Antiparasitic Effects of Plant Species from the Diet of Great Bustards

Background: Diets combine food types according to some trade-offs, as for example maximising nutrients and minimising toxins. But some diets include elements because of their activity against the host parasites and other pathogens. This so-called medicinal role of food is under-reported in the literature, either because toxic elements in diets of livestock and wildlife are infrequent, or because their activity against parasites and pathogens has not been fully documented. We contribute to ll this knowledge gap by testing the activity of extracts and essential oils from Papaver rhoeas and Echium plantagineum against a selection of laboratory pathogens. These plants are strongly selected by great bustards Otis tarda during the mating season. Results: During this season we found a signicantly higher frequency of P. rhoeas in male than in female faeces. The activity of different extracts of these plants against some laboratory models including a agellated protozoan (Trichomonas gallinae), a nematode (Meloidogyne javanica) and a fungus (Aspergillus niger) was evaluated. We found activity against nematodes and trichomonads in non-polar and polar extracts of the aerial parts of P. rhoeas, especially the extracts of owers and capsules, and E. plantagineum, especially the extracts of leaves and owers. Conclusions: Both plants showed anti-parasitic activity, a result compatible with the hypothesis that great bustards eat plants for non-nutritional purposes, likely to assist them in coping with parasites and other pathogens, and P. rhoeas could be especially helpful for males during the mating season, when their immune system is weakened by the investment in secondary sexual characters and sexual display. The self-medication properties of plants and animals included in diets should be considered in studies of foraging behaviour, habitat selection, and even conservation biology of wildlife.

In previous studies [9,14] we investigated the antiparasitic and antimicrobial effects of blister beetles (Meloidae), which are avoided by most animals but consumed by great bustards. Blister beetles contain cantharidin, a highly toxic monoterpene that could have positive effects in controlling the parasite load of the host [9]. We found that great bustards males consumed more blister beetles than females, which suggested that males could use cantharidin to reduce their parasite load and increase their sexual attractiveness [14].
Although great bustards eat beetles and other invertebrates in spring and summer, their diet is mostly vegetarian over the whole year, including mainly leaves of green plants [15][16][17]. Some of these plants could also have properties that might help great bustards control their parasite load. For example, Lane et al. (1999) found that most weeds were consumed in proportions expected by their abundance, but noted two exceptions, the common poppy Papaver rhoeas and the purple viper's bugloss Echium plantagineum. The diet of great bustards showed much higher foraging ratios [18] of these two plants in April compared to other months (Fig. 1). Since April is the month when males reach the peak of display activity and most matings occur [19,20], we hypothesize that consumption of these two plants could help them reduce the negative effects of parasites during this important phase of their reproductive cycle, as is the case with blister beetles. Males could bene t from the medicinal properties of these plants during a period when they are subjected to high stress and reduced immune resistance to infections due to their strenuous investment in sexual display [19][20][21][22]. Parasite over-loads may indeed affect the tness or reproductive status of the host [23,24], and chemical compounds helping to control parasites when the threat of parasitism is greater may have bene ts on health and reproductive success [25,26].
To test the hypothesis, our present study had the following objectives. First, we examined whether consumption of P. rhoeas and E. plantagineum is male-biased. A higher proportion of these plants in the diet of males would support the hypothesis that males use them to reduce their parasite load during their strenuous display season, in a similar way as that suggested for cantharidin [14]. Alternatively, or complementary to a self-medication function affecting only males, both plants might also have a prophylactic function for females during the peak copulation period in April when exposure to sexually transmitted diseases is highes. Second, we investigated whether extracts and essential oils (EOs) from P. rhoeas and E. plantagineum have antiparasitic activity against the agellated protozoa Trichomonas gallinae, the endoparasitic nematode Meloidogyne javanica, and the fungus Aspergillus niger. Third, we analysed the chemical composition of extracts to identify what components could be active against pathogens. The ful lment of these three objectives is not a demonstration of self-medication in Great bustards, but a necessary step towards it. A step that contributes to expand the frontiers of zoology.
The extracts were also tested on an endoparasite nematode model (M. javanica). The results are shown in Table 4. Freeze dried ower infusions from both plant species demonstrated strong nematicidal effects (up to 5% dilutions). Among organic extracts, the MeOH ower extract of E. plantagineum was also active against M. javanica (81.3%).

Cowlteropine
In short, P. rhoeas contains antiprotozoal and anthelmintic compounds, while E. plantagineum contains components with antiprotozoal, anthelmintic and, to a lesser extent, antifungal activity.

Polar extracts (EtOH / MeOH, infusion):
The HPLC-MS analysis of the polar extracts (leaves ethanolic, L-EtOH; owers methanolic F-EtOH; ower infusion lyophilized, F-IFD; dichloromethane extract of ower infusion, F-IDCM) are shown in Tables 5A (P. rhoeas) and 5B (E. plantagineum). To categorize the extracts, the peaks have been arbitrarily grouped as polar and medium polarity compounds according to their retention times (1-10 min and 10-50 min). The alkaloidal components identi ed in these extracts eluted between 10-50 min as part of the medium polarity class.
Among P. rhoeas extracts, the freeze-dried infusion of owers contained a high percentage of polar components (92%). The alcoholic extracts had similar amounts of polar (L-EtOH and F-MeOH, 69 − 68%) and medium polarity peaks (21 − 18%), with more alkaloids identi ed in the owers (14%). However, most of the identi ed alkaloids were concentrated in the F-IFD partitioned with dichloromethane (F-IDCM, 67%, see Table 5A).

Discussion
We found antiparasitic effects in Papaver rhoeas and Echium plantagineum, two plants strongly selected by great bustards during the mating season. From these plants we obtained two categories of active extracts: apolar (EO and Hex) and polar (alcoholic: EtOH / MeOH and aqueous infusion: IFD). We determined their chemical composition and tested their activity on a sample of common laboratory pathogens. Based on these results and a review of the antipathogenic properties of both plants, we infer that great bustards feed on them to reduce their load of pathogens.

Papaver rhoeas
Non-polar and polar extracts of aerial parts of P. rhoeas, especially owers and capsules, showed activity against nematodes and trichomonads. Among non-polar extracts, the ower essential oil (EO) and the capsule hexane extract showed strong trichomonacidal effects, the effect of EO being particularly powerful. A study of P. rhoeas EO from aerial parts collected in Turkey showed phytol (52.8%), tricosane (7.8%), 2-pentadecanone (6%) and heneicosane (5.3%) as the major compounds [32], while the EO studied here was characterized by alkanes such as n-hexatriacontane and 1-eicosanol and the fatty acid palmitic acid. The non-polar extract (Hex) of the capsules mostly contained methyl / ethyl linoleate, 1-eicosanol and palmitic acid. Palmitic acid has been reported as being the main component of non-polar (Hex) extracts of P. rhoeas leaves [33].
Previous results have shown trichomonacidal effects on T. vaginalis of Nigella sativa seed oil containing fatty acids [34]. However, this is the rst report on trichomonacidal effects of P. rhoeas lipid extracts.
Trichomonads are unable to biosynthesize fatty acids and cholesterol but can incorporate these compounds from the medium without further modi cation [35]. Therefore, externally supplied fatty acids and sterols could interfere with Trichomonads lipid metabolism.
The bioactivity of P. rhoeas polar extracts correlated with their content in polar compounds (peaks with retention times of < 10 min). The alcoholic ones (L-EtOH, F-MeOH ) showed trichomonacidal effects, while the F-IFD showed potent trichomonacidal and nematicidal effects. Nematicidal activity has been previously reported for aqueous extracts (4%) of P. roheas leaves against M. javanica [36]. These extracts contained molecular ions compatible with reported Papaver alkaloids [27], mostly concentrated in the dichloromethane partition of the ower infusion (F-IDCM), with rheagenine / criptopine, rhoeadine and hydroxy-N-methyl-coclaurine ions being the most abundant in the F-IDCM, F-MeOH and L-EtOH extracts.
Since the alkaloid-rich F-IDCM partition was not active, the trichomonacidal /nematicidal effects of P. rhoeas polar extracts cannot be attributed to these alkaloids, indicating that more polar components of the extracts are responsible for the observed effects. Among P. rhoeas components, alkaloids are the most representative, especially (+)-rhoeadine, along with N-methylasimilobine, rhoeagenine, epiberberine and canadine, depending on the plant origin [37]. Minor alkaloids included roemerine [37], with reported antibacterial, antifungal and anthelmintic activities [38,39]. Furthermore, alkaloids such as allocryptopine, potopine and berberine were nematicidal against Strongyloides stercolaris larvae [40]. P.
rhoeas also contains avonoids, phenols, organic acids and vitamin C [41][42][43]. Flavonoids may reduce the oxidative stress and enhance immunity, so they are selected by different bird species, presumably as a prophylactic drug [44] against pathogens. Polyphenols regulate immune and in ammatory responses during enteric bacterial and parasitic infections in livestock [45], and organic acids can signi cantly reduce microbial contamination in turkeys [46].
Corn poppy has been used since ancient times as a food ingredient and traditional remedy [37], but cases of poisoning with P. rhoeas in adults, children and animals have been described [47,48]. Poppy poisoning in humans can cause nausea, vomiting, altered mental state, headache, convulsions, miotic pupils, lethargy and disorientation [49]. Papaver species are actively toxic or narcotic and unpalatable to grazing animals. Animals are safe since the odour and taste of the plants render them obnoxious but there are reports of cattle and horses being poisoned by P. rhoeas [47]. Nonetheless, great bustards include Papaver rhoeas in their diet throughout most of the year and, to our knowledge, they are not poisoned by corn poppies. The diet composition of great bustards and the activity of P. rhoeas extracts shown here support the hypothesis of a self-medication function for this plant species during the bird's mating season [14].

Echium plantagineum
Extracts of the aerial parts of E. plantagineum (non-polar and polar) also showed trichomonacidal and nematicidal effects. The ower essential oil (EO) of E. plantagineum, with moderate trichomonacidal effects, was characterized by alkanes such as n-hexatriacontane and related substances, and the fatty acid ester 2-monopalmitin.
Similarly to P. rhoeas extracts, the bioactivity of E. plantagineum polar extracts also correlated with their content in high polarity compounds. The alcoholic F-MeOH showed trichomonacidal effects, while the freeze-dried infusion (IFD) showed potent trichomonacidal and nematicidal effects. Several molecular ions compatible with Pas reported in E. plantagineum [28][29][30][31] were identi ed in the polar extracts, mostly concentrated in the organic fraction of the owers' infusion (IDCM). Echimidine molecular ions were the most abundant in all extracts. Since the PA-rich IDCM partition was not active, the trichomonacidal /nematicidal effects of E. plantagineum polar extracts cannot be attributed to these alkaloids. However, nematicidal effects of PAs on Meloidogyne incognita have been reported for heliotrine, lasiocarpine and monocrotaline, but these effects were dependent on the PA structure and the exposure period (168 h) [50].
E. plantagineum produces different classes of secondary metabolites, including pyrrolizidine alkaloids (PA) in the aerial parts and seeds [29,51,52], echimidine and echiumine N-oxide being particularly abundant [30]. Pas are easily reduced to free bases and are metabolized by the herbivorous cytochrome P-450 oxidases, which give rise to pyrrol alkylating intermediates. Reactive pyrroles damage cellular DNA and are dangerous to cattle, horses, sheep, pigs, and rats, affecting also humans [53][54][55][56]. Harmful effects on bird health have also been described as a result of PA consumption [57,58].

Diet and health of great bustards
In this study we suggest that selection of P. rhoeas and E. plantagineum by great bustards could be based on the antipathogenic effects of these plants. The use of plants with active secondary metabolites for preventing or reducing parasite and pathogen loads (self-medication) has been described in invertebrates [59][60][61], mammals [62,63] and birds [64][65][66]. As for great bustards, Bravo et al. [14] described for the rst time a probable case of self-medication by ingestion of toxic insects. They found that bustards included two blister beetles of the family Meloidae in their diet. These beetles contain cantharidin, a highly toxic compound that can be even lethal for bustards if ingested in high doses [67]. Bravo et al. [14] found a male-biased consumption of blíster beetles, and interpreted it as a way to enhance their attractiveness to females by reducing their parasite load. Before selecting a mate for copulation, a female bustard carefully examines the cloaca of the displaying male and usually pecks it, probably looking for parasites. Bravo et al. [14] suggested that a higher consumption of blister beetles by males could be a sexually-selected mechanism to enhance their mating success. The hypothesis of self-medication in bustards was supported by Withman et al. [9], who demonstrated antiparasitic effects of extracts from blister beetles (Berberomeloe majalis) against different models (protozoan, nematodes, ticks and insects). Heneberg [68] proposed that bustard males self-medicated seeking sexual arousal rather than antipathogenic effects.
Regardless of the ultimate function of blister beetle selection, here we propose that there are more species with similar properties in the diet of great bustardst, and we present the antiparasitic results for two plant species as an example. Although the toxicity of P. rhoeas and E. plantagineum differs, great bustards show a marked selection for both plants during the mating season (Fig. 1), and during that season we found higher amounts of P. rhoeas in male than in female fecal samples (Table 1). Why could males be more interested in this plant than females and why during the mating season? Courtship is strenuous for males in most polygynous species and particularly in great bustard males, who show the most strongly skewed mating success reported among lekking birds, suggesting an extreme intensity of sexual selection in this species [20]. Males develop costly ornaments every spring and perform exhausting displays to attract females [19,20]. It is known that physiological investment in sexually selected characters competes with investment in immune response [69], a trade-off which is not as demanding for females, with the consequence of smaller loads of parasites and pathogens in this sex compared to males [70]. Great bustard males would hold higher load of parasites and other pathogens than females and still sire a number of descendants in the next generation if females chose to mate them, according to the Handicap Principle [71,72]. Attracting females while keeping pathogens may be quite demanding, so polar components in P. rhoeas' capsules and owers would help males control pathogens, reduce fatigue [sensu 68], or both. Measuring these effects in vivo is beyond any feasible experimental setup, at least with current designs and legal restrictions on experimenting with this vulnerable species. But inferring the causal links is reasonable, so we put forward the challenging hypothesis that great bustard males disproportionately foraged on P. rhoeas during the mating season due to the effects of some nonnutritional compounds present in this plant.
Activity of E. plantagineum against pathogens tested in the present study was noticeably higher than that of P. rhoeas, but the proportion of Echium plantagineum dry weight in faeces of great bustards was about 50 times smaller than that of P. rhoeas during the mating season, and seven times smaller in the nonmating season. The harmful effects of pyrrolizidine alkaloids on bird health described in previous studies [57,58] could explain the small amount of E. palntagineum in great bustard faeces, but not its higher proportional dry weight in males compared to females during the non-mating season, unless we admit that males would have a greater need of these compounds than females also outside the mating season. We cannot discard that males would also bene t more than females from the properties of E. plantagineum during the months we de ne as non-mating season (November-January, and July, see [17]). Males indeed start displaying and ghting to establish their dominance hierarchy in the male group in December-January [73], so these winter months could also be highly energy demanding. As for July, this is the hottest month of the year, when males suffer the debilitating effects of a much lower heat resistance compared to females [74], coinciding with the moult of the ight feathers.

Conclusions
In sum, we have shown that two plants consumed by great bustards are ingested in higher proportions by males, probably to cope with their higher stress levels due to their demanding display effort, and that both plants show activity against pathogen models in laboratory conditions. These results support to some extent the self-medication hypothesis [75] in these birds. Nonetheless, differentiating between a possible indirect effect of the ingestion of species rich in secondary metabolites or the consumption for their nutritional value is one of the di culties in the interpretation of self-medication in animals [76]. For example, alkaloids and other phytochemicals, in low doses and in appropriate mixtures, can be helpful for the health of some animals. Often, the limits between nutrients, drugs and toxins are determined by the dose ingested [77]. A recent hypothesis on self-medication behaviour in animals considers the increase in the consumption of species already present in the usual diet [78], in contrast to the traditional hypothesis that considered the ingestion of speci c compounds by animals only in response to infection [76].
Detailed eld or laboratory observations are required to shed light about this possibility, buth such studies will represent a di cult if not impossible task with great bustards due to legal restrictions on their capture and handling, and problems to keep them as experimental animals in captivity.

Plant collection
Flowering Papaver rhoeas and Echium plantagineum plants were collected in May 2019 during the mating season, near Valdetorres del Jarama (Spain) and at one of the largest great bustard leks of central Spain (Fig. 2, UTM: 40.708987, -3.495657) [79][80][81][82]. These plant species were chosen based on their positive selection by great bustards (Fig. 1) and their potential toxicity as reported in the literature Plant extracts and essential oils The plant parts were separated ( owers, leaves and capsules in the case of P. rhoeas), and dried at 40 °C during 48 h. Ground (40 g) owers were extracted with methanol; and 50 and 30 g of ground leaves and capsules were sequentially extracted with hexane, ethyl acetate and ethanol in a Soxhlet apparatus. The solvents were evaporated under vacuum to give dry extracts.
Essential oils were obtained by hydrodistillation for 2 h in a Clevenger-type apparatus according to the European Pharmacopoeia. Ground owers (100 g) were distilled with 2 l of water. The residual water (infusion) was freeze-dried (100 ml) or subjected to a liquid-liquid extraction (150 ml) with dichloromethane (DCM) (l50 ml x 3 times).
Antiprotozoal activity T. gallinae trophozoites isolated from a common wood pigeon (Columba palumbus) from Central Spain were employed to determine the antiprotozoal activity. The strain was maintained by serial passes in 10 ml tubes with TYM medium. The tests were carried out 48 hours after a serial pass when trophozoites were at exponential growth phase. One hundred and fty µl of a culture containing 500,000 trophozoites/ml were placed in each well of microtiter at-bottom plates. Extracts and EOs were tested at several concentrations (400, 200, and 100 µg/ml) dissolved in 2 µl DMSO. After 24 h incubation at 37 °C the antiprotozoal activity was obtained by the MTT colourimetric method. Brie y, plates were centrifuged and the TYM was carefully removed. Ten µL of MTT/PMS (1.25 mg/ml; 0.1 mg PMS) were added to each well. After incubation for 15 min to occur the reduction of MTT, 50 µl sodium dodecyl sulphate (10 g SDS; 31,5 µl HCL; 100 ml distilled water) were added to dissolve formazan crystals obtained as a result of the reduction of MTT. Once the crystals had dissolved (15-30 min), the plate was read on a spectrophotometer at 620 nm. The activity was calculated as percentage growth inhibition (%GI) as follows: where At is the absorbance of treated wells, Ac the absorbance of control wells (not treated) and Ab the absorbance of blank wells (culture medium and vehicle only). All assays were carried out in triplicate and were repeated at least three times independently to con rm the results.

In vitro nematicidal effects
The nematodes used to determine anthelminthic activity came from a population of Meloidogyne javanica maintained on Solanum lycopersicum plants (var. Marmande) in pot cultures at 25ºC and 70% RH. The bioassays were carried out as described by Andrés et al. [83]. Freeze-dried infusions were dissolved in distilled water at concentrations adjusted to water-infusion yield (27 and 8 mg/ml for P. rhoeas and E. plantagineum). Nematode inoculum (500 J2 in water) was ltered (25 µm), and the nematodes were suspended in 500 µL of lyophilized infusion treatments. Four aliquots (100 µL) of the nematode suspension (approximately 100 J2) and controls (water) were placed in 96-well plates. Organic extracts treatments were prepared by dilution in a DMSO-Tween solution (0.5% Tween 20 in DMSO) at 20 mg/ml and 5 µL of this solution were added to 95 µL of water containing 90-150 nematodes ( nal concentration of 1 mg/mL and 5% DMSO). Treatments were replicated four times. As a control, four wells were treated with the water/DMSO/Tween 20 in the same volume as the tests. The plates were covered to prevent evaporation and were maintained in the dark at 25 °C. After 24, 48, and 72 h, the dead J2 were counted under a binocular microscope. The nematicidal activity data are presented as percent dead J2s corrected according to Schneider-Orelli's formula [84].

Antifungal activity
Aspergillus niger (donated by Dr. K. Leiss, Wageningen University) was used to determine de antifungal activity [85].

Spore growth inhibition
To obtain the spores, 5 ml of saline solution were added in three-day PDA Petri dish of A. niger culture. The spores were resuspended in saline solution with a sterile swab and ltered to remove the mycelium.

Statistical analyses of plants abundance in fecal samples
The occurrence of Papaver rhoeas and Echium plantagineum was calculated as the percentage of the total dry weight in faeces of males and females in a previous study [17]. The sex effect in diet was calculated in that study a whole, including all plant species. Here we calculated generalized linear mixed models (GLMMs, binomial error, logit link function) for P. rhoeas and E. plantagineum consumption by great bustards, which is a new statistical analysis. Fecal samples were collected at nine different sites within the study area [17], but in the present study they were grouped for analysis in just two seasons: mating (April) and non-mating (November -January and July, respectively pre-mating and post-mating in [17]). Since the nine sites differed in availability of P. rhoeas and E. plantagineum, collecting site was included as a random factor in GLMMs. Likelihood-ratio tests were used to assess the signi cance effect of sex. Sample sizes were 81 female and 97 male faeces during the mating season, and 222 female and 223 male faeces during the non-mating season. We tested the effect of sex on the proportional consumption of each plant in fecal samples. The GLMM was repeated in the non-mating season as a complementary control of the main hypothesis: during mating males could eat both plants in higher proportions than females to counteract their higher parasite load due to stress, whereas in the non-mating season this sex difference could be smaller or even statistically non signi cant.  Monthly foraging ratios [18] of corn poppy (P. rhoeas, open circles) and purple viper's bugloss (E. plantagineum, lled circles) peaked in April, coinciding with maximum mating activity of great bustards (vertical bars, right axis). Figure composed with plant consumption and availability [16], and weekly mating attempts [86]. Vertical bars represent the number of mating attempts, de ned as males fulldisplaying in very close proximity to one or more females (<3 m) with at least one of them showing obvious precopulatory behavior, independently of whether these attempts were followed by effective copulations or not. The number of mating attempts is a reliable indicator of mating success, as indicated by the signi cant correlation between rates of effective copulations and copulation attempts [19].