Our l.iniii; Rcsiiiincx — luvertchnitcs 



171 



The Xerces Society started the Fourth olJuly 

 Buttertly Count (FJC) in 1975. sponsoring 

 it annually until 1993. when the North 

 American Butterfly Association (NASA) 

 assumed administration. The general methods 

 of the buttertly count are patterned after the 

 highly successful Christmas Bird Count (CBC), 

 founded in 1900 and sponsored by the National 

 Audubon Society (Swengel 1990). 



Painted lady iVcinessa cardui) nectaiing on showy milk- 

 weed [Asclepias speciosa). 



The results of the FJC. including butterfly 

 data, count-site descriptions, and weather infor- 

 mation on count day. are published annually. 

 The count was designed as an informal program 

 for buttertly enthusiasts and the general public. 

 These counts can never substitute for more for- 

 mal scientific censusing because data sets from 

 the counts have flaws that impair scientific 

 analysis. Nevertheless, the FJC program does 

 provide data that, with considerable caution, 

 can be useful for science and conservation 

 (Swengel 1990). FJC data have been used to 

 study the biology, status, and trends of both rare 

 and widely distributed species (Swengel 1990; 

 Nageletal. l991;Nagel 1992; Swengel. unpub- 

 lished data). 



Analysis and Application 



I reviewed FJC count reports and other pub- 

 lications for applications of FJC data to monitor 

 the status and trends of North American butter- 

 fly species. These studies varied considerably in 

 sample size, amount of data manipulation and 

 statistical analysis, and degree of variable con- 

 trol. Different methods of using FJC data 

 include, in order of ascending statistical refine- 

 ment; presence or absence of a species in a sub- 

 set of counts: highest observed number of a 

 species on a single count; individuals of a 

 species per count for a subset of counts in a 

 given year; and individuals of a species per 

 count hours or per count miles. The subset of 

 counts used to supply data for analysis also var- 

 ied from a single count to all counts in a certain 

 region or all counts ever reporting a given 



species during the study period. The sample sub- 

 set and statistical approach are best determined 

 by the nature and extent of available data. 



The rationales, methodologies, shortcom- 

 ings, and validity of analyzing FJC data have 

 been detailed elsewhere (Swengel 1990), but 

 are based on the substantial ornithological liter- 

 ature regarding the scientific use of CBC and 

 other types of survey data. As ornithologists 

 have clearly indicated, these kinds of data sets 

 must be used with great care because ( I ) the 

 sample sites and dates depend on when and 

 where volunteer observers choose to conduct a 

 count; (2) the quality of sampling and accuracy 

 of data vary among counts; (3) only certain 

 species are sampled adequately enough to allow 

 data interpretation; and (4) the species complex 

 can vary somewhat from year to year. Even with 

 such constraints, these data sets are valuable 

 because of the numerous sites surveyed, their 

 wide geographic scope, and the relatively low 

 cost of data acquisition. 



Interpreting Count Data 



For the first 1 1 years of the count program 

 (197.5-85). only a few dozen counts were held 

 annually, but since then the number of annual 

 counts has increased steadily to 209 in North 

 America in 1993. Each FJC annual report since 

 1982 has provided a table that details how many 

 counts reported each species and which single 

 count found the most individuals of each 

 species. Although informal, this table indicates 

 the frequency and abundance of buttertly 

 species as observed in the counts. 



Several rare species with federal status under 

 the Endangered Species Act have been sampled 

 in the counts, as reviewed in the introduction to 

 the 1993 FJC annual report (Opler and Swengel 

 1994). A researcher using FJC to study rare but- 

 terflies must be careful in inteipreting the data, 

 however. Unless a number of FJC counts are 

 specifically designed to sample rare species 

 well, it is unlikely that rare species will be sam- 

 pled adequately enough to allow scientific 

 analysis of status and trends. Even in these 

 cases, however, site data for rare species report- 

 ed in FJC remain useful as leads to follow in 

 status surveys of extant populations for these 

 species (Opler and Swengel 1994). Most likely, 

 the data should be considered as augmenting 

 additional, more formal scientific study and 

 should be confirmed, either by alternative sur- 

 vey means or by contacting the counters for 

 documentation. 



Because of the larger sample size, FJC data 

 may better demonstrate the population trends of 

 more abundant and widespread species. For 

 example, the painted lady (Vanessa cardui) is a 



Fourth of July 



Butterfly 



Count 



by 



Ann B. Swengel 



International Count Co-editor 



Monarch (Daiuiiis pitwippiis) nec- 

 laring on dwarf blazingstar [Liatris 

 cvlimlrcicea). 



