The size and fecundity and the choice of pleometrotic partners in the ant Lasius niger. An analysis of the influence

Contents

1 Abstract

2 Introduction
2.1 Cooperation Between Individuals
2.2 Altruism
2.3 Pleometrosis
2.4 Questions and Hypotheses

3 Material and Methods
3.1 Lasius niger
3.2 Collection and Storing
3.3 Decision Experiments
3.3.1 Random Grouping
3.3.2 Fecundity-based Grouping
3.4 Analysis

4 Results
4.1 The Influence of Fecundity
4.2 The Influence of Size
4.3 The Impact on Choosing Time and Chamber Visits

5 Discussion
5.1 Primary Study Goals
5.2 Fecundity as a Driving Factor in Pleometrotic Choice
5.3 Size does not Matter (mostly)
5.4 All in Good Time
5.5 Evolutionary Aspects
5.6 Error Analysis
5.7 Conclusion

6 Bibliography

1 Abstract

Cooperation between unrelated individuals provides benefits that could ensure the survival of species. This work investigates on pleometrotic choosing behavior in Lasius niger ant queens. Here, two or more unrelated and usually monogynous queens found a colony together right after their nuptial flight and cooperate up to the emergence of the first workers. Then they fight until only one of them remains. Since the surviving queen is usually more fecund and heavier than her former nestmates, it is hypothesized that she will choose smaller queens with a lower fecundity to cooperate with. Therefore, queens in special arenas with two opposite arranged branches were given different choices. Here, either both branches contained a queen of higher or lower fecundity or only one of them. Two different experiments were performed, where the queens were either grouped random or based on their fecundity. It was found that queens that were chosen over another queen later laid significantly less eggs than the rejected ones. Additionally, queens that chose a lower fecund queen were significantly faster at choosing than queens that chose a higher fecund one. Queens that chose a queen over an empty nest site also did this faster than queens that chose an empty nest site over a queen. There was no significant difference in thorax length between chosen and rejected queens. The differences in relative thorax length were only significant in one of two collection locations, where chosen queens were larger than rejected ones. These results lead to the hypothesis, that there is a mechanism in Lasius niger queens giving them the ability to assess the fecundity of another queen, probably through CHC's. The thorax length likely does not contribute to this mechanism.

2 Introduction

2.1 Cooperation Between Individuals

Cooperation between individuals is a widely spread phenomenon in animal behavior and topic of countless publications by behavioral ecologists. It can provide benefits that would be hard or even impossible to reach without cooperation (Dugatkin, 2002). In many cases, this cooperation is limited to related individuals. This is relatable from an evolutionary point of view, since relatives share the same gene pool.

Eusociality for example is a widespread phenomenon describing a special kind of social behavior in animals. It is defined by three main characteristics. Cooperative broodcare, overlapping and cohabiting generations and the division of labor between individuals. Societies which meet all characteristics are called states and are further divided into labor-specific castes (Crespi and Yanega, 1995). There are two gradations of eusociality, primitive and advanced. In primitive eusocial species the members of the castes only differ in their specific behavior and physiology, whereas in advanced eusocial species members of different castes show significantly modified morphological features (Sherman et al., 1995). If one or more castes of a species are sterile, to provide monopolized reproduction by the queen caste, the species sometimes is called hypersocial (Richards, 2019).

To our best knowledge, eusociality appears in three taxa. Mammals (Jarvis, 1981), crustaceans (Duffy, 1998) and insects (Ross, 1995). In mammals and crustaceans, it is not a very common phenomenon, only appearing in single species, but in insects it is far more widespread. Here the Hymenoptera contain most of the eusocial species such as bees (Apidae and Halictidae), wasps (Crabronidae and Vespidae) and ants (Formicidae). While bees and wasps appear to be eusocial in some cases, ants are nearly always eusocial (Hölldobler and Wilson, 1990).

The emergence and consistency of eusociality appeared to be an evolutionary paradox in the past. Charles Darwin (1895) already mentioned that sterile, or simply not reproducing individuals are not capable of passing on their genes to the next generation and therefore their eusocial characteristics should not persist over time. This apparent paradox could be solved later with the inclusive fitness theory by Hamilton (1964). It claims that the evolutionary fitness of an individual not only depends on the number and survival of their own offspring, but also on the offspring of their relatives, because they still share many genes. This leads to the occasion that the fitness of non-reproducing individuals indeed rises if they support their, of course very closely related, mother (Hamilton, 1964). Non-reproducing individuals in a eusocial society are therefore an extended phenotype of the reproducing individuals' genes, giving them a higher chance of survival and a higher evolutionary fitness, which finally leads to the consistence of eusocial behavior (Dawkins, 1982). The kin selection theory also supports the idea that supporting closely related relatives increases an individual's evolutionary fitness (Smith, 1964). Additionally, because of the haplodiploidy in many Hymenoptera species and the resulting relatedness structure, sisters are closer related to each other than they would be to their own offspring (Hamilton, 1964). Therefore, sterile female workers pass on more of their own genes if they raise their sisters, than they would raising their hypothetical offspring (Quinones and Pen, 2017).

2 Altruism

Apart from the very common phenomenon of cooperation between related individuals, there are also cases of cooperation between unrelated animals. This manifests in altruism, a behavior with high costs that nevertheless offers a lot of benefits. From a biological point of view, individuals act altruistic if they take an action which decreases their own reproductive fitness, in order to increase the fitness of another individual (Bell, 2008). The individual's intention, if there is any, does not matter here (Okasha, 2003). The fitness loss can result in a, more or less direct, fitness regain through evolutionary effects like kin selection. Here supporting closely related relatives can increase the success of an individual's own genes too, because the number of shared genes is high (Smith, 1964). This may result in obligate altruism, where the direct fitness loss remains permanent. For instance, sterile workers of eusocial insects evolutionary renounce their own reproduction in order to support their queen. Therefore, they can still spread their genes by raising their siblings (Davies et al., 2012).

But this does not explain altruistic behavior between unrelated individuals. Here the theory of reciprocal altruism takes place. It pictures altruism as a tradeoff situation. If the helping individual also receives help at a later point in time, the tradeoff pays off and will likely occur again. This increases the evolutionary fitness of both individuals in the long-term. If the altruistic actions remain one-sided, the helper will stop helping to protect itself from being exploited (Trivers, 1971).

2.3 Pleometrosis

Another, far more specialized kind of cooperation between unrelated individuals is pleometrosis. In social insects in general and Hymenoptera in particular, pleometrosis also shows that this cooperation is not limited to related individuals but also occurs between unrelated conspecifics.

Pleometrosis occurs where two or more unrelated queens of a social insect species found a colony together, even if they are monogynous. In Lasius niger the number of queens is normally reduced to one, shortly after the emergence of the first workers (Hölldobler and Wilson, 1977). Previous research found that in the field 18% of new founded Lasius niger colonies contain two or more queens (Sommer and Hölldobler, 1995).

The main reason why ants enter a pleometrotic association seems to be inter-colony competition. If queens cannot find a suitable nesting site with enough resources, because the density of founding queens and already existing colonies with workers is too high in their area, their probability to die without reproducing is high too (Nonacs, 1992). Older and lager colonies also often brood-raid younger colonies without workers to enlarge their own size (Stamps and Vinson 1991). Pleometrosis reduces the competition for food and other resources between founding queens and provides additional raid security, because pleometrotic colonies produce more workers in a shorter period than single queen colonies. This is also helpful when it comes to raiding other colonies. A large and early available worker force provides the possibility of taking control over resources and other colonies in the area and monopolizing them (Sommer and Hölldobler, 1992). This is a crucial factor for increasing the evolutionary fitness of the remaining queen, because the risk of dying alone seems to be higher than the risk of being rejected by the colony after the emergence of workers (Crespi and Choe, 1997). Therefore, pleometrosis is likely to be a density-dependent effect (Tschinkel and Howard, 1983).

To gain this enormous fitness advantage, the queen must survive the fight for the colony, occurring after the first workers emerged. Past research found that surviving Lasius niger queens are normally heavier (Aron et al., 2009) and lay more eggs (Medeiros et al., 1992) than their pleometrotic partners. Additionally, queens that lay more eggs are more likely to survive (Teggers et al., 2020). It is unknown if the queens can recognize these factors and make a choice based upon them.

Previous studies also found that the reduction of queens in a pleometrotic colony is probably influenced by workers behavior. It was observed that workers, shortly after their emergence, push not-dominant queens away from the eggs and probably induce aggressive behavior in the other queens (Bernasconi and Keller, 1996). Furthermore it is believed that the workers can not recognize their own mother in any way, leading to the hypothesis that they support the queen with the highest fecundity, in order to choose their mother by chance (Aron et al., 2009).

2.4 Questions and Hypotheses

The leading question of this work is whether Lasius niger queens choose their pleometrotic partners according to fecundity. The design of the experiments primarily addresses this question. In general, queens should always favor associations that most likely increase their chance of survival. Since the number of queens in a pleometrotic association in L. niger is reduced to one after the emergence of workers (Hölldobler and Wilson, 1977), the dominant queen gains all the evolutionary fitness of the colony in the future. Because the surviving queen usually lays more eggs (Medeiros et al., 1992) and the queen with the most eggs is more likely to survive (Teggers et al., 2020), it is hypothesized that queens laying less eggs will be chosen more frequently. Additionally, we will investigate on the influence of size on pleometrotic choosing behavior. The surviving queen in a pleometrotic association is usually heavier than her nestmates (Aron et al., 2009). Therefore, since size is expected to correlate with weight, we hypothesize that smaller queens will be chosen more frequently than larger ones.

To take a more detailed look into the behavior of the ant queens, the influence of the time they take to choose a partner is addressed too. The question here is, if the time needed to make a decision is influenced by queen size or fecundity. Referring to studies that connect a shorter decision time with a lower degree of conflict (Berlyne, 1957), it is hypothesized that queens choosing a lower fecund partner will be faster, since this decision should be preferred.

Finally, if the experiments reveal that Lasius niger queens choose their pleometrotic partners according to size or fecundity, it is hypothesized that they obey a mechanism giving them the ability to access the fecundity of other queens.

3 Material and Methods

3.1 Lasius niger

For this study, the black garden ant Lasius niger was used. It is a member of the subgroup Formicinae in the family of Formicidae. The species probably has existed for about 50 million years since it is considered morphologically identical with Lasius schiefferdeckeri which has been frequently found in Baltic amber (Larsson, 1978). It is mainly found in Europe and North America but also in South America and Asia. In central Europe it is the most common ant species (Collingwood, 1979). It has been reported that workers have a life expectancy of about one year with a maximum lifespan of about three years and queens can live for up to 30 years (Kramer et al., 2016; Hölldobler and Wilson, 1990).

Lasius niger normally is monogynous, meaning that there is only one queen per colony, but occasionally queens enter a pleometrotic association (Sommer and Hölldobler, 1995).

The mating, or nuptial flight usually occurs between June and August, typically on hot summer days. Time and place of nuptial flights heavily depend on the local weather. Whenever the conditions are in favor, sometimes thousands of males and queens fly together and fertilize each other (Brian et al., 1966). The males only live a few days after mating and then die (Eidmann, 1926). Shortly after mating, the queens remove their wings and digest the wing muscles to gain additional energy for raising the brood. Subsequently, they search for a suitable nesting site to lay their eggs and found a new colony (Turner, 1915). Lasius niger queens usually do not eat until the first workers hatch from their eggs to feed them (Eidmann, 1926). This normally takes eight to ten weeks. After the first workers emerged the queen also stops to take care of her brood and focuses on laying eggs which will be nursed by their workers from now on (Klotz et al., 2008).

The brood of Lasius niger has four consecutive stages of development: egg, larva, pupa and adult. The eggs are white with a sticky residue that allows them to be easily carried around in groups. After about 20 to 25 days larvae hatch from the eggs which need to be fed. The larvae grow continuously until they start to spin a cocoon around themselves. This is developmental stage is called pupa. Then a metamorphosis begins inside the cocoon and the shapeless larvae become juvenile workers. When this process is finished, the completely white ants emerge from the cocoon (Klotz et al., 2008). These newly hatched workers are entirely white in color and are called callows. They will darken and develop their colony-depending cuticular carbohydrate profile (CHC) within the first days. Before that, they can still be accepted by other queens than their own (Lenoir, 2009).

Lasius niger meets its nutritional needs mostly with sweet honeydew and protei n-rich insects (Donisthorpe, 1915). The nests, if the colony is not nesting under a stone or in rotten wood, consist of soil and are often build around plants that provide protection and stability. They usually reach about 50 centimeters above the ground (Wilson, 1955).

Among ant enthusiasts this species is a very popular one to keep in glass tubes or a formicarium at home. It is easy to care for, affordable and you can find mated queens among Europe and North America in the summer. Additionally, because Lasius niger usually hibernates, it can be kept at six to eight degrees Celsius for a longer period of time and will still be alive after removing it from the fridge (Haatanen et al., 2015). These characteristics make it very suitable for laboratory working conditions, where easy handling and cost-efficiency tremendously increase the practicability. Additionally, many studies investigating on pleometrosis used Lasius niger as a model organism (e.g.: Hölldobler and Wilson, 1977; Medeiros et al., 1992; Sommer and Hölldobler, 1995; Aron et al., 2009; Teggers et al., 2020). Since in this species only one queen of a pleometrotic association survives, it is more suitable to investigate on evolutionary effects than other species like Pogonomyrex californicus, where the state of multiple queens per colony remains permanent (Johnson, 2004). Here, the evolutionary costs of pleometrosis are not as high as in L. niger, since it is possibly beneficial for all associated queens in the long-term.

3.2 Collection and Storing

The Lasius niger queens that were used for the experiments come from two different locations in Mainz, Germany. All queens used for the random grouping experiments (Chapter 3.2.1) and half of the queens used for the fecundity-based grouping experiments (Chapter 3.2.2) were collected in the district of Bretzenheim (Coordinates: 49°58'50.8"N 8°14'21.3"E). The other half of the queens used for the fecundity-based grouping experiments were collected in the district of Hechtsheim (Coordinates: 49°58'36.7"N 8°15'55.8"E). Hereafter the two collection sites will be called Bretzenheim and Hechtsheim, respectively. The collection took place right after the nuptial flight, on the 4th of July 2020 from 15:00 to 21:00 local time in sunny weather with temperatures reaching from about 20°C to 25°C. Only queens that already lost their wings were collected, to make sure, that they were mated and able to produce eggs. The collection was done with the help of soft forceps and fluon-coated boxes (Polytetrafluorethylene), to prevent the queens from escaping.

After the collection, the queens were transferred into individual glass tubes (Figure 3). The tubes were filled halfway with water. On top of the water there was a piece of dentist cotton to hold it back and prevent the upper part of the tube from flooding. The tubes were sealed with another piece of cotton to prevent the queens from escaping.Tubes were put together in small, transparent plastic boxes with air permeable covers. Those were stored in groups of four in large, blue plastic boxes in a 21°C climate chamber with a humidity of 80%. On top of the small, transparent boxes, there were sheets of white printer paper and on top of the large, blue boxes, there were cardboard covers to provide darkness at all times.

Abbildung in dieser Leseprobe nicht enthalten

Figure 3: Scheme of a glass storing tube for Lasius niger queens Black: Walls of the tube. Blue: Water. Grey: Cotton. White: Space for the queen.

3.3 Decision Experiments

In general, the target of these experiments was to investigate on the influence of size and fecundity on the choice of a pleometrotic partner in newly mated Lasius niger queens.

To observe the choosing behavior we gave a queen different options. These queens, that were meant to make a decision will be called “choosing queens” from now on. They had the choice between either two queens or one queen and an empty nest site, depending on the experiment. These queens, that were meant to be selected by the choosing queen will be called “chosen queens”. If one of those chosen queens was rejected by the choosing queen, it will be called “not chosen queen”.

To achieve this, we planned two experiments aiming to get information about the choice of a pleometrotic partner. They differed in the selection and grouping of the queens that were used, as described in 3.2.1 and 3.2.2. Therefore, they will be referred to their grouping type, “Random” for 3.2.1 and “Fecundity-based” for 3.2.2. Additionally, each experiment had its own type of arena, also described in chapters 3.2.1 and 3.2.2. In general, the arenas consisted of a light area in the center, and two dark side arms that functioned as nesting sites. Depending on the experiment, chosen queens were attached to the arena by their abdomen using medical wire. These queens were either placed in only one of the nesting sites or in both. Then, a choosing queen was put into the arena to observe its choice.

The arenas were recorded with a Sony FDR-AX33 video camera for 12 hours. Recording took place in a dedicated room without windows and a constant temperature of 21°C, to avoid experimental disturbances. During recording, the lights in the room were switched on.

The thorax length of the queens was measured after the experiments to reduce the stress level. For this purpose, a Leica DFC425 digital microscope and the program Leica Application Suite (version 4.12.0) were used. The eggs were counted using a binocular with a 10x magnification, on day eight after the nuptial flight for the fecundity­based grouping and once a week for eleven weeks after the experiments for the random grouping. All measurements in this work were performed by a single person.

3.3.1 Random Grouping

This part of the Decision Experiments used randomized grouping of the collected queens to mimic a natural situation. This was realized by randomly choosing 135 queens and labelling their tubes from 1 to 135. Next, three queens were always grouped together, for example 1; 2; 3 or 7; 8; 9, ending up with 45 groups of three. There were 45 arenas (Figure 4) with a chosen queen attached to each nest site and a choosing queen that was added later.

The arenas consisted of a plastic petri dish as the center part (0=50mm) and two plastic test tubes (length=40mm) as the side arms. To provide darkness for the newly mated queens in order to mimic a natural nest site, we usedred foilto cover these side arms. As previous work has shown, the vision of Hymenoptera is insensitive to wavelengths humans consider red (Peitsch et al., 2004). The difference in lighting conditions between the center and the side arms was intended as an additional motivation to make a choice for one of the nest sites. To prevent the queens from escaping, the arenas were sealed with adhesive strips.

The arenas were recorded in groups of seven or eight. After the experiments, number of eggs, larvae, pupae and workers were counted for each queen once a week with a binocular until the first workers emerged.

Out of 45 trials, 16 were successful. Six trials failed because one or more queens escaped the arena, one additionally because a queen escaped the wire. Two trials failed after the experiments because a queen of the group died during the tracking period.In the other 20 trials the choosing queen did not made a distinct decision.

Abbildung in dieser Leseprobe nicht enthalten

Figure 4: Scheme of an arena for the random Grouping Decision Experiments with Lasius niger queens

Center: Petri dish, starting point of the choosing queen. Sides: Test tubes, chosen queens are attached here. Red rectangles: Red foil to provide darkness.

3.3.2 Fecundity-based Grouping

This experiment used grouping based on the number of eggs the queens laid as a measurement of their fecundity. More eggs indicate a higher fecundity. The eggs were counted with a binocular on day eight after the nuptial flight. Next, the queens were grouped as following, with the fecundity referring to the number of eggs.

The target was to create groups of five queens with two chosen queens and three choosing queens. The chosen queens stayed the same throughout all runs of the experiment, only the choosing queens changed. To give the choosing queens a choice between a higher and a lower fecund queen, the chosen queens had to have a corresponding fecundity.

High fecundity queens (HF) were the top 20% fecund queens, low fecundity queens (LF) the bottom 20%. Those were the chosen queens which were attached to the side arms of the arenas later. The remaining 60% were used as choosing queens (A, B, C) that were later added to the arenas to freely choose a nesting site. These three groups - HF, LF and choosing (A, B, C) - were each divided into an upper (U) and a lower (L) subgroup at the median (Table 1). Only queens from the same subgroup were paired, to make sure that their fecundity levels are far enough apart. The paired queens were chosen randomly within their subgroup.

Abbildung in dieser Leseprobe nicht enthalten

Out of 120 trials, 76 were successful. Eight trials failed because one or more queens escaped the wire. Five trials failed because a queen of the group died during the experiment. In the other 31 trials the choosing queen did not made a distinct decision. After grouping the queens like this, we ended up with 200 queens divided in 40 groups of five. Every group was tested three times, each time with a different choosing queen (A, B, C). Queen A was tested with both LF and HF queens, each in one side arm of the arena. Queen B was tested with the HF queen in one nest site and no queen in the other one. Queen C was tested with the LF queen in one nest site and no queen in the other one. Thus, we had 40 arenas with a queen attached to each nest site and 80 arenas with a queen attached to only one nest site. To prevent the probable effect of a side being preferred by the queens and therefore influences the results, the different fecundity level queens were in alternating sides of the arenas.

The arenas consisted of three stacked plexiglass panels resembling the floor, a milled- out space for the queens and the cover (Figure 5). The milled-out space had a center part (0=30mm) and two side arms (length=40mm; width=10mm). Red foil on top of the side arms provided darkness in the potential nest sites.There were always two arenas in one plexiglass block and they were recorded in a group of five blocks, respectively ten arenas.

Abbildung in dieser Leseprobe nicht enthalten

Figure 5: Scheme of an arena for the Fecundity-based Grouping Decision Experiments with Lasius niger queens

Blue: Plexiglass block. White: Milled-out space for the queens. Red rectangles: Red foil to provide darkness. Red lines: Medical wire to attach the queens to the sides.

[...]