ABSTRACT. In response to intra- and interspecific competition, ant colonies have evolved mechanisms to maximize early colony growth. Through the analysis of photographic images, this study describes the individual effect of two different strategies, pleometrosis and pupae transplantation, on the growth of Lasius niger colonies. Results showed that both methods had an overall
positive effect on colony production, albeit the magnitude at which pupae transplantation boosted colony growth was much larger than pleometrosis. After six weeks, colonies with two and three queens contained, respectively, 26% and 83% more brood than control colonies with only one queen. Queen association also led to overall decreased queen fecundity and death of all resident queens in over 60% of the colonies due to queen fights. After eight weeks, colonies that received 30 or 60 foreign pupae had produced on average 256% more brood than controls and nanitic workers that were 7% longer than those from colonies that did not receive transplanted pupae.
Keywords: pupae adoption, multiple queens, ant reproduction, colony founding, social insects, formicidae
Most colonies of Lasius niger are started by a single claustral queen that uses her stored energy reserves to raise her first brood. However, the high numbers of Lasius queens occurring after mating flights, coupled with the fact that queens avoid areas frequented by workers of established colonies (Sommer & Hölldobler 1995), leads to a high density of new nest foundations in the field.
In response to crowding and high inter-colony competition, incipient colonies may use two different strategies to gain competitive advantage.
One strategy is pleometrosis, where queens enter facultative associations with non-related queens during colony founding (Bernasconi & Strassmann 1999). Such pleometrotic association allows claustral species like L. niger, in which queens have a limited amount of resources to invest in reproduction, to produce more workers than single queens and to do so in shorter time (Waloff 1957, Bartz & Hölldobler 1982, Sommer & Hölldobler 1995). That may translate into (i) a higher foraging success, which will improve early colony growth and survival and (ii) an advantage when it comes to successfully defending the nest against brood raiding, usurpation (Bartz & Hölldobler 1982, Rissing & Pollock 1991, Tschinkel 1992b, Bales & Adams 1997) and predation (Jerome et al. 1998) by neighbouring colonies.
That said, an alternative reason for pleometrosis can also be a strong pressure for the newly mated queens to leave the soil surface and use any available holes, including those excavated by other queens (Tschinkel 1998). Despite the potential benefits of pleometrosis, there is also an important cost at the queen level, as the emergence of workers and the start of foraging activities generally trigger the onset of queen fights that continue until only one of the queens survives (Sommer & Hölldobler 1995, Bernasconi & Keller 1998, Aron et al. 2009).
A second strategy to improve competition by increasing early colony growth is the sequestration of foreign pupae and larvae through the raiding of brood from neighbouring colonies. This strategy provides ant colonies in the field with a competitive advantage derived from a rapid increase in the number of individuals in the nest, which leads to earlier colony maturation (Pollock & Rissing 1989, Rissing & Pollock 1991, Tschinkel 1992a, Gadau et al. 2003).
In addition to building up higher numbers of workers, it may also be advantageous for young colonies to produce larger individuals, as size influences the range of tasks a worker can perform (Wilson 1978, Mirenda & Vinson 1981, Arnan et al. 2011). For example, mature colonies in the field have not only a higher number of workers, but also larger individuals compared to incipient colonies (Porter & Tschinkel 1985, Hasegawa & Imai 2011), and worker polymorphism and task repertoire have been found to correlate with colony size (Anderson & McShea 2001). Whether increased colony growth via pleometrosis or sequestration of foreign pupae leads to production of larger workers, and in this way further improves the competitive abilities of incipient colonies of Lasius niger, remains to be tested.
In order to elucidate the potential competitive advantage colonies can achieve through pleometrosis and brood raiding, the present study focused on analysing the effects of multiple founding queens and pupae adoption on early colony growth and worker size of Lasius niger.
Experimental colonies Queens of Lasius niger were collected immediately after their mating flight on 17th July, 2014, in Aarhus, Denmark (56°11’04.9”N 10°07’01.1”E), and were kept individually in plastic test tubes containing water behind a cotton plug. The individuals necessary for the multiple queen experiment (pleometrosis) were placed in groups of 1, 2 or 3 queens per test tube less than 60 minutes after collection to minimize a build-up of territorial behaviour over time. The day after collection, queens were cooled to 16.5°C for 60 minutes for easier manipulation before being transferred into artificial nests.
The artificial nests were created in Petri dishes (5.5cm Ø). The lids of the dishes were perforated (single hole, 4mm Ø) to allow the movement of workers in and out of the nest; the bases were partly filled (approximately 2/3 of their volume) with a substrate consisting of a combination of Plaster of Paris and charcoal (8:1). A small piece of tape on the substrate allowed the provision of a drop of water for the queens to drink until the first workers emerged, at which point workers provided the queen with water. The nests
were then placed in a bigger Petri dish (14cm Ø) that contained a 10cm test tube filled with water and also served as an arena where the feedings would take place. Ad libitum sources of carbohydrate (20% sucrose solution in water) and protein were added when the first workers emerged.
Protein consisted of house flies, pieces of shrimp and a 1% solution of peptone from bovine meat (Sigma-Aldrich, product number 70175), alternated weekly. All arenas and nests were kept inside opaque paper boxes in a dark 25°C chamber for the duration of the experiment. Feedings and population assessments were performed outside the chamber, in light conditions and at room temperature (August-September, 2014, mean daily temperature approximately 18°C).
During the experiment, non-destructive assessments of colony sizes were based on photographic images taken every two weeks (camera: Canon EOS 1000D, lens: EF-S 18-55mm f/3.5-5.6 IS).
The software ImageJ (Open Source: http://imagej.nih.gov/ij/) was used to analyse the images in order to perform counts of brood and workers and to measure colony areas and lengths of nanitic workers. After the last assessment on week eight after the nuptial flight, the ant colonies were exposed to gradually decreasing temperatures (5 days at 20°C followed by 5 more days at 15°C).
Lower temperatures reduce the activity level of the colonies, making it possible to perform a more exact manual count of the number of pupae and workers that were used to correct the estimations obtained through photographic analysis. The number of eggs and larvae, however, could not be counted manually in this way due to the delicate nature of these stages and our need for keeping the individuals alive for future experiments.
Seven response variables were analysed in the two experiments. Four of them were readily obtained through analysis of photographic images.
These were (i) colony area (cm2) measured as the amount of substrate surface occupied by all brood, workers and queen(s), (ii) total number of imago workers (including transplanted individuals), (iii) total number of brood (i.e. eggs, larvae and pupae) and (iv) average length of intrinsically produced nanitic imago workers. For the brood transplantation experiment, total number of pupae and imago workers were corrected after a more accurate manual count was performed at the end of the experiment. These counts were not available for the multiple queen experiment, due to having to work with results from the analysis of images at week six, as explained above.
Per capita queen production was calculated in two different ways: (i) for the multiple queen experiment, the total number of eggs, larvae, pupae and imago workers at week six was divided by the number of queens in each colony, (ii) for the pupae transplantation experiment, as imago workers were a mixture of intrinsic an adopted workers, queen production was calculated as the total number of egg, larvae, pupae and workers at week eight, minus the number of transplanted pupae that survived.
Because an unknown fraction of the eggs in the nest may be consumed by the colony,
number of brood and per capita queen production were also analysed after excluding egg counts.
The software JMP 12.0 was used for statistical analysis. For data on nanitic length, ANOVA was used for overall analysis and TukeyKramer test for multiple comparisons between treatments within experiments. For data on colony areas, normality and variance homogeneity could not be achieved through data transformation and thus Kruskal-Wallis tests were used to analyse overall differences and paired Wilcoxon tests were used for multiple comparisons. Counts on brood and workers were analysed using generalized linear models with a Poisson log-linear model, adjusting p-values for pairwise comparisons with sequential Bonferroni correction (Holm 1979).