Size, not temperature, drives cyclopoid copepod predation of invasive mosquito larvae

During range expansion, invasive species can experience new thermal regimes. Differences between the thermal performance of local and invasive species can alter species interactions, including predator-prey interactions. The Asian tiger mosquito, Aedes albopictus, is a known vector of several viral diseases of public health importance. It has successfully invaded many regions across the globe and currently threatens to invade regions of the UK where conditions would support seasonal activity. We assessed the functional response and predation efficiency (percentage of prey consumed) of the cyclopoid copepods Macrocyclops albidus and Megacyclops viridis from South East England, UK against newly-hatched French Ae. albopictus larvae across a relevant temperature range (15, 20, and 25°C). Predator-absent controls were included in all experiments to account for background prey mortality. We found that both M. albidus and M. viridis display type II functional response curves, and that both would therefore be suitable biocontrol agents in the event of an Ae. albopictus invasion in the UK. No significant effect of temperature on the predation interaction was detected by either type of analysis. However, the predation efficiency analysis did show differences due to predator species. The results suggest that M. viridis would be a superior predator against invasive Ae. albopictus larvae due to the larger size of this copepod species, relative to M. albidus. Our work highlights the importance of size relationships in predicting interactions between invading prey and local predators.

Mesocyclops copepods were used to effectively control populations of Ae. aegypti and Ae. albopictus in six 91 communes in northern Vietnam (Kay et al., 2002). A semi-field study conducted in Bologna, Italy in 2007 found 92 that M. albidus reduced the Ae. albopictus density in experimental drums by greater than 99% (Veronesi et al., 93 2015). In the UK, laboratory experiments using M. albidus and Megacyclops viridis from Northern Ireland as 94 predators against Culex pipiens mosquito larvae from Surrey, UK have shown that these cyclopoid copepods are 95 effective control agents against Cx. pipiens and that the predators' attack rates tend to increase with temperature 96 (Cuthbert et al., 2018b). However, the use of UK copepods as predators of invasive Ae. albopictus larvae has not 97 been thoroughly examined over the range of temperatures that the invasive larvae are predicted to experience.

98
Due to the potentially negative impacts of exporting copepods to non-native regions for biocontrol purposes 99 (Coelho and Henry, 2017), it is important to investigate the performance of copepods local to the predicted sites 100 of Ae. albopictus establishment: London and South East England.

102
Populations of M. albidus and M. viridis cyclopoid copepods from the benthos of the Cumbrian lakes in the UK 103 have previously been studied at temperatures from 5 to 20°C (Laybournparry et al., 1988). Based on the dry 104 body masses of adult specimens, males were consistently smaller than females, and M. albidus copepods were 105 consistently smaller than M. viridis (Laybournparry et al., 1988). The general adult body length ranges for M. 106 albidus and M. viridis are 1.3 -2.5 mm (Einsle, 1993) and 1.2 -3 mm (Dussart, 1969, Einsle, 1988 107 1960), respectively. These are small enough to enable the distribution of copepod cultures to mosquito larval 108 habitats using "a simple backpack sprayer with a 5 mm hole in the nozzle," as has been previously 109 recommended (Marten, 1990b). Copepods reproduce sexually, and females that can produce new egg sacs every 110 3-6 days tend to predominate in mature populations (Marten and Reid, 2007). Cyclopoid copepods are 111 considered sit-and-wait ambush predators because of their attack behaviors, which have been described in six

115
Functional response curves were originally developed to relate the number of prey attacked by an invertebrate 116 predator to the prey density (Holling, 1966, Holling, 1959. Previous work has suggested that the best predators 117 to use in an "inundative release" biocontrol program are those that have type II functional responses to prey 118 density, so that the predators' attack rates are high even at low prey densities (Daane et al., 1996). Predators that 119 display a type III functional response against an invasive prey species -such as signal crayfish (Pacifastacus 120 leniusculus) against New Zealand mud snails (Potamopyrgus antipodarum)-may be able to limit the spread of 121 the invader, but cannot prevent it from establishing (Twardochleb et al., 2012

154
The copepods were kept in 3 L containers of spring water (Highland Spring, UK) at a 12:12 light/dark cycle, 155 and 20 ± 1°C, a temperature previously found to increase the portion of their life cycle spent in the reproductive 156 phase (Laybournparry et al., 1988). Chilomonas paramecium and Paramecium caudatum were provided ad 157 libitum as food for the copepods, and boiled wheat seeds were added to the containers to provide a food source 158 for the ciliates (Suarez et al., 1992). Adult copepods were identified as Macrocyclops albidus (Jurine, 1820) and   Adult non-gravid female copepods, identified by larger relative size, were removed from their culture and each 175 was placed in a Petri dish (diameter: 50 mm, height: 20.3 mm) holding 20 mL of spring water. At approximately 176 11am, the copepods were placed in three different controlled environments set to 15 ± 1, 20 ± 1, and 25 ± 1°C, 177 all at a 12:12 light/dark cycle to begin a 24 h starvation and temperature-acclimation period for the predators 178 ( Fig. 1). Each combination of copepod species (M. albidus or M. viridis) and temperature had 28 copepods, each one held in its own Petri dish; there were 168 copepods in total (Fig. S3a). Ten additional gravid females were 180 randomly selected from each species of copepod and preserved in 80% ethanol solution for size measurements.

214
Ae. albopictus larvae were hatched (Supplementary Material) and any residual food was diluted in spring water 215 following the same procedure used for the functional response experiments. Adult non-gravid female copepods 216 were each placed in a Petri dish at approximately 11am for a 24 h period of starvation and acclimation to three 217 different temperature settings: 15 ± 1, 20 ± 1, and 25 ± 1°C, all at a 12:12 light/dark cycle (Fig. 1). The largest 218 non-gravid copepods of each species were selected to minimize the risk of selecting males or immature stages.

219
The Petri dishes containing larvae were split into the three different temperature settings 12 h prior to the 220 introduction of copepod predators, and there was a 6 h period of predation between 11am and 5pm ( Fig. 1). At 221 the end of the 6 h period, the copepods were removed, and each was stored in 80% ethanol and labelled 222 according to its larval Petri dish. The number of surviving larvae in each Petri dish was recorded immediately 223 after removing the copepods.

225
At each of the three temperature settings, there was a total of 24 Petri dishes, each containing 24 larvae (1,728 226 larvae across all temperatures); the 24 dishes were divided into three groups (n = 8): one with M. albidus 227 predation, one with M. viridis predation, and one as a control (Fig. S3b). Every predator treatment was matched 228 to a control that had been held at the same temperature (Fig. S3b). Each temperature setting had a total of 192 229 control larvae. Larval background mortality was 5% at 15°C, 11% at 20°C, and 3% at 25°C. A total of 24 M. 230 albidus and 24 M. viridis copepods were used in this experiment. The body lengths of these copepods were 231 measured after each was preserved in 80% ethanol.

233
Functional response curve analysis:

235
The general shapes of the functional response curves were determined for each combination of copepod species 236 and temperature using predator-present data. When polynomial logistic functions are fit to describe the 237 relationship between the proportion of larvae killed and the initial larval density, a negative first-order term   where dN/dt is the change in prey density over time, b is the attack coefficient, q is an exponent that can fall 260 between a type II response (q = 0) and a type III response (q = 1), h is the handling time, P is the predator 261 density, and m is the mortality rate (Rosenbaum and Rall, 2018).

263
Since background prey mortality should not alter the general shape of the functional response curve 264 (Rosenbaum and Rall, 2018), the q parameter was fixed to either 0 or 1 based on the results of the "frair_test" Two linear regression models were fitted to explain predation efficiency. The first tested temperature and 283 copepod species as predictors (eqn S1), and the second tested copepod body mass in place of species (eqn S2).

284
Both models were also fitted without temperature predictors (Supplementary Material  (Table S1). Methods capable of accounting for background 306 mortality were used to graph the functional response curves, and to estimate the attack coefficient and handling 307 time parameters (Fig. 2).

309
Attack coefficient estimates are higher for M. viridis than for M. albidus in some instances. However, there are 310 no significant differences between any of the six estimates due to copepod species, temperature, or any 311 interaction of those variables (Fig. 2b). Similarly, the lower handling time estimates for M. viridis are not 312 significant; thus, no differences can be attributed to copepod species, temperature, or an interaction of those 313 predictors (Fig. 2c). For both parameter estimates, M. albidus predation at 25°C has the largest standard error; 314 this is likely due to the relatively high background mortality observed in this group (Fig. 2b, 2c).

327
A linear model testing copepod species and temperature category as predictors of predation efficiency explained 328 6.8% of the variance in predation efficiency (   372 species or temperature, tested at 15, 20, and 25°C (Fig. 2). A previous study of predator-prey interactions 373 involving sit-and-wait copepod predators found no significant difference in the attack parameter of the 374 functional response across three different temperatures: 18, 22, and 26°C (Novich et al., 2014). In addition, we 375 found that temperature categorical variables were not significant predictors of copepod predation efficiency 376 against Ae. albopictus larvae, and removing these predictors decreased the AICs and increased the adjusted R 2 377 values of our linear models (Tables 1-2, S2-S3).

379
Our handling time estimates (Fig. 2c) are likely over-estimations of ingestion time and may also include some of 380 the predators' search time. Although a cyclopoid copepod has previously been observed eating a chironomid 381 larva for more than 30 minutes (Fryer, 1957a), another study states that a mosquito larva is usually consumed within a few minutes (Marten and Reid, 2007). A previous study that compared directly-observed handling 383 times to handling times fitted by functional response calculations found that the directly-observed handling 384 times were significantly lower than the fitted handling time estimates (Poole et al., 2007). However, when 385 handling time was assumed to include both the time that the predator takes to move to the prey, and the 388 on bumping into … prey during the course of … meanderings" (Fryer, 1957a). Thus, it is plausible that 389 inefficient searching made up a large portion of our handling time estimates (Fig. 2c).

391
We found that M. viridis was a significantly more efficient predator of Ae. albopictus than M. albidus (Tables 1   392 & S2, Fig. 3). An analysis of the gut contents of English Lake District copepods previously found that 15.

407
The larger size of M. viridis contributes to the higher predation efficiency observed among that species against 408 Ae. albopictus (Fig. 3b). When the binary copepod species variable (Table 1, Table S2) was replaced with a 409 continuous copepod body mass variable ( Table 2, Table S3), the adjusted R-squared values increased, and the 410 AICs decreased. The linear regression model with the highest adjusted R-squared value, 16%, was the model 411 with copepod body mass as the sole predictor of predation efficiency (Table S3). This adjusted R-squared value 412 is still low, indicating substantial variation within the predation interaction. The positive relationship we 413 observed between copepod body mass and predation efficiency (Tables 2 & S3, Fig. 3b) is consistent with the 414 findings of a previous meta-analysis of crustacean predation of immature fish, which showed that the predation 415 rate was negatively related to the prey/predator size ratio (Paradis et al., 1996). According to the theory that, at 416 equilibrium, "the rate of food intake is equal to the rate at which food is leaving the gut" (Charnov and Orians,

428
Although we did not find any difference in functional response parameter estimates or predation efficiency due 429 to temperature, the temperature at which copepods are cultured impacts their size (Laybournparry et al., 1988  where Mass is a continuous variable referring to each copepod's body mass in mg; and all other notation is 738 identical to that used in the previous model.

740
Both linear regression models above were also fitted without any temperature predictors. Akaike information