Many invertebrates exhibit parental care, posited as a precursor to sociality. For example, solitary foundresses of the facultative social orchid bee Euglossa viridissima guard their brood for 6+ weeks before offspring emerge, when the nest may become social. Guarding comes at the fitness cost of foregoing the production of additional offspring. Yet it is unclear whether guarding (extended maternal care: EMC) can enhance offspring survival such that it outweighs those fitness costs, or if it is a consequence of the selective benefits of sociality, including extended female longevity. Experimental removal of solitary foundresses from nests of E. viridissima revealed an immediate fitness loss: decreased offspring survival. A mathematical model exploring the trade-off between EMC versus non-guarding revealed that EMC is advantageous if the nest established is greater than a threshold of 1.7-12.5 days. For much of the parameter space, EMC may be a precursor to sociality. Below this threshold, our model suggests that social fitness gains (acquiring helper daughters) need to be invoked to explain the evolution of EMC. Enhanced survival of offspring through guarding and nest inheritance may nevertheless ease conditions for the evolution of sociality by favouring extended adult longevity and brood care in incipient social species like E. viridissima.

Removal Experiment: Nests with females that had finished brood production were randomly assigned to one of two categories: treatment (T) or control (C) in a paired design (see below), with n = 11 pairs of nests. Females in control nests remained with their brood whilst females in treatment nests were removed. Nest entrances were always naturally closed at the time of experimental female removal, which created similar starting conditions for control (C) and treatment (T) nests. A pair of nests (one C, one T) was initiated at approximately the same time (+/- 7 days) to account for seasonal variation in environmental conditions. Nests were then monitored for adult brood emergence and the presence of intruders three times a day until all broods had emerged or until the last brood cell was considered to have failed.

Presence of Intruders: Nests of the Removal Experiment were monitored for the presence of Intruders three times a day until all brood had emerged or until the last brood cells were considered to have failed, as the number of days in which intruders were found within the nest divided by the number of observation days per nest. The number of days in which intruders were found in the nest, was used rather than the number of intruders per day to avoid counting the same intruder multiple times on the same day.

Intranidal Behaviour: A total of 12 solitary nests (5 control nests from the removal experiment plus 7 additional nests that were not included in the removal experiment) were observed, while females were in the guarding phase. Nests were brought into a darkroom inside a laboratory at ambient temperature. Females were observed under a red light through the glass roof of their nest, with the nest entrance always closed naturally from within. Observations commenced on the day after a female had finished brood cell provisioning, and each nest was observed on 2.75 ± 0.35 (mean ± SE) days (range 1-4 days). Once a new adult eclosed from the first brood cell, observations ceased. Observations were carried out across the day (09:00 to 17:00) to avoid diurnal bias and to record behaviours that might be related to daily weather conditions. Observations were performed at different stages of brood development. After bringing a nest into the lab, females were allowed to acclimatize to the new environment for five minutes, and then female behaviour was observed for two 15-minute periods, with a five-minute break between periods. Then a nest was returned to its original location in the field. All behaviours that occurred during observations were scored and their duration and number of occurrences were noted. Behaviours were then grouped into five categories: brood care (b) such as brood cell grooming and physical contact with brood cells, nest defence (d), including passive behaviours such as sitting at the nest entrance and active behaviours directed towards intruders, nest maintenance (m), self-grooming (g), and other (o).

Modelling Alternative Strategies: 
To evaluate whether the immediate solitary benefits of guarding are sufficient to account for its expression, we modelled the solitary pay-offs of two alternative strategies: guarding the brood by staying in the nest during the period of brood development (guarding strategy i.e. EMC), and deserting to found additional nests in which to produce more brood (deserting strategy).