THE DISPLACEMENT OF NATIVE ANT SPECIES BY THE INTRODUCED ARGENTINE ANT IRIDOJIYRMEX HUMILIS MAYR

INTRODUCTION Many authors have described how Iridomyrnex humilis Mayr has become a major pest throughout the world (Brun, I924; Zimmerman, 94o; Smith, 947; Morley, 953; Skaife, I96.I; Pasfield, 968). Once these ants become established in a. locality they will not tolerate the existence of any other species, of ants, and as the populations of each colony build up in density, they emigrate in all directions, consolidating as they go. and driving other species before them. Not only does I. humilis displace’ native ant species, but it has been shown to displace other introduced tramp species. The ant Pheidole megacephala F. is apparently a native to Africa and has been spread by commerce, to. almost all o.f the more humid parts of the world. It too, is a serious pest and displaces native species. However, in I852, in FUnchal, the capital o.f Madiera, this species was itself displaced by I. humilis (Stoll, 898; Wheeler, I9O6). The displacement of P. megacephala by I. humilis has also. been observed in the Hawaiian Islands (Wilson and Taylor, I967; Fluker and Beardsley,, 97o) and in Bermuda (Haskins and Haskins, I965 Crowell, 968). Vilson (I95I) reports that a local naturalist in Mobile, Alabama ob.served I. humilis displacing the imported fire ant Solenopsis saevissima richteri Forel, and Fluker and Beardsley (I97o) reported the displacement of S. geminata F. in Hawaii. Shapley (i92o a, b) describes an "intermittent war" between I. humilis and the native California species, which he eels would eventually eliminate, most of the native ant species. Tulloch (I93O) and Michener (i942) described the displacement of the California harvester ant Pogonomyrmex californicus by I. humilis.


INTRODUCTION
Many authors have described how Iridomyrnex humilis Mayr has become a major pest throughout the world (Brun,I924;Zimmerman,94o;Smith,947;Morley,953; Skaife, I96.I; Pasfield,968). Once these ants become established in a. locality they will not tolerate the existence of any other species, of ants, and as the populations of each colony build up in density, they emigrate in all directions, consolidating as they go. and driving other species before them. Not only does I. humilis displace' native ant species, but it has been shown to displace other introduced tramp species. The ant Pheidole megacephala F. is apparently a native to Africa and has been spread by commerce, to. almost all o.f the more humid parts of the world. It too, is a serious pest and displaces native species. However, in I852, in FUnchal, the capital o.f Madiera, this species was itself displaced by I. humilis (Stoll,898;Wheeler,I9O6). The displacement of P. megacephala by I. humilis has also. been observed in the Hawaiian Islands (Wilson and Taylor,I967;Fluker and Beardsley,, 97o) and in Bermuda (Haskins and Haskins, I965 Crowell,968). Vilson 196,,'3 1964 1964 1965 1965 1966 1966 1967 1967 ' 1968 1968 Figure The qeld was bordered by California Highway 78 on the west, dirt qeld roads on the east and south, and by a grass lawn on the north. Two additional fields of 5(2A) and 7(2B) hectares were located at the southern edge of the main study field. Here studies on colony size, foraging distance, and food preference were carried out (Erickson, I972 and in manuscript).
All colonies of P. californicus and I. humilis were individually marked with color-coded wooden stakes placed one meter i:rom the colony entrance. At intervals of approximately six months the position of each colony was noted on a large map, measured to. the nearest one meter from the colony entrance, using the reference Ps.vche [December markers placed at I'CO meter intervals throughout the field (see Fig. ). The line; on Fig. deliniate the limit of eastward expansion of I. hum.ilis colonies at the time. the survey was taken. The new area added was then calculated by subtracting the. total occupied area at the. previous sampling date from the new total area occupied by I. humilis. This area was. then divided by the number of colonies in the new added territory to. get the mean area per colony values..
At the sa.me time, quadrat sampling with 3o randomly placed 2 meter by 2 meter quadrats were carried out to determine the vegetation characteristics. Weather data were taken from a station 3 miles northeast of the field site and averaged to get monthly mean temperatures and precipitation.

RESULTS
The displacement of the three other ant species by I. humilis started slo.wly in October, I963, but increased to an almost constant rate from 3 May I964 to 4 October I968 (Table I for the whole occupied portion of the field increased during each sampling interval, whereas, the. mean area per colony in the newly displaced land was almost a constant 14oo m per colony. The number of colonies .of I. humilis increased during each displacement interval as the displacement proceeded whereas the number of colonies of P. californicus decreased except for a minor fluctuation due. to flooding between 3 October I965 and May 966 (Fig. 2). By October 968, not a single colony of P. grallipes was observed in the study field, and by 5 March 1969 all colonies of P. californicus and 17. pergandei were located outside the boundaries of the. study field. In their place remained 57 colonies of I. humilis.

DISCUSSION
To explain this phenomenon of displacement, some comparison is necessary of the basic biology of the ant species involved. The nests of I. humilis are situated wherever there is sufficient moisture and where light is excluded, as under rocks and logs (Woodword, 19o.5, 191o;Eckert and Mallis, 937, Smith, 1947) or in shallow nests in the soil (Cook, 1953). These ants occur in a wide variety of habitats-swamps, beaches, lawns and gardens, roadsides, houses, and various woodlands (Crowell, 1968 1963 1964 1964 1965 1965 1966 1966 1967 1967 1968 1968 Figure 2. The number of colonies in the entire field of P. californicus (o) and 1. humilis (e) from October 1963 to October 1968. of more favorable habitats (Wilson, 1971). Colonies of I. humilis contain a large number o.f queens with thousands o.f workers (Smith, I947) and proliferate by swarming of detachments of workers who accompany secondary queens out of the nest (Wheeler, I933; Wynne-Edwards, I963; Crowell, I968). They are highly omnivorous but tend to. seek sweet or fatty foods (Eckert and Mallis, I937; Creighton, I96o'; Cook, I953), and tend aphids and scale insects in orchards and gardens (Skaife,, I96I). In contrast with I. humilis, the California harvesters are large ants (4-6 mm lo.ng) which are primarily seed gatherers, but are also known to be slightly omnivorous (Van Pelt, I966). Colonies of P. californicus are .small in comparison to. I. kumilis and contain only one queen. Proliferation takes place, by large swarms of winged reproductives. The California harvester ant tends to nest in dryer semi-desert habitats and can tolerate much higher temperatures than I. humilig (Wheeler, I926; Cole, I932, I968; Michener, I942; Erickson, I972).
The relative reproductive potential of I. hunilis is probably much higher than P. californicus. This is most likely due to the large num-Psyche [Decembe ber o queens in each colony, the method of colony proliferation, and the omnivorous habits o these ants. These actors may also. account or the great and rapid spread o I. humilis throughout the temperate regions o the world.
Basic differences in o.od resources limit to. some extent the amount o competition between I. humilis and the three harvester ant species (Table 2). I. humilis is highly omnivorous whereas the three harvester ant species are' only slightly o.mnivorous, being basically seed gatherers. I. humilis not only monopolizes the proven ood sources but attempts to control the remaining oraging areas (Wilson,97I). In the main study field, t:ood, especially seeds, were very abundant.
The (o.od chambers o P. californicus and I. humilis were always full when the colonies were excavated. In qelds 2A and 2B, the area was supplemented with approximately qve pounds, o mixed grass seed per month to determine the oraging characteristics and distances or P. californicus. The seeds, colored with common ood coloring, we.re fully acceptable to the ants, making up 43 to 59% o the P. californicus ood stores and 9 to 7% o the I. humilis ood stores.
The variously colored seeds were spread in concentric circles. rom a nest o( P. californicus every 5 meters, to a distance of 3o meters. The maximum fo.raging distance or P. californicus was about o meters except in areas where there was an I. humilis colony in which case the harvester ants (oraged no arther than 5 meters even though the I. humilis colony was. 2o meters away. Even though both /]elds were supplemented with a little over 5o pounds o mixed grass seed per year, in one year P. californicus was displaced 76 meters (2A) and o9 meters (2B) by I. humilis. It does not appear that this displacement is. due to any overlap o a undamental o.od dimension.
At each sampling interval the mean area per colony o I. humilis in the newly displaced territory was approximately 4o.o m whereas the mean area per colony or the entire teld increased from 4oo m to. approximately 26.oo m during the rive year period (Table ).
There thus appears to be a minimal area o.r a colony o. I. humilis in the newly acquired areas and as these colonies become established and increase in population density, the. colo.ny requires a larger area.
Michener (942) working with P. californicus encountered a similar displacement by I. humilis. He described in detail how individual harvester ants would be set upon and killed by groups o [. humilis. When temperatures are cool, Pogonomyrmex species tend to. be sluggish and it is at this time. that the Argentine ants torment the harvester ants as they orage around the nest (Michener,942). sweet or fatty qq--k- Wheeler, 1910 foods, tends Eckert & aphids scale Mallis, 1937 insects, grains Creighton, 1950 Skaife, 1961 seed gatherers q- Wheeler, 1910Forel, 1928Wildermuth & Davis, 1931Cook, 1953Van Pelt, 1966Cole, 1968 sweet or fatty q--J-A- Wheeler, 1910 foods Forel, 1928 Mallis, 1937Cook, 1953 insects, fruits, q--k--Jr- Creighton, 1950 grains, flowers, Cook, 1953 vegetables insects, fruits, -f--f- Creighton, 1950grains Fluker & Beardsley, 1970 *Displaced in present study. +++ highly omnivorous; ++ moderately omnivorous slightly omnivorous Should ,a harvester ant come upon an Argentine ant during the warmer parts of the day, the former grasps the smaller ant with its mandibles and stings it to death (Michener,942). At dawn, sunset, or on a cloudy day the Argentine ants will attack and cling to the mandibles, legs, and antennae of the harvester ants and attempt to kill the larger ant. Observations made in the present study confirm ;Vfichener's discussion of the aggressive actions between the species.
There were no significant differences in the mean monthly temperature or precipitation from month to rnonth (i.e.all the Januarys, etc.) over the co.urse of the study. The vegetation studies Psyche [December similarly showed that there xvas no significant difference in the order of dominance of the six plants mentioned. It does not appear that P. californicus ameliorates the habitat as it does not .clear vegetation as many harvester ant species do. I. humilis does not utilize the same nest sites as the displaced species and in fact, not a single I. humilis colony was found within two meters of an abandoned harvester ant colony.
Pasfield (1968) found I. hztmilis displaced its neighbors at a maximum rate of 274 meters (3o0 yards) per year in Australia. This value, is higher than the OO to 2oo meters per year at Fort S'hafter on the island o.f Oahu between 194o and 1944 (Pemberton, 1944) or the average of lO.O meters per year in the present study (Fig. I). Fluker and Beardsley (I97O) observed I. humilis displace P. megaceth.ala in Hawaii at about 66 to lOO meter.s per year. All these values seem low when compared to the displacement rate. of 8 kilometers (5 miles) per year for native species by the fire ant S. saevissima in the Gulf states (Wilson and Brown,957).
The effectiveness of competition in nature is best demonstrated by the impact of an invading speci.es on the native, fauna. It appears that here, there is a tremendou.s competition for nest space, which is the general case f.or highly aggressive territorial ant species such as Pheidole, 8olenot)si.r, and Iridornyrmex (xvVilson, 1971 ). Three' aspects of the populations biology of I. humilis gives this species a distinct competitive advantage over the native harvester ants. The general aggressive nature of I. humilis as well as the large number of queens and method of proliferation allow these ants to. move in and establish new colonies in a very short time. Raiding columns of workers clear the way and pioneer groups of workers and queens follow into freshly opened nest areas.