Evolution, not transgenerational plasticity, explains the divergence of acorn ant thermal tolerance across an urban-rural temperature cline

Disentangling the mechanisms of phenotypic shifts in response to environmental change is critical, and although studies increasingly disentangle phenotypic plasticity from evolutionary change, few explore the potential role for transgenerational plasticity in this context. Here, we evaluate the potential role that transgenerational plasticity plays in phenotypic divergence of acorn ants in response to urbanization. F2 generation worker ants (offspring of lab-born queens) exhibited similar divergence among urban and rural populations as F1 generation worker ants (offspring of field-born queens) indicating that evolutionary differentiation rather than transgenerational plasticity was responsible for shifts towards higher heat tolerance and diminished cold tolerance in urban acorn ants. Hybrid matings between urban and rural populations provided further insight into the genetic architecture of thermal adaptation. Heat tolerance of hybrids more resembled the urban-urban pure type, whereas cold tolerance of hybrids more resembled the rural-rural pure type. As a consequence, thermal tolerance traits in this system appear to be influenced by dominance rather than being purely additive traits, and heat and cold tolerance might be determined by separate genes. Though transgenerational plasticity does not explain divergence of acorn ant thermal tolerance, its role in divergence of other traits and across other urbanization gradients merits further study.

change, few explore the potential role for transgenerational plasticity in this context. Here, we 23 evaluate the potential role that transgenerational plasticity plays in phenotypic divergence of 24 acorn ants in response to urbanization. F2 generation worker ants (offspring of lab-born queens) 25 exhibited similar divergence among urban and rural populations as F1 generation worker ants 26 (offspring of field-born queens) indicating that evolutionary differentiation rather than   Here we use a laboratory common garden experiment with a multi-generational breeding 84 design to test for the presence of transgenerational plasticity and also investigate the genetic 85 architecture underlying evolutionary divergence between urban and rural populations of the  However our design could not detect or rule out transgenerational plasticity as a mechanism 99 behind this adaptive divergence.

100
In this study, we tested the heat and cold tolerance of F2 offspring, lab-born workers of 101 lab-born queens. We tested the offspring of pure-type rural population matings, pure-type urban 102 population matings, and hybrids, both where the female was from the rural population and the 103 male from the urban population and vice versa. If transgenerational plasticity were responsible 104 for the increased heat tolerance and diminished cold tolerance we documented previously in 105 urban population acorn ants, then we would expect these phenotypic differences to disappear in

118
Colony collections 119 We collected queenright (queen present) acorn ant (Temnothorax curvispinosus) colonies and rural sites were categorized as 0% ISA (see Table S1 for site details).

125
Mating design and common garden rearing 126 We returned field-caught colonies to the lab and placed them under conditions to twine. This allowed us to check on the mating status of the paired alates, specifically whether the 143 ants had shed their wings (an indication that they have mated, (Herbers 1990) and whether mated 144 female alates began to lay eggs. All mated females began new egg production by 7 December 145 2017.

146
Once mated female alates began to produce eggs, we placed these newly established 147 colonies individually within 120 mL plastic cups and provided them with water and sugar 148 resource tubes and dead mealworms. We held these colonies at 25 °C (on a 14:10 photoperiod), 149 as this is the optimal temperature for brood development (Diamond et al. 2013). After the newly 150 laid eggs developed into workers, we tested the thermal tolerance of these workers (i.e., the 151 second generation of acorn ant workers reared entirely within the laboratory environment). per colony (and within each tolerance type, heat or cold tolerance) ranged from 5 to 24 with a 170 mean and SD of 13.6 and 5.17 (see also Table S2).

172
Statistical analyses 173 We fit linear mixed effects models with either CT max or CT min as the response variable 174 and the type of mating (the two pure types, urban-urban and rural-rural, and the two hybrid types 175 with the population origin of the mother listed first, urban-rural and rural-urban) as a predictor 176 variable (treated as a categorical variable with four levels). We used the lme function from the

198
We found no evidence that the evolutionary differentiation in heat and cold tolerance 199 among urban and rural acorn ant populations was driven by transgenerational plasticity.  Interestingly, hybrids were more similar to either the rural-rural or urban-urban pure 216 types depending on whether heat tolerance or cold tolerance was assessed. The heat tolerance of 217 hybrids more resembled the urban-urban pure type than the rural-rural pure type (Fig. 1A), 218 whereas the cold tolerance of hybrids more resembled the rural-rural pure type than the urban-219 urban pure type (Fig. 1B). This provides further evidence that divergence in thermal tolerance 220 does not stem from transgenerational plasticity, as the offspring did not consistently resemble a 221 specific parental population. Moreover, this differential matching of hybrids and pure types 222 11 across heat and cold tolerance potentially suggests some degree of genetic decoupling of the 223 traits. 224 We further explored the potential ecological consequences of the differential matching of 225 hybrid and pure-type heat and cold tolerance phenotypes by computing their tolerance breadth 226 (difference between colony mean CT max and CT min ). The likelihood ratio test for the effect of 227 mating type on tolerance breadth was statistically significant (χ2 = 9.22; P = 0.0266; df = 2).  Table   337 16 1). This expanded thermal performance (Figure 2, Table 1) breaks the expectation of 338 intermediate hybrid fitness for ecological speciation (Rundle and Nosil 2005) and suggests that 339 hybridization could lead to population collapse rather than population divergence. We then 340 expect that limited dispersal ability, shifted phenologies, or strong selection against migrants has 341 enabled the evolutionary divergence between urban and rural populations in light of the 342 expanded thermal breadth of their hybrid offspring. However, the evidence for any ongoing 343 speciation between urban and rural population is both preliminary and mixed at this time.

344
In summary, the results from a two-generation common garden rearing experiment 345 support the conclusion that acorn ants have evolved divergent heat and cold thermal tolerances in 346 response to the selective agent of the urban heat island in Cleveland, Ohio. Surprisingly, thermal 347 tolerance appears to be influenced by dominance rather than being a purely additive trait, and