Cerambyx cerdo and Cerambyx welensii Oak-Living Sympatric Populations Exhibit Species-Specific Responses to Face Ecological Factors in the Wild

Abstract

Oak open woodlands (dehesas) have outstanding socioeconomic and ecological values, sustain traditional agro-silvo-pastoral uses, provide high-value ecosystem services, and constitute key biodiversity hotspots. Cerambyx cerdo and Cerambyx welensii are two large, oak-living, wood-boring, sympatric longhorn beetles (Coleoptera: Cerambycidae) that may reach pest status in SW Spain, contributing to oak decline. Understanding species-specific habitat requirements determining occupancy–abundance patterns is needed to develop management or control strategies. We conducted a large-scale, four-year study using 1650 feeding traps to ascertain longhorn abundance and species-specific habitat suitability in relation to 18 ecological variables, 9 biotic (oak species, forest mass, trunk diameter, tree density, basal area, forest cover, shrub cover, ground cover, oak renewal), and 9 abiotic (bedrock outcrops, altitude, ground slope, aspect, mean temperature: annual/July/January, annual precipitation, insolation). Results showed that longhorn abundance was sensitive to most ecological variables and to many interactions between them. Interestingly, interactions between ecological variables and longhorn species were widespread, signifying that responses were species-specific and therefore predictive Generalized Linear Mixed Models (GLMMs) were different between species. Our research contributes to the understanding of the ecological factors that shape longhorn species-specific occupancy–abundance patterns, delves into their sympatric relationship, and contributes toward improving sustainable forest practices that will mitigate longhorn impact in oak open forests.

1. Introduction

Oak forests are widespread across the western Palaearctic and particularly in the Mediterranean Basin. In SW Iberia, oaks mostly appear as anthropogenic open woodlands, similar to the savannah landscape, the so-called dehesas in Spain and montados in Portugal. Oak open woodlands have outstanding socioeconomic and ecological values, sustain traditional agro-silvo-pastoral uses that model landscape multiplicity, provide high-value global ecosystem services (carbon sequestration, air purification, pollination, erosion prevention, climate regulation), and constitute key biodiversity hotspots in Europe [1,2,3,4,5,6,7,8]. Due to its high environmental significance, the dehesa/montado ecosystem is protected under the EU Habitats Directive [9].

Oak forests have experienced repeated decline and mortality events during the last three centuries, a circumstance that seems to have accentuated in the last decades [10,11]. Oak decline is currently viewed as a complex syndrome resulting from an array of interacting abiotic and biotic factors that can behave as predisposing, contributing, or inciting variables [12,13]. Among the pests that can impact the life of oak trees, bark and wood-boring insects have notable importance [10,14,15], including beetles in the family Cerambycidae, called longhorns due to their long antennae.

Two large longhorn species, Cerambyx cerdo L. (Cc) and Cerambyx welensii (Küster) (Cw), stand out due to their large body size, wide geographical distribution, and close trophic association with oak trees. They have been usually viewed as polyphagous [16,17,18], but host trees other than oaks are at present considered unusual [19]. In SW Iberia, their host range is almost restricted to oaks (Quercus spp.), mainly holm oak (Q. ilex L.), cork oak (Q. suber L.), and pyrenean oak (Q. pyrenaica Willd.) [7], species that constitute the main dehesa arboreal component. Cc and Cw are among the few saproxylic beetles whose larvae develop in living wood of healthy or decayed host trees [20,21,22], a specialized feeding regime in line with the narrow host range characterizing primary saproxylics [23,24]. Cc and Cw are comprised in the highly diverse assemblage of saproxylic insects, a functional group essential in the process of wood degradation and formation of cavities in trees, which are later used as shelter for an array of animals, including arthropods, reptiles, birds, and mammals. Therefore, they are viewed as “ecosystem engineers” [25,26] that improve forest biodiversity and are indicators of high-quality mature habitats [23,24,27,28,29,30].

Nevertheless, damage caused by larvae to oaks can be intolerable when longhorn populations become excessive. Larval damage initiates with subcortical galleries in the sapwood, which may alter sap flow and produce wilting, defoliation, vigor loss, and decline of oaks [8]. As the larvae grow, larger galleries in the sapwood and heartwood produce considerable physiological, mechanical, and structural damage, so that affected trees, with wood resistance debilitated by fiber breakage, may eventually collapse [8]. In addition, adult exit holes and other open larval galleries can function as gateways for oak pathogenic microorganisms [31].

Longhorn impact on oak open woodlands has risen noticeably in SW Spain in the last decades, so that it is thought to be a crucial variable in oak decline and a threat to the dehesa long-term conservation [8,32]. Longhorn damage infringed to oaks has customarily been attributed to Cw in SW Spain, but recent research shows that this statu quo is unrealistic, and that Cc is also a major driver to oak decline [7,8]. This is a non-trivial fact as Cc is an EU protected species [9], which seems to be against the simultaneous protection of the dehesa ecosystem in the Habitats Directive [33,34]. The situation is further complicated because Cc and Cw share a similar ecological habitat, often occur sympatrically, and may mate interspecifically and even hybridize given their close phylogeny [35]. Even so, a recent study by Torres-Vila et al. [8] shows that there is notable inter-specific variability in Cc and Cw occupancy–abundance patterns in SW Spain.

In such a complex phytosanitary and legal scenario, we need a deeper knowledge of those ecological factors driving Cc and Cw abundance, especially in oak forests in which both species co-occur. Understanding the mechanisms regulating the coexistence of sympatric species that compete for resources is a central issue to community ecology [22,36,37,38,39,40], but it is also essential to build risk analysis models and implement management strategies allowing the long-term conservation of the dehesa ecosystem [8]. These actions are especially necessary in a forest context in which tools to manage oak-associated bark and borer insects are almost lacking [14,41], and they are relevant in the current climate change scenario, which could aggravate oak decline both directly by the effect of global warming on trees and indirectly by intensifying the pressure from xylophagous species under warmer conditions [15,33,42].

In the described framework, the aim of this study was to investigate the ecological drivers determining the abundance of Cw and Cc in oak woodlands at a regional scale. The objective was twofold: (1) to assess the effect of an array of biotic and abiotic ecological factors on the abundance of Cw and Cc populations estimated with feeding traps, and (2) to identify those ecological and silvicultural factors that could shape species-specific differences in the occupancy–abundance patterns of both longhorns.

2. Materials and Methods

2.1. Study Area

The study area covered the whole region of Extremadura in SW Spain (41,634 km²), which houses extensive and diverse oak forests (more than 1 million ha), most of them open woodlands. Native oak species are, in order of importance, holm, cork, and pyrenean oaks. Climate type is Mediterranean with hot/dry summers (up to 40 °C) and mild/rainy winters. Mean annual temperature is close to 16 °C (range 10–18 °C) and mean annual precipitation about 700 mm (range 360–1500 mm) [43]. Altitude ranges from about 150 m to more than 2000 m, although the mentioned oaks seldom live above 1100–1200 m (holm and cork oak) or 1500–1600 m (pyrenean oak) [7].

2.2. Study Species

Cc and Cw life cycle and behaviour in the study area have been broadly described in previous papers of our research team [7,8,22,34,44,45,46,47]. In brief, these are univoltine species flying in May–August, adult diel activity being mainly crepuscular and nocturnal. Adults often feed on sapflows from trees. After mating, females lay almost 150 eggs on average, singly or in small groups, in the bark cracks of the host tree. Eggs hatch in 12–14 days, neonate larvae burrow the tree bark, and feed in the outer sapwood. As larvae grow, they enter the heartwood and pierce increasingly wide and long galleries. Larval development lasts 2–3 (4) years, depending on environmental conditions, and upon reaching maturity, larvae pupate inside a pupal cell in late summer-early autumn. The adult emerges 30–40 days later but overwinters inside its pupal cell until it leaves the tree in the spring of the following year through an exit hole, reinitiating the lifecycle. Adult longevity in the wild ranges 2–3 weeks (maximum 2 months) but longer lifespans (4–5 months) have been documented in laboratory conditions [34,44,45,46,47]. In oak open woodlands, adults are mostly sedentary and stay on the same tree for most of their lifespan, even though they can perform dispersal flights of several kilometers [21,47,48].

2.3. Abundance Estimates: Feeding Traps

We used adult catches obtained with feeding traps to assess the abundance of each longhorn species. Adult trapping rather than larval damage was chosen because the latter method is unsuitable for specific abundance evaluation as exit hole morphology has no taxonomic value [8,22]. Adult catch data were acquired for four years (2017–2020) using the traps from the SSV Regional Sampling Network [7,8]. Traps are built with 5 l cylindrical PET plastic containers cutting the upper part at the neck border and turning it around as a funnel. Bait consists of red wine, vinegar, sugar, salt (2 L, 100 mL, 500 g, 500 g), and water until completing 5 l of solution. Traps were set in the oak trunks at 1.4–1.5 m above ground, usually facing north to minimize bait evaporation, arranged in groups (n = 809 stands) of 2 or 3 (50 m apart), more rarely isolated or in larger groups, depending on the topography and forest structure (Table S1). We set 1657 traps but seven were lost, obtaining capture data from only 1650 traps. Traps were disposed covering the entire studied area over the altitude range of oaks (180–1550 m) and geolocated (WGS84 latitude/longitude). Trapping was operative from mid-May to late-August every year and traps were checked a variable number of times, from once per week to none (only trap setting and retrieval), depending on the year and the use of catch data for collateral studies. Between 0.5 and 2.5 L of bait was added to each trap, depending on the checking scheduled. Captured adults were taken to the laboratory for identification in labeled bags and plastic containers. See [7,8] for additional information on the trapping protocol.

2.4. Statistical Analyses

An array of 18 categorical/quantitative predictor variables were measured or estimated in each forest stand from different data sources [43,49,50] (Table 1). Generalized Linear Mixed Models (GLMMs) were computed to explore the effects of these ecological variables on longhorn abundance, with the trap nested to the year as random factors (the random structure was selected using the Akaike information criterion (AIC) in all models). Catch number per trap per year (sexes pooled within longhorn species) was the explained variable, which was considered a robust proxy of the abundance of each target species. Predictor variables were scored for subsequent GLMMs as either biotic (ref. 1–9) or abiotic (ref. 10–18) according to their forestry, dendrometric, physiographic, geological or climatic features (Table 1). The explained variable was log (ln) transformed to deal with zero-inflated data (trap subset without catches) and all GLMMs were computed with the glmmTMB library [51]. Poisson family errors were applied and tested for over- and under-dispersion. Residuals plots were checked to verify assumptions of the fitted models.

We first computed 18 two-way GLMMs to explore the effects on abundance of each predictor variable, the longhorn species, and their interaction (in order to assess species-specific responses) considering both factors with fixed effect. We also performed additional two-way GLMMs to resolve some interactions between predictors (e.g., ground slope x aspect), and especially to confirm the effect of some predictors when controlling for the confounding effect of correlated variables (e.g., precipitation effect controlling for altitude and oak species; or insolation effect controlling for temperature). Next, we built for each longhorn species two GLMMs with either biotic or abiotic variables, and then we built a total GLMM pooling the best predictor variables, irrespective of their biotic or abiotic nature. To do this, we first computed the full/saturated model, including all explanatory variables as well as their second- and third-order interactions, and then we followed a backward elimination process [52]. Once biotic and abiotic GLMMs were fitted, we pooled the two models within species to start a new backward elimination process. This was performed by only introducing those predictors fitted in the biotic and abiotic models to avoid overparameterization. Some variables were not considered in GLMMs (biotic, abiotic, or pooled) given their strong collinearity with other used variables: oak species (included in forest mass), basal area (a combination of trunk diameter and tree density), shrub cover (closely related to ground cover), and the three temperatures (all negatively correlated with altitude). Analysis of Deviance (Type II Wald Chi-square test) was used to assess the effect on abundance of the predictor variables and their interactions in all GLMMs. Post hoc Tukey contrasts (z values, p < 0.05) were used when necessary for multiple comparisons of means in order to establish homogeneous groups. All statistical analyses were computed with R 4.1.2. (R Core Team [53]).

3. Results

Results from GLMMs showed three important preliminary aspects: (1) seasonal catches per trap (a proxy of abundance) were overall higher in Cw than in Cc, (2) most ecological variables, both biotic and abiotic, had a significant impact on longhorn abundance, and (3) last but not least, the interactions between ecological variables and longhorn species were widespread, so that the response to the same ecological variable was usually species-specific (Table 2). Abundance showed a curvilinear rather than linear relationship with six predictor variables, either bell-shaped in the case of tree density and forest cover (higher abundances occurred with intermediate predictor values), asymptotic in the case of trunk diameter, or exponential-like in the case of the three temperatures (Figure 1, Table 2).

Oak species significantly affected longhorn abundance, but the response pattern of each longhorn was dependent on host oak tree (Cw: C > H > P; Cc: H > C = P; Figure 1A). A similar response was found when analyzing the forest mass, with a clear trend in mixed stands to show intermediate abundance values between the respective pure stands, especially in the case of Cw (Figure 1B). The effect on abundance of trunk diameter (a proxy for tree age) strongly depended on longhorn species: in Cw the abundance increased with trunk diameter until reaching a plateau in mature oaks (60–70 cm), while in Cc the pattern was just the opposite (Figure 1C), so that both species had a similar abundance in stands with trees 30–40 cm in diameter.

Tree density significantly impacted longhorn abundance following a bell-shaped relationship, with maximum abundance in Cw occurring in less dense woodlands (40 trees ha⁻¹) than in Cc (70 trees ha⁻¹) (Figure 1D). The effect of basal area on abundance was clearly dependent on the species, since Cw showed a slight rise in abundance with increasing basal area, while Cc showed a marked decline (Figure 1E). As observed with tree density, forest cover significantly affected longhorn abundance according to a bell curve, reaching both species maximum abundances with a 35–40% coverage (Figure 1F).

The effect of increasing shrub cover on longhorn abundance was noticeably opposite in both species, rising in Cc and falling in Cw (Figure 1G). A comparable trend was found with ground cover, as Cc was less abundant in stands with pure pasture and Cw in stands with undergrowth, even if abundance in both longhorns was slightly higher in those stands with mixed pasture-undergrowth (Figure 1H).

The effect of oak renewal (resprouting) on abundance was species-specific: while Cw showed just a slight decline with resprouting presence, Cc exhibited a marked rise (Figure 1I). Interestingly, a similar species-specific response was found when there were bedrock outcrops in the soil (Figure 1J).

Altitude was negatively related to abundance in both species without altitude x species interaction, despite the line slopes suggesting higher abundance of Cw than Cc below 600 m (Figure 1K). Similarly, abundance was negatively related to ground slope in both species, with no ground slope x species interaction (Figure 1L).

Aspect of the forest mass did not significantly affect longhorn abundance, with no aspect x species interaction, even if there was an underlying trend to increase abundance in warm aspects (Figure 1M). A complementary two-way GLMM showed that the effect on abundance of the interaction aspect x ground slope was significant in Cw (Wald test, Chi² = 18.78, df = 7, p < 0.01) but not in Cc (Chi² = 9.98, df = 7, p = 0.19 ns).

The relationship between longhorn abundance and temperature (T_m, T_m7 and T_m1, Table 1) was positive and curvilinear, the quadratic component of the models having a significant effect (Table 2). However, there were interesting interspecific differences depending on the temperature considered: with T_m (Figure 1N) and T_m7 (Figure 1O) there was temperature x species interaction while with T_m1 there was none (Table 2) even if curves suggest otherwise (Figure 1P).

Annual precipitation had a strong negative impact on longhorn abundance. There was precipitation x species interaction as the line slope was slightly (but significantly) steeper in Cw than in Cc (Figure 1Q). The adverse impact of rainfall on longhorn abundance remained significant when controlling for the confounding effect of altitude, as both predictors are positively correlated (R² = 0.51, p < 0.01). This occurred both in Cc (Wald test, precipitation: Chi² = 120.10, df = 1, p < 0.001; altitude: Chi² = 6.98, df = 2, p < 0.05; precipitation x altitude: Chi² = 2.087, df = 2, p = 0.24 ns) and Cw (Wald test, precipitation: Chi² = 37.85, df = 1, p < 0.001; altitude: Chi² = 73.98, df = 2, p < 0.001; precipitation x altitude: Chi² = 21.70, df = 2, p < 0.001) (altitude scored into three classes for these particular analyses: <400, 400–600 and >600 m).

The adverse impact of rainfall on longhorn abundance also remained significant when controlling for the confounding effect of oak species, as in mountainous areas (more rain) the pyrenean oak prevails while in valleys and lowlands (less rain) it is the holm oak, cork oak being usually more abundant in midlands (intermediate rain levels). This occurred both in Cc (Wald test, precipitation: Chi² = 114.26, df = 1, p < 0.001; oak: Chi² = 15.45, df = 2, p < 0.001; precipitation x oak: Chi² = 2.17, df = 2, p = 0.34 ns) and Cw (Wald test, precipitation: Chi² = 122.34, df = 1, p < 0.001; oak: Chi² = 53.86, df = 2, p < 0.001; precipitation x oak: Chi² = 1.38, df = 2, p = 0.50 ns).

Insolation was positively correlated with longhorn abundance, but abundance rise with increasing insolation was more pronounced in Cc than in Cw (Figure 1R). The positive effect of insolation on abundance persisted when controlling for the confounding effect of temperature (T_m), both in Cc (Wald test, insolation: Chi² = 76.00, df = 1, p < 0.001; T_m: Chi² = 148.24, df = 1, p < 0.001; insolation x T_m: Chi² = 0.21, df = 1, p = 0.65 ns) and Cw (Wald test, insolation: Chi² = 5.88, df = 1, p < 0.05; T_m: Chi² = 161.04, df = 1, p < 0.001; insolation x T_m: Chi² = 13.82, df = 1, p < 0.001) (T_m scored into two classes for these particular analyses, above and below 16 °C, the regional mean temperature). The Chi-square values of these complementary analyses also confirm the higher impact of insolation on Cc than on Cw. Comparable results were obtained when T_m7 was used in lieu of T_m.

The best GLMMs we found to predict Cc and Cw abundance are summarized in Table 3. Regarding biotic models, forest mass, trunk diameter, tree density, and ground cover contributed to the models of both longhorn species, as did some interactions between them or with other variables. Oak renewal contributed only to the Cc model while forest cover only contributed to the Cw model, with a significant quadratic effect. Forest cover by itself had no effect on Cc abundance, but this variable participated in the model because of its significant interactions with other variables such as forest mass and ground cover (Table 3).

Regarding abiotic models, bedrock outcrops (at the significance limit in the Cw model), altitude, and annual precipitation contributed to the models of both longhorn species, as did the bedrock outcrops x altitude interaction. Insolation contributed only to the Cc model while ground slope only contributed to the Cw model. Aspect by itself had no effect on the abundance of any longhorn species, but this variable was included in both models because of its significant interactions with either annual precipitation (Cc model) or altitude (Cw model) (Table 3).

In the final models that pooled all the predictor variables (biotic and abiotic), altitude and insolation came out of the Cc model and ground slope came out of the Cw model. Oak renewal only contributed to the Cc model, while forest cover and altitude only contributed to the Cw model. Forest cover and aspect by themselves had no effect on Cc abundance but participated in the model due to their interaction with other variables. The effects of tree density and bedrock outcrops, as well as that of forest mass x forest cover interaction, were just marginally significant in the Cw model (Table 3).

4. Discussion

Environmental factors are expected to strongly affect habitat suitability and demographic parameters, as animals actively select (or passively persist in) certain habitats, so that spatial distribution patterns are ultimately a consequence of survival and adaptive pressures shaping reproductive fitness [54,55,56]. The abundance of Cc and Cw was extremely sensitive to most ecological variables, as well as to many (sometimes complex) interactions between them. More interestingly, our results substantiate that the response to ecological variables of Cc and Cw, two sympatric and closely related longhorns, is often strongly species-specific, which determines distinctive spatial occupancy–abundance patterns [7,8]. A major advantage of our study was that Cw and Cc trapping data were obtained in the same spatiotemporal context in which both species coexisted and faced the same environmental pressures, so that acquired data were unbiased and directly comparable [47]. In some points of the discussion below, we will invoke interspecific differences in larval niche to interpret the species-specific effects of some ecological variables: larval niche preference in our target longhorns promotes larval stratification along host tree height, with Cw being prevalent in the fork/branches and Cc in the trunk/base [22,57,58,59,60,61].

The abundance of each longhorn depended on host oak species, which in turn largely determines their spatial distribution [8]. Data also revealed a “proportion-effect” of the forest mass because of oak species composition, so that abundance in mixed stands showed intermediate values between those of the respective pure stands. A similar fact was previously reported regarding oak shoot browning [13]. Host oak species-related factors regulating Cc and Cw abundance are poorly understood but they must include wood features such as hardness, durability, tendency to split, bark depth, and the perennial/deciduous nature of oaks [8,21,62]. The effect of trunk diameter (a proxy for tree age) on abundance was dissimilar between the two longhorns: Cw was more abundant in old oaks while Cc was more abundant in young oaks, this last result being contrary to what might be expected [62]. A broad-focused interpretation is needed here as oak species (especially pyrenean oak) have a high capacity for root growth and resprouting [63], so that an apparently young forest could actually be much older than expected and provide abundant underground wood [64,65], which will favor Cc more than Cw given their distinct larval habitat preference. The effect of increased basal area on abundance was also opposite between species (higher in Cw and lower in Cc), a response pattern similar to that obtained with trunk diameter. However, an increase in basal area may not represent an actual increase of wood volume per hectare available for longhorn larvae if old trees have the inner wood degraded or are even hollow, a frequent situation in SW Spain dehesas. The underlying effects of underground wood availability and aging-related hollow trunks question the idea that the number of trees required for the survival of Cc populations is lower at large tree sites than at small tree sites [62]. Moreover, our data challenge the widespread view held for decades that Cc performs better in old oak trees [21,57,61,62,66,67]. Such a discrepancy may be explained if old trees do provide better long-term deadwood-associated microhabitats for adults to protect themselves from adverse factors such as predators or unfavorable climate, while mature (but not old) host trees do provide more and better wood resources for feeding larvae. If so, those oak stands with presence of both mature and old trees would be optimal to balance larval and adult requirements, thus maximizing population abundance.

Longhorn abundance in open woodlands peaked with tree density ranging 40–70 trees ha⁻¹ and forest cover ranging 30–40%. Such a bell-shaped pattern suggests a compromise between (1) a minimum number of trees per surface unit providing a plenty wood volume for larval feeding, and (2) a not excessive number of trees per surface unit (in the extreme case with overlapping treetops) constraining adult flight and especially solar radiation. High tree densities/forest covers may hinder adult activities by demanding greater flight performance between/under treetops, or even by decreasing lighting below the forest canopy at dusk-early night when most adults are active, as longhorns are not skillful fliers [47]. More importantly, high densities/covers critically restrict trunk heating by direct solar radiation, especially in cold seasons (autumn-winter) when the sun ecliptic is lower. This is important because a heat deficit (see also temperature results below) may limit larval growth rate and decrease adult size, which impact fitness both in Cc [34] and Cw [46]. At worst, heat deficit may increase the number of years of the larval cycle and impact abundance if mortality factors intrinsic to a longer development increase. A higher longhorn occurrence in sun-exposed trees (isolated or in forest edges) and prevalence of exit holes in warmer trunk exposures (S, SE, SW, W) have been widely documented for a long time [21,29,57,61,62,68,69]. Our data also indirectly support the “wood warming” effect, as longhorn abundance increased with increasing insolation, although proportionally more in Cc than in Cw. This interspecific difference was attributed, other things being equal, to the insolation-derived heat intake being greater in the trunk/base (where Cc larvae prevail) than in the fork/branches (where Cw larvae prevail) due to the shading of the treetop itself, particularly in cold seasons with the low sun ecliptic. Additionally, one could speculate that insolation indirectly favors Cc more than Cw because the former tends to exhibit more diurnal or heliophilous habits than the latter, even if both species are mostly crepuscular and nocturnal.

Cc abundance tended to increase with shrub cover, ground cover, oak renewal (resprouting), and bedrock outcrops, while the same variables generally had less (or even negative) effect on Cw abundance. It is worth noting that shrub cover data must be interpreted with caution because they derive from digital estimates on orthophotographs in which it is difficult to assess shrub height or even plant woody nature; also, shrub cover in the dehesas can change substantially from year to year due to soil tillage. The species-specific effects of these four variables were tentatively interpreted again according to the vertical larval stratification in the host tree. Our rational is that in oak open woodlands, the occurrence of shrubs, oak renewal, and bedrock outcrops is usually linked to shallow soils in which tillage is occasional or impracticable, so that shrubs are just controlled by means of cattle or burning, very exceptionally (illegally) with herbicides. In this context, oak roots experience little mechanical aggression, increase nutritional reserves, resprout and grow over decades or even centuries, and accumulate substantial amounts of underground wood, which potentially benefit more Cc than Cw. We have empirical field evidence supporting this fact from very young oak pyrenean forests formed by clear-cutting (DBHs < 10–15 cm at best) with no visible aboveground evidence of Cc damage (absence of exit holes and larval frass), in which captures with feeding traps exceeded 125 adults/trap/season (about 500 adults/ha), so that Cc adults could only originate from underground wood. On the other hand, shading on host trunks by ivy and nearby shrubs has been proposed to explain a low Cc abundance in some forest contexts (see considerations about insolation above), as thermophilic larvae do not usually find a suitable warm microclimate inside shaded host trunks [21,57,61,70,71]. However, our results show that shrub presence in the study area, shading host tree trunks (or potentially constraining longhorn flight), had no measurable effect on abundance. This could be due to a higher temperature regime in SW Spain and because cattle breeders strongly limit shrub height and extent to enhance dehesa pastures.

All three temperatures used as predictors (T_m, T_m7 and T_m1) showed a positive quadratic relationship with abundance, in accordance with the thermophilic nature of both longhorns [59,71,72]. Note that T_m is an estimator of the stand thermicity, T_m7 is a predictor of adult activity during the flight period in summer, while T_m1 is more related to larval development/quiescence during winter. The steeper slope of the curves in Cw than in Cc (T_m and T_m7) supports the fact that Cw is more thermophilic than Cc, a fact in line with their respective latitudinal distribution range and abundance across Europe and the western Palaearctic realm. Moreover, under the current global warming scenario, temperature–abundance curves suggest a short-medium term increase in longhorn populations and northwards spread of the distribution ranges, especially in the case of Cw.

Abundance of both species decreased with increasing altitude, an expected result given the robust negative correlation between temperature and altitude, so that populations were in practice very scarce above 1200 m. Increasing ground slope promoted a decrease in abundance, which could derive from lower soil thickness with steeper slopes, limiting nutrient and water availability, root development, tree vitality and size [62], and particularly underground wood amount. Stand aspect by itself did not affect longhorn abundance, but aspect x ground slope interaction had an effect in Cw and not in Cc. South-facing aspects in combination with steep slopes (conditions in which sun falls upon the stand more time and more perpendicularly) promote warmer microclimates, but such an effect just favored Cw. We speculate that in these south-facing, steep-slope orographic conditions, the trunk/base of the trees (where Cc larvae prevail) could receive less sun radiation than the fork/branches (where Cw larvae prevail) due to cross-shading between nearby trees.

Longhorn abundance decreased markedly with increasing annual precipitation, both species being virtually absent above 1200 mm annual rainfall, which was a striking and unexpected result. In Mediterranean climates, rains are scarce in summer months, so it is unlikely that they significantly limit adult vital activities during the flight period (June–July). By contrast, abundant rainfall in the unfavorable season (autumn-winter-early spring) could be a key mortality factor for overwintering larvae and adults inside host trees [73,74]. Rainwater may be channeled from cavities and hollows in the tree fork and upper trunk (derived from bad pruning, longhorn damage or other causes) flooding the larval gallery network and causing longhorn death, either by direct drowning or favoring entomopathogens when moistening wood tissues. Our own field experience with dissecting oak trunks supports this view [22] as it is not uncommon to find dead/mummified larvae or adults cloistered in the galleries of damaged trunks with wet/soggy wood, sometimes covered with fungal mycelium [75].

The GLMMs fitted to predict longhorn abundance included many ecological variables, both biotic and abiotic. The effect of some of them was rather predictable (e.g., temperature and altitude) but in other cases was opposite to that expected (e.g., trunk diameter, a proxy of tree age). Results also revealed the importance of other more cryptic variables for which the effect on abundance was previously unknown or unproven (e.g., rainfall and bedrocks outcrops). Our results overall agree with a previous regional study in southern Spain estimating longhorn impact based on adult exit holes rather than adult catches (so Cc and Cw effects were indistinguishable), which showed that trunk diameter, tree density, and temperature were important predictors to estimate longhorn distribution [33]. It is important to note that typical structural features characterizing the dehesa/montado ecosystem closely fits longhorn habitat requirements: low tree density, open forest cover, sun-exposed trees, regularly pruned oaks, and limited understory [7,8,21,33,69], so that Iberian oak open woodlands somewhat appear to be a habitat “man-made on purpose” for these borers. Such a forest habitat suitability along with hot temperatures could account for the extreme negative impact of these longhorns in SW Spain, while on the contrary, populations are much scarcer or declining in central Europe.

Our research contributes to deciphering the “complex black box” of ecological factors that shape the species-specific occupancy–abundance patterns of Cc and Cw in oak open woodlands [7,8], and to delving into the sympatric relationship between both longhorns, especially when they behave as pests. Our results also contribute to improving sustainable forest practices to mitigate the longhorn impact in oak open forests, as effective tools to manage wood borer insects are almost non-existent. Insecticide control particularly is not only ineffective against wood-cloistered longhorn larvae, but also inappropriate because of the extreme sensitivity of the singular dehesa ecosystem to chemical-mediated external disturbances.

Author Contributions

Conceptualization, L.M.T.-V. and F.J.M.-D.; methodology, L.M.T.-V. and F.J.M.-D.; validation, L.M.T.-V., F.J.M.-D. and T.C.; formal analysis, T.C. and L.M.T.-V.; investigation, L.M.T.-V. and F.J.M.-D.; writing—original draft preparation, L.M.T.-V.; writing—review and editing, L.M.T.-V., F.J.M.-D. and T.C.; visualization, T.C. and L.M.T.-V.; supervision, L.M.T.-V. All authors have read and agreed to the published version of the manuscript.

Institutional Review Board Statement

Not applicable.

Informed Consent Statement

Not applicable.

Data Availability Statement

A full raw dataset is given in Table S1.

Acknowledgments

The authors are grateful to all companions and colleagues who provided assistance in either field or laboratory; to the owners, tenants, and managers of the dehesas and oak forests for their good disposition to conduct this study; to the companies Aguas del Suroeste (Los Riscos) and Gespesa (Ecoparque de Talarrubias) for the supply of PET containers; to Emiliano Zamora (Estacion Enologica de Almendralejo) for the continued supply of red wine (just to be used in traps!); and to four anonymous reviewers for improving the manuscript. This research was supported by the Plant Health Service (SSV), Junta de Extremadura, Spain.

Conflicts of Interest

The authors declare no conflict of interest.

Footnotes

Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content.

Figure and Tables

Figure 1. Effects of each of the 18 ecological variables studied on the abundance (adult catches per trap per year) of Cerambyx cerdo and Cerambyx welensii. (A): Oak species, (B): Forest Mass, (C): Trunk Diameter, (D): Tree Density, (E): Basal Area, (F): Forest Cover, (G): Shrub Cover, (H): Ground Cover, (I): Oak Renewal, (J): Bedrock Outcrops, (K): Altitude, (L): Ground Slope, (M): Aspect, (N): Mean Annual Temperature, (O): Mean July Temperature, (P): Mean January Temperature, (Q): Annual Precipitation, (R): Insolation. Vertical lines represent the standard error of the mean (SEM) and gray shaded areas the confidence intervals (95%) for categorical and quantitative predictor variables, respectively. See Table 2 and text for a detailed statistical analysis through GLMMs.

Table 1

Categorical or quantitative predictor variables measured or estimated in each oak stand. Variables were scored for subsequent GLMMs as either biotic or abiotic according to their forest, dendrometric, physiographic, geological, or climatic features.

Ref.	Predictor Variable	Variable Nature 1	Variable Type 2	Classes/[Units]/Definition	Data Souce 3
1	Oak Species	B	C	n = 3 classes: holm oak (H), cork oak (C), pyrenean oak (P) [main species in the forest mass]	t.s.
2	Forest Mass	B	C	n = 6 classes: holm oak pure mass (H), cork oak pure mass (C), pyrenean oak pure mass (P), cork-holm oak mixed mass (CH), holm-pyrenean oak mixed mass (HP), pyrenean-cork oak mixed mass (PC)	t.s.
3	Trunk Diameter	B	SQ	n = 10 classes (10-cm DBH classes): from 10–20 cm (young oaks) to 100–110 cm (old/veteran oaks); DBH: Diameter at Breast Height [cm]	t.s.
4	Tree Density	B	Q	[trees ha−1]	t.s.
5	Basal Area	B	Q	[m2 ha−1], cross-sectional area of trees at breast height in square meters per ha	t.s.
6	Forest Cover	B	Q	[%], surface covered by tree canopies per surface unit (=crown cover)	[49]
7	Shrub Cover	B	Q	[%], surface covered by shrubs per surface unit	[49]
8	Ground Cover	B	C	n = 3 classes: pastureland (P), pastureland-undergrowth (PU), undergrowth (U)	t.s.
9	Oak Renewal	B	C	n = 2 classes: presence or absence of oak root resprouting	t.s.
10	Bedrock Outcrops	A	C	n = 2 classes: presence or absence	t.s.
11	Altitude	A	Q	[m], elevation in meters above sea level	[50]
12	Ground Slope	A	Q	[%]	[50]
13	Aspect	A	C	n = 8 classes: N, NE, E, SE, S, SW, W, NW	[50]
14	Mean Annual Temperature	A	Q	Tm [°C]	[43]
15	Mean July Temperature	A	Q	Tm7 [°C], mean temperature of the warmest month	[43]
16	Mean January Temperature	A	Q	Tm1 [°C], mean temperature of the coldest month	[43]
17	Annual Precipitation	A	Q	P [mm], mean annual precipitation	[43]
18	Insolation	A	Q	[h], mean number of sunshine hours per year	[43]

¹ Variable nature, B: Biotic, A: Abiotic. ² Variable type, C: Categorical, Q: Quantitative; SQ: Semi-quantitative. ³ Data source, t.s.: this study; MITECO [49]; CNIG-IGN [50]; AEMET [43]. A 30-year historical series was used to calculate mean data of temperatures, precipitation, and insolation.

Table 2

Analysis of deviance table (Type II Wald Chi-square tests) from GLMMs showing the effect on the abundance for each Cerambyx species (C. cerdo and C. welensii) of the 18 predictor variables and the predictor x species interactions.

Ref.	Predictor Variable 1	GLMM 2	Cerambyx Species Effect			Predictor Effect [Quadratic Effect] 1			Predictor x Species Effect
			Chi2	df	p	Chi2	df	p	Chi2	df	p
1	Oak Species	GP	59.71	1	<0.001	168.12	2	<0.001	97.00	2	<0.001
2	Forest Mass	P	39.56	1	<0.001	195.50	5	<0.001	96.46	5	<0.001
3	Trunk Diameter	P	43.54	1	<0.001	5.91	1	<0.05	114.58	1	<0.001
4	Tree Density	P	50.31	1	<0.001	7.51	1	<0.01	11.57	1	<0.001
	Tree Density ^2					[18.18]	1	[<0.001]
5	Basal Area	GP	62.75	1	<0.001	0.08	1	0.77 ns	64.76	1	<0.001
6	Forest Cover	P	50.46	1	<0.001	40.44	1	<0.001	4.79	1	<0.05
	Forest Cover ^2					[57.51]	1	[<0.001]
7	Shrub Cover	P	50.47	1	<0.001	2.48	1	0.12 ns	5.10	1	<0.05
8	Ground Cover	GP	61.98	1	<0.001	66.70	2	<0.001	55.94	2	<0.001
9	Oak Renewal	GP	64.09	1	<0.001	0.11	1	0.74 ns	109.54	1	<0.001
10	Bedrock Outcrops	GP	61.72	1	<0.001	2.06	1	0.15 ns	22.93	1	<0.001
11	Altitude	P	50.58	1	<0.001	251.51	1	<0.001	0.05	1	0.83 ns
12	Ground Slope	GP	60.22	1	<0.001	113.10	1	<0.001	1.30	1	0.25 ns
13	Aspect	GP	60.90	1	<0.001	9.10	7	0.25 ns	5.43	7	0.25 ns
14	Mean Annual Temp.	P	50.37	1	<0.001	11.94	1	<0.001	5.56	1	<0.05
	Mean Annual Temp. ^2					[7.20	1	<0.01]
15	Mean July Temp.	P	50.33	1	<0.001	30.64	1	<0.001	7.24	1	<0.01
	Mean July Temp. ^2					[27.22	1	<0.001]
16	Mean January Temp.	P	50.62	1	<0.001	42.89	1	<0.001	0.31	1	0.58 ns
	Mean January Temp. ^2					[22.58	1	<0.001]
17	Annual Precipitation	P	46.52	1	<0.001	341.39	1	<0.001	21.55	1	<0.001
18	Insolation	P	48.33	1	<0.001	79.49	1	<0.001	36.08	1	<0.001

¹ The exponent “^{^2}” denotes the quadratic effect of the predictor variable concerned [statistics between brackets]. ² Family (link) of the best model fitted, GP: generalized Poisson model with logarithmic link [genpois(log)]; P: Poisson model with logarithmic link [poisson(log)].

Table 3

Analysis of deviance table (Type II Wald Chi-square tests) from the best GLMMs computed to predict the abundance of each Cerambyx species (C. cerdo and C. welensii) in the wild. Three models were built per species with the predictor variables selected within either biotic, abiotic, or pooled variables.

Model	Cerambyx Species
	C. cerdo				C. welensii
	Effects	Chi2	df	p	Effects	Chi2	df	p
Biotic	Forest Mass	170.19	5	<0.001	Forest Mass	78.70	5	<0.001
	Trunk Diameter	12.61	1	<0.001	Trunk Diameter	46.64	1	<0.001
	Tree Density	11.94	1	<0.001	Tree Density	4.10	1	<0.05
	Forest Cover	2.24	1	0.14 ns	Forest Cover	58.39	1	<0.001
	Ground Cover	10.75	2	<0.01	Forest Cover ^2	38.06	1	<0.001
	Oak Renewal	21.50	1	<0.001	Ground Cover	13.32	2	<0.01
	Forest Mass x Trunk Diameter	18.70	5	<0.01	Forest Mass x Forest Cover	16.76	5	<0.01
	Forest Mass x Forest Cover	33.98	5	<0.001	Forest Mass x Ground Cover	24.47	10	<0.01
	Forest Mass x Ground Cover	34.27	10	<0.001	Trunk Diameter x Ground Cover	6.99	2	<0.05
	Forest Mass x Oak Renewal	18.98	5	<0.01	Forest Cover x Ground Cover	5.99	2	<0.05
	Forest Cover x Ground Cover	8.54	2	<0.05
Abiotic	Bedrock Outcrops	42.71	1	<0.001	Bedrock Outcrops	3.34	1	~0.05
	Altitude	11.95	1	<0.001	Altitude	32.20	1	<0.001
	Aspect	9.44	7	0.23 ns	Ground Slope	6.25	1	<0.05
	Annual Precipitation	47.65	1	<0.001	Aspect	7.43	7	0.39 ns
	Insolation	14.30	1	<0.001	Annual Precipitation	34.44	1	<0.001
	Bedrock Outcrops x Altitude	5.70	1	<0.05	Bedrock Outcrops x Altitude	5.70	1	<0.05
	Aspect x Annual Precipitation	21.01	7	<0.01	Bedrock Out. x Annual Prec.	9.99	1	<0.01
					Altitude x Aspect	19.24	7	<0.01
Pooled	Forest Mass	33.38	5	<0.001	Forest Mass	52.33	5	<0.001
	Trunk Diameter	5.25	1	<0.05	Trunk Diameter	47.95	1	<0.001
	Tree Density	13.08	1	<0.001	Tree Density	3.42	1	~0.05
	Forest Cover	0.11	1	0.74 ns	Forest Cover	59.96	1	<0.001
	Ground Cover	13.50	2	<0.01	Forest Cover^2	33.06	1	<0.001
	Oak Renewal	16.22	1	<0.001	Ground Cover	9.36	2	<0.01
	Bedrock Outcrops	13.58	1	<0.001	Bedrock Outcrops	2.90	1	~0.05
	Aspect	5.74	7	0.57 ns	Altitude	26.72	1	<0.001
	Annual Precipitation	113.22	1	<0.001	Annual Precipitation	52.14	1	<0.001
	Forest Mass x Trunk Diameter	16.67	5	<0.01	Forest Mass x Forest Cover	10.61	5	~0.05
	Forest Mass x Forest Cover	21.01	5	<0.001	Trunk Diameter x Ground Cover	6.04	2	<0.05
	Forest Cover x Ground Cover	13.95	2	<0.001	Forest Cover x Ground Cover	11.11	2	<0.01
	Aspect x Annual Precipitation	19.40	7	<0.01	Bedrock Outcrops x Altitude	4.78	1	<0.05

Poisson model with logarithmic link in all cases [poisson(log)]. The exponent “^{^2}” denotes the quadratic effect of the predictor variable concerned.

Supplementary Materials

The following supporting information can be downloaded at https://www.researchgate.net/publication/369881228, Table S1: List of the 1657 feeding traps (seven traps were lost) used during 2017–2020 in Extremadura (SW Spain) to capture Cerambyx adults in oak forests, with indication of the trapping year, province, municipality, dehesa/site, stand (n = 809), number of traps per stand, WGS84 coordinates, the 18 ecological variables studied, and the number of captures per trap per species (sexes pooled). Cw: Cerambyx welensii, Cc: Cerambyx cerdo.

References

1. Médail, F.; Quézel, P. Biodiversity hotspots in the Mediterranean Basin: Setting global conservation priorities. Conserv. Biol.; 1999; 13, pp. 1510-1513. [DOI: https://dx.doi.org/10.1046/j.1523-1739.1999.98467.x]

2. Myers, N.; Mittermeier, R.A.; Mittermeier, C.G.; Da Fonseca, G.A.B.; Kent, J. Biodiversity hotspots for conservation priorities. Nature; 2000; 403, pp. 853-858. [DOI: https://dx.doi.org/10.1038/35002501] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/10706275]

3. Blondel, J. The ‘design’ of Mediterranean landscapes: A millennial story of humans and ecological systems during the historic period. Hum. Ecol.; 2006; 34, pp. 713-729. [DOI: https://dx.doi.org/10.1007/s10745-006-9030-4]

4. Moreno, G.; Pulido, F.J. The functioning, management and persistence of dehesas. Agroforestry in Europe; Rigueiro-Rodríguez, A.; McAdam, J.; Mosquera-Losada, M.R. Springer: Dordrecht, The Netherlands, 2009; pp. 127-160. [DOI: https://dx.doi.org/10.1007/978-1-4020-8272-6_7]

5. Bugalho, M.N.; Caldeira, M.C.; Pereira, J.S.; Aronson, J.; Pausas, J.G. Mediterranean cork oak savannas require human use to sustain biodiversity and ecosystem services. Front. Ecol. Environ.; 2011; 9, pp. 278-286. [DOI: https://dx.doi.org/10.1890/100084] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/36686577]

6. Ramírez-Hernández, A.; Micó, E.; Marcos-García, M.A.; Brustel, H.; Galante, E. The “dehesa”, a key ecosystem in maintaining the diversity of Mediterranean saproxylic insects (Coleoptera and Diptera: Syrphidae). Biodivers. Conserv.; 2014; 23, pp. 2069-2087. [DOI: https://dx.doi.org/10.1007/s10531-014-0705-7]

7. Torres-Vila, L.M.; Mendiola-Díaz, F.J.; López, R.; Sánchez, Á.; Ponce, F.; Fernández, F.; Zugasti, C.; de-Juan, J.M.; Echevarría-León, E.; Cáceres, Y. et al. Distribución actualizada del género Cerambyx Linnaeus, 1758 (Coleoptera: Cerambycidae) en Extremadura: Desde los registros históricos al muestreo a escala regional. Graellsia; 2022; 78, e169. [DOI: https://dx.doi.org/10.3989/graellsia.2022.v78.340]

8. Torres-Vila, L.M.; Mendiola-Díaz, F.J.; Moral-García, F.J.; Canelo, T. Large-scale geostatistical mapping and occupancy-abundance patterns of Cerambyx species threatening SW Spain oak forests. Eur. J. For. Res.; 2022; 141, pp. 1045-1057. [DOI: https://dx.doi.org/10.1007/s10342-022-01487-z]

9. CEC [Council of the European Communities]. Council Directive 92/43/EEC of 21 May 1992 on the conservation of natural habitats and of wild fauna and flora [Habitats Directive]. Off. J. Eur. Communities; 1992; 35, pp. 7-50.

10. Thomas, F.M.; Blank, R.; Hartmann, G. Abiotic and biotic factors and their Interactions as causes of oak decline in Central Europe. For. Pathol.; 2002; 32, pp. 277-307. [DOI: https://dx.doi.org/10.1046/j.1439-0329.2002.00291.x]

11. Haavik, L.J.; Billings, S.A.; Guldin, J.M.; Stephen, F.M. Emergent insects, pathogens and drought shape changing patterns in oak decline in North America and Europe. For. Ecol. Manag.; 2015; 354, pp. 190-205. [DOI: https://dx.doi.org/10.1016/j.foreco.2015.06.019]

12. Manion, P.D. Evolution of concepts in forest pathology. Phytopathology; 2003; 93, pp. 1052-1055. [DOI: https://dx.doi.org/10.1094/PHYTO.2003.93.8.1052]

13. Torres-Vila, L.M.; Echave-Sanabria, A.C.; Mendiola-Díaz, F.J.; Moral-García, F.J. Mapping oak shoot browning in SW Spain using online imagery as virtual prospecting tool. Ann. For. Sci; 2019; 76, 32. [DOI: https://dx.doi.org/10.1007/s13595-019-0818-y]

14. Evans, H.F.; Moraal, L.G.; Pajares, J.A. Biology, ecology and economic importance of Buprestidae and Cerambycidae. Bark and Wood Boring Insects in Living Trees in Europe, a Synthesis; Lieutier, F.; Day, K.R.; Battisti, A.; Grégoire, J.-C.; Evans, H.F. Springer: Dordrecht, The Netherlands, 2004; pp. 447-474. [DOI: https://dx.doi.org/10.1007/978-1-4020-2241-8_20]

15. Sallé, A.; Nageleisen, L.M.; Lieutier, F. Bark and wood boring insects involved in oak declines in Europe: Current knowledge and future prospects in a context of climate change. For. Ecol. Manag.; 2014; 328, pp. 79-93. [DOI: https://dx.doi.org/10.1016/j.foreco.2014.05.027]

16. Picard, F. Coléoptères: Cerambycidae. Faune de France 20; P. Lechevalier: Paris, France, 1929; 166p.

17. Villiers, A. Faune des Coléoptères de France 1. Cerambycidae. Encyclopédie Entomologique 42; Lechevalier: Paris, France, 1978; 612p.

18. Vives, E. Coleoptera Cerambycidae. Fauna Ibérica 12; Museo Nacional de Ciencias Naturales (CSIC): Madrid, Spain, 2000; 715p.

19. Miroshnikov, A.I. Review of the longicorn beetles genus Cerambyx Linnaeus, 1758 (Coleoptera, Cerambycidae) of the Caucasus [in Russian]. Moscow State For. Univ. Bull. (Lesnoy Vestnik); 2009; 5, pp. 43-55.

20. Escherich, K. Die Forstinsekten Mitteleuropas 2; P. Parey: Berlin, Germany, 1923; 663p.

21. Buse, J.; Schröder, B.; Assmann, T. Modelling habitat and spatial distribution of an endangered longhorn beetle–a case study for saproxylic insect conservation. Biol. Conserv.; 2007; 137, pp. 281-372. [DOI: https://dx.doi.org/10.1016/j.biocon.2007.02.025]

22. Torres-Vila, L.M.; Zugasti-Martínez, C.; Mendiola-Díaz, F.J.; De-Juan-Murillo, J.M.; Sánchez-González, Á.; Conejo-Rodríguez, Y.; Ponce-Escudero, F.; Fernández-Moreno, F. Larval assemblages of large saproxylic cerambycids in Iberian oak forests: Wood quality and host preference shape resource partitioning. Pop. Ecol.; 2017; 59, pp. 315-328. [DOI: https://dx.doi.org/10.1007/s10144-017-0592-3]

23. Speight, M.C.D. Saproxylic Invertebrates and Their Conservation. Nature and Environment Series 42; Council of Europe: Strasbourg, France, 1989; 79p.

24. Grove, S.J. Saproxylic insect ecology and the sustainable management of forests. Annu. Rev. Ecol. Syst.; 2002; 33, pp. 1-23. [DOI: https://dx.doi.org/10.1146/annurev.ecolsys.33.010802.150507]

25. Jones, C.G.; Lawton, J.H.; Shachak, M. Positive and negative effects of organisms as physical ecosystem engineers. Ecology; 1997; 78, pp. 1946-1957. [DOI: https://dx.doi.org/10.1890/0012-9658(1997)078[1946:PANEOO]2.0.CO;2]

26. Buse, J.; Ranius, T.; Assmann, T. An endangered longhorn beetle associated with old oaks and its possible role as an ecosystem engineer. Conserv. Biol.; 2008; 22, pp. 329-337. [DOI: https://dx.doi.org/10.1111/j.1523-1739.2007.00880.x] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/18261146]

27. Saint-Germain, M.; Drapeau, P.; Buddle, C.M. Host-use patterns of saproxylic phloeophagous and xylophagous Coleoptera adults and larvae along the decay gradient in standing dead black spruce and aspen. Ecography; 2007; 30, pp. 737-748. [DOI: https://dx.doi.org/10.1111/j.2007.0906-7590.05080.x]

28. Davies, Z.G.; Tyler, C.; Stewart, G.B.; Pullin, A.S. Are current management recommendations for saproxylic invertebrates effective? A systematic review. Biodivers. Conserv.; 2008; 17, pp. 209-234. [DOI: https://dx.doi.org/10.1007/s10531-007-9242-y]

29. Regnery, B.; Paillet, Y.; Couvet, D.; Kerbiriou, C. Which factors influence the occurrence and density of tree microhabitats in Mediterranean oak forests?. For. Ecol. Manag.; 2013; 295, pp. 118-125. [DOI: https://dx.doi.org/10.1016/j.foreco.2013.01.009]

30. Micó, E.; García-López, A.; Sánchez, A.; Juárez, M.; Galante, E. What can physical, biotic and chemical features of a tree hollow tell us about their associated diversity?. J. Insect Conserv.; 2015; 19, pp. 141-153. [DOI: https://dx.doi.org/10.1007/s10841-015-9754-9]

31. Martín, J.; Cabezas, J.; Buyolo, T.; Patón, D. The relationship between Cerambyx spp. damage and subsequent Biscogniauxia mediterranum infection on Quercus suber forests. For. Ecol. Manag.; 2005; 216, pp. 166-174. [DOI: https://dx.doi.org/10.1016/j.foreco.2005.05.027]

32. Domínguez, L.; López-Pantoja, G.; Cremades, D.; Paramio, A.; Hidalgo, P.J.; Sánchez-Osorio, I. Incidence of large wood borers in the conservation of dehesa islands forests in southwestern Spain. Forests; 2022; 13, 413. [DOI: https://dx.doi.org/10.3390/f13030413]

33. Duque-Lazo, J.; Navarro-Cerrillo, R.M. What to save, the host or the pest? The spatial distribution of xylophage insects within the Mediterranean oak woodlands of Southwestern Spain. For. Ecol. Manag.; 2017; 392, pp. 90-104. [DOI: https://dx.doi.org/10.1016/j.foreco.2017.02.047]

34. Torres-Vila, L.M. Reproductive biology of the great capricorn beetle, Cerambyx cerdo (Coleoptera: Cerambycidae): A protected but occasionally harmful species. Bull. Entomol. Res.; 2017; 107, pp. 799-811. [DOI: https://dx.doi.org/10.1017/S0007485317000323] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/28397625]

35. Torres-Vila, L.M.; Bonal, R. DNA barcoding of large oak-living cerambycids: Diagnostic tool, phylogenetic insights and natural hybridization between Cerambyx cerdo and Cerambyx welensii (Coleoptera: Cerambycidae). Bull. Entomol. Res.; 2019; 109, pp. 583-594. [DOI: https://dx.doi.org/10.1017/S0007485318000925]

36. Schoener, T.W. Resource partitioning in ecological communities. Science; 1974; 185, pp. 27-39. [DOI: https://dx.doi.org/10.1126/science.185.4145.27] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/17779277]

37. Chesson, P. Mechanisms of maintenance of species diversity. Annu. Rev. Ecol. Syst.; 2000; 31, pp. 343-366. [DOI: https://dx.doi.org/10.1146/annurev.ecolsys.31.1.343]

38. Agrawal, A.A.; Ackerly, D.D.; Adler, F.R.; Arnold, A.E.; Cáceres, C.; Doak, D.F.; Post, E.; Hudson, P.J.; Maron, J.; Mooney, K.A. et al. Filling key gaps in population and community ecology. Front. Ecol. Environ.; 2007; 5, pp. 145-152. [DOI: https://dx.doi.org/10.1890/1540-9295(2007)5[145:FKGIPA]2.0.CO;2]

39. Mayfield, M.M.; Levine, J.M. Opposing effects of competitive exclusion on the phylogenetic structure of communities. Ecol. Lett.; 2010; 13, pp. 1085-1093. [DOI: https://dx.doi.org/10.1111/j.1461-0248.2010.01509.x] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/20576030]

40. Satoh, T.; Yoshida, T.; Koyama, S.; Yamagami, A.; Takata, M.; Doi, H.; Kurachi, T.; Hayashi, S.; Hirobe, T.; Hata, Y. Resource partitioning based on body size contributes to the species diversity of wood-boring beetles and arboreal nesting ants. Insect Conserv. Divers.; 2016; 9, pp. 4-12. [DOI: https://dx.doi.org/10.1111/icad.12136]

41. Torres-Vila, L.M.; López-Calvo, R.; Sánchez-González, Á.; Mendiola-Díaz, F.J. Ecology of Oobius rudnevi, egg parasitoid of Cerambyx cerdo and Cerambyx welensii in oak forests. Entomol. Exp. Appl.; 2021; 169, pp. 646-656. [DOI: https://dx.doi.org/10.1111/eea.13059]

42. Allen, C.D.; Macalady, A.K.; Chenchouni, H.; Bachelet, D.; McDowell, N.; Vennetier, M.; Kitzberger, T.; Rigling, A.; Breshears, D.D.; Hogg, E.H. et al. A global overview of drought and heat-induced tree mortality reveals emerging climate change risks for forests. For. Ecol. Manag.; 2010; 259, pp. 660-684. [DOI: https://dx.doi.org/10.1016/j.foreco.2009.09.001]

43. AEMET. Guía Resumida del Clima en España (1981–2010), AEMET [Agencia Estatal de Meteorología]. Available online: https://www.aemet.es/es/conocermas/recursos_en_linea/publicaciones_y_estudios/publicaciones/detalles/guia_resumida_2010 (accessed on 15 January 2023).

44. Torres-Vila, L.M.; Sánchez-González, Á.; Ponce-Escudero, F.; Martín-Vertedor, D.; Ferrero-García, J.J. Assessing mass trapping efficiency and population density of Cerambyx welensii Küster by mark-recapture in dehesa open woodlands. Eur. J. For. Res.; 2012; 131, pp. 1103-1116. [DOI: https://dx.doi.org/10.1007/s10342-011-0579-0]

45. Torres-Vila, L.M.; Sánchez-González, Á.; Merino-Martínez, J.; Ponce-Escudero, F.; Conejo-Rodríguez, Y.; Martín-Vertedor, D.; Ferrero-García, J.J. Mark-recapture of Cerambyx welensii in dehesa woodlands: Dispersal behaviour, population density, and mass trapping efficiency with low trap densities. Entomol. Exp. Appl.; 2013; 149, pp. 273-281. [DOI: https://dx.doi.org/10.1111/eea.12133]

46. Torres-Vila, L.M.; Mendiola-Díaz, F.J.; Conejo-Rodríguez, Y.; Sánchez-González, Á. Reproductive traits and number of matings in males and females of Cerambyx welensii (Coleoptera: Cerambycidae) an emergent pest of oaks. Bull. Entomol. Res.; 2016; 106, pp. 292-303. [DOI: https://dx.doi.org/10.1017/S0007485315000747] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/26490198]

47. Torres-Vila, L.M.; Mendiola-Díaz, F.J.; Sánchez-González, Á. Dispersal differences of a pest and a protected Cerambyx species (Coleoptera: Cerambycidae) in oak open woodlands: A mark–recapture comparative study. Ecol. Entomol.; 2017; 42, pp. 18-32. [DOI: https://dx.doi.org/10.1111/een.12355]

48. Drag, L.; Cižek, L. Radio-tracking suggests high dispersal ability of the great capricorn beetle (Cerambyx cerdo). J. Insect Behav.; 2018; 31, pp. 138-143. [DOI: https://dx.doi.org/10.1007/s10905-018-9669-x]

49. MITECO. Mapa Forestal de España (MFE) de máxima actualidad [Extremadura], MITECO [Ministerio para la Transición Ecológica y el Reto Demográfico]. Available online: https://www.miteco.gob.es/es/cartografia-y-sig/ide/descargas/biodiversidad/mfe.aspx (accessed on 15 January 2023).

50. CNIG-IGN. MDT05: Modelo Digital del Terreno con Paso de Malla 5m (SRG: ETRS89), CNIG-IGN [Centro Nacional de Información Geográfica, Instituto Geográfico Nacional]. Available online: http://centrodedescargas.cnig.es/CentroDescargas/catalogo.do?Serie=LIDAR (accessed on 15 January 2023).

51. Brooks, M.E.; Kristensen, K.; Darrigo, M.R.; Rubim, P.; Uriarte, M.; Bruna, E.; Bolker, B.M. Statistical modeling of patterns in annual reproductive rates. Ecology; 2019; 100, e02706. [DOI: https://dx.doi.org/10.1002/ecy.2706] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/30916779]

52. Zuur, A.; Ieno, E.N.; Walker, N.; Saveliev, A.A.; Smith, G.M. Mixed Effects Models and Extensions in Ecology with R; Springer: New York, NY, USA, 2009; 574p.

53. R Core Team. R: A Language and Environment for Statistical Computing; R Foundation for Statistical Computing: Vienna, Austria, 2016; 2630p.

54. Southwood, T.R.E. Ecological Methods with Particular Reference to the Study of Insect Populations; Chapman and Hall: London, UK, 1978; 524p.

55. Seber, G. The Estimation of Animal Abundance; Macmillan Publishing Co.: New York, NY, USA, 1982; 654p.

56. Guisan, A.; Zimmermann, N.E. Predictive habitat distribution models in ecology. Ecol. Modell.; 2000; 135, pp. 147-186. [DOI: https://dx.doi.org/10.1016/S0304-3800(00)00354-9]

57. Rudnev, D.F. Der grosse Eichenbock, Cerambyx cerdo L., seine Lebensweise, wirtschaftliche Bedeutung und Bekämpfung. Z. Angew. Entomol.; 1936; 22, pp. 61-96. [DOI: https://dx.doi.org/10.1111/j.1439-0418.1936.tb00413.x]

58. Heyrovský, L. Fauna ČSR, Svazek 5. Tesaříkovití- Cerambycidae (Řád brouci.—Coleoptera); Nakladatelství Československé Akademie: Prague, Czech Republic, 1955; 346p.

59. Săvulescu, N. Cîteva observații ecologice asupra speciilor din Republica Socialistă România ale genului Cerambyx L. (Col. Cerambycidae). Comunicări de Zoologie, Prima Consfătuire Naţională de Entomologie, II-a; Societatea de Ştiinţe Biologice din RSR: Bucureşti, Romania, 1969; pp. 281-290.

60. Strojny, W. Badania nad biologia kozioroga debosza, Cerambyx cerdo L. (Coleoptera, Cerambycidae) zasiedlajacego deby szypulkowe, Quercus robur L. na Swojcu iw Wilczycach pod Wroclawiem w latach 1973–1976. Pol. Pis. Entomol.; 1977; 47, pp. 727-746.

61. Albert, J.; Platek, M.; Cižek, L. Vertical stratification and microhabitat selection by the great capricorn beetle (Cerambyx cerdo) (Coleoptera: Cerambycidae) in open-grown, veteran oaks. Eur. J. Entomol.; 2012; 109, pp. 553-559. [DOI: https://dx.doi.org/10.14411/eje.2012.069]

62. Platek, M.; Sebek, P.; Hauck, D.; Cižek, L. When is a tree suitable for a veteran tree specialist? Variability in the habitat requirements of the great capricorn beetle (Cerambyx cerdo) (Coleoptera: Cerambycidae). Eur. J. Entomol.; 2019; 116, pp. 64-74. [DOI: https://dx.doi.org/10.14411/eje.2019.007]

63. Valbuena-Carabaña, M.; Gil, L. Centenary coppicing maintains high levels of genetic diversity in a root resprouting oak (Quercus pyrenaica Willd.). Tree Genet. Genomes; 2017; 13, 28. [DOI: https://dx.doi.org/10.1007/s11295-017-1105-4]

64. May, M.R.; Provance, M.C.; Sanders, A.C.; Ellstrand, N.C.; Ross-Ibarra, J. A Pleistocene clone of Palmer’s oak persisting in southern California. PLoS ONE; 2009; 4, e8346. [DOI: https://dx.doi.org/10.1371/journal.pone.0008346] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/20041136]

65. Heggenstaller, D.J.; Zenner, E.K.; Brose, P.H.; Peck, J.E. How much older are Appalachian oaks below-ground than above-ground?. North. J. Appl. For.; 2012; 29, pp. 155-157. [DOI: https://dx.doi.org/10.5849/njaf.12-008]

66. Ratzeburg, J.T.C. Die Waldverderber und ihre Feinde: Ein Handbuch für Forstmänner, Landwirthe, Gärtner und alle mit Waldbäumen Beschäftigte; Nicolaische Verlags-Buchhandlung: Berlin, Germany, 1876; 523p.

67. López-Pantoja, G.; Domínguez Nevado, L.; Sánchez-Osorio, I. Mark-recapture estimates of the survival and recapture rates of Cerambyx welensii Küster (Coleoptera cerambycidae) in a cork oak dehesa in Huelva (Spain). Cent. Eur. J. Biol.; 2008; 3, pp. 431-441. [DOI: https://dx.doi.org/10.2478/s11535-008-0044-3]

68. Reitter, E. Fauna Germanica. Die Käfer des Deutschen Reiches 4; K.G. Lutz: Stuttgard, Germany, 1912; 236p.

69. Vodka, S.; Konvicka, M.; Cižek, L. Habitat preferences of oak-feeding xylophagous beetles in a temperate woodland: Implications for forest history and management. J. Insect. Conserv.; 2009; 13, pp. 553-562. [DOI: https://dx.doi.org/10.1007/s10841-008-9202-1]

70. Starzyk, J.R. Kozioróg dębosz Cerambyx cerdo L. w Puszczy Niepołomickiej. Chrońmy Przyr. Ojczystą; 1973; 29, pp. 22-30.

71. Strojny, W. Sędziwe dęby czy kozioróg dębosz?. Wszechświat; 1981; 1981, pp. 32-36.

72. Mulsant, É. Histoire Naturelle des Coléoptères de France: Longicornes; Magnin, Blanchard et Cie.: Paris, France, 1862–1863; 590p.

73. Gardiner, L.M. Survival of wood borers (Coleoptera: Cerambycidae, Buprestidae) in water-stored logs. For. Chron.; 1962; 38, pp. 459-462. [DOI: https://dx.doi.org/10.5558/tfc38459-4]

74. Hanks, L.M.; Millar, J.G.; Paine, T.D. Evaluation of cold temperatures and density as mortality factors of the Eucalyptus Longhorned Borer (Coleoptera: Cerambycidae) in California. Environ. Entomol.; 1991; 20, pp. 1653-1658. [DOI: https://dx.doi.org/10.1093/ee/20.6.1653]

75. Morales-Rodríguez, C.; Sánchez-González, Á.; Conejo-Rodríguez, Y.; Torres-Vila, L.M. First record of Beauveria bassiana (Balsamo) Vuillemin (Ascomycota: Clavicipitaceae) infecting Cerambyx welensii (Coleoptera: Cerambycidae) and pathogenicity tests using a new bioassay method. Biocontrol Sci. Technol.; 2015; 25, pp. 1213-1219. [DOI: https://dx.doi.org/10.1080/09583157.2015.1036004]