Cent. Eur. J. Biol. • 3(4) • 2008 • 431–441 DOI: 10.2478/s11535-008-0044-3

Central European Journal of Biology

Mark-recapture estimates of the survival and recapture rates of welensii Küster (Coleoptera cerambycidae) in a cork oak dehesa in Huelva ()

Research Article Gloria López-Pantoja*, Luis Dominguez Nevado, Sánchez-Osorio

Agroforestry Sciences Department, High Politecnic School, Campus of “La Rábida” Huelva University, 21819 Palos de la Frontera (Huelva), Spain

Received 21 January 2008; Accepted 10 July 2008

Abstract: The oak decline is probably the most severe plant health problem faced in the Mediterranean region which is one of the habitats of community interest under the EU’s environmental legal regime. More information on the role of Cerambycids species in this decay is still needed. This paper reports the apparent survival rate (Phi) and recapture rate (P) for a population of Küster (Coleoptera cerambycidae) in a highly degraded cork oak grove near the Doñana National Park (Huelva, Spain) as calculated using the mark-capture-recapture method. High and constants in the time values of apparent survival rates for males and females are de- tected. The male overall recapture rate (P) exceeded that of the female group with relatively low, but significant, values. The presence of transient individuals suggests a nucleus of population with many immigrants and emigrants in the study plot. The results are used to discuss various aspects of the biology, and the potential effect of the gradual deterioration of the studied ecosystem on the insect population it supports. Keywords: Cerambycids • Cerambyx welensii • Mark-recaptures • Oak decline • CJS model • Survival rate

© Versita Warsaw and Springer-Verlag Berlin Heidelberg.

1. Introduction biotic factors (pathogenic fungi and bacteria, ). There are a lot of insects implied in the decay process (triggering and/or worsening it), of which foliage-feeding Cork oak ( Linnaeus, 1753) forests insects Lymantria dispar Linnaeus, (1758) and Tortrix constitute one of the habitats of community interest viridana Linnaeus, (1758) are the most important [4]. under the EU’s environmental legal regime [1], which Wood-boring insects are mainly Platipus cilindrus compels the member states to take effective steps Fabricius, (1792) [5], Cerambyx welensii Küster (1846), towards their protection. These forest formations have Linnaeus (1758) and Prinobius myardi for centuries been exploited for mainly cork and ─ ─ Mulsant (1842) [6-13]. There is extensive knowledge on structured as open woodland forests locally called the biology and the population dynamics of L. dispar and “dehesas” in order to facilitate their use for agronomic T. viridana. Several control methods have already been and/or gaming purposes. The present decline in applied in order to reduce the damages caused by them. Quercus woodlands (oak decline) is probably the most The presence of P. cyindrus on Quercus forests is very severe plant health problem faced in the Mediterranean rare and local. In spite of its high pathogen potential, its region as it threatens to extinguish such representative, effects are insignificant. potential highly productive ecosystems [2,3]. This As far as the big cerambycid species concerned, degradation process is the result of a wide variety of their effects are known in weakening the structural abiotic (rainfall distribution, edaphic characteristics) and

* E-mail: [email protected] 431 Mark-recapture estimates of the survival and recapture rates of Cerambyx welensii Küster (Coleoptera cerambycidae) in a cork oak dehesa in Huelva (Spain)

resistance and plant health of trees in a lot of Linnaeus, (1753) and Q. suber masses. Their morphological similarity, semi-voltine character and non-specific damage have caused the indiscriminate management of their populations and estimation of their damage [14,15]. However, one of the species (C. cerdo) is protected by European laws (Annex IV of the Habitats Directive of 1992, which compels member states to control noxious agents for this type of habitat and actively conserve the populations of protected species). The original assumption that the action of these cerambycids is usually restricted to old trees in a decaying or poor physiological condition [12,16] has subsequently been challenged by the observation of young (70 years old) cork oak trees affected by these insects [14,17], which suggests that their ability to colonize new hosts may be dictated by factors other than the vigour status of the trees [17]. Ecologically, cerambycid species are functionally important to forest ecosystems because they influence the structure and composition of forests, and expand the available niches for a number of reptile, bird, mammal and invertebrate species acting as ecosystem engineers Figure 1. Couple of Cerambyx welensii Küster. [18,19]. length exceeds body length in only males. The yearly A number of ecological, economic and conservation flight period for adult individuals usually spans from reasons exist for predicting population dynamics in these mid May to late July (Figure 2). This species exhibits insects. However, most scientific studies on oak decline nocturnal and crepuscular habits. Although no accurate (“la seca”) have so far focused on specific mycological knowledge is available in this respect, direct field and phytoclimatic characteristics of the stands and little observations as well as bibliographical references on is known as a result about the insect fauna associated similar species (f.e. C. cerdo) suggest that they feed on to open woodland forest ecosystems (particularly in some nutritional substance in the trunk, leaf exudates or relation to xylophagous cerambycids). canker and also in mature fruits [20,21]. Tree damage is This paper reports part of the results of a serial caused during the larval stage and is associated with the study of the phenomena involved in oak decline that feeding, refuge and pupation habits of the species, which is performed within a global project funded by the constructs larval galleries in tree trunks and branches Environmental Council of the Andalusian regional (Figure 3). Therefore, the species can be included in the government. The aim of this work was to increase “strictly xylophagous” group defined by Hanks (1999) available knowledge about the population dynamics [22]. As larvae grow, their hosting trees become weaker of C. welensii, which is one of the species involved in and prone to breaking and decaying. The final size of the above-described processes. Interpreting the results the larvae is similar to that of C. cerdo (13 - 16 mm wide in terms of the phytosanitary status of the hosts can and 60 - 70 mm long) [12,23]. Their semi-voltine cycle help establish some reference for the development of requires 3 years to complete a generation. The larvae effective protection and management strategies. turn into pupa after summer (august-october) and the adults cross a phase of ripeness remaining inside the tree during the winter and spring (from august-october 2. Experimental Procedures to May). The main hosts of this species are trees of the genus Quercus (particularly Q. suber and Q. ilex) [24,25]. 2.1. Species Males compete strongly for females by engaging in C. welensii is a large insect. The studied males were violent fights that frequently result in amputation of legs 4.66 ± 0.45 cm (n= 72) and females 4.99 ± 0.13 cm and/or antennae. Adults are active fliers when fleeing or (n = 64) (personal observations) (Figure 1). They also moving to a new host (personal field observations). exhibit slight sexual dimorphism. Especially the antenna

432 G. López-Pantoja, L.D. Nevado, I. Sánchez-Osorio

90 80 70 60 50 40

Nº Insects /day 30 20 10 0 2-jul 8-jul 2-jun 8-jun 14-jul 20-jul 26-jul 1-ago 7-ago 14-jun 20-jun 26-jun 13-ago 15-may 21-may 27-may

2002 2003 2004

Figure 2. Flight period of C.welensii, during the study period. 2.2. Study area The present study was conducted in a Q. suber plot located in Almonte (Huelva, SW Spain). Although the area is in the vicinity of protected high biodiversity Figure 4. Location of the farm and sampled plot. spaces such as the Doñana Natural Park and Doñana National Park, anthropization of the area has resulted dispersal─ a sunflower (Helianthus annus Linnaeus, in increasing fragmentation of the fairly extensive forest 1754) plantation in the east and a fig tree (Ficus carica formations of cork oaks, interspersed with holm oaks, Linnaeus, 1753) plantation in the West (Figure 4). over the past fifty years. Also, cropped and urbanized The average monthly temperature for the sampling land has grown in parallel, which has considerably period in a typical year ranged from 17.9 to 24.8ºC reduced the original biodiversity. without any precipitation. The plot was 2.8 ha (120 m The studied plot was located on the “Dehesa boyal” long and 250 m wide) in size and contained 211 trees farm, which is largely devoted to cork production but (the density was thus 75 trees/ha). The diameters of the is also used for grazing by a small livestock of ovine trees were distributed as follows: 4.39% were less than cattle. On a yearly basis, the soil is sown with grasses 20 cm wide, 83.9% 20-60 cm wide and 11.7% wider and tilled by complete surface harrowing, therefore the than 60 cm. bush layer is virtually absent. Based on the damage The trees were markedly damaged. More than two- repeatedly caused by the machinery used, this tillage thirds had exit holes typical of drilling insects. Most have method seemingly makes the trees vulnerable to attack been stripped several times to obtain cork, but roughly by xylophagous cerambycids. one third have never been stripped. Trees were last The farm is surrounded by cropped land and the stripped in both the studied plot and those around it in nearest forest formations are more than 600 m away. the summer of 2003 (after field work on the studied plot The studied plot was a rectangular portion located at was completed). The use of a deficient cork removal the beginning of a narrow strip of land that extends to technique has resulted in wounds that have often the south of the farm and is bound by a regional road evolved into cracks in many trees throughout the farm. Other deficient manage­ment practices have potentially in the north ─a barrier that potentially hinders insect

Figure 3. Larvae of Cerambyx welensii and damage on trunk and branch.

433 Mark-recapture estimates of the survival and recapture rates of Cerambyx welensii Küster (Coleoptera cerambycidae) in a cork oak dehesa in Huelva (Spain)

facilitated colonization by cerambycids of unhealed Release as implemented in Mark 4.0 [28], and the wounds in pruned branches, which are present in high software U-Care [29]. The overall test consisted of four proportions in the studied plot. components that should be non-significant. Each of the four components was sensitive to different factors; 2.3. Mark-recapture study this allowed potential deviations from the model to be Field work was performed between 2002 and 2005. detected. Specifically, the significance of Test 3.SR The sampling frequency, which was chosen on the suggests the use of models considering a potential assumption that the sampled population was of influence on age on survival rates 30[ ,31]. The lack of the open type [26], was a visit on each second day fit of the general model was quantified by using the throughout the adult emergence period (mid May to variance inflation factor, c-hat (ĉ), which was calculated late July) (Figure 2). At least one living specimen was in four different forms and the most unfavorable found in 29 visits in 2002, 35 in 2003, 25 in 2004 and response adopted [32]. The procedures involved (a) 28 in 2005. Visits were done during the early nocturnal using the goodness of fit data provided by Release in hours (9.30 pm to 3.00 am), which was the period of the form proposed by Burnham et al. (1987) [33] and highest activity for (personal observation). In Lebreton et al. [34], (ĉ= χ2/df); (b) using data provided each visit, each of the 211 trees in the plot was visually by U-Care in the previous expression; (c) subjecting the inspected for the presence of insects, which were starting data to a Bootstrap resampling procedure as captured by hand. Captured insects were classified implemented in Mark; and (d) using the median c-hat as and subsequently marked with a customized colour proposed by Cooch and White [28]. code that was painted on their prothorax and a number The procedure used to calculate the survival and was written with a Paintmaker permanent marker from recapture rates involved the following [34-40]: Pilot (Tokyo, Japan) on their elytra. Though some of the (a) A global model compatible with the biology of marks were partially getting lost, the double code system the species concerned and a sampling design used allowed reconstructing the original marks. Therefore, to assess the fit. We used the above-mentioned CJS there is no evidence for any loss of information. The model and, included Time Since Marked (TSM) models marking technique used had no observable influence on if Test 3.RS was significant. While structurally, these insect behaviour. Marked insects were released on the are “age models”, they are used when all marked trees where they were captured, and each subsequent individuals are adults (with a single individual age), so recapture was recorded. any heterogeneity in the sample is not due to age, but The total number of insects studied over the four rather to the time elapsed since marking. In its most years was 1497, namely: 385 (152 females and 233 general form, the model is represented by the expression

males) in 2002, 511 (211 females and 300 males) in Phi[m(m2_t/t)‌‌f(m2_t/t)] P(g) (survival rate as a function of 2003, 249 (105 females and 144 males) in 2004 and the male (m) and female (f) group, each considered over

352 (132 females and 220 males) in 2005. All of them two sampling intervals (m2) and variable in time; variable were classified as C. welensii. There was no evidence recapture between groups. for the presence of other similar cerambycids species. (b) A selection of the most parsimonious model by applying Akaike’s Information Criterion (AIC) or 2.4. Analysis of data its corrected version (QAICc) for small samples if ĉ Whether a population is of the open or closed type suggests inadequate fitting. dictates the choice of method for analysing field data (c) An estimation of the maximum likelihood of the and estimating population parameters [27]. Of the many model parameters from precision estimators. The final methods available for open populations, we chose to estimates of the parameters were determined according use one that includes living individuals as recaptures to Seber and Schwarz [41]; this involved averaging the only. estimates for statistically equivalent models with an The original data was expressed in binary form appropriate weighting function. (i.e. 011001…) where 0 denoted the absence and 1 The recapture and survival rates of male and the presence of a given insect on a specific visit; this female individuals as calculated using mark-recapture is the usual choice for logistic regression in calculating techniques can check for behavioural hypotheses for basic population parameters. Analyses were started this species. Survival likelihood and recapture rate by checking whether the most general Cormack-Jolly- can be calculated by using the Cormack-Jolly-Seber Seber (CJS) model, Phi(t*g)P(t*g), fitted the field data; to (CJS) model as implemented in the software Mark this end, we used a modified version of the programme v. 4.0, which processes capture-mark-recapture data. This method allows parameters to be estimated with

434 G. López-Pantoja, L.D. Nevado, I. Sánchez-Osorio

2002 2003 2004 2005 Cerambyx welensii Küster male female male female male female male female N(0,1) statistic for transient >0 0.767 –0.344 3.138 1.134 2.276 2.609 2.319 1.141 Test3.SR P-level two-sided test 0.443 0.712 0.002* 0.257 0.02* 0.01* 0.02* 0.254 Quadratic chi2 0.000 – 6.456 0.000 3.294 0.000 0.000 – Test3.SM P-level 1.000 – 0.994 1.000 0.993 1.000 1.000 – N(0,1) signed statistic for trap dependence –0.370 –1.464 –3.549 –0.145 –0.459 –0.731 –1.205 0.000 Test2.CT P-level two-sided test 0.711 0.143 0.000* 0.885 0.646 0.465 0.228 1.000 Quadratic chi2 5.856 1.872 19.713 0.708 6.793 2.706 4.264 0.936 Test2.CL P-level 0.951 0.759 0.762 1.00 0.963 0.745 0.978 0.988

Table 1. Results for Cerambyx welensii Küster provided by U-Care.

2002 2003 2004 2005 Based on ĉ, the data for all four years exhibited a Release 1.192 1.154 1.489 1.453 good fit. Irrespective of the method, ĉ was very close to 1 (perfect fitting) and rather different from 3 the acceptable U-care 0.587 0.803 0.501 0.490 upper limit (Table 2). A ĉ value of 1.19 for 2002, 1.45 for Bootstrap 1.191 1.454 1.348 1.439 2004, 1.49 for 2004 and 1.45 for 2005 was adopted to Median C-hat 0.928 1.051 1.373 1.401 correct for the studied models. Table 2. Variance inflation factor obtained each year using various computation methods. 3.2. Selection of the best model a minimal residual deviation [34]. Also, it provides the Table 3 shows the models exhibiting the lowest QAICc confidence intervals for the estimates and a solid base values and a difference equal to or less than 2.5 in this to check hypotheses concerning survival likelihood and parameter. A total of 5 models fit the data optimally. None recapture between groups [42]. of them involved any changes in the target parameters The TSM models were used to calculate the survival with time. Also, models resulted in differences between rate over the first sampling interval (the transient groups (sexes), especially in recapture rate. Classical -individuals that are seen once but never recaptured- CJS models proved as effective as the TSM models survival rate, Phi ) and for the other intervals (resident 1 for all four years. The former expose the presence of survival rate, Phi ). These parameters were processed 2 transients in the sampled population. Three of the five according to Cooch and White [28], and Perret et al. models assumed the presence of such individuals [43], in order to calculate the proportions of transient among males and/or females. Fitting them to the data individuals present in the sample (τ) and the overall revealed differences between years (Table 3). population (T). 3.3. Survival Based on the results, each year the survival rate of 3. Results the resident individuals was constant over time for both females and males. There were, however, some 3.1. Fitting of data differences between years, and also between the groups The data provided by U-Care suggested the presence (Table 4). In any case, such differences were insignificant of some heterogeneity in the study sample (Table 1). for both males and females. Thus, as can be seen from Specifically, Test 3.SR was significant (α = 0.005) for Figure 5, the confidence intervals for the survival rates three of the four studied years (p ≤ 0.02 in 2003, 2004 overlapped (see [44], for a statistical interpretation of and 2005), which indicates the simultaneous presence this result). The overlap was especially strong among of resident and transient individuals in the sample. females. By contrast, the males exhibited less marked Therefore, the population can be heterogeneous in coincidence (especially in the result for 2003). at least one of the target parameters: the survival The male group exhibited survival rates slightly likelihood. exceeding those of the female group. The difference, Specifically, the heterogeneity appears to be due to however, was also insignificant. This parameter ranged the males. Thus, their test was significant for three of the from 83.7% (0.017 SE) in 2002 to 91.7% (0.010 SE) in four years; by contrast, the test for the females exhibited 2003 for the male group, and from 82.8% (0.037 SE) in no clear-cut trend and was only significant for one year 2005 to 89.8% (0.031 SE) in 2004 for the female group (2004). (Table 4).

435 Mark-recapture estimates of the survival and recapture rates of Cerambyx welensii Küster (Coleoptera cerambycidae) in a cork oak dehesa in Huelva (Spain)

QAICc Model QAICc ∆QAICc No. par weights {Phi(.) p(g)} 756.995 0.000 0.391 3

2002 {Phi(m(.) f(m2./.)) p(g)} 757.559 0.564 0.295 4 {Phi(g) p(g)} 758.257 1.262 0.208 4 {Phi(m(m2 ./.) f(m2./.)) p(g)} 759.587 2.592 0.107 5 QAICc Model QAICc ∆QAICc No. par weights

2003 {Phi(m(m2./.)f(.) p(g)} 1396.560 0.000 0.651 5 {Phi(m(m2./.) f(m2./.) p(g)} 1398.586 2.026 0.236 6 QAICc Model QAICc ∆QAICc No. par weights 2004 {Phi(m(m2./.) f(m2./.) p(g)} 492.382 0.000 1.000 6 QAICc Model QAICc ∆QAICc No. par weights {Phi(.) p(g)} 526.382 0.000 0.530 3 2005 {Phi(m(m2./.) f(.)) P(g)} 527.916 1.534 0.246 5 {Phi(g) p(g)} 528.105 1.723 0.224 4

Table 3. CJS and TSM models best fitting the QAICc results for Cerambyx welensii Küster in the studied period.

Males Females Year Phi 95% C.I. Year Phi 95% C.I. 2002 0.837 0.801–0.868 2002 0.853 0.781–0.905 2003 0.917 0.895–0.934 2003 0.858 0.792–0.905 2004 0.910 0.867–0.939 2004 0.898 0.82–0.945 2005 0.841 0.791–0.881 2005 0.828 0.743–0.889

Table 4. Mean estimates of the apparent survival rate (Phi) for resident individuals. 2004. Based on the model fitting the data, one would have expected the presence of transient females in 2002 and their absence in 2005. The transience rates for the male group increased over the studied period. The year 2004 showed the highest transience rates for both groups (T = 0.174, 0.168-0.179, 95% CI for males; and T = 0.392, 0.364-0.421, 95% CI for females).

3.5. Recapture The recapture rate also remained constant throughout the flight period and it was constant within each group. Therefore, the recapture likelihood of both females and males was independent of their emergence time. Both groups exhibited a slight difference in recapture rate Figure 5. Apparent survival rate (and 95% confidence interval) for resident individuals as calculated over the entire yearly between years. The differences, however, were almost flight period (∆ males ■ females). negligible for the male group. The standard errors for 3.4. Transience the estimates strongly overlapped each other. The Time Since Marked models (TSM) were used to differences between years for the female group were calculate the proportions of transients present in the similar to those for the male group except in 2004, where sample (τ) and the overall population (T) except in both the recapture rate (P = 0.218) and its confidence 2002 and 2005 for the females (Tables 5 and 6). The interval (α= 0.05, 0.074-0.372) departed from those of proportion of transient individuals for the female group the other years (Figure 6). The overall recapture rate of changed dramatically from 1.1% in 2003 to 39.2% in the male group exceeded that of the female group. The

436 G. López-Pantoja, L.D. Nevado, I. Sánchez-Osorio

τ Year Phi1 95% C.I. Phi2 95% C.I. 95% C.I. Males 2002 0.826 0.754–0.898 0.837 0.820–0.854 0.013 0.000–0.103 2003 0.800 0.756–0.844 0.919 0.910–0.928 0.129 0.079–0.180 2004 0.678 0.604–0.752 0.912 0.893– 0.931 0.257 0.158–0.355 2005 0.716 0.643–0.789 0.857 0.834–0.880 0.165 0.063–0.266 Females 2002 – – 0.853 0.826–0.880 – – 2003 0.842 0.673–1.000 0.847 0.827–0.881 0.014 0.000–0.216 2004 0.468 0.468–0.378 0.889 0.851–0.929 0.474 0.121–0.828 2005 – – 0.814 0.757–0.857 – –

Table 5. Parameter values provided by the TSM.

Phi1: apparent survival between the first two sampling visits; Phi2: apparent survival between the other sampling visits; τ: proportion of tran- sient individuals in the sample

Males Females Year T 95% C.I. Year T 95% C.I. 2002 0.009 0.005–0.013 2002 – – 2003 0.072 0.071–0.074 2003 0.011 0.000–0.023 2004 0.174 0.168–0.179 2004 0.392 0.364–0.421 2005 0.122 0.117–0.127 2005 – –

Table 6. Transience index (T) for the female and male groups.

Males Females Year P 95% C.I. Year P 95% C.I. 2002 0.186 0.137–0.247 2002 0.039 0.018 –0.082 2003 0.190 0.153–0.233 2003 0.078 0.043–0.139 2004 0.187 0.125–0.270 2004 0.218 0.074–0.372 2005 0.175 0.111–0.264 2005 0.071 0.032–0.151

Table 7. Mean estimates of the recapture rate (P). effects by restricting most of their activity to the nocturnal hours and taking shelter in holes or hollows during the day. In this way, they mimic the behavioural patterns of saproxylic insects [45], the life strategy of which adjusts to the environmental stability of their habitat [46-48]. The combination of a partly cavernous life and nocturnal activity habits additionally reduces the already small number of potential predators even in an environment close to such a highly biodiverse protected Figure 6. Recapture rate (and 95% confidence interval) as calcu- natural space as the Doñana National Park and Doñana lated over the entire yearly flight period (∆ males ■ fe- Natural Park. males). The survival rates found during the study can be difference between the two was significant. However, deemed high, especially if one considers the presence the highest rate found was that for the female group in of transient individuals among the population. This was 2004 (Table 7). the case with both males and females, and invariably resulted in underestimated survival rates [49]. The survival rates for the female group might be further 4. Discussion underestimated by low recapture rates [50]. These results are consistent with the absence of limiting Diurnal temperatures in the Mediterranean region can factors, whether endogenous or exogenous, controlling be extreme and increase the risk of dehydration or the population via an increased mortality. reduce food resources such as liquid exudates from The little change in survival rate between years trees. However, adult insects avoid these unfavourable indicates that fluctuations in biotic and abiotic factors

437 Mark-recapture estimates of the survival and recapture rates of Cerambyx welensii Küster (Coleoptera cerambycidae) in a cork oak dehesa in Huelva (Spain)

in the study area are either small or uninfluential As a result of this situation, the structure of the forest on mortality in the studied population [51,52]. The formation can collapse as a habitat within one hundred greatest environmental oscillations detected during the years if young trees in the plot are not colonized by the studied period (personal observations) were those in insects. If the present mortality rate persists, all currently temperature, especially as regards the time distribution living trees in the plot can die and the insect population of the maximum values (in the form of an advanced or can disappear within the next 180 years. The virtual delayed summer). Such oscillations do seem to affect isolation of the farm reduces the possibility of colonization the emergence cycle of the adults [9,53]. of nearby vegetation patches and increases the chance The insignificant differences in survival rates of young trees being infested by the insects. between males and females indicates that factors such The presence of transient individuals throughout as weakening from fights between males or additional the studied period is suggestive of two different life efforts during laying in females make no mortality strategies. A major strategy that involves the resident differences between the two sexes. The survival rate of individuals and a minor but significant strategy involves some proterandric insects is reduced by the increased the transient individuals. The detection of phoresic effort and risk involved in males waiting for females to pseudoscorpions associated with cerambycids is emerge in order to copulate. This tendency, which has an accurate reflection of the active character of the been preferentially observed in lepidoptera [54-56], individuals of these species [59]. has also been reported for some cerambycids such as The transience rate of males tended to increase with Tetraopes tetraophthalmus Forster (1771) [57], but no time. No clear-cut trend could be inferred in females, evidence exists of its presence in C. welensii. however, it could not be determined each studied year. Both the gradual increase in the proportion of There is no biological explanation for this trend except infested trees in the studied plot [44% in 2002 and for a potential effect of stripping of the tree cork in 70% in 2005 (unpublished results)] and the presence 2003 and 2004 on the number of available laying and of infested young trees can indicate that the population sheltering sites. Because the characteristics of the trunk level exceeded that of its initial colonization stage. are essential elements of the habitat structure, their Under these conditions, the self-improvement alteration can be expected to affect male and female obtained by making new holes both adult shelter and movements [60]. eggs lay, as previously observed in xylophagous and In this respect, studies on Tetraopes tetraophthalmus saproxylic species [19,48], can be near of its limit by (Col. Cerambycidae) suggest that a variable abundance the way of the habitat collapse. Whether the present of mates and competitors facilitates the rapid evolutionary pattern of the habitat can by itself destabilize development of local mobility in males. By contrast, a the insect population is unknown. However, the relative stable habitat favours sedentary habits in young females constancy of the mortality rate over the studied period [60]. C. welensii could follow these patterns since and suggests that the insects are potentially capable of living altered environment seemingly compels females to be on suboptimal resources such as the young trees and more active and also males in response. moving to more favourable areas. Based on previous studies such as the one by Under conditions of fragmentation and isolation, the Prieto et al. [61], the recapture rate can be used as time a habitat takes to reach a residual state leading an indirect indicator of the character of a population. to the local extinction of an insect species can be as When P > 60%, the population concerned is assumed long as a few centuries for some species living on to be confined (i.e. immigration and emigration are both decaying wood [48,58]. However, the time needed for negligible) and, even if it is of the open type, it behaves the gradual deterioration and eventual death of trees to as if it were not. Our P values were much lower than substantially reduce available resources for larvae or 60%, so our population is not confined. Rather, there is even adults of drilling species with larvae developing continuous exchange of individuals with other portions in living trees can be much shorter. Trees in the study of the study area. area exhibit fast decay. Thus, they are disappearing by The mobility of the individuals of this population is effect of cerambycids breaking trunks and branches at also confirmed by their site loyalty, estimated from the a rate above 0.40 trees/ha*yr (personal observations). survival rate as recommended by some authors [62]. In addition to the intrinsic effects of the population In this context, the term “survival” is replaced with dynamics, deficient forestry practices and the withdrawal “residency” [60] as no distinction between migration or in situ burning of large dead branches and trunks and death is possible in calculating this parameter are leading to an increased instability of the habitat for [63,64]. The residence time for insects in the studied drilling insects. plot as calculated from the expression -ln (Φ)-1

438 G. López-Pantoja, L.D. Nevado, I. Sánchez-Osorio

[40,42,55,60,65,66], was 7 days on average and never Acknowledgements exceeded 12 days. Compared to the 48 days of maximal survival derived from the capture-recapture observations We want to aknowledge the Environmental Council during the study, these values can be deemed typical of of the Andalusian regional government for financial a moving population. support. We thank Dr Carl Schwarz for his scientific Males and females markedly differ in recapture rate. support and his advice about popan models’ use. We Biologically, this confirms a disparate behaviour in the are very gratefull to Toñi Paramio, Miguel Angel Gay and two groups that influences the recapture rate. Thus, David Cremades for helping the field work. Finally we one can expect one sex to be more fleeting or exhibit a want to thank Rául Tapias and Patricia Alesso for the higher survival rate than the other making its recapture their technical and field support at the beginning of the more likely. However, the survival rates for males and work. females were very similar, which suggests that the latter are markedly fleeting. This is consistent with field observations where females were seen laying their eggs or even copulating in sheltered or hidden places (e.g. trunk cracks, old cerambycid holes).

References

[1] Council of the European Communities, 1992, Coun- Ibérica, islas Baleares e islas atlánticas: Canar- cil Directive 92/43/EEC on the conservation of ias, AÇores y Madeira, Bol. Soc. Ent. Ar., 2007, 40, natural habitats and of wildfauna and flora (Habi- 595-596, (in Spanish) tats Directive), Official Journal of the EEC L206 [9] López G., Domínguez L., Sánchez I., Tapias R., Cre- 22/07/1992:7–50 mades D., Paramio A., et al., Population ecology of [2] Sánchez M.E., Navarro R. M., Trapero A., Fernán- xylophagous beetles (Coleoptera: Cerambycidae) in dez P., La “seca” de encinas y alcornoques: una Mediterraneam Quercus forest (southwest of Iberian visión histórica, Montes, 2000, 62, 29–39, (in Span- Peninsule). Incidence on oak trees health (Quer- ish) cus ilex L. spp ballota and Quercus suber L.), In: [3] Soria F.J., Villagrán M., Cárdenas A.M., Distribución Arianoutsou- Papanastasis (Ed.) Ecology, Conser- e incidencia de los principales perforadores de la vation and Management of Mediterraneam Climate encina en el Parque Natural de las Sierras Subbé- Ecosystems. Proceedings 10th MEDECOS Confer- ticas (Córdoba, Spain). Zoologica Baetica, 1996, 7, ence (April 25– May 1 2004, Rhodes, ) 33-43, (in Spanish) [10] Moral del J., Casado D., Gallego M., Rey J. M., [4] Extremera F.M., Cobo A., Pérez Rodríguez M.C., Presencia de insectos parásitos del grupo Cer- Pérez Guerrero S., Vargas Osuna E., El complejo ambyx cerdo en la dehesa extremeña, Phytoma- de lepidópteros defoliadores de Quercus en la pro- España, 1994, 59, 44-52, (in Spanish) vincia de Córdoba, Bol. San. Veg. Plagas, 2004, 30, [11] Rodríguez-Molina M.C., Santiago R., Blanco A., 203-209, (in Spanish) Pozo J.D., Colino M.I., Palo E.J., et al., Detec- [5] Ferreira M.C., Ferreira G.W.S., Platipus cylindrus F. ción de Phytophthora cinnamomi en dehesas de (Coleóptera, Platypodidae), plaga de Quercus suber Extremadura afectadas por “seca” y su compor- L., Bol. San. Veg. Plagas, 1989, 15, 301-306, (in tamiento in vitro, Bol. San. Veg. Plagas, 2003, 29, Spanish) 627-640, (in Spanish) [6] Bense U., Longhorn Beetles. Illustrated Key to [12] Romanyk N., Cadahia D., Plagas de Insectos en the Cerambycidae and Vesperidae of Europe, las Masas Forestales Españolas. Coedición: So- Margrat Werlag, Weikersheim,1995 ciedad Española de Ciencias Forestales. Ediciones [7] Gonzáñez Peña F.,Vives i Noguera E., Sousa Zu- Mundi-Prensa. Madrid, Spain, 2002, (in Spanish) zarte A.J., Nuevo Catálogo de los Cerambycidae [13] Sama G., Atlas of the Cerambycidae of Europe (Coleoptera) de la Península Ibérica, islas Baleares and Mediterranean area, Vol. 1, Vit Kabourek, Zlín, e islas atlánticas: Canarias, AÇores y Madeira, 2007, 2002 Monografías SEA, Vol. 12, (in Spanish) [14] Martín J., Cabezas J., Buyolo T., Patón D., The [8] González Peña F.,Vives i Noguera E., Sousa Zuzarte relatonship between Cerambyx spp. damage and A.J., Addenda et corrogenda al Nuevo Catálogo de subsequent Biscogniauxia mediterranum infec- los Cerambycidae (Coleoptera) de la Península tion on Quercus suber forest, Forest Ecol. Manag.,

439 Mark-recapture estimates of the survival and recapture rates of Cerambyx welensii Küster (Coleoptera cerambycidae) in a cork oak dehesa in Huelva (Spain)

2005, 216, 166–174 [27] Schwarz C.J., Seber G.A.F., Estimating [15] Montoya J.M., Mortandad de quercíneas: la per- abundance: review III. Statistical Science, 1999, spectiva selvícola y los antecedentes climáticos. 14, 427-456 La cuestión de Hypoxylon mediterraneum en el al- [28] Cooch E., White G., Program Mark “A gentle in- cornocal de Mamora (Marruecos), Ecología, 1992, troduction”, 2006, http://www.phidot.org/software/ 6, 123–130, (in Spanish) mark/docs/book/ [16] El Antry S., Biologie et dègâts de Cerambyx cerdo [29] Choquet R., Reboulet A.M., Pradel R., Giménez mirbecki Lucas (Coléoptère, Cerambycidae) en O., Lebreton J.D., User’s manual for U-Care. Mim- subéraie de la Mamora (Maroc), Integrated Protec- eographed document, CEFE/CNRS, Montpellier, tion in Oak Forests IOBC, Bulletin, 1999, 22, 59–64 2003, ftp://ftp.cefe.cnrs-mop.fr/biom/Soft-CR/ (in French) [30] Pradel R., Hines J.E., Lebreton J.D., Nichols J.D., [17] Moral del J., Cerambyx spp., historia de una plaga Capture–recapture survival models taking account de las dehesas extremeñas, Phytoma-España, of transients, Biometrics, 1997, 53, 60-72 1994, 60, 18–24, (in Spanish) [31] Prévot-Julliard A.C., Lebreton J.D., Pradel R., Re- [18] Buse J., Ranius T., Assmann T., An endangered evaluation of adult survival of black-headed gullls longhorn associated with old oaks and its (Larus ribidundus) in presence on recapture het- possible role as an ecosystem engineer, Conserv. erogeneity, Auk, 1998, 115, 85-95 Biol., 2008, 22, 329-337 [32] Conroy M.J., Senar J.C., Doménech J., Analysis [19] Faytl Ph., Dufrêne M., Branquart E., Hastir P., Pon- of individual- and time-specific covatiate effects on tégnie P., Henin J., et al., Contrasting responses survival of Serinus serinus in north-eastern Spain, of saproxylic insects to focal habitat resources: the J. Appl. Stat., 2002, 29, 125–142 example of longhorn beetles and hoverflies in Bel- [33] Burnham K.P., Anderson D.R., White G., Brownie gian deciduous forests, J. Insect Conserv., 2006, C., Pollock K.H., Design and analysis methods for 10, 129–150 fish survival experiments based on release–recap- [20] Dupont P., Zagatti P., (2005). Cerambyx cerdo Lin- ture, Am. Fish. Soc., 1987, 5, 1–437 né, 1758. http://www.inra.fr/Internet/Hebergement/ [34] Lebreton J.D., Burnham K.P., Clobert J., Ander- OPIE-Insectes/observatoire/coleos/cerambyx/c_ son D.R., Modeling survival and testing biological cerdo.htm hypotheses using marked : a unified ap- [21] Sánchez-Osorio I., Orientación olfativa de Cer- proach with case studies, Ecological Monographs, ambyx welensii Küster y Prinobius germari De- 1992, 62, 67–118 jean, principales cerambícidos xilófagos de enci- [35] Anderson D.R., Link W.A., Johnson D.H., Burnham na (Quercus ilex L. subsp. ballota) y alcornoque K.P., Suggestion for presenting the results of data (Quercus suber L.), para la localización de hos- analyses, Journal of wildlife management, 2001, pedantes, PhD thesis, Huelva University, Huelva, 65, 373-379 Spain, 2005, (in Spanish) [36] Andrés J.A., Cordero A., Survival rates in a natural [22] Hanks L.M., Influence of the larval host plant on re- population of the damselfly Ceriagrion tenellum: ef- productive strategies of cerambycid beetles, Annu. fects of sex and female phenotype, Ecol. Entomol., Rev. Entomol., 1999, 44, 483–505 2001, 26, 341-346 [23] Strojny W., Studies on the biology of the great cap- [37] Blomberg S.P., Shine R., Modelling life history ricorn beetle Cerambyx cerdo L. (Coleoptera: Ce- strategies with capture–recapture data: Evolution- rambycidae) living on pedunculate oaks –Quercus ary demography of the water skink Eulamprus tym- robur L – in Swojec and Wilczyce near Wroclaw in panum, Austral. Ecol., 2001, 26, 349–359 1973–1976, Polskie Pismo Entomologiczne, 1977, [38] Jamieson I.G., Forbes M.J., McKnight E.B., Mark– 47, 727-746 recapture study of mountain stone weta Hemideina [24] Vives E., Coleoptera, Cerambycidae. Fauna Ibéri- maori (Orthoptera: Anostostomatidae) on rock tor ca. Museo Nacional de Ciencias Naturales. CSIC. ‘islands’, New Zeal. J. Ecol., 2000, 24, 209–214 Madrid, Spain, Vol. 12, 2000, (in Spanish) [39] Joyce1 S.J., Jamieson I.G., Barker R.,. Survival [25] Vives E., Atlas fotográfico de los Cerambícidos Íbe- of adult mountain stone weta Hemideina maori ro–Baleares. Editorial Argania. Barcelona, Spain, (Orthoptera:Anostostomatidae) along an altitude 2001, (in Spanish) gradient as determined by mark–recapture, New [26] Southwood T.R.E., Henderson P.A., Ecological Zeal. J. Ecol., 2004, 28, 55–61 methodology, Blackwell Science, Ltd. Oxford, UK, [40] Leisnham P.T., Cameron C., Jamieson G., Life cy- 2000 cle, survival rates and longevity of an alpine weta

440 G. López-Pantoja, L.D. Nevado, I. Sánchez-Osorio

Hemideina maori (Orthoptera: Anostostomatidae) [54] Aukland J.A., Debinski D.M., Clark W.R., Sur- determined using mark–recapture analysis, New vival, movement, and resource use of tha butter- Zeal. J. Ecol., 2003, 27, 191–200 fly Parnassius clodius, Ecol. Entomol., 2004, 29, [41] Seber G.A.F., Schwarz C.J., Capture–recapture: 139–149 before and after EURING 2000, J. Appl. Stat., [55] Schtickzelle N., Le Bolulengé E., Baguette M., 2002, 29, 5–18 Metapopulation dynamics of the bog fritillary but- [42] Anholt B.R., Vorburger C., Knaus P., Mark–recap- terfly: demographic processes in a patchy popula- ture estimates of daily sulvival rates of two dam- tion, Oikos, 2002, 97, 349-360 selflies (Coenagrion puella and Ischnura elegans), [56] Wiklund C., Torbjörn F., Why do males emerge be- Can. J. Zool., 2001, 79, 895–899 fore females?, Oecologia, 1977, 31, 153-158 [43] Perret N., Pradel R., Miaud Grolet O., Joly P., Tran- [57] Yanega D., Field guide to Northeastern longhorned sience, dispersal and survival rates in newt patchy beetles (Coleoptera:Cerambycidae). Illinois Natu- populations, J. Anim. Ecol., 2003, 72, 567-575 ral History Survey, Manual 6, October, 1996 [44] Hilton B.Jr., Miller M.W., Annual survival and re- [58] Whitlock M.C., Nonequilibrium population structure cruitment in a Ruby Throated Hummigbird popu- in forked fungus beetles:extinction, colonization, lation, excluding the effect of transient individuals, and the genetic variance among populations, Am. The Condor, 2003, 105, 54–62 Nat., 1992, 139, 952-970 [45] Dajoz R., Entomologia forestal: los insectos y el [59] Ranius T., Douwes P., Genetic structure of two bosque, Mundi-prensa, Madrid, Spain, 2001, (in pseudoscorpion species living in tree hollows in Spanish) Sweden, Anim. Biodivers. Conserv., 2002, 25, [46] Jonsson M., Dispersal Ecology of Insects Inhabit- 67-74 ing Wood-Decaying Fungi, PhD Thesis, Uppsala, [60] Lawrence W.S., Movement ecology of the rad Sweden, 2002 milkweed beetle in relation to population size and [47] Ranius T., Constancy and asynchrony of Osmoder- structure, J. Anim. Ecol., 1988, 57, 21–35 ma eremita populations in tree hollows, Oecologia, [61] Prieto C.,Takegami C., Rivera J.M., Estructura po- 2001, 1126, 208-215 blacional de Morpho sulkowskyi Kollar 1850 (Lepi- [48] Ranius T., Population ecology and conservation of doptera: Nymphalidae) en un sector de la cordillera beetles and pseudoscorpions living in hollow oaks occidental, departamento del Cauca (Colombia), in Sweden, Anim. Biodivers. Conserv., 2002, 25, Entomotropica, 2005, 20, 15-22, (in Spanish) 53-68 [62] Bart J., Flinger M.A., Notz W.I., Sampling and sta- [49] Pradel R., Transient animals in a resident popula- tistical methods for behavioral ecologists, Cam- tion of snow, J. Appl. Stat., 1995, 22, 695-710 bridge University Press, Cambridge, UK, 1998 [50] Begon M., Ecología animal. Modelos de cuantifi- [63] Bergman K.O., Landin J., Population structure cación de poblaciones, Ed. Trillas, Mexico, 1989, and movements of a threatened butterfly (Lopinga (in Spanish) achine) in a fragmented landscape in Sweden, [51] Covas R., Brown C.R., Anderson M.D., Brown Biol. Conserv., 2002, 108, 361–369 M.B., Juvenile and adult survival in the sociable [64] Knight A., Brower L.P., Williams E.H., Spring remi- weaver (Philetairus socius), a southern–temperate gration of the monarch butterfly, Danaus plexippus colonial cooperative breeder in Africa, Auk, 2004, (Lepidoptera: Nymphalidae) in north-central Flori- 121, 1199–1207 da: estimating population parameters using mark– [52] Oro D., Aguilar J.S., Igual J.M., Louzao M., Model- recapture, Biol. J. Linn. Soc., 1999, 68, 531–556 ling demography and extinction risk in the endan- [65] Fisher R.A., Ford E.B., The spread of a gene in gered Balearic shearwater, Biol. Conserv., 2004, natural contitions in a colony of the moth Panaxia 116, 93-102 dominula L., Heredity, 1947, 1, 143–174 [53] Muñoz C., Cobos P., Martínez G., Soldevilla C., [66] Cook L.M., Brower L.P., Croze H.J., The accurancy Díaz M., Micoflora y patología del alcornoque of a population estimation from multiple recapture (Quercus suber L.). Ministerio de Agricultura, Pes- data, J. Anim. Ecol., 1967, 57–60 ca y Alimentación. Madrid, Spain, 1996, (in Span- ish)

441