The tawny crazy ant, Nylanderia fulva (Mayr) (Hymenoptera: Formicidae), develops extremely large populations that dominate landscapes and invade buildings in states along the Gulf of Mexico (Florida, Alabama, Mississippi, Louisiana, and Texas), Georgia, and the island of St. Croix, US Virgin Islands (Klotz et al. 1995; Wetterer & Keularts 2008; MacGown & Layton 2010; Gotzek et al. 2012; Wetterer et al. 2014; Gochnour et al. 2019). There is evidence that populations of N. fulva throughout the USA form a single supercolony (Eyer et al. 2018; LeBrun et al. 2019; Lawson & Oi 2020). Control of invasive ants, such as N. fulva, requires a thorough understanding of their biology to guide the development of effective measures of suppression or eradication. For example, phenology and annual population abundance cycles may be used to determine optimal treatment application times (Markin 1970; Hoffmann et al. 2011; Hoffmann 2015). Examining the seasonal prevalence of N. fulva reproductives, Oi (2020) reported the presence of numerous queens within nests during the winter, but curiously these nests contained minimal amounts of brood. To further understand these observations, this study examined the seasonal reproductive status of N. fulva queens collected in north central Florida.

The reproductive status of N. fulva queens was determined by estimating the number of eggs, or large developing eggs, observed in their ovaries and examining spermathecae for sperm. Three N. fulva colonies, or more appropriately called colony fragments because of its supercolony status, were collected monthly from 2 sites in Gainesville, Florida, USA (a private residence [29.6302°N, 82.4638°W]; and a wooded area near Petra Design Inc. [29.7001°N, 82.3335°W]), from Sep 2013 to Feb 2015. Colony fragments (hereafter colonies) were extracted from isolated (> 5 m apart) decaying branches found on the ground and from nested plant nursery containers using the methods of Sharma et al. (2019). Ten dealate queens then were separated from each colony and either dissected immediately or frozen live at –80 °C. Ovaries were dissected from 10 queens per colony, and the number of developing and mature eggs in the ovaries of each queen was estimated with a rating scale used for Solenopsis invicta Buren (Hymenoptera: Formicidae) where 1 = no eggs; 2 = 1 to 10 eggs; 3 = 11 to 50 eggs; and 4 = > 50 eggs (Valles et al. 2013). Whereas the fecundity of S. invicta queens is much higher than N. fulva based on oviposition rates (Tschinkel 1988; Arcila et al. 2002; McDonald 2012), this scale could distinguish changes in N. fulva fecundity among seasons. In addition, the spermathecae of queens were examined by microscopy for the presence of spermatozoa to determine if queens were inseminated (Glancey & Lofgren 1985) (Fig. 1).

Ovary ratings and percentage of inseminated queens per colony were determined for each mo, then grouped into seasons. Using temperatures recorded by the Florida Automated Weather Network ( https://fawn.ifas.ufl.edu/) for Alachua County from Sep 2013 through Feb 2015, seasonal categories were determined from the average minimum and maximum daily temperatures (Oi 2020). The months of Dec to Feb had the lowest average daily minimum and daily maximum temperatures of 6.6 °C (range: –4.6–17.8) and 19.1 °C (range: 3.4–28.3), respectively. These mo were designated as the winter season and the succeeding 3 mo intervals comprised the remaining seasons of spring (Mar–May), summer (Jun–Aug), and fall (Sep–Nov).

The association of the frequency of the 4 ovary ratings per queen with the 4 seasons was determined with a 4 × 4 contingency table and chi-square test. Similarly, the association between the 4 ratings and the number of inseminated and number of uninseminated spermathecae was evaluated by chi-square test of a 4 × 2 contingency table (Proc FREQ, SAS Version 9.4, SAS Institute, Cary, North Carolina, USA). Queens with undetected spermathecae were considered to be uninseminated because previously observed uninseminated spermathecae usually were very difficult to discern.

The frequency of the individual ovary ratings was not randomly distributed across seasons (χ² = 56.129; df = 9; P < 0.0001), where 73 to 91% (n = 90 and 150, respectively) of the queens had ovary ratings ≥ 3 per season (Fig. 2A). Among the 4 seasons, 81 to 92% of the queens were inseminated, indicating a significant preponderance (χ² = 12.166; df = 3; P = 0.0068) of mated female reproductives throughout the yr. The seasonal distribution of uninseminated queens was 19% (n = 155) and 14% (n = 90) for winter and spring, respectively, and 8% each for summer (n = 150) and fall (n = 126). The association between ovary ratings and the inseminated and uninseminated queens also was not random (χ² = 81.686; df = 3; P < 0.0001). All 4 ovary ratings were distributed evenly among uninseminated queens, ranging from 20 to 32% (n = 65). In contrast, 89% (n = 456) of the inseminated queens had ovary ratings ≥ 3 (Fig. 2B). Arcila et al. (2002) reared N. fulva queens from larvae in queenless colony fragments consisting of larvae and workers, and thus could be a source of uninseminated queens. However, whether these queens were dealates or alate gynes was not specified.

The combination of high numbers of N. fulva queens with low brood levels in permanent nest sites, and few, small, ephemeral nests during the winter suggested a winter consolidation of colonies (Zenner-Polania 1990; Oi 2020). Because of the lack of brood in the winter nests despite the presence of many queens, it was hypothesized that queens in the winter were unmated and ovaries would have fewer eggs. The results of this study showed that over 80% of the queens had inseminated spermathecae in each season. In addition, ovary rating frequencies for each season were skewed toward the ratings of 3 and 4 (10–50 eggs and > 50 eggs, respectively) indicating that eggs were present in queens throughout the yr.

The percentages of queens with > 50 eggs increased from 26% in winter to a peak of 68% in summer. The occurrence of queens with 10 to 50 eggs declined from 55% in winter to a low of 28% in summer, while queens with < 10 eggs was below 20% for all seasons (Fig. 2A). These trends suggested that egg production increased in queens as the seasons progressed toward the summer. More dramatic changes in ovary status have been reported in other ants such as Dolichoderus mariae Forel (Hymenoptera: Formicidae) where queen ovaries were inactive in Jan but had increasing numbers of eggs in Apr and Jun (Laskis & Tschinkel 2008).

Fig. 1.

Ovaries dissected from Nylanderia fulva queens. (A) Ovary rated as a “2” (1–10 eggs) where the spermatheca is present. (B) Ovary rating of “4” (> 50 eggs).

Ovaries of N. fulva queens were primed with eggs in the winter, but this did not reflect the absence of brood in nests examined in the winter (Oi 2020). It is likely that the warmer temperatures during spring and summer and the associated increase in food availability, for example, honeydew producing insects (Sharma et al. 2013), will initiate and sustain the production and oviposition of eggs as seen in other invasive ants (Markin 1970; Tschinkel 1993; Abril et al. 2007; Abril & Gomez 2014; Hoffmann 2015). If queens are not actively producing and laying eggs during the winter, their ingestion of food and bait could be minimal and thus limit the impact of the bait on the queens (Collins et al. 1992; Hoffmann 2015). In Argentine ants, the dominant reproductive queens ingested more bait than non-dominant queens allowing the latter queens to survive and potentially allowing colonies to recover (Hooper-Bui et. al. 2015).

Targeting bait applications to winter nesting sites where N. fulva queens have coalesced represents an enticing control strategy. Indeed, removal of winter nests has resulted in temporary suppression of Argentine ants (Diaz et al. 2014), but bait applications are logistically more feasible than excavating nests over large areas. However, given the potential for limited bait ingestion by queens in the winter, it would be prudent to specifically examine seasonal queen development and empirically determine when and how N. fulva queens could be most efficiently controlled.

Truly appreciated was the detailed technical assistance of E. Mena, J. Dietz, and M. Perdomo (all formerly USDA-ARS), and M.Y. Choi (USDA-ARS) for providing training on ovary dissections. Gratitude is extended to J. Hogsette for providing access to an infested site and to S. Valles (USDA-ARS) for his helpful review of the manuscript. Mention of trade names or commercial products in this article are for the information and convenience of the reader and does not imply recommendation or endorsement by the US Department of Agriculture.

Fig. 2.

Percent frequency of Nylanderia fulva queens with specified ovary ratings: 1 = 0 eggs; 2 = 1 to 10 eggs; 3 = 10 to 50 eggs; 4 = > 50 eggs. (A) among seasons (winter: Dec–Feb, n = 155; spring: Mar–May, n = 90; summer: Jun–Aug, n = 150; fall: Sep–Nov, n = 126), and (B) between uninseminated (n = 65) and inseminated (n = 456) queens over all seasons.