Roadside vegetation may provide examples of indigenous vegetation which has largely disappeared from the region (E.G. Stuwe and Parsons, 1977; Yugovic et al., 1985); it may include rare species of plants (Beauglehole, 1980 Appendix 5; Hussey, 1987); it may serve as a habitat or refuge for a wide range of faunal species (Way, 1977; Middleton, 1980; Krohn, 1981, Arnold et al., 1987; Newbey and Newbey, 1987) and it has the potential to serve as a corridor for the dispersal of wildlife (Getz et al., 1978; Breckwoldt, 1983).

At Naringal in south-western Victoria, clearing and fragmentation of the forests over the past 100 years has resulted in loss of tree cover from 90% of the region.

Presently, patches of forest less than 100 ha in size, loosely linked by strips of forested vegetation along roadsides and creeks, provide the only natural habitat for the native fauna.

As part of a broader study of the biogeography and conservation of mammals within the fragmented forest environment at Naringal (Bennett, 1987a, b), an investigation was made of the potential role of roadside strips of vegetation as corridors to facilitate the dispersal of wildlife between otherwise-isolated populations in forest patches. Initially, a survey was carried out to determine the mammals were present at a number of roadside sites varying in width, isolation and road usage. These results, supplemented with additional observations of mammals on roadsides are presented here to illustrate the range of mammal species which utilise roadside vegetation in this locality.

Roadside Vegetation at Naringal

Remnant forest vegetation occurs along many road reserves in the Naringal area as a narrow strip between the paved road surface and the adjoining fence-line, often forming an essentially continuous habitat, punctuated only by crossroads or gateways for farm access.

Along narrow secondary roads, where the width of the road reserve is approximately 20 m, there may be only a single row of trees together with scattered shrubs and field vegetation totalling less than 5 metres in width. However, along the wider reserves (of up to 60 m total width), the belt of vegetation may extend to 40 m in width. Strips of forest vegetation are also frequently present along "unmade" roads --- road reserves approximately 20 m in width which have been surveyed and fenced, but which lack a developed road surface.

The roadside vegetation is chiefly low open-forest or open-forest dominated by Eucalyptus Obliqua and E. Ovata, the latter species being present in poorly-drained situations.

Acacia melanoxylon and E. viminalis may also be present, particularly on volcanic soils in the north of the area.

At roadsides where native vegetation remains dominant, the shrub stratum is usually well represented, and the field stratum is dense.

In the better-drained situations, the canopy is dominated by E. Obliqua, and common plant species in the understorey include:

Acacia myrtifolia, A.stricta, A.verticillata, Banksia marginata, Gonocarpus tetragyna, Helichrysum dendroidium, Lomandra longifolia, Leptospermum juniperinum, Pteridium esculentium, Tetrarrhena juncea and Tetratheca ciliata.

In wetter, poorly drained situations, E. Ovata is the dominant canopy tree, and the understorey may include:

Juncus spp. Lepidosperma laterale, L. juniperinium, Lepyrodia muelleri, Melaleuca squarrosa and Viola hederacea.

Introduced weeds and grasses are common along the roadsides, particularly at the edge of the forested strip.

Prominent introduced plants include:

Anthoxanthum odoratum, Holcus lanatus, Hypochierus radicatus, Paspalum dilatatum, Sporobolus africanus and Trifolium spp.

Mammalian fauna using roadside vegetation:

Observations of mammals using roadside vegetation at Naringal were from two sources, as described below

1. Survey of roadside sites.

Fifteen roadside sites were surveyed for the presence of mammals (Fig. 1).

Severely disturbed vegetation dominated by introduced grasses and avoided; but otherwise, sites were selected to represent the range of road types and a range of distances from the nearest forest patch. All sites had experienced past disturbance from tree felling or clearing of forest vegetation, and from deliberate burning.

Most sites had a regenerating canopy with trees estimated to be less than 40 years old; and at two sites the vegetation was regenerating from an apparent loss of trees within the previous 10 years.

A standard total of 60 trap-nights was recorded for each site.

The identity, sex, and weight of each captured animal were noted before it was released at the point of capture.

Nocturnal observations were made on two occasions at each site (Approximately 0.2 hrs. on each occasion), by walking slowly along the road and searching the canopy and shrub strata with a 50-watt spotlight. Daylight searches were made for characteristic signs such as diggings, tracks, faecal remains or skeletal material.

2. Other Observations

Incidental observations (1979-82) at other roadsides throughout the study area, and additional records from further trapping (1980-1982) at four of the survey sites, also contributed to a better understanding of the range of mammals and their frequency of occurrence in roadside vegetation

Results:

Characteristic features of each roadside survey site, including the width of the road vegetation, the canopy tree species and the distance to the nearest forest patch.

Fourteen species of mammal were recorded in remnant forest vegetation at these roadside sites during the survey, and a further four species (Southern Brown Bandicoot, Isoodon obesulus, Koala, Phascolarctos cinereus, Common Brushtail Possum, Trichosurus vulpecula, and Brown Hare, Lepus Capensis) were recorded from other observations.

Thus a total of 18 of the 23 species mammal (including bats) known to presently occur in forest remnants within the Naringal area (Bennett, 1987a), also use roadside vegetation in some way. Those species not recorded from roadside vegetation where the Feathertail Glider Acrobates pygmaeus, Eastern Grey Kangaroo, Macropus giganteus, and Red-necked Wallaby, Macropus rufogriseus, and two aquatic species, the Platypus, Ornithorhynchus anatinus and Water-rat, Hydromys chryogaster.

The most widespread and abundant native mammal was the Bush Rat, Rattus fuscipes, which were caught at 14 of the 15 roadside sites.

The high abundance of this species may be attributed to the dense ground cover present at most sites. Most other native mammals were recorded from a small number of sites. The Swamp Rat, Rattus lutreolus, was trapped at three sites where there was a moist field stratum of grasses, sedges and sedge-like plants (E.G. Lomandra longifolia, Lepidosperma laterale, Dianella tasmanica)

The Long-nosed Potoroo, Potorous tridactylus, was captured at four sites, but probably also occurred as a further three sites where diggings and runaways similar to those made by this species were noted. The species is locally common in forest remnants where there is a dense field stratum, which often comprises dense cover of Lepidosperma laterale in shallow depressions with adjacent sclerophyllous vegetation (P. Esculentum, L juniperinum, B. marginata, Acrotriche serrulata, Tetratheca ciliata and Leucopogon australis) on drier soils.

At three sites, the Short-beaked Echidna, Tachyglossus aculeatus, was recorded by the presence of characteristic diggings made when foraging for ants. Individuals were irregularly observed elsewhere on roadsides (on at least four occasions), foraging as they moved along the forested strip.

The most frequently observed arboreal mammal was the Common Ringtail Possum, Pseudocheirus peregrinus with single animals observed at five sites. This species is common in remnant patches of forest in the area. The Common Brushtail Possum was not observed at any of the survey sites, but individuals were recorded as road-kills at two locations close to farm houses.

At two sites, both within 100m of an adjacent forest patch, the occurrence of the Swamp Wallaby, Wallabia bicolor, was initially inferred from the presence of faecal pellets; and individuals were later seen at each site. Unidentified tracks of a macropod at two further sites may also have been from this species.

Hair sampling tubes proved useful in recording the Brown Antechinus, Antechinus stuartii, from two sites where individuals had not been trapped, and the single record of the Sugar Glider, Petaurus Breviceps, was also obtained from a hair sample.

The occurrence of the latter species in roadside vegetation was further substantiated by two road-killed specimens and by later spotlight observation at another roadside.

In June, 1981, a single Koala was observed on a narrow roadside verge of E. ovata.

The origin of this animal is uncertain. This species was historically present in the area, but had not been sighted for at least 60 years. Later, in November 1981, a group of 10 Koalas was introduced to the Ralph Illidge Sanctuary, one of the larger remnants of forest in the study area, by the Fisheries and Wildlife Service ("Warrnambool Standard", 4/11/1981).

Discussion:

Use of narrow forested strips along roadsides at Naringal was not limited to a few common or generalist species, but included a wide range of mammals.

At least 18 of the 23 species of mammals (excluding bats) which are known to occur in forested vegetation in the area were recorded from roadside vegetation. This total includes species which are uncommon within the area (E.G. Southern Brown Bandicoot, Long-nosed Bandicoot), as well as those which are abundant (E.G. European Rabbit, Bush Rat).

The occurrence of such a large proportion of the native mammalian fauna in roadside vegetation in this locality can be attributed to two main factors: the retention of native vegetation along road reserves, and the nearby presence of tracts of forest vegetation amongst the farmland.

Although relatively narrow, many of the strips of roadside vegetation have a well-developed understorey of shrubs and ground cover, which are necessary components of the habitat of most terrestrial animals. Heavily disturbed sites, where the vegetation is dominated by introduced grasses and weeds, were not sampled; however, incidental observations and field experience suggest that few native mammals utilise such disturbed vegetation.

The close proximity of larger patches of forest and the continuity of many roadside strips with adjacent forest, provide a source for colonists to move into and through the roadside vegetation. In most instances, strips of roadside vegetation are part of a continuous habitat which may extend in a narrow belt for more than a kilometre. These attributes contribute to greater numbers of species being recorded at roadside survey sites than from a small discrete patches of forest of comparable size.

Observations during this study suggest that mammals use the roadside vegetation at Naringal in a variety of ways.

Roadside vegetation may provide some species with an additional area in which to forage (E.G. Short-beaked Echidna, Common Brushtail, Fox.); It may serve as a refuge from disturbance (E.G. Swamp Wallaby, European Rabbit, Cat.); and for some species it is a habitat In which they can live (E.G. Common Ringtail Possum, Long-nosed Potoroo, Bush Rat, Swamp Rat).