DOES BIOTURBATION BY THE TADPOLE SHRIMP LEPIDURUS PACKARDI

PROMOTE CRUSTACEAN ABUNDANCE AND TAXONOMIC RICHNESS IN

CALIFORNIA VERNAL POOLS?

A Thesis

Presented to the faculty of the Department of Biological Sciences

California State University, Sacramento

Submitted in partial satisfaction of the requirements for the degree of

MASTER OF SCIENCE

In

Biological Sciences

by

Russell Croel

SPRING 2014

DOES BIOTURBATION BY THE TADPOLE SHRIMP LEPIDURUS PACKARDI

PROMOTE CRUSTACEAN ABUNDANCE AND TAXONOMIC RICHNESS IN

CALIFORNIA VERNAL POOLS?

A Thesis

by

Russell Croel

Approved by:

______, Committee Chair Jamie Kneitel, Ph.D.

______, Second Reader Ronald M. Coleman, Ph.D.

______, Third Reader James W. Baxter, Ph.D.

______Date

ii

Student: Russell Croel

I certify that this student has met the requirements for format contained in the University format manual, and that this thesis is suitable for shelving in the Library and credit is to be awarded for the thesis.

______, Graduate Coordinator ______Jamie Kneitel, Ph.D. Date

Department of Biological Sciences

iii

Abstract

of

DOES BIOTURBATION BY THE TADPOLE SHRIMP LEPIDURUS PACKARDI

PROMOTE CRUSTACEAN ABUNDANCE AND TAXONOMIC RICHNESS IN

CALIFORNIA VERNAL POOLS?

by

Russell Croel

Ecosystem engineers are increasingly recognized for their potential in facilitating habitat restoration efforts. An example of ecosystem engineering in aquatic habitats is bioturbation, the disruption of sediment at the water-sediment interface by animal activity. Among the varying effects they have on aquatic communities, bioturbating animals can facilitate zooplankton recruitment by digging up buried, resting eggs and returning them to the sediment surface, where they have a higher probability of hatching.

Such facilitation has been demonstrated in studies involving lake and permanent-pond ecosystems, but the effects of bioturbation in temporary ponds, such as California vernal pools, have largely been overlooked. Vernal pools are home to a strong bioturbator, the endemic notostracan Lepidurus packardi. I hypothesized that bioturbation by L. packardi facilitates the hatching of buried, resting eggs by returning them to the sediment surface. I tested this hypothesis by removing L. packardi from mesocosms filled with natural vernal pool soil and comparing the resulting crustacean communities to those in unmanipulated

iv

mesocosms. I predicted that mesocosms with fewer L. packardi would have fewer crustacean individuals and/or taxa in the active community, because fewer buried eggs would be returned to the sediment surface. I directly tested L. packardi’s digging abilities by conducting complementary microcosm experiments where I buried propagules

(resting eggs and plant seeds) at different depths and added freely roaming or caged L. packardi. These experiments also allowed me to determine whether L. packardi can influence the hatching of resting eggs through kairomones (chemical signals).

I found no support for my hypothesis. In the mesocosm experiment, four taxa were actually more abundant, not less, in mesocosms with fewer L. packardi. This indicates that L. packardi was suppressing these taxa in the Control mesocosms, most likely through predation. In the microcosm experiments, I found that L. packardi did not translocate propagules buried ≥ 0.5 cm deep, and that it also consumed eggs (but not seeds) lying on the sediment surface. I further found no evidence for kairomones. Results from the microcosm experiments additionally suggest that i) egg translocation was not masked by egg predation; and ii) propagule translocation simply did not occur. I conclude that bioturbation by L. packardi does not facilitate crustacean recruitment in California vernal pools, and that this taxon influences other crustacean taxa primarily through predation on both resting and active stages.

______, Committee Chair Jamie Kneitel, Ph.D.

______Date

v

ACKNOWLEDGEMENTS

I thank the outstanding faculty and staff of the Department of Biology for providing me with support and encouragement throughout my academic career at CSUS.

I especially thank Dr. Ronald Coleman and Dr. James Baxter for their constructive comments and criticisms of this research. Their guidance and coursework have made me a better scientist, teacher, and writer, and I am grateful that they were on my committee.

Dr. Jamie Kneitel, my advisor, deserves singular recognition. He exemplifies what it means to be a great mentor and ecologist. I feel honored to know him and to be part of his scientific pedigree.

None of this would have been possible without the unending support of my wife. Not only did she encourage me and provide me with uninterrupted study time over the years, but she also helped me collect data for this project when I was sidelined by a leg injury.

In a very literal sense, I could not have completed this project without her.

I dedicate this to my father, Philip Miles Croel.

vi

TABLE OF CONTENTS

Page

Acknowledgements...... vi

List of Tables ...... viii

List of Figures...... ix

INTRODUCTION ...... 1

METHODS ...... 8

Mesocosm Experiment...... 8

Microcosm Experiments ...... 15

RESULTS ...... 21

Mesocosm Experiment...... 21

Microcosm Experiments ...... 28

DISCUSSION...... 30

Literature Cited ...... 43

vii

LIST OF TABLES

Tables Page

1. Summary of crustacean taxa abundance (excluding L. packardi)

observed in each treatment group ...... 22

2. Independent-samples t-test results for evenness, total abundance,

and per-taxon abundances in mesocosm experiment...... 25

viii

LIST OF FIGURES

Figures Page

1. Number of L. packardi captured at each treatment application...... 11

2. NMDS ordination of mesocosms...... 23

3. Comparison of Bosmina sp., Cypris sp., and Limnocythere

ceriotuberosa abundances in Control (unmanipulated) mesocosms

vs. mesocosms from which L. packardi was removed weekly...... 26

4. ANCOVA comparison of Eucypris sp. abundance in Control (unmanipulated)

mesocosms vs. mesocosms from which L. packardi was removed

weekly...... 27

5. Comparison of nauplii and seedling abundances in microcosm

experiments ...... 29

ix

1

INTRODUCTION

Ecosystem engineers are organisms that modify their physical habitat in ways that affect resource availability to other organisms (Jones et al., 1994). Such habitat modification can promote biodiversity by creating habitat space or ameliorating abiotic stresses. For example, beaver dams can increase habitat complexity in riparian zones, thereby increasing plant species richness (Bartel et al., 2010), and seaweed canopies can reduce thermal and desiccation stresses in intertidal zones, enhancing recruitment and survival of intertidal organisms (Bertness et al., 1999). Because of the positive impacts ecosystem engineers can have on biodiversity, they are increasingly being recognized for their potential in facilitating habitat restoration efforts (Byers et al., 2006).

A key goal of habitat restoration is re-establishing biodiversity in ecosystems degraded by human activity (Palmer et al., 1997). An obvious first step in using ecosystem engineers in restoration efforts is identifying engineer species and assessing how their activities affect the community. If an engineer species is found to positively influence the establishment or persistence of other species, its focused use in restoration efforts could contribute to the success of those efforts. For example, lakes that are subject to nutrient loading from human activity can shift from a pristine, clear-water state to a turbid, algae-dominated state. Planting macrophytes in such lakes can facilitate a return to the clear-water state (Byers et al., 2006), in part because macrophytes provide zooplankton, which consume algae, refuge from predators.

One example of ecosystem engineering is bioturbation. In aquatic habitats, bioturbation is the disruption of sediment at or below the water-sediment interface due to

2 burrowing and foraging by animals. Species that disrupt the sediment can be large in size, such as sting rays and manatees, but small benthic invertebrates are considered the dominant bioturbators in aquatic ecosystems, due largely to their sheer numbers

(Meysman et al., 2006). Such taxa include annelids (Mermillod-Blondin and Lemoine,

2010), echinoderms (Lohrer et al., 2004), and crustaceans (Gyllström et al., 2008).

Bioturbation can have both positive and negative effects on other aquatic species. On the negative side, suspended sediment resulting from bioturbation can reduce macrophyte cover by creating turbid conditions that occlude light (Croel and Kneitel, 2011).

Suspended sediment can also clog the feeding apparati of suspension feeders, such as bivalves (Pillay et al., 2007) and cladocerans (Kirk, 1991). Further, bioturbation by relatively large invertebrates can bury or displace smaller invertebrates, limiting their access to food resources (Brenchley, 1981; Pillay et al., 2007). On the positive side, bioturbation has been shown to enhance nutrient cycling (Mermillod-Blondin and

Rosenberg, 2006), contribute benthic food resources to food webs (Creed et al., 2010), and increase the densities of epibenthic invertebrates by changing sediment topography

(Sun and Fleeger, 1994). Bioturbation can also stimulate macrophyte growth by mixing relatively oxygen-rich water into anoxic sediments (Mermillod-Blondin and Lemoine,

2010).

One way that bioturbating animals can positively affect zooplanktonic taxa is by digging up their buried, resting eggs and returning them to the sediment surface (Marcus and Schmidt-Gengenbach, 1986). Resting eggs are embryos that undergo a period of obligate dormancy called diapause. They are produced in large numbers by many

3 zooplanktonic taxa, such as small crustaceans in highly variable habitats. Resting eggs accumulate in the sediment and form an egg bank, a process analogous to the formation of seed banks in terrestrial habitats (De Stasio, 1989; Cáceres, 1998; Brendonck and De

Meester, 2003). Crustacean eggs in diapause can remain viable for years, and even centuries (Hairston et al., 1995; Cáceres, 1998; Frisch et al., 2014). The eggs hatch, or break diapause, when abiotic cues such as photoperiod, temperature, and salinity collectively signal favorable environmental conditions (Brendonck, 1996; Brendonck and

De Meester, 2003; Vandekerkhove et al., 2005). Eggs must be at or near the sediment surface to experience the full effects of these cues (Cáceres, 1998). Eggs covered by as little as 0.5 cm of sediment become insulated from hatching cues and generally will not hatch (Brendonck and De Meester, 2003; Gleason et al., 2003). Bioturbation can positively affect crustacean abundance by translocating these buried-but-viable resting eggs to the sediment surface, where they have a much higher probability of hatching

(Marcus and Schmidt-Gengenbach, 1986; Brendonck and De Meester; 2003; Gyllström et al., 2008).

In addition to potentially increasing crustacean abundance, bioturbation might also increase taxonomic richness in the active community (Brendonck and De Meester, 2003).

Some sediments exhibit greater richness, in the form of diapausing eggs, than the active zooplankton community in any given season (Hairston and Kearns, 2002; Brendonck and

De Meester, 2003; Vandekerkhove et al., 2005). This mismatch can be attributed to a life- history strategy called bet hedging, which posits that the hatching of a relatively small fraction of eggs prevents all offspring from dying should abiotic conditions lead to

4 abortive hatchings (i.e., hatched individuals do not survive long enough to reproduce;

Hildrew, 1985; Simovich and Hathaway, 1997; Belk, 1998; Vanschoenwinkel et al.,

2010). Cumulative hatching fractions for some temporary-pond crustaceans, for example, have been found to be as low as 3% (Hildrew, 1985; Belk, 1998; Simovich and

Hathaway, 1997). Although hatching fractions vary widely (Brendonck, 1996), in general resting eggs accumulate faster than they hatch (De Stasio, 1989; Cáceres, 1998).

Bioturbation might contribute to taxonomic richness of the active aquatic community by promoting recruitment of species that would otherwise remain in diapause.

Several studies have demonstrated that bioturbation can return buried, resting eggs back to the sediment surface. Marcus and Schmidt-Gengenbach (1986), for example, buried copepod eggs and found that a polychaete bioturbator returned a sizeable fraction

(~15%) of them to the sediment surface. The eggs hatched and were recovered from the water as nauplii. Consequently, Marcus and Schmidt-Gengenbach (1986) suggested that bioturbation may be an important recruitment mechanism for zooplanktonic taxa. Kearns et al. (1996), using small, plastic beads as proxies for eggs, reported similar upward translocation, primarily due to bioturbation by oligochaetes. More recently, Gyllström et al. (2008) showed that bioturbation by an amphipod increased the taxonomic richness of zooplankton in the active community. They attributed this result to the amphipod digging up buried, resting eggs, which facilitated their hatching. Other studies have shown that the translocation of resting stages is not limited to crustaceans. Ståhl-Delbanco and

Hansson (2003), for instance, showed that recruitment of dormant algae cells in the sediment was dramatically higher in the presence of an amphipod bioturbator.

5

The effects of bioturbation on crustacean communities have been examined primarily in marine, lake, and permanent-pond ecosystems. Few studies have examined the effects of bioturbation in temporary ponds, let alone the effects specifically on crustacean communities in these unique ecosystems. Temporary ponds occur where seasonally variable climatic conditions produce annual cycles of inundation and desiccation. These cycles result in distinct aquatic and terrestrial communities in pond basins at different times of year. Temporary pond ecosystems tend to be characterized by high levels of species richness and endemism and are generally considered biodiversity hotspots

(Holland and Jain, 1981; King et al., 1996; Semlitsch and Bodie, 1998; Simovich, 1998).

In Mediterranean climate regions, such as the Central Valley of California, temporary ponds are known as vernal pools. California vernal pools contain many plant and invertebrate species that are listed as threatened or endangered (Federal Register, 2003).

Despite their global ubiquity (Keeley and Zedler, 1998), temporary pond systems worldwide are in steady decline due to habitat loss and invasive species (Holland, 1978;

Blaustein and Schwartz, 2001; Brinson and Malvárez, 2002; De Meester et al., 2005). In

California, for example, only one-tenth of the historical expanse of vernal pool habitat in the Central Valley remains due to agriculture and urbanization (Holland, 1978; Holland,

1998). The loss of vernal pools combined with their high levels of endemism makes this habitat a focus of conservation efforts in the state. Understanding the factors that contribute to and maintain biodiversity in vernal pools is a pre-requisite for their successful management and restoration (Simovich, 1998). For example, knowing whether

6 one species was particularly important in maintaining crustacean diversity in California vernal pools could inform and help guide management and restoration strategies.

The tadpole shrimp Lepidurus packardi (Crustacea: Branchiopoda: Notostraca:

Triopsidae) is a benthic omnivore that grows up to 8 cm long. It is endemic to California vernal pools and, like other members of the family Triopsidae (Yee et al., 2005;

Waterkeyn et al., 2011b), is a strong bioturbator (Croel and Kneitel, 2011). It occurs as far north as Shasta County and as far south as Tulare County, with the most known occurrences in Sacramento County (U.S. Fish and Wildlife Service, 2007). It is also one of the four large branchiopods federally listed as endangered (Federal Register, 2003).

Lepidurus packardi forages by sifting the sediment surface and underlying layers for food items, often digging shallow depressions (~0.5 cm deep; personal observation) in the process. Because of this bioturbation, L. packardi is recognized as an ecosystem engineer

(Croel and Kneitel, 2011). Although the presence of L. packardi is associated with higher crustacean taxonomic richness compared to pools without L. packardi (King et al., 1996;

Croel and Kneitel, 2011), the impact of its bioturbation on the crustacean community in

California vernal pools has not been experimentally tested.

The goal of this study was to experimentally test the hypothesis that bioturbation by

Lepidurus packardi positively affects other crustacean taxa in California vernal pools. I tested this hypothesis by reducing the density of L. packardi in vernal pool mesocosms and comparing the resulting crustacean communities to those in unmanipulated mesocosms. I predicted that mesocosms with reduced L. packardi densities would have lower total crustacean abundance, taxonomic richness, and per-taxon abundances,

7 because fewer buried eggs would be returned to the sediment surface in these mesocosms. I also compared water physicochemisty between treatments to assess how bioturbation by L. packardi affected general abiotic conditions in the mesocosms. I conducted complementary microcosm experiments in which I buried fairy shrimp eggs at different depths. These microcosm experiments allowed me to directly test L. packardi’s digging and egg-translocation abilities. They also allowed me to assess whether L. packardi influenced resting-egg hatching via kairomones (chemical signals released by predators; Lass and Spaak, 2003), which otherwise might be a confounding factor in this study. Combined, these microcosm experiments would help me interpret the results of the mesocosm experiment and shed greater light on how L. packardi affects other crustacean taxa in California vernal pools.

8

METHODS

Mesocosm experiment

In December 2011, I arranged 20 mesocosms in a 4 × 5 array in an outdoor area on the campus of California State University, Sacramento. A white, canvas drop cloth (3.7 ×

4.6 m) was placed beneath the array for weed control. The mesocosms were black, circular, plastic pond liners (diameter = 0.6 m, depth = 0.18 m, volume = 51 L). I filled each mesocosm with 10 kg of dry soil collected from vernal-pool basins in the Elder

Creek watershed of Sacramento County (Kneitel and Lessin, 2010). This amount of soil yielded a soil layer about 4 cm deep. The soil surface in each mesocosm was smoothed by hand to ensure an even depth throughout. Before being added to the mesocosms, the soil was mixed in a cement mixer to homogenize the egg bank. Previous studies using this soil (Kneitel and Lessin, 2010; Croel and Kneitel, 2011) demonstrated that it contained eggs for many of the crustacean taxa typical of California vernal pools, e.g., cladocerans (water fleas), ostracods (seed shrimp), copepods, anostracans (fairy shrimp), and notostracans (tadpole shrimp).

Due to lack of natural rainfall, I inundated the mesocosms with well water from a hose on 1 January, 2012, to initiate the experiment. I filled each mesocosm to the top using a gentle spray. To prevent the spray from disrupting the soil, I covered the surface of the soil with bubble wrap prior to filling (Roast et al., 2004). After this initial filling, water levels in the mesocosms were regulated solely by natural weather conditions. The mesocosms always contained ample water during the experiment, never dropping below

9 about two-thirds full. Throughout the experiment, floating leaves from nearby trees were removed by hand when observed in the mesocosms. Submerged leaves were left in place.

The experiment included two treatment groups: an unmanipulated group (hereafter, the Control group) and an experimental group in which I reduced the density of L. packardi (hereafter, the Removal group). Two replicates of each group were placed in an alternating arrangement in each row of four mesocosms. I removed L. packardi from the

Removal group by gently sweeping the water column with a 15 × 20 cm, 500-µm mesh aquarium net. The sweeping procedure consisted of one circular sweep counterclockwise, one circular sweep clockwise, and then one linear sweep through the diameter, keeping the bottom of the net as close as possible to the sediment surface without touching it. The contents of the net were transferred to a shallow pan, in which I located and removed L. packardi individuals via forceps or pipette. The remaining contents of the pan were returned to the source mesocosm. The Control group was not manipulated, except that I applied the sweeping procedure described above to these mesocosms as a disturbance control. The contents of the net underwent the same handling as in the Removal group, but all individuals, including L. packardi, were returned to the mesocosms. Treatments were applied once per week beginning on 27 January, which was when I first observed nauplii in the mesocosms. The final treatment was applied on 18 March, for a total of eight treatments. I performed the sweeping procedure once in each mesocosm for the first four treatments. Thereafter, I repeated the procedure twice in each mesocosm to increase capture efficiency.

10

Croel and Kneitel (2011) used the same sweeping procedure and were able to greatly reduce L. packardi densities in their mesocosms, in most cases to zero. Although I did not necessarily expect to achieve complete removal of L. packardi in the present study, I anticipated removal efficiencies comparable to Croel and Kneitel (2011). This was not the case, however. During most treatment applications, the number of tadpole shrimp removed from the Removal mesocosms approximated the number captured in the Control mesocosms (Figure 1). This indicates that tadpole shrimp were hatching continuously from resting eggs in the Removal mesocosms, which was unexpected. Although L. packardi abundance was similar in both treatment groups for most of the experiment, this does not necessarily mean that the two groups were identical with respect to L. packardi.

Compared to the Control group, individuals in the Removal group were generally smaller because most were ~7 days old or less. In any event, not being able to achieve complete removal of tadpole shrimp makes the experiment more conservative, as it lessens any differences between the groups.

On 9 January, seven of the 20 mesocosms (three Control replicates and four Removal replicates) were visited by waterfowl. I excluded these seven mesocosms from the experiment because their sediments had been substantially disrupted, which could have confounded my results. To maintain a balanced design, I randomly chose one of the remaining Control replicates and also excluded it from the experiment. As a result, all analyses in this study are based on six replicates per group rather than 10. To prevent further visitation by waterfowl, I covered the mesocosm array with fine-filament bird netting (1.2-cm aperture mesh), suspended 1.8 m above the array. The netting was left in

11

Figure 1. Number of L. packardi captured at each treatment application. After being counted, L. packardi individuals were returned to mesocosms in the Control group, but were not returned to mesocosms in the Removal group. In the Control group, the number captured reflects the population size on that date. In the Removal group, the number captured is the number of L. packardi permanently removed from the mesocosms on that date. The first treatment was applied on 27 January, but this date is not shown because no

L. packardi were captured. The black arrow at 18 March indicates the final treatment application. The gray arrow at 24 March indicates the end of the experiment, when all mesocosms were destructively sampled. Error bars show ± 1 SD.

12 place for the remainder of the experiment. On two other occasions I observed evidence of animal visitation in the form of paw prints on mesocosm rims. However, there was no evidence that the sediment in the visited mesocosms had been disturbed, and so these mesocosms were retained in the study.

On 23 March, one day prior to terminating the experiment, I collected water physicochemistry data. I used an Oakton pH/CON 300 Meter to measure water temperature (oC), conductivity (µS), and pH, and an Oakton pH/DO 300 Meter to measure dissolved oxygen (mg/L). These measurements were taken in situ in the early afternoon. I also collected 50-mL water samples for laboratory analyses of turbidity

3- - (NTU), PO4 (as orthophosphate, mg/L), and NO3 (mg/L). Lab analyses were completed within 8 h of sample collection. All water samples were first filtered through a 243-µm sieve to remove coarse particulates. I then used a LaMotte 2020i Turbidity Meter to

3- - quantify the turbidity of the filtered samples. For the PO4 and NO3 analyses, I further processed the water samples by centrifuging them at 4000 rpm for 10 minutes to pellet the suspended sediment, which can interfere with testing reagents. The supernatant was then syringe-filtered through a 0.45-µm nylon membrane. I used a Hach DR2800

3- - spectrophotometer to quantify PO4 (via Method 8048) and NO3 (via Method 8171) in the filtered supernatant.

I destructively sampled the mesocosms for crustaceans on 24 March by performing the previously described sweeping procedure four times in each mesocosm. I counted L. packardi individuals and returned them to the mesocosms. All other crustaceans were preserved in 95% ethanol. I sorted and counted the preserved crustaceans under a 25x

13 dissection microscope. I identified crustaceans to the lowest taxonomic level possible according to Thorp and Covich (2010). For ostracods, I also used a 100x light microscope as well as the identification key in Pennak (1989). Due to the large number of crustaceans captured (~20,000 individuals), I did not key-out each individual. Rather, I used the identification keys to identify three to five individuals per taxon per mesocosm, and then

I sorted the remaining individuals based on morphological similarity to the identified taxa. Most taxa were identified to genus; two were identified to species. I did not collect data for non-crustacean invertebrates, but the abundance of such taxa in the preserved samples was low (~5 or fewer total individuals per mesocosm).

I conducted an Analysis of Similarities (ANOSIM) to test for a treatment effect on crustacean-community similarity among treatments. This analysis used 5000 permutations and was based on Bray-Curtis dissimilarities calculated from a matrix of taxa abundances in each mesocosm. I interpreted a significant result with a Similarity

Percentage (SIMPER) analysis to reveal which taxa were contributing most to the dissimilarity between treatments. To visualize the communities in ordination space, I used a 2D non-metric multidimensional scaling (NMDS) plot, which was based on the same data matrix as the ANOSIM. I saved the dimension scores from this plot and used

Pearson’s correlations to examine relationships between these scores and water physicochemistry variables.

I conducted additional tests to determine whether specific aspects of community structure were affected by L. packardi removal. I used independent-samples t-tests to compare evenness and total crustacean abundance between treatments. The measure of

14 evenness used was Pielou’s evenness (E), calculated as E = H/lnS, where H is Shannon’s

Diversity Index (−Σpi[lnpi], where pi is the proportion of individuals of taxon i relative to the total number of individuals in all taxa) and S is taxonomic richness (Pielou, 1966).

Total abundance was the total number of all crustacean individuals in each mesocosm. I did not compare taxonomic richness because both treatments had a mean richness of 10.

To determine how individual taxa responded to reduced L. packardi density, I conducted an independent-samples t-test on the abundance of each taxon, with the exceptions of

Cypris sp. and Eucypris sp. Data for Cypris sp. did not satisfy parametric assumptions, even after transformation, and so were analyzed non-parametrically via the Mann-

Whitney U-test. Eucypris sp. abundance was compared with Analysis of Covariance

(ANCOVA), using water temperature as the covariate, because preliminary analyses showed that this taxon’s abundance was strongly correlated with water temperature in the mesocosms (Pearson’s r = −0.85, P = 0.001). Prior to this analysis, I confirmed that

ANCOVA assumptions were satisfied (i.e., homogeneity of regression slopes and covariate independence of treatment effect).

I used Multivariate Analysis of Variance (MANOVA) to compare abiotic conditions

(i.e., water physicochemistry) between treatments. Dependent variables were turbidity,

3- - pH, conductivity, and water temperature. Data for PO4 and NO3 concentrations were invariable and thus excluded from this analysis. Dissolved oxygen was strongly correlated with water temperature (Pearson’s r = 0.84, P = 0.001) and so was also excluded to preserve degrees of freedom.

15

Lepidurus packardi was excluded from all analyses. Two other taxa, Moina sp. and

Pseudocandona sp., were also excluded from all analyses because they were extremely rare in the mesocosms in both abundance and occurrence. Prior to the parametric analyses, I used Shapiro-Wilk’s test and Levene’s test to check data for normality and equal variance, respectively. Data were square-root or log transformed as needed to satisfy parametric assumptions. All figures show untransformed data for clarity. I used

PAST 1.94b (Hammer et al., 2001) for the analyses based on Bray-Curtis dissimilarities.

All other analyses were conducted in SPSS 21.0.

Microcosm experiments

I conducted three microcosm experiments that directly tested whether L. packardi could dig up buried resting eggs. I also considered whether plant seeds could be translocated, as an added measure. This allowed me to assess whether propagules (i.e., both eggs and seeds) in general could be translocated by L. packardi. These experiments also tested for any kairomone-related influences exerted by L. packardi. Accounting for the potential influence of kairomones was necessary because they might otherwise confound the mesocosm results.

All three microcosm experiments were set up at the same time and conducted simultaneously. These experiments used clear, plastic, food-storage containers (14 cm square × 8 cm tall, volume = 1.5 L) as microcosms. Each was filled with 350 mL (~ 440 g) of dry soil collected from vernal-pool basins in the Gill Ranch Conservation Bank in

Sacramento County. This amount of soil yielded a soil layer about 2 cm deep. I glided a

16 straight edge over the soil surface to make it smooth and uniform. I then sprinkled 7 mg of fairy shrimp resting eggs (~2000 eggs) as evenly as possible on the soil surface. The eggs belonged to the freshwater taxon Streptocephalus sp. and were obtained from a commercial supplier (www.arizonafairyshrimp.com). Preliminary studies showed that these eggs hatched quickly upon rehydration (within 24 h) under laboratory conditions.

Because of time constraints, using this quick-hatching taxon as an indicator of egg translocation was more practical than relying on the soil’s natural egg bank, as crustacean eggs in California vernal pools generally take at least 2 – 4 weeks to hatch after inundation (Ahl, 1991; Gallagher, 1996; personal observation). This constraint did not apply to seeds because germination of some vernal-pool plant species is rapid (≤ 3 days;

Bliss and Zedler, 1998; personal observation). Also, preliminary studies confirmed that seeds in the soil used in these experiments reliably germinated within a few days following inundation. It was thus more practical to leverage the seeds naturally present in the soil rather than add them.

In one microcosm experiment, the propagules were allowed to remain at the soil surface (the 0 cm-burial experiment). In the other two experiments, I buried the propagules under 0.5 cm- and 1 cm-deep layers (respectively) of vernal pool soil that had been autoclaved for 90 min. The autoclaved soil acted as a barrier to hatching and germination once the microcosms were inundated (Gleason et al., 2003; personal observation). The 0.5 cm-burial and 1 cm-burial experiments allowed me to directly test whether L. packardi could dig through the soil barrier and translocate propagules to the sediment surface. To create the 0.5-cm and 1-cm deep layers, I added 100 mL (~114 g)

17 and 200 mL (~228 g), respectively, of autoclaved soil atop the propagules. The autoclaved soil layer was smoothed with a straight edge to achieve as uniform a depth as possible.

After adding the eggs and autoclaved soil, I laid a piece of bubble wrap atop the soil

(Roast et al., 2004) and then gently inundated each microcosm with 1 L of deionized tap water. The bubble wrap prevented the spray from disrupting the soil. Once the microcosms were inundated, I added the treatment, which was the influence of L. packardi. Levels for this treatment were L. packardi present, absent, and caged. Each treatment was replicated four times, for a total of 12 replicates per experiment. For the L. packardi-present treatment, I added one L. packardi to each replicate. The movement of

L. packardi in these replicates was unrestricted; individuals were allowed to freely roam and bioturbate. The L. packardi-absent treatment had no L. packardi added. For the L. packardi-caged treatment, one L. packardi was added to each replicate but was sequestered in a 6 × 6.5 cm, 53-µm mesh cage constructed from an aquarium net. The cage was suspended above the sediment in one corner of the microcosm and prevented L. packardi from contacting the sediment. The handle of the cage was bent over the corner of the microcosm and taped to the outer sidewall. The cage’s mesh was fine enough to prevent nauplii from swimming into the cage, where they might be eaten by L. packardi.

Sequestering L. packardi in cages allowed me to test for kairomone-related influences this taxon may have on propagules (Waterkeyn et al., 2012). The L. packardi used for these experiments had a mean (± SD) carapace length of 1.19 ± 0.13 cm and were obtained from surplus mesocosms immediately before being added to the microcosms.

18

I prepared two additional groups of microcosms as controls. I included an autoclave- control group to ensure that the autoclaved soil added atop the propagules contained no viable eggs or seeds of its own. This group contained two replicates, each filled with 350 mL of autoclaved soil only. A cage-control group, containing four replicates, was also included. The microcosms in this group were identical to the L. packardi-caged replicates in the 0 cm-burial experiment, except that no L. packardi were added to the cages. This controlled for any effect of the cage itself on hatching or germination.

Once all 42 replicates (12 replicates in each of three experiments, plus six control replicates) were filled and all L. packardi added, I placed the microcosms in an environmental chamber and maintained them under a 24oC/14 h light:19 oC/10 h dark regime. Nine replicates (one replicate of each L. packardi treatment from each experiment) were placed on each of four shelves. The six cage- and autoclave-control replicates were placed on a fifth shelf.

I removed all L. packardi from the microcosms after 24 h. To remove them from the

L. packardi-present replicates, I plucked them out with forceps, taking care not to touch the sediment. I briefly dipped the forceps in all other microcosms as a control. For the L. packardi-caged replicates, I detached and removed the cages from the microcosms. I removed the cages from the cage-control replicates at the same time. I held each cage over a separate container and rinsed the mesh with water to dislodge any nauplii that may have been stuck to the mesh. I visually inspected the rinse water for nauplii, and if any were observed I added them to the total number of nauplii captured at the end of the experiment (I observed nauplii in the rinse water for only one microcosm). Seven of the

19

12 L. packardi individuals in the replicates with cages did not survive the 24-h period, presumably due to starvation. One individual from an L. packardi-present replicate also did not survive. I returned the microcosms to the environmental chamber once all L. packardi and cages had been removed.

Three days after removing L. packardi, I siphoned the water of each microcosm through a 53-µm sieve to capture fairy shrimp nauplii. The siphon used 5-mm (internal diameter) plastic tubing. Nauplii were immediately preserved in 95% ethanol and counted under a 25x dissection microscope. The siphoning removed all standing water, but the soil remained saturated and was conducive to seed germination. Accordingly, I returned the microcosms to the environmental chamber for an additional eight days, after which I counted the number of seedlings in each microcosm.

In the 0-cm burial experiment, I predicted that nauplii and seedlings would be similar in abundance across L. packardi treatments, because the propagules would be fully exposed to hatching and germination cues. In the 0.5-cm burial experiment, I predicted that nauplii and seedlings would be most abundant in microcosms containing freely roaming L. packardi. In these microcosms, I expected L. packardi to dig up many of the buried propagules and translocate them to the sediment surface, where they would hatch/germinate. I expected no nauplii or seedlings in the L. packardi-caged or L. packardi-absent treatments, because the propagules would stay buried and hence remain insulated from emergence cues (Gleason et al., 2003; personal observation). Similarly, I predicted that no nauplii or seedlings would be found in the 1-cm burial experiment, reasoning that this is too deep for L. packardi to dig (Gleason et al., 2003; personal

20 observation). I further expected to find no evidence of kairomones in these experiments

(Waterkeyn et al., 2012).

I used one-way ANOVAs to analyze abundance data for each experiment. These tests used L. packardi treatment (present, absent, or caged) as the fixed factor and nauplii abundance or seedling abundance as the dependent variable. Data for the 0.5 cm-burial experiment were square-root transformed to meet parametric assumptions. These analyses were conducted in SPSS 21.0.

21

RESULTS

Mesocosm experiment

The preserved samples contained 19,690 crustaceans belonging to 12 taxa (Table 1).

Most taxa were found in all six mesocosms of each treatment group, with three exceptions. Moina sp. occurred in just one mesocosm, Pseudocandona sp. occurred in three mesocosms (two Control replicates and one Removal replicate), and Simocephalus sp. occurred in four mesocosms of each treatment group. The number of crustacean taxa in each mesocosm ranged from 9 to 11, with a mean richness of 10 in both treatment groups.

Weekly removal of L. packardi from the Removal group had no significant effect on crustacean-community similarity between treatments (ANOSIM: R = 0.35, P = 0.12).

Because this result was not significant, I did not proceed with the SIMPER analysis.

Consistent with the ANOSIM result, the NMDS plot showed no discernable patterns in community composition related to L. packardi removal (Figure 2). The stress value for this ordination was low (stress = 0.091).

I found a strong positive relationship between Dimension 1 scores of the NMDS plot and water temperature (Pearson’s r = 0.88, P < 0.001). Dimension 1 scores were also strongly negatively related to the abundance of Eucypris sp. (Pearson’s r = −0.97, P <

0.001), which was the numerically dominant taxon in all mesocosms (Table 1). This indicates that variation in Eucypris sp. abundance was responsible for most of the dissimilarity observed along Dimension 1. No other taxon’s abundance was significantly related to Dimension 1 scores. Dissolved oxygen was also positively related to

22

Table 1. Summary of crustacean taxa abundance (excluding L. packardi) observed in each treatment group. For each group, n = 6. Removal group = mesocosms from which L. packardi was removed weekly. Control group = unmanipulated mesocosms. Total = total number of individuals captured.

Mean ± SE abundance

Taxonomic group Taxon Removal group Control group Total

Anostraca Linderiella occidentalis 25.7 ± 4.9 28.0 ± 4.3 322

Cladocera Bosmina sp. 15.5 ± 1.3 6.2 ± 1.6 130

Daphnia sp. 264.0 ± 89.0 408.7 ± 177.6 4036

Moina sp. 1 0 1

Simocephalus sp. 17.2 ± 9.8 15.7 ± 8.1 197

Copepoda Hesperodiaptomus sp. 24.2 ± 4.8 32.5 ± 2.9 340

Leptodiaptomus sp. 276.2 ± 47.1 288.0 ± 31.9 3385

Ostracoda Candona sp. 21.2 ± 5.6 15.8 ± 3.6 222

Cypris sp. 74.7 ± 9.3 34.5 ± 6.2 655

Eucypris sp. 941.8 ± 237.3 775.0 ± 181.0 10,301

Limnocythere ceriotuberosa 13.0 ± 4.4 3.3 ± 1.0 98

Pseudocandona sp. 1 2 3

Total 12 taxa 1674 ± 270 1608 ± 171 19,690

23

Figure 2. Non-metric multidimensional scaling ordination of mesocosms. The stress value for this ordination was 0.091.

24

Dimension 1 scores (Pearson’s r = 0.66, P = 0.020). Dimension 2 scores were related to turbidity (Pearson’s r = 0.74, P = 0.006) and Daphnia sp. abundance (Pearson’s r = 0.78,

P = 0.003).

Reducing L. packardi density had no significant effect on evenness or total crustacean abundance, but it did significantly affect the abundances of four taxa (Table 2). Contrary to my hypothesis, however, all four affected taxa were more abundant in mesocosms from which L. packardi had been removed. The affected taxa were the cladoceran

Bosmina sp. (t = 4.56, df = 10, P = 0.001; Figure 3A) and the ostracods Cypris sp. (U =

1.00, P = 0.004; Figure 3B), Limnocythere ceriotuberosa (square-root transformed, t =

2.37, df = 10, P = 0.039; Figure 3C) (Table 2), and Eucypris sp. (square-root transformed, F1,9 = 5.70, P = 0.041; Figure 4). Water temperature also had a significant effect on Eucypris sp. abundance (F1,9 = 45.00, P < 0.001). Adjusted for covariation with water temperature, Eucypris sp. abundance in the Control treatment was (back- transformed, estimated marginal mean ± 95% CI) 645.0 ± 182.8 individuals, compared to

949.8 ± 221.8 individuals in the Removal treatment.

Overall abiotic conditions in the mesocosms, as measured by the water physicochemistry variables, were not significantly different between treatments (Pillai’s

Trace = 0.57; df = 4,7; F = 2.32; P = 0.16).

25

Table 2. Independent-samples t-test results for evenness, total abundance, and per-taxon abundances in mesocosm experiment. Eucypris sp. abundance was analyzed separately because it covaried with water temperature. Taxonomic richness was not analyzed because both groups had the same mean richness. df = 10.

Variable t P-value

Evenness 0.55 0.60

Total abundance 0.21 0.84

Linderiella occidentalis 0.36 0.73

Bosmina sp. 4.56 0.001

Daphnia sp. 0.73 0.48

Simocephalus sp. a 0.05 0.97

Hesperodiaptomus sp. 1.48 0.17

Leptodiaptomus sp. 0.21 0.84

Candona sp. b 0.67 0.52

Cypris sp. c − 0.004

Limnocythere ceriotuberosa a 2.37 0.039

a square-root transformed b log transformed c Mann-Whitney U-test; U = 1.00

26

Figure 3. Comparison of Bosmina sp., Cypris sp., and Limnocythere ceriotuberosa abundances in Control (unmanipulated) mesocosms vs. mesocosms from which L. packardi was removed weekly. (A) Comparison of Bosmina sp. abundance. (B)

Comparison of Cypris sp. abundance (analyzed non-parametrically). (C) Comparison of

L. ceriotuberosa abundance. Panel C shows untransformed data for clarity. Error bars show ± 1 SE.

27

Figure 4. ANCOVA comparison of Eucypris sp. abundance in Control (unmanipulated) mesocosms vs. mesocosms from which L. packardi was removed weekly. This figure shows untransformed data for clarity.

28

Microcosm experiments

In the 0 cm-burial experiment, nauplii abundance was significantly lower in the L. packardi-present treatment compared to the L. packardi-caged treatment (F2,9 = 6.55, P =

0.018; Figure 5A). There was no significant difference in seedling abundance between treatments (F2,9 = 0.41, P = 0.68; Figure 5B).

In the 0.5 cm-burial experiment, burying propagules sharply reduced their hatching and germination. There was no significant difference between L. packardi treatments in either nauplii abundance (square-root transformed, F2,9 = 0.15, P = 0.87; Figure 5C) or seedling abundance (square-root transformed, F2,9 = 3.95, P = 0.059; Figure 5D), although the latter difference approached significance, trending toward more seedlings in the L. packardi-present treatment compared to the other two treatments.

I did not analyze data from the 1 cm-burial experiment because only 3 nauplii and 2 seedlings were observed among all 12 replicates.

No nauplii or seedlings were found in the control group containing autoclaved soil only, indicating that this soil contained no viable eggs or seeds. Also, in the cage-control group, there was no significant difference in either nauplii abundance (ANOVA: F1,6 =

0.015, P = 0.91) or seedling abundance (ANOVA: F1,6 = 4.31, P = 0.083) compared to the L. packardi-caged replicates in the 0 cm-burial experiment. This indicates that the cage itself did not affect hatching or germination.

29

Figure 5. Comparison of nauplii and seedling abundances in microcosm experiments. (A,

B) 0 cm-burial experiment. (C, D) 0.5 cm-burial experiment. Lepidurus packardi treatment: Present = freely roaming L. packardi, Absent = no L. packardi added, Caged =

L. packardi present but sequestered in cage. Nauplii and seedlings were counted 3 days and 11 days (respectively) after removing L. packardi from the microcosms. In A, different letters above bars indicate a significant difference between treatments (P =

0.018). All other differences are non-significant (P ≥ 0.059). C and D show untransformed data for clarity. Error bars show ± 1 SE.

30

DISCUSSION

I found no support for my hypothesis that bioturbation by L. packardi returns buried, resting eggs to the sediment surface and facilitates their hatching. In the mesocosm experiment, I predicted that removing L. packardi (or at least reducing its density) would reduce crustacean taxonomic richness, total abundance, and per-taxon abundances.

However, none of these variables decreased in the Removal treatment. Contrary to my hypothesis, in fact, four taxa were significantly more abundant in mesocosms where L. packardi density had been reduced. This shows that L. packardi was suppressing these taxa in the Control mesocosms. In the microcosm experiments, which were in part designed to directly test L. packardi’s digging ability, I found that L. packardi did not dig up propagules buried ≥ 0.5 cm deep. Overall, my results suggest that bioturbation by L. packardi does not facilitate the hatching of buried propagules. They also indicate that L. packardi’s influence on other crustacean taxa is generally negative.

There are at least five possible, non-mutually exclusive explanations for L. packardi’s suppression of the four taxa in the Control mesocosms. The first and most likely explanation is predation on these taxa by L. packardi. Although I know of no studies that have examined the feeding habits of L. packardi specifically, several studies have demonstrated that notostracans consume adult crustaceans (Christoffersen, 2001; Boix et al., 2006; Waterkeyn et al., 2011b) and resting eggs in the sediment (Waterkeyn et al.,

2011a). My own observations of L. packardi in aquaria verify that it too consumes adult crustaceans. The results of my 0-cm burial microcosm experiment show that L. packardi also consumes resting eggs. Given that tadpole shrimp are known predators of both active

31 and resting crustacean stages, it is likely that predation played a large role in suppressing the four affected taxa in the Control mesocosms. I cannot determine from my data whether resting or active stages were predominantly consumed, but it was probably the latter. If resting stages had mainly been consumed, I would expect all taxa to decline, because resting-egg consumption by tadpole shrimp appears to be indiscriminate

(Waterkeyn et al., 2011a). Further, three of the four affected taxa were ostracods, which, similar to L. packardi, are benthic and substrate feeders. It seems logical that L. packardi would preferentially consume active-stage ostracods, because it would encounter them more often than it would open-water taxa (but see Boix et al., 2006). Both Waterkeyn et al. (2011b) and Yee et al. (2005) showed that notostracans in the genus Triops negatively affected ostracod taxa, with the former study directly showing consumption of adult ostracods as the main cause. The decline of the cladoceran Bosmina sp. is more difficult to explain in terms of predation on active stages. Notostracans are able to capture and consume cladocerans (Christoffersen, 2001; Waterkeyn et al., 2011b), but why this cladoceran taxon alone was affected is unclear. It could be that this cladoceran simply spent more time near the sediment than other cladoceran taxa, and thus was disproportionately vulnerable to predation by L. packardi.

The second possible explanation for the reduced taxa abundances in the Control mesocosms is competition with L. packardi. Tadpole shrimp and ostracods share the same general feeding habit, i.e., foraging along the sediment surface. The larger size of L. packardi may have allowed it to more effectively compete for benthic food resources. In addition, L. packardi’s near-constant movement along and within the sediment could

32 have resulted in interference competition by displacing food items or ostracods directly

(sensu Brenchley, 1981). However, competitive interactions would probably not explain why a cladoceran taxon also declined in abundance, as cladocerans are pelagic suspension-feeders.

A third possible explanation is that the suspended sediment caused by L. packardi’s bioturbation may have clogged the feeding apparati of the affected taxa in the Control mesocosms (Newcombe and MacDonald, 1991). Such clogging has been shown to negatively affect certain suspension feeders, including cladocerans (Kirk, 1991).

However, in the present study only one suspension feeder (Bosmina sp.) out of six

(excluding Moina sp.) was negatively affected. If clogging were an issue, more suspension feeders would likely have been affected. The other taxa affected in this study were ostracods, which are not suspension feeders, although they do filter/strain food particles to some extent (Pennak, 1989). Several observational studies have reported a negative relationship between ostracods and suspended sediment (e.g., Cohen et al.,

1993; Ruiz et al., 2013), but it is unclear whether this was due to clogging or some other factor associated with suspended sediment (Donohue and Molinos, 2009).

The fourth possible explanation also involves suspended sediment, which may have inhibited the growth of planktonic and periphytic algae through light limitation. These algae are a major food resource for temporary-pond crustaceans, including notostracans

(Boix et al., 2006). This explanation is doubtful, however. Although the Control mesocosms had noticeably higher turbidity than the Removal mesocosms (Control-group turbidity [mean ± SD]: 268 ± 67 NTU; Removal-group turbidity: 196 ± 29 NTU; data not

33 shown), turbidity in the Control mesocosms was probably not high enough to have caused light limitation. The results of Croel and Kneitel (2011) support this idea. Using the same mesocosm containers, they found that it took over four times the turbidity observed in the present study (~1150 NTU vs. ~270 NTU, respectively) to cause only a relatively small (25%) reduction in macrophyte cover. Thus, light limitation in the

Control mesocosms was probably nominal compared to the Removal mesocosms.

Finally, although I found no evidence of kairomone-mediated hatching suppression in the microcosm experiments, strictly speaking this finding applies just to the anostracan eggs used in those experiments. Therefore, it is still possible that the eggs of the affected taxa were able to detect the presence of L. packardi via kairomones (Lass and Spaak,

2003) and subsequently remain in diapause (Waterkeyn et al., 2012). This is an unlikely explanation for my mesocosm results, however. Kairomones would have been present in all mesocosms, albeit at different concentrations depending on treatment, because tadpole shrimp were never completely absent from the Removal group. And while there is some evidence that the eggs of certain crustacean taxa can detect kairomones and delay their hatching, the evidence so far is limited to vertebrate predators that target only adult crustaceans (Blaustein, 1997; Spencer and Blaustein, 2001; Bozelli et al., 2008).

Waterkeyn et al. (2012) also found no evidence that crustacean eggs can detect notostracan kairomones. They further note that a strategy of delayed hatching would not benefit crustaceans in the presence of notostracan predators, because their eggs would still be vulnerable to predation in the sediment.

34

Assuming that predation was a factor in my results, it could be posited that egg translocation occurred in the mesocosm experiment but was offset by egg consumption, such that the net effect on crustacean abundance was neutral or negative. The merits of this idea depend on L. packardi’s rate of predation on active and resting stages relative to the number of eggs in the mesocosms’ soil. To my knowledge, there are no published studies reporting predation rates for L. packardi in either natural pools or artificial containers. However, studies have quantified predation rates for the notostracans Triops cancriformis (Waterkeyn et al., 2011a; Waterkeyn et al., 2011b) and Lepidurus arcticus

(Christoffersen, 2001). These studies show that notostracans can exert considerable predation pressure on crustacean adults and eggs, at least in small, experimental containers that offer no other food options. To illustrate, Waterkeyn et al. (2011a) extrapolated their laboratory results to estimate that a dense population of Triops sp. (300 individuals/m2) could completely strip an egg bank of resting eggs in 100 days or fewer.

If bioturbation by tadpole shrimp returned eggs to the sediment surface and facilitated their hatching, this could very well be undetectable if more eggs were being eaten than were being translocated. However, these studies likely overestimate predation rates, for two reasons. First, the researchers imposed artificially high encounter rates on predator and prey by using small containers (200 mL – 10 L) from which the prey could not escape. Second, these studies did not include the broad array of food items normally consumed by notostracans in their natural habitats. Rather, the researchers experimentally limited their subjects’ food options to one or a few prey items. Under such controlled conditions, a high rate of predation is not surprising. Notostracans indeed consume

35 animal prey, but under natural conditions they also rely heavily on detritus and plant matter (Boix et al., 2006). If notostracans were as voracious a predator (of either resting or active stages) in natural temporary ponds as these studies suggest, they would quickly become the only crustacean taxon found in these water bodies. Yet, L. packardi co-occurs with as many as 26 other crustacean taxa in California vernal pools (King et al., 1996).

Therefore, the predation pressure exerted by notostracans in habitats containing their natural assortment of food items is arguably much less than in small, laboratory containers with restricted food options. Under realistic field densities, for instance,

Christoffersen (2001) estimated that one adult Lepidurus sp. would consume a maximum of just six cladocerans per 24 h period. In contrast to putatively low predation rates in natural temporary ponds, the sediments in such ponds can contain a huge number of resting eggs, e.g., 105 – 107 eggs/m2, in their uppermost layers (Hairston and Kearns,

2002; Brendonck and De Meester, 2003). The soil I used in the mesocosms was probably no exception. In principle, these eggs represent a massive potential food resource for L. packardi. In practice, however, these eggs were probably just a fraction of L. packardi’s diet in the mesocosms because plants, algae, and detritus were also available (because natural vernal pool soil was used). Therefore, any signal of egg translocation should have been only marginally muted by egg predation in the mesocosm experiment.

The results of the microcosm experiments can also shed light on the potentially masking effect of egg predation. In the 0-cm burial experiment, L. packardi clearly ate eggs (Figure 2A) but not seeds (Figure 2B). In the 0.5-cm burial experiment, freely roaming L. packardi could have translocated eggs but then eaten them (which might

36 explain why I found no significant difference in nauplii abundance between groups;

Figure 2C). However, if freely roaming L. packardi had translocated eggs in this experiment, it almost surely would have also translocated seeds. Since L. packardi was not eating seeds (i.e., Figure 2B), I should have found a significant difference between treatments in seedling abundance in the 0.5-cm burial experiment, if propagule translocation had in fact occurred. There was no such difference, however (Figure 2D, although this difference approached significance). This indicates that seeds had not been translocated. If seeds had not been translocated, it is unlikely that eggs were. From the microcosm experiments alone, therefore, I infer that egg translocation was not masked by consumption because it did not occur in the first place. This argument rests on the assumption that eggs and seeds have the same potential to be translocated. The anostracan eggs used in the microcosm experiments each had an estimated mass of ~4 µg.

This is less than seeds of vernal-pool plants, which can range from ~20 µg (Linhart,

1974) to ~200 µg (Zammit and Zedler, 1990). Even so, both types of propagule are very small relative to adult or juvenile notostracans. It is reasonable to expect that eggs and seeds in vernal pools have a similar potential for translocation. My microcosm experiments therefore provide evidence that propagule translocation simply did not occur. It is possible that the propagules were buried too deep for L. packardi to reach, but my own observations refute this. I observed freely roaming L. packardi digging ~0.5 cm deep in the microcosm containers. Waterkeyn et al. (2011a) additionally showed that the notostracan Triops cancriformis, similar in size to my L. packardi, can dig as deep as 1.5 cm. It is therefore doubtful that the propagules were out of reach.

37

One weakness with the mesocosm experiment was that I was unable to achieve a sizeable difference in tadpole shrimp density between treatments until the end of the experiment. This indicates that tadpole shrimp were hatching continuously from resting eggs in the Removal mesocosms. This was not observed in a prior study using the same mesocosms, soil, and netting procedure (Croel and Kneitel, 2011), and so was unexpected. Although the continuous presence of juveniles throughout a season has been previously reported (Ahl, 1991; Gallagher, 1996; Alexander and Schlising, 1998), it was thought that these juveniles probably hatched from subitaneous eggs (i.e., eggs that hatch in the same season they are produced; Ahl, 1991). Lepidurus packardi needs a minimum of 41 days, and an average of 54 days, to produce eggs of either type (subitaneous or resting; Helm, 1998). Moreover, I did not observe ovisacs on any individuals until the 18

March treatment. These observations mean that all L. packardi captured in the Removal mesocosms prior to and likely including the 10 March treatment could only have come from resting eggs (assuming that some of the nauplii observed on 27 January were L. packardi). The divergence in L. packardi abundance beginning with the 18 March treatment (Figure 1) probably reflects the hatching of the first cohort of subitaneous eggs in the Control mesocosms. Continuous hatching from resting eggs presumably also occurred in the Control mesocosms, but L. packardi density in these mesocosms may have been kept in check by intraspecific competition or cannibalism from larger, older individuals (Golzari et al., 2009; Waterkeyn et al., 2011b). This might explain why the weekly abundances of L. packardi in the two treatment groups were similar, despite the consistent removal of this taxon from one group. While this issue would have lessened

38 any differences between the two groups of mesocosms, it did not necessarily compromise my ability to detect egg translocation. The L. packardi individuals removed from the

Removal mesocosms were typically smaller than those observed in the Control mesocosms (although I did not collect size data) because they were younger, most having hatched only within the previous ~7 days. Assuming that larger L. packardi individuals disrupt sediment more extensively than smaller ones, more eggs would potentially have been translocated in the Control mesocosms than in the Removal mesocosms. The fact that the Control treatment had 37% greater turbidity than the Removal treatment (data not shown) supports the idea that larger individuals in the former group were causing more sediment disruption. Although this quantitative difference in turbidity may have manifested only within the final week or two of the experiment, when subitaneous eggs began to hatch, my qualitative, weekly observations of turbidity indicate that sediment disruption was greater in the Control mesocosms for most of the experiment.

Water temperature appeared to contribute strongly to community dissimilarity between mesocosms, as suggested by the correlation between temperature and Dimension

1 scores of the NMDS plot. However, dissimilarity along Dimension 1 was driven almost entirely by variation in Eucypris sp. abundance. The strong negative association between

Dimension 1 scores and Eucypris sp. abundance shows that these variables are basically one and the same. Therefore, the positive correlation between water temperature and

Dimension 1 can be interpreted as simply reflecting the negative correlation between water temperature and Eucypris sp. abundance. The temperature gradient itself probably originated from a large tree to the west of the study area that partially shaded the

39 mesocosm array in the morning hours for much of the experiment. It is curious that

Eucypris sp. was the only taxon to exhibit a relationship with water temperature, given that this abiotic variable is a reliable indicator of habitat suitability for temporary-pond crustaceans (Brendonck and De Meester, 2003). It may be that there was not enough variation in the abundances of the other taxa to allow their relationships with water temperature to be detectable. The positive relationship between dissolved oxygen and

Dimension 1 scores was probably a by-product of the positive relationship between dissolved oxygen and temperature; warmer mesocosms likely had higher rates of photosynthesis and lower rates of respiration (because Eucypris sp. was less abundant in warmer mesocosms), hence they had more dissolved oxygen. Turbidity was the only abiotic correlate of Dimension 2, and the only taxon related to this dimension was

Daphnia sp. Taken together, these results indicate that the two most abundant taxa in this study, along with temperature, dissolved oxygen, and turbidity, were mainly responsible for the dissimilarity in crustacean communities between mesocosms.

Based on an analysis of data published in King et al. (1996), Croel and Kneitel (2011) reported that vernal pools containing L. packardi had over twice as many crustacean taxa as pools without L. packardi. Croel and Kneitel (2011) suggested that this might be due to bioturbation by this taxon, but the results of the present study cast doubt on this suggestion. Indeed, my results suggest that L. packardi suppresses crustacean taxa via predation rather than facilitates them via bioturbation. The lack of support for my hypothesis was somewhat surprising. Lepidurus packardi clearly digs within and disrupts the sediment as it forages (Croel and Kneitel, 2011; personal observation). It is difficult

40 to imagine that buried eggs (and seeds) are not returned to the sediment surface in L. packardi’s ejecta. It could be that the L. packardi individuals observed in this study were not large enough to dig up buried eggs. Individuals in the Control mesocosms were smaller than normal for their age, reaching maximum carapace lengths of only ~1.2 cm despite being as old as 7 weeks. Adults this age are usually about twice as long as this

(personal observation). The relatively small size of adults, as well as their continuous hatching from resting eggs, were perhaps both tied to the unusually dry early-winter conditions of 2012, when the mesocosm study was initiated. In any event, I observed these ~1.2 cm-long individuals actively digging in the sediment in the microcosm experiments, so size (or lack thereof) probably had little to do with my results. The short timeline of the microcosm experiments, on the other hand, may have been a factor.

Tadpole shrimp were in the microcosms for just 24 hours, and although I observed the freely roaming L. packardi digging ~0.5 cm deep soon after being added, they perhaps were not in the microcosms long enough to measurably translocate eggs. Hatching phenology could be another reason why I found no evidence of egg translocation, at least in the mesocosm experiment. By the time L. packardi individuals in that experiment had grown large enough to extensively bioturbate and dig up eggs, it may be that the window of opportunity for the hatching of other crustacean taxa was closed because the requisite abiotic cues were no longer optimal (Brendonck, 1996; Brendonck and De Meester,

2003). Rather than hatching, therefore, eggs translocated to the sediment surface might stay in diapause and hatch in a future season. Their return to the sediment surface would be undetectable in a single-season experiment unless the eggs were collected and directly

41 counted. This was beyond the scope of this study, but directly counting eggs in the sediment post-bioturbation (e.g., Marcus and Schmidt-Gengenbach, 1986) might be a more effective way of assessing translocation than counting individuals in the active community. These temporal considerations could be addressed in a longer-term experiment that tracked the abundances of crustaceans in the active community, as well as eggs in the sediment, following exposure to L. packardi. Ideally, such an experiment would include treatments where L. packardi individuals of different sizes were added to mesocosms immediately after inundation and left in for varying amounts of time over the course of multiple seasons. It might then be possible to determine if an egg-translocation signal emerges over a longer time scale, and/or if a size threshold for L. packardi needs to first be attained.

Sediment disruption in California vernal pools occurs from other animals besides L. packardi. For example, cattle frequently wade into vernal pools for forage and water

(Brendonck and De Meester, 2003; Marty, 2005; personal observation) and humans wade into them to sample invertebrates for research purposes. Cattle and humans would obviously mix the sediment much more extensively than L. packardi owing to their larger mass. Such mixing would undoubtedly involve bidirectional transport of eggs and seeds; each step down would plunge propagules on the surface deeper into the sediment, while each step up would bring buried propagules to the sediment surface. Not only could bioturbation by large animals affect recruitment within a single pool, it could also affect recruitment in other pools as well. Waterkeyn et al. (2010), for instance, showed that humans may be important, albeit inadvertent, vectors of local dispersal for temporary-

42 pond crustaceans by transporting their eggs in mud stuck to boots and vehicles.

Vanschoenwinkel et al. (2008, 2011) demonstrated the potential for similar egg transport in mud stuck to the fur and skin of large mammals that wallowed in temporary ponds.

Clearly, eggs in the sediment can potentially be translocated over different spatial scales by the activity of various animals.

The results of the present study indicate that L. packardi is not one of these animals, however. I found no evidence that it translocates crustacean eggs and facilitates their hatching, despite this taxon being a strong bioturbator in California vernal pools. How L. packardi can dig so vigorously through the sediment and yet not translocate buried eggs is unclear, but investigating the mechanism behind this presents opportunities for future research. It could be that L. packardi eats eggs as it encounters them while bioturbating, but my results suggest that consumption did not mask translocation in this study.

Although my hypothesis was not supported, my results do shed light on L. packardi’s role as a predator of crustaceans in vernal pools. This study joins the growing body of research showing that notostracans in general have negative effects on other crustacean taxa in temporary ponds, due mainly to their role as predators in these unique bodies of water.

43

LITERATURE CITED

Ahl, J. S. B. 1991. Factors affecting contributions of the tadpole shrimp, Lepidurus

packardi, to its oversummering egg reserves. Hydrobiologia 212:137–143.

Alexander, D. G., and R. A. Schlising. 1998. Patterns in time and space for rare

macroinvertebrates and vascular plants in vernal pool ecosystems at the Vina Plains

Preserve, and implications for pool landscape management. Pages 161–168 in C. W.

Witham, E. T. Bauder, D. Belk, W. R. Ferren Jr., and R. Ornduff, editors. Ecology,

conservation, and management of vernal pool ecosystems – proceedings from a 1996

conference. California Native Plant Society, Sacramento, California, USA. 1998.

Bartel, R. A., N. M. Haddad, and J. P. Wright. 2010. Ecosystem engineers maintain a rare

species of butterfly and increase plant diversity. Oikos 119:883–890.

Belk, D. 1998. Global status and trends in ephemeral pool invertebrate conservation:

implications for Californian fairy shrimp. Pages 147–150 in C. W. Witham, E. T.

Bauder, D. Belk, W. R. Ferren Jr., and R. Ornduff, editors. Ecology, conservation, and

management of vernal pool ecosystems – proceedings from a 1996 conference.

California Native Plant Society, Sacramento, California, USA. 1998.

Bertness, M. D., G. H. Leonard, J. M. Levine, P. R. Schmidt, and A. O. Ingraham. 1999.

Testing the relative contribution of positive and negative interactions in rocky

intertidal communities. Ecology 80:2711–2726.

Blaustein, L. 1997. Non-consumptive effects of larval Salamandra on crustacean prey:

can eggs detect predators? Oecologia 110:212–217.

44

Blaustein, L., and S. S. Schwartz. 2001. Why study ecology in temporary pools? Israel

Journal of Zoology 47:303–312.

Bliss, S. A., and P. H. Zedler. 1998. The germination process in vernal pools: sensitivity

to environmental conditions and effects on community structure. Oecologia 113:67–

73.

Boix, D., J. Sala, S. Gascón, and S. Brucet. 2006. Predation in a temporary pond with

special attention to the trophic role of Triops cancriformis (Crustacea: Branchiopoda:

Notostraca). Hydrobiologia 571:341–353.

Bozelli, R. L., M. Tonsi, F. Sandrini, and M. Manca. 2008. A Big Bang or small bangs?

Effects of biotic environment on hatching. Journal of Limnology 67:100–106.

Brenchley, G. A., 1981. Disturbance and community structure: an experimental study of

bioturbation in marine soft-bottom environments. Journal of Marine Research 39:767–

790.

Brendonck, L. 1996. Diapause, quiescence, hatching requirements: what we can learn

from large freshwater branchiopods (Crustacea: Branchiopoda: Anostraca, Notostraca,

Conchostraca). Hydrobiologia 320:85–97.

Brendonck, L., and L. De Meester. 2003. Egg banks in freshwater zooplankton:

evolutionary and ecological archives in the sediment. Hydrobiologia 491:65–84.

Brinson, M. M., and A. I. Malvárez. 2002. Temperate freshwater wetlands: types, status,

and threats. Environmental Conservation 29:115–133.

45

Byers, J. E., K. Cuddington, C. G. Jones, T. S. Talley, A. Hastings, J. G. Lambrinos, J. A.

Crooks, and W. G. Wilson. 2006. Using ecosystem engineers to restore ecological

systems. Trends in Ecology and Evolution 21:493–500.

Cáceres, C. E. 1998. Interspecific variation in the abundance, production, and emergence

of Daphnia diapausing eggs. Ecology 79:1699–1710.

Christoffersen, K. 2001. Predation on Daphnia pulex by Lepidurus arcticus.

Hydrobiologia 442: 223–229.

Cohen, A. S., R. Bills, C. Z. Cocquyt, and A. G. Caljon. 1993. The impact of sediment

pollution in Lake Tanganyika. Conservation Biology 7:667–677.

Creed, R. P., A. Taylor, and J. R. Pflaum. 2010. Bioturbation by a dominant detritivore in

a headwater stream: litter excavation and effects on community structure. Oikos

119:1870–1876.

Croel, R. C., and J. M. Kneitel. 2011. Ecosystem-level effects of bioturbation by the

tadpole shrimp Lepidurus packardi in temporary pond mesocosms. Hydrobiologia

665:169–181.

De Meester, L., S. Declerck, R. Stoks, G. Louette, F. Van de Meutter, T. De Bie, E.

Michels, and L. Brendonck. 2005. Ponds and pools as model systems in conservation

biology, ecology and evolutionary biology. Aquatic Conservation: Marine and

Freshwater Ecosystems 15:715–725.

De Stasio, B. T., Jr. 1989. The seed bank of a freshwater crustacean: copepodology for

the plant ecologist. Ecology 70:1377–1389.

46

Donohue, I., and J. G. Molinos. 2009. Impacts of increased sediment loads on the

ecology of lakes. Biological Reviews 84:517–531.

Federal Register. 2003. Endangered and threatened wildlife and plants; final designation

of critical habitat for four vernal pool crustaceans and eleven vernal pool plants in

California and Southern Oregon; final rule. Federal Register 68:46684–46762.

Frisch, D., P. K. Morton, P. Roy Chowdhury, B. W. Culver, J. K. Colbourne, L. J.

Weider, and P. D. Jeyasingh. 2014. A millennial-scale chronicle of evolutionary

responses to cultural eutrophication in Daphnia. Ecology Letters 17:360–368.

Gallagher, S. P. 1996. Seasonal occurrence and habitat characteristics of some vernal

pool branchiopoda in northern California, U.S.A. Journal of Crustacean Biology

16:323–329.

Gleason, R. A., N. H. Euliss Jr., D. E. Hubbard, and W. G. Duffy. 2003. Effects of

sediment load on emergence of aquatic invertebrates and plants from wetland soil egg

and seed banks. Wetlands 23:26–34.

Golzari, A., S. Khodabandeh, and J. Seyfabadi. 2009. Some biological characteristics of

tadpole shrimp, Triops cancriformis, from seasonal pools of West Azarbaijan (Iran).

Journal of Agricultural Science and Technology 11:81–90.

Gyllström, M., T. Lakowitz, C. Brönmark, and L.-A. Hansson. 2008. Bioturbation as

driver of zooplankton recruitment, biodiversity and community composition in aquatic

ecosystems. Ecosystems 11:1120–1132.

47

Hairston, N. G., Jr., and C. M. Kearns. 2002. Temporal dispersal: ecological and

evolutionary aspects of zooplankton egg banks and the role of sediment mixing.

Integrative and Comparative Biology 42:481–491.

Hairston, N. G., Jr., R. A. Van Brunt, and C. M. Kearns. 1995. Age and survivorship of

diapausing eggs in a sediment egg bank. Ecology 76:1706–1711.

Hammer, Ø., D. A. T. Harper, and P. D. Ryan. 2001. PAST: Paleontological Statistics

Software Package for Education and Data Analysis. Palaeontologia Electronica 4:1–9.

http://palaeo-electronica.org/2001 1/past/issue1 01.htm.

Helm, B. P. 1998. Biogeography of eight large branchiopods endemic to California.

Pages 124–139 in C. W. Witham, E. T. Bauder, D. Belk, W. R. Ferren Jr., and R.

Ornduff, editors. Ecology, conservation, and management of vernal pool ecosystems –

proceedings from a 1996 conference. California Native Plant Society, Sacramento,

California, USA. 1998.

Hildrew, A. G. 1985. A quantitative study of the life history of a fairy shrimp

(Branchiopoda: Anostraca) in relation to the temporary nature of its habitat, a Kenyan

rainpool. Journal of Animal Ecology 54:99–110.

Holland, R. F. 1978. The geographic and edaphic distribution of vernal pools in the great

Central Valley, California. California Native Plant Society, Special Publication # 4,

Sacramento, CA.

Holland, R. F. 1998. Great Valley vernal pool distribution, photorevised 1996. Pages 71–

75 in C. W. Witham, E. T. Bauder, D. Belk, W. R. Ferren Jr., and R. Ornduff, editors.

Ecology, conservation, and management of vernal pool ecosystems – proceedings

48

from a 1996 conference. California Native Plant Society, Sacramento, California,

USA. 1998.

Holland, R. F., and S. K. Jain. 1981. Insular biogeography of vernal pools in the Central

Valley of California. The American Naturalist 117:24–37.

Jones, C. G., J. H. Lawton, and M. Shachak. 1994. Organisms as ecosystem engineers.

Oikos 69:373–386.

Kearns, C. M., N. G. Hairston Jr., and D. H. Kesler. 1996. Particle transport by benthic

invertebrates: its role in egg bank dynamics. Hydrobiologia 332:63–70.

Keeley, J. E., and P. H. Zedler. 1998. Characterization and global distribution of vernal

pools. Pages 1–14 in C. W. Witham, E. T. Bauder, D. Belk, W. R. Ferren Jr., and R.

Ornduff, editors. Ecology, conservation, and management of vernal pool ecosystems –

proceedings from a 1996 conference. California Native Plant Society, Sacramento,

California, USA. 1998.

King, J. L., M. A. Simovich, and R. C. Brusca. 1996. Species richness, endemism and

ecology of crustacean assemblages in northern California vernal pools. Hydrobiologia

328:85–116.

Kirk, K. L. 1991. Suspended clay reduces Daphnia feeding rate: behavioural

mechanisms. Freshwater Biology 25:357–365.

Kneitel, J. M., and C. L. Lessin. 2010. Ecosystem-phase interactions: aquatic

eutrophication decreases terrestrial plant diversity in the California vernal pool

ecosystem. Oecologia 163:461–469.

49

Lass, S., and P. Spaak. 2003. Chemically induced anti-predator defences in plankton: a

review. Hydrobiologia 491:221–239.

Linhart, Y. B. 1974. Intra-population differentiation in annual plants I. Veronica

peregrina L. raised under non-competitive conditions. Evolution 28:232–243.

Lohrer, A. M., S. F. Thrush, and M. M. Gibbs. 2004. Bioturbators enhance ecosystem

function through complex biogeochemical interactions. Nature 431:1092–1095.

Marcus, N. H., and J. Schmidt-Gengenbach. 1986. Recruitment of individuals into the

plankton: the importance of bioturbation. Limnology and Oceanography 31:206–210.

Marty, J. T. 2005. Effects of cattle grazing on diversity in ephemeral wetlands.

Conservation Biology 19:1626–1632.

Mermillod-Blondin, F., and D. G. Lemoine. 2010. Ecosystem engineering by tubificid

worms stimulates macrophyte growth in poorly oxygenated wetland sediments.

Functional Ecology 24:444–453.

Mermillod-Blondin, F., and R. Rosenberg. 2006. Ecosystem engineering: the impact of

bioturbation on biogeochemical processes in marine and freshwater benthic habitats.

Aquatic Sciences 68:434–442.

Meysman, F. J. R., J. J. Middelburg, and C. H. R. Heip. 2006. Bioturbation: a fresh look

at Darwin’s last idea. Trends in Ecology and Evolution 21:688–695.

Newcombe, C. P., and D. D. MacDonald. 1991. Effects of suspended sediments on

aquatic ecosystems. North American Journal of Fisheries Management 11:72–82.

Palmer, M. A., R. F. Ambrose, and N. L. Poff. 1997. Ecological theory and community

restoration ecology. Restoration Ecology 5:291–300.

50

Pennak, R. W. 1989. Fresh-water invertebrates of the United States: Protozoa to

Mollusca, 3rd ed. John Wiley & Sons, Inc. New York, New York, USA.

Pielou, E. C. 1966. The measurement of diversity of different types of biological

collections. Journal of Theoretical Biology 13:131–144.

Pillay, D., G. M. Branch, and A. T. Forbes. 2007. The influence of bioturbation by the

sandprawn Callianassa kraussi on feeding and survival of the bivalve Eumarcia

paupercula and the gastropod Nassarius kraussianus. Journal of Experimental Marine

Biology and Ecology 344:1–9.

Roast, S. D., J. Widdows, N. Pope, and M. B. Jones. 2004. Sediment–biota interactions:

mysid feeding activity enhances water turbidity and sediment erodability. Marine

Ecology Progress Series 281:145–154.

Ruiz, F., M. Abad, A. M. Bodergat, P. Carbonel, J. Rodríguez-Lázaro, M. L. González-

Regalado, A. Toscano, E. X. García, and J. Prenda. 2013. Freshwater ostracods as

environmental tracers. International Journal of Environmental Science and Technology

10:1115–1128.

Semlitsch, R. D., and J. R. Bodie. 1998. Are small, isolated wetlands expendable?

Conservation Biology 12:1129–1133.

Simovich, M. A. 1998. Crustacean biodiversity and endemism in California’s ephemeral

wetlands. Pages 107–118 in C. W. Witham, E. T. Bauder, D. Belk, W. R. Ferren Jr.,

and R. Ornduff, editors. Ecology, conservation, and management of vernal pool

ecosystems – proceedings from a 1996 conference. California Native Plant Society,

Sacramento, California, USA. 1998.

51

Simovich, M. A., and S. A. Hathaway. 1997. Diversified bet-hedging as a reproductive

strategy of some ephemeral pool anostracans (Branchiopoda). Journal of Crustacean

Biology 17:38–44.

Spencer, M., and L. Blaustein. 2001. Hatching responses of temporary pool invertebrates

to signals of environmental quality. Israel Journal of Zoology 47:397–417.

Ståhl-Delbanco, A., and L.-A. Hansson. 2003. Effects of bioturbation on recruitment of

algal cells from the “seed bank” of lake sediments. Limnology and Oceanography

47:1836–1843.

Sun, B., and J. W. Fleeger. 1994. Field experiments on the colonization of meiofauna into

sediment depressions. Marine Ecology Progress Series 110:167–175.

Thorp, J. H., and A. P. Covich. 2010. Ecology and classification of North American

freshwater invertebrates, 3rd ed. Academic Press, San Diego, California, USA.

U.S. Fish and Wildlife Service. 2007. Vernal pool tadpole shrimp (Lepidurus packardi)

5-year review: summary and evaluation. Sacramento Fish and Wildlife Office,

Sacramento.

Vandekerkhove, J., S. Declerck, L. Brendonck, J. M. Conde-Porcuna, E. Jeppesen, L. S.

Johansson, and L. De Meester. 2005. Uncovering hidden species: hatching diapausing

eggs for the analysis of cladoceran species richness. Limnology and Oceanography:

Methods 3:399–407.

Vanschoenwinkel, B., A. Waterkeyn, T. Nhiwatiwa, T. Pinceel, E. Spooren, A. Geerts, B.

Clegg, and L. Brendonck. 2011. Passive external transport of freshwater invertebrates

52

by elephant and other mud-wallowing mammals in an African savannah habitat.

Freshwater Biology 56:1606–1619.

Vanschoenwinkel, B., A. Waterkeyn, T. Vandecaetsbeek, O. Pineau, P. Grillas, and L.

Brendonck. 2008. Dispersal of freshwater invertebrates by large terrestrial mammals: a

case study with wild boar (Sus scrofa) in Mediterranean wetlands. Freshwater Biology

53:2264–2273.

Vanschoenwinkel, B., M. Seaman, and L. Brendonck. 2010. Hatching phenology, life

history and egg bank size of fairy shrimp Branchipodopsis spp. (Branchiopoda,

Crustacea) in relation to the ephemerality of their rock pool habitat. Aquatic Ecology

44:771–780.

Waterkeyn, A., B. Vanschoenwinkel, S. Elsen, M. Anton-Pardo, P. Grillas, and L.

Brendonck. 2010. Unintentional dispersal of aquatic invertebrates via footwear and

motor vehicles in a Mediterranean wetland area. Aquatic Conservation: Marine and

Freshwater Ecosystems 20:580–587.

Waterkeyn, A., J. Vanoverbeke, N. Van Pottelbergh, and L. Brendonck. 2011a. While

they were sleeping: dormant egg predation by Triops. Journal of Plankton Research

33:1617–1621.

Waterkeyn, A., N. Van Pottelbergh, J. Vanoverbeke, B. Vanschoenwinkel, L. De

Meester, and L. Brendonck. 2012. Constitutive but no Triops-induced differences in

bet-hedging strategies for hatching in Daphnia. Hydrobiologia 715:29–35.

53

Waterkeyn, A., P. Grillas, M. Anton-Pardo, B. Vanschoenwinkel, and L. Brendonck.

2011b. Can large branchiopods shape microcrustacean communities in Mediterranean

temporary wetlands? Marine and Freshwater Research 62:46–53.

Yee, S. H., M. R. Willig, and D. L. Moorhead. 2005. Tadpole shrimp structure

macroinvertebrate communities in playa lake microcosms. Hydrobiologia 541:139–

148.

Zammit, C., and P. H. Zedler. 1990. Seed yield, seed size and germination behaviour in

the annual Pogogyne abramsii. Oecologia 84:24–28.