DRAGLINE-FOLLOWING BY MALE LYCOSID SPIDERS

Wolf spiders stalk or ambush their prey rather than build preycapture webs. Production of silk by lycosids is important during aerial dispersal via ballooning (Richter, 1970b) and, in some, during post-immobilization wrapping of prey (Rovner and Knost, 1974), and construction of egg sacs or sperm webs (Richter, 1970a). During locomotion both sexes lay down silk draglines that are generally assumed to have a stabilizing or security function (Richter and Van der Krann, 1970). In addition, female draglines induce courtship behavior in male conspecifics and, in some species, males are reported to have followed a female trail, but no systematic analyses of the dragline-following behavior have been attempted (Bristowe and Locket, 1926; Engelhardt, 1964; Kaston, 1936). The present study is concerned with the trail-following behavior of male Lycosa rabida and L. punctulata. Specific areas of investigation include determination of the use of various appendages during following, analysis of cues involved in initiating and maintaining trail-following, and an interspecific comparison of trailfollowing behavior.

containing males were located on shelves along one wall of the laboratory that received indirect sunlight.Laboratory humidity was not controlled (range 36-56% RH) and temperature ranged from 23-28C.Cages containing females were placed in a con- trolled environmental chamber (Frease model 818) at 15C night and 20 C day and 53-56% RH, with a photoperiod approximating natural conditions.The lower temperatures were necessary to re- tard female development and to delay onset of eggsac production, which even occurs eventually in unmated females, making them useless for further experimentation.Females were removed from the environmental chamber and allowed sufficient time to warm to room temperature before experimental use.
In order to lay draglines along predetermined trails spiders were anesthetized with CO2.A thread leash, secured to a hemostat, was tied around the cephalothorax of each spider between the second and third pairs of legs.Upon recovery, the spider was led along a path such that the dragline was laid either directly on the sub- stratum to form "ground lines" or suspended above the substratum by being laid across horizontally placed glass rods to form "aerial lines" (Fig. 1).Lines connecting parts of an artificial plant made of wood dowels were also called "aerial lines."Male draglines or imitation draglines (nylon thread or human hair) were laid in either a continuous or discontinuous manner with the female trail during some experiments.
The test trail was then surrounded by glass walls to provide an arena (54 64 cm), the position of which was rotated in a random manner between experimental runs.The position, length, and direc- tion of the dragline, including attachment disks (produced by the spider to fix the dragline to a substratum) were recorded on a dia- gram of the completed path.Test males were of two physiological conditions: (1) unprimed males, which had not encountered female silk for the previous 72 hr and (2) primed males, which were induced to court in response to substratum silk from a female cage, and then tested hr later for dragline-following behavior.Experimental trials began when a male released into the arena contacted and explored a female line with his palps.Trials lasted for 5 min, during which time data were recorded in shorthand notation from a distance of 0.5-1.0m from the experimental males.Tietjen Dragline-Following by Spiders 167 Female dragline Imitation male dra line Figure 1.Diagrammatic summary of a typical trail including attachment disks (D) and female aerial lines (A).Female ground lines (G) and imitation draglines are laid contiguously (C) and discontiguously (DC) with the female dragline.Small strips of masking tape (T) attach imitation dragline to the wall of the arena and to 5-mm diameter glass rods (R).Glass rods were usually parallel and 9-11 cm apart.
When artificial "plants" (P) were incorporated into the experiment, female trails up to 236 cm in length were constructed.
dopteran silk or conspecific male or female silk.Male L. punctu- lata were also exposed to female L. rabida silk.Clean cage cards were used as controls.A positive response by males to the silk consisted of exploring the silk and initiating courtship behavior within 5 min.Statistical analyses of data were performed according to the methods of Conover (1971) and Sokal and Rohlf(1969).All means are accompanied by their standard errors.

COMPARISONS OF YEARLY SAMPLES
No difference was found in trail-following behavior between 1974  and 1976.Unprimed male L. rabida showed no difference in fre- quency of following-behavior (X 0.0, df 1, P > .50)and in Psyche [June total distance followed during the 5-min test period (Mann-Whitney test, T= 173,P > .80).Primed male L. rabida showed no difference in frequency of following (X2 .504,df 1, P > .70)or in total distance followed (Mann-Whitney test, T--229.5,P > .56).Male L. punctulata showed no difference in the frequency of following (X 2.5, df 1, P >.90) or in total distance followed (Mann- Whitney test, T 235.5, P 3> .84).Therefore, data for the two years were pooled for analysis.
DESCRIPTION OF DRAGLINE-FOLLOWING Males of both species followed conspecific female lines.Al- though high speed (36-180 fps) cinematographic analysis indicates that there are slight differences in the use of appendages during dragline-following (Tietjen, unpublished data), the two species ex- hibit many similarities.Upon contacting the silk line for the first time, males tend to examine it with alternating movements of the palps such that the dorsal surface of the palp contacts the line and moves anteriorly along it.Such behavior is soon terminated if the line is a male's or is an imitation.Dragline-following results if the line is that of a female conspecific.. Males usually wander to the edge of the arena and walk around the periphery if chemo-and mechano- exploratory behavior is terminated.
In both species, dragline-following is characterized by the male straddling the silk so that it passes medially beneath him.Although the first pair of legs may occasionally pluck at the line, the palps are the most important appendages used.The palps are alternately moved forward and back, making contact with the silk on their medial surface rather than on the dorsal surface, which is used dur- ing the initial contact with the line.Experiments conducted under red light (Kodak safelight Filter No. 1) and with blinded animals indicated that vision is not necessary for dragline-following.Males may accompany their dragline-following behavior with elements of courtship or, more typically, interrupt dragline-following with sep- arate bouts of courtship behavior.DRAGLINE-FOLLOWING BY L YCOSA RABIDA Both primed and unprimed male L. rabida followed female aerial draglines.Unprimed males exhibited a lower incidence of draglinefollowing, with 84.6% of primed and 65.2% of unprimed following (X 5.83, df 1,P < .025).Comparing only those males that Imitation and male draglines number followed 7 4 0 number not followed 6 12 16 Ground lines number followed 0 0 13 number not followed 18 9 4 followed draglines, no significant difference was found in the total distance followed between primed and unprimed animals (Mann-Whitney test, T= 1069.5, P .72).No males of either group fol- lowed ground lines, although they would usually examine such lines with their palps (Table I).None of the male L. rabida followed isolated imitation draglines; and chemoexploratory behavior to such lines was exhibited only five times.Males in the process of following a female line would occasionally follow imitation and male draglines laid contiguously with the female trail (Table I).Once, a male following a female line continued on and followed the entire length of a human hair that constituted an 8.5 cm interruption in the female trail, losing no momentum in his following behavior.Male draglines were fol- lowed four times, human hairs were followed six times and nylon thread was followed once.
The polarity of each male's trail was calculated as the difference between total distance traveled in the same direction as the female Psyche [June trail was laid and total distance traveled in the opposite direction.Primed males exhibited a positive polarity by following the correct direction more often than the incorrect direction (X2 7.36, dr= 1, P < .01),and by traveling further in the correct direction along the female trail (Wilcoxon test, T 252.5, P < .01).Unprimed males did not show a directional preference (X 0.0, dr--1, P .50; Wilcoxon test, T 525, P > .64;Table I).
Primed and unprimed male L. rabida showed no difference in the frequency of courtship during dragline-following (X .613,df 1, P .50)or in the elapsed time during a test before court- ship was observed (= courtship latency)(Mann-Whitney test, T 70.0, P > .54;Table I).Primed males that courted showed no correlation between total distance of dragline followed and court- ship latency (Spearman's rho, 0 + .026,P > .53).Unprimed males courting early in the test showed decreased total draglinefollowing (Spearman's rho, 0 + .692,P < .01).Comparing primed and unprimed males that both courted and followed draglines, no difference was found in the total amount of dragline- following (Mann-Whitney test, T-747.0,P .65).
DRAGLINE-FOLLOWING IN L YCOSA PUNCTULA TA L. punctulata was a more difficult species to study experimentally than L. rabida.Females often would not lay the required lines and both sexes tended to break the lines by catching a claw on them, and the males by running through the lines.
Incidence of courtship did not differ between the two groups with primed males courting during one and unprimed males courting during three of the trials (X 1.25, df-1, P .70).Data for the two groups were pooled for further analysis.
In 1976 only unprimed males were used.During 34 trials 23 males followed a mean total distance of 52.4 + 11.3 cm.No dif- ferences were found between pooled 1974 data and the 1976 (all unprimed) data for incidence of following (X 2-3.41, df-1, P .05),or for distance followed (Mann-Whitney test.T-235.5, P > .77).Unprimed males in 1976 exhibited courtship three times during dragline-following; no difference in courtship frequency was found between the pooled 1974 data and those for 1976 males (X 0.34, df= 1, P .70).In order to have a larger data base the data were pooled for both years.
Unlike male L. rabida, male L. punctulata initiated dragline- following at ground lines and appeared to use their palps in a manner similar to chemoexploratory behavior.Male L. punctulata rarely examined isolated male lines or imitation lines; of 16 that did palpate these lines, none followed them (Table I).Males ex- hibited a positive polarity (X 4.33, df-1, P .05),and traveled a greater distance in the correct direction along the female trail (Wilcoxon test, T-255.5, P < .03).
Males that courted during dragline-following showed no corre- lation between courtship latency and distance followed (Pearson's r, r-+ .700,P > .90).No difference in distance followed was found between males that courted during dragline-following and those that did not (Mann-Whitney test, T 93.0, P .58).Court- ship latency of males placed on a female substratum was affected after dragline-following; those which had recently followed a drag- line exhibited a mean courtship latency of 0.66 + .29 min (N 23) compared with 2.30 _+ 0.25 min (N-20) for those which had not (Mann-Whitney test, T-52.0,P < .0001).
COURTSHIP RESPONSES TO VARIOUS TYPES OF SILK Male L. rabida courted in response to silk in a female cage in 50% of the trials.Courtship latency in response to female silk was 2.30 + .25 min.The incidence of courtship in response to female silk differed from the courtship frequency in response to a clean cage, lepidopteran silk and male L. rabida silk (X 33.99, df 3, P < .001).The incidence of courtship did not differ among the latter three conditions (X 2.05, df-2, P > .70). (Table II).
Male L. punctulata courted in response to female conspecific silk in 61.1% of the trials and exhibited a mean courtship latency of 2.55 min _+ .44 min.The incidence of courtship in response to female conspecific silk, female L. rabida silk, lepidopteran silk, a clean cage and male L. punctulata silk differed significantly (X2 35.93, df-4, P .001),while courtship frequency did not differ among the latter four test conditions (X 2.10, df 3, P .50).
To simulate the effects of dew under natural conditions, silk trails of females of both species were sprayed with a fine mist of distilled water and were air dried (= washed lines).treated lost elasticity and drooped between the supporting glass rods.Both unprimed male L. rabida (N 10) and unprimed male L. punctulata (N 10) exhibited short bouts of chemoexploratory behavior at washed lines, but neither species exhibited courtship or trail-following behavior during the 5-min trials.

DISCUSSION
Trail-following has been observed in a variety of arthropods.Terrestrial chemical trails are often employed by eusocial insects, but such trails are detected by the receiver as a rapidly decaying vapor cloud above the substratum and provide no directional in- formation (Wilson, 1971).Lepidopteran larvae (Malacosoma, Hyponomeuta and Thaumetopoea) have been observed to follow silk trails from their nest to a food source; but tactile information provided by the silk is unimportant, and the pheromone decays within a few minutes (Fitzgerald, 1976;Wigglesworth, 1966).Wolf spider trails differ from the above examples in function (mating vs. food), in the use of a contact rather than an olfactory phero- mone, and in the presence of an inherent polarity.Silk lines in the laboratory up to 4 days old will elicit following-behavior in male L. rabida and male L. punctulata; and a cage substratum with female silk elicits courtship behavior for over a month (Tietjen, unpub, data), indicating that the spider pheromone does not decay over a relatively long period of time.
Female lycosid spiders exhibit low motility during the breeding season and adopt a "sit and wait" reproductive strategy; the highly active males seek out the females (Hallander, 1967a, 1967b; Hol- lander, 1972; Muma, 1973; Richter, et al., 1971).Mathematical modeling of females laying no silk trails and silk trails of various lengths indicates that the presence of silk effectively increases the size of the female target as perceived by the searching male (Tietjen, unpub, data).Thus, the ability of male L. rabida and L. punctu- lata to follow conspecific female draglines to the source increases the reproductive success of these males over spiders that depend on chance encounters in a heterogeneous environment.Female silk also alerts the male to the presence of a female conspecific and, as shown in L. punctulata, decreases the courtship latency.The latter two functions effectively reduce the likelihood of intraspecific can-  nibalism (Platnick, 1971).
Results from the present study support the hypothesis that a phe- romone associated with the female dragline elicits courtship in re- ceptive males, i.e., these behaviors are not dependent exclusively upon tactile information provided by the silk.Both L. rabida and L. punctulata are found with a variety of other spider species.Dependence on a pheromone to elicit sexual behavior in males would reduce the chances of males following draglines of other species and thereby adversely affecting the time budget allocated for reproductive purposes.
The propensity of male L. rabida not to follow ground lines which cannot be plucked by the palps (thereby providing little tac- tile input) and their occasional following of imitation draglines, which provide no chemical cues, indicate that the pheromone is responsible only for initiating the following behavior in males.Once a male is following a dragline, tactile cues appear to be ade- quate.Males are expected, however, to occasionally sample the trail for female pheromone, as only 38% of male and imitation Psyche [June dragline contacts during female dragline-following resulted in fol- lowing by males, and most imitation draglines were not followed the entire length.Male L. punctulata followed ground lines, but did not follow male and imitation draglines.This species appears to depend upon the female pheromone for initiating and for maintaining dragline- following.Apparently, male L. punctulata are either more sensi- tive to tactile input provided by ground lines, or more sensitive to pheromonal cues than male L. rabida, or both.
The differential sensitivity to tactile and chemical cues exhibited by the two species may be related to microhabitat preferences.L. punctulata is found in the lower levels of the herbaceous stratum (Eason and Whitcornb, 1965) in which one expects a greater di- versity and density of spiders (Whitcomb, Exline and Hite, 1963)  and consequently more draglines.Many lycosids of similar or larger size are found within this microhabitat of which male L. punctulata could be potential prey, including L. helluo and L. caro- linensis.Continuous sampling for female pheromone during dragline following reduces the likelihood that males will follow another species' trail in a high silk density microhabitat.In addition, due to the greater foliage density at this level, many draglines could be expected to be laid directly on a substratum.A greater sensitivity to tactile and/or chemical cues could allow male L. punctulata to follow such ground lines.
L. rabida, on the other hand, is found high in the herbaceous stratum (Eason and Whitcomb, 1965; Kuenzler, 1958) and is asso- ciated with lower interspecific silk densities derived mainly from smaller species such as salticids and oxyopids.Foliage density is less, and a greater volume of open space is found at this level.
Fewer female silk trails would be expected to be found directly on this substratum.Under these conditions the sensitivity to tactile and/or chemical cues exhibited by L. punctulata would not be required for efficient dragline-following by male L. rabida.
Silk lines are destroyed mechanically in the field by wind and the activities of animals.Moisture resulting from rain and early morning dew inactivates the female pheromone found on unbroken lines (Dondale and Hegdeker, 1973; Hegdeker and Dondale, 1969).
All of these effects reduce the likelihood that males will follow draglines laid the previous day.
Male L. punctulata and primed male L. rabida may be able to extract directional information from the female dragline, though it is unlikely that a pheromone concentration gradient could provide the directional information since the pheromone remains active over long periods of time.Examination of female L. rabida silk under a light microscope (450) as well as scanning electron micro- scope inspection of female Lo punctulata silk (1000) indicate no evidence of structural features on the dragline which could provide directional information to males.
Spiders are able to detect slight changes in web tension through vibration receptors during prey capture, web building, and court- ship (Robinson, 1969; Walcott, 1969; Witt, 1975).Such variations in tension, related to the direction in which the dragline was laid, could provide directional information to dragline-following males.If present, the tension differential may be related to the structure of the attachment disk by which the silk is fixed to a substratum.
As the dragline enters the attachment disk it is tightly wound, whereas the silk leaving the attachment disk is composed of single threads which later join and wind to form the dragline.Single silk threads leaving an attachment disk would be expected to damp transients produced along the line more readily than a complete dragline entering an attachment disk.If this were the case, drag- line-following toward the pole with the higher resonant frequency would constitute movement in the direction the silk was laid (Tietjen, unpub, data).Such a system would require that males im- mobilize a portion of the dragline, perhaps with a palp, while plucking on either side of the fixed portion.Male spiders en- countering a female line often wipe the dorsal surface of each palp alternately over the silk and often pluck at the silk with the first legs as they swivel their bodies along a short length of the trail.This behavior may represent both a sampling for female phero- mone and a testing for differential mechanical properties of the line related to the direction in which it was laid.Unprimed male L. rabida did not exhibit a polarity in dragline- following, but a positive correlation between courtship latency and total distance followed was observed.The above correlation indi- cates that unprimed males probably have a lower threshold for chemical cues than do primed males.If unprimed males are at- tending more to chemical cues, they would be less likely to follow Psyche [June a dragline in a preferential direction if directional cues were de- pendent on the mechanical properties of the line.

SUMMARY
The dragline-following behavior of male wolf spiders was ob- served in response to male, female and imitation draglines.Male L. rabida followed female draglines suspended above the sub- stratum but not those laid directly on the substratum.These males followed male silk and imitation lines if laid contiguously with female draglines, suggesting that they depend on tactile cues dur- ing dragline-following.On the other hand, male L. punctulata followed both aerial lines and ground lines laid by females but did not follow imitation or male lines.This suggests that male L. punctulata are more sensitive to tactile and/or chemical cues than are male L. rabida.Microhabitat preferences of the two species may explain the above differences.Data also suggest that males of both species are able to extract directional information from the dragline, perhaps by tensional cues.

Table I :
Performance of Male Lycosa rabida and Male L. punctulata During Dragline-Following of ConspecificMale, Female, and Imitation Silk.

Table II .
Courtship Responses to Various Types of Silk by Male Lycosa rabida and Male L. punctulata.