This methodological paper evaluates the productivity of a social wasp colony by examining the number of meconia per 100 cells of comb, to estimate the total number of adults the wasps produced. The associated video describes how to search for Vespula wasp nests, a method developed by amateur wasp chasers.
For vespine wasps, colony productivity is typically estimated by counting the number of larval cells. This paper presents an improved method that enables researchers to estimate more accurately the number of adults produced, counting the number of meconia (the stools left in the cells by wasp larvae when pupating into adults, per 100 cells) in each comb. This method can be applied before or after colony collapse (i.e., in active or inactive nests). The paper also describes how to locate wild Vespula wasp colonies by "flagging" wasp baits and chasing the wasp collecting them, using a method traditionally performed by local people in central Japan (as illustrated in the associated video). The Vespula chasing method described has several advantages: it is easy to reinitiate the chase from a point where the forager flying back to the nest was lost, and it is easy to pinpoint the nest location as marked wasps often lose their flag at the nest entrance. These methods for estimating colony productivity and collecting nests can be valuable for researchers studying social wasps.
Every species is thought to develop an optimal strategy for survival and reproduction among a vast array of possible strategies. In natural selection, individuals with traits that maximize an individual's reproductive success will leave more offspring (and genes) to the next generation. Therefore, the number of offspring produced by an individual can be used as an indicator of the individual's relative evolutionary fitness. In a given ecological context, the comparison of the number of offspring produced relative to alternative behavioral strategies can help researchers predict the best strategy for optimizing fitness1.
Social Hymenoptera (such as wasps, bees, and ants) have a system of three different castes, which are workers (sterile females), queens (gynes), and males1. Only new queens (gynes) and males count toward fitness in social Hymenoptera. Worker production does not directly contribute to fitness since the worker is infertile. On the other hand, a queen that can produce a higher colony productivity (such as a higher number of total cells or a heavier nest) is considered to have a higher fitness in social Hymenoptera, regardless of the number of actually produced new queens and males (see, e.g.,Tibbetts and Reeve2 and Mattila and Seeley3). In general, it is difficult to precisely count the number of offspring produced by a colony of social Hymenoptera. In fact, the queens of many social insects live for more than 1 year (e.g., leaf-cutter ant queens can live >20 years4 and honeybee queens may live for 8 years5). In addition, one queen may produce thousands of reproductive offspring over the course of several weeks or months, even in annual species of genera Vespa and Vespula6,7,8. Furthermore, the lifespan of workers is shorter than that of their mother queen, and workers often die away from their nests. Hence, even if one could accurately count all adults in a nest at any given point in time, such a count would not accurately depict the number of offspring produced. Therefore, the number of offspring produced has been roughly estimated from the size of the nest, the number of workers in the nest, or the weight of the nest at a given point in time3,9,10. The number of larval cells could result in an overestimation of the offspring production when some cells are empty. The same method could also result in a potential underestimation of the offspring production because combs of small cells that contain worker brood can produce two or three cohorts of larvae6,7,11.
The first aim of this work is to provide an improved method for estimating vespine wasp colony productivity in terms of the number of adults produced. Yamane and Yamane suggested that the best way to estimate the number of offspring produced by a colony is to count the meconia in the nest12. The meconia is the fecal pellet comprising larval cuticle, gut, and gut contents that a larva leaves in its cell when pupating (Figure 1A). The total number of meconia produced per comb is calculated by multiplying the total number of cells present by the average number of meconia per cell. There are often multiple layers of meconia in a cell, and each meconia indicates that an individual successfully pupated in that cell6,11 (Figure 1B). When estimating the mean number of meconia per cell, if the number of cells examined is small (a small sample size), the standard error (SE) increases, and as a result, the error for the total number of meconia per comb becomes higher than if the sample size was larger. The SE of the mean (SEM) is a measure of the dispersion of sample means around the population mean. Therefore, in this study, I focus on the SEM of the number of meconia per cell to estimate the population (the number of adults produced) from the sample mean (the average number of meconia per cell). This study attempts to determine how many samples are required to obtain an SE rate of less than 0.05 per cell. To do this, a numerical simulation is performed with real data on the number of meconia per comb, to determine the minimum sample size (for both worker and queen combs) needed to estimate this value accurately within the defined SE of 0.05.
Vespine wasp colonies live in concealed nests (underground or aerial) composed of multiple horizontal combs, built in series from top to bottom6,7,11. The average size of the cells increases from the first (top) to the last (bottom) comb. In the bottom combs, a sudden shift in the average cell size can be seen. These wider cells are built for the development of new queens. Hence, a more accurate estimate of colony productivity (i.e., the number of individuals produced) can be obtained when the total number of meconia in the worker cells (small cells) and queen cells (large cells) are considered. In order to estimate fitness at the colony level, researchers could estimate the number of queens produced and focus on the meconia in the queen cells alone. As for reproductive males, these are reared either in worker or queen cells, depending on the species. Thus, it may be difficult to estimate the male production of a colony, except in species where males have a third, unique cell size13 (e.g., Dolichovespula arenaria).
The second aim of this work is to present a useful technique for locating wild vespine wasp colonies in the field and transplanting them into laboratory nest boxes. Although some researchers obtain wasp nests from pest control calls (i.e., people reporting them as pests14,15), this method is not always possible or desirable. Researchers might need to collect nests in wild and inhabited areas where pest controllers do not operate, or to conduct their research by more flexibly obtaining nests at specific times. Interestingly, people living in the mountainous areas of central Japan traditionally collect and rear wasps (Vespula shidai, Vespula flaviceps, and Vespula vulgaris) for food. Therefore, collecting and artificial rearing techniques for these wasps are well developed in those areas17.
This paper also summarizes the methods employed to rear Vespula wasps. The experimental organism for this study was V. shidai, a social, ground-nesting wasp inhabiting western Asia and Japan. V. shidai possesses the largest colony size among all Japanese vespine wasps, with a total of 8,000 to 12,000 cells per nest, with a maximum of 33,400 cells14,18. Workers of V. shidai have an average wet weight of 67.62 ± 9.56 mg. Males are usually reared in worker cells; in contrast, new queens are reared in specially constructed, wider queen cells14.
Figure 1: Meconium in a larval cell. (A) Cross section of a comb of Vespula shidai. Meconia is indicated by red arrows. (B) Two meconia are layered. Each blue arrow indicates one meconium. Please click here to view a larger version of this figure.
1. Evaluation of Colony Productivity
2. Finding Vespula Nests
Figure 2: Providing wasps with a flagged meat bait. (A) Baiting wasps with meat attached to the tip of a stick. (B) The piece of meat is tied with a thread to a plastic flag. (C) The wasp holds on the meat which is tied to the flag. Such “flagged” baits will increase the visibility of the flying forager. The photos in panels B and C were taken by Fumihiro Sato. Please click here to view a larger version of this figure.
3. Transfer of the Nest
Figure 3: Carrying box. (A) Box for carrying nests collected in the field. (B) A bamboo grid is placed at on the bottom of the box. The two boxes in the image on the right are upside down. Please click here to view a larger version of this figure.
4. Rearing Vespula
Figure 4: Laboratory setup. (A) Setting a carrying box into a nest box used for a long-term study. Before placing the carrying box in the nest box, the wood board at the bottom of the carrying box was removed, leaving only the newspaper to cover the bottom of the nest. (B) A series of nest boxes with food resources hanging from a wire line. Please click here to view a larger version of this figure.
One goal of this study was to determine how many samples are required to obtain an SEM of the number of meconia per cell which is less than 0.05. In this study, a comb with an average cell size of <20 mm2 was defined as a worker comb, whereas larger combs were defined as queen combs. I counted the number of cells for queen combs and worker combs (in this study, counts were made of six queen combs and six worker combs from five V. shidai colonies). The actual number of cells per comb was estimated from these data via extrapolation (Table 1).
ID | State | Collection date | Area (mm2) | Estimated number of cells (ENC) | Actual number of cells (ANC) | Actual number of meconium (ANM) | Mean number of meconium in a cell | ANM /ENC |
WW-Kb01 | Alive | 18-Oct-16 | 27756.7 | 1599.9 | 1433 | 2430 | 1.70 | 1.52 |
WW-Kb02 | Alive | 18-Oct-16 | 4098 | 381.9 | 347 | 494 | 1.42 | 1.29 |
WW-Kb02 | Alive | 18-Oct-16 | 22439.3 | 1118.9 | 986 | 1317 | 1.34 | 1.18 |
WR-Ksb | Collapse | 3-Nov-16 | 19094.9 | 1098.6 | 1,181 | 974 | 0.82 | 0.89 |
WR-Ksc | Collapse | 27-Nov-16 | 38,933.40 | 2,198.70 | 2,455 | 4,321 | 1.76 | 1.96 |
WR-Kb05 | Collapse | 29-Nov-16 | 10970 | 860 | 763 | 1315 | 1.72 | 1.53 |
QW-Kb01 | Alive | 18-Oct-16 | 29186.2 | 1094.4 | 1095 | 759 | 0.69 | 0.69 |
QW-Kb01 | Alive | 18-Oct-16 | 36920.5 | 1361.6 | 1341 | 1075 | 0.80 | 0.79 |
QW-Kb02 | Alive | 18-Oct-16 | 37295.9 | 1047.2 | 1080 | 1068 | 0.99 | 1.02 |
QR-Ksb | Collapse | 3-Nov-16 | 24811.2 | 1011.9 | 893 | 701 | 0.78 | 0.69 |
QR-Ksc | Collapse | 27-Nov-16 | 33352.8 | 1384.5 | 1241 | 1069 | 0.86 | 0.77 |
QR-Kb05 | Collapse | 29-Nov-16 | 25157.6 | 1071.4 | 922 | 572 | 0.62 | 1.97 |
WW = a worker comb from a wild nest, WR = a worker comb from a rearing nest, QW = a queen comb from a wild nest, QR = a queen comb from a rearing nest. Alive = viable wasp larvae in cells, Collapse = no viable larvae in cells. |
Table 1: The actual and estimated numbers of cells in six worker combs and six queen combs and the number of meconia per comb. WW = a worker comb from a wild nest, WR = a worker comb from a rearing nest, QW = a queen comb from a wild nest, QR = a queen comb from a rearing nest. Alive = viable wasp larvae in cells, Collapse = no viable larvae in cells.
An analysis of the relationship between the sample size and the SEM of the number of meconia per cell demonstrated that the sample size should be established using a bootstrap approach based on the number of meconia counted (from real data). Using real data, the mean and standard deviation (SD) of the number of meconia per cell were calculated, with the number of samples set at 1,000 for each sample size (the number of cells to be examined were 1 to 500; Figure 5). I did not allow for an iterative extraction from the data at sampling. The SEM for the number of meconia per cell was calculated for each sample size for each set of real data. Then, the sample size at which the SEM was less than 0.05 was examined. All calculations were made using software R.3.2.4.19 This analysis showed that the SEM was <0.05 when the sample size was 100 cells (for both worker and queen combs) (Figure 5). Therefore, the following results are based on examining the number of meconia per 100 cells per comb.
The actual and estimated numbers of cells in six worker combs and six queen combs and the number of meconia per comb are shown in Table 1. The estimates of the number of cells in the worker combs, based on comb area measurements, were both higher and lower than the true count. The mean number of meconia in the cells of worker combs, which represents the number of workers produced, ranged from 1.96 times more than the number of estimated larval cells to 0.89 times less than the estimated number of cells (Table 1). In the queen combs, the actual number of cells was often less than the estimated number of cells. The number of meconia in the queen combs, which can represent a component of fitness (i.e., a part of the reproductive success of the founding queen), was 0.53 to 1.02 times the estimated number of cells.
All cells and meconia were counted in six randomly selected worker combs and six randomly selected queen combs from the five nests (Table 1). The total number of cells counted in the worker combs was 7,165, whereas the number of meconia counted in the worker combs was 10,851. The average number of cells per comb was 1,194.2 ± 720.3 (average ± SD), whereas the average number of meconia in the worker combs was 1,808.5 ± 1,368.2. In the queen combs, the total number of all cells was 6,572, whereas the number of all meconia was 5,244. The average number of cells per comb in the queen combs was 1,095.3 ± 174.820, whereas the average number of meconia was 874.0 ± 223.8. Meconium layers in worker cells ranged from zero to three, whereas the queen cells had either one or no meconium layer.
Figure 5: The relationship between the sample size and the standard error (SE) relative to the number of meconia counts. (a) Meconia per cell in worker combs. (b) Meconia per cell in queen combs. Each circle depicts an SE relative to the number of meconia per cell obtained via simulation with actual data. Color differences represent the data from each sampled nest. Simulating the SE for the number of meconia per cell in comb WWkb02 (worker comb) was accomplished with a sample size of 300 because that comb only had 347 cells. Please click here to view a larger version of this figure.
The colony productivity of bees, ants, and wasps has been estimated previously by the number of workers and cells in nests or by the weight of the nests3,9,10. This study shows that the estimate of the number of meconia provides a better estimate of the overall number of individuals produced (i.e., a better indicator of colony productivity). In fact, it was found that, for both worker and queen combs, the number of meconia ranged from 0.53 to 1.96 times the number of larval cells in the comb. These findings quantify how inaccurate the determination of the number of workers and queens produced can be when it is based on the number of cells in a comb. Despite being more labor-intensive, estimating the number of meconia in a nest seems to guarantee a more precise evaluation of colony productivity. On the other hand, in this study, it was not evaluated how accurately the number of meconia represents the number of individuals produced.
This paper shows how many cells of a V. shidai nest should be examined to estimate colony productivity, based on the results of a bootstrap simulation approach using sample data on the number of meconia in the nest. Based on these results, it would be appropriate to investigate 100 cells per comb of both worker and queen cells. The method for counting meconia can also be applied to a nest after it has collapsed (i.e., is inactive), which can be advantageous for researchers: the reproductive period of vespine wasp colonies is quite long8 and studying a nest after it has collapsed means that the total number of adults produced over the entire reproductive period can be estimated. Such colonies are also easier to collect.
To collect nests of V. shidai, some researchers have followed either marked (e.g., coated with fluorescent powder) or unmarked wasps21. The nest location method presented here (feeding the wasps "flagged" meat) facilitates following wasps to their nests. This approach is also helpful if a tracked wasp is lost because the same wasp will eventually return to the bait along the transect. Provide new flagged bait to this wasp and carry it to the point where it was last lost, thus allowing chasers to resume the chase from that point onward (closer to the nest). Some of the flags brought to the nest are dislodged at the nest entrance, which also facilitates finding ground nests. However, this method is not suitable for rainy days because markers tend to stick to branches and leaves when they get wet. Although chasing flagged wasps is useful for V. shidai, V. flaviceps, and V. vulgaris in Japan, this method could not be applied to Vespula rufa because these wasps do not come to the bait and do not grab flagged bait. The nest location method can probably not be used for some Vespula wasps.
More sustainable diets are needed by an ever-increasing global population. In addition, the demand for edible insects increases daily. Many edible insects, which are being consumed locally and traditionally throughout the world, have been identified by the Food and Agriculture Organization of the United Nations21 as a promising alternative protein source for overcoming food insecurity worldwide. Larvae and pupae of Vespula have traditionally been used as food in mountainous areas of Japan16, and so, they could be used to provide a source of protein elsewhere in the world. The set of protocols developed in this study is likely applicable to locating nests of other yellowjacket species. Therefore, the protocols outlined in this paper will be useful for collecting yellowjackets as an edible resource and studying wasp behavior.
The authors have nothing to disclose.
The author would like to thank Katsuyuki Takahashi, Hiroo Kobayashi, Fumihiro Sato, Daikichi Ogiso, Toshihiro Hayakawa, and Hisaki Imai for teaching him the traditional wasp hunt method. The author would like to offer special thanks to Kevin J. Loope and Davide Santoro for carefully proofreading the manuscript. The author is grateful to Masato Abe, Yasukazu Okada, Yuichiro Kobayashi, Masakazu Shimada, and Koji Tsuchida for their discussion. The author wants to thank Yuya Shimizu and Haruna Fujioka for their technical assistance with evaluating colony productivity. The author would like to thank Tsukechi black bee club for supporting the video shooting. The author wishes to thank three anonymous reviewers for their comments on an early version of this paper. This study was supported in part by Takeda Science Foundation, Fujiwara Natural History Foundation, Funding of the Nagano Society for The Promotion of Science, Shimonaka Memories Foundation, Takara Harmonist Fund, and the Dream Project by Come on UP, Ltd.
cuttlefish | Any | fresh/ as a bait | |
dace | Any | fresh/ as a bait | |
chichken heart | Any | fresh/ as a bait | |
plastic bag (polyethylene) | Any | as a flag | |
bamboo skewer | Any | ||
industrial sewing thread | FUJIX Ltd. | King polyester, No.100 | |
paint marker pen | Mitsubishi pencil | UNI, POSCA, PC5M | |
fishing rod | ANY | ||
carrying box | made of wood | ||
nest box | made of wood |