A correlational and factor analysis of anticipatory and consummatory measures of sexual behavior in the male rat

A correlational and factor analysis of anticipatory and consummatory measures of sexual behavior in the male rat
of 12
All materials on our website are shared by users. If you have any questions about copyright issues, please report us to resolve them. We are always happy to assist you.
Related Documents
  P1 doeM koaiu~ogy. Vol 15. No. 5 6,W.p 329-340.1990 0306-4530i90 53.00 .00Prinied irn mz Urghm CM9 awm= P ep A CORRELATIONAL AND FACTOR ANALYSIS OF ANTICIPATORY AND CONSUMMATORY MEASURES OF SEXUAL BEHAVIOR IN THE MALE RAT JAmEs G. PFAus, Scorr D. MUMNELSON and ANTHoNY G. PHILLIPS Department of Psychology, University of British Columbia, Vancouver. British Columbia, Canada  Received April 1990; ginfinalform 9 Jun 1990) SUMMARY This study investigated the relationship among measures of anticipatory and consummatorysexual behavior displayed by male rats in the bilevel chambers designed by Mendelson andGorzalka (1987). Normative data from a standard test of sexual behavior were gathered from 80 intact, sexually experienced male Long-Evans rats and subjected to multiple correlational and factoranalyses. The correlational analysis confirmed that several consummuatory measures of copulationwere related significantly, whereas the anticipatory me-asure, level changing, was statistically inde- pendent of consummatory measures. Factor analysis using orthogonal rotations revealed five factorsthat accounted for 95 of the intersubject variance for all measures: Copulatory Rate, Initiation, Hfit Rate, Mount Count, and Anticipation. These results indicate that at least five conceptual mecha- nisms are required in any theoretical description of male sexual behavior in the bilevel chamber. In particular, the extraction of separate anticipation and initiation factors indicates that these aspects of male sexual behavior are distinct. The use of bilevel chambers thus may facilitate the identification of potential neurochernical or endocrine mechanisms associated with different aspects of malesexual motivation. In addition, several statistical techniques are discussed with the aim of reducingthe elevated experiment-wise error that can occur when related measures of sexual behavior are analyzed independently. INTRODUCTION MucHi of the theoretical description of male sexual behavior is besed on analyses of the copula- tory patterns of male rats (e.g., Beach, 1956; Larscon, 1956; Sachs & Barfield, 1976; Sachs, 1978: Toates, 1986; Sachs & Meisel, 1988). These analyses have led to the formulation of hypothetical models of male sexual behavior, many of which are useful heuristically for deter- mining potential neural and endocrine mechanisms.Several models have been derived from the statistical relationship among measures of male sexual behavior. For example, Beach (1956) calculated multiple correlations among normative measures of copulatory behavior. The relationship among those measures suggested that aminimum of two conceptually distinct mechanisms controlled the sexual behavior of male rats: Address correspondence and reprint requests to: Dr. Anthony G. Phillips, Department of Psychology.University of British Columbia, Vancouver, British Columbia V6T I Y7, CANADA. *Drs. Pfaus and Mendelson are now at T~he Rockefeller University. New York, New York 10021 U.S.A.329  a sexual arousal mechanism that controlled both precopulatory investigation of the female andthe initiation of mounts, and an intromission and ejaculation mechanism that focused the male's responses on genital stimulation until ejaculation was achieved.Sachs (1978) and Dewsbury (1979) used multiple correlations and factor analysis to exa- mine the statistical relationship among eight standard measures of copulatory behavior in malerats: the mount and intromission latencies, the ejaculation latency, the number of mounts and intromissions preceding ejaculation, the 'hit rate (intromission ratio), the interintromission interval (Il), and the postejaculatory interval (PEI). Sachs (1978) also analyzed the absolute and relative phases of the PEI. In Sachs' analysis, the correlations among those measures were examined separately in three data sets from the laboratories of three investigators, Sachs(n = 18). Barfield (n =23), and Bermant (n = 74). Most of the data were derived from the rat second test with receptive females. In Dewsbury's analysis, the intercorrelations were exa- mined separately in three successive tests of copulatory behavior (N = 312). Although littleinformation about the testing conditions was provided in either study, it may be assumed that the rats were tested in small square or circular unilevel testing chambers under relatively optimal conditions (e.g., during the dark phase of the circadian light/dark cycle). Similar correlations among the measures of copulatory behavior were significant in bothstudies. The mount and intromission latencies were highly correlated, as were the number ofmounts and the intromission ratio. The III was correlated with the ejaculation latency, the absolute phase of the PEI, and the total PEI. Ukewise, the ejaculation latency was correlatedwith the absolute phase of the PEI and the total PEI. Finally, the total PEI was correlated with its absolute and relative phases. Dewsbury (1979) found 17 significant correlations in two ofthe three tests that were not reliably significant in the data samples analyzed by Sachs (1978). For example, the mount latency was correlated positively with the ejaculation latency, the 1m and the PEI. The intromission latency was correlated positively with the number of mounts and intromissions, the HI, the PEI. and negatively with the intromission ratio. The ejaculation latency was correlated positively with the number of mounts and negatively with the intromis-sion ratio. However, several of those correlations were quite small and accounted for relatively small proportions of the variance. Sacks (1978) subjected the correlational matrix of each of the data samples to principal components analysis with orthogonal rotations. This analysis revealed four factors thataccounted for over 80% of the variance for all measures in the three data samples. Thesefactors were given tentative names based on the measures that loaded most heavily on each factor (ie., whose variance was accounted for by each factor). The factor that accounted for the largest proportion of the variance (40%) was termed the copulatory rate factor. This factorwas loaded by the M.I the ejaculation latency, and the absolute and total PEI. The factor that accounted for the second largest proportion of the variance (16%) was called the initiation factor. This factor, reminiscent of Beach's sexual arousal mechanism (1956), was loaded heavily by mount and intromission latencies, and moderately by the relative and total PEI. An equal proportion of the variance (16%) was accounted for by the hit rate factor. This factor was loaded heavily, by the number of mounts and the intromission ratio. The final proportion of the variance considered (14%) was accounted for by an intromission count factor. This factor, reminiscent of Beach's intromission and ejaculation mechanism (1956), was loaded by the number of intromissions and the relative refractory period. Using principal componentsanalysis, Dewsbury (1979) extracted three factors that were nearly identical in composition to Sachs' copulatory rate, initiation, and hit rate factors. These three factors accounted for over 330 J. Q PFAUS et at  STAT151CAL ANALYSIS OF MALE RAT SEXUAL BERAVIOR 331 75 of the variance in the three tests. However, Dewsbury did not find evidence of an intro- mission count factor.Although the similarity of the factors indicates that certain measures of copulatory behavior are relatively stable in male rats, there were several inconsistencies regarding the factorloadings in these studies. For example, in the sample provided by Bermant, the number ofmounts, the MI, and the ejaculation latency loaded onto the initiation factor, whereas in Sachs'and Barfield's sirriples they did not. In two of the three tests conducted by Dewsbury, the PEI loaded onto the initiation factor and the ejaculation latency and number of intromissions loadedonto the hit rate factor. In one of the three tests the mount latency also loaded onto the copula-tory rate factor. It is impossible to determine whether the different number of subjects in each factor analysis, or any possible differences in testing procedure, contributed to the inconsistent factor loadings. Sachs (1978) also noted that none of the analyses took the contribution of the female's behavior into consideration. Given evidence that the solicitation and pacing behavior of the female can alter the copulatory behavior of the male (Madlafousek & Hlhinak, 1978; McClintock, 1984). it is important to examine measures of male copulatory behavior in condi-tions where the females can control the sexual interactions.The bilevel chambers designed by Mendelson and Gorzalkca (1987) provide an opportunityto assess the copulatory behavior of male rats in a situation where the female can pace the cop-   ~~~ulatory contact. Male rats typically chase the females from level to level following each intro- mission. This behavior is reminiscent of the chasing behavior displayed by male rats givenaccess to receptive females in a large open field (McClintock, 1984). In addition, the bilevelchambers allow a measure of anticipatory sexual behavior, level changing (Mendelson & Pfaus,1989), to be recorded during a 5-min period prior to the introduction of the receptive female. To further evaluate the sexual behavior of male rats in the bilevel chambers, we derived correla-tions among 11 measures of sexual behavior displayed by 80 sexually experienced male rats. The underlying structure of the correlation matrix then was examined by principal components analysis and the factors interpreted with respect to both theoretical and practical aspects of male sexual behavior. SUBJECTS AND METHODS Animals and Surgery Femiale and male Long-Evans rats wer: obtained from Charles River Canada. Inc., St. Constant, Quibec. They were separated by sex and housed in groups of six in 18 x 25 x 65 cm wire mesh cages in a colony roommaintained on a reversed 12 hr light:12 hr dark cycle at approximately 211 C. Food and water were continu-ously available in the home cages. ibe females were bilaterally ovariectornized via lumbar incisions undersodium pentobarbital anesthesia (Somnital, 40 mgfkg) at least 4 wk before testing. Sexual receptivity was induced in the females by subcutaneous injections of estradiol benzoate (10 ptg) and progesterone (500 pg) 48 hr and 4 hr, respectively, before each test in which they participated as stimulus females. Apparatus The testing chambers were similar to those designed by Mendelson and Gorzalka (1987). Each chamberwas constructed of Plexiglas, with outside dimenisions of Si x 70 x IS cm (Fig. 1). The narrowness of the chambers forced animals to mate with their flanks positioned towaird the observer, the optimal orientation for evaluating the sexual behavior of rats. However, the chambers were not so narrow as to prevent animals from turning about freely. A platform, 40 cm in length, centered, and set 28 cm above the floor, divided eachchamber into two levels. A set of ramps and narrow landings at each end of the chamber allowed animals to move from one level to the other.  332 1. C. PPAUS t al. Fo 1: 7The bilevel chambers used In the present study. A standard test of sexual behavior con- sisted of an antictpatory phase. during which the males were placed in the chambers for a 5-witn test of level changing (Merndelson Ffaus. 1989); and a consummatory phase. during which sexu-ally receptive females were placed in the chambers with the males for a 20-min test of copulation. Behav'ioral Testing Procedures Eighty males (450-700 z ere given 10 baseline tests of sexual behavior in the bilevcl chambers. Thesetests were 25 win in duration and were conducted every four days during the middle third of the dark cycle.During these tests, each male was placed in a bilevnI chamber for a 5-win test of level changing. after which a sexually receptive female was placed in the chamber for 20 min. The occurrence of each level change, mount, intromission, and ejaculation was recorded on a computerized event recorder which derived the following standard measures of male sexual behavior: (1) number of level changes (LC) in the 5-min period; (2) latencyin seconds to initiate level changes (LCL); (3) mount and (4) intromission latencies (ML and II_ time in seconds from the introduction of the femnale to the first mount or intromiission); (5) ejaculation latency (EL; time in seconds from the first intromission to ejaculation); (6) PEI (time in seconds from ejaculation to the next intromission); (7) number of mounts (NM) and (8) number of intromissions (NI) prior to each ejaculation; (9) III (the ejaculation latency/number of intromissions); (10) intromission ratio (IR) (number of mounts with intromission / total number of mounts with and without intromission); and (1 1) total number of ejaculationsachieved in the 20-min test (NE). By the tenth baseline test, all 80 males had reached the criteria for inclusion:stable rates of level changing during the 5-min adaptation period; intromission within 2 win of the presentation of the female; ejaculation within 15 min of the first intromission; and the reinitiation of intromission within IO win of ejaculation on each of the last three baseline tests. Data from a final test of copulatory behavior, con-ducted 4 days after the last baseline test, were subjected to statistical analysis. SiadstncalAnalysis Means standard errors, and multiple correlations among each of the I I measures of sexual behavior were calculated with a commercially available statistical package (Complete Statistical System, StatSoft Tulsa OK, USA). Two-tailed determinations of significance were made for each correlation coefficient, p < .O5. Principal components analysis was conducted with the same statistical package, with normalized orthogonal(varimax) rotation of the matrix of correlation coefficients (Conmrey, 1973). This reproduced the method usedby Sachs (1978) and Dewsbwry (1979). 5   I  STATISTICAL ANALYSIS OF MALE PAT SEXUAL BEHAV OM RESULTS The means, standard errors, and correlation coefficients among the II measures of sexual behavior are shown in Table I. Correlation coefficients with a statistical significance of p<0.05 are shown in Table HI. The most interesting finding in this analysis was that LC did not corre- late significantly with any of the other measures of sexual behavior. This distinguishes LCfrom measures that are typically used to infer sexual motivation, e.g., the ML or EL. The LCLwas correlated positively and significantly with the ML, NM, and NI, although the correlationcoefficients were small (<c26). Thbis indicates that the initiation of level changing and mounting may not be independent processes. Consistent with' Sachs' (1978) and Dewsbury's (1979) analyses, the ML and IL were highly positively correlated. Thbis was expected because highlytrained rats usually gain vaginal intromission during their first or second mount. Sachs and Dewsbury reported that the EL ~was positively correlated with the III. the PET, and the NM. Those three correlations were replicated, although the coefficients were generally smaller in our sample. Consistent with Dewsbury's analysis, the EL was also negatively corre- lated with the NE and the IR. The first of these was expected, because the number of ejacula- tions possible in a fixed time limit decreases as the latency to each ejaculation increases. However, the negative correlation between the EL and the IR requires a more extensive expla-nation. At first glance, this correlation would seem to indicate that the relative number of intro- missions predicts the latency to ejaculate. However, the EL was also positively correlated withthe NI in our study. This correlation, together with the positive correlation between the EL andthe NM, suggests that a relatively larger increase in mounts over intrornissions leads to the TABLE I. MEANS, STANDARD ERRORS OF THE MEANS (SEM), AND CORRELATIONS AMONG THE 11 MEASURES OF SEXUAL BEHAVIOR DISPLAYED BY MALE RATS IN THE ElILEVEL CHAMBERS LC LCL ML IL EL PEI NM NI NE IR III Mean 11.21 18.00 13.80 17.31 286.66 270.04 4.22 10.75 2.47 0.74 26.99 SEM 0.35 1.46 1.64 1.96 14.09 5.87 0.32 0.37 0.07 0.01 1.17 Correlations LC - LCL -. 18 ML -. 04 +.25 EL -. 01 +.17 +.87 - EL -. 06 +.11 -. 05 -. 07 - PEI *.12 -. 06 +.05 +.01 +.43 - NM -. 10 +.22 4.01 +.04 +.42 +.27 -- NI -. 03 .. 25 -. 13 -. 12 +.54 +.II +.46 - ME -. 10 -. 02 -. 15 -. 16 -. 50 -. 28 -. 24 -. 35 - IR ..15 -. 19 -. 15 -. 20 -. 26 -. 26 -. 84 -. 06 +.12 - III -. 08 -. 01 4.10 +.07 +.76 +.44 +.23 -. 09 -. 29 -. 32 LC (number of level changes): LCL (latency to first level change): ML (latency to first mount): IL (latency to first utntomlsslon): EL (timne from the first intromission to the first ejaculation); PEt (time from the first ejaculation to the next Intromission): NM (number or mounts): NI (number of tntromlasions); NE (total number of ejaculations): IR (Intrornission ratio, defined as the number or mounts with Intrornisslon divided by the total number of mounts with and without tntromnlssion): and IlI (Inerintromlssion Interval.defined as the ejaculation latency divided by the number of Intromnissions). All latencies and intervals are in seconds. 'Me NM. NI. MR nd III are calculated prior to the first ejaculation. 333
Similar documents
View more...
Related Search
We Need Your Support
Thank you for visiting our website and your interest in our free products and services. We are nonprofit website to share and download documents. To the running of this website, we need your help to support us.

Thanks to everyone for your continued support.

No, Thanks