Mites’ cultures
Neoseiulus californicus and N. cucumeris were obtained from a laboratory culture rearing on Tetranychus urticae Koch. and Tyrophagus putrescentiae. S. newsami and P. citri were collected from citrus/lemon orchards in 2019 and cultured for 6 months before starting the experiment. The predator–prey synchronizations were gained by rearing N. californicus, N. cucumeris, and S. newsami on P. citri. The eggs of predators were separated into new leaf (lemon) discs with enough prey mites to get same-aged adult females. All this rearing process was carried out under controlled conditions with a temperature of 26 ± 1 °C, photoperiod of 16: 8 h (Light: Dark), and relative humidity of 75 ± 1%.
Experimental conditions and methodology
The experiment was performed using a lemon leaf disc (lower leaf surface up) (3.5 cm diameter) by placing it on a water-saturated sponge. All leaf discs were stripped with absorbent paper on their edges to prevent mites from escaping and retain moisture. Mites attached to the absorbent paper were not included in the study data. Each 4 tested units were considered one test group by placing into a plastic box (15 cm × 15 cm × 6.5 cm) and replicated 10 times (10 × 4). All plastic boxes were filled with water just below leaf discs and water was maintained regularly. All experimental procedure was done under controlled conditions as described above. Adult females (> 24 h old) were collected from the stock cultures and isolated from the colony before gaining maturity for uniform age. All collected females were allowed to mate with males for a period of 24 h. After mating, females were starved for 24 h before transferring to the experimental units. Injured or less moved females were excluded from the experiments.
Different densities of P. citri eggs were obtained from placing the 10–20 adults female on the lemon leaf discs and removed after 24 h under controlled conditions. The number of eggs was counted as per density by removing excess eggs. Immature densities were obtained by placing more than 50 adult females on the lemon leaf and allowing them to lay eggs for 5–7 h. before removing the adults. For further growth, the lemon leaf with eggs was placed under controlled conditions for further development. As soon as the immature stage (> 90% proto-nymph) was obtained, they were shifted to the experimental unit.
The selection of prey densities against the 3 predators (N. californicus, N. cucumeris, and S. newsami), a pre-consumption of immatures and adults of P. citri was carried out. The pre-densities selection test revealed that offering predators more than 200 and less than 4 motile stages of P. citri reduced their searching performance and reduced fecundity rate. A maximum of 160 and a lowest of 5 density of each stage was offered to the predator based on the pre-test data, as endorsed by Xiao and Fadamiro (2010).
A functional response experiment was performed using 6 densities (5, 10, 20, 40, 80, and 160) of eggs, immatures, and adults of P. citri by offering to a single female of N. californicus, N. cucumeris, and S. newsami. The prey eggs, immatures, and adults consumed by the predators were replaced by new ones during the experiment. Data of prey eaten and eggs laid by all predators were recorded after 24 h.
Statistical analysis
Data from functional responses are analyzed in 2 steps (Juliano 2001). In the first step, maximum likelihood estimates (intercept (P0), linear (P1), quadratic (P2), and cubic (P3) coefficients) were calculated, using a logistic regression/polynomial function by comparing the proportions of prey eaten to the initial density and resulted in functional response type.
$$N_{e} {/}N_{0} {\text{ = exp(}}P_{0} { + }P_{1} N_{0} { + }P_{2} N_{0}^{2} { + }P_{3} N_{0}^{3} {\text{)/[1 + exp(}}P_{0} { + }P_{1} N_{0} { + }P_{2} N_{0}^{2} { + }P_{3} N_{0}^{3} {)]}$$
where “Ne/No” is the prey consumption probability and “intercept (P0), linear (P1), quadratic (P2), and cubic (P3)” are regression coefficients. A significant functional negative and positive value of P1, the predator gave functional response types II and III curve, respectively (Juliano 2001).
Since it accounts for prey loss per density, the non-linear least square regression was used in the second step to estimate handling time and attack rate coefficients.
$$N_{e} = N_{0} \left\{ {1 - \exp \left[ {\alpha \left( {T_{h} N_{e} - T} \right)} \right]} \right\}$$
where “Ne” stands for the number of prey killed, “No” stands for initial prey density, “Th” stands for the handling time, and “T” stands for the total time given to the predator (24 h). The attack rate “α” and handling time “Th” parameters were calculated using a nonlinear least-squares regression procedure.
Curve fitting using the polynomial equation and Holling’s disc equation was performed using R (RCore-devlopment-team 2019). ANOVA analyzed the prey density-dependent effects on consumption and number of eggs laid by predators, and then the Tukey test was used to compare the results (P = 0.05). The abundance data of the 3 predatory mites were obtained from a greenhouse augmented biological control experiment by releasing the predatory mites (n = more than 50/plant) at the highest infestation of P. citri and other insect pests. Abundance data (Mean ± SE) from highest (predators released) to almost zero P. citri was used. The fecundity (%) of predators (per adult female) was calculated by using the below formula by feeding on different life stages of P. citri;
$${\text{Fecundity}} = \left( {{\raise0.7ex\hbox{${{\text{Eggs}}\;{\text{production}}}$} \!\mathord{\left/ {\vphantom {{{\text{Eggs}}\;{\text{production}}} {{\text{prey}}\;{\text{consumed}}}}}\right.\kern-\nulldelimiterspace} \!\lower0.7ex\hbox{${{\text{prey}}\;{\text{consumed}}}$}}} \right) \times 100$$
The relative control potential (RCP) was calculated to compare their efficacies based on a maximum feeding rate (1/h), as a result of their functional response and abundance of predator in greenhouse or fecundity (%) as calculated above. The RCP was calculated as below formula;
$${\text{RCP}} = \left( {\frac{{{\text{FR}}_{{(1/{\text{h}})}} {\text{predator}} - A}}{{{\text{FR}}_{{(1/{\text{h}})}} {\text{predator}} - B}}} \right) \times \left( {\frac{{X\;{\text{predator}} - A}}{{X\;{\text{predator}} - B}}} \right)$$
where FR is the maximum feeding rate (1/h), and “X” is proxies for the responses (a measure of predator greenhouse abundance or fecundity of predator). If the value of RCP = 1, no difference between the 2 predators; for RCP < 1, predicted predator-A less efficient than predator-B, and RCP > 1, predator-A is predicted as greater efficacy than predator-B. RCP indicating the predicted increase or decrease relative efficacy of predator-A compared to predator-B (Cuthbert et al. 2018).
