Publications
Jill Landsberg, Craig D. James, Stephen R. Morton, Trevor J. Hobbs, Jacqui Stol, Alex Drew and Helen Tongway
CSIRO Division of Wildlife and Ecology
January 1997
Published by Environment Australia and CSIRO
© Commonwealth of Australia, 1997
ISBN 0 642 27010 4
Increaser and decreaser response groups were determined separately for each gradient and each plant group. Full listings of the response groups determined for all the species recorded at each gradient are given in Appendix 4. Of the total of 673 plant species identified, many (around 75%) were recorded at one gradient only. They include annual and perennial grasses and forbs, shrubs and trees, from many families. There are too many species to consider their identities individually. If they could be classified by their functional attributes, rather than their taxonomic identities, we could determine which attributes increaser and decreaser species have in common. The development of such a scheme is desirable (Sections 4.2.3 and 4.3.3), but is outside the scope of this report.
Some species were recorded at two or more gradients, however. For this subset of species it is instructive to consider the identities of those that showed the same pattern of response on different gradients (e.g those that were classified as decreasers on each gradient where they occurred) with those that showed different patterns of response on different gradients (e.g. those that were classified as decreasers on some gradients and increasers on others).
Around 130 of the 478 plant species growing in the understorey were recorded at two or more gradients. Moderate numbers were classified as decreasers on at least two gradients and were not classified as increasers at any gradient, although some were classified "not determined" at some gradients (Table 3.4.1.1). These "repeat decreasers" included forbs, grasses and shrubs from many families (Appendix 4). Some, such as the grasses Enneapogon polyphyllus and Monachather paradoxa, are known to be highly palatable and prone to decrease under sustained grazing (Appendix 1; Cunningham et al. 1992). Others, such as woollybutt grass, Eragrostis eriopoda, and purple lovegrass, E. lacunaria, are more usually considered to be hardy and persistent under grazing and some, such as the copper burrs Dissocarpus paradoxus and Sclerolaena decurrens are considered to be seldom eaten by stock. Others, such as Centipeda thespidioides, are known to be palatable to stock (Cunningham et al. 1992) but have not been reported as liable to decline under grazing. Very few of the repeat decreasers are commonly recognised "indicator species" for range assessment (e.g. none is listed in the extensive lists presented by Pringle 1994). One of the repeat decreasers was an exotic but its abundance was very low. Many other species of repeat decreasers were also locally uncommon or rare, however; whether this is a cause or a consequence of their decreaser trends is a moot point.
We identified very few species that were increasers at several gradients and decreasers at none (Table 3.4.1.2). Two of the five repeat increasers, Carrichtera annua and Schismus barbatus, are exotic weeds that have become widespread in the semi-arid regions of southern Australia (Section 2.2.3). They were also the most abundant repeat increasers (Appendix 4). One of the other repeat increasers is a common and widespread copperburr, Sclerolaena diacantha, which is considered one of the more valuable copper burrs for pastoralism because of its moderate palatability (in contrast to the reputed lower palatability of the three decreaser copper burrs), coupled with an ability to colonise eroded areas (Cunningham et al. 1992). The occurrence of Sclerolaena species in both increaser and decreaser response groups shows clearly that the characteristics associated with persistence under disturbance may be quite subtle and need not necessarily correlate very closely with taxonomic relationships.
Table 3.4.1.1 Understorey plant species that were repeat decreasers
ie., species that were classified as decreasers at two or more gradients and were not classified as increasers at any gradient.
Species Inc Dec ND Total Mn freq% *Malvastrum americanum 2 1 3 0.07 Abutilon malvifolium 2 1 3 0.07 Amaranthus mitchellii 2 2 0.415 Aristida obscura 2 2 1.69 Brachiaria gilesii 2 2 0.105 Bulbine alata 2 2 28.44 Calandrinia eremaea 3 1 4 1.043 Calotis cuneifolia 2 1 3 0.217 Calotis inermis 2 2 0.01 Centipeda thespidioides 2 3 5 0.25 Dactyloctenium radulans 3 3 0.973 Dissocarpus paradoxus 2 1 3 6.18 Enneapogon polyphyllus 2 2 4 2.245 Eragrostis eriopoda 2 3 5 0.512 Eragrostis lacunaria 2 2 0.02 Erodium cygnorum 2 2 0.105 Goodenia berardiana 2 2 0.21 Goodenia cycloptera 2 2 1.875 Lepidium oxytrichum 2 2 11.565 Lepidium phlebopetalum 3 2 5 0.796 Maireana villosa 2 2 0.53 Malacocera tricornis 2 2 0.01 Marsdenia australis 2 2 0.01 Monachather paradoxa 2 1 3 0.23 Portulaca filifolia 2 1 3 0.07 Portulaca oleracea 2 2 11.875 Rhagodia spinescens 2 1 3 0.07 Rhodanthe maryonii 2 2 0.01 Rhyncharrhena linearis 2 2 0.01 Schoenia ayersii 2 2 0.105 Sclerolaena decurrens 2 2 6.46 Sporobolus caroli 2 2 0.23 Triglochin calcitrapum 2 2 3.96 Triraphis mollis 2 2 1.77 Vittadinia eremaea 3 3 0.21
Inc = the number of gradients where classified increasers; Dec = the number where classified decreasers; and ND = the number where classified not determined. Total is the total number of gradients where recorded. Mnfreq% is the mean frequency across all gradients. *=Exotic species.
Table 3.4.1.2 Understorey plant species that were repeat increasers
ie., species that were classified as increasers at two or more gradients and were not classified as decreasers at any gradient. Column headings as in Table 3.4.1.1
Species Inc Dec ND Total Mn freq% *Carrichtera annua 2 2 4.01 *Schismus barbatus 2 2 11.04 Osteocarpum 2 2 0.05 dipterocarpum Sclerolaena diacantha 2 1 3 2.237 Solanum esuriale 2 2 1.165
Also of great significance, however, is the number of species that showed different response trends at different gradients (Table 3.4.1.3). Some of these species could have been mis-classified; this is particularly likely to occur when a species was locally rare or patchily distributed. For example, ruby saltbush, Enchylaena tomentosa, is a widespread shrub that is generally considered a decreaser species (e.g. Pringle 1994). It was classified in the decreaser group at three of the gradients where it occurred, and as an increaser at one gradient only, where it was locally rare (Appendix 4). Similarly, blue crowfoot, Erodium crinitum, a widespread herb, is generally considered highly palatable to stock (Cunningham et al. 1992) and therefore would be expected to show a decreaser response. It was classified as a decreaser at the four gradients where it was more common and an increaser at only one gradient, where it was locally rare (Appendix 4). In these instances it is possible that the apparent increaser trends in abundance may have arisen by chance or been due to responses to underlying environmental variations, rather than responses to distance from water.
For many of the species with mixed responses, however, their grouping as decreasers in one environment and increasers in another may be a real result of differences in environmental context. This is most likely to be the case where they were classified as decreasers in several gradients and increasers in several others (Table 3.4.1.3). The expression of plant traits and grazing animal preferences are both highly context-specific. For example, a grazing animal may prefer a plant species in one community (e.g. if it is the only perennial grass on offer) and avoid it in another (e.g. if it is a very fibrous grass in a community dominated by softer grasses). Similarly a plant species may be very rare and therefore vulnerable in one plant community (e.g. near the edge of its range) but very abundant and therefore more likely to persist in another. Thus just as taxonomic relationships (such as species belonging to the same genus) do not necessarily correlate with ability to persist in the face of disturbance, the ability of a species to persist in one environment does not necessarily correlate with its ability to persist in a different environment.
Table 3.4.1.3 Understorey plant species that were mixed responders
ie., species that were classified as decreasers at one or more gradients and were also classified as increasers at one or more gradients. Column headings as in Table 3.4.1.1
Species Inc Dec ND Total Mn freq% *Sonchus oleraceus 1 2 1 4 0.005 Abutilon otocarpum 2 1 1 4 0.058 Acacia aneura 2 1 1 4 0.213 Aristida contorta 2 2 1 5 3.42 Atriplex holocarpa 1 1 2 1.875 Atriplex vesicaria 1 2 1 4 9.895 Brachycome ciliaris 1 2 1 4 0.99 Calotis hispidula 2 3 2 7 0.654 Calotis plumulifera 1 1 2 9.48 Chamaesyce drummondii 1 2 3 6 7.257 Chenopodium cristatum 3 1 4 0.163 Chenopodium melanocarpum 2 3 1 6 0.007 Convolvulus erubescens 1 2 1 4 0.833 Crassula colorata 2 1 3 4.513 Daucus glochidiatus 1 4 2 7 0.239 Digitaria brownii 1 1 1 3 6.75 Dissocarpus biflorus 1 1 2 0.125 Enchylaena tomentosa 1 3 2 6 0.368 Enteropogon acicularis 2 1 3 0.007 Eragrostis dielsii 1 2 1 4 0.053 Eremophila gilesii 1 1 2 0.22 Erodium crinitum 1 4 1 6 0.972 Euphorbia eremophila 1 2 3 12.64 Fimbristylis dichotoma 1 1 1 3 0.417 Gnephosis arachnoidea 2 2 1 5 0.088 Hyalosperma semisterile 1 1 2 1.98 Nicotiana occidentalis 2 1 3 0.013 Nicotiana velutina 1 1 2 1.665 Pimelea simplex 1 1 2 0.335 Plantago turrifera 1 2 1 4 0.52 Ptilotus gaudichaudii 1 2 3 3.487 Ptilotus helipteroides var. 1 1 2 0.105 helipteroides Ptilotus obovatus 2 2 1 5 0.004 Rhodanthe floribunda 1 4 1 6 3.618 Rhodanthe stricta 1 2 3 0.763 Salsola kali 1 4 2 7 0.274 Senna artemisioides artemisioides 2 1 3 0.277 Sida fibulifera 2 3 5 0.584 Solanum ellipticum 2 2 1 5 0.066 Solanum quadriloculatum 1 3 1 5 0.02 Sporobolus actinocladus 1 2 3 1.057 Tetragonia eremaea 1 1 2 2.51 Thysanotus baueri 1 1 2 0.52 Tragus australianus 2 1 3 0.277 Tripogon loliiformis 1 3 1 5 3.334 Zygophyllum simile 2 1 3 5.493
Several of the species identified as repeat decreasers or repeat increasers in the seedbank showed the same pattern as understorey plants growing in the field (Tables 3.4.1.4-5). For example, the grasses Enneapogon polyphyllus and Eragrostis eriopoda were identified as decreasers at more than one gradient in both seedbank and field (Tables 3.4.1.4 & 1). Similarly the exotic weed Carrichtera annua was consistently identified as an increaser in the seedbank and field at a number of gradients (Tables 3.4.1.5 & 2).
Table 3.4.1.4 Seedbank plant species that were repeat decreasers
ie., species that were classified as decreasers at two or more gradients and were not classified as increasers at any gradient. Column headings as in Table 3.4.1.1, except that abundance is shown by average number of individuals per gradient. U indicates species that were also classified as repeat decreasers growing in the understorey.
Species Inc Dec ND Total No. of indiv.. *Rostraria pumila 2 1 3 42 Aristida contorta 4 4 74 Aristida sp1620 2 1 3 2 Calandrinia balonensis 2 2 1 Calandrinia polyandra 2 2 8 Calandrinia sp2712 2 2 1 Calocephalus ?knappi 2 2 38 UCalotis inermis 3 3 2 Crassula sieberiana 3 3 449 Crassula spC7 2 1 3 9 Cyperus sp2199 2 1 3 4 Digitaria brownii 2 2 19 UEnneapogon polyphyllus 2 2 263 UEragrostis eriopoda 3 1 4 5 Mollugo cerviana 2 2 16 Paspalidium rarum 2 2 1 Pseudognaphalium 2 1 3 7 luteoalbum Rhodanthe floribunda 2 1 3 13 Sclerolaena sp1891 2 2 11 unidentified sp2011 2 2 9 Wahlenbergia gracilenta 2 2 87 Zygophyllum spC2 2 2 2
Several of the species identified as repeat decreasers in the seedbank (e.g. Calandrinia balonensis, C. polyandra) are reported to be very palatable to livestock (Cunningham et al. 1992) and were not detected growing in the field. Others, such as Digitaria brownii, were classified as mixed responders in the field (Table 3.4.1.3) but are recorded as highly palatable and liable to decrease in degraded areas (Cunningham et al. 1992); thus their inclusion as decreasers in the seedbank is consistent with what is known of their persistence under grazing. Rhodanthe floribunda, however, has previously been reported as becoming abundant in heavily grazed areas (Cunningham et al. 1992), although our surveys suggest otherwise: it was classified as a decreaser in the understorey of four gradients and an increaser at only one gradient (Table 3.4.1.3); and was classified as a repeat decreaser in the seedbank (Table 3.4.1.4).
Table 3.4.1.5 Seedbank plant species that were repeat increasers
ie., species that were classified as increasers at two or more gradients and were not classified as decreasers at any gradient. Column headings as in Table 3.4.1.4. U indicates species that were also classified as repeat increasers growing in the understorey.
Species Inc Dec ND Total No. of Indiv.. *UCarrichtera annua 2 2 1 Bergia trimera 2 2 1113 Calotis hispidula 3 2 5 35 Centipeda thespidioides 2 2 5 Chenopodium cristatum 3 1 4 1 Chenopodium melanocarpum 5 5 2 Sclerolaena patenticuspis 2 2 35 Synaptantha tillaeacea 2 1 3 1
Several of the seedbank repeat increasers (e.g. Chenopodium cristatum C. melanocarpum, Sclerolaena patenticuspis) are reported to be unpalatable to stock and/or indicators of severely grazed sites (Appendix 1; Cunningham et al. 1992); their identification as increasers in the seedbank is consistent with this description. Bergia trimera, one of the species identified as an increaser in our seedbank studies, has also been identified as increasing in the seedbank at sites close to water in another grazing study (Navie et al. 1996). The classification of Centipeda thespidioides as a repeat increaser in the seedbank is puzzling, however, since it was classified as a repeat decreaser when growing in the understorey (Table 3.4.1.1). Its close relative C. cunninghamii is reported to be unpalatable to stock much of the time, but utilised to a moderate extent in dry times (Cunningham et al. 1992). Perhaps this is also the case for C. thespidioides. If so, the pattern of decreasing abundance in the field might reflect recent grazing patterns resulting from depletion and drying off of more preferred species, while its pattern of increasing abundance in the seedbank might reflect more long-term grazing patterns.
Only one species of shrub was classified in the same response group at different gradients (Appendix 4). It was Pimelea microcephala, Shrubby Rice-flower, and was classified as an increaser at the two gradients where it occurred. Although not generally considered as prone to increase it is known to be unpalatable to stock and very hardy. It has also been suspected of poisoning stock (Cunningham et al. 1992).
Among bird species, the Galah, Crested Pigeon, Yellow-throated Miner, and corvids (crows and ravens) were the most consistently identified increaser species, in both our own and previous studies (Table 3.4.2.1; Appendix 4; Curry and Hacker 1990; Saunders and Curry 1990), and were often very abundant within only 1 km of water. We also identified as increasers, a number of other species that have been reported as increasing in abundance in heavily grazed rangelands in Western Australia. Examples include Bourke's Parrot and Zebra Finch (Saunders and Curry 1990). However, Bourke's Parrot has also been identified as having undergone a pronounced decrease in abundance in New South Wales (Smith and Smith 1994) and responded as a decreaser at one of our Queensland gradients (Appendix 4). The Spiny-cheeked Honeyeater was another decreaser identified at this gradient, that has apparently increased in abundance and/or range in Western Australia (Saunders and Curry 1990). These and some other species that were only found on a few gradients, or where classed as increasers for some of the gradients on which they occurred (e.g., Australian Ringneck, Blue Bonnet, Common Bronzewing) are likely to be abundant close to water due to their dependence on drinking water.
Many of the species identified as having increased in abundance and/or geographic range in the arid zone since European occupation (Reid and Fleming 1992) were recorded during our surveys (Table 3.4.2.1). Most of these species were recorded as increasers or not determined, however this was not always consistent when the species occurred on a number of gradients. The occasional recording of one of these species in a decreaser group along one gradient may be a chance statistical occurrence, or an anomaly of the particular habitat of the gradient or the time of sampling.
(from Table 8 in Reid and Fleming 1992.) EX = excluded from analyses INC = increaser; DEC = decreaser; ND = not determined. Rare = 1-10; Common = 11-100; Abundant >100.
Species Number of Sites Abundance Response groups
gradients
Australian Kestrel 4 1-6 Common EX on 2; ND on 2
Banded Lapwing 2 1-2 Common INC
Common Bronzewing 5 1-6 Common INC on 3; ND on 1;
DEC on 1
Crested Pigeon 6 1-6 Abundant INC on 5; ND on 1
Bourke's Parrot 2 2-6 Common DEC
Galah 7 1-6 Abundant INC on 5; ND on 2
Spiny-cheeked 6 1-6 Abundant INC on 2; ND on 4
Honeyeater
Yellow-throated 5 1-6 Abundant INC on 4; ND on 1
Miner
White-plumed 1 1,2,6 Common INC
Honeyeater
Yellow-rumped 5 2-6 Common ND on 3; DEC on 2
Thornbill
Southern Whiteface 6 1-6 Abundant ND on 4; INC on 1;
DEC on 1
Willie Wagtail 6 1-6 Common INC on 3; DEC on 2;
ND on 1
Black-faced 7 1-6 Abundant ND on 5; DEC on 2
Woodswallow
Australian Raven 4 1-6 Common EX on 2; DEC on 1;
ND on 1
Little Crow 3 1-5 Common INC on 2; ND on 1
Torresian Crow 1 1,3,4,6 Rare ND
Richard's Pipit 4 1-6 Common ND on 3; INC on 1;
DEC on 1
Brown Songlark 3 1,3,5,6 Common INC on 1; DEC on 1;
ND on 1
Species of birds detected during the study that have been recognised as having declined or being at risk (Reid and Fleming 1992) are shown in Table 3.4.2.2. Many of these species were rare and so their abundance profile along the gradient could not be determined. Where the species were classified into a response group, they were mostly identified as decreasers. For those species that could not be classified, they were usually sighted on sites away from water (ie sites 4-6). Comparison of Tables 3.4.2.1 and 3.4.2.2 shows that birds that have been identified as having increased in abundance or geographic range generally occurred on more gradients and were more abundant than species that have been identified as rare or threatened, or having declined in abundance, or being at risk.
Table 3.4.2.2 Birds seen during the surveys that have been identified as rare or threatened
(Garnett 1992), or having declined in abundance or being at risk (from Table 6 in Reid and Fleming 1992). EX = excluded from analyses; INC = increaser; DEC = decreaser; ND = not determined. Rare = 1-10; Common = 11-100; Abundant >100.
Species Number of Sites Abundance Response groups
gradients
Little Button-quail 3 3-6 Rare EX on 2, ND on 1
Pink Cockatoo 1 5 Rare ND
Scarlet-chested 1 3-4 Rare ND
Parrot
White-winged 4 1-6 Abundant ND on 3, DEC on 1
Fairy-wren
Grey Honeyeater 1 6 only Rare DEC
Pied Honeyeater 1 4 Rare DEC
Redthroat 3 1-6 Common ND on 2, DEC on 1
Rufous Fieldwren 2 1-6 Common ND
Red-browed Pardalote 2 2-3 Rare ND and INC
Chiming Wedgebill 2 1-6 Common ND and DEC
Hall's Babbler 2 4-6 Common ND
Cinnamon Quail-thrush 2 1-6 Common ND and DEC
It is not surprising that species dependent on water were classed consistently as increasers because their current distribution and abundance throughout the arid and semi-arid zones is a direct reflection of the provision of artificial water. In contrast, species that are not dependent on regular drinks occurred throughout these landscapes before the advent of pastoralism and the provision of water. The factors influencing the distribution and abundance of species that do not depend on water are likely to be more complex. Hence, in these species we see less consistency among gradients in their classification in relation to distance from water.
Some of the factors that may influence the distribution and abundance of species in relation to a piosphere are: (1) the presence of aggressively-dominant bird species near a water point; (2) disturbance and opening-up of the habitat under heavy grazing; (3) changes to vegetation composition and structure; and (4) changes to food resources. No patterns of species displacement by known aggressively-dominant bird species (e.g., Yellow-throated miners - Grey 1996) were obvious in our gradient data, although displacement of some species possibly occurs. There was a suggestive anomaly in the bird data from the NSW mulga gradient. At this gradient the reference site was lightly grazed because it was a long way from the nearest water accessible to large grazing animals, but for birds, water was readily available across the dingo-proof border fence (Section 1.3.4). Yellow-throated Miners, which were abundant all along this gradient were particularly prominent at the reference site. Presumably this was in direct response to the availability of water, rather than as an indirect consequence of grazing impact. No other bird species were common at the reference; therefore it is possible that other species may have been displaced by the aggressively-dominant Miners.
The second and third factors incorporate a range of changes in the vegetation and ground surface and no common response to these is apparent (or expected). Ground-dwelling species have been suggested as one group at high risk from modification of their habitat because of grazing (Reid and Fleming 1992) and our data show that ground-feeders were a larger proportion of the decreaser species than they were for increaser species (Table 3.4.2.3) In contrast, canopy and shrub feeders were a larger proportion of the increaser species than they were for decreaser species. There was no predominance of taxa such as bush birds among the decreasers or pigeons and parrots among the increasers, nor is there any relationship apparent between the structure of the dominant vegetation on a gradient (woodland vs. shrubland) and the proportion of decreaser bird species identified there.
Foraging level % of decreasers % of
increasers
Aerial 8 2
Canopy and shrub 32 47
Ground 61 52
Total no. 59 60
species
As for birds, attributes pre-disposing some species of reptiles to decline are not readily apparent. In general, the clarity of the response groups for reptiles was not as obvious as it was for other plant groups and animal taxa. That is, many species were relatively uniformly spread along gradients, suggesting that they may be relatively insensitive to piosphere effects. James et al. (1995b) suggested that diurnal species closely associated with shrub leaf-litter, such as Morethia boulengeri and Ctenotus regius, may be disadvantaged by heavy grazing if the leaf-litter is disrupted. However, we could only detect a decline in leaf litter close to water at two gradients (NSW and Qld mulga gradients; Section 3.1) and neither species was identified among the decreaser species there or elsewhere (Appendix 4). There were no obvious associations between particular guilds (nocturnal vs. diurnal) or families of reptiles and membership of different response groups.
Many species of ants occurred on more than one gradient (Appendix 4), but only a few of these were repeatedly classified as increasers (or decreasers) at each of the gradients where they occurred (Table 3.4.4.1). Of the 11 species that were repeatedly classified in the same response group, eight were species that showed a decreaser pattern. In addition to the three species that were repeatedly classified as increasers (and never classified as decreasers; Table 3.4.4.1) several other species were predominantly classified as increasers, ie. they occurred on more than two gradients and were classified as increasers more often than not. They were Camponotus ephippium, C. spAF and C. spAH, Iridomyrmex agilis, I. galabanus, and I. spCJ, Rhytidoponera spAA, R. spAE, and Tapinoma spAA. Several other species of ants were predominantly classified as decreasers; they included Calomyrmex spAA, Cerapachys spAB and Leptogenys centralis.
Functional groupings of ants have been discussed in the literature on a number of occasions (Greenslade and Greenslade 1985; Andersen 1990, 1992, 1995a&b) and provide a convenient method of collapsing large numbers of species into a few genus-based groups from which behavioural and ecological inference can be drawn. Increaser and decreaser groups in mulga habitats were mostly composed of hot-climate specialists (Melophorus and Meranoplus) sub-dominant camponotini (Camponotus, Opisthopsis and Polyrachis) and generalised myrmicines (Monomorium, Pheidole and Crematogaster) (Table 3.4.4.2). In chenopod habitats the increaser and decreaser groups were dominated by different functional groups: hot-climate specialists and generalised myrmicines made up most of the species of increasers, and opportunists (Rhytidoponera, Tetramorium, and Odontomachus), generalised myrmicines and dominant dolichoderines (Iridomyrmex) made up most of the species of decreasers.
Percentages are calculated as the number of species in a functional group over the total number of species in the whole response group, and averaged across gradients. Functional group names are from Andersen (1992).
Mulga Chen
----------------- -----------------
INC DEC ND INC DEC ND
Dominant dolichoderines 13% 9% 14% 12% 21% 15%
(1)
Sub-dominant componotini 22% 24% 18% 7% 6% 10%
(2)
Hot climate specialist 36% 34% 37% 33% 18% 29%
(3a)
Cold climate specialist 0% 2% 0% 4% 0% 0%
(3b)
Cryptic and sub-cryptic 7% 4% 8% 9% 9% 5%
(4a, b)
Opportunists (5) 7% 18% 20% 16% 27% 17%
Generalised myrmicines 16% 12% 18% 20% 26% 24%
(6)
Solitary/ specialist 10% 4% 0% 5% 5% 3%
predators (7)
Based on functional groups, we may expect a number of patterns in the distribution of ant species along a gradient of disturbance by grazing. First, that dominant dolichoderines (Iridomyrmex spp) may be dominant in disturbed habitats (ie near water points) and exclude most other ants except the sub-dominant camponotini and some generalised myrmicines (Andersen 1995a; Scougal et al. 1993; Bestelmyer and Wiens 1996). This dominance does not appear to be manifested in higher species richness of Iridomyrmex in increaser groups (Table 3.4.4.2) but may be manifested in higher biomass of Iridomyrmex which has not been examined in this study.
Second, there may be a trend for more species of hot-climate specialists in increaser groups because of the more open habitat near water points. They were not more speciose in increaser groups in mulga habitats but were so in chenopod habitats. This is consistent with the changes in cover that occurred at these gradients: shrub cover tended to increase close to water at mulga gradients, but to decline close to water at chenopod gradients (see Section 3.1).
Third, opportunists are poor competitors and generally predominate in disturbed locations where there is structural simplification (Greenslade 1978) so they may be more speciose in habitats near water points. Our data contradict this prediction because they show that opportunists were more speciose in decreaser groups compared to increaser groups in both habitats. The presence of larger numbers of species of opportunists in decreaser faunas suggests that groups such as dominant dolichoderines, sub-dominant camponotini, and generalised myrmicines may exclude them from areas closer to water.
Finally, solitary and specialist predators have been found to be more speciose in less-disturbed sites (Bestelmyer and Wiens 1996). In our study they were more speciose in increaser groups than decreaser groups but only in mulga habitats.
Although many of the predictions for functional groups seem to be contradicted by our data, this analysis only looks at the relative proportion of species in each functional group and does not take into account the abundance of the species. It is not clear whether species richness or biomass is the main factor that responds to a gradient of disturbance or to interactions with other ants. It is likely that both species composition and the abundance of species are key structuring components and understanding how assemblages of ants respond to grazing disturbance will require further data collection (e.g., refining taxonomic and behavioural information) and analysis that is beyond the scope of this report. However, the data collected during this project will form the foundation of future analyses aimed at functional classification of ants to incorporate functional variation within genera, particularly for those species that repeatedly show the same responses to disturbance.