Robert R. Jackson and Sandra Smith.
Aggregations of Mallos and Dictyna (Araneae, Dictynidae): Population Characteristics.
Psyche 85:65-80, 1978.

Full text (searchable PDF, 1352K)
Durable link: http://psyche.entclub.org/85/85-065.html

The following unprocessed text is extracted from the PDF file, and is likely to be both incomplete and full of errors. Please consult the PDF file for the complete article.

AGGREGATIONS OF MALLOS AND DICTYNA
(ARANEAE, DICTYNIDAE):
POPULATION CHARACTERISTICS
BY ROBERT R. JACKSON' AND SANDRA E. SMITH North Carolina Division of Mental Health Services Research Section
P.O. Box 7532
Raleigh, North Carolina 2761 1
Generally web spiders are solitary animals. Usually only a single spider occupies a single web, with the relatively common exception of webs jointly occupied by a male-female pair or females with their young. In most species, occupied webs are usually not connected to other occupied webs by silk lines. However, aggregations of various types and degrees occur in some species (reviewed by Buskirk, 1975; Krafft, 1970; and Shear, 1970). These include cases in which occupied webs are in close proximity (sometimes touching), but occupied by single spiders, as well as cases in which single webs are occupied by groups of spiders of all sex/age classes. The species involved are sometimes referred to as "social spiders" (e.g., Burgess, 1978; Kullmann, 1972). At present relatively little is known about the social adaptations of these spiders. Some of the first things we would like to know about animals that occur in aggregations are the size, composition and spacing characteristics of naturally occurring groups. Data concerning these basic social characteristics will be presented in this paper for several species of dictynid spiders. The Dictynidae are a group of web-building cribellate spiders, most of which have body lengths less than 5 mm. In the closely related genera Mallos and Dictyna, there are species with different types of social organization. Most are solitary; at least three are communal, territorial; and at least one is communal, non-territorial. Each individual web of a solitary species consists of a mesh in which prey is captured and a nest within the mesh. Communal, territorial 'Present address:
Department of Zoology, University of Canterbury, Christchurch 1, New Zealand. Manuscript received by the editor June 10, 1978. Pzwhf Μφ5:65-8 (1978). hup Ylpsyche rnlclub orgt8SY85-065.htinl



================================================================================

66 Psyche [March
species (M, trivittatus Banks, Dictyna calcarata Banks, D. albopilosa Franganillo) live in web complexes. Web complexes consist of variable numbers of web units, each consisting of a nest and a mesh; and web units are connected to each other by strands of silk in the interstitial web. Evidently, web units are treated by the occupants as defended territories (Jackson, 1978a). In the communal, non- territorial species (M. gregalis Simon), groups of spiders live in communal webs which consist of surface sheets with holes leading into the interior. The interior contains nests, tunnels and supporting lines connecting the sheet with the substrate beneath (see Burgess, 1976; Diguet, 1909a, 1915). Communal webs are not divided into defended units as in web complexes. More detailed information concerning social organization, web characteristics, and habitats is provided elsewhere (Jackson, l978a).
DATA FROM NATURAL POPULATIONS
For census areas, locations were chosen which seemed to have particularly high densities of spiders. The census areas for Mallos niveus 0. P. Cambridge were one-dimensional (horizontal lines). All other census areas were rectangular. For D. calcarata and Censuses 3-6 for M. trivittatus, complete web complexes of relatively small size were used; and the boundaries of the census areas were chosen to closely correspond to the edges of the web complexes. The distance from each spider in the complex to the nearest conspecific in the vicinity but not in the complex was greater than twice the diameter of the complex in each of these census areas. For Census No. 1 and 2 for M. trivittatus, each area censused was part of a large web complex.
The corners of each census area were marked with pieces of tape. Each web or web unit was completely searched, a process that destroyed the web unit. The occupants of the units were recorded; and a piece of tape, with a code number, was placed at the location of each occupied nest before it was destroyed. The distance of each of these nests to its nearest neighboring nest was measured. The exact locations of spiders on the sheet and interstitial webs proved excessively difficult to measure accurately. This was because they were frequently obscured from view by silk and debris; and the process of locating spiders inevitably disturbed them, leading to



================================================================================

19781 Jackson & Smith - Mallos and Dictyna 67 changes in their positions before they could be recorded. Only the component of the web (nest, mesh, interstitial web) occupied by the spider could be recorded with confidence. When counting the number of spiders per web unit, those in the interstitial web were included in the record of the web unit to which they were in closest proximity.
Spacing and density data were not collected for D. albopilosa. The web complexes of this species were particularly difficult to disassemble in the field because they were three-dimensional and located on the leaves, stems and roots in dense growths of herbaceous plants. Those of D. calcarata and M. trivittatus, in contrast, were two-dimensional and usually built on more accessible flat surfaces. At San Anton Falls (Morelos, Mexico), 23 occupied web units of D. albopilosa were searched, and the occupants were recorded, with no attempt to assess density or spatial relations. In addition to the spiders in the census areas, the occupants of another 48 web units of D. calcarata and 152 of M. trivittatus were recorded. Each of these was in the general vicinity of the census areas. Data for M. trivittatus were collected in June; those for M. niveus, D. calcarata, and D. albopilosa were collected in September. These species seem to feed predominantly in the late afternoon and early evening (Jackson, 1977a). Data in this study were collected at mid-day, when the spiders were less active. LABORATORY STUDIES - Mallos gregalis
In the laboratory, M. gregalis in large communal webs were maintained on a diet of houseflies (Musca domestica), provided at approximately 5-day intervals. Temperature was maintained at approximately 24OC, and the light cycle was approximately 13 L: 11 D. These large colonies were begun from fewer than 200 spiders collected by Burgess (see Burgess, 1976) in Guadalajara, Mexico, 3 yr previous. The webs in the laboratory were on plants and other objects, and they were not enclosed.
Two census methods were used. 1. Small discrete communal webs and relatively discrete portions of larger communal webs were completely disassembled in the spring and early summer. All occu- pants were recorded; and the condition, structure, and size of the webs were recorded. 2. The surface of a large communal web, at least 2 years old, was censused with 30cm transects in the fall. Each transect consisted of a string laid haphazardly onto the web, and all



================================================================================

6 8 Psyche [March
spiders within 1 mm of the string were counted. The web was arbitrarily divided into 3 sections. Two transects were placed onto each section on each census-day; and there were 11 census days, each 2 to 3 days apart. A "census" was defined as the sum of the spiders found on the two transects on a given section on a given census-day.
Mallos niveus
Three aggregations of M. niveus were censused in Guanajaato (Guanajaato, Mexico), each under a different window sill on exterior walls of buildings in the city. The webs were arranged in horizontal straight lines, and the length of the aggregation was defined as the distance between the first web at one end and the last web at the other end (Census Area No. 1, 54 cm; No. 2, 87 cm; No. 3, 273 cm). The distance to the nearest conspecific not in the aggregation was greater than length of the aggregation in each case. Mallos trivittatus
Census Areas Nos. 1, 2, and 3 were at East Turkey Creek in the Chiracahua Mountains (Arizona, USA). The creek passes through a 1.8 m diameter metal culvert, with an interior surface area of 88m2. Almost the entire interior of the culvert was covered by a single enormous web complex. The bottom of the culvert was covered by water; and in the immediate vicinity of the water, web units were scarce. Subtracting these parts of the culvert, the web complex was conservatively estimated to be 79 m2 in area. Census Areas Nos. 1 and 2 were each inside the culvert, within 2 m of opposite entrances and 1.5 m above the bottom of the culvert. Census Area No. 1 was 56 cm X 50 cm; No. 2,60 cm X 56 cm. Census Area No. 3 was a web complex (54 cm X 42 cm) 30 m from the culvert, on a large boulder (ca. 2 m high) beside the creek. The boulder was overhanging the ground, forming a shallow cave (ca. 1 m from ground to top; 0.5 m from entrance to back). The web complex was on the upper surface of the cave.
Census Area No. 4 (56 cm X 32 cm) and No. 5 (56 cm X 56 cm) were web complexes on rock ledges near Estes Park (Colorado, USA). Each was 1 to 2 m above the ground. Census Area No. 6 was a web complex (124 cm X 34 cm) on the upper surface of a shallow



================================================================================

19781 Jackson & Smith - Mallos and Dictyna 69 cave (ca. 1 m high and 1 m deep) made by an overhanging rock ledge in Big Thompson Canyon (Colorado, USA).
Dictyna calcarata
Census Area No. 1 (24 cm X 24 cm) and No. 2 (54 cm X 46 cm) were web complexes on the external walls of buildings in Chapala (Jalisco, Mexico).
POPULATION SIZE AND DENSITY
Densities for M. trivittatus and D. calcarata were relatively similar (Table 1). Since the M. niveus aggregations were linear (one- dimensional) it is problematic to compare densities in this species with the densities found in the other species. The webs of M. niveus in the census areas were oriented such that the shorter distance across each was vertical. The areas given in Table 1 were obtained by multiplying the length of the aggregations by 3 cm, the mean widths of M. niveus webs found on buildings (Jackson, 1978a). M. gregalis tends to live in higher densities. The webs censused by Method No. 1 (Table 2) had a mean density of 0.5 k 0.66 spiders per cm2 and 0.2 2 0.14 per cm3.
Based on many hours spent in the culvert at E. Turkey Creek, it seems that the densities obtained in Census Nos. 1 and 2 for M. trivittatus were representative of the large complex in the culvert. Using the mean of the two densities for spiders (14.23 per 1000 cm2), webs (9.79 per 1000 cm2), and occupied webs (8.79 per 1000 cm2) as estimates for the densities for the complex as a whole, this web complex contained 10,200 spiders, 7,700 webs, and 6,500 occupied webs. This was by far the largest web complex of this or any other species in this study.
Web units of M. trivittatus tend to be 5 cm X 4 cm; those of D. calcarata tend to be 3 cm X 2 cm; and webs of M. niveus on walls of buildings tend to be 4 cm X 3 cm (Jackson, 1978a). In the culvert on E. Turkey Creek, the interstitial area between web units tended to be quite small, since the mean for nearest neighbor distance was not much greater than the diameter of web units (Table 1). In all other censused web complexes, web units tended to be 2 to 4 web unit diameters apart; and aggregated M. niveus webs tended to be 2 to 4 web diameters apart also.




================================================================================

Table 1. Censuses of web complexes (M. trivittatus and D. calcarata) and aggregations of individual webs (M. niveus) in the natural habitats of the spiders.
~allos trivittatus Dictyna calcarata Mallos niveus Census 1 2
Area (Cm*) 2800 3360
Number of Spiders 48 38
Number of Web Units or Webs 29 31
Occupancy Rate' 93 90
Density (Spiders per 1000 Cm2) 17.14 11.31 Inter-Nest Distance2 6.11 5.43
310.68 k0.50
'No. of web units (or webs) occupied by dictynids/total No. of web units (webs) X 100. 2Mean k S.D. for distance between each nest and the nearest neighboring nest. Only occupied nests considered.



================================================================================

19781 Jackson & Smith - Mallos and Dictyna 7 1 Some of the larger communal webs of M. gregalis in the laboratory had surface areas and volumes estimated to be as large as 1300 cm2 and 1500 cm3. If the densities in these large webs were comparable to those in smaller webs, they can be estimated to have contained 300 to 700 spiders. Based on the descriptions of Diguet (1909a, b) and Burgess (1976), 4 m2 will be used as an approximate estimate for the surface area of larger communal webs on trees in Mexico. If densities in nature are comparable to those in the laboratory then these very large webs may have contained as many as 20,000 spiders.
WEB OCCUPANTS
Solitary species, including M. niveus, usually occurred one spider per web (Jackson, 1978a), and this was true of all censused webs in the aggregations of M. niveus. However, each occupied web unit of the three communal, territorial species usually contained a set of several spiders. These could be almost any combination of females, males, and immatures of varied sizes (Table 3), except that web units were never shared by two adults of the same sex. In the only instances in which male-male or female-female pairs were seen together in the same web units, aggressive interactions took place, after which one individual departed. In contrast, large numbers of M. gregalis belonging to all sex/age classes shared the same communal webs without aggressive behavior or cannibalism. DISTRIBUTION OF WEB OCCUPANTS
Large proportions of each sex/age class of M. trivittatus occupied the nests (Tables 3 and 4); but adult females were especially prone to be inside nests, and there was a preponderance of immatures in the interstitial web. These data were consistent with more casual observations of M. trivittatus in other locations and of the other communal, territorial species. During the day, M. niveus and other solitary species were most often, but not always, inside nests. All M. niveus in censuses were inside nests.
Many hours of casual observation in the laboratory gave the impression that adult males and immatures of M. gregalis were more prone to be on the outer surfaces of webs, and females were more prone to be in the interior. (Data related to this will be presented below.)




================================================================================

Psyche [March
Table 2. Data from complete disassembly of 19 M. gregalis webs or subunits of webs (census Method No. 1, see text). Area and volume are estimates, since borders of webs were irregular. Small immature: ca. 1 mm in body length (color: uniformly brown). Large immature: body length ca. 2 mm (gray markings present on body). Adult females: ca. 3 mm. Adult males: ca. 3 mm. Web class A: 3-D and lacking evidence of deserted areas. B: 3-D and containing conspicuous deserted areas. C: 2-D and lacking conspicuous deserted areas. D: 2-D (but not as flat as C) and containing conspicuous deserted areas. E: 3-D, small discrete web on irregular substrate. A-D: census areas were subunits of large webs, except for some of C that were discrete webs. 3-D: web occupies relatively much space in each of 3 dimen- sions. 2-D: web confined predominately to 2 dimensions ("flat"). "Deserted" areas: no longer occupied by M. gregalis, not substantially adhesive, and covered by fly carcasses and dust. Note: as population size increases, composition of population shifts toward decreasing percentages of females and increasing percentages of im- matures. Kendall rank correlation (Sokal and Rohlf, 1969) for increasing percent- ages of small immatures with increasing population size (P < 0.001). Sex/ Age Class Composition
(%) of Population
Popu-
lation
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
Area
(cm2)
144
144
105
150
144
7 5
70
7 0
150
70
144
70
64
150
155
100
400
45
150
Volume Web
(cm3) Class
Popu-
lation
Size
5
7
9
10
10
11
2 1
23
25
27
3 1
34
40
47
53
74
8 2
114
288
Adult
Female
60.0
71.4
0.0
90.0
30.0
27.3
19.0
65.2
4.0
48.2
12.9
11.7
27.5
23.4
32.1
17.5
9.7
7.0
9.3
Adult
Male
0.0
0.0
11.1
10.0
0.0
0.0
0.0
17.4
4.0
18.5
0.0
5.8
0.0
6.3
0.0
1.3
2.4
0.0
2.0
Large
I m ma-
ture
20.0
14.2
44.4
0.0
10.0
27.3
19.0
17.4
8 .O
7.4
0.0
32.3
40.0
21.3
20.7
22.2
7.3
12.3
9.7
Small
Imma-
ture
20.0
14.2
44.4
0.0
60.0
45.4
61.0
0.0
84.0
25.9
87.0
50.0
32.5
48.9
47.2
55.4
80.4
80.4
78.8




================================================================================

19781 Jackson & Smith - Mallos and Dictyna 73 SEXIAGE CLASS RATIOS
Based on the spiders in the two censuses in the culvert at E. Turkey Creek, this large web complex was estimated to have had a population consisting of 27% adult females, 9% adult males, and 64% immatures. The adult sex ratio (Ma1e:Female) was approxi- mately 1:3. Based on the censuses at Chapala, populations of D. calcarata in web complexes on buildings were composed of 14% adult females, 5% adult males, and 81% immatures; and the adult sex ratio was 1:3, as for M. trivittatus. No males were found in the censuses of M. niveus. Generally only small numbers of males of the solitary species were found, and these were predominantly ones in webs with females (Jackson, 1977b).
Based on the 19 censuses of M. gregalis by Method No. 1 (disassembly), population composition was 29.8 k- 25.49% adult females, 4.1 IL 5.97% adult males, 17.6 2 12.17% large immatures, and 48.2 k 27.41% small immatures. Immatures comprised a greater proportion of the spiders in webs with large population size (Table 2). The average adult sex ratio was approximately 1:7. For the 33 transect censuses (Method No. 2), population composition was 12.3 k- 7.44% adult females, 3.0 2 3.85% adult males, 38.0 k- 12.96% large immatures, and 46.2 2 13.1 1 % small immatures (adult sex ratio, ca. 1:4), with a total of 16.8 k 2.34 spiders occurring in each census. The three sections of the web were comparable. AGGREGATION SIZE AND DENSITY -
EVOLUTIONARY CONSIDERATIONS
The occurrence in certain areas of very large populations of some species, especially M. gregalis and M. trivittatus, and the great densities of spiders within these populations contrast markedly with the majority of Dictyna and Mallos species. Most species are solitary, with individuals living in relatively widely spaced indi- vidual webs on stems and leaves of vegetation (Jackson, 1978a); and although density censuses were not carried out for these, our impression is that an area of 50 to 100 m2 in a population of a solitary species would usually contain fewer than a dozen indi- viduals. In contrast, the largest web complex of M. trivittatus