Insectes Sociaux

, Volume 58, Issue 1, pp 17–22

Digging effort in leaf-cutting ant queens (Atta sexdens rubropilosa) and its effects on survival and colony growth during the claustral phase


  • R. S. Camargo
    • Laboratory of Social Pest Insects, Department of Vegetal Production, Faculty of Agronomic ScienceUNESP
  • L. C. Forti
    • Laboratory of Social Pest Insects, Department of Vegetal Production, Faculty of Agronomic ScienceUNESP
  • R. T. Fujihara
    • Laboratory of Social Pest Insects, Department of Vegetal Production, Faculty of Agronomic ScienceUNESP
    • Department of Behavioural Physiology and SociobiologyUniversity of Würzburg, Biozentrum, Am Hubland
Research Article

DOI: 10.1007/s00040-010-0110-5

Cite this article as:
Camargo, R.S., Forti, L.C., Fujihara, R.T. et al. Insect. Soc. (2011) 58: 17. doi:10.1007/s00040-010-0110-5


Nest foundation in the leaf-cutting ant Atta sexdens is claustral, and the single queen completely relies on its body reserves throughout, approximately, 9 weeks until the first workers emerge and initiate foraging. Nest digging is much time- and energy-consuming, and it is an open question how queens decide on the length of the tunnel they dig and therefore the depth of the initial chamber. Shallow founding nests may be energetically cheaper to dig, but queens may be more exposed to changing environmental variables. Deeper nests, on the other hand, may be climatically more stable and suitable, but more expensive to dig. We hypothesized that the maximal nest depth excavated by Atta founding queens may represent the outcome of an evolutionary trade-off between maximizing nest depth and minimizing energy expenditure during digging, so as to save energy for the long claustral phase. We tested this hypothesis by comparing the fitness consequences of increased digging effort in queens that were experimentally stimulated to excavate a complete founding nest either once, twice or three times consecutively compared to control queens that did not dig. Fitness was quantified as mortality rates, rates of egg-laying and offspring production, and size of the fungus garden until the emergence of the first workers. Results showed that, in contrast with the initial expectations, fungus growth, egg-laying rates and offspring production were not affected by the increased digging effort in the experimentally induced successive excavations. However, a significant higher mortality was observed in queens with increased digging effort, i.e., those that dug two or three nests consecutively. It is argued that in queens a behavioral mechanism for the control of nest depth has evolutionary been selected for as a trade-off between maximizing nest depth, to favor protection of the queen against unsuitable environmental variables, and minimizing energy expenditure during digging, which significantly affects survival.


Leaf-cutting antsEgg-layingAttaNest diggingClaustral


Nest foundation in Atta leaf-cutting ants is claustral (with an exception in A. cephalotes, which was mentioned to be semi-claustral; Weber, 1972), so that queens utilize their body reserves to maintain their activities until the first workers emerge and initiate foraging (Hölldobler and Wilson, 1990). Claustral queens tend to be larger in size due to storage of copious lipids, carbohydrates and proteins for the activities required in the claustral phase (Boomsma and Isaaks, 1985; Brown and Bonhoeffer, 2003; Keller and Passera, 1989; Seal, 2009; Seal and Tschinkel, 2007). Nest digging by the queen demands a great expenditure of energy and time, and little is known as to how queens decide on the length of a tunnel they dig and therefore the final depth of the initial chamber. The digging of a superficial chamber is expected to require less energy, although the queens would be exposed to environmental stress like changing temperatures or desiccation. Deep chambers, on the other hand, may be microclimatically more stable, but more costly to dig, and queens would in addition be considerably exposed to predators during excavation.

Why do leaf-cutting ant queens not dig large tunnels, i.e., deeper founding nests? It can be hypothesized that in evolutionary terms, the maximal depth excavated by queens may have resulted from a trade-off between maximizing nest depth, so as to avoid extreme environmental conditions, and minimizing energy expenditure during digging. If the body reserves to be used during the claustral phase are limited, energy savings during digging may positively influence both the rates of offspring production and queen survival in the many weeks the queen cares alone for the developing brood and the initial fungus garden. In this evolutionary scenario, natural selection may have acted against longer tunnel lengths due to reduction of available energy reserves that may compromise foundation success. A straightforward way to test this hypothesis would involve the comparison of the fitness consequences in queens that excavate tunnels of different lengths. Considering that it is not possible to force queens to excavate beyond their selected tunnel length, we took advantage of the fact that Atta queens, after having completed the excavation of the founding nests, are known to immediately initiate the digging of a new, similar nest if they are removed from the first one and offered again a digging arena filled with soil (Ribeiro, 1972, 1995). As a consequence, we were able to test the hypothesis outlined above by evaluating the fitness consequences of increased digging effort in Atta queens that were stimulated to excavate a nest either once, twice or three times consecutively compared to control queens that did not dig at all. Digging effort was measured as the excavated tunnel lengths and the times spent digging. Fitness was quantified as mortality rates, rates of egg-laying and offspring production and size of the initial fungus garden until the emergence of the first workers, approximately 9 weeks after nest digging.

Materials and methods

Digging effort by queens and colony growth

The rationale of the experimental approach was to compare the performance, measured as both fungus growth and offspring production over time (numbers of eggs, larvae, and pupae raised until the emergence of the first workers) of queens with normal and increased digging efforts. Digging effort was quantified as both the excavated tunnel lengths and the total time spent digging in single or successive excavations of the founding nest by the same queen.

Queens were collected immediately after the nuptial flight on 4 October 2008 at the Lageado Experimental Farm, FCA/UNESP, Botucatu, Brazil, before they started to dig their nests. Queens were individually maintained in small plastic containers (11 cm in diameter and 8 cm in height), with a 1-cm floor of humid plaster. All experiments were performed at the Laboratory of Social Pest Insects, FCA/UNESP, Botucatu.

Each experiment was initiated by placing a single queen, immediately after the nuptial flight, on the surface of a tube filled with soil (25 cm in height and 10 cm in diameter), where it could initiate digging. The soil used originated from the University Campus (Latosol) and was collected at a depth of 60 cm (soil density = 1.6 g/cm3; water content: 5.4%).

Four different experimental series were performed, as follows:
  1. 1.

    Without digging: Queens did not dig at all, and were directly placed singly on a plastic container for further observations (body mass: 685.28 ± 57.55 mg, mean ± SD, N = 15).

  2. 2.

    Single digging: Queens were allowed to dig in the tube filled with soil, as described above, until they sealed the excavated tunnel, which is an indication that they excavated the founding chamber (Eidmann, 1935) (body mass: 682.17 ± 27.15 mg, mean ± SD, N = 15).

  3. 3.

    Double digging: Single queens, after finalizing the excavation of the founding nest, were immediately removed from the dug chamber and again confronted with a new digging tube, in which they started to excavate a second founding nest (body mass: 661.73 ± 24.20 mg, mean ± SD, N = 15).

  4. 4.

    Triple digging: It is the same as described above, but single queens dug successively three founding nests (body mass: 668.34 ± 48.81 mg, mean ± SD, N = 15).


Queen belonging to all groups were finally removed from the excavated chambers (by opening the digging tubes; the fungus regurgitated by the queen was also collected), and the tunnel lengths and the digging time were recorded. Digging time was defined as the time elapsed since the beginning of tunnel digging until the first soil pellets were deposited by the queen at the tunnel entrance for closing. At that point, it could be estimated that tunnel construction had been completed, as queens could turn around after having laterally enlarged the end of the tunnel for the construction of the founding chamber (Eidmann, 1935).

After digging (single or repeated), queens were immediately placed in small plastic containers as described above for counts of offspring production. Queen mortality, egg-laying rates and rates of offspring production were quantified weekly throughout the 9 weeks of the execution of the experiments. At the end of this phase, the mass of the fungus gardens was recorded on a semi-analytical balance to the nearest 0.1 mg.

Statistical analysis

Data on the excavated tunnel lengths and digging times were compared by the Mann–Whitney U test (α = 0.05). For more than two groups, comparisons using Kruskal–Wallis H test with the Mann–Whitney U test or Student–Newman–Keuls test as a post hoc (α = 0.05) were used. Data on offspring production (eggs, larvae, pupae and adults) and fungus garden masses were analyzed using ANOVA (α = 0.05). Data on queen mortality were submitted to a survival analysis (Likelihood ratio test). Statistical analyses and graphs were processed by BioEstat 5.0, SAS 9.1.3 for Windows, R 2.9.0 for Windows and SigmaPlot 11.0.


Digging behavior of queens and increased effort

Under the experimental conditions, single queens were observed to excavate their nests as already described (Eidmann, 1935; Cunha, 1968). Briefly, the excavation of the nest structure requires two consecutive digging patterns. At first the queen bites the soil with its mandibles and digs a small depression, from which a tunnel as wide as the queen’s body diameter results. The behavioral sequence involves the removal of a portion of soil, aggregation and compression to form a soil pellet and its transport. Pellets are placed in a circle around the developing tunnel. The position of the queen while transporting pellets changes over the course of the excavation. At the beginning, the queen enters head first and climbs up backwards loaded with a pellet to exit the tunnel. The second phase is initiated when the selected tunnel length is reached: the queen switches from tunnel to chamber digging by enlarging the end of the tunnel towards one side. Because of the gain in space, the queen is able to turn around, and so it walks forwards when removing further pellets and starts to close the tunnel with them, hence initiating the claustral founding phase.

Queens belonging to the group “single digging” excavated a nest as described above, with a tunnel of 15.1 ± 3.3 cm length by spending 306.9 ± 74.8 min (mean ± SD, Fig. 1). Additional 55 min were used in average until complete closure of the tunnel. The excavated chambers were in average 4.6 ± 0.7, 2 ± 0.2 and 3 ± 0.3 cm (chamber length, height and width).
Fig. 1

Lengths of the excavated tunnels (cm, left) and total digging time (minutes, right) as recorded across the different experimental series, for queens that excavated either once, twice, or three times consecutively. Data are median, quartiles and ranges. Black points represent outliers. For statistics, see text

Queens experimentally stimulated to dig successively a second founding nest dug significantly shorter tunnels (Fig. 1, left; Mann–Whitney U test, N = 27, U = 34.50, p < 0.005). For queens excavating three nests consecutively, differences occurred between the first versus the second and the third excavation (Fig. 1; Kruskal–Wallis H test with Student–Newman–Keuls test, first vs. second: N = 29, p < 0.0001; first vs. third: N = 27, p < 0.005). No significant differences were found between the second and the third tunnel lengths (Kruskal–Wallis H test with Student–Newman–Keuls test, N = 26, p = 0.2763).

In relation to digging time, there were no statistical differences between the successive excavations in the group “double digging” (Fig. 1, right; Mann–Whitney U test, N = 27, U = 79.00, p = 0.2957), nor among the successive excavations in the group “triple digging” (Fig. 1; Kruskal–Wallis H test, H = 0.7985, df = 2, p = 0.6708).

Colony growth and queen survival as a function of digging effort

Newly deposited eggs were already present in the first week after nest founding in all experimental series. Larvae were observed from the fourth week and pupae from the sixth week onwards. Workers emerged in the eighth and ninth week (Fig. 2). There were no significant differences across the experimental series in the average production of eggs (ANOVA, F3,32 = 0.0110, p = 0.9981), larvae (ANOVA, F3,24 = 0.0301, p = 0.9928) and pupae (ANOVA, F3,12 = 0.1315, p = 0.9395).
Fig. 2

Number of offspring (mean, SE) counted weekly in the colonies initiated by the Atta queens from the different experimental series

The fungus garden masses at the end of the 9-week period varied from 56.8 to 77.9 mg, but there were no statistically significant differences among the four treatments (no digging, single, double and triple digging: ANOVA, F3,39 = 1.777, p = 0.1674).

Survival rates of queens belonging to the four experimental groups are presented in Fig. 3. Survival decreased markedly in the first week after nest founding, averaging 58%, and continued decreasing over the 9-week period of observations until an average value of 27%. Although there appears to be a tendency of reduced survival with increasing digging effort, the four experimental groups did not differ statistically (Fig. 3a, Likelihood ratio test = 1.89, df = 3, p = 0.5965), probably because of the relatively low N values (N = 15 each). In order to assess the effects of increased digging effort on survival, data from queens with no or with normal digging effort (no digging + single digging; N = 30) were pooled and compared with data from queens with increased effort (double + triple digging; N = 30). Queens in the first pooled group survived significantly longer than queens with experimentally increased digging effort (Fig. 3b, Likelihood ratio test = 3.95, df = 1, p = 0.0441).
Fig. 3

a Survival (%) of Atta queens over 9 weeks; b survival curves after pooling of data in two groups, to address the effects of increased digging effort on survival: queens with no digging effort plus queens that dug once (N = 30) and queens with increased digging effort (double + triple digging, N = 30). See text for statistical analyses


Both tunnel lengths and digging times of Atta sexdens queens in the laboratory were in the range reported in previous field (between 6 and 10 h, Autuori, 1942) and laboratory studies (between 7 and 11 h and 8.5–15 cm, Ribeiro, 1972).

Queens that dug successively two or three founding nests excavated significantly shorter tunnels than those of their first nest, although the time spent did not differ. It follows that the queens excavated the successive nests at a slower pace. The observation of invariant digging times for the successive nest excavations (Fig. 1) suggests that some kind of time control may underlie the control of nest depth, but the involved mechanisms remain elusive and are the focus of current investigations.

Through successive excavations, queens were forced to spend a total digging time that was three times longer that under natural conditions, with the expected concomitant large increase in energy expenditure. However, such large differences in effort did not influence the rates of offspring production and fungal growth (Fig. 2). Offspring production was in the range of previous reports (Autuori, 1942), and similar to that of Atta capiguara (Pereira-da-Silva, 1979). It can be concluded that Atta queens, with approximately 40% of their body mass comprised of fat (Seal, 2009), possess enough energy reserves to maintain their rate of offspring production high over 9 weeks, independent of the experimentally increased energy expenditure at the day of nest founding. In fact, queens were shown to lose ca. 40% of their body mass after 60 days (Della Lucia et al., 1995), so reaching their minimal body mass, before the first workers start to forage. When foraging is initiated, the queen regains mass because of feeding and lays eggs at a rate that depends on colony’s nourishment (Della Lucia et al., 1990).

However, the survival of queens was significantly affected by the increased digging effort (double and triple digging, Fig. 3b). Interestingly, the increased mortality in the queens that successively excavated two or three nests occurred in the first week after digging, while the mortality rates thereafter remained independent on the initial digging effort, with both survival curves decreasing with a similar pattern (Fig. 3b). Mortality rates due to increased digging effort were marked, averaging one-third of the existing queens in the first week (a reduction from 63 to 40%, Fig. 3b). As a consequence, and also considering the arguments presented above, it can hardly be argued that queen mortality was based on the exhaustion of body reserves. Two consequences of increased effort on queen survival appear likely. First, an excessive cuticular abrasion due to increased digging, with the concomitant water losses as described for seed harvester ants (Johnson, 2000; Johnson and Gibbs, 2004) and second, the accumulation of oxidative damage associated with the intense initial digging activity, as known for flying insects (Magwere et al., 2006; Sohal and Buchan, 1981), may have been a significant main proximate cause of the mortality observed shortly after digging.

Why do Atta queens not excavate deeper tunnels, and so escape the more severe environmental conditions expected to occur at the superficial soil levels? We suggest that the maximal nest depth excavated by Atta queens appears to be the outcome of an evolutionary trade-off between maximizing nest depth, to favor protection of the queen against environmental variables, and minimizing energy expenditure during digging, which was shown to significantly affect survival. To what extent a minimization of predatory risks has in addition selected for shorter tunnels, because of the reduction in digging time and associated queen exposure, remains to be evaluated as an additional hypothesis. In fact, mortality of semi-claustral queens during foraging has been suggested as the main selective pressure favoring the evolution of claustral colony founding in ants (Brown and Bonhoeffer, 2003).


FR dedicates this work to Martin Bollazzi, who shared all phases of euphoria and hopelessness of the queen project over the years. The authors are indebted to Kerstin Fröhle for inspiring work on Atta vollenweideri queens, to Liciana Vaz de Arruda Silveira, Carlos Matos and Christian Jost for help with statistical analyses, to Nelson Carneiro, Antonio Marcos de Lima and Donizete de Almeida for soil collection and help with the preparation of the experimental tubes, and to two anonymous reviewers for helpful comments. This study was performed in the framework of the co-operation agreement between the UNESP-Botucatu and the University of Würzburg (Ref. 910-2007). Thanks are also due to The São Paulo State Research Foundation (FAPESP) for the post-doctoral scholarship to the first author (2007/04010-0) and for a research aid (2007/07091-0). FR was supported by funds from the Deutsche Forschungsgemeinschaft (DFG, grant SFB 554/TP E1) and FAPESP (2008/05434-0).

Copyright information

© International Union for the Study of Social Insects (IUSSI) 2010