Pest and pathogens are known to cause major damage in commodity production landscapes, resulting in not only great financial losses, but also other socio-economic crises, such as decreased job and food security (Wingfield et al. 2008; Gregory et al. 2009). The polyphagous shot hole borer (PSHB, Euwallacea fornicatus) and its symbiotic fungus, Fusarium euwallaceae, have infested deciduous fruit orchards in the Western Cape province of South Africa (de Jager and Roets 2023; Engelbrecht et al. 2024). Recently, PSHB was observed also infesting cultivated European olive trees (Olea europaea subsp. europaea) in the same landscape (Fig. 1). South Africa is the largest olive oil producer in sub-Saharan Africa, of which 95% of production is in the Western Cape province (SA Olive 2023). In addition, the native African olive, Olea europaea subsp. cuspidata, often occur in patches of indigenous vegetation in the matrix of cultivated olive plantations and may also be attacked by PSHB (Fig. 1). The potential negative impact of this paninvasive beetle–fungus complex could have far-reaching consequences for both commodity production and species conservation.

Cultivated Olea europaea is known to be attacked by PSHB in California, Israel, and South Africa (Mendel et al. 2021; van Rooyen et al. 2021), and early on has been assumed as fusarium disease susceptible (Eskalen et al. 2013). A recent study from Israel, however, showed that F. euwallacea growth on fresh olive sawdust was relatively inhibited (Freeman et al. 2019). In South Africa, many fungi associated with cultivated and wild olives have been isolated (Spies et al. 2020), some subsequently showing signs of pathogenicity (van Dyk et al. 2021). However, Fusarium isolates were not considered in these studies, as they were not considered clear dieback or decline pathogens (Spies et al. 2020). Therefore, there are no published records describing the pathogenicity of Fusarium euwallaceae to Olea species, wild or cultivated.

Fig. 1
figure 1

Symptoms of attack by Euwallacea fornicatus (PSHB) and inoculation with Fusarium euwallaceae on African olive (Olea europaea subsp. cuspidata) and European olive trees (Olea europaea subsp. europaea) trees. (A) African olive with three entry holes of PSHB on trunk (arrows) that resulted in gumming. (B) European olive with four entry holes of PSHB on trunk (arrows) that resulted in gumming. (C) European olive, with bark removed to show entry holes (arrows) through bark and cambium. (D) European olive, with cambium removed to show staining in xylem caused by F. euwallaceae around PSHB entry hole (arrow). (E) Exposed gallery of PSHB (arrows) within a branch of European olive

Full size image

In this study, three isolates were used in pathogenicity trials of F. euwallaceae towards European and African olives in South Africa. Two were obtained from infested pear trees in an orchard adjacent to the PSHB-infested olive grove (CMWIA6007 and CMWIA6008 from Engelbrecht et al. 2024) on Lourensford Estate (-33°35′36″S, 18°55′53″E) and one (FR0420) originated from the gallery of a PSHB infested olive tree at this same site. The identity of this isolate was confirmed by comparisons of its elongation factor 1-α DNA region, which was identical, to that of the ex-holotype strain of F. euwallaceae (GenBank JQ038007) following methods of De Jager and Roets (2023). Eight mature individuals of Olea europaea subsp. europaea cv. Mission and eight individuals of Olea europaea subsp. cuspidata, growing in pots in a nursery (TreesSA, Stellenbosch, South Africa), were inoculated with each of the three F. euwallaceae isolates in August 2021 using F. euwallaceae-colonised toothpicks, following De Jager and Roets (2023). Uncolonized toothpicks served as controls. After four weeks of incubation, the outer layer of the bark was removed around the inoculation points using a sterile knife to expose and measure the resulting lesions (brown staining of vascular tissues). Re-isolations were made from these lesions following De Jager and Roets (2023) to confirm F. euwallaceae as the causal agent of the lesions.

Lesion length data (mm) were analysed using R software (version 3.6.3). Data were non-normally distributed (Shapiro–Wilk test, W = 0.6198, p < 0.001) and subsequently evaluated with a generalized linear mixed-effects model fitted to a gamma distribution (lme4 package, Bates et al. 2014). Treatment, host species, and their interaction were used as fixed effects and tree individual was used as a random effect in the full model. The best model was selected from this full model by evaluating all combinations of variables and selecting the combination that resulted in a model with the lowest AIC value. Significant main effects in the selected model were separated using pairwise testing (emmeans package; Lenth et al. 2023).

The best model contained treatment, host species, and their interaction as explanatory variables. This model had an AIC value of 164.7 and a deviance of 472.89 at 63 degrees of freedom. Pairwise testing revealed that lesion lengths after inoculation with F. euwallaceae on the cultivated olives were significantly longer than those on the wild olives and the controls but similar in length compared to each other (Fig. 2). Lesions caused by controls on both hosts were the same length and were also similar to those caused by inoculation of F. euwallaceae on wild olives (Fig. 2). Fusarium euwallaceae was consistently re-isolated from all inoculated wounds on the cultivated olives but not their controls. Fusarium euwallaceae was only rarely re-isolated from inoculated wounds on the wild olive (45.83% recovery after 4 weeks) but was never isolated from the controls.

Fig. 2
figure 2

Lesion lengths after four weeks of incubation of three isolates of Fusarium euwallaceae inoculated into cultivated and wild olive trees. Isolate1 = CMWIA6007; Isolate2 = CMWIA6008; and Isolate3 = FR0420. Centre lines show the medians; box limits indicate the 25th and 75th percentiles as determined by R software; whiskers extend 1.5 times the interquartile range from the 25th and 75th percentiles; and outliers are represented by dots. Different letters indicate significant differences in lesion lengths (p < 0.05) (n = 8). Figure created using BoxPlotR (Spitzer et al. 2014)

Full size image

This is the first report of F. euwallaceae as pathogenic to cultivated European olive trees. The fungus is not pathogenic to the closely related African olive, a native species that may occur in patches of indigenous vegetation adjacent to European olive orchards. While this result bodes well for species conservation, the potential threat to commercial production of olive oil and olives is highlighted.

PSHB beetles bore straight into xylem, and when F. euwallaceae establishes and grows, it occludes the affected vessels (Freeman et al. 2013). It is known that percentage loss in xylem conductivity (water transport) due to pests and pathogens is a more rapid contributor to stand-level mortality when compared with other pest and pathogen damage e.g., to phloem (Dietze and Matthes 2014). Hydraulically impaired trunks and branches could lead to yield losses or tree dieback; this is especially relevant given that the semi-arid and Mediterranean-type climates of regions where olives are planted in in South Africa are becoming increasingly drought prone (Naik and Abiodun 2020). There currently is no information about the interaction between F. euwallaceae pathogenicity and the actual levels of plant stress experienced by the olive trees. Moreover, PSHB may create tens to hundreds of entry holes in tree trunks, inviting co-infection by opportunistic fungi that are known to be pathogenic to European and African olives in South Africa (van Dyk et al. 2021). More research is warranted to understand the potential impacts of these infestations on commercial European olive orchards.