Vertical Distribution Patterns of Crustaceous Zooplankton Associated with Invasive Bythotrephes longimanus and Cercopagis pengoi in Lake Champlain (U.S.A.)

Abstract

Bythotrephes longimanus (spiny waterflea) and Cercopagis pengoi (fishhook waterflea) are large invasive predatory cladocerans that alter the composition, density, and behavior of native zooplankton communities. Lake Champlain was invaded by Bythotrephes and Cercopagis in 2014 and 2018, respectively. This study was conducted to determine the changes in crustaceous zooplankton diel vertical migration (DVM) associated with the presence of these two invasive species. Daytime and nighttime zooplankton samples were collected from vertical net tows at 5 m intervals using 153 µm and 250 µm closing plankton nets at a 50 m deep site in Lake Champlain during the month of August (2013–2016, 2019, 2023, and 2024). Sampling dates encompassed years before and after each invader entered the lake. The results show increased DVM activity in several native zooplankton taxa associated with invasion years, including Daphnia retrocurva, Bosmina longirostris, and Diacyclops thomasi. Zooplankton in Lake Champlain appear to occupy deeper depths during the daytime after Bythotrephes and Cercopagis invaded than in previous years. Alterations associated with Bythotrephes were more extreme, whereas changes associated with Cercopagis were longer lasting. These shifts in DVM behavior have potential implications for trophic dynamics in Lake Champlain by altering competitive interactions and foraging behavior of zooplankton and their predators.

1. Introduction

Aquatic invasive species introductions have been attributed to a variety of ecological alterations in various freshwater systems. Beyond negative impacts on recreational activities and economic losses, aquatic invasive species can modify lake food webs and decrease biodiversity [1,2,3]. Anthropogenic activity has been the primary cause of invasive species introductions for centuries. The Great Lakes have been the primary recipient of aquatic invasive species brought to North America via ballast water transfer from intercontinental vessels [4]. Once established, waterway interconnectivity and overland transport by recreational boaters can facilitate further spread of aquatic invasive species [5,6].

Lake Champlain is the largest lake in the northeastern United States, excluding the Great Lakes. The historical features and recreational opportunities provided by the lake support a productive tourism economy [7]. Lake Champlain has received comparatively fewer aquatic invasive species than the Great Lakes. Because of its relative inaccessibility for large intercontinental vessels, Lake Champlain likely has not received invasive species through ballast water transfers [6]. However, due to manmade canals, Lake Champlain is directly connected to the Great Lakes and other systems [6]. The entrance and establishment of invasive species in Lake Champlain have altered its lower food web through both temporary and lasting declines in native zooplankton taxa [8,9]. The invasion of Dreissena polymorpha in the mid-1990s resulted in significant declines in rotifer abundance and species richness [1,8].

Diel vertical migration (DVM) is a widespread zooplankton behavior that involves migration deeper in the water column during the day and returning to the surface at night. Although this behavior is influenced by a variety of factors, such as food availability and water transparency, the ultimate driving factor is widely considered to be the avoidance of visual predators [10,11,12]. DVM behavior varies across taxa and waterbodies and is generally associated with metabolic costs [13,14,15]. Migration often extends to depths beyond the thermocline, where productivity and temperatures are lower, inhibiting growth and reproduction [14,16].

Lake Champlain was invaded by two morphologically similar large planktivorous cladocerans from the family Cercopagidae, Bythotrephes longimanus (spiny waterflea) and Cercopagis pengoi (fishhook waterflea), in 2014 and 2018, respectively [9,17]. Both species are predators capable of extensive feeding on smaller zooplankton taxa [9,18,19] and have been linked to declines in several native taxa in the Great Lakes and Lake Champlain [9,18,20,21]. Based on studies conducted on Bythotrephes and morphological similarities between the two invaders, both ceropagids likely detect prey primarily through visual cues but may rely on mechanoreception when light levels are reduced [22,23]. Despite their larger body sizes making them more visible for planktivorous fish, Bythotrephes and Cercopagis are not believed to undergo DVM in invaded waterbodies, remaining in the epilimnion or metalimnion at all times [18,20,21]. Both species possess an elongated tail-spine that may interfere with predation by small fish, reducing risks associated with occupying illuminated surface waters [24,25]. Bythotrephes exhibit higher predation rates in illuminated waters [22] and may not be able to feed efficiently at deeper depths where light penetration is limited.

Lake Champlain zooplankton community composition in the Main Lake historically has been representative of deep oligotrophic lakes in temperate regions [8,26,27]. The community currently consists of 30 crustaceous zooplankton taxa. The most abundant cladocerans include Bosmina longirostris and Daphnia retrocurva, and the most common copepods are Diacyclops thomasi and Leptodiaptomus [8]. An early study on Lake Champlain zooplankton DVM suggests that the most common Cladocera in the lake preferentially inhabit the epilimnion both day and night and do not undergo DVM. However, there was limited DVM in D. thomasi and strong DVM, mostly restricted to the hypolimnion, observed in Leptodiaptomus sicilis [28]. Daphnia retrocurva, B. longirostris, and D. thomasi have experienced density declines following invasion by cercopagids in Lake Champlain and other large freshwater lakes, likely as a result of direct predation [9,18,29,30]. Daphnia retrocurva and D. thomasi populations in the Great Lakes appear to alter their vertical structure to avoid Bythotrephes and Cercopagis [18,30] and likely exhibit similar behavior in Lake Champlain.

The goal of this study was to determine the impacts of Bythotrephes and Cercopagis on the diel vertical distribution of crustaceous zooplankton in Lake Champlain. Bythotrephes only occurred in high densities in 2014 and 2015, dropping to such low densities in 2016 that they were not detected in routine Lake Champlain sampling for two years (2017 and 2018). Although Bythrotrephes was detected again in 2019, they were present in very low densities (

Table 1. Invasion timing for Bythotrephes (SWF) and Cercopagis (FHWF) in Lake Champlain based on Cutter et al. [9]. N/D = not detected during the month of August or September. Low = low-density populations detected. High = high-density populations detected. Incr = low population density in August, increasing to high population density in September. Bolded columns represent years in which zooplankton diel vertical structure samples were taken. Peak late-season density (#/m³) is shown for each taxon.

Invader	2013	2014	2015	2016	2017	2018	2019	2020	2021	2022	2023	2024
SWF	N/D	Incr	High	Low	N/D	N/D	Low	Low	Low	Low	Low	Low
Peak density	0	15	14.1	0.10	0	0	0.05	0.20	0.10	0.52	0.20	0.62
FHWF	N/D	N/D	N/D	N/D	N/D	Incr	High	High	High	High	High	High
Peak density	0	0	0	0	0	36.4	49.8	22.8	26.5	38.0	40.0	53.1

, [9]). Cercopagis, however, have remained in high densities since their 2018 invasion through 2023. We collected zooplankton vertical structure data in a 50 m deep Main Lake study site before and after both the Bythotrephes and Cercopagis invasions in Lake Champlain, allowing the potential impacts of both invasive predators to be assessed in this study.

2. Materials and Methods

2.1. Field Data Collections

Samples were collected at a 50 m depth site at the SUNY Plattsburgh Lake Champlain Data Buoy just south of Valcour Island in the Main Lake segment of Lake Champlain (Figure 1). This site is used for intensive monitoring of lake thermal structure, with water column temperature profiles collected by the data buoy at 1–2 m depth intervals every 15 min.

Zooplankton and phytoplankton net tow samples were taken at 5 m depth intervals 12 h apart within the same diel periods using closing plankton nets retrieved at 1 m/s⁻¹. Sampling was conducted in August during low moon phases in 2013–2016, 2019, 2023, and 2024 between the hours of 1100–1300 (daytime) and 2300–0100 (nighttime). Standard zooplankton samples were collected with a 153 μm, 30 cm diameter closing net. Larger zooplankton, including invasive cercopagids, were taken using a 250 μm, 50 cm diameter closing net. Zooplankton samples were condensed into bottles and preserved with a 5% Rose Bengal formalin solution.

2.2. Laboratory Processing

Zooplankton samples collected in the standard (153 μm) zooplankton nets preserved in 125 mL bottles were processed in 1 mL subsamples until a minimum of 100 organisms were counted or five subsamples were processed. Subsamples were taken with a Hensen–Stempel pipette and processed in a Sedgewick Rafter Counting Cell using a Leica inverted microscope. Zooplankton species were identified using A revised key to zooplankton of Lake Champlain [31] and Zooplankton of the Great Lakes: A guide to the identification and ecology of common crustacean species [32]. All Bythotrephes, Cercopagis, and Leptodora were counted in the 250 μm net samples under a dissecting scope.

2.3. Data Analysis

The weighted mean depth (WMD) of species and species groups was calculated for each sampling date and diel period based on calculated densities from samples at each depth interval using the following formula, where n equals abundance and d represents depth [33]: WMD = ${\displaystyle \frac{\sum n_i{d}_i}{\sum n_i}}$.

Proportions of taxa in each water layer were determined using calculated zooplankton densities in each interval (individuals/m³). The metalimnion was calculated by determining the steepest thermal gradient in the water column within 24 h of each sampling event, where water above and below the ends of the gradient was considered the epilimnion and hypolimnion, respectively. The proportions of taxa in the groups in the 3 water layers for each sampling event were determined by dividing the density of taxa within each respective water layer by the total density of taxa within the entire water column.

A redundancy analysis (RDA) was conducted using the R package ‘vegan’ [34,35] to explain differences between the day and night depth distribution of the most common Cladocera (Daphnia retrocurva, Bosmina longirostris, B. coregoni, Diaphanosoma birgei, and Ceriodaphnia) and Copepoda (Diacyclops thomasi, Mesocyclops edax, Tropocyclops prasinus mexicanus, Leptodiaptomus minutus, and L. sicilis) collected in this study. Although relatively common, immature copepod stages were not included in the RDA because they were not identified to species. The RDA was performed following the methods used by Daase et al. [36], with species and time of day acting as categorical constraints. Year was added as an additional categorical constraint, and Hellinger-transformed abundances were used for each depth to reflect the relative abundance of individuals of each species at depth and account for zeros in the data. Due to extremely low densities among all species, 2015 was not included in the RDA. A permutation test (999 permutations) was used to assess the significance of explanatory variables.

3. Results

A summary of invasion dynamics for Bythotrephes and Cercopagis in Lake Champlain can be found in Table 1. Each taxon was first detected in 2014 (Bythotrephes) and 2018 (Cercopagis) in lake monitoring samples, followed by increased density in subsequent years. Both species were detected in low densities in August, quickly reaching high densities by September during their respective invasion years. Despite consistent sampling efforts, Bythotrephes were not detected in August of 2016 and were in very low abundance in September of 2016 in the Main Lake [9]. After not being detected in 2017 or 2018, Bythotrephes returned to low density levels in 2019–2024. Cercopagis has remained at high densities since initial detection and invasion in 2018 through 2024 (Table 1). In this study, 2013 is considered a pre-invasion reference year for Bythotrephes, and 2016 is considered a reference year for the Cercopagis invasion.

3.1. Vertical Zooplankton Structure: Pre- and Post-Invasion

Diel vertical migration (DVM) responses of zooplankton taxa associated with the invasion years of both Bythotrephes and Cercopagis varied among taxonomic groups and species. Most crustaceous zooplankton species inhabited the epilimnion during both daytime and nighttime before the 2014 invasion of Bythotrephes. In general, reference years (2013, 2016) showed less overall diel differences in relative community abundance in Lake Champlain than post-invasion years (2014, 2015, 2019, 2023, 2024). There were strong migration patterns in most of the common taxa associated with the Bythotrephes invasion in 2014. When Bythotrephes reached their peak summer density in the Main Lake in 2015 (>15/m³ in July and >5/m³ August, [9]), most taxa were reduced to low densities, making vertical migration trends less apparent (see

Table 2. Average water column densities (individuals/m³) of native zooplankton taxa during the month of August from each sampling event.

Taxon	2013	2014	2015	2016	2019	2023	2024
Bosmina coregoni	5198.6	1808.5	0.0	183.9	2242.9	11,511.7	1458.9
Bosmina longirostris	14,240.9	3557.7	61.3	20,107.8	11,483.7	27,796.1	14,403.5
Ceriodaphnia	3819.2	76.6	0.0	490.4	182.7	1379.3	3519.1
Daphnia retrocurva	10,559.6	4259.6	0.0	8705.2	26,826.4	14,402.5	17,969.8
Diacyclops thomasi	29,218.6	11,162.5	490.4	31,878.1	2651.6	17,765.4	6531.2
Diaphanosoma birgei	3457.6	0.0	0.0	2268.3	1237.9	1179.4	2690.9
Leptodiaptomus minutus	8893.2	7549.6	147.1	6191.7	2428.6	2620.2	548.2
Leptodiaptomus sicilis	1597.0	2344.9	2611.6	4046.1	4982.1	2900.2	1494.3
Mesocyclops edax	1414.1	9662.6	306.5	613.0	2055.3	3569.7	1341.0
Tropocyclops prasinus mexicanus	5574.6	1996.5	527.2	5701.3	143.4	695.6	512.8

). DVM behavior increased after the Cercopagis invasion in 2019 and was still evident in 2023 and 2024. During years when taxa were collected in low densities, their DVM behavior should be interpreted with caution. Average water column densities of native zooplankton taxa from each sampling event can be found in Table 2. Note the low densities of several zooplankton taxa in 2015.

3.1.1. Cladocera

No more than half of the Bosmina longirostris population migrated during any of the sample dates. Bosmina longirostris DVM did not occur in the 2013 or 2016 reference years, where the entire population remained in the epilimnion. Approximately half of the population migrated below the epilimnion during the daytime in 2014 after the Bythotrephes invasion. There was no detectable population of B. longirostris at our sampling location in 2015 sampling during very high lakewide Bythotrephes abundance [9]. DVM was observed in 2019 when Cercopagis were present at high densities, with about 20% of the population in the epilimnion during the day versus 80% at night. In 2023, the population continued to exhibit DVM; individuals that remained in the hypolimnion at night occurred higher in the water column than during the day (Figure 2).

Most Bosmina coregoni remained primarily in the epilimnion during reference years (2013, 2016), exhibiting little to no DVM behavior. No individuals were collected during Bythotrephes peak abundance in 2015, and very few were collected in the 2016 reference year. When Cercopagis was established in 2019, the B. coregoni population began to exhibit DVM, with most of the daytime population collected in the metalimnion. DVM was less pronounced in 2023 and 2024, with a proportion of the population occupying the hypolimnion at night (Figure 3).

The vertical distribution of Daphnia retrocurva, the most common large cladoceran in Lake Champlain, experienced several shifts during the sample years. Daphnia retrocurva exhibited little to no DVM in the 2013 reference year, remaining in the epilimnion both day and night. In the 2014 Bythotrephes invasion year, D. retrocurva exhibited a strong DVM pattern, with the entire population (99%) below the epilimnion during the day and within the epilimnion at night. There was no detectable D. retrocurva population in 2015, when Bythotrephes reached their peak abundance. The population returned to the pre-invasion strategy in 2016 when Bythotrephes abundance was very low; D. retrocurva remained in the epilimnion both day and night. Two-thirds of the population exhibited DVM, with daytime migration below the thermocline, in 2019 following the Cercopagis invasion. DVM patterns in 2023 saw 75% of the population in the hypolimnion during the day and 73% migrating to the epilimnion at night. Most of the population that migrated to the epilimnion in 2023 inhabited the 5–10 m strata, avoiding the surface, unlike previous years. About a third of the population exhibited DVM into the hypolimnion in 2024 (Figure 4). Both invasions resulted in altered DVM behavior for D. retrocurva, with daytime populations moving deeper in the water column in response to the invasion.

Ceriodaphnia populations mostly remained in the epilimnion during the sampling dates (Figure 5). They were collected in very low densities in 2014 and 2019 and were not collected at all in 2015. In 2013 and 2024, the vertical distribution did not change between daytime and nighttime. In 2023, Ceriodaphnia exhibited a weak reverse migration pattern, with the nighttime hypolimnion proportion 20% higher than that of the day. Populations exhibited a weak normal migration pattern in 2024 (Figure 5).

Diaphanosoma displayed very short migration amplitudes throughout this study, largely remaining in the epilimnion. Limited DVM activity in the 2013 reference year had a quarter of the population moving from the metalimnion to the epilimnion at night. The population was concentrated in the epilimnion in the 2016 reference year during both diel periods. Individuals were not collected during the Bythotrephes invasion year or peak abundance year (2014 and 2015). Following the Cercopagis invasion, approximately half of the population migrated from the metalimnion to the epilimnion at night in 2019. Although the proportion of individuals in the epilimnion in 2023 slightly increased at night, the proportion in the surface (0–5 m) strata decreased. Additionally, in 2023, 22% of the Diaphanosoma population remained in the hypolimnion at night. A similar trend was observed in 2024, with about 30% of Diaphanosoma occupying the hypolimnion at night (Figure 6).

Leptodora are the only native predatory cladocerans in Lake Champlain. Most individuals occupied the top 20 m of the water column throughout this study. About a third of the Leptodora population exhibited DVM in 2013, spreading throughout the hypolimnion during the day and migrating to the epilimnion at night; around two-thirds of Leptodora did not exhibit DVM in 2013 (pre-invasion), remaining at the surface of the epilimnion (0–5 m deep) both day and night. Leptodora were not collected during the 2014 Bythotrephes invasion year. Only a small portion of the population exhibited DVM in 2016 prior to the Cercopagis invasion, with most remaining near the surface both day and night. Few individuals were collected during the day in 2019, but those collected were in the epilimnion. DVM was observed in a larger proportion of the population in 2023 than in previous sample years, with almost half of the Leptodora collected during the day coming from below the thermocline, migrating to the epilimnion at night. Partial migration was still evident in 2024, but individuals did not migrate into the hypolimnion (Figure 7).

3.1.2. Copepods

Diacyclops thomasi DVM behavior was impacted by both predatory invasions in Lake Champlain. Individuals exhibited a weak DVM in August of the 2013 reference year, when their daytime depth distribution was widely distributed within the water column and the majority of the population moved into the epilimnion at night. Following the Bythotrephes invasion (2014), DVM activity increased as most of the population migrated below the epilimnion in the daytime, returning to the epilimnion at night. Densities were greatly reduced in the 2015 Bythotrephes peak abundance year; around 60% of the remaining population stayed in the epilimnion at all times, with about 30% of the population migrating from the hypolimnion to the metalimnion at night. DVM activity of D. thomasi was reduced in 2016 after Bythotrephes abundance declined, with populations more evenly distributed in the water column during the day and a large majority (85%) migrating to the epilimnion at night. In 2019, following the Cercopagis invasion, almost no individuals occupied the epilimnion during the daytime; the bulk of the DVM involved migration from the hypolimnion to the metalimnion at night, with less than a quarter of the population reaching the epilimnion. Daytime populations were almost exclusively in the hypolimnion in 2023; although most migrated up to the epilimnion at night, about a quarter of the population remained in the hypolimnion. Strong migration behavior continued in 2024, with fewer individuals reaching the surface strata at night (Figure 8).

Another taxon with strong DVM-related impacts from the invasive predators was Mesocyclops edax. Individuals exhibited limited DVM in 2013 (pre-invasion), with the majority of the population occupying the epilimnion during both daytime and nighttime. There was a major increase in DVM during the 2014 Bythotrephes invasion year, when only 10% of the population inhabited the epilimnion during the day compared to 95% at night. Populations were greatly reduced in 2015 when Bythotrephes reached their peak abundance. Mesocyclops edax populations in 2016 migrated from the upper portion of the deep metalimnion (starting at 22 m) during the day to the top of the epilimnion at night. DVM occurred in 2019, following the Cercopagis invasion, with populations distributed in the 5–25 m depth strata of the water column during the day and 90% moving to the surface at night. Daytime DVM into the deep strata continued with high Cercopagis abundance in 2023 and 2024 (Figure 9).

Tropocyclops prasinus mexicanus remained near the surface (epilimnion or metalimnion) during both reference years (2013, 2016) and exhibited a similar strategy during post-invasion years. As for many other taxa, low densities occurred in the 2015 and 2019 post-invasion years. Populations showed a partial reverse DVM pattern in 2023 and 2024, where 30% migrated from the epilimnion to the hypolimnion at night (Figure 10).

The most common calanoid copepods in Lake Champlain from the family Diaptomidae (Leptodiaptomus spp.) showed variable shifts in their vertical migration patterns across the sample year. Leptodiaptomus minutus exhibited several shifts in their DVM patterns in response to cercopagid presence. The population exhibited a weak vertical migration during the 2013 reference year, with the proportion of individuals occupying the epilimnion at night 20% greater than during the day. Nearly the entire population avoided the epilimnion during the day in the 2014 Bythotrephes invasion year, with three-quarters of the population migrating up to the epilimnion at night. The species was reduced to extremely low densities in 2015. About half of the L. minutus population migrated from below to above the epilimnion at night in the 2016 reference year. Leptodiaptomus minutus shifted their DVM pattern in 2019, following the Cercopagis invasion, with most of the population in the metalimnion during the day, moving to the epilimnion at night. Vertical migration patterns shifted again in 2023, with continued high Cercopagis density; while 40% of the population migrated to the epilimnion at night, over a quarter of the population remained in the hypolimnion. Leptodiaptomus minutus occurred in low densities in 2024 (Figure 11).

Nearly the entire population of Leptodiaptomus sicilis was in the hypolimnion during all sampling dates. A relatively small proportion of the population migrated into the metalimnion at night before and during the first year of the Bythotrephes invasion (2013 and 2014, respectively). Despite typical daytime densities, few L. sicilis individuals were collected at night in 2015, leading to a patchy vertical distribution. The patchy nighttime vertical distribution following the Cercopagis invasion in 2019 was not a result of a reduced population density. In 2023 and 2024, the population did not migrate above the hypolimnion at night (Figure 12).

Bythotrephes and Cercopagis exhibited opposite DVM trends in years when they were collected in high densities. Bythotrephes were only collected in the daytime in 2014, when the entire population was at the surface (0–5 m), exhibiting low densities (around 1/m³). During their peak density year (2015), the majority of the Bythotrephes population was in the epilimnion during the day, exhibiting a reverse DVM pattern, moving to deeper strata at night. Although abundant in Lake Champlain at the time [9], Cercopagis were collected in low densities both day and night in 2019 and exhibited a low amplitude reverse DVM strategy. In 2023, Cercopagis clearly exhibited a normal DVM strategy, with most of the population below the epilimnion during the day, returning closer to the surface at night. DVM continued in 2024, but daytime individuals did not venture beyond the bottom of the metalimnion (Figure 13).

3.2. Community Structure

The ordination analysis (RDA) revealed differences in migration patterns across sample years for the ten most common species. The horizontal plot axis explained about 33% of the constrained variance, whereas the vertical axis explained about 8% (Figure 14). Depth followed a gradient on the horizontal axis of the plot, with samples higher in the water column appearing further to the left and deeper samples appearing to the right side of the plot. Deeper samples (>30 m, L. sicilis only) on the right side of the horizontal axis followed a vertical gradient, with points representing shallower depths higher up on the vertical axis. For taxa that exhibited strong migration responses to cercopagid invasions, such as D. thomasi and D. retrocurva, there are increased distances between day and night samples relative to distance during years when cercopagids were absent. The 2014 Bythotrephes invasion year showed the greatest distance between daytime and nighttime points for most species. Samples from after the 2018 Cercopagis invasion generally showed greater distances between day and night samples than the reference years (2013 and 2016). Ceriodaphnia and T. prasinus mexicanus did not exhibit noticeable migration patterns, remaining at the surface day and night for all sample years. Beyond migration distances, other changes in vertical distribution are also revealed by the ordination. When looking at the 2023 sample year, the day/night points are closer on the ordination than those of the 2016 reference year because the populations did not return to the surface at night, as seen by the nighttime sample appearing further right on the horizontal axis than the daytime samples of the 2013 and 2016 reference years for several species. Daytime depths for migrating taxa in 2023 are similar to those of other years where DVM was apparent (i.e., 2014 and 2019, Figure 14). Information on species richness and the relative abundance of taxa at each depth interval during the sampling events can be found in Appendix A.

4. Discussion

Our results suggest that both Bythotrephes and Cercopagis have influenced the zooplankton vertical community structure in Lake Champlain. Although interannual and seasonal variation in zooplankton DVM is apparent [37,38], Cercopagidae invasion years mark clear shifts in the zooplankton vertical structure in Lake Champlain during late summer thermal stratification (August). Responses for most taxa related to Bythotrephes were immediate but only lasted for the two years in which Bythotrephes were abundant (2014, 2015). Responses related to the 2018 Cercopagis invasion have continued through 2024, slightly increasing in their intensity over time for cyclopoid copepods. Changes in DVM magnitude of the most affected species (B. coregoni, B. longirostris, D. retrocurva, D. thomasi, M. edax) are summarized in Figure 15. Although DVM alterations related to Cercopagis have continued since their initial invasion, alterations related to Bythotrephes were more intense, especially for D. retrocurva (Figure 14 and Figure 15).

4.1. Vertical Migration Patterns During the Reference Years

During the two years when Bythotrephes and Cercopagis were absent, 2013 and 2016, Cladocera remained at the surface both day and night. Surface waters generally have higher food abundance and warmer temperatures that are essential for population fitness [39]. Even in cases where peak phytoplankton densities occur deep, cladocerans need to migrate to warmer surface waters for optimized growth and development [40]. The negative effects of remaining in lower temperatures are apparently more detrimental than low food abundance [14]. Because of their small body sizes, the most common Cladocera in Lake Champlain, Bosmina and D. retrocurva, likely did not undergo DVM to avoid fish predation prior to the cercopagid invasions. Both taxa can respond to planktivorous fish by reducing their body size at maturity to reduce predator detection and may not have been susceptible to intense enough fish predation in Lake Champlain to warrant migration [41].

Copepods exhibited a very weak DVM pattern during reference years, 2013 and 2016, with notable differences in their overall distributions. The two taxa most influential in determining distributions were Diacyclops thomasi and Diaptomidae, the most common cyclopoid and calanoid, respectively. Diacyclops thomasi spread throughout the water column during the day in 2013 and 2016, migrating to the surface at night both years. The same trend was observed for D. thomasi during the 1976 sampling events [28]. During 2016, Diaptomidae occurred more frequently at deeper depths than in previous years. This trend is partially related to species composition. The two Leptodiaptomus species found in Lake Champlain have different depth preferences. Leptodiaptomus minutus prefer to inhabit the upper strata of the water column, whereas L. sicilis prefer to stay deep [32]. Although L. sicilis was found to migrate into the epilimnion at night in Lake Champlain in 1976 [28], the species did not exhibit that tendency in 2013 or 2016. The metalimnion was both deeper and thicker on the 2016 sampling date compared to other years, potentially influencing observed migration patterns.

4.2. Vertical Distribution Shifts Associated with the Bythotrephes Invasion

There was a marked shift in the daytime distribution of Cladocera during the 2014 and 2015 Bythotrephes invasion years, despite the invader being first detected in low densities during the month of August in 2014. Nearly all Cladocera were collected at the surface at night across all sample years that they were present, regardless of their daytime distribution. Less common herbivorous cladocerans, Diaphanosoma and Ceriodaphnia, did not appear to shift their daytime vertical distribution in response to Bythotrephes, remaining at the surface. Bosmina species, however, reacted post-invasion by distributing more evenly in the water column during the day. The most prominent migration response was that of Daphnia retrocurva, which completely avoided the epilimnion during daylight in 2014 in response to the Bythotrephes invasion. Almost no herbivorous cladocerans of any species were collected in this study during August 2015, when Bythotrephes reached their peak density and were highly concentrated in the epilimnion, suggesting that different vertical migration strategies were ineffective at avoiding Bythotrephes predation at these extremely high densities. The 2015 Bythotrephes densities in Lake Champlain are among the highest recorded abundance levels in the scientific literature [9]. In other parts of the Main Lake in 2015, modest declines occurred for B. longirostris and other small cladocerans, while D. retrocurva were virtually eliminated [9]. Bythotrephes prefer Cladocera prey, such as Daphnids [42], and D. retrocurva lack the swimming ability to escape predation [43]. In summary, Bythotrephes appear to have reduced or nearly eliminated many cladoceran taxa from the pelagic waters of Lake Champlain in the first year following their invasion, resulting in altered migration behavior, with some taxa moving to deep waters during the day. The Bythotrephes invasion in Lake Champlain was short lived, with high population densities observed only in late 2014 and 2015. By 2016, cladoceran DVM behavior had largely returned to pre-invasion strategies, with most taxa in the epilimnion day and night.

Although copepods exhibited some form of DVM pre- and post-invasions, they exhibited stronger DVM following the Bythotrephes invasion. The shift was likely the result of the avoidance of Bythorephes by the most common taxa in the lake [44]. Both D. thomasi and M. edax did not migrate as far down in the water column as Daphnia retrocurva in 2014, suggesting that they may have been more capable of avoiding predation in the reduced light levels present in the metalimnion compared to the cladoceran. Although both cyclopoid species are omnivores, they preferentially prey on smaller plankton taxa [45,46,47] that likely did not migrate in response to Bythotrephes. Cyclopoid population reductions that occurred when Bythotrephes reached their peak abundance in 2015 may have been a function of both predation by Bythotrephes and competition for smaller invertebrate prey, such as Bosmina and rotifers [45,48]. Half of the reduced D. thomasi population remained in the epilimnion during the day in August of 2015, suggesting that they may have needed to risk Bythotrephes predation in order to forage sufficiently. Less mobile nauplii and juvenile cyclopoids exhibited small increases in DVM activity in 2014 (unpublished data). Although nauplii are not a preferred prey item for Bythotrephes, they can be consumed at fairly high rates if nothing else is available [42].

Diaptomid responses to the Bythotrephes invasion varied by species in Lake Champlain. Bythotrephes have a lower selectivity for faster swimming prey, such as Diaptomids, and exhibit lower capture rates, even in well-illuminated conditions [43]. Reductions in mean Diaptomid body size related to the alewife invasion [8] may have reduced the swimming speed of individuals in Lake Champlain, making them less capable of evading Bythotrephes predation, particularly the epilimnetic species, L. minutus, which increased its DVM behavior in response to the Bythotrephes invasion and had its population reduced to nearly undetectable numbers when Bythotrephes reached their peak abundance in 2015. Leptodiaptomus sicilis, however, remained in deep water and did not change its DVM behavior in response to Bythotrephes. Changes in the vertical distribution of L. minutus and not L. sicilis have also been observed in Lake Michigan [44]. Given both its mobility [43] and habitat preferences [32], there was likely no need for L. sicilis to alter DVM behavior that already provides minimal overlap with Bythotrephes.

4.3. Vertical Distribution Shifts Associated with Cercopagis Invasion

The crustaceous zooplankton community responded to the Cercopagis invasion with increased DVM behavior, similar to the response to the Bythotrephes invasion. Over 50% of the population was found below the epilimnion during the day, moving into the epilimnion at night during the first invasion year (2019). Less than 20% of the community exhibited this behavior in the 2016 reference year, suggesting avoidance of the epilimnion during the day was a response to Cercopagis presence. The relative shift in DVM activity was much more prominent in cladocerans than in copepods. Given the smaller size of Cercopagis relative to Bythotrephes, it is thought to have more of an impact on smaller zooplankton prey taxa, such as Bosmina and Ceriodaphnia [19]; both taxa responded with migration into the metalimnion during the day following the Cercopagis invasion in Lake Champlain. Although Cercopagis is considered to be less capable of handling larger prey items, such as Daphnia [19], D. retrocurva also responded to the invasion by shifting their daytime vertical distribution lower in the water column. Due to their high abundance, size reductions associated with the alewife invasion in Lake Champlain [8], and slower swimming speeds compared to other daphnids [43], D. retrocurva is likely susceptible to Cercopagis predation, as suggested by observed density reductions associated with Cercopagis in Lake Ontario [19,21].

In general, copepods showed increased DVM into the hypolimnion during the day following the Cercopagis invasion compared to the reference year (2016). The difference in daytime vertical distribution was largely a product of movements of D. thomasi, the most common copepod in Lake Champlain. Diacyclops thomasi migrated deeper on average than smaller Cladocera species did in 2019, suggesting that this movement was to avoid predation by Cercopagis rather than foraging on prey displaced by Cercopagis. Reductions in D. thomasi densities associated with Cercopagis have been observed in Lake Ontario [21,30,49]. Despite having similar morphology and feeding habits as D. thomasi, M edax did not exhibit the same level of DVM activity when Cercopagis were abundant in 2019 as D. thomasi. Mesocyclops edax are generally larger than D. thomasi in Lake Champlain (mean body lengths in the Main Lake in August 2023 were 1.07 and 0.77 mm, respectively, n = 8 each), possibly making them more difficult prey items for Cercopagis to process and reducing the need for vertical migration. Tropocyclops prasinus mexicanus did not exhibit DVM in 2016 or 2019, even with density reductions related to the Cercopagis invasion [9]. As the smallest copepod species in Lake Champlain, readily consumed by larger cyclopoids [45], T. prasinus mexicanus may have faced equal or greater predation risk by migrating deeper during the day than remaining in illuminated surface water with abundant Cercopagis; this is reflected in the weak reverse DVM they exhibited during years prior to the Cercopagis invasion, a strategy used to avoid predators exhibiting normal DVM patterns [50].

4.4. Vertical Distribution Trends with Continued High Abundance of Cercopagis

DVM responses varied after several years of high Cercopagis densities in Lake Champlain. Increased daytime migration depths were most prevalent in D. retrocurva and copepods, primarily in 2023. Both Bosmina species, however, reduced their DVM activity in 2023 and 2024. Our results suggest that Cercopagis may preferentially seek larger prey items or that smaller prey items are less susceptible than previously thought in Lake Champlain. Although Cercopagis can occur in relatively high proportions in and above the metalimnion, they are not thought to vertically migrate in Lake Ontario and did not occur below the thermocline in previous studies [20,21,30]. But, a portion of the Cercopagis population exhibited daytime vertical migration in Lake Champlain in 2023 and 2024. The reduced density of Cercopagis in the epilimnion during the day in 2023 and 2024 may explain the reduced migration activity of Bosmina during that year. Due to their small size, Bosmina have more invertebrate predators than larger taxa and are less susceptible to fish predation [51,52,53].

Given the limited number of observations of Cercopagis DVM in North America [54], it is difficult to determine the underlying causes for the observed behavior in Lake Champlain in 2023 and 2024. Although they exhibit extensive DVM in their native range [17], studies suggest Cercopagis generally do not exhibit strong migration patterns in Lake Ontario or Lake Michigan [20,21,30,55,56]. Cercopagis DVM in Lake Champlain could be related to predation risk or food availability. Cercopagis are preyed on extensively by Baltic herring (Clupea harengus membras) in the Baltic Sea [57,58]. Although alewife incorporate Cercopagis into their diet in Lake Ontario, they have not been shown to impact Cercopagis densities or induce Cercopagis vertical migration [25,59]. Storch et al. [60] found that Cercopagis were selected against by alewife and were only consumed because they were widely available. The prevalence of alewife in Lake Champlain also does not appear to have impacted Cercopagis densities, suggesting that their vertical migration may not be related to alewife. Additionally, movement into the hypolimnion by Cercopagis may increase interactions with hypolimnetic predators, such as rainbow smelt and mysids [25,59]. Given their morphological similarities to Bythotrephes, Cercopagis should be a more efficient predator in illuminated surface waters [22] and would likely exhibit reduced feeding rates in the darker hypolimnion. Cercopagis continued to vertically migrate deeper during the daytime in 2024, but did not enter the hypolimnion. The density of Cercopagis was an order of magnitude greater in the 2023 sampling event than in 2019 or 2024, suggesting that their increased migration in 2023 could have been a density-dependent strategy to avoid competition or cannibalism.

4.5. Implications for Higher Trophic Levels

The mysid (Mysis diluviana) population in Lake Champlain experienced a large density crash, correlating with the D. polymorpha invasion and rotifer declines in the 1990s [8,61]. Lake Champlain mysids incorporate crustaceous zooplankton into their diet, appearing to prefer daphnids and calanoid copepods [62,63]. Mysids prefer water temperatures between 6 and 8 °C [64], likely restricting them to the hypolimnion in Lake Champlain during thermal stratification. Although mysids can tolerate higher water temperatures (>15 °C) for short periods of time to feed if they detect abundant prey in the area [64], their mortality rates increase and their feeding rates are significantly reduced with increasing temperature [65]. Zooplankton DVM to deeper waters induced by invasive Cercopagidae likely resulted in more spatial overlap between zooplankton and mysids in Lake Champlain. Although mysids exhibit normal DVM patterns in lakes, migrating from the benthos at night, they can ascend much more rapidly in the water column than smaller crustaceous zooplankton [66,67,68], potentially providing temporary spatial overlap for feeding on zooplankton within their preferred temperature range during thermal stratification. Additionally, mysids seasonally select for Cercopagis as prey in Lake Ontario [69]. In Lake Champlain, Cercopagis exhibited DVM below the thermocline in 2023 while also correlating with higher abundances of taxa moving below the thermocline during the day. Diet analysis studies on Lake Champlain mysids will help determine if they have been influenced by the Cercopagis invasion.

Fisheries management on Lake Champlain has long been focused on coldwater species, particularly lake trout (Salvelinus namaycush) and Atlantic salmon (Salmo salar). Both species were extirpated from the lake by the 1900s due to anthropogenic alterations [27,70]. Prior to coldwater fish restoration programs initiated in the 1970s, the coldwater sportfishery was restricted to rainbow smelt (Osmerus mordax), which now act as the primary coldwater forage fish [27,71]. As coldwater stenotherms, smelt and salmonids cannot tolerate the warm temperatures in the epilimnion of Lake Champlain during late-summer thermal stratification, restricting their foraging to what is available deeper in the water column. Smelt feed on crustaceous zooplankton and mysids [72,73] and could benefit from higher densities of prey items in the hypolimnion associated with DVM. A barrier to lake trout restoration efforts in Lake Champlain has been the survival of wild juvenile fish. Diet analysis revealed that wild age 0 lake trout in Lake Champlain feed almost exclusively on mysids and zooplankton [74]. Increased DVM activity in zooplankton related to the cercopagid invasion may have resulted in increased availability of zooplankton in the hypolimnion, potentially influencing juvenile lake trout productivity, acting both as a direct food source and as a food source for their coldwater prey (mysids and rainbow smelt).

Alewife (Alosa psuedoharangus) invaded Lake Champlain in the mid-2000s and were present throughout this study. As pelagic planktivores, alewife foraging behavior may be influenced by the vertical distribution of their prey. Adult alewife select Bythotrephes as prey in the Great Lakes [60], and alewife predation was a potential cause of the sudden decline of Bythotrephes in Lake Champlain occurring in 2016. Conversely, alewife select against Cercopagis in the Great Lakes [60], which may have allowed Cercopagis to persist in high densities since their initial invasion in Lake Champlain. Given the tendency of the zooplankton community to remain near the surface both day and night in 2013, six years after alewife became abundant in Lake Champlain [6], it appears that alewife did not cause any long-term shifts in zooplankton DVM behavior. An exploration of DVM of Lake Champlain forage fish conducted by Simonin et al. [75] suggests that alewife in the Main Lake did not undergo vertical migration prior to the cercopagid invasions, remaining near the surface both day and night [75].

Zooplankton DVM associated with the cercopagid invasions in Lake Champlain appears to have reduced daytime alewife prey abundance in the epilimnion. In Lake Ontario, alewife exhibit DVM, following vertically migrating crustaceous zooplankton [76]. If alewife adopted similar DVM behavior to chase their prey in Lake Champlain, spatial overlap with smelt and lake trout would have increased. Given that alewife make up a substantial portion of juvenile lake trout diet in Lake Champlain [74], this could be another example of how cercopagid-induced DVM could impact the coldwater fishery. Beyond changes in spatial distributions observed in this study, long-term monitoring data revealed that extreme alterations of the Lake Champlain lower food web have occurred since the alewife invasion, related to an extreme flooding event in 2011 [47] and the cercopagid invasions in 2014 and 2018 [9]. Although these alterations have greatly influenced the composition and availability of prey for alewife in Lake Champlain, they have not been considered in recent modeling efforts linking alewife to lake trout restoration [77].

4.6. Potential Patterns Associated with Climate Change in Lake Champlain

The warming climate will influence the thermal structure of Lake Champlain, influencing the vertical distribution of multiple trophic levels. Warming may result in a deeper, thicker summer epilimnion and a deeper, thinner hypolimnion. The depth of the thermocline can determine the depth selection of zooplankton taxa [78]. Van Gool and Ringelberg [79] suggest that the kairomones that induce DVM cannot pass through the thermocline, and the concentration of kairomones in the epilimnion is inversely correlated with the thermocline depth. If predator biomass remains the same, DVM responses may be reduced in a larger epilimnion. Additionally, if there is a thicker epilimnion, DVM to escape visual predators may not require movement into the hypolimnion to avoid illuminated water, reducing its costs associated with lower temperatures for several taxa [14]. Daphnia that do not migrate have higher protein content than migrating individuals because they always remain in warm food food-rich waters [80]. The warming climate may continue to shift the Lake Champlain phytoplankton community toward increased dominance by cyanobacteria taxa adapted to warmer waters [81], which could alter interspecific competition of zooplankton grazers based on differences in feeding modes [82,83].

5. Conclusions

Both the Bythotrephes and Cercopagis invasions altered the zooplankton vertical community structure in Lake Champlain by altering DVM behavior in several taxa. Deeper average daytime depths of the most common taxa in Lake Champlain (D. retrocurva, B. longirostris, and D. thomasi) can temporarily increase the relative abundance of non-migrating taxa (Ceriodaphnia and T. prasinus mexicanus) while simultaneously decreasing overall zooplankton density in the epilimnion. These alterations in zooplankton vertical structure can affect multiple trophic levels. Density reductions of phytoplankton grazers, such as Daphnia, by Cercapagids may also result in increased algal blooms. Even if herbivorous zooplankton densities were maintained, their migration into deeper strata should temporarily relieve grazing pressure for phytoplankton in the epilimnion. Additionally, zooplankton migration into deeper strata may transport epilimnion-sourced energy that was previously unavailable to deep water planktivores. The DVM alterations associated with Bythotrephes were more intense than those associated with Cercopagis but short lived. However, the changes associated with Cercopagis appear to have been maintained for at least five years. By inducing DVM in several zooplankton taxa, continued high Cercopagis densities may ultimately boost the productivity of higher trophic levels in the hypolimnion and improve the coldwater fishery in Lake Champlain. Consequently, Cercopagis may also indirectly result in increasing surface algal biomass by temporarily reducing the abundance of herbivorous zooplankton at the surface during the growing season. Future studies will help determine the long-term top-down and bottom-up effects of increased DVM in Lake Champlain.