Global change drivers are imposing novel conditions on Earth's ecosystems at an unprecedented rate. Among them, biological invasions and climate change are of critical concern. It is generally thought that strictly asexual populations will be more susceptible to rapid environmental alterations due to their lack of genetic variability and, thus, of adaptive responses. In this study, we evaluated the persistence of a widely distributed asexual lineage of the alfalfa race of the pea aphid, Acyrthosiphon pisum, along a latitudinal transect of approximately 600 Km in central Chile after facing environmental change for a decade. Based on microsatellite markers, we found an almost total replacement of the original aphid superclone by a new variant. Considering the unprecedented warming that this region has experienced in recent years, we experimentally evaluated the reproductive performance of these two A. pisum lineages at different thermal regimes. The new variant exhibits higher rates of population increase at warmer temperatures, and computer simulations employing a representative temperature dataset suggest that it might competitively displace the original superclone. These results support the idea of a superclone turnover mediated by differential reproductive performance under changing temperatures.

Sampling and laboratory rearing

In a previous study [47], A. pisum individuals were sampled from alfalfa fields in the spring and summer of 2017-2018 in central Chile, the same region previously surveyed by Peccoud et al. [23] in 2006-2007 (latitude 35 to 40 ºS, see [47]). This study showed that there were only two MLGs in this area (named APG2 and APG3 in [47]), with 96% of the genotyped individuals belonging to APG3 and the remaining 4% to APG2. Clonal lineages were established in the laboratory from these two MLG by rearing them separately on broad bean plants (Vicia faba L.) inside 4 L transparent plastic buckets, kept in a climatic chamber at 20°C (± 0.5°C) with a long photoperiod (LD = 16:8) to ensure their sustained parthenogenetic reproduction [46, 50]. Five aphids from each lineage were taken before performing the experimental procedures on the thermal response (see below), and their DNA was extracted to confirm their genetic profile.

Genetic profiles and secondary symbionts assessment

Following Peccoud et al. [23], we used fifteen highly polymorphic genetic markers (microsatellites) to identify MLGs, including the seven original markers from [23] and eight additional markers from Nouhaud et al. [51] (Table 1). We extracted total genomic aphid DNA by the “salting-out” method, using proteinase-K digestion and precipitation by ethanol [52]. Then, the DNA was resuspended in ultrapure water and set to a concentration of 30 ng/mL. Each locus was amplified separately with fluorescent-labeled primers following the same methodology performed by Peccoud et al. [23]. The electrophoresis of amplified fragments was completed using the same capillary sequencer ABI PRISM 3130xl (Applied Biosystems, U.S.A.), and the same laboratory set-up. MLG characterization was made after the automatic allele calls given by the software GeneMapper (version 3.5 and 3.7, Applera Corp., U.S.A.) and the visual checking of the allele sizes.

Besides carrying the obligate endosymbiont bacteria Buchnera aphidicola, aphids can host a range of facultative endosymbionts that can produce diverse phenotypic effects on the aphid host, including heat tolerance [32, 49]. Hence, we screened five facultative bacterial symbionts commonly reported in A. pisum by randomly selecting pea aphid samples (n = 5) from each clonal lineage. Aliquots of DNA extracted from live samples of APG2 and APG3 lineages were used to amplify diagnostic PCR fragments for Spiroplasma (1500 bp), Serratia symbiotica (1140 bp), Regiella insecticola (840 bp), Rickettsia sp. (600 bp), Hamiltonella defensa (480 bp), and B. aphidicola (270 bp), the obligate aphid symbiont, used as control [42, 43]. The amplification steps were an initial DNA denaturation at 94 °C for 5 min, 30 cycles of 94 ºC for 30 sec, 58 ºC for 30 sec, and 72 ºC for 1 min, and a final extension step at 72 ºC for 5 min. Amplicons were separated through electrophoresis in 1.5% agarose gels, using a GeneRuler 100 bp Plus DNA Ladder (Thermo Fisher Scientific, Waltham, Massachusetts, USA). The presence of each facultative endosymbiont was visually identified according to the size of the diagnostic amplicon in base pairs (bp) obtained from the DNA of each aphid [46].

Temperature and heat-shock effects on life-history traits

To evaluate the performance of individuals from APG2 and APG3 lineages following exposure to short and extreme heat events, we used an already-established protocol used by Montlor et al. [48] and Russell and Moran [53] on the pea aphid using two extreme temperatures registered in central Chile (i.e. 38 and 42 °C; [36-41]). In brief, 40 to 60 age-synchronized aphid nymphs of each lineage were separately kept on V. faba seedlings in three separated climatized chambers at 20ºC (± 0.5°C, LD = 16:8). After 30 min of acclimation, trials started by increasing the temperature of two chambers to 38°C and 42ºC (heat-shock treatments, heating rate = 0.15 and 0.18 °C/min respectively) in the two first chambers respectively, while the third chamber was kept with no temperature change (20ºC constant, control). For each treatment, the temperature rose for 2 h until the trial temperature was reached. Once reached, this temperature was maintained for 4 h and then lowered to 20ºC in the next 2 h (total heat challenge time = 8.5 h). After the trials, individuals were returned to their original acclimation temperature (20°C ± 0.5°C, LD = 16:8) for 48 additional hours to allow for recovery and avoid over-manipulation. Subsequently, after checking that all the plants were well, surviving aphids were counted to obtain the proportion of alive/dead individuals and randomly transferred to new V. faba seedlings until they were reproductive adults. Furthermore, we replicated the control treatment at constant rearing temperatures of 15 and 25ºC (n = 6 individuals per temperature).

Comparisons of reproductive performance were made by using intrinsic growth rates (r_m), calculated as a derivation from the Euler-Lotka equation, where r_m = ln (R₀)/t (see Birch [54]). Fecundity (R₀) of the survivor nymphs was measured as the number of nymphs produced by each focal individual aphid in a lapse of ten days since the birth of its first nymph since this period generally comprises 95% of the total progeny of an aphid [55]. The developmental time (t), in turn, was measured as the number of days elapsed from the focal individual's birth to the birth of its first nymph. R₀ and t were recorded individually for the maximum number of survivors after the trials, for each MLG and heat challenge combination. To summarize, five experimental treatments were used: three constant acclimation temperatures (15, 20, and 25 °C) and two heat-shock temperatures (38 and 42 °C), totaling 154 individual measures of r_m, a consistent number given the clonal nature of these aphid lineages (see [56]).

Data analysis

Since the MLGs used in this study came from the same region previously surveyed by Peccoud et al. [23], a Neighbor-Joining (NJ) tree of MLGs was built with the MLGs found in this study and those found by Peccoud et al. [23] using the seven original microsatellite markers to make both studies comparable. NJ tree was constructed based on a distance matrix of allele shared distances (D_AS) with bootstrapping of 1000 replicates and computed by the software Populations v1.2.31 [57].

To analyze and compare the intrinsic growth rate of each lineage within their thermal niche, we fitted a linear model after a visual diagnostic of the data using constant acclimation temperature and MLG as independent variables and intrinsic growth rate as dependent variable. Then, predicted values of r_m were used to show how detected differences in r_m between genotypes could be translated into an ecologically relevant scenario. We assessed how daily population growth and cumulative population size are comparatively predicted to change as a function of mean diurnal temperatures (08:00 – 18:00 h). This time range was selected to better capture the influence of heat-temperature episodes recorded in the field. First, we obtained temperature data at hourly intervals for the Quinta Normal weather station (33°25’S, 70°42’W) in Santiago for the years 2018, 2019, and 2020, sourced from the Chilean Agricultural Research Institute (https://agrometeorologia.cl/, data downloaded on 12 May 2022). We chose this locality because the effects of thermal change are representative of central Chile [44], this location shows a marked seasonality and it is a centered spot considering the wide latitudinal distribution of the pea aphid along central Chile. Temperature was converted into Kelvin degrees (adding 273°) and then returned to the Celsius scale to avoid possible confounding effects of below-zero temperatures. Then, predicted r_m based on these thermal conditions were separately converted into daily population growth for APG2 and APG3, and subsequently into cumulative population size over a three-year period (1096 days in total, considering one leap year). For simplicity, calculations assume that initial population sizes were identical for both genotypes and negative r_m predicted from the linear regressions were set to zero (i.e., aphids are assumed not to die due to cold, which is likely realistic during the growth period of aphid populations). On the other hand, to estimate how different MLG responded to the heat-shock treatment, we performed a multiple regression including rm as the dependent variable, genotype as a categorical factor, and heat-shock temperature as a covariable.