Expansion of the rare Eucalyptus risdonii under climate change through hybridization with a closely related species despite hybrid inferiority

Abstract

Background and aims: Hybridization is increasingly recognized as an integral part of the dynamics of species range expansion and contraction. Thus, it is important to understand the reproductive barriers between co-occurring species. Extending previous studies that argued that the rare Eucalyptus risdonii was expanding into the range of the surrounding E. amygdalina by both seed and pollen dispersal, we here investigate the long-term fitness of both species and their hybrids and whether expansion is continuing.

Methods: We assessed the survival of phenotypes representing a continuum between the two pure species in a natural hybrid swarm after 29 years, along with seedling recruitment. The performance of pure species as well as of artificial and natural hybrids was also assessed over 28 years in a common garden trial.

Key results: In the hybrid zone, E. amygdalina adults showed greater mortality than E. risdonii, and the current seedling cohort is still dominated by E. risdonii phenotypes. Morphologically intermediate individuals appeared to be the least fit. Similar results were observed after growing artificial first-generation and natural hybrids alongside pure species families in a common garden trial. Here, the survival, reproduction, health and growth of the intermediate hybrids were significantly less than those of either pure species, consistent with hybrid inferiority, although this did not manifest until later reproductive ages. Among the variable progeny of natural intermediate hybrids, the most E. risdonii-like phenotypes were the most fit.

Conclusions: This study contributes to the increasing number of reports of hybrid inferiority in Eucalyptus, suggesting that post-zygotic barriers contribute to the maintenance of species integrity even between closely related species. However, with fitness rapidly recovered following backcrossing, it is argued that hybridization can still be an important evolutionary process, in the present case appearing to contribute to the range expansion of the rare E. risdonii in response to climate change.

Keywords: Eucalyptus amygdalina; Eucalyptus risdonii; fitness; hybrid inferiority; hybrid swarm; hybrid zone; hybridization; outbreeding depression; post-zygotic barriers.

Fig. 1.

Morphology of the pure species Eucalyptus amygdalina (left) and E. risdonii (right), and a first-generation hybrid (middle). Photographs highlight the narrow, green and alternate leaves of E. amygdalina (left), in contrast to the broad, glaucous and fused leaves of E. risdonii (right). The intermediate leaf phenotype (H), with intermediary leaf shape, level of glaucousness and opposite leaves with petiole, is shown in the middle picture. This intermediate is a putative F₁ hybrid found in open-pollinated seed collected in the wild from a pure E. amygdalina phenotype. We thank Robert Wiltshire for providing the photograph of E. amygdalina.

Fig. 2.

Long-term climate patterns for the Eucalyptus amygdalina and E. risdonii hybrid zone at Government Hills (left) and the common garden trial (right). The graphs show 5 year moving average mean annual temperature (top graphs) and the standardized precipitation evapotranspiration index (SPEI) (bottom graphs). SPEI values below zero indicate water deficit and above zero indicate water surplus. The grey shading corresponds to the period between the first and last monitoring of the trees at the hybrid zone (1990–2019) and the trial growing period (1991–2019). The red line represents the overall average for the monitored sites (1911–2019), the blue line represents the historical climate average prior to the detectable signature of climate change in the southern hemisphere (Abram et al., 2016) – also referred as the pre-industrial warming climate period (1911–1959), the green line corresponds to the climate average often used to represent the contemporary climate (1976–2005) and the grey line corresponds to the climate average during the monitoring period at the hybrid zone (1990–2019) and common garden trial (1991–2019). Arrows indicate known wildfire events that burnt the E. amygdalina × E. risdonii hybrid zone.

Fig. 3.

Location and composition of the two studied sites in Tasmania. The right-hand map shows the area comprising the common garden trial (blue rectangle). The bottom map represents the studied area at Government Hills. The hills are dominated by Eucalyptus amygdalina which ranges from a tree form on the southern slopes to a low mallee (2–3 m high) on the northern slopes. The yellow dashed line delineates the boundary of the main population of pure E. risdonii close to the hybrid zone. White squares, circles and triangles represent the location of the trees that were assessed in this study within the E. amygdalina, E. risdonii and hybrid zones, respectively. The orthophoto and simple basemaps used in this figure were obtained from theLIST ©State of Tasmania.

Fig. 4.

Survival rates of Eucalyptus amygdalina, E. risdonii and hybrids located at a natural hybrid zone and species pure zones at Government Hills since 1990. The significance of the overall Pearson’s χ² test comparing all phenotypes within the hybrid zone is shown on the upper right-hand corner. Significant differences (P < 0.05) in survival rates were found between E. amygdalina and E. risdonii from the pure zones and between H phenotypes and both backcrosses (AH + RH) and R in the hybrid zone. Leaf shapes are shown to demonstrate the large morphological differences between these eucalypt species and their hybrids (Dungey et al., 2000). Phenotype abbreviation: A = E. amygdalina-like; AH = hybrid with phenotype lying outside the range of the pure species but closer to E. amygdalina; H = hybrid with phenotype equally unlike either pure species (intermediate); RH = hybrid with phenotype lying outside the range of the pure species but closer to E. risdonii; R = E. risdonii-like.

Fig. 5.

Seedling recruitment of Eucalyptus amygdalina, E. risdonii and hybrids located in the natural hybrid zone at Government Hills in 2019. Bar plots indicate the percentage of seedlings (solid fill) and mature plants (striped) assessed to be in the E. amygdalina-like (A + AH), intermediate (H) and E. risdonii-like (R + RH) phenotypic classes. The significance of an overall Pearson’s goodness-of-fit χ² test comparing frequencies of phenotype in seedlings and the putative maternal cohort is shown in the upper right-hand corner. Statistically significant differences (P < 0.05) were found between E. risdonii-like (RH + R) and E. amygdalina-like (AH + A) phenotypes and between E. risdonii-like (RH + R) and the intermediate (H) phenotypes when comparing the observed number of seedlings with the expected number of seedlings based on the frequencies of the surrounding adults.

Fig. 6.

Percentage survival of Eucalyptus risdonii, E. amygdalina and their hybrids in the common garden trial over 28 years by (A) cross type and (B) phenotypic class within the natural advanced generation hybrids (F₂op). The significance of the log-rank test for the difference in percentage survival between cross types or phenotypic classes at age 28 years is indicated in the upper right-hand corner of each plot. Cross types are: A = pure E. amygdalina; F₁ = first-generation hybrid between E. risdonii × E. amygdalina; F₂op = natural advanced hybrids between E. risdonii × E. amygdalina. Phenotypes are: A = E. amygdalina-like; AH = hybrid with phenotype lying the outside the range of the pure species but closer to E. amygdalina; H = hybrid with phenotype equally unlike either pure species (intermediate); RH = hybrid with phenotype lying outside the range of the pure species but closer to E. risdonii; R = E. risdonii-like.

Fig. 7.

Hazard ratios for Eucalyptus amygdalina, E. risdonii and their hybrids in the common garden trial. The numbers and plots represent the relative risk of death for each (A) cross type and (B) phenotype within the advanced hybrids (F₂op) compared with the least fit pure species E. amygdalina. Asterisks indicate significant differences (*P < 0.05, **P < 0.01; ***P < 0.001). The overall test for the difference between cross types or phenotypic classes is indicated in the bottom left-hand corner of each graph (global P-values). Cross types are: A = pure E. amygdalina; F₁ = first-generation hybrid between E. risdonii × E. amygdalina; F₂op = natural advanced hybrids between E. risdonii × E. amygdalina. Phenotypes are: A = E. amygdalina-like; AH = hybrid with phenotype lying outside the range of the pure species but closer to E. amygdalina; H = hybrid with phenotype equally unlike either pure species (intermediate); RH = hybrid with phenotype lying outside the range of the pure species but closer to E. risdonii; R = E. risdonii-like.

Fig. 8.

Reproduction and health assessments of Eucalyptus amygdalina, E. risdonii and their hybrids located in the common garden trial. Bar plots represent the percentage of healthy (A) and reproductive (C) individuals per cross type and per phenotypic class within natural advanced hybrids (B, D). Numbers inside the bar plots represent the total number of individuals in each cross type or phenotypic class. The significance of the overall Pearson’s contingency χ² test is indicated on the upper right-hand corner of each graph. Cross types are: A = pure E. amygdalina; F₁ = first-generation hybrid between E. risdonii × E. amygdalina; F₂op = natural advanced hybrids between E. risdonii × E. amygdalina. Phenotypes are: A = E. amygdalina-like; AH = hybrid with phenotype lying outside the range of the pure species but closer to E. amygdalina; H = hybrid with phenotype equally unlike either pure species (intermediate); RH = hybrid with phenotype lying outside the range of the pure species but closer to E. risdonii; R = E. risdonii-like.