suggested that the P rich films are probably the low melting point phase Fe3P or Fe3(Mn)P, which can remain liquid down to temperatures as low as 950°C. Some back diffusion of P into the grain interior is possible if the strain rate is reduced and/or at high testing temperatures during the 5 min hold prior to testing. This allowed some improvement in ductility to occur in the lower P containing steels by reducing the amount of the low melting point phase at the boundaries.

The hot ductility of an as cast austenitic stainless steel slab having a coarse columnar grain structure emanating from the surface to the interior has been established in the temperature range 700-1000°C. Tensile specimens were taken in three directions, namely, parallel and transverse to the casting direction in a plane close to the top surface of the slab, and in the through thickness direction. The average dimensions of the columnar grains in these directions were 1, 2, and 8 mm, respectively. The ductility was highest in the through thickness direction and lowest parallel to the casting direction, this being the direction of stressing during the straightening operation in continuous casting. For the lowest ductility failures, the cracks were always intergranular. These observations can be explained in terms of the crack size being related to the grain size perpendicular to the tensile axis, this being greatest for samples tested parallel to the casting direction. In this direction, cracks can readily propagate a long distance before meeting a new grain and having to change direction. It is recommended that to restrict transverse cracking, columnar grains should be avoided. © 2000 IoM Communications Ltd.

The importance of dynamic recrystallisation in restoring ductility at the high temperature end of the hot ductility trough has been examined in an austenitic stainless steel Compression testing was used to establish the critical strain for dynamic recrystallisation εc for the temperature range 1000 to 850°C using the same strain rates of 3 × 10-2, 3 × 10-3, and 3 × 10-4 s-1 as had been used in previous work to establish the hot tensile ductility curves. Specimens were heated first to 1175 or 1000°C to give a coarse (240 μm) and finer (80 μm) grained steel respectively. The flow stress data from the compression tests on the coarser grained material were used to obtain the strain to the peak stress εp which could then be used to calculate the curve of εp versus temperature for use in establishing the temperature at which dynamic recrystallisation would first occur in a tensile test. For the coarse grained steel, the hot tensile tests had given ductility troughs for each strain rate with minimum ductility occurring at 850°C, the trough deepening and widening with decreasing strain rate. The trough was found to be associated with the presence of coarse carbides situated at the boundaries. Below 850°C, ductility recovered because grain boundary sliding was reduced. Above 850°C, ductility improved since fewer carbides were precipitated at the boundaries, facilitating dynamic recrystallisation. Recovery in ductility at the high temperature end of the trough in the coarse grained condition was shown to occur at a temperature close to that at which the base of the trough in the ductility curve intersected the curve εc versus temperature, i.e. when dynamic recrystallisation was possible but only for a strain rate of 3 × 10-3 s-1. At higher strain rates, the reduction in the rate of grain boundary sliding was sufficient to improve ductility without the necessity for dynamic recrystallisation. At lower strain rates cracks were able to develop to such a degree that dynamic recrystallisation was not effective in improving ductility. Refining the grain size eliminated the trough for all the strain rates examined. In this case the calculated curve εp or εc versus temperature intersected the hot ductility curves at temperatures below the range examined, indicating that dynamic recrystallisation always occurred. It was concluded that dynamic recrystallisation can have a major influence in restoring ductility at the high temperature end of the trough. However, it must often be well advanced to be effective and ductility can recover without the necessity for dynamic recrystallisation, by increasing the strain rate, thus reducing the amount of grain boundary sliding. © 1999 IoM Communications Ltd.

The steels C-Mn-Al and C-Mn-Nb-Al have been chosen so that their compositions would give a wide range of Ae3 and Ar3 (undeformed) temperatures. The steels were heated to 1330°C and cooled at 60 K min-1 to test temperatures in the range 1050 to 600°C and strained to failure at a strain rate of 3 × 10-3 s-1. In all cases troughs were obtained in the curve of reduction of area versus test temperature, the width varying markedly with composition. Although the Nb containing steel with the lowest transformation temperature had the widest trough, the depths of the troughs were similar. This relative insensitivity of the depth of the trough to the presence of Nb is believed to be owing to the high P levels in these steels reducing the amount of Nb(CN) precipitated in the region of the grain boundaries. Below the Ae3 temperature, the troughs were owing to the presence of a thin film of deformation induced ferrite allowing strain concentration to occur around MnS inclusions. Above the Ae3 temperature continuation of the trough was caused by grain boundary sliding in the austenite. Recovery of ductility at the high temperature end of the trough corresponded to the onset of dynamic recrystallisation and this was delayed in the Nb containing steels so that the trough was extended to higher temperatures. Recovery of ductility at the low temperature end always corresponded to the presence of a large amount of ferrite (∼50%) in the structure. In some cases this required having the ferrite present before deformation, as in the lower transformation steels having 1.4%Mn and 0.085%C, but for the higher transformation steels having 0.6%Mn and 0.15%C, deformation induced ferrite formed very readily so that ductility recovered just below Ae3. For these lower Mn and C steels, this ability for the thin films of ferrite to progress into the matrix giving enhanced ductility may be associated with the presence of a high Si level (0.5%). The commercial implications of the work are discussed. © 1998 the Institute of Materials.

An attempt has been made to predict the tensile hot ductility curves for simple C-Mn steels heated into the austenite phase to produce a coarse grain size and cooled to test temperatures in the range 700-1100°C. The depth of the trough can be calculated from a knowledge of the inclusion volume fraction and the volume fraction of the ferrite films surrounding the γ grains. The temperature at which ductility recovers at the high temperature end can also be determined from the point at which the base of the trough intersects the curve of peak strain v. temperature. This temperature is that at which dynamic recrystallisation becomes possible. Full recovers in ductility at the low temperature end of the trough occurs when large amounts of ferrite are formed; this is either 30-40 K below the Ar3 (undeformed) or the Ae3 depending on whether ferrite is present before or forms during deformation. The relevance of the hot ductility curve to the problem of transverse cracking is also discussed. © 1998 the Institute of Materials.