P. concinna seasonal and spatial distribution
P. concinna is the most abundant Euchaetidae in the study area, comprising more than 70% of the total Euchaetidae population (Wong unpublished data). It is one of the four species of Euchaetidae (P. concinna, E. rimana, P. plana and E. indica) recorded in Hong Kong’s coastal waters
[20, 24, 32, 33]. Along the Chinese coast, P. concinna has been reported from the East China Sea, the Taiwan Strait and the South China Sea
[18, 19, 34]. The species is considered to be limited to the Indo-Pacific in distribution
Our results are in agreement with previous studies that dense populations of P. concinna appear only in winter and spring in Hong Kong
[20, 22, 32, 33], when the water temperature is low (16–24°C). In comparison, Euchaetidae in the temperate oceans of the East China Sea and the Yellow Sea exhibits an opposite seasonal pattern. Both P. concinna and E. rimana are carried by the Kuroshio Current into the coastal areas of the East China Sea, and reach peak densities in summer and autumn
. P. concinna occurs throughout the year in the northern Taiwan coast which is influenced by the Kuroshio Current year-round
, but their density is higher in autumn and winter when water masses from the East China Sea flow toward Taiwan
The abundance and distribution of copepods are known to be influenced by hydrographic conditions
[36, 37]. The density of P. concinna is lower in the semi-enclosed Tolo Harbour than the east and southeastern waters that are fully exposed to water currents from the South China Sea. The low densities at most times of the year, together with the pattern of shoreward decrease in density, suggest that P. concinna is transported into the coastal seas of eastern Hong Kong by water currents from the South China Sea, and the scarcity of P. concinna in the inner parts of Tolo Harbour can be explained by loss during physical transport.
On the other hand, trophic interactions may also influence the abundance of P. concinna in Tolo Harbour. The copepod community in Tolo Harbour is dominated by small copepods such as Parvocalanus crassirostris and Paracalanus parvus. Such a plethora of small calanoid copepods suggests that carnivorous copepods such as P. concinna will not be limited by food
[8, 9, 38, 39]. Chang et al. proposed that the density of large copepods, such as P. concinna, is controlled directly by fishes, while the density of small copepods is controlled by invertebrate predators. Several observations suggest that fish predation may contribute to the scarcity of large copepods such as P. concinna in Tolo Harbour and other nearshore areas of Mirs Bay. First, dense populations of the planktivorous Ambassis (glassfish) occur in Tolo Harbour throughout the year
. Secondly, copepods are known to be the major prey of larval and juvenile fishes which appear in large numbers in Tolo Harbour during late winter and early spring
Both male and female P. concinna exhibit DVM. The congener Paraeuchaeta norvegica has been observed to perform DVM over depths of hundreds of meters
[43–45], but there are also reports of non-migrating populations of this species
[12, 16]. The vertical distribution of P. norvegica and Paraeuchaeta elongata[10, 46] has been shown to be affected by ovigerity. The reproducing females of the calanoid copepod Eurytemora hirundoides have also been found to occupy deeper waters in the Archipelago Sea
. Ovigerous Eurytemora are highly selected by the planktivorous Baltic herring
, and Vuorinen
 has proposed that visual predation is the most probable cause of the difference in vertical distribution between ovigerous and non-ovigerous females. However, the results of this study show that both ovigerous and non-ovigerous females of P. concinna perform DVM.
DVM in crustacean zooplankton is widely considered to be an antipredator behavior
[15, 45, 50–53]. Small pelagic organisms have no means to hide from visual predators, such as fish, in well-lit surface waters during the daytime, and must minimize predation risk by seeking refuge in deeper depths that have less light penetration
. Light is, therefore, commonly considered as a proximal cause or signal for zooplankton DVM
[55, 56]. Light intensity drops dramatically in the first 5 m, and very little light penetrates into waters below 15 m. The WMD of P. concinna is below 15 m during daytime. This suggests that P. concinna adjusts its vertical distribution to achieve an optimal balance between the energy cost of DVM and the benefit of reducing visual predation risks. Vuorinen
 suggested that it may be a viable strategy for copepods to stay in deeper waters at all times if the temperature and food supply do not vary greatly between the deep and surface layers. While there was little temperature change with depth in the well-mixed, shallow waters of eastern Hong Kong, more than 50% of the Acrocalanus and Paracalanus/Parvocalanus populations remained in the upper 0–5 m during both day and night, so the need for food may force P. concinna females to ascend to the surface at night. On the other hand, while adult male P. concinna do not feed, they still perform DVM and migrate to the surface at night. Such DVM behavior may therefore be an intrinsic behavior not driven by the need to feed, but other factors such as light. It is also possible that the P. concinna males follow the migrating females to increase the chance of mating, as male copepods of various species, e.g. Temora longicornis, Centropages typicus, Pseudocalanus elongatus, and Acartia tonsa, have been found to track and pursue female copepods through pheromones or hydromechanical signals for mating
Diel patterns in gut fullness suggests that female P. concinna only feed actively at night. Diel feeding rhythms have been reported in Paraeuchaeta in both laboratory
 and field studies
[13, 16]. Marked increase in the gut fullness of female P. concinna appears to coincide with their nocturnal ascent in the early part of the night, so the prevalence of individuals with relatively empty guts during the day may be due to a lack of access to prey. On the other hand, diel variations in gut fullness may not be associated with DVM. Female P. concinna stay in significantly deeper waters than Acrocalanus and Paracalanus/Parvocalanus during the day, but appear to share the same depths with Canthocalanus copepodites during both day and night. The absence of Canthocalanus in the guts of female P. concinna in daytime therefore suggests that the predators feed only at night, and implies that the diel variations in gut fullness was not the result of the predators migrating into and out of the prey-rich surface layer. Other studies have found that copepods exhibit diel variations in the gut contents regardless of whether they perform DVM. The gut fluorescence of both surface dwelling and migrating zooplankton in the Bedford Basin reaches peak values only at night
. Diel feeding rhythm in two species of Calanus in the Bering Sea is also independent of DVM
, and though Calanus pacificus in Dabob Bay performs DVM and enters into the surface layer 2.5 h before sunset, its gut pigment content only increases substantially after sunset
Diel feeding rhythm in zooplankton is often regarded as a strategy to avoid the accumulation of pigments in daytime
[62, 63], the purpose of which is to reduce the risk of attack by visual predators. Laboratory and field experiments with the rainbow trout Salmo gairdneri have confirmed that visual predators select the most pigmented calanoid copepods as prey
. Bollens and Stearns
 reported that the gut fullness of Acartia hudsonica is lower in the presence than in the absence of fish. A. hudsonica and Acartia tonsa reduce their gut fullness in the presence of fish exudates, and the response is observed only when the light level is sufficient for visual predation
. These results suggest that P. concinna females may stop feeding during the day to reduce the chances of being attacked by fishes.
Prey composition and selectivity
Small calanoid copepods constitute ~40% of the prey in the guts of female P. concinna. The estimate is considered to be conservative as some unidentified items found in the guts may also be the remains of calanoid copepods. The importance of calanoid copepods as food for female P. concinna is in accordance with previous findings on the natural diets of Paraeuchaeta species
[9, 38]. Copepods including Metridia gerlachei, Calanoides acutus, Euchaeta spp., Oncaea spp., and Oithona spp. made up 80–90% of the prey consumed by adults and copepodites (C5) of Paraeuchaeta antarctica in Antarctic waters
. Øresland and Ward
 also reported that copepods form 46–99% of the diets of P. antarctica, Paraeuchaeta farrani, Paraeuchaeta rasa, and Paraeuchaeta biloba in South Georgia.
P. concinna females feed on a variety of small calanoid copepods including Canthocalanus, Centropages, and Subeucalanus, but show strong preference for Acrocalanus, Parvocalanus, and Paracalanus. In laboratory feeding experiments, Yen
 found that prey size is an important factor in the dietary selectivity of P. norvegica. The prosome length of preferred copepod prey is usually ~70% the length of the second basipodal segment of the maxilliped of the predator. A similar proportion of 65% was found for P. antarctica, which exhibits the highest feeding rates on copepods with prosome length of 1.2 mm
. In this study, the prosome length of Acrocalanus (~0.6–0.7 mm) is about 65–75% the length of the second basipodal segment of the maxilliped (~0.92 mm) of P. concinna females, agreeing with the optimal prey size proposed by Yen
. On the other hand, the prosome lengths of Paracalanus and Parvocalanus (~0.3–0.4 mm) are only 33–40% the length of the second basipodal segment of the maxilliped of P. concinna. These smaller copepods may therefore be more suitable prey for the smaller late copepodite stages of P. concinna (C4 and C5). The prosome length of the negatively selected Canthocalanus copepodites (~0.8–1.0 mm) is ~85% the length of the second basipodal segment of the predator maxilliped.
P. concinna females in this study did not feed on cyclopoid copepods, even though small cyclopoid copepods including Oithona and Oncaea were common in the gut of P. antarctica and may be the major prey of this predatory copepod, especially the copepodite stages
[9, 38, 68]. Yen
 reported that the cyclopoid copepods Oithona and Corycaeus are not preferred prey of P. antarctica, and proposed that the intermittent and darting movement of cyclopoid copepods are not easily detected by tactile predators. The copepod community in Mirs Bay is dominated by calanoid copepods (> 80% of the total copepod population), and Oithona and Oncaea species only comprise ~ 7% and 5% respectively of the total copepod populations
. The reason for the absence of cyclopoid copepods in the gut of female P. concinna may be a combination of the prey’s low abundance, small size (prosome length ~0.2–0.3 mm), and less detectable swimming behavior.
Digestion time and feeding rate
Tönnesson et al.
 found a gut evacuation rate of 0.080 h-1 or digestion time of 12 h at 15°C for P. norvegica. This is much longer than the digestion time of ~ 5 h recorded at 18°C for P. concinna in this study. The relationship between the predator size and digestion time in copepods is unclear, although some studies have found no relationship between digestion time and predator size
[69, 70]. A common conclusion from previous studies is that digestion time varies with temperature
[71–75] and prey size
. The higher water temperature in the subtropical seas of Hong Kong may allow shorter digestion times for female P. concinna, but it can also be argued that Paracalanus and Pseudocalanus, the copepod prey used to estimate the digestion time of P. norvegica, are bigger than Acrocalanus and Paracalanus/Parvocalanus, the natural prey of P. concinna. On the contrary, Yen
 reported a considerably higher gut evacuation rate of 0.43 h-1 and correspondingly very short digestion time of ~ 2 h for P. norvegica feeding on cod larvae at 7.5°C. As fish larvae lack the chitinous exoskeleton of copepod prey, they may be more easily digestible.
While the daily feeding rate of copepods changes with temperature and prey availability
[35, 71–73], our estimated daily feeding rates for female P. concinna feeding on Acrocalanus (~4.7 prey predator-1d-1) and Paracalanus/Parvocalanus (~4.4 prey predator-1d-1) are within the range of values reported for P. norvegica. Olsen et al.
 reported a feeding rate of 3.6 prey predator-1d-1 at 7–10°C in the laboratory. Tönnesson et al.
 reported in situ feeding rates of 1.4–5.2 prey predator-1d-1 at 5°C and 6.2–8.6 prey predator-1d-1 at 15°C.
P. concinna females remove ~ 4.3% of both the Acrocalanus and Paracalanus/Parvocalanus standing stock daily in the coastal sea of eastern Hong Kong. The estimates were conservative as only predation by adult females was considered. Øresland
 found that the feeding rates of copepodites are higher than that of adults in Antarctic waters. The predation impact by the entire P. concinna population may therefore be higher as Paracalanus and Parvocalanus are small enough to be eaten by late copepodite stages, which are frequently more abundant than adults (Wong unpublished data). Predation impact estimated for female P. concinna in this study is within the range of values reported for other predatory copepods. Tönnesson et al.
 reported a predation impact of 2.0–6.5% on small copepods by P. norvegica in the Skagerrak. The predation impact of Tortanus spp. on small copepod populations ranged from 1% in San Franscisco Estuary
, to 2.7% in Fukuyama Harbour
Copepods are also the major prey of chaetognaths. While the predation impact of chaetognaths on copepods can reach 6% in Northern Chile
 and 7.8% in the eastern Mediterranean
, a recent study conducted in Tolo Harbour and Mirs Bay showed that the predation impact of chaetognath on the copepod population is < 1%
. This finding suggests that P. concinna may play a more important role than other invertebrate predators in regulating the populations of small calanoid copepods in the eastern waters of Hong Kong, especially during winter and spring.