Effect of Drying Period and Soil Moisture on Egg Hatch of the Tadpole Shrimp (Notostraca: Triopsidae) LISA L. FRY

AND

MIR S. MULLA

Department of Entomology, University of California, Riverside, California 92521

J.

Econ. Entomol. 85(1): 65-69 (1992)

ABSTRACT Tadpole shrimp (Triops spp.) are potential biological control agents against larval mosquitoes in temporary ponds and flood-irrigated fields. In some rice field situations, however, they may become pests that uproot and eat young rice plants. In cursory observations, it has been reported that tadpole shrimp eggs do not readily hatch on flooding when the soil or substrate containing eggs is moist before flooding. The relationship between drying (moisture content) of soil and tadpole shrimp hatch was determined in studies conducted in mesocosms at the University of Cali fomi a Aquatic and Vector Control Research facilities at Riverside and at Oasis in the Coachella Valley of southern California. In laboratory hatching trials, an increase in hatch of Triops longicaudatus (LeConte) with declining soil moisture content was demonstrated (t = 8.4, P < 0.001; ,.z. = 0.76). In field trials in mesocosms at Riverside, egg hatch increased with increased drying period and declining soil moisture content (G = 29.8, P < 0.01). No hatch of eggs occurred after 3 d of drying, when soil moisture content was high, but a high level of hatching occurred after 7 and 14 d of drying, when soil moisture declined to low levels. At Oasis, soil moisture did not decrease with drying time because of porous soil and capillary action of water from adjacent flooded mesocosms and thick vegetation covering the pond bottoms. Therefore, hatch rates at Oasis were not associated with the length of the drying period (G = 35, P > 0.05). The effect of flooding regimens on tadpole shrimp hatch is an important facet of shrimp biology that may be used to predict or control their densities in flood-irrigated habitats. KEY WORDS

THE

TADPOLE

SHRIMP,

Crustacea, Triops longicaudatus,

Triops longicaudatus

(LeConte), has evolved a mode of life that depends on intermittent flooding and drying of the habitat in which it lives. Triops spp. have a worldwide distribution and are commonly found in ephemeral bodies of water and flood-irrigated fields. Tadpole shrimp have a high reproductive potential and produce resistant eggs that may lie dormant in soil for extended periods, remaining viable for several years. In various situations, these crustaceans have been described as pests or as beneficial organisms. T. longicaudatus are sporadic pests in rice fields in the United States (Grigarick et al. 1961, Rosenberg 1947), and T. cancriformis (Bose) are pests of rice seedlings in Italy, Spain, and France (Longhurst 1955). Damage occurs when the shrimp feed upon and uproot young seedlings, and the characteristic increase in suspended silt as a result of their burrowing action decreases the photosynthetic capacity of the rice plant (Rosenberg 1947). Tadpole shrimp can be controlled after hatch by chemical applications (Grigarick et al. 1961), but nonchemical control methods have not been examined. On the other hand, tadpole shrimp have been considered to be beneficial in rice fields in Japan, where cul-

hatch rates, population management

tural practices differ from those in the United States. In Japan, rice seedlings are transplanted from nurseries to fields at a size where the shrimp are no longer able to dislodge or harm them. This rice culture practice promotes the use of tadpole shrimp for weed control because they feed upon and uproot young weed plants (Takahashi 1977). Tadpole shrimp have been noted to devour mosquito larvae (Mail 1934, Maffi 1962, Scott & Grigarick 1979, Tietze & Mulla 1989), which led to the supposition that they could be used in mosquito control efforts. In recent studies, Tietze & Mulla (1989, 1990, 1991) have demonstrated reduction in mosquito populations by T. longicaudatus by active predation. Tietze & Mulla (1991) also reported that the surface swimming behavior of T. longicaudatus interferes with the oviposition activity of gravid mosquitoes on the water surface. In larval development habitats, such as date gardens and flooded ponds and pastures, tadpole shrimp show promise as mosquito control agents. Pond or field flooding regimens may directly affect hatch rates of tadpole shrimp eggs by providing soil conditions either unsuitable for or amenable to hatching. As in other floodwater or-

0022-0493/92/0065-0069$02.00/0

© 1992 Entomological

Society of America

66

JOURNAL

OF ECONOMIC

ganisms, desiccation of eggs is an integral part of T. longicaudatu$... biology in natural situations. Qualitative obserVations suggest that their eggs do not readily hatch on flooding when the soil or substrate containing eggs is moist (Igarishi 1970). Therefore, in intermittent flooding systems with controlled irrigation, soil moisture content may be managed by manipulation of flood intervals. Potentially, control of soil moisture content may be used to maximize shrimp numbers or to eliminate their hatch altogether. To determine the relationship between drying (moisture content) of soil after flooding and tadpole shrimp hatches, field-collected soil samples of varying moisture contents were brought into the laboratory where they were reflooded and tadpole shrimp hatches assessed. Field experiments relating drying time, soil moisture, and hatch rates were then completed at two locations: the University of California Aquatic and Vector Control Research facilities at Riverside, and at Oasis in the Coachella Valley of southern California. Materials and Methods Laboratory Hatching Experiment. To examine the relationship between soil moisture and hatching of T. longicaudatus, hatch rates of eggs collected from soil samples varying in moisture content were determined. Three experimental mesocosms in Riverside were drained on 13 August 1990 and allowed to dry for 16 d. Standing water disappeared from these mesocosms within 3 d of the time water was shut off. Daily soil samples were taken from the bottom of each mesocosm beginning on 18 August (5 d after water was shut off) and continuing for 11 consecutive d. After 11 d, the soil was completely dry and crumbly and could not be collected in the coring device. Six samples were collected daily from random locations in the centers of each of three mesocosms where dead shrimp were found. The centers of the mesocosms have been shown to yield the highest egg numbers (L.L.F., unpublished data). The top 2 cm of soil were gathered in a brass coring pipe 4.3 cm in diameter. Each core contained 29 cm3 of soil, and each sample consisted of two cores (totaling 58 cm3 of soil), providing an adequate number of eggs in the soil sample for subsequent hatching analysis. Samples were placed in sealable plastic bags for transport to the laboratory. To determine hatch rates in the laboratory, three soil samples from each mesocosm were immediately flooded with distilled water (2226°C) in the plastic bags in which they were transported and held for 48 h, after which the number of hatched shrimp in each bag was counted. At this length of time following hatching, the tadpole shrimp were in a metanauplius

ENTOMOLOGY

Vol. 85, no. 1

larva stage of development discernible by the naked eye. The remaining three samples from each mesocosm were weighed in the plastic bags in which they were transported, and the weight of the bags was subtracted. The bags containing the samples were then opened and placed in an Imperial II radiant heat oven at 35 ± 2°C. After 48 h drying time, they were removed and weighed, again subtracting the weight of the bag. The percentage water content in the soil at the time of collection was determined by calculating the water-weight loss after the drying period as a percentage of the initial weight. To obtain a standard hatch rate, these dried samples were then flooded and held for 48 h, after which the numbers of shrimp hatched were counted. The data were log-transformed [In(n + 1)] and analyzed using multiple regression to determine the relationship between soil moisture and hatch rates. The standard consisted of the numbers hatched from the samples that were dried for 48 h in the oven before flooding. Differences among the three replicate mesocosms were examined. Field Studies. Mesocosms at two locations in southern California were used in these experiments to determine the effect of drying intervals on the hatch of natural populations of eggs of T. longicaudatus. Six mesocosms at the Riverside facility were used beginning 7 July 1989. These mesocosms have silty soil and are devoid of vegetation. Twelve mesocosms at Oasis were used beginning 18 August 1989. Located in the lower desert, these experimental mesocosms are covered with low vegetation and are subject to high temperatures in the summer months. The soil type in these mesocosms is sandy loam. The mesocosms at each site were randomly placed in three groups (two replicates in Riverside, four in Oasis) and flooded in a pretrial regimen to determine the extent of hatch of natural populations of tadpole shrimp eggs. Four days after flooding, tadpole shrimp densities were assessed. All sampling was done between 1130 and 1430 hours (PST) at both locations. To sample the shrimp, three D-net drag samples were taken along fixed lines in each mesocosm. Maximum and minimum daily water temperatures and pH were recorded. After sampling, the water was turned off, and the mesocosms were allowed to dry by evaporation and percolation. It took from 48 to 96 h at Riverside and from 72 to 96 h at Oasis for the standing water to disappear completely. The drying periods were then measured from the time all the surface water disappeared to the time of the next flooding. Immediately before reflooding, soil samples were collected. In Riverside, one 50-g soil sample was collected from the top 1 cm of soil in each of six mesocosms. The samples were taken from the dampest area of the bottom of the mesocosm. At Oasis, four 50-g soil samples were

February

1992

FRY & MULLA: SOIL MOISTURE RELATED TO HATCH OF TADPOLE SHRIMP

taken from the centers of quadrants in each of the 16 mesocosms and the mean percentage water content of soil of each mesocosm was determined. The soil water content was determined by the method described above; however, these samples were dried at 50°C for 48 h. After drying for a 3-, 7-, or 14-d interval, the mesocosms at both sites were reflooded. Drying intervals were assigned randomly to mesocosms before the start of the experiment. Four days after flooding, shrimp densities in each mesocosm were assessed with D-net samples as described above. Data were analyzed using a G test for goodness of fit to determine the effect of drying period on shrimp hatch. Regression analysis was used to determine the linear relationship between soil moisture and the length of the drying period. Fifteen days following the conclusion of the experiments in Riverside and 23 d after completion at Oasis (to ensure complete drying), all the test mesocosms were reflooded for the third time to attempt replication of the experiment. Because of lack of adequate hatch on this flood, the mesocosms at Oasis were dried, and soil samples were taken from the center of each mesocosm to determine egg densities after the last flooding. Soil was collected to 5 cm depth for a total volume of 452 cm3. Eggs were collected from the soil samples by flotation, and the eggs present in each sample were counted. Results

67

70

1 .c •• u ••

-ID ID_

60 50

.cu •• II) II) ••

40

.1:1::0

E::o.!!

30

=El•.

20

co c E

10 0 1-9

10-18

19-27

28-36

37·45

"" soli moisture

Fig. 1. Mean number of eggs of T. longicaudatus hatching from samples in each of five soil moisture classes ranging from 1 to 45% moisture. Bars represent standard errors.

when soil moisture was >25.5%, but after 7 d (soil moisture, 4-16%) and 14 d (soil moisture, 0.05). Moisture did not decline at Oasis because the soil was very porous and adjacent mesocosms were flooded throughout the drying

Laboratory Hatching Experiment. Standard hatch rates were obtained for eggs retrieved in core samples from the Riverside mesocosms. The oven-dried samples were subjected to hydration in the laboratory. Regression analysis showed that the hatch rates from these dried samples remained constant throughout the duration of the experiment (rZ ::::0.16, F ::::1.89, P > 0.1). The analysis also showed that mesocosm 2 had fewer eggs than mesocosms 1 and 3 (t ::::2.25, P < 0.05). Table 1. Relalionship between drying periodo and egg Multiple regression showed that hatch rates in halch of T. longicaudatw in mesocosms al Riverside, Calif.b the laboratory were inversely related to soil moisture content from the field (t :::: 8.4, P < Mean no. shrimpc 0.001; ,2 ::::0.76). The regression equation is y :::: Mesocosm % Soil moistured Pretrial After drying 4.7 - O.llxI - O.85x2' where Xl is soil moisture Drying period: 3 d and X2 is the egg number in mesocosm 2. The 25.7 0.0 29.2 1 range of soil moisture percentages determined 25.5 12.3 0.0 2 from daily sampling was divided into five equal Drying period: 7 d classes. The mean numbers of eggs hatching in 22.0 22.3 16.0 3 4 7.7 52.7 4.1 the samples from each moisture class are shown Drying period: 14 d in Fig. 1. It is obvious that hatch rates are highest 1.7 0.08 0.33 5 when soil moisture is 24.3% moisture from either Riverside or Oasis mesocosms (Tables 1 and 2). This is a particularly interesting result, as the average soil moisture in the field studies was probably overestimated for the entire mesocosm because of the concavity of the bottoms. The elevated edges of the mesocosms were likely to have areas drier than the estimates given. However, the centers (where the moisture was the highest and where the majority of eggs was found) were most accurately represented by the soil moisture figures shown (Tables 1 and 2). Although laboratory results indicate that very limited hatching can occur at high soil moisture contents (up to 45%) (Fig. 1), suppression of tadpole shrimp in the field can be achieved by not allowing moisture content to decrease