Dispersion and temperature-associated orientation of Chinese mantis (Mantodea: Mantidae) egg cases.
INTRODUCTIONPraying mantids (Mantodea: Mantidae) are tritrophic predatory insects living in oldfield habitats, where they consume herbivorous and carnivorous arthropods (Hurd el al., 2015), as well as some pollen from flowers (Beckman and Hurd, 2003). Temperate-zone mantids, such as the Chinese mantis Tenodera aridifolia sinensis (Saussure), are univoltine, hatching in spring, maturing by late summer, and ovipositing in autumn before being killed by hard frosts. Although some temperate mantids require a cold period to complete their development, the Chinese mantis does not. This was first seen when Butler (1966) obtained six egg cases (and later nymphs) at intervals of 11-19 d from a well-fed female in the lab. Hurd el al. (2004) confirmed this observation by recording oviposition dates in the field, later collecting and refrigerating egg cases until 1 April and then reporting that earlylaid oothecae hatched significantly earlier than those laid later, indicating a development period that begins at fertilization and oviposition. Winter halts development, suggesting that any part of development can occur before or after winter dormancy. If the 41-45 d development period (Butler, 1966) is completed before the killing frost, nymphs will hatch but soon be killed by the cold, resulting in complete loss.
During our studies in an old field in the process of becoming a pine forest, we observed a number of features that seemed unusual for an oldfield insect, including the placement of most mantid egg cases in trees, often high in trees, and few on herbaceous vegetation as would be expected. We expected egg cases to be randomly placed in relation to compass orientation, but when that was not true, additional collections were made to explore this unusual pattern further. We also performed a hatching study in order to better understand the possible forces of selection that may account for his egg-laying pattern in Chinese mantids.
The objectives of our study of the Chinese mantis were to determine the: (1) types and heights of vegetation used for oviposition, (2) compass orientation of egg cases on host plants, (3) dispersion of egg cases, (4) relationship between compass orientation and days to hatching of young, and (5) relationships among mass of egg case, number of young, and orientation/height.
MATERIALS AND METHODS
The study site was an old field in Chesapeake, 36[degrees]37'47.5"N, 76[degrees]8'32.3"W, Virginia, removed from agricultural practice since 2000, shortly after the 11 ha property was acquired by The Nature Conservancy (TNC). The shallow drainage ditch that bisected our study site had a 1 m wide band of 4 to 6 m sweet gum (Liquidambar slyraciflua) trees in and next to the ditch, just beyond the plow zone. Soil bulldozed to plug ditches created low places and other low areas were dug as vernal ponds, resulting in patches in which wetland plants such as wool grass (Scirpus cyperinus), cattails (Typha latifolia), soft rushes (Juncus effuses and J. tenuis), and asters (Aster spp.) dominated. TNC planted swamp chestnut oaks (Querais michauxii) but other trees were volunteers, mostly loblolly pines (Pinus taeda) and sweet gums, derived from the soil seed bank or carried there by wind or other means. This early successional habitat, dominated for 3 y by little bluestem (Schizachyrium scoparius), provided conditions for the colonization and expansion of a population of Tenodera aridifolia sinensis. The site is rapidly becoming a forest of volunteer pine and sweet gum trees, with some other woody plants in the mix. As the years pass, the herbaceous plants are squeezed out and the woody plants become more and more dominant.
In February 2005 we established a 9 m by 9 m study grid with 12.5 intervals and an area of 1.266 ha, composed of 81 "cells." In order to collect information on dispersion of egg cases on the grid, we divided each cell using strings measured from the mid-point of each side into four sub-cells, each 6.25 m per side.
In February and March 2005, we searched each sub-cell for mantid oothecae, recording the species and height of vegetation on which an egg case had been deposited and its approximate compass orientation. We also estimated five categories of microhabitat openness (100, 75, 50, 25, and 0%) in each sub-cell. One-hundred percent open means no trees and 0% open is closed forest. Compass orientation was recorded. The location and orientation of the grid near an east-west road, with the baseline of the grid parallel to the road, made an approximation of NSWE compass directions feasible.
In all 425 egg cases of Chinese mantids were recorded, comprising the 2005 data set. Preliminary analysis revealed a preponderance of egg cases oriented in a southerly direction. However, more precision using compasses was required to confirm this finding.
On 28 January 2006, we collected 167 egg cases from the grid. For each egg case, we recorded the other information as before: species and height of vegetation on which each ootheca was deposited plus precise ([+ or -]5[degrees]) orientation with a surveyor's compass. Each egg case was weighed in the lab, placed in a quart mason jar with gauze-covered lid, kept at 20 C, and examined each day to test the hypothesis that south-facing egg cases hatch earlier, because of solar heating, than those oriented in other compass directions. In effect, each egg case was merely completing its development in a standard 20 C environment, where any differences could be attributable to its earlier long or shorter development time while in nature.
Jars with newly emerged nymphs were set aside for a day to ensure all young had hatched and then placed in a freezer overnight to kill the young. Next, each jar was placed in a 40 C drying oven for 48 h. Afterward, all hatchlings from each ootheca were counted and weighed to the nearest mg using an electronic balance. This information, together with hatching date, was merged with the information on height, host plant, compass orientation, and provided a test of the null hypothesis that hatching time is independent of compass orientation.
On 28 March 2007, we recorded height and vegetation type of oviposition and also orientation using compasses for 149 mantid egg cases. All searches for oothecae were conducted in late winter, when the probability of detection was greatest. In 2005 each egg case was assigned to one of the eight principle compass directions based on an approximation of the direction in which it was oriented, whereas in 2006 and 2007 each compass reading recorded in the field was later assigned to the closest cardinal direction for analysis. Analysis of variance (ANOVA), t-tests, correlation analysis, and [chi square] tests were performed using Number Cruncher Statistical software (NCSS v. 2002) and EXCEL.
Later (2013, within 5 km of grid), we examined temperature variation on 5 y old pine trees in late autumn (the period when second or later egg cases are laid) in order to determine whether temperatures differed between south- and north-facing locations on the trees. We placed six ibuttons (temperature data-loggers: Thermochrons by Embedded Data Systems, model DS1921D-F5#) on each of 10 pines; three were south-facing near the surface (2-5 cm above ground), at 1 and 2 m, and three north-facing at similar elevations. Temperature ([+ or -] 0.25 C) data were recorded at 30 min intervals between 20 September and 10 November and then analyzed with EXCEL.
RESULTS
TYPES AND HEIGHTS OF VEGETATION FOR OVIPOSITION (2005-2007)
In 2005 we recorded the locations and heights of 425 egg cases of Chinese mantis. Egg cases, located on 13 species of plants including eight shrubs and trees (Table 1), were placed mostly on woody plants (91.3%), especially on sweet gum (34.6%) and loblolly pine (29.9%), the most common and tallest trees. Of the 37 egg cases deposited on herbaceous plants, 19 were on goldenrods. None was found on either of the tallest herbaceous plants, giant plume grass (Erianlhus ravennae), or wool grass; the latter was common in the wettest places on the grid.
In 2006 and 2007, the searches for egg cases were not as intensive as in 2005, probably resulting in lower proportions being observed in low-lying herbaceous vegetation (Table 2). Nevertheless, the patterns were similar to 2005, with the majority of egg cases found on the tallest and most common trees: sweet gums and loblolly pines.
The mean heights of oviposition (Tables 1 and 2) in the tallest trees differed among years (ANOVA: F = 7.70, P = 0.0005), increasing from 1.45, 1.65, and 1.73 m. The distribution of heights at which egg cases were deposited (Fig. 1) indicated that although most were laid between 0.9-2.3 m, in all years the central tendency shifted toward greater heights. The proportions oviposited in pines and sweet gums were comparable among years (64.5, 71.2, and 61.1%, respectively).
[FIGURE 1 OMITTED]
[FIGURE 2 OMITTED]
COMPASS ORIENTATION OF EGG CASES ON THE HOST PLANT
Oothecae were nonrandomly deposited (in a southerly direction) in all 3 y (Fig. 2). The [chi square] values for the tests of random orientation of egg cases were 385.1, 124.1, and 147.9, respectively, with a critical value of [24.32.sub.P = 0.05] for 7 df, P < 0.001; therefore, the deviations were significant in all years. Although the expectation is for 37.5% of egg cases to be oriented in the combined SE, South, and SW directions, these proportions were 53.92, 63.02, and 66.44% for the 3 y, respectively.
In previous oldfield studies of Chinese mantids, herbaceous plants such as goldenrod, <1 m tall, usually were selected as oviposition sites and no mention of non-random orientation of placement was reported. To test the importance of elevation in the placement of egg cases, we put the egg case data for all years combined and sorted into the eight cardinal directions into two groups, those deposited [less than or equal to] 1-m and those placed >1 m. The results of ANOVAs indicated random compass orientations in the [less than or equal to] 1-m group (F = 1.2904, P = 0.255). By contrast significant differences (F = 2.1586, P = 0.0366) were observed for egg cases placed >1 m in vegetation.
DISPERSION OF EGC CASES ON THE 1.266 HA GRID (200S)
On the grid, 17 (21.0%) of the 81 cells and 156 (48.1%) of 324 sub-cells had zero egg cases. The spatial distribution within the grid, using sub-cells as units, was examined with Moran's-1; the Z-score (1.70) was 0.10 > P > 0.05, indicating a tendency toward but no significant clustering of egg cases at the scale of sub-cell.
A further way to evaluate factors affecting dispersion of oothecae on the grid is to examine the possible relationship between degree of openness, as recorded in 2005, and placement of egg cases. To test whether Chinese mantises deposit egg cases randomly in heterogeneous habitat, we used a goodness-of-fit test (Zar, 1999) to compare observed and expected numbers of egg cases in sub-cells in the five categories of openness. The significant results (Table 3: [chi square] = 136.59, 4 df, P < 0.001) were unexpected, in part because the smallest mean number, 0.47 egg case per sub-cell, was from 100% open sub-cells (no trees). The mean in 0% open patches, i.e., closed forest, was 1.36, nearly three times greater than 0.47. For sub-cells judged to be 50% and 25% open, many more egg cases were observed than expected (Table 3).
COMPASS ORIENTATION, DAYS TO HATCHING, AND NUMBERS OF HATCHLINGS (2006)
In all 21,521 hatchlings emerged from the 167 egg cases from days 2 to 89 in the lab, the mean being 52.98 d (Table 4). There was no difference in mean days to hatching in relation to compass orientation (ANOVA: F = 1.00, P = 0.436). Therefore the hypothesis south-facing oothecae hatch sooner was not supported. Mean days to hatching ranged from 44.38 [+ or -] se 4.9 d to 56.08 [+ or -] se 4.0 d for West- and East-oriented oothecae, respectively. The means of the next longest days to hatching after East were SE, SW, and South, in that order (Table 4). Therefore, despite no significance, south-facing egg cases tended to produce hatchlings later, not earlier, than expected, suggesting they might have been laid later (if development duration is more or less fixed).
Hatchling number ranged from 2-228. No relationship was observed between number of hatchlings and compass orientation (ANOVA: F = 1.31, 7 df, P = 0.249). The grand mean number of hatchlings was 128.87, with the smallest mean (Table 4) from SW-oriented egg cases (116.08 [+ or -] sf. 11.0) and the largest mean from N-oriented oothecae (157.45 [+ or -] sf. 16.3). However, because south-facing (SE, S, SW) oothecae were so numerous, comprising 61.77% of egg cases compared to an expected percentage of 37.5, the number of young produced (12,770, or 59.34%; Table 4) also was greater than expectation ([chi square] = 207.63, 7 df, P < 0.001).
We also examined the live biomass of young by subtracting the mass of empty egg case from its initial mass (i.e., when collected; Table 4). ANOVA results were significant (F = 3.47, 7 df, P = 0.0018), with SW and South having the lowest average live biomasses (645.29 [+ or -] SE 51.0 and 713.16 [+ or -] SE 35.7 mg, respectively), followed by East (781.00 [+ or -] se 72.1 mg) and SE (southeast: 836.32 [+ or -] SE 49.9 mg). These results closely conform to those of mean days to hatching and mean number hatched (above). As a group, then, egg cases oriented East, SE, South, and SW had the lowest live biomasses and smallest mean numbers of hatchlings (with one exception) but took longer to hatch compared to those in the other half of the compass rose.
Again using the 2006 hatching data, we examined the relationships among mass of egg case, orientation, and both number and dry mass of hatchlings, dry mass being a surrogate for body size of hatchlings. Pearson's correlation coefficient (r = 0.686) between initial mass of the egg case and number of hatchlings indicated a modest level of association. Next, we subtracted the mass of each empty egg case from initial egg case mass to get a value for live biomass of young (No. 1, Table 5). Live biomass differed among compass directions (ANOVA: F = 3.200, P = 0.0026), with those from E, SE, S, and SW (hereafter termed 'south,' as shown in underlined type in Table 4) all lighter than those from the W, NW, N, and NE (='north'). The initial masses of egg cases (No. 2, Table 5) also differed among compass directions (ANOVA: F = 2.071, P = 0.0037), and when combined into two compass groups (No. 6, Table 5) egg cases from north were significantly heavier (1313.61 [+ or -] SE 49.0 mg vs. 1128.68 [+ or -] SE 29.8 mg) than those oriented south (t = -3.699, P = 0.0001). The proportion of egg cases <1000 mg was 14.3% for north and 34.5% for south. Therefore, many small egg cases were oriented south.
Next we examined with ANOVA the possible relationships of the eight compass orientations on dry mass of all young, number of young, and mass of individual young; all were not significant (Nos. 3-5, Table 5).
However, when these values were grouped by north and south and evaluated with one-tailed t-tests, dry mass per egg case of all hatchlings from north (158.71 [+ or -] SE 11.4 mg) was significantly greater than for hatchlings from south (136.88 [+ or -] SE 6.83 mg), and number of young (147.86 [+ or -] SE 7.49 vs. 127.53 [+ or -] SE 4.57) was significantly greater for north than for south (Nos. 7 and 8, Table 5). In contrast, the mass of individual young was nearly identical (north = 1.0561 mg per individual vs. south = 1.0556 mg; No. 9, Table 5). In sum egg cases from north were heavier initially, their live biomass was greater, and their number of young and dry mass of young greater, all significantly greater than those from south. However, the dry mass of individual young was nearly identical (1.056 mg), regardless of source.
Each of the 60 data-loggers yielded more than 2100 temperature values across the nearly 7 wk period. The temperatures at the three elevations on north and south sides of pines were significantly (P < 0.001) different, and at the 2 m elevation the temperature differences were even greater (one-tailed t-test: P = 1.87 E -17). South-facing locations at 2 m elevation averaged ca. 0.45 C warmer than similar north-facing locations, despite 2 m ibuttons usually cooling down faster and for longer periods than those closer to the ground surface. By the end of the study, most data-loggers had recorded one or more 0 C value, but the few below 0 C values were for 1 or 2 h periods and mostly at -0.5 to -1.5 C temperatures.
DISCUSSION
Chinese mantises oviposit on goldenrods or similar herbaceous vegetation (Hurd, 1999). By contrast, in our study >90% of egg cases were deposited in shrubs and trees, mostly in tall pines and sweet gums (Tables 1 and 2). Of the 475 egg cases found on pines and sweet gums, <5 were attached to the trunk; almost all were deposited near the ends of branches.
The placement of an egg case in tall vegetation is potentially adaptive, reducing chances of its being flooded or buried in snow and in part because when nymphs hatch in low vegetation they disperse from the same point source. By contrast the higher the egg case is placed in a tree the better chance emerging hatchlings will become more widely dispersed after walking along branches before dropping to the ground by the thread from the tip of the abdomen. On windy days the >100 hatchlings might be dispersed over an area of several [m.sup.2], giving them better chances of finding the food needed to get them to the next instar. With highest levels of mortality during the first-instar nymphal stage (Hurd and Eisenberg, 1984), a broad dispersion of hatchlings is adaptive.
It is noteworthy that no egg cases were found either on wool grass and giant plume grass, the tallest herbaceous plants on the grid. Both plants, most often found in wet patches, grow to 2 m by the time imagos were present in the population. On 8 and 9 September 2006, when we spent 13 h searching for and photographing 108 undisturbed adults on the grid, we frequently observed mantids hunting or perched on the floral heads of wool grass, but we never found egg cases on this common plant.
In the only published study on dispersion of oothecae of the Chinese mantis in North America, Eisenberg and Hurd (1990) found them to be contagiously distributed in Delaware; 49-71% was within the same 2 [m.sup.2] area as another egg case and 29-61% was within 1 [m.sup.2]. Although we did not record the proximity of any ootheca to any other when doing the searches in 2005, we did not see such obvious clustering of egg cases as Eisenberg and Hurd (1990) report; only in blackberry thickets were egg cases occasionally clumped in such proximity. Further, our analysis of possible contagious dispersion, while not significant using the Moran test, evaluated the dispersion of egg cases in 6.25 m square sub-cells. Consequently, the pattern of clustering was on a much coarser scale than that used by Eisenberg and Hurd (1990). The sub-cells with the highest numbers of egg cases (10-13) were those with tall gum trees in or close to the ditch.
To evaluate how numbers of egg cases compared to those reported in other studies, we calculated a density estimate of 0.0336 per [m.sup.2] for 2005, the only year we searched the entire grid. Densities three times higher (0.10 and 0.12 per [m.sup.2]) are reported by Eisenberg and Hurd (1990) for 0.12 ha and 0.07 ha areas in large old fields. For the other years, we have only number of egg cases found per hour of searching, 6.68 per h in 2006 and 16.60 per h in 2007, the latter by more experienced searchers.
Mantids avoided 100% open areas when ovipositing (Table 3), with lowest mean numbers there. Density of egg cases was higher in closed forest and the greatest mean density was seen in 25% open areas. These results were unexpected, because as putative oldfield animals, Chinese mantises and their egg cases are expected to be most dense in 100% open areas, and perhaps nearly absent in 0% open areas (closed forest). The sub-cells with many more egg cases than expected were those with half or more of the area with trees (Table 3). These results may suggest the Chinese mantis is an edge species, foraging in open areas where their arthropod prey is abundant in low-lying herbaceous vegetation but when possible seeking woody vegetation when not feeding, perhaps as a means to reduce predation.
The egg cases collected in late January 2006 yielded a mean of 128.87 hatchlings per egg case, which is consistent with the hatching potential of the sizes of egg cases in our population (see Table 2; Eisenberg et al, 1981). Eisenberg and Hurd (1977) found a significant positive correlation (r = +0.74, P < 0.01) between mass of egg case and number of hatchlings and so did we. The 167 egg cases ranged in size from 320 mg to 2135 mg ([bar.x] = 1191.4 mg); only four were >2000 mg. By contrast Hurd et al (1978) report a mean mass of 1790 mg for 224 egg cases in northern Delaware. No statistical test is necessary to support the assertion that egg cases are smaller in eastern Virginia.
The significantly nonrandom (southerly) orientation of egg cases, previously unreported but observed in all 3 y, is an unexpected result and prompted our hatching study and the later ibutton/temperature study. Although the finding of significant differences in live biomass (Table 4) is consistent with the nonrandom compass orientation of the oothecae (Fig. 2), no differences were seen in either days to hatching or number of hatchlings in the lab study (Table 4). Although the t-test results in Table 5 (nos. 6, 7, 8) are suggestive, our hatching study failed to detect reasons for the preponderance of egg cases being deposited in the southern half of the compass rose.
There are several (not mutually exclusive) explanations for the observed nonrandom compass orientation of egg cases in the 3 y of our study in the habitat rapidly converting to a pine forest. One is that the rate of nymphal development may respond somehow to a southerly orientation of egg cases. We tested this possibility with our hatching study and obtained no significant results. An alternative would be to record hatching dates in the field. Later, using the regression equation method on empty oothecae, the number of hatchlings could be calculated and these values could be related to compass orientation. This protocol would require daily inspection of dozens of egg cases in the field over a period of 4-6 wk in spring. The resulting information would be most valuable if the dates of oviposition were known. A third explanation is that females may lay egg cases in random compass directions early in the oviposition season but shift to nonrandom (mostly southerly) later in autumn, when the daytime temperatures are cooler and the sun is lower in the sky. This oviposition pattern would result in a preponderance of south-oriented oothecae, especially in geographic locations with a long autumn, where multiple egg cases could be laid by well-fed females. Collection of egg cases at frequent intervals and from a measured area or from numbered trees during the oviposition season would provide support for this possibility.
Another explanation, and the one we favor, is that females orient preferentially to the south and at the tips of branches towards the end of the growing season. By doing so they elevate their body temperatures and in turn their metabolic rates, enabling them to convert food into eggs more efficiently, and enhancing their chances of laying second or later clutches of eggs when food availability and the weather permit. The observed placement of almost all oothecae near the tips of tree branches supports this possibility. Behavioral studies of marked imagos would provide further support. Our temperature data-logger results support the higher temperatures on the south- than on north-facing sides of pine trees, and also showed that the killing hard frosts may not occur until well into November in our region of eastern Virginia. We have observed adult females as late as 31 October.
In conclusion our population of Chinese mantises in eastern Virginia exhibited several unusual features. Egg cases were randomly dispersed in structurally diverse habitat, with open (treeless) patches having the fewest egg cases, not the most, as would be expected of an oldfield insect. Egg cases, 91% of which were deposited on woody plants, mostly from 0.9-2.3 m, were laid in nonrandom (predominately southerly) orientations. Our hatching study produced equivocal results and failed to explain the nonrandom orientation of egg cases. However, temperature data-loggers showed that south-facing locations on pines were significantly warmer than similar elevations on the north sides of trees, supporting that autumnal females may select the south sides of pines, thereby raising their body temperatures and metabolic rates, and sometimes their ability to deposit late and usually small clutches.
Acknowledgment.--We thank file Nature Conservancy for the use of their land as a study site, A. Rose for assistance with the hatching study, T. Georgel, W. Spaur, S. Selden, and especially C. Day for assistance in the field, colleagues C. Binckley and D. Waller for reading earlier drafts of the manuscript, and three reviewers for useful comments.
Literature Cited
Beckman, N. and L. E. Hurd. 2003. Pollen feeding and fitness in praying mantids: the vegetarian side of a tritrophic predator. Environ. Entomol., 32:881-885.
Butler, L. 1966. Oviposition in the Chinese mantid, Tenodera aridifolia sinensis (Saussure) (Orthoptera: Mantidae). J. Georgia Entomol. Soc., 1:5-7.
Eisenberg, R. M. and L. E. Hurd. 1977. An ecological study of the emergence characteristics for egg cases of the Chinese mantis (Tenodera aridifolia sinensis Saussure). Am. Midi. Nat., 97:478-482.
-- and --. 1990. Egg dispersion in two species of praying mantids (Mantodea: Mantidae). Proc. Entomol. Soc. Washington, 92:808-810.
--, --, and J. A. Bartley. 1981. Ecological consequences of food limitation for adult mantids (Tenodera aridifolia sinensis Saussure). Am. Midi. Nat., 106:209-218.
Hurd, L. E. 1999. Ecology of praying mantids. In: F. R. Prete, H. Wells, P. H. Wells, and L. E. Hurd (eds.). The praying mantids. Johns Hopkins University Press, Baltimore, Maryland.
--, P. A. P. deHart, J. M. Taylor, M. C. Campbell, and M. M. Shearer. 2015. The ontogenetically variable trophic niche of a praying mantis by stable isotope analysis. Environ. Entomol., 1-7 (2015):DOI: 10/1093/ee/nov004.
--, and R. M. Eisenberg. 1984. Experimental density manipulations of the predator Tenodera sinensis (Orthoptera: Mantidae) in an old-field community. I. Mortality, development and dispersal of juvenile mantids./. Anim. Ecol., 53:269-281.
--, --, and J. O. Washburn. 1978. Effects of experimentally manipulated density on field populations of the Chinese mantis (Tenodera aridifolia sinensis Saussure). Am. Midi. Nat., 99:58-64.
--, R. E. Malus, K. C. Bui.ka, and A. M. Jones. 2004. Life history, environment, and deme extinction in the Chinese mantid Tenodera aridifolia sinensis (Mantodea: Mantidae). Environ. Entomol., 32:182-187.
NCSS 2002. 1996. Jerry Hintze. Kaysville, Utah.
Zar, J. 1999. Biostatistical analysis, 3rd ed. Prentice-Hall Inc., Upper Saddle River, New Jersey.
Submitted 5 June 2015
Accepted 12 January 2016
ROBERT K. ROSE (1) and A. SCOTT BELLOWS
Department of Biological Sciences, Old Dominion University, Norfolk, Virginia 23529; Virginia Space Grant Consortium, 600 Butler Farm Road, S-2200, Hampton 23666
(1) Corresponding author: 757-489-1702; Fax: 757-683-5283; e-mail: [email protected]
TABLE 1.--Oviposition sites for 425 egg cases deposited on woody or
herbaceous vegetation in autumn 2004 and evaluated in February-March
2005 during searches on a 1.266 ha grid, based on ca. 350 h of
searching. No oothecae were found on Japanese honeysuckle or Virginia
creeper, both common vines, and none on these common herbs: hemp
dog-bane, panic or foxtail grasses, wool grass, or giant plume grass
Woody N % Total Mean ht (m)
Sweet gum 147 34.6 1.6
Lobl. pine 127 29.9 1.2
Red maple 51 12.0 0.8
Oaks 33 7.8 1.5
Blackberry 20 4.7 0.8
Amer, cane 8 1.9 0.5
Red elm 1 0.2 1.3 *
Groundsel 1 0.2 1.2 *
Totals 388 91.3
Herbaceous N % Total Mean ht (m)
Goldenrods 19 4.5 0.5
Soft rushes 8 1.9 0.3
Bluestem 6 1.4 0.4
Asters 2 0.5 0.3
Dog fennel 2 0.5 0.7
Totals 37 8.7
* Not a mean (n = 1)
TABLE 2.--Oviposition sites for egg cases deposited on woody and
herbaceous vegetation as determined by searches in January 2006 (25 h
searching by the authors and three others) and March 2007 (9 h
searching by the authors)
2006 N % Total Mean ht (m)
Sweet gum 56 33.5 1.7
Lobl. pine 63 37.7 1.6
Red maple 6 3.6 0.9
Oaks 10 6.0 1.5
Blackberry 11 6.6 1.0
Amer, cane 6 3.6 1.4
Red elm 4 2.4 1.5
Groundsel 2 1.2 1.7
Willow 2 1.2 2.4
Bald cypress 1 0.6 --
Vines 2 1.2 0.9
All herbs 4 2.4 0.9
Totals 167
2007 N % Total Mean ht (m)
Sweet gum 74 49.7 1.7
Lobl. pine 17 11.4 1.8
Red maple 13 8.7 1.3
Oaks 17 11.4 1.6
Blackberry 19 12.8 1.0
Amer, cane 1 0.7 --
Wax myrtle 6 4.0 2.2
Groundsel 1 0.7 --
All herbs 1 0.7 --
Totals 149
TABLE 3.--Observed and expected numbers of egg cases in relation to
the degree of openness (2005), estimated as 100%, 75%, 50%, 25%, or
0%, the latter being enclosed forest, of the habitats in each of 324
sub-cells. Also given are the partial [chi square] values and the
mean numbers of egg cases per sub-cell. The grand mean is 1.31 egg
case per sub-cell (425/324)
Openness Partial Mean
(%) N Observed# Expected# [chi square] #/sub-cell
100 114 54 149.54 61.04 0.47
75 92 122 120.68 0.01 1.33
50 64 137 83.95 33.52 2.14
25 29 78 38.04 41.98 2.69
0 25 34 32.79 0.94 1.36
Totals 324 425 425 136.59 [bar.x] = 1.31
TABLE 4.--Results of hatching trials conducted in spring 2006 with
egg cases collected in late January and returned to the lab.
Underlined values are those for egg cases from compass directions
that took longest to hatch, yet had fewest hatchlings (with one
exception) and the lowest live biomass (defined as initial egg case
mass minus mass of empty egg case)
Mean Mean Mean live Total
Compass days to number biomass number
orientation N hatching hatched (mg) hatched
N 11 49.82 157.45 875 1732
NE 13 51.54 137.15 936 1783
E 12 56.08# 122.08# 781# 1465
SE 26 55.35# 142.96# 836# 3717
S 53 54.09# 118.25# 713# 6267
SW 24 54.33# 116.08# 645# 2786
W 8 44.38 129.38 948 1035
NW 20 49.55 136.80 855 2736
Totals 167 21,521
Note: Underlined values are those for egg cases from compass
directions that took longest to hatch are indicated with #.
TABLE 5.--Results of statistical tests of hatching experiment
conducted in 2006. Analyzed separately by ANOVA for 8 "cardinal"
compass directions were live biomass (initial egg case mass minus
mass of empty case), initial mass of egg case, dry biomass of
hatchlings, and mass of individual hatchlings. Later, data were
combined into two groups (North and South): 'north,' which included
W, NW, N, and NE, and 'south,' which included E, SE, S, and SW, and
analyzed by two-sample (-tests
Analysis Grouping Test P
1. Live biomass 8 cardinal ANOVA 0.0026
2. Initial mass of egg case 8 cardinal ANOVA 0.0037
3. Mass of clutch after drying 8 cardinal ANOVA 0.1414
4. Number of hatched yg per egg case 8 cardinal ANOVA 0.0796
5. Mass of individual young 8 cardinal ANOVA 0.2097
6. Mean mass of egg case N vs. S t-test 0.0001
7. Mass of clutch after drying N vs. S t-test 0.0446
8. Number of hatched yg per egg case N vs. S t-test 0.0088
9. Mean mass of individual young N vs. S t-test 0.4961
Analysis Crit stat df
1. Live biomass F = 3.200 158
2. Initial mass of egg case F = 2.071 158
3. Mass of clutch after drying F = 1.593 158
4. Number of hatched yg per egg case F = 1.862 158
5. Mass of individual young F = 1.399 158
6. Mean mass of egg case t = -3.699 157
7. Mass of clutch after drying t = -1.710 157
8. Number of hatched yg per egg case t = -2.398 157
9. Mean mass of individual young t = -0.098 157
 Printer friendly
 Cite/link
 Email
 Feedback
| |
| Author: | Rose, Robert K.; Bellows, A. Scott |
|---|---|
| Publication: | The American Midland Naturalist |
| Article Type: | Report |
| Geographic Code: | 1U5VA |
| Date: | Jul 1, 2016 |
| Words: | 5621 |
| Previous Article: | Golden-cheeked warbler behavior in relation to vegetation characteristics across their breeding Range. |
| Next Article: | Temporal trends of channel catfish population characteristics in the Missouri River, Nebraska. |
| Topics: | |