Non-choice vs. Free-choice: when and how Cydnoseius negevi (Swirski and Amitai) female attack its prey?

Background Cydnoseius negevi (Swirski and Amitai) (Acari: Phytoseiidae) is a promising bio-agent in the Mediterranean region. It attacks a wide range of pest species. Two fundamental questions have been raised when C. negevi was a promising candidate for biological control applications in a hydroponic system. This study was carried out to answer the raised questions: (1) how long a predatory female takes to attack its prey in a non-choice biological control application? and (2) what do the predatory females prefer when offered a multiple choice (free-choice) of pest preys?, to understand the potentialities of C. negevi attacking behaviour. Results Two sets of experiments were conducted for 240 min under laboratory conditions of 27 ± 2 °C, 60 ± 5% RH, and 18:6h L:D photoperiod. The tested prey species were Tetranychus urticae Koch, Bemisia tabaci (Gennadius), Frankliniella occidentalis (Pergande), Parlatoria ziziphi (Lucas), and Icerya aegyptiaca (Douglas). Non-choice experiments tested the time (min) needed to the 24 h starved female to attack its prey. The free-choice experiments tested the female’ choices and the time recorded for each female during the full experimental time. Significant differences were recorded in both sets at confidence level of 95%. The spider mite T. urticae recorded the shortest predatory female’s response in both non-and free-choice sets. On the contrary, the insect I. aegyptiaca had the least female’s choice in both sets of experiment.


Background
Prey choice behaviour is considered as one of the theoretical keys that support the knowledge of predator-prey interaction (Abrams 2000).Predators act in a way that maximises their energy intake by preferentially preying on the most beneficial resources (Samaras et al. 2021).The nutritional value of the prey and how it is easy to be consumed are two attributes of the prey that affect a predator's preference (Zhang et al. 2022).Predation on prey of the highest nutritive value boosts the predator's fitness (increased survival and reproduction rates in a short time) (Dicke et al. 1990).
Prey selection does not happen randomly among available species (Ishii and Shimada 2010).The tendency of a predator to choose a given prey type over another may change as the relative frequencies of the prey species, sizes, ages, or sex classes in the predator's environment change (Whitney et al. 2018).Rahmani et al. (2016) concluded that in the first and second nymphal stages of Neoseiulus californicus (McGregor), when given the option to choose between the first larval instar of T. tabaci and immature of T. urticae, they clearly preferred the second option; in the presence of T. urticae females, N. californicus stages directly switched their preference to T. tabaci.
Also, Alatawi et al. (2018) reported that C. negevi, when reared on date palm pollen, preferred and consumed juveniles of O. afrasiaticus more than T. urticae.However, the field race of C. negevi did not show significant feeding preference between both species of spider mites in the choice test.The observed and expected rates of predation in mixed food species varied significantly, and the observed differences cannot be attributed to behavioural changes in the prey species as a result of being together.Consequently, it likely appears that the predator's behaviour in selecting prey differed depending on whether the prey species was presented in groups or alone (Alatawi et al. 2018).Switching among preys is an important aspect in defining a predator's individual fitness since it ultimately impacts hunting success and a predator's ability to gain and keep body reserves (Jaworski et al. 2013).Indeed, through prey choice, predators can alter the distribution of life history features over generations and impact the evolution of their prey.
It has been hypothetically asked: When several prey species are presented together, will the predator's preference change?Therefore, this study was targeted to answer two raised questions: (1) how long a predatory female takes to attack its prey in non-choice biological control application?and ( 2) what do predatory females prefer when offered a multiple choice (free-choice) of pest preys?, in order to be a prospective candidate for biological control applications in the hydroponic system.

Predatory species collection
Cydnoseius negevi individuals were extracted from lowgrowing weeds associated with the hydroponic farm at Gamaiet Ahmed Orabi.Five females were collected after extraction and mounted on Hoyer's medium for microscopic identification according to the most recent keys of the Egyptian phytoseiid mites by Abo-Shnaf and de Moraes (2014).Fifteen specimens were randomly collected from the stock colony and have been taxonomically confirmed by Dr. Ismail Döker, Çukurova University, Turkey.

Stock colonies
Fifty gravid C. negevi females and 20-25 active males were transferred by fine brush into five arenas/colonies of 12 cm in diameter (cm Ø) upside-down acalypha, Acalypha marginata L. (Euphorbiaceae) in 15 cm Ø Petri dish with wet-saturated cotton pad at the Acarology laboratory, National Research Centre (NRC, Egypt).Colonies of C. negevi were reared on fresh pollen grains of castor bean, Ricinus communis L. (Euphorbiaceae), provided every 3 days.Culture maintenance, cleaning, and changing were carried out weekly on fresh leaves of acalypha.Laboratory conditions were 27° ± 2 °C, 60 ± 5% RH, and 16:8 photoperiod.Cultures were established for five weeks before starting the experimental trials.

Pest species
Active stages of the two-spotted spider mite T. urticae were collected from heavily infested L. sativa var.capitata and B. oleracea var.palmifolia leaves.Eggs and crawlers of the tomato whitefly, B. tabaci, infested on B. oleracea var.palmifolia leaves.The first larval stage of the western flower thrips, F. occidentalis, infested on leaves of L. sativa var.capitata and B. oleracea var.palmifolia.Eggs and crawlers of the black parlatoria scale insect Parlatoria ziziphi (Lucas) (Insecta: Hemiptera: Diaspididae) infested on lemon leaves, Citrus limon (L.) (Rutaceae).Also, egg clusters of the Egyptian fluted scale, Icerya aegyptiaca (Douglas) (Insecta: Hemiptera: Monophlebidae), were collected from heavily infested lemon seedlings and hibiscus shrubs, Hibiscus rosa-sinensis L. (Malvaceae), found in the associated biodiversity around the same hydroponic location.

Experimental procedures
Trial sets were conducted to estimate the females' attacking time and aggressiveness when offered (1) non-choice and (2) free-choice natural preys.The methodologies of Bazgir et al. (2020) were followed with some modifications.Lettuce, kale seedlings and leaves of lemon and hibiscus were sampled twice weekly to collect the selected pest stages.For T. urticae, 20 gravid females were transferred by a fine brush onto fresh lettuce leaves and kept under laboratory conditions for egg depositing.After egg hatching, larvae were kept for feeding and moulting.The newly emerged protonymphs were transferred individually onto experimental arenas (3 cm Ø).
The 1st larval instar of F. occidentalis was directly collected from infested lettuce and kale seedlings and set as prey in the experimental trials.Eggs and crawlers of P. ziziphi were removed from lemon leaves by a fine needle.A disc of ½ cm 2 of B. oleracea var.palmifolia leaves containing circle-like cluster of B. tabaci eggs and crawlers was cut with a sharp cutter and placed onto experimental arenas.The egg clusters of I. aegyptiaca were removed softly by a fine brush from lemon leaves, then separated individually in wet filter paper, and transferred into the experimental arenas.

Predatory female
Fifty deutonymph individuals of the predatory mites were collected from mite colony and transferred to 7 cm Ø upside-down lettuce, L. sativa var.capitata, leaf discs in 10 cm Petri dish with wet-saturated cotton pad with excessive pollen of R. communis under the previous mentioned laboratory conditions.After maturity, predatory females were mated with newly emerged males.Twenty 5-day-old gravid females were transferred individually into the experimental arenas (3 cm Ø lettuce leaf discs) and starved for 24 h.After starvation period, two sets of experiments have been conducted: (i) Non-choice test: Each pest species was introduced individually to the 24h starved predatory female (n = 20 females × 5 pest species).(ii) Free-choice test: five individuals from each tested pest species were introduced to 24 h starved females (n = 20).
Both sets were kept for 240 min.The experimental time was divided into 8 phases/intervals: 0-15 min, 16-30 min, 31-45 min, 46-60 min, 61-90 min, 91-120 min, 121-180 min, and 181-240 min.The first phase started after loading the prey into the experimental arena.In the free-choice trials, the time whenever C. negevi female took to attack and consume its first choice and the other later choices during the experimental phases were recorded (Fig. 1).

Statistical analysis
Data normal distribution was tested using Kolmogorov-Smirnov and Shapiro-Wilk at 95% confidence level.Preliminary statistical tests were performed to experiment the normal distribution using Leven's test using an IBM SPSS computer program v.26 (SPSS 2019).Data were arranged and grouped using Microsoft office Excel ver.2019 for each single prey species in both non-and freechoice tests (attacking time/min).Data were not transformed.Figures and data visualisation were performed using SPSS (SPSS 2019).Mean comparison of predatory females data in both non-choice (attacking time of prey/ min) and free-choice (attacking time of x preys/min) tests was made and analysed using one-way analysis of variance (ANOVA).Mean differences were compared for all experiments with Duncan's test at 95% confidence level using SPSS (SPSS 2019).

Responses of C. negevi female to the non-choice prey preference
Only T. urticae recorded non-significant F-value due to C. negevi's rapid response which was recorded in the 1st phase, where 100% of the females attacked T. urticae protonymphs in less than 5 min.During the 1st phase, 45% of C. negevi tested females attacked B. tabaci as a second choice, then 15% attacked F. occidentalis, while 10% were attacked both I. aegyptiaca and P. ziziphi (Table 1).
Equally in the fourth (46-60 min) and fifth (61-90 min) phases, 40% of the reached females attacked the 1st larval instars of F. occidentalis, then 20% attacked I. aegyptiaca, and 15% attacked P. ziziphi.After that, in the Within columns, small bold italic similar letters are not significant at P ≤ 0.05 (Duncan's test of homogeneous subsets).Within rows, similar capital letters are not significant difference at confidence level of 95% (Duncan's test of homogeneous subsets).ANOVA * significant, *** highly significant, and ns non-significant at confidence level of 95% (P ≤ 0.05)  1).
Figure 2 illustrates the significant responses of C. negevi females when kept with each pest species individually during the non-choice experimental duration (240 min).The first choice was T. urticae with an average time about 3.30 min, the second choice was B. tabaci (average of 11.60 min), the third choice was P. ziziphi (average of 40.20 min), the fourth choice was I. aegyptiaca (average of 51.75 min), while the last choice was F. occidentalis (average of 70.30 min) all at P ≤ 0.05.
Figure 3 illustrates the relationship between the experimental time and C. negevi female's free-choice preferences at P ≤ 0.05.In the 1st phase (0-15 min), C. negevi's first choice was T. urticae, then B. tabaci and P. ziziphi were equally the second choice, F. occidentalis was the third choice, and the last one was I. aegyptiaca (Fig. 3a).In the 2nd phase (16-30 min), the first choice was F. occidentalis, then I. aegyptiaca was the second choice, B. tabaci was the third, P. ziziphi was the fourth, and T. urticae was the fifth choice (Fig. 3b).In the 3rd phase (31-45 min), C. negevi's first choice was T. urticae, the second was F. occidentalis, while B. tabaci and P. ziziphi were equally the third choice, and the fourth choice was I. aegyptiaca (Fig. 3c).

Discussion
The present results showed that C. negevi attacking behaviour has been categorised in seven stages: (1) inspection, (2) patrol, (3) invasion, (4) prey penetration, (5) searching, (6) resting, and (7) grooming (Mirza et al. 2018).Accordingly, C. negevi female showed an aggressive behaviour towards the spider mite T. urticae protonymphs in both non-and free-choice experiments.It has taken few minutes (3.30 ± 0.52 in non-choice and 6.73 ± 0.81 in free-choice sets) to inspect its prey.During this period, predatory females went in circles patrolling inside the experimental arena before prey invasion and penetration.Functionally, C. negevi's handling time (T h ) and attacking rate (α) were clarified to different stages of T. urticae (Rasmy et al. 2014) and O. afrasiaticus (Alatawi et al. 2018).Concerning T h analysis, which is the time (in hours) for inspecting an individual prey, investigating and invading the prey as well as the time spent in penetrating the prey body.Besides, α describes the predator's response distance (patrolling behaviour), the predator-prey chasing and escaping rapidity, and the quantity of predatory female's successful attacks (Sulek et al. 2023).Together, these functional values are covering the response categories of the predatory female.Consequently, T. urticae protonymps were significantly preferable than other choices in the free-choice experiments.Due to its fragile bodies, and having less anti-predator behaviour than the adult stages (Dittmann 2015).In addition, Fagan et al. (2002) stated that the body softness of preys was a recommended option by the predatory female, as an easy source for nutrients and water supply, which helped retaining the needed resource after 24h starvation either for survival or egg production (Sulek et al. 2023).Moreover, data showed there was no much difference in C. negevi's selection during the 1st experimental phase (0-15 min).B. tabaci and P. ziziphi recorded the second female's choice in the non-and/or free-choice experiments.Accordingly, C. negevi fed on eggs of B. tabaci, its immature stages could successfully survive, and females recorded high significant life table parameters (Elkholy et al. 2022).Additionally, C. negevi females recoded higher preference to B. tabaci eggs (as an extra-guild prey) than the offered intra-guild preys of C. negevi larvae and/or A. swirskii eggs (Momen et al. 2013).The recent results hypothesised that learning plays an important role in the genetic recognition of suitable preys (Ishii and Shimada 2010).There were no many data about the biological potentialities and life table parameters of C. negevi when offered coccoid insects, as these species considered a minor importance for phytoseiid predators (Gerson et al. 1990).However, some indigenous species such as E. scutalis (El-Sharabasy et al. 2017) and N. zaheri (Rasmy et al. 2003) were reported as successfully survive, reproduce, and consume P. ziziphi eggs (and/or eggs + crawlers).Additionally, it was proved that mealybugs have low nutritional values, which had negatively affected the predatory mite A. swirskii potentialities (Ragusa and Swirskii 1977).On the contrary, E. scutalis recorded high life table parameters when it offered preys such as I. purchase as mentioned by Yousef and El-Halawany (1982) and I. aegyptiaca by Momen and Abdel-Khalek (2008).
It is an assumption that after the post-penetration phases of resting and grooming, C. negevi females were went to make another round to get fullness.It has been observed that C. negevi females aggressively scanned the arena during the inspection and patrolling phases.In order to build a chemo-physical response of the targeted prey, the predatory mite uses its fore legs before making a decision of attack and penetration.For searching, female mite used its chemoreceptors on its body (Su et al. 2021).The present observations confirm Mirza et al. (2018), who mentioned C. negevi behaviour during resting and grooming: when predatory females were cleaning their tarsus using chelicerae and palps, it can be stated that this action may occur as a learning behaviour of the outer habitat.Hypothetically, it may be proposed that C. negevi performs searching images to the spatial properties of prey species while searching, resting, and grooming phases (Ishii and Shimada 2010).In this regard, it has been recorded the priorities of C. negevi female: when offered non-and/or free-choice prey, it put I. aegyptiaca in the 'less important' image.Due to its lower nutritional value (Gerson et al. 1990), additionally, the 1st larval body has very long legs and antennae (Mani and Shivaraju 2016).
Nevertheless, F. occidintalis larvae forced the predatory female to get backward to the grooming phase and retrying the process after a while.However, F. occidintalis was a targeted prey for C. negevi in biological control applications in field (Sanad and Hassan 2019).Unfortunately, there were no data available on C. negevi's functional and/or numerical responses to thrips pests.However, the current observations showed an anti-predator behaviour of F. occidintalis 1st larval instars when kept together with a starved female of C. negevi in both experimental sets.The larva of the prey physically defended itself by swinging its body ends aggressively, when being attacked to hit the attacking predatory female.This physical counter-attack behaviour was reported with F. occidentalis and T. tabaci and was clarified by Beretta et al. (2024).The counter-attack of F. occidintalis larvae forced the predatory female get backward to the grooming phase and retrying the process after a while.As a result, escape predation extends the inspection and patrolling durations.Therefore, to delay the invasion and penetration phases up to an average of 70 min (in non-choice) and ~ 80-85 min in free-choice sets (current study), the behaviour aspects of C. negevi need more investigations to completely understand its potentialities as a prospective bio-agent in the biological control application.Additionally, being a common species in the Mediterranean area makes C. negevi as a prospective candidate in different ecosystems such as the hydroponic system.

Conclusion
The phytoseiid species C. negevi recorded significant positive responses to both non-and free-choice tests.The 24 h starved female arranged its choices as images of priorities, as follows: (1) T. urticae, (2) B. tabaci + P. ziziphi, (3) F. occidintalis, and (4) I. aegyptiaca.This list of prey images indicated the aggressiveness of C. negevi as a generalist predator, in the meantime, recommending it as a promising bio-agent in biological control programme on a wide scale.Further behavioural studies need to be performed to fully understand its responses to several stresses.

Table 2
Mean time recorded (min ± SE) of C. negevi females when offered free-choice prey species during the experimental duration of 240 min a Post hoc tests are not performed for F. occidentalis and I. aegyptiaca because at least one group has fewer than two cases at 95% confidence interval for means Within rows, similar small letters following means are not significant difference.The group sizes are unequal using Duncan's test of homogeneous subsets ANOVA at P ≤ 0.05 where F-values and degrees of freedom were varied