The taxonomic portion of this study followed the usual procedures. Initially, we surveyed the existing literature for references to California desert ants. Except for those
describing new California forms, these records are often based on misidentifications. When we were unable to verify the identification or locality, we did not include the record.
      Based on this literature survey and the collections of the Natural History Museum of Los Angeles County (LACM) we assembled a list of 25 genera and 81 species for our area. Of these, three genera were ultimately eliminated (one, however, through transfer of its sole representative to another genus, already included) as were five species. We, therefore, had records for 88% of the genera and 77% of the species in our area.
      Our next phase was a survey of existing collections. We visited and examined collections in a number of institutions. Here, we identified all pertinent ant specimens that we could locate and recorded the data. In general, the results were disappointing (See COLLECTION ASSESSMENT below). Several field trips were made to selected areas which, on the basis of data then available to us, required attention.
      The time-honored field method for the myrmecologist is to wander about, locating ant nests, digging into them and collecting samples. This was, and remains, the primary method. The berlese funnel technique is of very limited effectiveness in desert habitats because there is so little litter suitable to this method. Berlese samples contributed fewer than 20 specimens to this study.
     Ethylene glycol pit traps have also been utilized. These can be effective but unless placed with care can also be a waste of time and energy as a survey method. They are probably a good technique for use by the nonmyrmecologist, however.
     Pitfall traps did not yield any species not already collected by conventional methods by an experienced myrmecologist. In some instances they are not as effective, for some small ants either do not enter them or are able to walk out without falling into the ethylene glycol.
     Baits were also used. Our most frequently used baits were wheat germ, honey-water, and tuna fish. Other food items, such as peanut butter, fruits and fruit juices, bits of meat, were also utilized on an opportunistic basis. Baits are effective for sampling an area, especially for those species which forage on the surface of the soil and seem particularly good for crepuscular and nocturnal species.
     While in the field we photographed habitats and nest tumuli. Some ants construct quite characteristic tumuli. Excellent examples are those of Messor pergandei (Fig. 301), Acromyrmex versicolor (Fig. 307) and Myrmecocystus mexicanus (Fig. 318). However, the usual tumulus is a simple crater of excavated soil and the ultimate appearance is dependent upon the type of soil available, vegetation, weathering, and slope. Another factor is the age of the colony. Thus, mature colonies of M. pergandei construct rather large, low rings of soil surrounded or covered by a ring of chaff; the tumulus of a new colony is a simple crater indistinguishable from that of many other ants. Furthermore, in the same area one nest of any given species may be surmounted by a normal crater and another a few meters away may not have any tumulus.
      Our conclusion on tumuli as a means of identification is: don't attempt it. We found that, in general, they are not dependable except for a few species. In the field, as experienced observers of ants, we were wrong as often as right, once the obvious species, such as those listed above, were discounted. We fared better at the generic level than at the specific.
     While collecting in the field we observed ant behavior: when they were foraging, temperatures at which they were and were not surface active, how far they foraged, forage items, and interactions with ants and other organisms. Forage items were collected for identification. We also collected samples of nest debris which, of course, included remains of forage items discarded once they had been cleaned of any food.
     Such debris samples can be very informative. In several instances we were able to recover remains of ant species that we had not located at a specific site. These debris samples can be utilized as an adjunct to normal survey methods, for ants are very effective scavengers. In another instance a debris sample yielded hundreds of termite head capsules. Since all were those of the soldier caste, it seems safe to infer that the ants had gained access to the nest of the termites.
     Remains of other ants were a conspicuously large part of these debris samples. At first, it was assumed that these were remains of prey items. This interpretation is surely only partly correct. For all the time we've spent in the field over the last fifteen years, we have witnessed very few aggressive encounters between ants, certainly not enough to explain the relatively large number of fragments in the debris. Another explanation now seems more plausible. In collecting samples from returning foragers we found that remnants of other ants were certainly well represented. Generally these were quite dry, and obviously did not represent a fresh "kill". Rather, they were scavenged items from debris piles of neighboring community members (George, in preparation).
     Workers of such well-known scavengers as Solenopsis, Forelius, Dorymyrmex, and Myrmecocystus are often seen on and around the mounds of other ants. Usually, their presence is ignored by the resident species and direct confrontation is rare.
The "visitor" rummages through the debris and eventually finds something to carry off. Rarely, the "visitor" indulges in brigandage and actually snatches an item from a returning forager of the resident. Species of Dorymyrmex are especially noted for
such behavior.
      All data gathered in the field were recorded on data cards and were also recorded in notebooks. These are part of the data file on ants maintained by RRS at the LACM. Specimens are provided with field numbers that correspond to file card numbers.
     Preparation of keys and illustrations was conventional. Our only departure from convention is that we have separate keys for workers, females, and males in many cases. In other cases, keys for the sexual forms were not feasible because these are still too poorly known. Spot maps represent every verified locality for each species. The taxonomic portion was the work of RRS and he assumes sole responsibility for all taxonomic decisions, whether good or bad. So, too, with the figures and maps.
     In citing geographic data we have continued to use the English system. Thus, miles instead of kilometers, feet instead of meters for elevations. Currently, all topographic and general maps are based on this system and until new maps with metric coordinates are available, this seems the only sensible approach. Otherwise, all measurements are metric. Temperatures are in °C; air temperature is abbreviated AT, surface temperature, ST.

     COLLECTION ASSESSMENT
     Normally, authors do not make comment on the collections studied. We choose to be exceptional. Ecologists are increasingly aware of the importance of ants within arid
ecosystems. This means that, increasingly, these organisms will be subjected to study. Inevitably, ecologists will find it necessary to refer to systematic collections and taxonomists to provide identifications for their subjects.
     Groups popular to collectors are usually well represented in several collections. Not so for ants. Ants are mostly medium-sized or smaller, usually unobtrusive and not very
spectacular to look at. As a rule ants are collected only in passing or not at all.
     Historically, it has been difficult to get identifications on miscellaneous ant material, for ant taxonomy is formidable to the uninitiated and specialists are not much prone to look at a lot of ordinary, casually collected material. Further collecting tends to be discouraged. Consequently, there is not a great deal of ant material in the more important western collections. We have found, in general, that most collections had either very little in the way of ants or that they had more exotic than
native ants. The collections of the California Academy of Sciences (CAS), University of California at Berkeley (UCB) and at Davis (UCD) must be ranked as minor, as far as their holdings of southwestern ants are concerned. We do not disparage these collections, for they have important material. And, such collections can only reflect the interests of their curators, past and present.
     One of the largest collections, not surprisingly, is that of the California Department of Food and Agriculture (CDFA). Since ants often are considered to be pests, any collection emphasizing pests would accumulate many ants. The material is adequately
prepared, but mostly unidentified.
     The very large collection amassed by George and Jeanette Wheeler (GJW) reflects their long interest in, and involvement with, ants. It is extensive, well curated and, in the case of material collected by the Wheelers, accompanied by field notes on file cards. This collection was invaluable to our study.
      A few specimens from the collections of the Los Angeles County Agricultural Commisssioner's Office (LACA) were examined and included in this study.
     The collection of the Los Angeles County Museum of Natural History (LACM) is the largest collection of western American ants, especially of the southwestern United States and adjacent Mexico. It, of course, houses the material collected by RRS. In
addition, it includes the very important collections of W. F. Buren, A. C. Cole, Jr., and W. S. Creighton, as well as their field notes. Here, too, is a large library of ant-oriented
literature, with particular emphasis on systematics.

Next