Introduction
The olive moth, Prays oleae (Bernard) (Lepidoptera: Praydidae), is one of the most destructive pests of the Mediterranean olive groves. This insect develops three generations per year that damage different components of the olive tree (Alvarado et al., Reference Alvarado, Civantos, Durán, Barranco, Fernández-Escobar and Rallo2008). The leaf or phyllophagous generation develops to adulthood from September to March, consuming the inner side of the leaves. The flower or anthophagous generation, present from March to June, feeds on the flower bud devouring anthers, stigma and ovaries. The fruit or carpophagous generation occurs from June to September, when the larvae penetrate the developing fruit and feed on the gelatinous tissue of the developing stone causing its premature fall (Bento et al., Reference Bento, Torres, Cabanas, Pereira and Azevedo2007; Gonçalves et al., Reference Gonçalves, Oliveira and Torres2007).
Although olive moth populations can be controlled by a rich complex of antagonists, it sometimes assumes a key pest status. It is recognized that this situation is largely due to these antagonists not being sufficiently effective due to, a spatial or temporal lack of food sources (such as pollen and nectar) (Jervis et al., Reference Jervis, Kidd, McEwen, Campos and Lozano1992).
The use of plants to provide nectar and pollen resources for natural enemies through habitat management is an increasing focus of conservation biological control. Some authors argue that this can be expedited by sowing flowering plants in the crop or in its vicinity (Maingay et al., Reference Maingay, Bugg, Carlson and Davidson1991; Lövei et al., Reference Lövei, Hodgson, Macleod, Wratten, Corey and Milne1993). Moreover, studies also suggest that, when properly selected, autochthonous plants can be as valuable as exotic species to manage the habitat, with the advantage of not only being locally adapted, but also due to the fact that their use can also contribute to increasing the natural biodiversity and multifunctionality (Nentwig, Reference Nentwig, Pickett and Bugg1998; Nicholls et al., Reference Nicholls, Parrella and Altieri2000; Fiedler & Landis, Reference Fiedler and Landis2007a , Reference Fiedler and Landis b ) with a consequent local economic return (Bertacchini, Reference Bertacchini2009).
Many adult parasitoids and predators benefit from shelter and food through combinations of flowering plants (Root, Reference Root1973; Hickman & Wratten, Reference Hickman and Wratten1996; Harmon et al., Reference Harmon, Ives, Losey, Olson and Rauwald2000; Landis et al., Reference Landis, Wratten and Gurr2000). Floral resources help the fitness of a wide range of agriculturally beneficial invertebrates, including lacewings (Robinson et al., Reference Robinson, Jonsson, Wratten, Wade and Buckley2008), parasitoids (Irvin et al., Reference Irvin, Hoddle and Castle2007), coccinellids (Bertolaccini et al., Reference Bertolaccini, Nunez-Perez and Tizado2008), spiders (Taylor & Bradley, Reference Taylor and Bradley2009) and hoverflies (Van Rijn et al., Reference Van Rijn, Kooijman and Wäckers2006; Pinheiro et al., Reference Pinheiro, Torres, Raimundo and Santos2013a , Reference Pinheiro, Torres, Raimundo and Santos b ).
The selection of the most appropriate plant species for habitat management may be influenced by multiple factors, such as: nectar availability (Idris & Grafius, Reference Idris and Grafius1995) and accessibility (Wäckers et al., Reference Wäckers, Björnsen and Dorn1996), duration of flowering (Lövei et al., Reference Lövei, Hodgson, Macleod, Wratten, Corey and Milne1993) and time of flowering (Wratten & Van Emden, Reference Wratten, Van Emden, Glen, Geaves and Anderson1995). Negative aspects, such as the potential for pest populations also benefiting, resulting in an increased crop damage, must also be considered (Gurr et al., Reference Gurr, Van Emden, Wratten and Barbosa1998).
An important aspect of nectar exploitation by insects is the morphological fit between their mouthparts and the floral architecture (Jervis, Reference Jervis1998). Based on the structure of the flowers and the mouthparts of the pest and its natural enemies, nectar accessibility can be theoretically studied (Wäckers et al., Reference Wäckers, Van Rijn and Bruin2005; Vattala et al., Reference Vattala, Wratten, Phillips and Wackers2006). Natural enemies with short mouthparts may be unable to reach flowers’ nectar when there are deep corollas, and therefore benefit from flowers that have open, shallow corollas (Wratten et al., Reference Wratten, Lavandero, Tylianakis, Vattala, Cilgi and Sedcole2003).
Accessibility is important to ensure not only that the natural enemies can access floral resources, but also ensures that energy is not wasted through futile foraging (Winkler et al., Reference Winkler, Wäckers, Kaufman and Van Lenteren2009). Hence, assuming that the system is resource-limited, ensuring that the plants are correctly distributed spatially and temporarily is important to guarantee the optimal provision of the floral resources for the fulfillment of pest control goals (Tompkins, Reference Tompkins2010).
Clearly, the flowering period needs to coincide with the presence of the natural enemy in the system (Landis et al., Reference Landis, Wratten and Gurr2000). Notwithstanding this, there is also the need to consider when the natural enemy is most resource limited, as it is then that supplementation of its diet would allow for its continued presence in the system (Jonsson et al., Reference Jonsson, Wratten, Robinson and Sam2009). Based on this without a thorough understanding of an agricultural system's invertebrate complex (and the various abiotic factors interacting) the effectiveness of conservation biological control manipulations will be unpredictable.
Identifying floral resources native to each country, which can also enhance natural enemy fitness, would add their value within agricultural landscapes and provide more Ecological Services ‘for the price of one’ (Tompkins, Reference Tompkins2010). These services include for example, crop pollination and pest control, which help to maintain agricultural productivity and reduce the need for pesticide inputs (Isaacs et al., Reference Isaacs, Tuell, Fiedler, Gardiner and Landis2009)
The present work was developed to identify a set of weed species from olive groves, that may enhance the impact on both the olive moth pest, and its natural enemies such as the hymenopteran parasitoids species Ageniaspis fuscicollis (Dalman), Apanteles xanthostigma (Haliday), Chelonus elaeaphilus Silvestri and Elasmus flabellatus (Fonscolombe) and the chrysopid predator species Chrysoperla carnea (Stephens), according to the guidelines of Böller et al. (Reference Böller, Häni and Poehling2004).
These natural enemies were selected because they are known to be important in the control of the olive moth, either due to the frequency at which they occur in the Mediterranean region (Torres, Reference Torres, Torres and Azevedo2007), namely in the interior Centre and North East regions of Portugal (Bento et al., Reference Bento, Lopes, Campos and Torres1998; Soares et al., Reference Soares, Gomes, Simão, Veiga, Bento and Torres2005).
In the Mediterranean region, A. fuscicollis is the most important parasitoid of P. oleae showing parasitism rates that can sometimes exceed 80% (Torres, Reference Torres, Torres and Azevedo2007). A. xanthostigma is a parasitoid whose biological cycle is synchronized with that of the phytophagous stage of P. oleae (Campos & Ramos, Reference Campos and Ramos1981) and leads to parasitism of up to 30% of the population (Tominic, 1969 cit. by Stavraki, Reference Stavraki1984). C. elaeaphilus is a common parasitoid of P. oleae, in Portugal in regions such as Trás-os-Montes and Beira Interior (Bento et al., Reference Bento, Lopes, Campos and Torres1998; Soares et al., Reference Soares, Gomes, Simão, Veiga, Bento and Torres2005), where it can parasitize up to 79% of the carpophagous generation population of the pest (Bento et al., Reference Bento, Cabanas, Pereira, Torres, Herz and Hassan2004). E. flabellatus can parasitize up to 11% of the P. oleae phyllophagous generation (Campos & Ramos, Reference Campos and Ramos1981; Bento, Reference Bento1999). C. carnea was the most common chrysopid found in Portuguese olive groves (Bento, Reference Bento1999), and the larvae are the major oophagous predators of the olive moth accounting for over 90% of egg predation in some regions and years (Ramos & Ramos, Reference Ramos and Ramos1990).
Material and methods
Insects
P. oleae and parasitoids were obtained in olive groves from Beira Interior and Trás-os-Montes regions (respectively, Centre and North Eastern of Portugal) between March 2009 and September 2012. Samples of leaves and flowers with larvae (3rd, 4th and 5th instars) and pupae of the P. oleae were collected, in March, during the overwintering (phyllophagous) and the first (anthophagous) generations in June. In September, insects of the carpophagous generation were obtained from recently fallen fruit.
In the laboratory, mined leaves, flowers and fruit with larvae and pupae of P. oleae, were maintained in a climate controlled chamber at 25 ± 2°C, 60 ± 5% relative humidity (RH), photoperiod of 16:8 (L:D) h and daily observed until emergence of either P. oleae or parasitoids adults, identified and separated by species and gender.
C. carnea was obtained from the Koppert Biological Systems (The Netherland) and reared in UV-sterilized frozen eggs of Ephestia kuehniella Zeller (Lepidoptera: Pyralidae) purchased from Biotop (Livron-sur-Drôme, France), in a climate-controlled chamber under the same conditions as those describe above for P. oleae.
Plant species
Twenty-one plant species were chosen (table 1) to be characterized and evaluated in terms of their interests for habitat management, from floristic inventories carried out in olive groves in the Beira Interior region (Nave et al., Reference Nave, Crespí, Campos and Torres2009), based on a group of criteria listed in the bibliography namely: (a) flowering phenology (Rebek et al., Reference Rebek, Sadof and Hanks2005; Winkler, Reference Winkler2005), (b) potential role in supporting natural enemies (Bugg et al., Reference Bugg, Ellis and Carlson1989; Maingay et al., Reference Maingay, Bugg, Carlson and Davidson1991; Lavandero et al., Reference Lavandero, Wratten, Shishehbor and Worner2005; Ambrosino et al., Reference Ambrosino, Luna, Jepson and Wratten2006) and (c) multifunctionality (Rogers & Potter, Reference Rogers and Potter2004; Shrewsbury et al., Reference Shrewsbury, Lashomb, Hamilton, Zhang, Patt and Casagrande2004; Fiedler et al., Reference Fiedler, Landis and Wratten2008).
Plant species selected in this study, corresponding criteria and flowering period.
1 (1) Flowering phenology, (2) Increasing the role natural enemies, (3) Multifunctional role.
Twelve families were represented, in the plants considered in most cases with one or two species except for Apiaceae and Asteraceae family, which had three and four species, respectively, given their importance in feeding natural enemies.
Associating the characteristics of flowering (duration, period of the year) with the success of parasitoids, it was our decision to select species with different flowering periods.
Stipulating three flowering periods, from February to April, from May to July and from August to November, there were seven species (33% of total) in the 1st of these periods, 19 species (86% of total) in the 2nd period, and 17 species (81% of total) in the 3rd period. Further six of the seven flowering plants from the first period were also found to be present in the other two periods; these were: C. bursa-pastoris, L inaria saxatilis, Raphanus raphanistrum, Silene gallica, Spergula arvensis and Trifolium repens. There were also species with shorter flowering periods concentrated in the warmer months of the year, including Foeniculum vulgare, Hypericum perforatum and Lavandula stoechas.
Measurements of insects
Male and female insect morphology was measured, under a stereoscopic microscope with an ocular micrometer using the computer software Digital Imaging Solutions (Olympus Soft Imaging Solutions GmbH, Olympus Europa).
To measure P. oleae proboscis length (fig. 1a) (N = 66, 33 per sex) the proboscis of individuals was unrolled with a needle and measured from the tip to the front of the compound eye. For C. carnea (fig. 1b) and parasitoids (fig 1c–f) that do not have elongated mouthparts, the head width or the thorax width are the limiting factor in exploiting nectar from deep, narrow flower corollas. Thus, the predator and parasitoid head and thorax (N ≥ 12) were measured between the extreme lateral margins of, respectively, the eyes and thorax.
Measurements of P. oleae and its studied natural enemies morphological parameters.
Flower morphology
At least, twelve samples of flowering plants were measured for each plant species. Corolla width and corolla depth of the flowers were measured. The latter was taken from the most proximal point of the corolla to the observed location of the nectar at the corolla base. Corolla width was measured at the width of the narrowest part of the corolla tube (fig. 2) (Baggen et al., Reference Baggen, Gurr and Meats1999; Winkler et al., Reference Winkler, Wäckers, Valdivia, Larraz and Van Lenteren2003; Krenn et al., Reference Krenn, Plant and Szucsich2005; Vattala et al., Reference Vattala, Wratten, Phillips and Wackers2006).
Measurements of the flower morphology, shown for the case of R. raphanistrum.
Data analysis
Statistical analyses were performed with IBM-SPSS statistics, version 20.0.0 (SPSS Inc., IBM Company, 2010). All data were checked for normality and homogeneity of variance with the Kolmogorov–Smirnov Test and the Levene's Test, respectively. Male and female measurements were compared, using a Student's t test. All statistical tests were performed at 5% significance level. Data are presented as mean values ± standard error (SE).
Results
Measurements of insects
In all the species studied, the measurements show that the thorax width is larger than the head width and, in both cases, they are higher in females (table 2).
Measurements of insect morphology of the insect species studied.
N, number of insets measured.
* Not significantly different (P > 0.05).
1 Insects not separated by sex, due to their small size.
The head width varied between 0.34 ± 0.01 mm in A. fuscicollis and 1.21 ± 0.01 mm in C. carnea and thorax width varied between 0.48 ± 0.01 mm in A. fuscicollis and 1.59 ± 0.08 mm in C. carnea.
Measurements of flowers
Of the 21 plants species studied, eight had flat corollas (no depth), 12 had deep corollas, and one was intermediate, with its nectar being partially hidden in small, cub-shaped flowers (table 3).
Flower species used and floral architecture measurements.
N, number of plants measured.
In those species with exposed floral nectaries, the smallest corolla opening was found in C apsella bursa-pastoris (2.08 ± 12.11 mm) and the largest in Malva neglecta (19.46 ± 0.65 mm).
The corolla depth was between 2.31 ± 0.09 mm in T. repens and 34.09 ± 0.82 mm in Lonicera hispanica, while the corolla width was between 0.42 ± 0.03 mm in Sonchus asper and 5.79 ± 0.29 mm in Echium plantagineum.
R. raphanistrum is the only example where the nectar is partly protected in small cub-shaped flowers (corolla depth: 7.55 ± 0.37 mm and corolla width: 3.07 ± 0.13 mm).
Theoretical nectar accessibility
The combinations of the floral measurements with the insect mouthpart structure and the thorax width of the insect species, allowed conclusions to be drawn on the theoretical nectar accessibility amongst the species considered (table 4).
Theoretical accessibility to floral nectar of the plant species studied by the insect species.
Theoretical access to floral nectar: Y, possible. N, not possible.
1 Proboscis length in the case of P. oleae, head width for the others insects.
2 Male.
Of the selected plants, eight were found to have exposed nectaries, which allow their nectar to be easily accessed by all the studied insects. These comprised three Apiaceae (C onopodium majus, Daucus carota, F. vulgare), one species of Asparagaceae (Asparagus acutifolius), one species of Brassicaceae (C. bursa-pastoris), one species of Caryophyllaceae (S. arvensis), one species of Hypericaceae (H. perforatum) and one species of Malvaceae (M. neglecta).
For P. oleae its relatively short proboscis (1.38 ± 0.02 mm), prevented access to all of the protected and partly protected nectaries. Thus theoretically, P. oleae would not be able to access nectar from 13 out of 21 plant species considered, i.e. Anarrhinum bellidifolium, Andryala integrifolia, Calamintha baetica, Chondrilla juncea, Dittrichia viscosa, E. plantagineum, L. hispanica, L. saxatilis, L. stoechas, R. raphanistrum, S. gallica, S. asper and T. repens. However, the relatively narrow width of the thorax would probably allow it to access the nectar of A. bellidifolium, C. baetica, E. plantagineum, L. saxatilis, L. hispanica, R. raphanistrum, S. gallica and T. repens.
Data from either the head or the thorax width of C. elaeaphilus and A. xanthostigma show that both species can theoretically access the nectar of all the selected plants, with the exception of three Asteraceae (A. integrifolia, C. juncea and S. asper).
A. fuscicollis, due to its small head width, is also theoretically capable to access the nectaries of all plant species studied, but the access to S. asper and C. juncea would be prevented by the A. fuscicollis thorax width.
Regarding E. flabellatus, either the head or the thorax width would prevent it access to S. asper, while its thorax width would prevent it from accessing to A. integrifolia and C. juncea.
The C. carnea head width would prevent access to the nectar of four Asteraceae, A. integrifolia, C. juncea, D. viscosa and S. asper, while the thorax width prevents its access to three other plant species, i.e. A. bellidifolium, L. stoechas and T. repens.
Discussion
The suppression of weeds associated with olive groves, practiced in many regions, negatively affects the performance of pest enemies, and consequently their biological control potential. Hence regional research efforts are needed to screen plants for their potential to conserve beneficial arthropods.
The use of native plant species ensures local adaptation, reduces the risk of one plant becoming particularly weedy or invasive, and increases the potential success of conservation efforts (Alpert et al., Reference Alpert, Bone and Holzapfel2000; Isaacs et al., Reference Isaacs, Tuell, Fiedler, Gardiner and Landis2009). Bugg & Waddington (Reference Bugg and Waddington1994) and (Boller, Reference Boller2001) have observed that the natural ground cover used as ecological infrastructure can have several positive aspects in agroecosystems.
This study, focusing on the theoretical accessibility to P. oleae and six of its most important natural enemies (the parasitoids C. elaeaphilus, A. xanthostigma, A. fuscicollis, E. flabellatus and the predator C. carnea), of nectar in 21 native plants found in the olive grove ecosystem, allowed to include these plants into four groups.
One group consisted of 14 plant species whose nectar is theoretically accessible to all the studied natural enemies as well as to P. oleae. These are, A. acutifolius, C. bursa-pastoris, C. majus, D. carota, F. vulgare, H. perforatum, M. neglecta and S. arvensis, that have an open corolla, as well as E. plantagineum, R. raphanistrum, L. hispanica, S. gallica, C. baetica and L. saxatilis.
A second group consists of three plant species (D. viscosa, L. stoechas and A. integrifolia), which are importance because their nectar is accessible to the parasitoids studied but not to P. oleae. However they have a disadvantage in that C. carnea cannot also access them. This situation has been reported by Van Rijn (Reference Van Rijn2012) for the Asteraceae, who found that the mouthparts of Chrysoperla lacewings (that do not show any adaptation towards feeding on floral nectar) may prohibit entry into the narrow tubular florets of these plants, which are also partly blocked by the anthers.
A third group comprises A. bellidifolium and T. repens. These species while they could well improve the populations of parasitoids they may also benefit P. oleae. Also their nectaries are not accessible to C. carnea. This finding is consistent with the finding of Van Rijn (Reference Van Rijn2012), who stated that this insect cannot apparently use Fabaceae flowers. Further, species of Fabaceae family, such as T. repens, can theoretically provide accessible nectar but only in case where the parasitoids are able to push parts of the corolla (Winkler, Reference Winkler2005).
The fourth group comprises two species, C. juncea and S. asper, whose nectar is not accessible to any of the insects studied, so they offer little potential in terms of increasing populations.
Other studies involving optimizing the value of hoverflies and lacewings, have shown useful to dedicate a substantial part of the field margin vegetation to flowers with well exposed nectaries (depth <1 mm) (Van Rijn, Reference Van Rijn2012). In this study, it was been shown that the same approach could be taken with eight of the species studied, namely A. acutifolius, C. majus, C. bursa-pastoris, D. carota, F. vulgare, H. perforatum, M. neglecta and S. arvensis. However, this group could probably be extended further. Some of the potentially valuable plant species identified here have also been reported in the literature as sources of nectar, pollen, honeydew and even as an alternative food sources for beneficial insects (Heil, Reference Heil2015). These are for example, D. carota and F. vulgare (Bugg & Waddington, Reference Bugg and Waddington1994; Bugg et al., Reference Bugg, Anderson, Thomsen, Chandler, Picket and Bugg1998; Domínguez Gento et al., Reference Domínguez Gento, Roselló Oltra and Aguado Sáez2002; Fiedler et al., Reference Fiedler, Landis and Wratten2008; Sivinski, Reference Sivinski2014). In which concern C. majus, C. bursa-pastoris, R. raphanistrum, S. arvensis and L. saxatilis, because they bloom early, they can play an important role in the early establishment of natural enemy populations (Nentwig, Reference Nentwig, Pickett and Bugg1998). Also Böller et al. (Reference Böller, Häni and Poehling2004) reported that some of these species can be used in mixtures of seeds for ground cover, namely D. carota, F. vulgare, H. perforatum and M. neglecta. As stated by Van Rijn (Reference Van Rijn2012), since flowers with easily accessible nectar are much less common than those with accessible pollen, raising them in or around agricultural fields will likely relieve an important limitation to natural pest control.
It should be noted that, in addition to the accessibility, the concentration of nectar can also prevent the insect to feed it, since due to proboscis morphology, Lepidoptera often have difficulties in imbibing high viscosity nectars (Daniel et al., Reference Daniel, Kingsolver and Meyhofer1989).
Results from this study provide information on the suitability of native olive grove flowering species that can probably enhance conservation biological control of P. oleae. Nineteen of the 21 weed species are potentially able to enhance conservation biological against P. oleae, namely: C. majus, D. carota, F. vulgare, A. acutifolius, A. integrifolia, D. viscosa, E. plantagineum, C. bursa-pastoris, R. raphanistrum, L. hispanica, S. gallica, S. arvensis, T. repens, H. perforatum, C. baetica, L. stoechas, M. neglecta, A. bellidifolium and L. saxatilis. However, further research is required to verify that these plant species do indeed increase the effectiveness of the studied natural enemies. Conversely for plant species whose nectar was found to be accessed by P. oleae, the risk of benefiting the pest should also be carefully evaluated.
Acknowledgements
This study was supported by FEDER Funds throughout Programa Operacional Factores de Competitividade – COMPETE and National Funds throughout FCT – Fundação para a Ciência e Tecnologia, within project PTDC/AGR-AAM/100979/2008 –‘Increasing functional biodiversity in olive groves to enhance conservation biological control of insect pests’, as well as PhD grant, SFRH/BD/34394/2008 granted to the first author. This manuscript is part of A. Nave's Ph.D. dissertation. The authors are grateful to the olive grove owners who permitted access to their groves for this study, with support to Associação de Agricultores para Produção Integrada de Frutos de Montanha. Thanks are due to Köppert Biological Systems that kindly provided Chrysoperla carnea used in this study.
