Antipredator responses of three Daphnia species within the D. longispina species complex to two invertebrate predators

Abstract Prey communities in natural environments face a diverse array of predators with distinct hunting techniques. However, most studies have focused only on the interactions between a single prey species and one or more predators and typically only one of many induced defense traits, which limits our understanding of the broader effects of predators on prey communities. In this study, we conducted a common garden experiment using five clones each of three Daphnia species (D. cucullata, D. galeata, and D. longispina) from the D. longispina species complex to investigate the plasticity of predator‐induced defenses in response to two predators in a community ecology setting. Five clones from each species were subjected to predator kairomones from two closely related invertebrate predators that are common in several European lakes, Bythotrephes longimanus or Leptodora kindtii for a duration of 10 days, and the morphological traits of body size, head size, spina size, and the presence of spinules on the ventral and dorsal carapace margins were measured. We show that among the species within this species complex there are different antipredator reactions to the invertebrate predators. The induced responses exhibited were species, trait, and predator‐specific. Notably, D. galeata and D. cucullata developed distinctive helmets as defensive mechanisms, while microdefenses were induced in D. galeata and D. longispina, but not in D. cucullata. This demonstrates that the expression of micro‐ and macrodefenses across species was unrelated, highlighting the possible independent evolution of microstructures as defensive modules in Daphnia's antipredator strategies. This study is the first to document both micro‐ and macrodefensive phenotypic plasticity in three co‐occurring Daphnia species within the D. longispina species complex. The differences in inducible defenses may have a substantial impact on how these three species cohabit with Bythotrephes and Leptodora.


| INTRODUC TI ON
Predation is a crucial evolutionary force that shapes ecological communities and drives the development of antipredator defenses in many prey species (Tollrian & Harvell, 1999).The evolution of antipredator defenses is determined by several factors.One of these factors is the presence of a reliable cue that indicates the proximity of a threat and activates a defense response in the prey.
Additionally, the effectiveness of the prey's defenses against the predator, and the balance between the costs and benefits of developing defenses play crucial roles in this evolutionary process (Tollrian & Harvell, 1999).Prey employ various inducible defenses against their predators, such as morphological changes (Laforsch & Tollrian, 2004;Octorina et al., 2022;Sperfeld et al., 2020), behavioral modifications (Stich & Lampert, 1981), and life-history traits (Kruppert et al., 2017;Stibor, 1992) to enhance their fitness and reduce predation risk, which are essential for their survival in the environment (Riessen, 2012).
A wide range of organisms, including, for example, protozoans, vascular plants, rotifers, arthropods, and vertebrates have been observed to possess inducible antipredator defenses (Tollrian & Harvell, 1999).In natural environments, prey encounter various predators, each employing distinct hunting and capturing techniques (Laforsch & Tollrian, 2009).Prey can gather and evaluate information about the risk of predation, for example, through chemical cues emitted by predators or their conspecifics (Turner, 2008;Weiss et al., 2012).They can also differentiate between predator species, as seen in Daphnia pulex and a freshwater snail Radix balthica exhibiting predator-specific traits when facing multiple predators (Beckerman et al., 2010;Lakowitz et al., 2008;Miner et al., 2005).
Nevertheless, most studies examining predator-prey interactions have been limited to investigating interactions between one prey and one predator species (Sperfeld et al., 2020;Tams et al., 2018;Weiss et al., 2015) and single-induced traits.Even while there has been some research conducted on prey with multiple predators and traits (Diel et al., 2021;Herzog et al., 2016;Laforsch & Tollrian, 2004), there are still surprisingly few studies that concentrate on coexisting prey alongside their complete predator assemblage.This limited scope hinders a comprehensive understanding of the breadth of predator-induced responses in the prey species that comprise natural ecological communities (Laforsch & Tollrian, 2004;Miner et al., 2005).
Despite this huge amount of research done during the last decades (Diel et al., 2020), new types of inducible defenses, particularly morphological defenses, in Daphnia continue to be discovered, including the alteration of tiny microstructures (Diel et al., 2021, Ritschar et al., 2020).According to Diel et al. (2021), changes in an organism's microstructure might either be developmentally connected to more obvious changes in its features or represent independent defensive structures to fine-tune protection against a predator.The first reasoning would suggest that changes at the microstructural level are strongly related to the more significant induced changes that take place during development.For instance, Laforsch and Tollrian (2004) noted that D. cucullata displays a strengthened carapace and extended helmet in response to various invertebrate predators, a positive correlation between the macro (helmet) and micro (carapace structure) induced defense.However, in accordance with the "concept of modality" of predators, specialized defenses have generally surpassed general ones throughout evolution.In order to improve the overall defense strategy, subtle changes may thus be important.For example, spinules, micro spike structures along the carapace, might additionally play a role in refining or fine-tuning the induced phenotype to provide optimal protection against a particular predator with distinctive methods of capture or manipulation (Diel et al., 2021).
In many European lakes, Daphnia of the Daphnia longispina complex, D. longispina, D. galeata, and D. cucullata co-occur and dominate the grazing pressure on phytoplankton.In many of these lakes, these Daphnia species are exposed to one or two cladoceran predators, Leptodora kindtii and Bythotrephes longimanus.These two predators differ in predation modes: Leptodora is a tactile predator and Bythotrephes is a visual predator (Octorina et al., 2022).Bythotrephes relies on mechanoreceptors or its large medial compound eye to detect its prey.It then uses its long feeding appendages to grasp the prey and subsequently shred it (Manca et al., 2008).In contrast, Leptodora employs a strike tactic for capturing prey and requires direct contact with the prey before initiating an attack (Browman et al., 1989;Manca et al., 2008) and uses a "trap basket" for capturing prey (Branstrator, 2005).Unlike Leptodora, Bythotrephes are likely capable of successfully feeding on larger prey items because they are not constrained by a feeding basket (Manca et al., 2008).
Previous studies have shown that L. kindtii induces significantly longer helmets and tail spines on D. cucullata (Laforsch & Tollrian, 2004), while B. longimanus induces a typical helmet on D. galeata (Octorina et al., 2022), but the responses of D. longispina to both predators have not been explored.
Here, we investigate the predator-induced defenses in a community ecology context, evaluating responses among three species within the D. longispina species complex which co-occur with the two invertebrate predators in Lake Constance.We examine plasticity in head, respectively helmet, body, and spina size, as well as in proposed microstructures, the extension of the ventral and dorsal spinule areas, and the length of the ventral spinule (Diel et al., 2021) in response to the two predators.
Both D. galeata and D. longispina are morphologically similar and adult individuals of these species typically range in body size from 1.2 to 2.5 mm.As D. cucullata is smaller (Ogorelec et al., 2022;Stich et al., 2005) compared with the other two species, we expect that small D. cucullata might be more vulnerable to invertebrate predation.Besides size, the species also differ in antipredator behavior.D. longispina performs diel vertical migration (DVM) (Geller, 1986;Stich & Lampert, 1981), while D. cucullata and D. galeata stay in the upper water strata (0-20 m) (Geller, 1986;Ogorelec et al., 2022;Stich & Lampert, 1981).As the two invertebrate predators also inhabit the upper water layers (Stich, 1989), D. galeata and D. cucullata (not performing DVM), but not D. longispina (performing DVM) are exposed to the two predators throughout a 24-h day.
Given this background and the potential for co-variation between macro and micro induced defense traits, we formulated and tested four hypotheses: (1) based on the differences in predation behavior between the two predator species, we anticipate distinct responses in the daphnids; (2) we expect variations in the expression of defensive traits among the Daphnia species, with the most vulnerable D. cucullata showing the strongest responses, while the least vulnerable D. longispina displaying only weak responses to predator kairomones; (3) given the "connectedness sensu Diel et al. (2021)" between small-scale defensive qualities and the large-scale defensive traits against invertebrate predators, we expect that all three species will display microdefenses in response to both predators; (4) finally, we expect that the expression of microdefenses would be directly related to the expression of macrodefenses, emphasizing the complexity of the defensive strategies.

| Origin of Daphnia clones
Daphnia clones were collected in Upper Lake Constance, located in Central Europe and bordering Germany, Switzerland, and Austria, which is the main basin of the large peri-alpine Lake Constance (Güde & Straile, 2016).The number of Daphnia species has changed there during the last century.While until the mid-20th century only D. longispina was found in the lake, D. galeata was able to invade the lake with eutrophication in the 1950s (Straile, 2015).More recently, D. cucullata established large densities (Ogorelec et al., 2022) and currently all three species co-occur in the lake.Both predators, B. longimanus and L. kindtii, have been regularly found in the lake since more than one century (Straile & Geller, 1998).Clones of Daphnia galeata were hatched from ephippia, which were collected from the upper 20 cm of a sediment core taken from the lake.In contrast, Daphnia longispina and Daphnia cucullata clones were collected in Lake Constance using a plankton net (mesh size 140 μm) drawn from 40 to 0 m.The lineages established from these individuals of D. cucullata and D. longispina are thus isolates, and it is not 100% certain that they represent different clones.However, given the large clonal diversity of Daphnia spp. in Lake Constance (Beninde, 2021), and our low numbers of isolates obtained from 100's sampled, we consider it unlikely that two randomly picked individuals from several hundred individuals in one sample share their clonal identity.Consequently, in this paper, we referred to them as clones, acknowledging the potential that two isolates may share the same clonal identity.The three species were cultured in the laboratory for several months before the experiment.The third clutch from the third generation of maternal lines was collected as study organisms.In order to reduce the influence of extraneous variables, all stock cultures of clones were kept under the same conditions (e.g., feeding volume, feeding schedules, temperature, light).Maternal cultures were kept in 1 L glass jars with 800 mL lake water in 20°C (range: 19-21°C) with a 16:8 L:D photoperiod cycle.
These cultures were fed with 2 mg C/L green alga Tetradesmus obliquus (Culture Collection of Algae, University of Göttingen, Germany, SAG 276-3a) three times per week and were transferred to a new medium until they reached the third generation.Lake water was prepared for both Daphnia culturing and the experiment by undergoing a process involving filtration through a 0.2μm mesh sieve.Subsequently, the filtered water was aerated and allowed to age for 24 h to prevent expression of morphological defenses in Daphnia cultures.

| Experimental setup and measurements of morphological traits
Since our focus was on studying species differences rather than clonal variations within species, we conducted replications at the species level, with five clones each, rather than at the clonal level.Five predators were added to each jar in the predator treatments, which were collected from the lake using a plankton net towed behind a boat.These predators were placed in a cylindrical plastic cage (4 cm diameter and 6 cm height) with a 140μm nylon mesh (Octorina et al., 2022) and were not fed with Daphnia.In the control treatment, the plastic cage was left empty.Every second day, daphnids were transferred to new jars filled with lake water with T. obliquus and freshly collected predators.Throughout the experiment, daphnids were fed with 1 mg C/L of the green alga T. obliquus for the first 7 days and 2 mg C/L for the remaining 3 days to ensure also that large and matured daphnids have unlimited food.The experiment was conducted at a constant temperature of 20°C (range: 19-21°C) and a 16:8 L: D photoperiod cycle.
At an age of 10 days, the daphnids were preserved in 70% ethanol following shock treatment in 95% ethanol (Black & Dodson, 2003).We chose age 10 for all three species because this age is sufficient to guarantee that all individuals were mature in all treatments.Subsequently, morphological defenses of Daphnia individuals were measured in R version 4.2.2 (R Core Team, 2021) with a self-written script using photographs of the daphnids taken with a Bresser MikroCamII 12 MPu mounted to a Stemi 2000-C stereomicroscope (Carl Zeiss Werk, Göttingen, Germany).For each Daphnia, we measured body size, head size, spina size, and the spinules on the ventral and dorsal carapace margins (Figure 1).Body size was determined from the base of the spina to the mid-eye, while spina size was measured from the end of the tail spine to the base of the spina.Head size was measured in parallel to the body axis from the top of the head or top of the helmet until the line perpendicular to the body axis crossing the midpoint of the eye.All Daphnia individuals were examined for the presence of spinules in the dorsal and ventral carapace margins and for the presence of neckteeth.The ventral spinules bearing area (ventral SBA) was measured from the first ventral spinule to the base of the tail spine, while the dorsal spinules bearing area (dorsal SBA) was measured from the first dorsal spinule to the base of the tail spine.Additionally, the length of the ventral spinules was measured from the base of the spinule to its tip, with five spinules per animal measured (Diel et al., 2021).Unfortunately, one D. longispina and eight D. galeata pictures were not suitable to measure the microstructures.This reduced the number of observations, but not the number of independent experimental units in our dataset regarding microdefenses.

| Test of hypotheses 1 and 2: Predator-specific responses and different level of defensive trait expressions in Daphnia
Variations in body size, head size, spina size, dorsal SBA, ventral SBA, and ventral spinule length were then analyzed using linear mixed-effect models (Kuznetsova et al., 2017) with body size, treatment, and species as fixed effects and clonal identity as a random effect.This accounts for non-independence of the daphnids from each clone which were kept in one jar.Body size was centered in the model to eliminate co-variation between body size and Daphnia species identity.Hence, centered body size in the models predicts within-species trait variability, whereas between-species trait variability is predicted by species identity.We used likelihood ratio tests (ANOVA and RANOVA functions) (Kuznetsova et al., 2017) to evaluate the significance of fixed and random effects.If significant treatment effects were found, we conducted pairwise post hoc comparisons using the Tukey method (Lenth et al., 2023).

| Test of hypothesis 3: Microdefenses in all three species
To test the predictions regarding the occurrence of microdefenses in all three species, we calculated the mean trait residual for dorsal SBA, ventral SBA, and ventral spinule length, and compared these values among the Daphnia species.The linear mixed-effect model for dorsal SBA, ventral SBA, and ventral spinule length was then utilized to analyze treatment and species effects, followed by post hoc comparisons.

| Test of hypothesis 4: Co-variation of macrodefenses and microdefenses
Finally, to examine co-variation between macrodefenses and microdefenses, we first calculated the residuals of regressions between each trait and body size separately for each species.Next, F I G U R E 1 Measurements of various morphological traits of Daphnia species and the presence/absence of dorsal and ventral spinules.
we calculated the mean residual values for each trait and clone and determined the correlation for each combination of pairwise traits and across Daphnia species.This allowed us to assess the presence of trade-offs versus positive co-variation between the different traits.For multiple comparisons, we corrected the p-values using a Benjamini-Hochberg adjustment (Benjamini & Hochberg, 1995).
Post hoc testing, where appropriate, was carried out using the emmeans package (Lenth et al., 2023).Figures were generated using We did not observe any neckteeth for any Daphnia.All other traits were influenced by predator kairomones in a species-specific manner, and except for the ventral spinule length, all trait lengths were related to body length (Figures 2 and 3).This relationship held even when body lengths were centered for each species (Table 1).
Regarding body size, head size, and spina size, significant clonal variability was found, indicated by significant random intercepts (Table 1).
The influence of kairomones on spina length depended on body size and Daphnia species (significant three-way interaction).This was mainly due to the response of D. galeata in the Leptodora treatment, where the spina length-body size slope was steeper than in the Bythotrephes and control daphnids, resulting in large spines only for large daphnids (Figure 2a).The spina length-body size slopes were comparable between the control and Bythotrephes treatments for D. galeata and larger spines were observed in the Bythotrephes treatment compared with the control treatment at similar body sizes (t 68.5 = 3.939, p = .0006).

F I G U R E 2
Relationship between body size and spina size, head size, dorsal SBA, ventral SBA, and ventral spinule length for the three Daphnia species.Each treatment condition is shown with distinct letters (Bythotrephes (B), Control (C), and Leptodora (L)).

| Hypothesis 3: Microdefenses in all three species
We found qualitative differences between the three species regarding microdefenses: D. cucullata did not develop microstructures in the presence of predators in contrast to the other two species.Consequently, we used only the data for D. galeata and D. longispina in the mixed models for microstructures.The effect of kairomones on dorsal SBA depended on body size and Daphnia species (significant three-way interaction).This threefold interaction was mostly caused by the Bythotrephes treatment response, where dorsal SBA for D. longispina increased more steeply with body size than SBA of D. galeata (Figure 2c).Ventral SBA and ventral spinulae lengths did not differ between the two species, but were larger in both kairomones treatments compared to the control (Figure 3e

| Hypothesis 4: Co-variation of macrodefenses and microdefenses
Morphological traits adjusted for body size (Figure 4) co-varied across species and treatments in 4 out of 10 trait combinations.
Across all species and treatments, Daphnia with species and body size-specific large head/helmet were found to have also a

| DISCUSS ION
The invertebrate predators Bythotrephes longimanus and Leptodora kindtii are able to induce morphological defenses within all three species of the D. longispina complex.We found that the morphological responses of D. longispina species complex were trait-, Daphniaspecies, and predator-species-specific.However, our results only partially support our expectations.D. longispina was indeed less responsive to the two predators as compared to the other two species but only regarding macrodefenses.Microdefenses were induced in F I G U R E 3 Responses (means ± 2 standard errors) of the three Daphnia species to three treatments (B: Bythotrephes, C: Control, L: Leptodora) for the traits body size, residual head size, residual spina size, residual dorsal SBA, ventral SBA, and residual ventral spinule length.Small dots represent the clonal means for the different treatmentspecies combinations.
D. longispina and D. galeata, but not in D. cucullata.Hence, microdefenses were expressed independently of macrodefenses across the three Daphnia species.
We did neither observe the induction of an expanded head form (helmet) nor spina elongation for D. longispina.The absence of helmet induction corresponds to the typical appearance of D. longispina in Lake Constance (Güde & Straile, 2016).This suggests that diel vertical migration of this species in Lake Constance seems to provide sufficient protection against the predation pressure of both invertebrate predators against the visual predator Bythotrephes but also against the tactile predator Leptodora.The lack of strong morphological responses here suggests a trade-off may exist between behavioral and morphological defenses and suggests different strategies to deal with predation risk can evolve among species in this complex.The supposed behavioral protection seems to remove potential fitness benefits of elongated heads and spines, causing absence of helmets not only in situ, but also in our laboratory environment, in which daphnids were exposed to kairomones and light during daytime.Hence, helmet induction seems to be not a part of the defensive repertoire of D. longispina and is not only switched off in situ by DVM induced reduction of light availability.This is in contrast to life-history antipredator strategies of D. magna in response to fish kairomones, which expression depends on light intensity (Effertz & von Elert, 2014).
Absence of D. longispina morphological defenses was also observed in Norwegian deep lakes (Sperfeld et al., 2020).However, D. longispina clones from small and shallow water bodies in Norway developed longer spines and neck teeth in response to Chaoborus kairomones (Sperfeld et al., 2020).Interestingly, some D. longispina with neck teeth have also been observed occasionally in Lake Constance (Güde & Straile, 2016) despite the absence of Chaoborus in this deep lake.As no neck teeth were observed in our experimental animals this might suggest that either a different predator induces neck teeth induction in Lake Constance D. longispina, or that neck teeth expression is highly clone-specific.In the latter case, neck teeth were not part of the defensive repertoire of the five clones used in our study.This reasoning is consistent with the rarity of neck teeth observations in Lake Constance.
The strong response of D. galeata observed in this study confirms our recent study showing that Bytotrephes and Leptodora will induce helmets and longer spines in this species (Octorina et al., 2022).In the same study, using eight clones with replicates at the clonal level, the age at first reproduction was observed to be the latest and demographic costs, characterized by a reduced number of offspring, were found to be the highest in the presence of Bythotrephes treatment.Helmet formation of D. galeata is also typically observed in Lake Constance during the summer season (Güde & Straile, 2016).Likewise, helmet formation of D. galeata-mendotae has been experimentally induced by other invertebrate predators such as Chaoborus and Notonecta (Dodson, 1988).However, D. galeata clones from various lakes, including Lake Constance, did not produce a helmet in response to fish kairomones (Tams et al., 2018).
Longer spines in response to copepod kairomones have been TA B L E 1 Summary of the linear mixed-effects model on the morphological trait responses of Daphnia community against two co-occurring invertebrate predators.Note: The analysis included fixed effects such as centered body size, head size, spine size, dorsal spinule bearing area (SBA), ventral SBA, and ventral spinule length using type III analysis of variance with Satterthwaite's method to examine the significance of these fixed effects.Clonal identity was treated as a random effect, and Likelihood ratio tests were employed to evaluate its impact on the model."Tr" refers to the three treatments: Bythotrephes, Control, Leptodora.The model employed an equal count of observations (n = 101) for body size and head size, whereas the count differed for spina size (n = 96) because data points were excluded for daphnids with broken spines in the Bythotrephes treatment.Additionally, the observation count varied for the microdefenses model (n = 71) due to the absence of these defenses in D. cucullata.Significance levels are given as: ns, not significant; *p < .05;**p < .01;***p < .001.
Induction of D. cucullata helmet formation in response to invertebrate predators is widespread and has been experimentally demonstrated in response to Chaoborus, copepod, and Leptodora kairomones (Laforsch & Tollrian, 2004).Likewise-and in contrast to our results-spina elongation was demonstrated in response to Chaoborus and Leptodora kairomones (Laforsch & Tollrian, 2004).
Unfortunately, we lost two D. cucullata clones within the Leptodora treatment, and thus had a lower statistical power to detect effects of Leptodora kairomones for this species.We can therefore not exclude the possibility that also Lake Constance D. cucullata might elongate their spines in response to Leptodora and Bythotrephes.
Our study is the first to show the induction of three micro- presence and length induction of microspines in D. longispina despite the absence of morphological macrodefenses.
The multitude of Daphnia induced defenses have been suggested to be uncoupled (Boersma et al., 1998) following a modular concept, which favors the evolution of defenses tailored toward individual predator species (Herzog & Laforsch, 2013).We found that body size-adjusted traits in the Daphnia longispina complex co-varied positively with other body size-adjusted traits both between and within Daphnia species.For example, adjusted head, respectively, helmet sizes, were significantly related to all other traits besides dorsal SBA.
In contrast, the expression of adjusted dorsal SBA was not related to any other morphological trait.This might suggest that specific combination of traits might be especially favorable in some predator environments and might act synergistically.In contrast, other traits might provide only additive benefit when exposed to predators, and thus show no co-variation with other traits.

| FUTURE RE S E ARCH
Within the world of plasticity research, one of the ideas that receives too little attention is how among trait relationships (tradeoffs or positive covariation) vary across environments and among species (Reger et al., 2018;Stearns, 1989).This idea that tradeoffs or positive co-variation patterns are themselves plastic is important because if the traits involved relate strongly to fitness, as predator defense traits do, then context-dependent variation in these relationships will be important to coexistence under variable environments.
While we cannot analyze our data to explore bivariate reaction norms across environments as proposed by Stearns (1989), we do have enough data to propose a compelling hypothesis that relationships between traits are predator-environment dependent and that this dependency may vary among co-occurring prey species that share two common predators.
For example, in Figure 5, we can see that patterns of G × E vary among species (the variation and sign of the relationship for each species change in different ways across the treatments).Variability in particular seems to manifest in different ways among species for each predator.For example, head size is more variable under Bythotrephes whereas spina size is more variable under Leptodora.
One might argue that only with Bythotrephes as a predator, we see The five clones of each Daphnia species, D. cucullata, D. galeata, and D. longispina were placed in three different treatments: Bythotrephes (B), Control (C), and Leptodora (L).For each clone and treatment, we added three to four 1-day old Daphnia individuals to 200-mL glass jars containing 150 mL lake water.The number of experimental animals at the start of the experiment was 3 species × 5 clones (5 jars) × 3 treatments × 3-4 daphnids per jar, in total of 45 jars and 164 individuals.Unfortunately, 63 daphnids died during the experiment resulting in 101 daphnids used for morphological measurements.Mortality was mostly caused by the transfer of daphniids into new jars every second day and was highest for fragile D. cucullata (65% of all mortality).This caused the loss of two clones of D. cucullata in the Leptodora (L) treatment.The analysis of spina length variability was conducted using only daphnids with intact spines (n = 96).This exclusion involved the removal of two clones of D. cucullata in the Bythotrephes (B) treatment due to broken spina.Hence, the number of independent experimental units was 43 (3 species × 5 clones × 3 treatments-2) for analyses of body and head sizes and microstructures, and 41 for the analysis of spina size.
the ggplot2 package(Wickham, 2016) and the pairs R function.3| RE SULTS3.1 | Hypotheses 1 and 2: Predator-specific responses and the differences in the defensive trait expressions in DaphniaAfter a growth period of 10 days, it was evident that the body size of D. cucullata was significantly smaller, mostly below 1000 μm, compared to the other two species, which generally reached body sizes between 1500 and 2000 μm (Figures2 and 3a).The presence of predators led to an increase in the body size of D. galeata (B -C: t 84.7 = 2.769, p = .0187;C -L: t 84.2 = −3.355,p = .0034)(Figures2 and 3a), while the effects of predator kairomones on the body size of the other two species were not as clear (Table1a, significant treatment × species interaction).
species-and size-specific large spina (r = .47,p < .05).Likewise, residual head size and residual ventral SBA (r = .49,p < .05),residual head size and residual ventral spinule length (r = .52,p < .05),and residual ventral SBA and residual ventral spinule length (r = .69,p < .0001)were significantly and positively related after correction for multiple testing.No co-variation of trait expressions was observed between residual dorsal SBA and any other trait (Figure4).Correlations conducted within species suggest that D. galeata had positive significant co-variation between residual head size and spina size (r = .62,p < .05), between residual head size and residual ventral spinule length (r = .70,p < .05), between residual head size and residual ventral SBA (r = .70,p < .05), between residual spina size and residual ventral spinule length (r = .58,p < .05),and between residual ventral SBA and residual ventral spinule length (r = .71,p < .05).No significant co-variation was found between traits of D. longispina and between residual head size and residual spina size of D. cucullata.
defenses in D. longispina and D. galeata, and the absence of these defenses in D. cucullata.Such induction of microdefenses has been reported previously in several Daphnia species, e.g., D. barbata, D. similis, D. magna, and D. longicephala(Herzog & Laforsch, 2013;Ritschar et al., 2020) in response to the invertebrate predators Triops cancriformis and Notonecta maculata suggesting that these defenses are widespread among Daphnia.However, no increase of dorsal and ventral SBAs, nor of spinule lengths of D. magna have been found in response to kairomones of a fish species, Leucaspius delineatus, suggesting that microstructure induction might also be a predator-specific adaptation(Diel et al., 2021).Similarly, in D. cucullata, we found no dorsal or ventral spinules in any of the treatments, including the control and the two predator treatments.This suggest that also spinules are not expressed in all Daphnia species.However, we detected that spinules were present in neonates of D. cucullata in all treatments (data not shown), suggesting that the expression of microdefenses likely depends on developmental stage.Such stage-specificity has been shown for neck teeth development of D. pulex in response to Chaoborus kairomones(Tollrian, 1995).However, the strongest evidence for microstructures as independent modules of the defense strategy and not merely by-products of other defensive structures is the F I G U R E 4 Relationships among various residual mean traits.The colored letters in the figure correspond to three treatments, Bythotrephes (B), Control (C), and Leptodora (L) showing the mean of the clone among three Daphnia species.The r values in each box in the figure displays the Pearson correlation coefficient for the respective pair of residual traits, and asterisks indicate significant co-variation between the respective traits (*p < .05,**p < .01,***p < .001(Ross, 2017)).
more variability in both residual head and spina sizes, possibly indicating a positive relationship for two species (D. galeata and D. longispina).Of course these data, with only five genotypes, are too sparse to draw firm conclusions via statistics, but our core analyses highlight substantial genetic variation and across species variation suggesting that increasing the number of genotypes assayed can help reveal the context dependency of trait relationships.Such an effort could start to shed light on how trade-offs and positive relationships among predator defense traits are tied to coexistence among multiple prey species sharing two predators.This is the first study which explores induced defenses of Daphnia in a community context involving multiple Daphnia prey species, and multiple predators-all occurring in one lake ecosystem.Overall, our work demonstrates the presence of dorsal and ventral SBA microdefenses, highlighting the complexity of defensive mechanisms and their multifaceted contributions to survival strategies.Our study further reveals a captivating contrast between two Daphnia species.D. cucullata has a big helmet but no microdefenses.In contrast, D. longispina deploys microdefenses rather than macromorphological ones, which is consistent with modularity theory, but challenges the view that microdefenses are mere developmental byproducts of other induced morphological changes.The patterns in our data among three prey species and two predator species reveal a multifaceted set of strategies linking macro-morphological characteristics and smaller-scale defenses, which is further modified by the potential role of DVM behavioral responses in D. longispina.Clearly, further investigation is warranted to unveil the strategies defined across multiple types of induced traits among multiple species facing multiple predators.AUTH O R CO NTR I B UTI O N S Marjohn Yucada Baludo: Conceptualization (equal); data curation (equal); formal analysis (equal); investigation (equal); methodology (equal); writing -original draft (lead); writing -review and editing (equal).Pelita Octorina: Data curation (equal); investigation (equal); methodology (equal); writing -review and editing (equal).Andrew Beckerman: Formal analysis (equal); writing -review and editing F I G U R E 5 Relationship between residual head size and residual spina size among clonal means of each species in Lake Constance separately shown for the three predator treatments (Bythotrephes (B), Control (C), and Leptodora (L)).The lines represent the linear regressions of each species' residual head size as a function of its residual spina size.