Survivorship and Walking Behavior of Inquilinitermes microcerus ( Termitidae : Termitinae ) in contact with host workers and walls from host nest

Constrictotermes sp. nests are frequently inhabited by colonies of Inquilinitermes microcerus . In this association, I. microcerus colonies usually establish their colonies spatially isolated from Constrictotermes colonies. Here, we investigated whether the apparent spatial isolation of I. microcerus colonies in Constrictotermes nests should be related to their needs ( e.g. feeding) in relation to the central part of the nest or to a possible stress provoked by the presence of the host. For this, survival and walking behavior bioassays were performed to test the hypothesis that the survivorship of inquilines is: (i) reduced in the presence of host, mainly of those from different nests, (ii) increased in contact with inner walls compared with external walls; and that the distance walked and walking velocity of inquiline is: (iii) increased in the presence of the host and (iv) reduced in contact with the internal walls compared with external walls of host nest. The mean time to death of inquiline workers is lower in contact with host (independently from the same or different nest) compared with control and the mean time to death of inquiline workers is lower in contact with external walls of host nest compared with control group and the inner walls. The distance walked and walking velocity of inquiline workers in contact with their hosts (from the same or different nest) did not differ from control, however, these parameters were reduced when workers were in contact with inner and external walls compared with control. In general, our results showed that I. microcerus adopt behavioral strategies to avoid perception by its host.

by symbionts species (see Hughes et al., 2008).Colonies of social insects, stand out in relation to the great diversity of symbionts that cohabit their nests, including parasites, mutualists and commensal species (Hughes et al., 2008).Termite nests, for example, can house a range of species, such as vertebrates (Brightsmith, 2000;Dechmann et al., 2009) and invertebrates (Cunha & Brandão, 2001;Costa et al., 2009), called as termitophilous.Among these cohabitants, other termite species than that building the nest are included, which are called inquilines (Redford, 1984).Some inquiline species are obligatory, since they do not have the ability to build their own nests and depend intrinsically of the host nest for its survivorship.These examples are found in species of the genus Inquilinitermes Mathews, 1977 (Termitidae: Termitinae) (e.g.I. fur, I. inquilinus and I. microcerus), which inhabit nests of Constrictotermes sp.(Mathews, 1977) and in Serritermes serrifer (Hagen & Bates, 1858) (Serritermitidae), which is obligatory inquiline of Cornitermes cumulans (Emerson & Krishna, 1975) (Termitidae: Syntermitinae).Differently from observed in Hymenoptera, the high incidence of cohabitation in termite nests is intriguing (Campbell et al., 2016;Marins et al., 2016).Like other social insects, in termites, colonies are able to recognize nestmate from non-nestmates, and also invest energy in the production of a specific caste to nest defense (Šobotník et al., 2008).
Since obligatory inquilines have a closer relationship with their hosts, they are expected to exhibit strategies that allow the temporal stability of this interaction.Although inquilines benefit from the nest of host colony, with no costs to constructing this structure (Jirošová et al., 2016), a series of evidences suggest that the inquiline seems to avoid (physically and chemically) its host.Studies have shown that the colonies of Inquilinitermes sp. are small in size relative to their host and they are usually found in the central parts of the host nest where individuals of host colony are rarely observed (Cunha et al., 2003).The aggregation of the inquilines in the central part of host nest may be related to the fact that they use this structure as a food resource (Mathews, 1977;Bourguignon et al., 2011;Cristaldo et al., 2012;Florencio et al., 2013;Barbosa-Silva et al., 2016).In addition, recent studies have shown evidence that individuals of I. microcerus can use chemical cues as a strategy to avoid conflicts, since they avoid chemical cues of the whole body, as well as the trail and alarm signals emitted by its host, C. cyphergaster (Cristaldo et al., 2014;2016).However, in spite of such advances in the understanding this interaction, the final balance for both partners still need to be unraveled.
In the present study, we investigated whether the apparent spatial isolation of I. microcerus colonies in Constrictotermes sp.nests should be related to their needs (e.g.feeding) in relation to the central part of the nest or to a possible stress caused by the presence of host colony.For this, survival and walking behavior bioassays were performed to test the hypothesis that the survivorship of inquilines is: (i) reduced in the presence of host, mainly of those from different nests, (ii) increased in contact with inner walls compared with external walls; and that the distance walked and walking speed of inquiline is: (iii) increased in the presence of the host, mainly of those from different nests and (iv) reduced when in contact with the internal walls of nest compared with external walls.

Study site and termite collection
Nests of Constrictotermes sp.(N= 9) were collected in June 2016 at Campus Rural of Federal University of Sergipe, in São Cristóvão (11º 01'S e 37º 12'O), Northeast, Brazil.The mean monthly precipitation and temperature in the region are 1.200 mm and 25 ºC, respectively.According to Köppen classification, the clime is As' type (tropical with rainy winter and dry summer) (Pidwirny, 2011).
Collection consisted in the complete removal of nests from the field using spades and picks.The nests were taken to the laboratory and fragmented to verify the presence of inquiline colonies, which were found in three of the nests sampled.The identification of species was conducted in comparison with samples from the Isoptera Collection of University of Brasília (UnB), in which samples were deposited.The host was identified as Constrictotermes sp.(#10745 and 10747) and the inquiline as Inquilinitermes microcerus (#10746 and 10748).

Bioassays
Bioassays were conducted to check whether the survivorship and walking behavior of I. microcerus workers are affected by the presence of its host (from the same or different nest) and by the contact of inquilines with its host nest walls (inner and external walls).

Survival bioassays
To check whether the presence of host affects the survivorship of inquiline workers, the following treatments were established: (i) 20 inquiline workers alone (control), (ii) 10 inquiline workers with 10 host workers from the same nest (SH) and (iii) 10 inquiline workers with 10 host from different nest (DH).Bioassays were conducted in a Petri dish (Ø 9 cm x 1.5 cm [height]) covered with a filter paper.
To check whether the contact with host nest wall affects the survivorship of inquiline workers, the following groups were established: (i) 10 inquiline workers alone (control), (ii) 10 inquiline workers with 7 g of substrate from inner wall of host nest (IW) and (iii) 10 inquiline workers with 7 g of substrate from external wall of host nest (EW).Fragments from inner (easily recognized by its dark coloration) and external walls of host nest were ground with a mortar and pestle and sieved through a 6-mesh sieve.Then, the material was weighed in a precision balance and distributed at the bottom of the Petri dish on the filter paper.In control group, Petri dishes were covered only with filter paper.
For both survival bioassays, three replicates were performed per nest for each treatment, totalizing 27 replicates.The mean value of three replicates/nest were then used in the statistical analyses.Petri dishes were kept in a biochemical oxygen demand (BOD) incubator (25°C, without light) and the quantification of dead individuals was performed at onehour intervals in the first day and then two-hours intervals until the death of all individuals.

Walking behavior bioassays
Walking behavior bioassays were conducted in a Petri dish (Ø 9 cm x | 1.5 cm) covered with a black paper with the same treatments described above.However, in this bioassay, only one individual was placed in the control group and nest wall treatments and two individuals in the host treatments.Three replicates were performed per nest and treatment, totalizing 27 replicates.The mean value of three replicates/ nest were then used in the statistical analyses.
To allow the visualization and recorded of walking behavior of inquiline workers, individuals placed in the Petri dish were previously marked with a mixture of white gouache and glue (2:1) following the procedure described in Marins et al. (in press).The movement of individuals was video-recorded for 10 min using a camera (Panasonic SD5 Superdynamicmodel WV-CP504), equipped with Spacecom lens (1/3 "3-8 mm) coupled to a computer.The distance walked and walking velocity were captured by EthoVision XT ® software (version 8.5, Noldus Integration System, Sterling, VA) and the data were analyzed using the Studio 9 software (Pinnacle Systems, moutain View, CA).

Statistical Analyses
To check whether survivorship of I. microcerus workers (y-axis) is affected by the presence of their hosts (from the same or different nests) or by the contact of host nest wall (inner or external wall), data were submitted to Survival Analysis using Weibull distribution.The mean time to death of all individuals in the Petri dish was calculated in each treatment x nest and then, data were submitted to Analysis of Deviance (ANODEV).
To check whether the distance walked (mm) and the walking velocity (mm/s) of inquiline workers (y-axis) is affected by the presence of their hosts (from the same or different nests) or by the contact of host nest wall (inner or external wall), data were submitted to Analysis of Deviance (ANODEV).
All analyses were performed in the R statistical software (R Development Core Team 2015) using Generalized Linear Modelling (GLM) followed by residual analyses to check the suitability of distribution choose.Statistical simplification among treatments was performed via t test using multcomp package.

Survival bioassays
The survivorship of I. microcerus workers was significantly lower in the treatments with presence of host than in the control (F 2,6 = 6.905,P = 0.027; Fig 1A).However, no significant differences in the mean time to death of inquilines was observed in the treatments with host from same or different nests (P = 0.65).
Inquiline workers in contact with external wall of host nest (EW) showed a lower mean time to death compared with inquiline workers in contact with the inner wall of host nest (IW) and with control (inquiline workers alone) (F 2,6 = 28.070;P< 0,001; Fig 1B).Control and inquiline workers in contact with the inner wall of host nest (IW) did not show significant differences between them (P = 0.83).

Walking behavior bioassays
Most representative trails of I. microcerus workers in all treatments are showed in Fig 2 .The presence of host (from same or different nest) did not affect significantly the distance walked (F 2,6 = 1.13,P = 0.32) and the walking velocity (F 2,6 = 0.19, P = 0.82) of I. microcerus workers.
The distance walked (F 2,6 = 7.31, P < 0.001) and the walking velocity (F 2,6 = 0.654, P = 0.02) of inquiline workers were significantly reduced when in contact with host nest walls (inner or external) compared with the control group (Fig 3A -B).However, there were no significant differences in the distance walked and the walking velocity of inquiline workers in contact with the inner or external walls of host nest (P = 0.10).

Discussion
In the present study, we investigated how the presence of the host species (Constrictotermes sp.) as well as the contact with host nest wall can modulate the survivorship and the walking behavior of its obligatory inquiline I. microcerus.In general, our results suggest that I. microcerus adopt strategies to avoid perception by its host, as already showed by previous studies.
The mean time to death of I. microcerus workers was lower in the presence of host than in the absence of it (Fig 1A).Interestingly, the mean time to death of inquiline workers was not dependent of host origin (i.e. from same or different nest).Such result can indicate that inquilines do not acquire the odor of the host colony, which could minimize the stress with its host colonies.In fact, C. cyphergaster and its obligatory inquiline I. microcerus do not share the same profile of cuticular hydrocarbons (Cristaldo, unpublished data), a strategy usually used by inquilines of social insects.In Hymenoptera, for example, the integration of social inquiline in the host nests is facilitated by the acquisition of host colony odor by inquilines [e.g.bees (Dronnet et al., 2005), wasps (Sledge et al., 2001), and ants (Lenoir et al., 2001)].Thus, it is expected that the relationship of obligatory inquilinism observed here is mediated by other evolutionary stable strategies, allowing a faster establishment of I. microcerus in nests of Constrictotermes sp.
The fact that inquilines apparently did not acquire the odor of its host colony suggests that they should use strategies to avoid meeting the host.However, here we found that the distance walked and the walking velocity of inquiline workers have not been changed in contact with host workers.This lack of changes in the walking behavior of inquiline workers in the presence of host can be a strategy to not becoming apparent, which would reduce possible conflicts.Thus, inquiline workers seem to minimize any reaction of perception and aggressiveness on the part of its host.Ours results are in agreement with recent studies that showed that I. microcerus is able to recognize the trail pheromone of its host C. cyphergaster.By doing so, the inquiline can avoid galleries occupied by its host, facilitating the coexistence between them (Cristaldo et al., 2014).Inquilines are also able to recognize the alarm pheromone of its hosts, using it for their own benefit, either to avoid spaces inhabited by hosts or to escape from predators (Cristaldo et al., 2016).Recently, Jirošová et al. (2016) showed that soldiers of I. inquilinus, an obligatory inquiline of C. cavifrons, produce chemical substances that are repellent to its host, suggesting that spatial separation of colonies is chemically mediated, which allow the coexistence by reduction of direct conflicts.To avoid the host, inquiline individuals could also contribute to reduce the chances of selection of strategies of counterattack by the host.In addition, reports in the literature suggest that inquiline colonies are usually found spatially isolated in the host nest and that they use galleries that do not overlap with those used by its host colonies (Mathews, 1977;Cunha et al., 2003;Cristaldo et al., 2012).This apparently spatial segregation could occur to avoid direct contact with the host (see Mathews, 1977;Florencio et al., 2013).
The inner wall of host nest is formed not only by excrement of host but also by excrement of inquilines.Contrary to our hypothesis, inquiline workers in contact with inner wall of host nest did not survival more than those in the control group (Fig 1B).This absence of significant variation can be related with two factors: i) the inability of inquiline workers to ingest the substrate of inner wall, as it was used in our bioassay and ii) the obstacle represented by the substrate.Once the wall substrate has been macerated, it can be assumed that the inquiline workers were unable to ingest such particles, thereby reducing a possible effect of food on the increment in the survival of them.In addition, the presence of this substrate may have been an obstacle to walking, increasing the energy expenditure of these individuals, thus masking a possible positive effect of this treatment.This last explanation is supported by the fact that the walking behavior was reduced when in contact with the inner and external walls in the same way when compared to the control group (Fig 2 and 3A-B).On the other hand, the contact with the external wall (EW) of host nest significantly reduced the survivorship of inquiline workers compared to IW and the control group (Fig 1B).The external wall of host nest should present less signs that indicate to the inquiline workers the presence of a known environment and it should limit the contact among the individuals of its colonies, which would explain the greater mortality of this treatment when compared to the control.This suggests that the reduction of walking behavior may be more related to some obstacle to any behavioral change due to chemical signals in these structures.On the other hand, survival seems to have been more influenced by the presence of chemical signals of the colony.
A series of hypothesis have been proposed to explain the cohabitation in termite nests: (i) inquilines become imperceptible in the nest (e.g.perceiving the chemical cues of its host or using different spaces in the host nest) (Cristaldo et al., 2012;2014;2016), (ii) inquilines have ability to repel their hosts (Jirošová et al., 2016) or (iii) inquilines do not overlap diet with their host (diet segregation; Florencio et al., 2013).The present study supports the idea that avoidance strategies used by inquilines can be the primary factor of coexistence between the obligatory inquiline I. microcerus and its host Constrictotermes sp.Previous studies have already indicated that avoidance strategy seems to occur by perception of chemical cues from host, here we showed that I. microcerus also have behavioral mechanisms that can avoid possible conflict easing the cohabitation in its host nests.

Fig 1 .
Fig 1. Survival of Inquilinitermes microcerus workers in different treatments.A) Control (inquilines workers alone), SH (inquiline workers with host workers from the same nest), DH (inquiline workers with host workers from different nest; B) control (inquiline workers alone), IW (inquiline workers in contact with inner wall of host nest), EW (inquiline workers in contact with external wall of host nest).Different letters in each graphic indicates significant difference among treatments.

Fig 2 .
Fig 2. Most representative trails of Inquilinitermes microcerus workers in the following treatments: A) inquiline workers alone, B) inquiline workers with host workers from the same nest, C) inquiline workers with host workers from different nest, D) inquiline workers in contact with substrate from the inner wall of host nest and E) inquiline workers in contact with substrate from the external wall of host nest.

Fig 3 .
Fig 3. Distance walked (A) and walking velocity (B) of Inquilinitermes microcerus workers in the following treatments: control (inquiline workers alone), IW (inquiline workers in contact with substrate from the inner wall of host nest) and EW (inquiline workers in contact with substrate from the external wall of host nest.Different letters in each graphic indicates significant difference among treatments.