User:Alisterhaz/sandbox
Tasmanian alpine vegetation communities and the effect of fire
Fire appears to be a rare event in the Tasmanian alpine vegetation, suggesting that its effects might be more permanent than in most lowland vegetation types of south-eastern Australia [1].There have been suggestions that despite the infrequent occurrence of fires in the alpine zone especially historically and pre-European settlement, that the Australian alpine biota is resilient to infrequent large fires [2]. In their study for the paper ‘Decades-scale vegetation change in burned and unburned alpine coniferous heath’ Kirkpatrick et al (2010) demonstrate that several dominant species of Tasmania’s alpine vegetation (ie. endemic conifers and Nothofagus gunnii) have in fact a very low resilience to a single fire even though most of the associated species assemblage have a moderate to high resilience to fire events.
Tasmanian alpine vegetation
The true alpine zone for vegetation is defined as the space between the climatic treeline and the permanent snow and ice (nival zone). The climatic limit for the growth of trees varies a little species that form the tree line, but it has a proximity across the world to an imaginary line that joins places with a mean temperature of the warmest month of 10°c [3]. In many parts of the world for example the European Alps there is a very abrupt boundary between the forest and the alpine vegetation which corresponds with this 10°C isotherm.
In Tasmanian alpine regions this abrupt treeline rarely exists where often eucalypt forest or sub-alpine rainforest gradually merge into alpine heath with no real distinct sharp boundary. The climatic treeline can vary in Tasmania from 750 m in the south-west to around 1400 m in the north-eastern mountains indicating that climate and the harsh glazing winds can have an influence on the treeline boundary also (Crowden). In the eastern mountains of Tasmania a boundary can roughly be defined by the lower altitudinal limits of the subalpine woodlands of Eucalyptus coccifera, E.gunnii and E.archeri [4]. The alpine vegetation in Tasmania is therefore defined as the mostly treeless vegetation communities above the climatic treeline of the subalpine Eucalyptus woodlands which itself contains a diverse range of vegetation types including coniferous heath, deciduous heath, bolster cushion heath, alpine heathland and alpine sedgeland.
The alpine vegetation is unusual in a global context in that it is largely dominated by small and hard-leaved scleromorphic shrubs and hard cushion (bolster) plants as opposed to the normally herb and/or grass dominated alpine vegetation found elsewhere in the world [5].
In Tasmania, the alpine vegetation communities exist as more than 50 isolated habitat islands separated by floristically diverse lowland forest and sedgeland [6]. Approximately 300 or more of the 1500 vascular plants species of Tasmania occur above the climatic treeline in the alpine zone. Only 10 species are confined to the alpine zone but many others which are not obligate to the alpine zone have their core distribution there (Crowden). The alpine zone species exhibit a high level of endemism with almost half of the 300 species that occur there being endemic to Tasmania and in turn accounting for half of the total number of endemic plant species to Tasmania [7].
Description of alpine vegetation in Tasmania
(*not all vegetation communities are described – focus is on communities most affected by fire in the alpine zone)
Alpine coniferous heath (HCH Tasveg)
Described as long unburnt alpine heathland containing a significant cover of dwarf conifers, with other alpine heathland and graminoid species. There may be scattered emergent alpine Eucalyptus species and/or copses or scattered single trees of Athrotaxis species (Athrotaxis cupressoides and Athrotaxis selaginoides) and/or Nothofagus gunnii. Coniferous heath only occurs in unburnt, high-altitude areas on exposed, well-drained, gentle slopes on dolerite with high rainfall.
This community typically has many species in common with adjacent heathlands from the Proteaceae, Ericaceae and Asteraceae families, but is characterized by large numbers of any of the dwarf conifer species: Microcachrys tetragona, Microstrobus niphophilus, Diselma archeri and Podocarpus lawrencei. Microcachrys tetragona is the most common dwarf conifer, forming prostrate mats between other low growing shrubs, herbs and grasses. Diselma archeri may form scattered emergent. Microstrobus niphophilus is confined to a few small areas on dolerite plateaus, usually in association with Diselma archeri and Microcachrys tetragona. Podocarpus lawrencei may also be present.[8]
Associated heathland species commonly include Orites acicularis, Leptecophylla juniperina, Planocarpa petiolaris, Richea sprengelioides, Pentachondra pumila, Leptospermum rupestre, Boronia citriodora, Ozothamnus rodwayi, Helichrysum milliganii and Poa Gunnii. [9]
Athrotaxis cupressoides open woodland (RPW Tasveg)
Occurs on dolerite generally above 1000m elevation in which Athrotaxis cupressoides 5 to 8m tall forms small woodland patches or appears as copses or small scattered trees through alpine heathland and along creek and lake margins. Athrotaxis cupressoides may be the only tree species present over a generally heathy understory with the occasional emergent Eucalyptus coccifera and/or Nothofagus gunnii. It seems able to withstand frost and icy winds and it may be found on all but the most exposed heights on dolerite mountains with a long fire-free history. Athrotaxis cupressoide may form fringes around the base of steep rocky slopes or on dolerite block fields where it occurs with dwarf coniferous species such as Podocarpus lawrencei and Dislema archeri along with a range of other heath species such as Orites revoluta, Orites acicularis, Planocarpa petiolaris, Bellendena Montana, Leptospermum rupestre, Olearia pinifolia, Richea sprengelioides, Richea scoparia and Tasmannia lanceolata.[10]
Deciduous heath - Athrotaxis cupressoides-Nothofagus gunnii short rainforest (RPF Tasveg)
Dense closed montane short forest and scrub in which Athrotaxis cupressoides emerges above a tangle of Nothofagus gunnii. Generally occurs on but not confined to dolerite geology and between 900 and 1200 m elevation in a range of fire-protected sites, including head walls or cirques, south-facing slopes, cliffs and scree fields, and the edges of tarns, creeks and lakes [11].
The understory is a dense tangle dominated by Nothofagus gunnii over a coniferous heathy understory that can include Diselma archeri, Microstrobos niphophilus, Microcachrys tetragona, Orites acicularis, O.revoluta, Richea scoparia, R.pandanifolia, R.sprengelioides and Tasmannia lanceolata. Small open patches also occur where Gleichenia alpina, Empodisma minus, Astelia alpina and Eurychorda complanata are prominent [12].
Bolster cushion moorland (HCM Tasveg)
Cushion moorland is treeless alpine vegetation where greater than 50% of the ground is covered by cushion plants of at least one of the five main bolster heath species (ie. Donatia nova-zelandiae, Abrotanella forsteroides, Pterygopappus lawrencei, Dracophyllum minimum, Phyllachne colensoi and Schizacme archeri). It is usually found in poorly-drained,high rainfall areas on peaty soils above around 1000 m elevation.
Some cushion species, most notably Abrotanella forsteroides recover strongly after fires (ie. on Mt Field, Mt Wellington and eastern central plateau), but other species develop best if left unburnt for long periods of time. On the eastern central highlands above 1100 m the cushion moorland has regrown since extensive fires in 1962 [13].
Alpine heathland (HHE Tasveg)
Alpine heathland occurs on rocky well-drained gentle to steep slopes over dolerite or sedimentary rock. Drainage, exposure and fire history usually determine the dominant species with both Orites revoluta and Orites acicularis often being the most prominent species along with a suite of species such as Leptecophylla juniperina, Cyathodes straminea, Boronia citriodora, Leptospermum rupestre, Baeckea gunniana and Epacris serpyllifolia. More rocky areas are dominated by Richea sprengelioides along with Exocarpus humifusus, Coprosma nitida and Planocarpa petiolaris. As drainage decreases, Richea scoparia becomes the dominant and the community quickly merges from heathland to sedgeland (HSE) [14].
Alpine sedgeland (HSE Tasveg)
Found in alpine areas throughout the dolerite and sedimentary bedrock areas of central, eastern and southern Tasmania between 700 and 1200 m on poorly-drained flats where frost and poor-drainage may limit tree growth.
Alpine sedgeland is a fairly uniform, dense community up to 1 m in height. It may be dominated by any of Gleichenia alpina, Astelia alpina, Baloskion austral or a combination of all three. Empodisma minus is present at most sites. Emergent shrubs are not common but may include Baeckia gunniana and Richea scoparia and there are other scattered shrubs of Orites species, Ozothamnus species, and Richea species.
Alpine sedgelands are often found in mosaics with alpine heathland or fringing Athrotaxis cupressoides woodland [15].
Effect of fire on alpine vegetation in Tasmania
Fire appears to be a rare event in the Tasmanian alpine vegetation, suggesting that its effects might be more permanent than in most lowland vegetation types of south-eastern Australia [16].
Falling embers from hot wild-fires in lowland forests have probably always been a cause of spot-fires in alpine regions. However, it is certain that the damage to the alpine vegetation has increased significantly since European settlement, due to the burning practices of graziers operating leaseholds on the high country [17] as well as from regeneration burns from surrounding clearfelling forestry operations (Reid personal comment 2013)[18]. One fire in the summer of 1961-62 burnt nearly 128 000 ha with large proportions of the Athrotaxis forest and coniferous shrubberies being killed [19]. Between 1960 and 1980 16.1% of alpine vegetation in Tasmania burned in wildfires [20].
There have been suggestions that despite the infrequent occurrence of fires in the alpine zone especially historically and pre-European settlement, that the Australian alpine biota is resilient to infrequent large fires [21]. In their study for the paper ‘Decades-scale vegetation change in burned and unburned alpine coniferous heath’ Kirkpatrick et al (2010) demonstrate that several dominant species of Tasmania’s alpine vegetation (ie. endemic conifers and Nothofagus gunnii) have in fact a very low resilience to a single fire even though most of the associated species assemblage have a moderate to high resilience to fire events.
Past studies on the effect of fire on alpine vegetation in Tasmania
There have been relatively few temporal studies of succession after disturbance in alpine vegetation especially in Tasmania [22]. The two most significant studies were carried out by Kirkpatrick and Dickinson in 1984 focusing on the effect of fire on Tasmanian alpine vegetation and soils and by Kirkpatrick et al in 2010 looking at the decades-scale vegetation change in burned and unburned alpine coniferous heath. Both of these papers attempt to measure the response to fire in the floristic composition of the community, the structure of community, the resilience of endemic dominant species of coniferous heath to fire and the stages of temporal succession in the vegetation community up to 69 years post-fire.
Species regeneration classes
Kirkpatrick and Dickinson (1984)[23] developed the notion of species regeneration classes of particular species based on characteristics of the species such as their mode of reproduction (ie. seed or vegetatively), and how they responded to a fire in terms of the speed of regeneration in burned plots.
| Species regeneration class | Characteristic of species and how they respond to disturbance such as fire | Examples of species from regeneration class |
|---|---|---|
| 1 | Most often absent from recently burned plots. Adult trees are readily killed by fires and the seed seldom survives. | All gymnopserms ie. Athrotaxis cupressoides, Microcachrys tetragona, Nothofagus gunnii and Orites milligani |
| 2 | Found in recently burned areas with a much lower cover than adjacent unburned areas. Regenerates from seed. | Shrub and herb species from the Proteaceae and Ericaceae families such as Orites acicularis, Bellendena Montana, Richea scoparia, Cyathodes spp, Epacris serpyllifolia, Sprengelia incarnata and from other families such as Pimelea sericiea and Coprosma nitida |
| 3 | Recovers vegetatively yet does not attain the cover exhibited in unburned areas. | Usually shrubs and herbs. Nothofagus cunninghamii, Orites revoluta, Richea spregeliodies, Empodisma minus and Oreobolus pumilio. |
| 4 | Attains greater cover in recently burned areas than in unburned areas. Establishes by seed. | Many members of the Asteraceae family such as Olearia spp. and Helichrysum spp, as well as members of the Cyperaceae (Carpha spp., Uncinia compacta, Schoenus calyptratus) and Poaceae families (Austrodanthonia spp., Agrostis spp.). Another important species from this group is Bauera rubioides |
| 5 | Attains greater cover in recently burned areas than in unburned areas. Establishes vegetatively. | Members of the Poaceae family (Poa gunnii, Microlaena tasmanica), the Cyperaceae family (Oreobolus acutifolius, Oreobolus distichus) and Astelia alpina which tends to dominate post burn in some alpine situations. |
The nature of succession and long-term vegetation change in Tasmanian alpine vegetation communities
Using the notion of species regeneration classes (table 1) and what they measured in the field, Kirkpatrick and Dickinson (1984) developed a model of successional phases in the Tasmanian alpine vegetation for Mt Field and Mt Read.
Species regeneration class 3,4 and 5 were the first to recover and were present within a year after the burn and often dominating until at least 20 years post-burn.
After 20 years there was sufficient time to allow the recolonisation of some group 1 species. The most successful of those were the bird dispersed species of Podocarpus lawrencii and Microcachrys tetragona while the least successful were the wind dispersed species of Nothofagus gunnii and Athrotaxis cupressoides[25].
10 to 20 years post fire
Where bolster heaths (ie. Donatia novae-zelandiae) have high cover in coniferous heath or other heath communities in the surrounding unburned vegetation, they are generally dominant in the burned area, having the ability to reproduce vegetatively.
Where bolster heaths are absent, grassland, herbland or shrubland replaces coniferous and deciduous heath with the species Astelia alpina and Helichrysum backhousii being the most frequent dominants in the herbland and shrubland respectively[26].
20 to 40 years post fire
Between 20 to 40 years post-burn the group 2 species (species from the Proteaceae and Ericaceae families) establish themselves and become the more dominant vegetation in an alpine heath community along with some of the more dominant long lived group 3,4 and 5 species such as Bauera rubioides, Astelia alpina and Orites revoluta[27].
Grasses and herbaceous species decreased in abundance during this time due to the availability of open bare ground decreasing over time (from 31.3% cover in 1981 to 19.2% in 1998 to 17.3% cover in 2010 at Mt Field and Mt Read) and also being shaded out by the growth of larger heath shrubs [28].
40 to 50 years post fire
In some alpine areas burned greater than 40 years ago there is no invasion from directly adjacent stands of group 1 species. This implies that the successional process requires the senescence of group 2 species which dominate the alpine vegetation 40 to 50 years post-fire before the group 1 species such as the coniferous heath dominants can proceed in short steps to reestablish themselves in the burned plots[29].
Vegetative spread did occur at the edge of the fire scare or boundary between burned and unburned plots, but at an exceptionally slow rate such as 1m from the fire boundary [30] Naturally, this required a patch of unburned coniferous heath directly adjacent to the burned patch for vegetative regeneration to occur.
Temporal sequence of dominance
Kirkpatrick et al (2010) further developed the idea of succession based on their data with a temporal sequence of dominance depending on the hydrology and geology of the site.
On poorly drained sites the temporal sequence of dominance was:
mosses⇒graminoids⇒herbs & grasses⇒bolster dominance⇒gymnosperms dominance
On well-drained sites the temporal sequence of dominance was:
mosses⇒graminoids⇒herbs & grasses⇒shrubs (Proteaceae & Ericaceae) ⇒gymnosperms dominance
Data suggests that the upright shrubs of Proteaceae, Ericaceae and Asteraceae replaced the cushion plants in the temporal sequence of dominance when the ground was well drained [31]
Discussion on the impact of fire on alpine vegetation
(Based on the two studies Kirkpatrick and Dickinson (1984) and Kirkpatrick et al (2010))
Several dominant species of Tasmanian alpine vegetation have a very low resilience to a single fire, although most of the more common associated plant species have a moderate to high level of resilience to a single fire event[32]. There is no evidence of a ‘fire-requiring’ element for the vegetation of the Tasmanian alpine region especially for the coniferous and deciduous heath communities [33].
Where bolster cushion plants are absent, grassland, herbland and shrubland replaces coniferous and deciduous heath 10 to 20 years post fire with Astelia alpina and Helichrysum backhousii being the most frequent dominants in the herbland and shrubland respectively [34].
The mean species richness (alpha diversity) was higher in burned plots on all mountains measured, but fire eliminates or reduces much of the high mountain endemic shrub elements of the coniferous and deciduous heath while favouring more widespread species not only limited to the alpine environment from the families Proteaceae, Ericaceae and Asteraceae. This effect reduces spatial heterogeneity and leads to accelerated erosion of shallow alpine soils [35].
Regeneration rates of many of the endemic dominant species of alpine vegetation are extremely slow. Unless there is a dramatic change to regeneration rates and modes of succession, it will be many centuries before the gymnosperms will attain their pre-fire cover in the coniferous heathlands of alpine Tasmania [36]. In fact, there are many examples of where the endemic Gymnosperm species and Nothofagus gunnii are totally absent from many alpine habitat islands even though they are within their climatic and edaphic range, and populations exist on adjacent mountains, disappearing from the pollen records after the evidence of fire [37].
The recovery and succession of the vegetation of alpine Tasmania does not appear to accelerate with time, or not at least in the first 69 years post-fire. Gymnosperms increased from 1.3% cover (1981) to 2.1% (1998) to 2.4% (2010). In Ericaceae and Proteaceae shrubs the cover increased from 3.0% (1981) up to 17.4% (1998) before slowing down and dropping to 15.6% cover (2010). In burned coniferous heath the rate of recovery may have slowed because of the lack of bare ground to colonise which reduced over the time of the study from 31.3% (1981) to 19.2% (1998) to 17.3% (2010) [38].
Recovery will be limited to the area within the dispersal range of gymnosperms and deciduous beech. Kirkpatrick et al (2010) found that the dispersal range for Athrotaxis cupressoides was around 100m from the parent plant.
Figure 1(left). The location of individuals of Athrotaxis cupressoides around a single survivor of fire unknown age in subalpine vegetation, Mt Field NP Tasmania [39]
Other species such as Podocarpus lawrencei and Microcachrys tetragona which are also species regeneration class 1 but which possess fruits with sweet and fleshy arils dispersed by birds did not fare all that much better in dispersing widely with a range of only 70 m from the parent tree and a maximum of 130 m[40].
Very few Tasmanian alpine species lack all of the ability to recover vegetatively, store viable seed, and disperse rapidly into burned areas. However, those that do lack all of those attributes to re-invade a burned alpine environment are in fact the major dominants of coniferous heath and deciduous heath communities [41].
Example: Fire scar above Backhouse Tarn on the Tarn Shelf, Mt Field NP
Our methodology
In order to measure species richness of a burned area of coniferous heath/Athrotaxis cupressoides woodland compared to an adjacent unburned area we undertook the following:
A transect line was run 15 m into the unburned coniferous heath/Athrotaxis cupressoides woodland starting at the fire boundary to measure the unburned vegetation.
A second transect line was run 15 m upslope away from the fire boundary to measure species richness of the burned vegetation 50 years post-fire.
Results
| Unburned Athrotaxis cupressoides woodland and coniferous heath | Fire scar – burned in 1961-1965 |
|---|---|
| Description: Trees approximately 8 to 10m in height with a dense shrubby heath understorey | Description: Heathy shrubs dominant 0.5 to 1m in height |
| Athrotaxis cupressoides | Orites acicularis |
| Microstrobus niphophilus | Orites revoluta |
| Diselma archerii | Bauera rubioides |
| Richea pandanifolia | Richea scoparia |
| Leptospermum rupestre | Leptospermum rupestre |
| Tasmannia lanceolata | Rubus gunnianus |
| Orites revoluta | Archerii comberi |
| Astelia alpina | Astelia alpina |
| Sprengelia incarnata | Epacris serpyllifolia |
| Coprosma nitida | |
| Euphrasia gibbsiae | |
| Oreobolus sp. | |
| Empodisma minus | |
| Sprengelia incarnata | |
| Richea sp. (hybrid) | |
| Tasmannia lanceolata | |
| 9 species in total (measured in a quick 10 minute survey) | 16 species in total (measured in a quick 10 minute survey) |
Discussion from fire scar above Backhouse Tarn, Mt Field NP
No regenerating gymnosperms were observed in our randomly selected transect. However from 10 years ago Athrotaxis cupressoides along with other dominant conifers of the coniferous heathland began to appear as regenerating seedlings in the fire scar adjacent to unburned areas (Reid personal comment 2013).
_seedling_in_fire_scar.jpg)
This indicates that it was only beyond 50 years post-fire that the dominant gymnosperm species that for the coniferous heath began to regenerate, which follows the ‘nature of succession model’ that Kirkpatrick and Dickinson (1984) suggested[42]. As there was an absence of bolster cushion plants in the transect zone the period between 20 and 40 years post-fire as suggested by the ‘temporal sequence of dominance’ Kirkpatrick et al (2010), the vegetation that was surveyed was dominated by tall upright shrubs of the Proteaceae and Ericaceae families that formed a dense heathland.
Mean species richness (alpha diversity) was greater in the burned transect than the unburned. There was also greater habitat diversity with more bare ground and inter-shrub space in the burned transect than the unburned which had close to 100% cover and a dense almost impenetrable heath understory.
Given the proximity to adjacent unburned coniferous heath and Athrotaxis cupressoides woodland, it is anticipated that the burned plots will follow the temporal succession model of Kirkpatrick and Dickinson (1984) with class 1 species such as the endemic and previously dominant conifer species establishing between 40 and 50 years post fire, following the senescence of some of the class 2 species.
Conservation of alpine vegetation in relation to fire
Many recent fires in alpine areas have been started by recreational users such as hikers and fisherman. Kirkpatrick (1997) argues that a policy of fuel stove only for all alpine areas (not just the Tasmanian Wilderness World Heritage Area) is necessary to reduce the threat that warm-season recreational users pose to Tasmanian alpine environments[43]. Many fires have and will burn the alpine areas from land use practices below such as escapes from regeneration burns in surrounding eucalypt forests, such as the fire that burned the coniferous heath at Backhouse Tarn, Mt Field NP (Reid personal comment 2013)[44]. Kirkpatrick (1997) suggests that regeneration burns for forestry practices should be timed so that the alpine vegetation is too wet to burn and that resources are made available to prevent any escape fires from burning into the alpine zone[45].
There is no ‘fire requiring’ element of the alpine vegetation and it will take several hundreds of years (or perhaps a thousand or more) before the coniferous heath and Athrotaxis cupressoides woodlands will attain their pre-fire condition [46], [47]. It is strongly recommended that all management actions relating to Tasmania’s alpine vegetation aim to exclude fire from long unburned areas of alpine areas to maintain the spatial heterogeneity of the vegetation communities and to ensure the long-term survival of the unique and endemic flora that finds it’s core distribution in Tasmanian alpine environments.
References:
Balmer J (1991) Chapter 9 in 'Tasmanian Native Bush: Management Handbook', edited by Kirkpatrick JB, Tasmanian Environment Centre, Hobart.
Crowden RK (1999)Chapter 15 in 'Vegetation of Tasmania', edited by Reid JB, Hill RS, Brown MJ and Hovenden MJ, Commonwealth of Australia, Canberra.
Harris S and Kitchener A (2005). ‘From Forest to Fjaeldmark: Descriptions of Tasmania’s Vegetation’. Department of Primary Industries, Water and Environment, Printing Authority of Tasmania. Hobart
Jackson WD (2000) 'Nutrient stocks in Tasmanian vegetation and approximate losses due to fire', Royal Society of Tasmania vol 134
Kirkpatrick JB (1997) ‘Alpine Tasmania: an illustrated guide to the flora and vegetation.’ Oxford University Press: Melbourne
Kirkpatrick JB, Bridle KL and Dickinson JM (2010) 'Decade-scale vegetation change in burned and unburned alpine coniferous heath.'Australian Journal of Botany 58, 453-462
Kirkpatrick JB, Bridle KL and Wild AS (2002) 'Succession after fire in alpine vegetation on Mount Wellington, Tasmania.' Australian Journal of Botany 50, 145-154
Kirkpatrick JB and Dickinson KJM (1984) ‘The impact of fire on Tasmanian alpine vegetation and soils.’ Australian Journal of Botany 32, 613-629
All photos are the authors own (Hazeldine 2013) unless photo credit is given.
- ^ Kirkpatrick et al (2010)
- ^ Kirkpatrick et al (2010)
- ^ Kirkpatrick (1997)
- ^ Kirkpatrick (1997)
- ^ Kirkpatrick (1997)
- ^ Kirkpatrick et al (2010)
- ^ Kirkpatrick (1997)
- ^ Harris S and Kitchener A (2005)
- ^ Harris S and Kitchener A (2005)
- ^ Harris S and Kitchener A (2005)
- ^ Harris S and Kitchener A (2005)
- ^ Harris S and Kitchener A (2005)
- ^ Harris S and Kitchener A (2005)
- ^ Harris S and Kitchener A (2005)
- ^ Harris S and Kitchener A (2005)
- ^ Kirkpatrick et al (2010)
- ^ Jackson(2000)
- ^ Reid(2013) personal comment
- ^ Jackson(2000)
- ^ Kirkpatrick & Dickinson (1984)
- ^ Kirkpatrick et al (2010)
- ^ Kirkpatrick et al (2010)
- ^ Kirkpatrick & Dickinson (1984)
- ^ Kirkpatrick & Dickinson (1984)
- ^ Kirkpatrick and Dickinson (1984)
- ^ Kirkpatrick and Dickinson (1984)
- ^ Kirkpatrick and Dickinson (1984)
- ^ Kirkpatrick et al (2010)
- ^ Kirkpatrick and Dickinson (1984)
- ^ Kirkpatrick et al (2010)
- ^ Kirkpatrick et al (2010)
- ^ Kirkpatrick et al (2010)
- ^ Kirkpatrick and Dickinson (1984)
- ^ Kirkpatrick et al (2010)
- ^ Kirkpatrick & Dickinson (1984)
- ^ Kirkpatrick et al (2010)
- ^ Kirkpatrick et al (2010)
- ^ Kirkpatrick et al (2010)
- ^ Kirkpatrick and Dickinson (1984)
- ^ Kirkpatrick and Dickinson (1984)
- ^ Kirkpatrick & Dickinson (1984)
- ^ Kirkpatrick and Dickinson (1984)
- ^ Kirkpatrick(1997)
- ^ Reid(2013) personal comment
- ^ Kirkpatrick(1997)
- ^ Kirkpatrick and Dickinson (1984)
- ^ Kirkpatrick et al (2010)