> ||qh|l ’39.... “VI; ' H >ul“"" '.‘.','f"l1.‘,l‘(q... J. l .. . :’I -i' luff} rant.- u a 11‘- V :3:ch .1. ‘Ezwn‘ ‘ ' Al 1' 6.2 ‘!‘. “u".‘L‘J. “.1 ~ LL 3: 9‘ H"? '1 ”1313‘“? l" “‘1” UL A“? . V 1“1:31;,1‘rvufiL- ‘ I1 14Mfi‘fl‘i‘kéi Vi ' ur'v u ’ L' w * - n ‘ mx‘?‘:."'.a ""“E ‘w‘rli ‘ ‘ '3 nun-um ;:.t;mj‘ 1 .- . ~ . ”fin-r ,3: - I 3+.“ {sag} 3:11;.“ ' ‘11n’IqMfin4 m A . mm: “a: ‘4. an ‘x’w .' ht. {gab V , ‘ , . n' ‘. V _ . 'VV , ‘i:71"- % ~ 'LM “3:!“ "5.“13‘3‘1 a". '79-’1'1v‘fi‘ $141.7” , ‘ M31 K! A 151.sz MW {1‘ a2“. Wag—7n!)- . V ,~ gal wk‘dsag H'm-JQ‘N 31‘:er 1‘ V d , V 1 ' ’ ”r: 1:7th ' 23.5““ aw" J: 5- .E’JJRI. ‘ J. .p‘. 'o ‘ ~ .1 ‘ x x @3333; ‘ 1" ‘6 3.3; V‘ ‘ A CMfigL 2: KER-m :31: 9:150 n l ‘ l L "11' ‘:=~'~ . "nigh A. 3‘31 '1}... a”:- ‘1’ka fix? ”if. 1; :15. 4&1”). 1: ‘ ' m ' I ~ 3%”, dig?» 1.445. p s‘,,- m,‘ V .17, 7,.V Hid." figmégg‘fiéfifidfi “:3. “i‘n " 1.. ~ ' “3)! 7 ‘0‘ "hi-i .. £3.33... 1 ~ ‘ , 41:. v ‘1 P 3?}; a . . k n '3 ‘3 '1’ .- 7.1' ‘Vm’w‘rrflii‘a"? g '1 ‘ w. .1. A: 1‘23” AL ‘3‘ . 3‘" 3.‘ my L455 ,t r . . 1-. 23"}. V .. "*1. 'z M41411»! M2“ A. l «s.- L . Y It: r211 . ii“, . -.'~ 0-! 1E7: m: “ L11: 73“»..0191": I' “1.15%..” ‘- LEE? :1. i‘gfiw V W fi::: «WI-A! ‘ MK Aft?“ N‘- “ «a "- mt; .3: ru- ‘v‘fi " ~ -.- . - .~-- 0-:-.-n- —.-—. -.--r ‘owa .. . ls. 1, ‘ - _I E 40‘ Z O 2 30' 20‘ l0- Figure 2. Rainfall recorded at Farm—8, Changuinola, Bocas del Toro, Panama during the years 1970-1971. Dotted line indicates 45 year average. 2a’C« 27°c‘ T EMPERATURE 25° C4 MONTHLY 24°c« AVERAGE 17 Figure 3. Ch 24 t! c. .m 3 >1 3' > a: (D Z Average monthly temperature at Balboa Heights (solid line), Changuinola (dashed line), and Capira (circles). This statistic was obtained by averaging daily maxima and minima. 18 associations recognized by Holdridge and Budowski (1956) dependent on the height of the land above the water table and the salinity of the water. The mangrove association is found in the lowest areas where the water is more sa— line and consists of the red mangrove (Rhizophora mangle), the black mangrove (Avicennia marina), and the white man- grove (Langucularis racemosa). Nearly pure stands of orey (Campnospermia panamensis) which is tolerant of brackish water occurs extensively over the area just above sea lev— el in the orey association. On slightly higher ground where alluvial fresh water stands much of the time, there are large forests of the silica palm (Raphia taedegera) which form the silica palm association. Still farther inland is another swamp association, the cerillo—sangrillo association where cerillo (Symphonia globulifera) and san- grillo (Pterocarpus officinalis) are common in the upper story and the coquillo palm (Manicaria saccifera) is com— mon in the lower canopy. Mile—2 Station consists of a small wooden house used occasionally by Gorgas Memorial Laboratory as a field sta— tion and located on the west side of the railroad tracks. North and south of the station there are numerous fresh- water marshes which become more and more brackish as one approaches Almirante. The freshwater marshes are sur— rounded by dense impenetrable stands of Calathea, Helico- nia, Cecropia, and Piper among others. A portion of the silica palm association can be seen immediately across 19 the railroad tracks east of the station. This association then dominates the region between the railroad and the coast to the east. Behind the station to the west the ground slopes gradually upward and is covered with cerillo and sangrillo in the upperstory and the coquillo palm (Manicaria saccifera) in the understory. The upperstory trees are thinly distributed here allowing a dense under— growth including the coquillo palm and numerous vines and shrubs. Most trapping was done along the railroad tracks, and along the edges of the clearings and marshes in stands of bamboo and Heliconia. Unlike Santa Rita the vegetation in Almirante is lush and green the year around. Fauna of the Almirante study area Mammals trapped on the Almirante site and returned to Panama City are shown in Table 1. This fauna was limi— ted by both trap choice and trap size. Santa Rita study area Santa Rita is a small rural community located on the Pacific slope. This region lies between 50 and 200 meters above sea level and is characterized by low rolling hills cut by frequent seasonal streams emptying into the Rio Caimito which has water the year around. Table 1. 20 Mammals trapped on the Almirante and Santa Rita study areas. Numbers of individuals brought back to the laboratory for study are indicated in parenthesis. Almirante study area Marsupials Philander opossum (l) Didelphis marsupialis (10) Rodents Oryzomys caliginosus (30) Nectomys alfari (27 Sigmodon Eispidus (80) Rattus rattus (5) Proechimys semispinosus (176) Santa Rita study area Marsupials Caluromys derbianus (36) Marmosa robinsoni (10) Philander opossum (85) MetaChirus nudicaudatus (3) Didelphis marsupialis (175) Chironectes minimus (l) Rodents Sciurus granatensis (l) Sciurus variegatoides (1) Liomys adspersus (0) Nectomys alfari (2) Zygodontomys microtinus (24) Sigmodon Hispidus (127) Rattus rattus (6) Proechimys semispinosus (963) Diplomys labilis (377 * Scores were trapped but released immediately without be— ing tallied. 21 Climate of the Santa Rita study area No climatic data were available for Santa Rita; how— ever, there was good data for Balboa Heights, Canal Zone, 35 kilometers east and 5 kilometers north of Santa Rita and enough data from Arraijan 24 kilometers east and 5 kilometers north of Santa Rita and from Capira l6 kilome- ters south of Santa Rita to establish that the region sur— rounding Santa Rita has a similar rainfall pattern to Balboa Heights (Figure 4). Comparison of Figures 2 and 4 shows that the verano dry season (mid—December through mid-April) in the Santa Rita area is more severe than it is in the Almirante area. At the end of 1971 the ten year average rainfall for Balboa Heights was 181.6 centimeters per year compared to a 45 year average of 248.7 centime- ters per year for Farm—8, Changuinola. Temperatures for the Santa Rita area are shown in Figure 3 and varied from a low of 200 Celsius to a high of 340 Celsius while the monthly averages of the daily low temperature varied only 10 Celsius (22—23O Celsius) and the high only 20 Celsius (32—34O Celsius). Available data for Capira indicate a temperature regime almost iden- tical to Balboa Heights. The temperature varied from a low of 200 Celsius to a high of 330 Celsius during 1971 while the monthly averages of the daily low temperature varied 2O Celsius (20—22O Celsius) and the high 10 Celsius (31—33O Celsius). 22 100‘ 90* 80‘ ’E :5 70+ 3 E 60 Z < 1 “1 .50 >. i‘ h, 40* Z C) 2 130< 2(L IOT Figure 4. Rainfall recorded in the area of Santa Rita de la Chorrera during 1970—1971 (solid line) and 10 year means for Balboa Heights (dotted line). Squares and circles represent data from Capira and Arraijan, respectively. 23 Other climatic data from Balboa Heights show that during the dry season there is more sunshine, the relative humidity drOps, and wind velocities increase. The reduced rainfall coupled with increased winds and sunshine inten— sifies the dry season in Santa Rita and causes the soil there to harden and crack. Vegetation of the Santa Rita study area Santa Rita is included in the Tropical Moist Forest of Holdridge and Budowski (1956:94) as is Almirante. In Santa Rita most of the forest has been removed and only small patches of the original forest remain on some hilltops and along a few streams and rivers. The area is covered instead, by numerous small farms where citrus, avocados, mangos, bananas, pineapples, rice, corn, yucca, name, and sugercane are cultivated. Some of the more prosperous farms have a few dairy or beef cattle. The most conspicuous trees on the farms are citrus trees, the corozo palm, maintained for thatch, and the mango tree. Along the streams, the conspicuous vegetation includes espavé (Anacardium excelsum) and several species each of Ficus, Inga, Cecropia, Heliconia, Bursera, Piper, Luehea, Capparis, Chrysophyllum, Terminalia, B353, Amaran— 3223’ Carica, and Bambusa. Grass is common about pastures around dwellings, along fence rows, and surrounding cultivated fields. Although the annual rainfall in the study area is sufficient to maintain some of the species 24 of plants indicative of the Tropical Moist Forest, the months of January, February, and March are generally very dry. Much of the herbaceous vegetation dies including grasses and many of the trees drop their leaves. The area remains dry until the rains return in April. Although Holdridge and Budowski (1956) placed both the Almirante and Santa Rita areas in their Tropical Moist Forest, they recognized that the dry season is more marked on the Pacific side than on the Caribbean side. The se— verity of the dry season can be observed by examining the total rainfall for the months of February and March. The average for Cristobal (on the Caribbean coast of the Canal Zone) is 76mm, the average for Barro Colorado (in the Canal) is 90mm, Capira registered 76mm, but Arraijan registered 18mm and Balboa Heights averages 32mm of rain- fall. Moist forests are found in Cristobal, Barro Colora— do, and Capira, and dry forests are found in Arrijan and Balboa Heights. During the same period Almirante averages 394 mm of rainfall. Although the area between Capira and Arraijan appears much different than the Almirante area in the dry season, they do have several tree species in common including Anacardium excelsum, Cecropia obtusi- folia, and Luehea seemannii. These are found along streams in the Santa Rita area. The dry season is shorter and less severe between Capira and Arraijan because here the elevated ”backbone" of the Isthmus of Panama is low allow— ing for some Caribbean rains to reach this part of the 25 Pacific coast of Panama. To the west of Capira and to the east of Arraijan, however, the dry season is much longer and more severe, and many trees typical of the Moist Tropical Forest such as Anacardium excelsum cannot survive there. Fauna of the Santa Rita study area Mammals trapped on the Santa Rita study site and returned to Panama City are shown in Table 1. Like the Almirante site this reported fauna was limited by both trap size and trap bait choice. 26 METHODS AND MATERIALS Initial effort was made to find a site on the wet Caribbean side of the Isthmus of Panama and one on the dry Pacific side where Proechimys occurred in large num— bers. Wet site Almirante (Figure 1) was chosen as the wet site be— cause of the abundance of Proechimys and because this was the home of a reliable collector that had worked part time for Gorgas Memorial Laboratory. Small mammals were col- lected intermittently from late 1970 to early 1972 and sent via railroad to Changuinola and then by air to Panama City. Dry site Attempts to find large populations of Proechimys near Panama City failed. In trapping near Tocumen, 20 kilome— ters northeast of Panama City during August and September of 1970 and near Juan Diaz 11 kilometers northeast of Pana— ma City during October 1970, I found only small popula- tions of Proechimys. Similarly, trapping at Juan Mina 36 kilometers north and 16 kilometers west of Panama City during September and October of 1970 revealed only a small 27 population of Proechimys. During September 1970 a part—time animal collector for some of the staff members at Gorgas Memorial Labora- tory reported large numbers of Proechimys near Santa Rita (Figure 1). Santa Rita is 55 kilometers from Panama City by automobile and can be reached the year around since 40 kilometers of the way is an all—weather highway and the last 15 kilometers is accessible by a gravel road which proved to be passable even in the wet season. An animal collector who lived on a small finca one kilometer west of Santa Rita was hired to obtain small mammals in the area. Animals were trapped from late 1970 to early 1972 in 15x15x18 and 23x23x70 centimeter folding live traps baited with banana. Traps were set within walking distance of the finca and transported using a home-made cart. After being carted to the finca, mammals were placed in l4xl8x26 centimeter cages hand-made out of one—half inch hardware cloth and sheet metal. These were kept in a spacious thatched shelter open on four sides and made specially for this purpose. Rice hulls were used as bed— ding and animals were given Wayne Lab—Blox(:)and water Ed libitum. Their diet was supplemented with oranges, yucca, name, and bananas. These animals were picked up and transported to Panama City once each week. 28 Field data After arriving in Panama City animals were taken to the Gorgas Memorial Laboratory annex located behind the United States Embassy on Calle 37. Females were palpated for pregnancy and held until parturition if pregnant. The remaining females were anes— thetized with ether to relax their abdominal muscles and palpated again. Nonpregnant females were euthanized by chloroform and autopsied. Some females and males were selected for breeding in the laboratory. Later in the course of the study as laboratory space declined, even obviously pregnant females were euthanized and autopsied. The length of one inguinal nipple, one lateral nip— ple, and the length and width of the clitoris were record- ed to the nearest millimeter. The mammary tissue was ex- amined and recorded as not developed or well-developed. The vaginal opening was recorded as closed, partly Open, or open. If the vaginal area was swollen and dry or if swollen and wet, this was recorded. Among several fe- males, vaginal smears were made by sampling the vaginal cells using a physiological saline-soaked cotton swab. These smears were examined under a microscope and the cells present were recorded. Since vaginal smears made with saline often form fern patterns when dry and since it has been suggested that these ferns may indicate estro- gen levels in the female, these were photographed for la- ter study. 29 The reproductive tracts of females were fixed in AFA and cleared using the technique of Orsini (1962). After being cleared, embryos and recent scars were counted and their distribution in the horns of the uterus was record— ed. Males were routinely euthanized and autopsied unless held for breeding. They were identified as testis scro— tal, cauda epididymus scrotal, or nonscrotal (see discus— sion on the scrotum in Reproduction in field captured males). The testes were removed, the length of the left testis was measured to the nearest millimeter, and the testes were preserved by fixing them in 10 per cent forma- lin and later placed in 70 per cent alcohol. Testes were analyzed among field—caught males by first grouping the material into Almirante and Santa Rita data. These two groups were each divided into material collected October through January and material collected February through September. Within each of the resulting four groups the testes lengths and body lengths were trans— formed using natural logarithms. The length of the left testis was then regressed by simple linear regression on body length. Within each site—season group the testes measurements were adjusted to the mean body length by the expression ya =yi—bl(xi—x) where bl is the slope of the regression equition. The adjusted testes lengths are sum- marized for each of the four groups by means and 95% con— fidence intervals. 30 Dead animals were weighed to the nearest tenth of a gram on a triple—beam balance, and measured to the near- est millimeter. All measurements were made by the inves- tigator and taken on relaxed animals just after death. Measurements included total length, tail length, length of hindfoot, and height of ear from the notch. These are standard measurements used by mammal preparators and are described in Hall (1981). The age of field animals was estimated by comparing their body length (total length minus tail length) with the mean body length of known-age laboratory-reared ani- mals of the same sex and of stock originating from the same site. See Growth rate for equations relating age and body length among laboratory born spiny rats. The month of conception for each field-captured spiny rat was estimated by noting its date of capture, estimating its age using laboratory growth-data, and then summing its estimated age and 64 days gestation and subtracting this from its date of capture. Both males and females were examined for lesions on the extremities and if lesions were present, these animals were referred to investigators at Gorgas Memorial Labora- tory studying leishmaniasis. Laboratory data Animals held in the laboratory were placed in 14x18x26 centimeter stainless steel cages and placed in 31 a spacious room open and covered by a wire screen on the north and west (but shaded from the sun by a porch roof). Here animals were exposed to ambient temperature and natu— ral photoperiod. Chopped wood, rice hulls, or Absorb—Dry ®was used as bedding. Animals were fed Wayne Lab—Blox® and given water ad libitum. Their diet was supplemented with spinach greens, oranges, and bananas. Cages were cleaned by washing them with soap and water and replacing the litter once each week or earlier if soiled. In the laboratory Proechimys were generally kept l or 2 animals to a cage. Other mammals collected at Santa Rita or Almi— rante were also transported to Panama City and kept in the same room as Proechimys. All animals were exposed to similar temperatures dur— ing their confinement in the laboratory. The temperature extremes in Panama City are generally 10 Celsius lower and 10 Celsius higher in the dry season then they are in the wet seasons. The monthly mean percent relative humi— dity (based on the mean of the daily averages of the maxi— mum and minimum values) averages 11 percent lower in the dry season. All animals were handled as uniformly as pos— sible to keep variation at a minimum. Pregnant females held in the laboratory were checked for young three times daily, and new born were counted, sexed, and toe—clipped for identification. Young were weighed to the nearest tenth of a gram and measured to the nearest millimeter on days 0, 5, 10, every ten days 32 until 100 days of age and then every 20 days until 200 days of age. Length measurements included total length, body length, hindfoot length, and ear length as described previously. The progress of the post—juvenile molt was fully described on the days of measurement. Observations were made to determine when the testes of males descended and when the vaginal closure membrane of females was first perforate. This membrane is perforate only at estrus and parturition (Weir, 1973). The length of gestation was determined by selecting females with perforate vaginas and pairing them with adult males and placing them in 25x25x35 centimeter cages. Va- ginas were checked each day of the pairing for the pre- sence of spermatozoa by microscopic examination of vaginal smears made by swabbing vaginas with cotton swabs soaked in physiological saline. When spermatozoa were observed in the vaginal smears, males were separated from the fe- males. Corroborative evidence of length of gestation was obtained by noting the length of the time interval between successive litters when males were separated from the fe- male on the day of birth. Body length growth data from spiny rats reared in the laboratory were used to estimate the age, month of birth, and month of the conception (using 64 days as the length of gestation) of every field captured male and fe- male. Occasionally rats were bled to search for blood 33 parasites. Rats were supplied at various time for feeding mosquitos or provided to MARU (Middle American Research Unit) in the Canal Zone for virus studies and to the Virus Department at Gorgas Memorial Laboratory for similar work. Others were occasionally provided for the pathology de- partment at Gorgas to be sent to the United States. Because of the small dispersion of many of the mea— surements of length and weight, the standard errors are expressed to more decimal places than the number of signi— ficant figures in the means; this does not imply that these estimates are more precise than the original meas— urements. Many of the measurements considered in this study are discrete integers and can only be approximately normally distributed. Two—tailed tests are used through— out for all statistical analyses. All calculations were performed on a hand—held TI—59 calculator. Statistical tables provided in Steel and Torre (1960) were used to determine significance levels. LABORATORY RESULTS Gestation Period A total of 50 female Proechimys were paired with males and checked daily for sperm for a total of 1100 rat— days. Sperm was observed in the vaginas of 6 Almirante females and 5 Santa Rita females that produced litters of young in the laboratory. Santa Rita spiny rats were born an average of 64.2 days (63, 64, 64, 64, and 66) af— ter sperm was observed and Almirante young were born after 64.7 days (63, 64, 65, 65, and 66) of gestation. When males were left with females through the gesta— tion period and removed the day after parturition, peri— partum mating occurred and a second litter was produced after an average interval of 63.5 days (63 and 64) for Santa Rita females and 64.0 days (64, 64, 64, and 64) for Almirante females. Eighteen Santa Rita females and 11 Almirante females bred in the laboratory without being observed and gesta- tion could not be determined. Litter size Among pregnant spiny rats trapped near Santa Rita and brought to the laboratory the mean litter size was 34 35 3.0(SE=0.12; R=l—6; N=68) for those giving birth and the mean embryo count was 3.0(SE=0.09; R=l—7; N=126) for those autopsied. There was no significant difference among these means (£=0.01; p>0.05; d.f.=l,l92). Among pregnant spiny rats trapped near Almirante and brought to the la— boratory the mean litter size was 2.0(SE=0.21; R=l—5; N: 24) for those giving birth and the mean embryo count was 2.1(SE=0.15; R=l—4; N222) for those autopsied. There was no significant difference among these two means F=0.27; p>0.05; d.f.=l,44). Among pregnant spiny rats trapped in Santa Rita and Almirante, litter size and the number of embryos were larger for Santa Rita (F=l9.34; p<0.005; d.f.=l,90; and F=l3.37; p<0.005; d.f.=l,l46 for litter size and embryo count, respectively). However, litter size in both populations is significantly related to ma- ternal weight after parturition (F=8.32; p<0.01; d.f.=l,26 and F=7.69; p<0.025; d.f.=l,l8 for Santa Rita and Almiran- te respectively), and embryo count is significantly rela— ted to maternal body length at autopsy for Santa Rita (E: 10.26; p<0.005; d.f.=l,123) but not Almirante (Fz4.05; p>0.05; d.f.=l,23). Furthermore Santa Rita mothers were heavier than Almirante mothers at parturition (F=24.88; p<0.005; d.f.=l,46), and Santa Rita mothers were larger than Almirante mothers in body length at autopsy £28.00; p<0.005; d.f.=l,l48). Covariance analysis indicated that the litter size differences between Almirante and Santa Rita merely reflected the differences in maternal body 36 weight (5:0.14; p>0.05; d.f.=l,45). The same analysis of the number of embryos found at autopsy, however, showed a significant difference in embryo count independent of maternal body length (§=l3.97; p<0.005; d.f.=l,l47). Growth and development Ninety-three female spiny rats were pregnant when captured and gave birth to litters in the laboratory. Another 37 females conceived in the laboratory and gave birth to live litters. Size at birth There were no significant sexual differences in the average individual body weights of newborn from Santa Rita F=0.54; p>0.05; d.f.=l,86 litters) or Almirante (§=l.43; p>0.05; d.f.=l,27 litters). Average individual body weights of Santa Rita newborn were significantly negative— ly correlated with litter size (£23.4z—0.46; E=-2.53; p<0.02; n—k=28 litters) but not significantly correlated with maternal weight (£24.3=—0.006; £=—0.03; p>0.05; n-k= 28 litters). Average individual body weights of Almirante newborn were significantly negatively correlated with lit- ter size (£23.4=—0.53; 5:2.76; p<0.02; n—k=20 litters) but not significantly correlated with maternal weight ( —0.07; £=—0.32; p>0.05; n-k=20 litters). Covari— 524.3: ance analysis of mean individual body weight of newborn versus litter size showed that newborn from Santa Rita 37 were significantly lighter in weight than those from Almi— rante (5:4.93; p<0.03; d.f.=l,94 litters) independent of litter size. Growth rate Growth in body weight between birth and twenty days of age was not significantly correlated with weight at birth, litter size, or maternal weight (Table 2). There were no significant sexual differences in body weight growth between ages 0 and 20 days for either Santa Rita (F=2.32; p>0.05; d.f.=l,86 litters) or Almirante (F=0.ll; p>0.05; d.f.=l,27 litters). There were no site differ— ences in body weight between ages 0 and 20 days for either males (£21.64; p>0.05; d.f.=l,55 litters) or females (F=0.55; p>0.05; d.f.=l,59 litters). Analysis of growth by linear regression of 1n(body length) on 1n(age) resulted in the following equations: Santa Rita 0.94 0.97 Males y=4.408(0.008)+0.204(0.002)x r Females y=4.407(0.019)+0.195(0.002)x r Almirante Males y=4.47l(0.013)+0.197(0.003)x r2=0.95 Females y=4.469(0.011)+0.l84(0.003)x r2=0.95 Growth rates, estimated by the slopes of the above regression equations, were significantly greater for males for both Santa Rita (5:3.51; 240.001; d.f.=l49l) and Almi— rante (£=2.82; p<0.01; d.f.=422). Males were sugufianmly 38 m No.HI mq.0| eq.01 oH.OI NH wN.o oH.o NH 0N.o oN.o m NN.H wq.o Nq.o mH.o NH HH.HI Om.o| NH qH.O qo.o m mo.o Ho.o mm.o MH.o NH mq.H mm.o NH oo.H oN.o XIC u H u 9 xi: u H xi: u u monEom monz monEom monz oucmHHEH< wuHm mucmm .uLwHoB Hocuoume Adv one .oNHm HmuuHH Amv “LuHHn um uanoB zoos ANV .mhmp ON pan 0 mowm Cmozuon uanoB xoon CH SDBOHw AHV How muCoHonwooo conmonoh HmHuwmm MN.aHu sm.maa sm.NHa .N mHnme 39 larger than females by day 30 in Santa Rita (F=8.83; B<0.005; d.f.=l,96 litters) and significantly larger by day 80 in Almirante(§=12.04; p<0.005; d.f.=l,25 litters). Growth rates, estimated by the slopes of the above regres- sion equations, showed that females from Santa Rita grew significantly faster than Almirante females (£=3.29; p< 0.01; d.f. 966), but no significant differences in body length were found between Almirante and Santa Rita females for any age. There was a significant difference in growth rate between Santa Rita and Almirante males (3:2.01; p<0.05; d.f.=875), and Almirante males were significantly larger (e.g., F=5.86; p<0.025; d.f.=l,60 litters on day 5 and §=9.01; p<0.005; d.f.=l,50 litters on day 100). Molt Molt comparisons between males and females from Santa Rita and Almirante are shown in Table 3. The molt began earlier among females than among males, but there was no discernible difference between sites. Sexual maturity The earliest vaginal perforations observed among Al— mirante and Santa Rita females born and reared in the la— boratory were 60 days and 70 days of age, respectively. The percentages of females found perforate for the first time on each day of observation are shown in Table 4. A significantly greater proportion of Almirante females 40 .owma uxoc Go UoDcHucoo $m No .Louma Hmmuoo oz Rm Nd .mummadw umsm Loumd Hmmhoa NmN Nos qu .souaa Hamaoe .50 «can on m>aa Noe NAN Ham Hem .mmenm co soc mmema unse< Nmm New xw fiqH .mxcmHm mH Honoumom use HHm co omeod uH3o< No NN .mmema oase< co Hmw HooH Hwa New .nouma mehoe HHQEm m mH oposu use mumo ou BHoz ow Nan Hwa NMN Hm .mmam on no: use: Nwm NAN Nmm .mmNm m>onm uaoz NNN Hon Nam .masm awesome uHoz ON New fiHN Noe .mumo ocm moxo osu panopm mozoumd mum mumzu one EDuumou mo aHu co umdm uHoz Nmm Ham Nos .mmsm and on so Nashua: EDpumou a: commonOHa mm: uHoz co Nam HooH .csmmn was SHoz cm fim fimn .cswon mm: uHoz oq monE monEom monE monEom muemuHeHa mocmuHeH< munm macaw «one macaw owmucmouom uHoe mo owmum ow< .mhmp CH oohsmmoe mH ow< .oucmuHEH< pom wuHm modem 50pm monk xcHam onEow paw onE How UHOE oHHco>Dqum0d wo mums osu mo mCOmHHmQEoo .m oHan 41 NH wH mm Nw po>pomno mHmEHcm No Honesz ma mm mON qu mcoHum>pomno wo Honesz $0 NN .moon udooxo owwHod UHDU< NNH NON NMN $9 .moon mo uoHpoumoa udooxo omeoQ UHDU< wa New Nmo Nmo .mmema SHse< QNH NNN Nmm NHN NmH .souma Hmmuoe Eu made On m>aa emu New Now Has .mwmn no moUHm udooxo omeod uH3o< NNH NMN fie .moon wo ponoumOQ udooxo omeod uHDo< fie $NH .mon xomn odoUXo omeod uH3p< xNH $N XoH .omeod uH3U< ooH monE monEow monE monEom oucmeEH< oucmuHEH< muHm mucmm muam modem owmucoopom OHOE mo oumum ow< .ie.ucooc m magma. 42 Table 4. Percentage of laboratory reared female spiny rats with perforate vaginas seen each observation day. . . 1 Age Alm1rante Santa Rita gadj. p 60 33% (1/3) 0% (0/23) 0.96 >0.05 70 27% (3/11) 6% (1/18) 1.14 >0.05 80 67% (4/6) 0% (0/9) 5.38 ((0.025 90 40% (2/5) 7% (3/44) 1.76 >0.05 100 70% (7/10) 13% (6/45) 10.04 ‘<0.005 110 100% (1/1) 33% (3/9) 0.04 >0.05 120 78% (7/9) 34% (15/44) 4.18 (0.05 l. The 2x2 test of independence using the G—statistic with Yate's correction (Sokal and Rohlf, 1969:591). 43 were perforate on three of these observation days. By the age of 120 days, a significantly greater prOportion of Almirante females (ll/l8) than Santa Rita females (10/35) had been observed perforate on some previous ob— servation day (Eadj.=3'96; p<0.05; d.f.=1). The mean day of perforation could not be determined because observa- tions were not made after 120 days and many Santa Rita females were still imperforate at that age (Table 4). Field data (Table 5) indicated that females from both populations became reproductively active (perforate, preg- nant, or with placental scars) at approximately the same size. In the laboratory, 7 pairs of laboratory reared spiny rats from Almirante and 11 pairs of laboratory reared ani- mals from Santa Rita bred successfully. The earliest mating among Santa Rita rats was 120 days of age when two siblings bred successfully. One Almirante female bred at 120 days of age and one male bred at 130 days after the two were paired at 100 days of age. Testes descended earlier among Santa Rita males (Table 6). Growth and development New born Proechimys are covered by a thick pelage. The head is dark but the back is gray because of white hairs (about 4 millimeters long mixed in with the short 44 Table 5. Fraction of field captured females reproductively active (perforate, pregnant, or with placental scars) on arrival at the laboratory. Body length Almirante Santa Rita gadj.l p 180—189 mm 0/3 0/13 190—199 mm 1/2 2/11 0.005 >0.05 200—209 mm 3/4 5/16 1.04 >0.05 210—219 mm 3/3 13/23 0.76 >0.05 220-229 mm 5/5 9/16 1.87 >0.05 230+ mm 22/27 171/202 0.02 >0.05 l. The 2x2 test of independence using the G—statistic with correction (Sokal and Rohlf, 1969:591). 45 Table 6. Percentage of laboratory born males with descended testes at four ages. . . 1 Age Santa Rita Alm1rante gadj. p 70 4% (2/45) 0% (0/14) —0.02 >0.05 80 20% (8/39) 7% (1/14) 0.58 >0.05 90 56% (26/46) 12% (1/8) 3.91 (0.05 100 75% (33/44) 89% (8/9) 0.23 >0.05 l. The 2x2 test of independence using _ with Yates' correction (Sokal and Rohlf, 1969:591). the G—statistic 46 dark hairs (about 1 millimeter in length). The ventral surface is pink and covered with short white hairs with a few longer ones (about 4 millimeters long). Vibrissae (25 to 30 millimeters long) are conspicuous on the rostrum and above the eyes (15 to 22 millimeters long). Within minutes of birth the ear pinnae unfold revealing the already perforate auditory meatus. The eyes open soon after birth, but occasionally one or both eyes will remain closed for a few hours. The tips of the toenails are always white at birth and this marking persists for a day and was used occasionally to time births not observed directly. The incisors have usually erupted before birth but in two instances were still covered by a thin membrane at birth. Although no cheek teeth are visible at birth spiny rats will nibble at solid food within hours of birth. Molt Juvenile Proechimys semispinosus have a brown pelage of thin aristiforms (guard hairs) and setiforms (underfur). The post-juvenile molt begins as agouti—colored hairs ap- pear on the anterior rostrum, below the eyes and ears. The agouti coloration moves up the rostrum as the patches expand. The back is still covered with juvenile pelage with soft aristiforms raised above the surface, but if 47 the skin of the back is examined by parting the fur, short black hairs can be seen just erupting. The agouti colora- tion continues to expand on the head as agouti hairs ap— pear above the eyes and in front of the ears. At this stage of the molt the adult pelage has advanced up the ros- trum until it is almost to the eyes. The agouti-colored pelage continues moving up the rostrum until it is between or even above the eyes to the ears. Next, the agouti—col- ored setiforms and flattened dark-colored aristiforms ap— pear on the back forming a conspicuously dark shiny oval patch. The agouti—colored adult pelage continues up the head until it is between the ears. The dorsal patch ex— pands, moving to the head and down over the shoulders to the front legs and down and back over the sides until fin- nally the rump, hips, and hind legs are covered by the agouti-colored adult pelage. Sex ratio Laboratory born spiny rats from Almirante and Santa Rita exhibited a 50:50 sex ratio (Table 7). Lactation All females observed but one had three pairs of nip— ples, one inguinal pair cephalad from the prominent clito- ris, and two lateral pairs just above the midline, one immediately in front of the femur and one at mid—thorax. 48 Table 7. The number of male and female spiny rats captured in the field and born in the laboratory (includ— ing both field and laboratory conceptions). Males Females X p Spiny rats trapped near Almirante 92 82 0.58 >0.05 Spiny rats trapped near Santa Rita 471 474 0.01 >0.05 Almirante births in the laboratory 37 35 0.06 70.05 Santa Rita births in the laboratory 109 113 0.07 70.05 Total 713 718 0.02 70.05 49 One laboratory—born female had the 6 nipples described above plus a pair on the abdomen 4.5 centimeters cephalad of the inguinal pair. During the period of lactation the nipples are swol— len and elongated and the mammary tissue is thick and con— spicuous. It is especially conspicuous around the ingui— nal nipples where the hair is thin and where it forms a thick l.4x4.5 centimeter patch extending from the clitoris forward and laterally under the nipples. The mammary tis— sue around the lateral nipples is well—developed but less conspicuous because of the thick hair on the thorax. The areas immediately surrounding the lateral nipples are de— void of hair during the period of lactation when the nip- ples are erect and prominent. Young Proechimys were observed eating soft fruit on the day of birth and one youngster gained 1% of its weight on day 9 when removed from its mother for a day. When left with the mother for longer periods young will spend a great deal of time suckling for the first three weeks. If the young are left with the mother longer, the inguinal mammary tissue recedes but the young will continue suck- ling on the lateral nipples. Well—developed lateral mam- mary tissue was observed as late as 46 days after birth and young were observed suckling as late as day 51. Most young have quit suckling by day 50 and the nipples have receded in size. Some young, if left confined with their 50 mother, continue suckling even though the mammary tissue is no longer conspicuous. In such situations the nipple is still enlarged and was observed as late as day 66. Young spiny rats were observed suckling longer in this study than has been reported elsewhere. Enders (1935) reported lactation lasting 46 days in one case and Maliniak and Eisenberg (1971) noted that young are nursed until 40 days of age in captivity. It is not known if lactation lasts as long in the field, but Enders (1935) estimated by the size of trapped young that they stay with their mother until 2 to 2% months of age. Young spiny rats may be able to find food long before this age since in this study young did eat solid food at an early age. Other investigators have re- ported similar observations. Enders (1935) reported young eating solid food by the age of 11 days and Maliniak and Eisenberg (1971) remarked that young begin to eat solid food almost from the first day. Since most male spiny rats were measured on arrival, estimates can be made of their ages (Table 8). These es— timates suggest than some young are moving about at an early age. 51 Table 8. Frequency distribution of age estimates of male spiny rats trapped in Almirante and Santa Rita. Estimated Frequency of Frequency of age(days) Almirante males Santa Rita males 10—20 1 2 20—30 4 3 30-40 1 6 40—50 3 6 50—60 5 5 60-70 0 2 70-80 4 12 80—90 2 6 90—100 2 9 100-120 8 21 120—140 6 26 140-160 2 11 160-180 6 4 180—200 6 14 N 0‘ 200+ 173 FIELD RESULTS Reproduction in field captured males Proechimys like other caviomorphs do not have a true scrotum (Weir, 1974; Pocock, 1922), but the testes will descend from the inguinal canal until the caudae epididymi or even the testes themselves are just beneath the skin. The position of the testes was labile in field—captured males and was not used for determining the fertility of field-trapped males. When handled, most field captured males pulled the testes into the inguinal canals with only portions of the caudae epididymi exposed. Males that were held in the labortory longer were more likely to have external testes when handled. The testes of laboratory born males with descended testes remained external when examined. The testes of most male spiny rats trapped in Santa Rita were significantly smaller during the months of Octo— ber, November, December and January (Figure 5). In Santa Rita this period of four months juxtaposes the two wettest months with the first two months of the dry season. There was insufficient evidence to Show any seasonal change in the testis size of Almirante males (Table 9). 52 53 .mNHm oHaEmm Ucm newsmu n.H.o Nmm mowsHocH .pomeH Mo LuwcoH mpon CH EEONN monE wUHm wucmm mo mHumou umoH mo HEEV LawCoH cam: .m opstm J . H ova >02huoo amm m3< Haw cSh aw: Mm< he: Lou CMW .fi mm EEOHI «H z .3 EEONL MH 4 m i 54 .ponEoumomlxpw3Hnom wcsto memE oucmeEH< mo omocu Cmnu nowme >HucmoHMchHm ohm monE muHm mucmm Mo mnuwcoH moumweke .%MMDthIHonouoo onto wonEoudom prmdhnom wcHHDU HomeH %HquoHMchHm ohm monE muHm mucmm mo wLOMCoH woumoH « AHVNM AchN iwvom.oaww.am iwvow.muma.fim +oom Amcmm.muoo.om AHva ANNVom.HHom.am AaVNN.Naow.Hm omN-owN AmVNN.oHoo.wN Amvmq.HHmm.Hm iANNqu.HHmm.sm AvaNm.Hflmq.oN aNN-oNN «aiocmH.Naom.NN AHVMN «Amqvmo.Hamo.Nm ASNVNH.HHom.wN sowuooN aaflocoo.NHom.mN Amvom.NHow.mN iimquH.Hwa.Om AmHvaw.Haow.oN amNIOAN Aavwm.NHqH.©N Accao.m+om.mm «AQNVOH.HHH©.0N Amqum.HHow.qN asmloqm «aiovmw.mhmw.mw Amvmo.qfloo.qN aANHVmS.HHMN.NN Amwcmm.Huow.HN ammlOMN ANVOm.wH Amado.oHHmm.0H AOHVSH.NHNS.HN AOHVNH.mHoH.wH ONN-ONN Amvso.NHoN.OH AHVaH Aquou.quwN.om lovem.aamm.oH oHNIOHN Amva.onm.HH AHVm ANVNm.meN.MH Amvom.NHow.NH oomloom dominom thluoo dominom cmhluoo LOMCoH xoom oucmuHEH< mUHm mucmm .mHmosu Icowmd CH oNHm deEmm LuHB whouoEHHHHE EH Go>Hw ohm mucoEopsmon .mumu xcHam onE Ummamhu UHon wo moumou ummH mLu wo .H.o Nmo H sowCoH cmoz .o oHan 55 Testis size increases with body length. The natural logarithm of testis size was significantly correlated with the natural logarithm of body length for Almirante and Santa Rita males for the period of October through January (5:10.93, p<0.05, d.f.=49; £=18.26, p<0.05, d.f.=210) and February through September (5:6.04, p<0.05, d.f.=25; 5:16.14, p<0.05, d.f.=l43) respectively. When testes lengths were adjusted to the mean body length for each of the four site—season combinations (see Methods), the mean testes lengths, measured in millimeters, for the periods of October through January and February through September were 24.97:0.63 (n=56) and 29.56:0.65 (n=64) respectively for Santa Rita and 23.66i1.64 (n=20) and 23.98il.00 (n=29) respectively for Almirante. There— fore even after testes lengths were adjusted to remove the effects of body size this analysis supported the argu- ment that the testes of Santa Rita males captured during February through September were significantly larger than those of males captured during October through January. No seasonal change in testis size was observed in Almiran— te males after testis size was adjusted to body size. This analysis also revealed that the testes of Santa Rita males captured during February through September were larger than those of Almirante males regardless of season of capture even though Almirante males were larger than Santa Rita males. 56 Analysis of a small sample of testes taken from known age laboratory born males supported the argument that the testes of Santa Rita males were larger than those of Almi— rante males (Figure 6). Maximum testis length and the body length when testis growth rate stops increasing and begins decreasing can be estimated by fitting the data of Table 10 to the logis— tic equation y=a/l+eb+cx. The mean testis lengths were fitted by nonlinear least squares to the midpoint of the body length interval using Marquardt's algorithm as de- scribed by Conway gg g1. (1970). The three partial deri— vatives of the above logistic equation described by Conway gg g1. (1970) are incorrect, however, and should read: Y/a=l/l+eb+CX Y/b=—aeb+CX/(l+eb+cx)2 Y/c=—axeb+CX/(l+eb+cx)2 Calculations were performed on a TI—59 hand—held calcula— tor using a program written by the author. The three re— sulting logistics have the asymptopes (maximum testes sizes) 40.1 millimeters, 34.5 millimeters, and 33.8 milli— meters respectively for males captured in Santa Rita dur— ing February through September, during October through January, and in Almirante in both seasons. This reflects the previous conclusion that testes are larger for males captured in Santa Rita during February through September but that the testes of males captured in Santa Rita 57 30 _ D Almirante 0 Santa Rita 25 - 0 (225mm) E E .5 go 20 ’ 0 (234mm) [:1 o (240mm) v-l (I) '3 0 (214mm) 8 D (238mm) H 15 . 0 (175mm) 40 70 100 130 Age (days) Figure 6. Testis length of young known age spiny rats from Santa Rita and Almirante. Body lengths are in parenthesis. q Hoo.HVo.Nm N Hovo.0m «ONImmN mH HNo.ovN.mm o HwH.va.Hm o Awo.HVN.om qulmwN MN ANq.oVN.qM «H HNo.oVH.om m Aom.va.NN qulmNN om Hom.oVH.qm Nm Hoo.ovm.oN w ANm.HVo.wN qNNImoN Hq Amq.ovo.Hm wH HHN.ovN.NN HH HqN.ovN.oN qulmmN am Hom.ovo.0m NH HNm.HVN.mN oH Hmo.ovo.mN «lequ oH Ado.ovH.wN @H ANm.ovm.NN N HNo.va.MN quImMN oH Adm.ovq.mN qH ANN.HVN.ON MH Amw.ovq.HN QmNImNN 8 «H HNo.HVo.mH OH AOH.va.wH m Haw.ovo.oH qNNImHN 5 HH Hem.qu.oH o HHN.ovm.qH N ANVMH qHNImON N HHH.qu.HH m Awm.ovo.MH q AoH.va.oH quImoH m Hoq.ovo.m N AHVo o qulme m AHm.ovo.w q AmN.oVN.¢ quImNH m Hmm.ovm.N H o N Hom.va N «NHImoH m Hovo.N H m «OHImmH m Hovo.N N AHVN quIqu c Ammvm c Hmmvm : Ammvm Hm>pmucH Illllllllllllll IIIIIIIIIlIIIIII LuwcoH HonEouaomINHmsunom NpmscwhlpoQOuoo MonEmooQINpmscmw Npom SE Scam anemia: .oquHHEH< one wuHm wucmm CH powdudmo msmocHdeEom mNEHLooopm wo moumou umoH mo LuwCoH new: .oH oHan 59 during October through January are similar to those of males trapped in Almirante in both seasons. The body lengths at which inflection (body length when the rate of testes growth begins decreasing) occurs was determined by fitting the parameters obtained by solv— ing the logistic for the second derivative: 2 b+cx(l+eb+cx_2eb+cx) y”=-ac e ——F—7___ (1+e +cx) This analysis revealed that the points of inflection were remarkably similar, 219 millimeters, 217 millimeters, and 217 millimeters for Santa Rita during February through September, Santa Rita during October through January, and Almirante during the whole year. In order to better understand the meaning of changes in testis size with season, histological sections of the testes from animals captured both in the dry season and the wet season will have to be studied. Some information is provided by the availability of testis sections of ten adults caught in Santa Rita during October and November. Among these, one with an adjusted testis length of 30 millimeters had enlarged seminiferous tubules with many cell layers, few interstitial spaces, and many spermatozoa. The remaining nine testes 25 millimeters in length or less when adjusted to mean body length had small seminiferous tubules with few cell layers, large interstitial spaces, 60 and few to no spermatozoa. If the fluctuations in testis size do reflect fertil— ity among males, then one would expect to find more preg- nant females during the months of February through Septem- ber in Santa Rita than in the months of October through January but no seasonal difference in Almirante. We will see in the next section that the reproductive activity of female spiny rats in Santa Rita and Almirante sup- ported this proposition. Reproduction in field captured females During the months of October, November, December, and January only 26% (34/131) of females (estimated by body length to be 120 days of age or older) trapped near Santa Rita were pregnant on arrival to the laboratory (Table 11). During the remainder of the year 89% (188/ 211) were pregnant. This difference is very highly signi- ficant [p(§ l47.9)<0.005, n=l). Juveniles (estimated by body length to be less than 120 days of age) and young adults (estimated by body length to be between 120 and 200 days of age) were more commonly trapped in the months of October through January. This suggested that the sea— sonality in the proportion of pregnant females found in the field was merely a result of a seasonal change in the age structure of the population. Although age estimates for females that gave birth to live or still born young 61 NaN wN HH m d N ponemooo NmH mm o m H N HonEo>oz NNm oH m m N N MoDOOoO Nam N NH o N CH nonEoudom .NNo a w o H N umswsa. NOS o S H w mm 33. NNa N «N H N 0H ocsm Nam a 3 q o S .32 New N mm H E 3 ENE NOOH o mN m a HH Lopez NNw q wH m o @ hhmsunom .Nmm 2 m N m m GEES ow>wwmno LupHn o: uzn Hm>HHHm Hm>HHhm Hm>HHHm woumaHma wcsoz wcsox kmmounm co co Co No CMonHHHum o>HH ou um ucwcmopm ucmcwoua pamcwoha ou LuHHn o>mw Luth ucmcwopd unoohom uoz HmDOH umcu HonEDz o>mo HoQEDZ Lucoz .MoUHo Ho mxww ONH o uwCo %Uon % UoDMEHumo one wuHm mqum New: popsuamo mzmocH mHEom m EHnomopm onEow mo COHqucoo o>Huospowdop mHnucoe mo Nuweesm .HH mHan 62 were not available, they were available for all other fe— males including pregnant and nonpregnant. Analysis of these data revealed the same seasonality in the proportion of pregnant females for all age classes (Table 12). Females from Almirante appear to be reproductively active the year around (Table 13). From October through January 93% (13/14) were pregnant on arrival and during the remainder of the year 92% (33/36) were pregnant. Among 16 juvenile sized individuals captured near Almi— rante, 4 were pregnant when captured in January, February, May and June. Estimates of the month of conception (see Methods and materials) for each field trapped spiny rat are summa- rized in Figures 7 and 8. The large sample of spiny rats from Santa Rita provided the opportunity to analyze males and females separately. This allowed two independent esti— mates of the months of conception for field caught spiny rats from Santa Rita. The monthly pattern for 420 Santa Rita males was very similar to that for 378 Santa Rita fe- males (Figure 7). Most conceptions were estimated to have occurred in May, June, and July and the fewest occurred in the months of October, November, December, and January. The observation that males and females analyzed sepa- rately show a similar pattern increases confidence in these estimates and further supports the argument that breeding among Santa Rita Proechimys is concentrated into 63 Table 12. The proportion of pregnant to nonpregnant females revealed by autopsy of spiny rats trap— ped in the months October through January and February through September in Santa Rita. Age estimates Pregnant/nonpregnant Oct—Jan Feb—Sep Between 90 and 120 days 1/5(20%) 6/12(50%) Between 120 and 200 days 2/40(5%) l3/l6(8l%) Greater than 200 days 6/6l(10%) 106/115(92%) .5-‘-: mxj‘i LE.- 64 NOOH o o o o m HonEoooo I o o o o o HmnEo>oz NNo H N o N o Honouoo NooH o N o N o HonEoudom I o o o o o umzws< .83 o H o H o Nst NHN N m o H q wand .33 o q o H m .32 .83 o H o o H HHS... NooH o m H o q Loam: .N.Ha H 0H m m N 3838 .33 o N o o H 3.268 Uo>Homno LHHHQ oc ujn Hm>HMHm Hm>HHHm Hm>HHHw poundea wcso% wcnox deoqu co co :0 Ho CHOLHHHum o>HH ou um ucmcwopa namcmohd ucmcwoha ou LHHHD m>mw LOHHQ ucmcwopd ucoopom uoz kuOH umfiu Honesz o>mo HonEDz Ludo: .Hmeo Ho mxm ou suwco N 0 NH woumEHumo can oucmHHEH< who: Umpsudmo msmocH mHEom m EHcooOHm onEow Mo CoHqucoo o>Huodpouaop NHLHCOE mo NHmEEDm .MH oHan 65 Figure 7. Santa Rita males Santa Rita Females -------- Estimate based on pregnant females The monthly distribution of conceptions estimat— ed to occur throughout the year at Santa Rita. The circle represents 8.3% or the expected monthly proportion if breeding occurs uniformly throughout the year. Figure 8. 66 Almirante males ................. Almirante females ________ Males and females combined The monthly distribution of conceptions estimat- ed to occur throughout the year at Almirante. The circle represents 8.3% or the expected monthly proportion if breeding occurs uniformly throughout the year. 67 the months of February through September. Obviously mortality is a confounding factor here. If young are born during the period of October through January and their mortality is high they will not be cap— tured. If it is assumed that (1) the number of adult fe— males is constant the year around, (2) the pregnancy pattern is that shown in Table 11, and (3) the average female‘s pregnancy was conceived one month (half the gestation period) earlier, then the estimated pattern of conceptions would be as shown in Figure 7. Comparison of this estimate with the estimated month of conception of field trapped males and females supports the argument that the reproductive season in Santa Rita was restricted but also suggests a higher mortality for spiny rats con— ceived in the verano (January, February, March, and April) since the proportion of conceptions estimated by pregnancy patterns of captured females exceeds the proportion of conceptions estimated by the body lengths of males and females during these four months. Among 84 males and 52 females from Almirante most conceptions were estimated to have occurred in the four months of July to October and the fewest in November, December, and January. The sample for Almirante was small— er and less reliable than the Santa Rita sample. Comparison of the two Figures 7 and 8 suggests that breeding in Almirante was shifted so that breeding 68 was reduced a month later than it was in Santa Rita. Re— call that the heaviest rains occurred in Almirante one month later than they did in Santa Rita and lasted into January. If Proechimys semispinosus in Almirante do breed equally the year around (Table 12), then Figure 8 suggests that mortality was highest among those spiny rats con— ceived in November, December, and January. Additional field results Distribution of embryos in the uterus Analysis of the distribution of embryos in the horns of the uterus of Proechimys was prompted by a remark by Fleming (1969:131) that implantation tended to occur more frequently in one uterine horn of the spiny rats that he examined. One caviomorph, the mountain viscacha (Lagidium peruanum), is known to nearly always (97% in one sample) carry embryos only in the right horn of the uterus (Pear— son, 19492155). The numbers of embryos found implanted in the left and right horns of the uteri fromn pregnant spiny rats trapped near Santa Rita and Almirante are shown in Table 14. Neither the data from Santa Rita nor that from Almi— rante showed any significant; difference from an equal distribution of embryos among the two horns [p(§223.43)= 0.064, d.f.=l and p(§;0.51)=0.475, d.f.=l, respectfully]. 69 Table 14. The distribution of embryos in the uterine horns of pregnant spiny rats trapped near Santa Rita and Almirante. Site Number of Uterine Total number pregnant horn of embryos females Santa Rita 119 Right 161 Left 196 Almirante 24 Right 22 Left 27 70 When the data for both of these sites were pooled however, the number of embryos implanted in the left horns was significantly greater than those implanted in the right horns [2(5223.94)=0.047, d.f.=l]. Unilateral pregnancies among females with embryo counts of two or more were not more frequent than would be expected if an ovule was just as likely to come from the left as well as the right ovary at each ovulation and there was no migration of embryos from one horn to another (Table 15). Unilateral pregnancies among females with embryo counts of 3 or more were less frequent than would be expected by chance 0.005uHHDumE onE HoumH LuBOHw Hoonm moxhnEo Hozom wouwcooc HomeH wcHoooHn HMCOmmom< wcnox mwo>mw coHuanhumHo ow< NqujumE mHmEom HoHHme uHoz mNHm HmuuHH coHumumoo oHo mHo>mw coHuanHumHo ow< NuHusumE onEow HoumH monE UHDom HoHHmEm NHHHDOME onE HmHHHmm Luzopw Moummm mowhnEo oHo: monocooc HoHHmEm wcHoooHn HQCOmmom oodouomeo oz mmHsumom oouooHomlz deSumom oouooHomIH QUCQHHEH< muHm macaw mCoHHMHDQom oquHHEH< oCm .msmocHdeEom mkEHfiooonm mo wuHm mucmm Ho mohsumow NuoumHSIoMHH mo NHmEESm .ON oHan 88 is dependent on the taxonomic level examined. Comparisons among higher taxonomic categories offer strong support but support is weak for intrageneric or intraspecific com— parisons (Stearns, 1977, 1980, 1983a, 1984a). Stearns (1984bz264) has argued that microevolution could not have produced the pattern because it results from differences among higher taxa that occurred long ago and has not changed significantly within lineages since then. I be— lieve the present study refutes this view and shows that this pattern involving r— and K-selection characteristics can be explained by microevolutionary forces. In this study intraspecific tactics were perceptible and trade— offs among r- and K-selection attributes did occur. Due to the logical structure of statistical infer— ence, hypotheses can be rejected but not confirmed. The data presented above is consistent with r— and K—selection theory but does not confirm it. In fact, these data are also consistent with the hypothesis that there has been selection for larger size for Almirante rats and/or selec— tion for smaller size for Santa Rita rats. Larger size can help a population avoid some predation, it increases the variety of food choices available, and increases the efficiency of energy acquisition (Armitage, 1981:43). Once there has been selection for size, allometric growth will lead to life history changes consistent with r— and K-selection (see Theory of r- and K—selection; Blueweiss, 89 gg gl., 1978; Western, 1979; Stearns, 1983a, 1984a). Stearns (1983a) believes that many patterns attributable to r— and K—selection among broad taxonomic groups is due to selection on size followed by coadaptive shifts in life— history attributes. If this is true, then not all corre— lations between the life-history attributes are adaptive. The life history of Proechimys semispinosus in Santa Rita and Almirante is also consistent with the hypothesis that density-dependent mortality applied to all age class— es is greater in Santa Rita, or that adult mortality is high, variable, or unpredictable in Santa Rita and Almi— rante, and/or juvenile mortality is high, variable or un— predictable in Almirante (Stearns, 1983bz601; Parsons, 1983:12). In this study it was assumed that the selection of the life—history parameters of Proechimys semispinosus are mediated through the density of Proechimys with re- spect to resources. Obviously any other forces impinging on the density of Proechimys or its resources like preda- tion or interspecific competition will require modifica— tion of the r- and K—selection model. Many Panamanian predators feed on fruit in season but turn to animal prey when fruit is less abundant (Fleming, 1969:63; Smythe, 1970, 1978, 1982). Interspecific competition may also change seasonally (Smythe, 1970). These are just a few of the many possible dimensions of a life—history pattern 90 (Wilbur, gg gl., 1974; Parry, 1981). Until the covariance structure of the life—history attributes can be identified by study of other populations of Proechimys semispinosus, one cannot identify the life history features under strongest selective pressure. The life history of Proechimys semispinosus in Panama appears to be consistent with r— and K—selection theory but before this hypothesis can be termed the cause of this pattern with any degree of confidence, the rival hypothe— ses must also be tested and the density of populations measured directly. 91 LITERATURE CITED Aldrich, J.W. and B.P. Bole, Jr. 1937. The birds and mam- mals of the western slope of the Azuero Peninsula (Repub— lie of Panama). Sci. Publ. Cleveland Mus. Nat. Hist. 7:1- 196. Allen, J.A. and F.M. Chapman. 1893. On a collection of mammals from the island of Trinidad, with descriptions of new species. Bull. Amer. Nat. Hist. 5:203-234. Armitage, K.B. 1981. Sociality as a life—history tactic of ground squirrels. Oecologia 48:36-49. Baker, R.H. 1983. Sigmodon hispidus. Pp. 490—492 in: D.H. Janzen (ed.), Costa Rican Natural History. Univ. Chicago Press, Chicago. Berry, R.J. 1964. The evolution of an island population of the house mouse. Evolution 18:468—483. Biggers, J.D., C.A. Finn, and A. McLaren. 1962. Long— term reproductive performance of female mice. II. Vari— ation of litter size with parity. J. Reprod. Fertil. 3: 313—330. Birch, L.C., Th. Dobzhansky, P.O. Elliott, and R.C. Lewon— tin. 1963. Relative fitness of geographic races of Dro— sophila serrata. Evolution 17:72—83. Blueweiss, L., H. Fox, V. Kudzma, D. Nakashima, R. Peters, and S. Sams. 1978. Relationships between body size and some life—history parameters. Oecologia 37:257—272. Bonaccorso, F.J., W.E. Glanz, and C.M. Sandford. 1980. Feeding assemblages of mammals at fruiting Dipteréx ana— mensis (Papilionaceae) trees in Panama: see pre ation, dispersal, and parasitism. Rev. Biol. Trop., 28:61—72. Bonner, J.T. 1965. Size and Cycle: An essay on the structure of biology. Princeton Univ. Press, Princeton, N.J. Bonoff, J.B. and D.H. Janzen. 1980. Small terrestrial rodents in eleven habitats in Santa Rosa National Park, Costa Rica. Brenesia 17:163—174. 92 Bowdre, L.P. 1968. An examination of the ecology of sev— eral small terrestrial mammals of three life zones in Costa Rica. Unpublished Organization for Tropical Studies Report. Bowdre, L.P. 1971. Litter size in Sigmodon hispidus. Southwestern Natur. 16:121—128. Bunge, M. 1967. Scientific Research: The search for system. Springer—Verlag, Berlin. Cain, A.J. and C.A. Harrison. 1958. An analysis of the taxonomist's judgement of affinity. Proc. Zool. Soc. London 131:85—98. Camin, J.H. and P.R. Ehrlich. 1958. Natural selection in water snakes (Natrix si edon L.) on islands in Lake Erie. Evolution 1 : —51I. Caughley, G. 1966. Mortality patterns in mammals. Eco— logy 47:906-918. Christian, J.J. and C.D. Lemunyan. 1958. Adverse effects of crowding on reproduction and lactation of mice and two generations of their progeny. Endocrinology 63:517—529. Cody, M.L. 1966. A general theory of clutch size. Evo- lution 20:174-184. Cole, L. 1954. The population consequences of life his— tory phenomena. Quart. Rev. Biol. 29:103-137. Conway, G.R., N.R. Glass, and J.C. Wilcox. 1970. Fitting nonlinear models to biological data by Marquardt's algo— rithm. Ecology 51:503—507. Crow, J.F. and M. Kimura. 1970. An introduction to popu— lation genetics theory. Burgess Publ. Co., Minneapolis. Darwin, C. 1859. The origin of species by means of na- tural selection. John Murray, London. Deevey, E.S., Jr. 1947. Life tables for natural popula— tions of animals. Quart. Rev. Biol. 22:283-314. Delany, M.J. and B.R. Neal. 1969. Breeding seasons in rodents in Uganda. J. Reprod. Fert. (Suppl.6):229-235. Dobzhansky, Th. 1950. Evolution in the tropics. Amer. Sci. 38:209—221. Dunmire, W.W. 1960. An altitudinal survey of reproduc— tion in Peromyscus maniculatus. Ecology 41:174-182. 93 Ehrlich, P.R. and J.H. Camin. 1960. Natural selection in middle island water snakes (Natrix sipedon L.). Evolu- tion 14:136. Emmons, L.H. 1982. Ecology of Proechim s (Rodentia, Echimyidae in southeastern Peru. Iropical Ecology 23:280— 290. Enders, R.K. 1935. Mammalian life histories from Barro Colorado Island, Panama. Bull. Mus. Comp. 201. 78:385- 502. Endler, J.A. 1973. Gene flow and population differentia— tion. Science 179:243—250. Everard, C.O.R. and E.S. Tikasingh. 1973. Ecology of the rodents, Proechimys u annensis and Oryzomys capito 5%:875—886. on Trinidad. J. Mamm. Fisher, R.A. 1930. The genetical theory of natural se— lection. Clarendon Press, Oxford. Fleming, T.H. 1969. Population ecology of three species of neotropical rodents. Ph.D. Thesis, Univ. Mich. Fleming, T.H. 1970. Notes on the rodent faunas of two Panamanian forests. J. Mamm. 51:473—490. Fleming, T.H. 1971. Reproductive patterns of 40 Pana— manian mammals. Handbook for tropical biology in Costa Rica. Organization for Tropical Studies, San Jose, Costa Rica. Fleming, T.H. 1972. Three Central American bat communi- ties: Structure, reproductive cycles, and movement pat— terns. Ecology 53:555-569. Fleming, T.H. 1973. The reproductive cycles of three species of opossums and other mammals in the Panama Canal Zone. J. Mamm. 54:439—455. Fleming, T.H. 1974. The population ecology of two spe- cies of Costa Rican Heteromyid rodents. Ecology 55:493— 510. Foster, R.B. 1982. The seasonal rhythm of fruitfall on Barro Colorado Island. Pp. 151-172 in: E.G. Leigh, Jr., A. Stanley Rand, and D.M. Windsor (eds.), The ecology of a tropical forest: Seasonal rhythms and long—term changes. Smithsonian Inst. Press, Washington, D.C. Frankie, G.W., H.G. Baker, and P.A. Opler. 1974. Com- parative phenological studies of trees in tropical wet and dry forests in the lowlands of Costa Rica. J. Ecol. 62:881—919. 94 Glantz, W.E. 1982. Adaptive zones of Neotropical mam— mals: A comparison of some temperate and tropical pat— terns. Pp. 95—110 in: M.S. Mares and H.H. Genoways (eds.), Mammalian biology in South America. Special Publ. Pymatuning Lab. Ecol. #6. Gliwicz, J. 1984. Population dynamics of the spiny rat Proechimys semispinosus on Orchid Island (Panama). Bio— tropica 16:73—78. Goertz, J.W. 1965. Reproductive variation in cotton rats. Amer. Midl. Nat. 74:329—340. Goldman, E.A. 1920. Mammals of Panama. Smith. Misc. Coll. 69:1—309. Green, P.M. 1964. Density of population as a regulating factor in the reproductive potential of Sigmodon hispidus. Ph.D. Diss. Abstr., Oklahoma State Univ., Stillwater. Guillotin, M. 1982. Activity rhythms and diet of Pro- echimys cuvieri and Oryzomys capito velutinus (Rodentia) in french Gu1ana. Terre Vie 36:337—371. Hairston, N.G., D.W. Tinkle, and H.M. Wilbur. 1970. Natural selection and the parameters of population growth. J. Wildl. Mgmt. 34:681—690. Hall, E.R. 1981. The mammals of North America. Ronald Press, N.Y. Handley, C. 1966. Checklist of the mammals of Panama. Pp. 753—795 in: R.L. Wenzel and V.J. Tipton (eds.), Ecto— parasites of Panama. Field Mus. Nat. Hist., Chicago. Hempel, C.G. 1966. Philosophy of natural science. Pren— tice-Hall, Englewood Cliffs, N.J. Hess, R., W.C. Allee and K.P. Schmidt. 1951. Ecological animal geography. John Wiley, N.Y. Hoffman, R.S. 1958. The role of reproduction and mortal- ity in population fluctuations of voles (Microtus). Ecol. Monogr. 28:79—109. Holdridge, L.R. and G. Budowski. 1956. Report of an eco— logical survey of the Republic of Panama. Caribbean For- ester 17:92—110. Hooper, E.T. 1968. Classification of Peromyscus. Pp. 27—69 in: J.A. King (ed.), Biology of Peromyscus (Roden— tia). Amer. Soc. Mamm., Special Publ. #2. 95 Jackson, W.B. 1965. Litter size in relation to latitude in two Murid rodents. Amer. Midl. Nat., 73:245—247. Jewell, P.A. 1966. Breeding season and recruitment in some British mammals confined on small islands. Symp. Zool. Soc. London 15:89-116. Kilgore, D.L. 1970. The effects of northward dispersal on growth rate of young, size of young at birth, and litter size in Sigmodon hispidus. Amer. Midl. Nat. 84:510—520. King, J.A. 1958. Maternal behavior and behavioral devel- opment in two subspecies of Peromyscus maniculatus. J. Mamm. 39:177-190. Kleiman, D.C., J.F. Eisenberg, and E. Maliniak. 1979. Reproductive parameters and productivity of Caviomorph rodents. 1n Vertebrate ecology in the northern Neotro- pics, ed. J.F. Eisenberg, pp. 173-183. Washington, D.C.: Smithsonian Institution Press. Lack, D. 1954. The natural regulation of animal numbers. Oxford Univ. Press, Fair Lawn, N.J. Lack, D. 1965. Evolutionary Ecology. J. Anim. Ecol., 34:223-231. Leck, C.F. 1970. The seasonal ecology of fruit and nec- tar eating birds in lower Middle America. Ph.D. Thesis, Cornell Univ., Ithaca, N.Y. Lewontin, R.C. 1965. Selection for colonizing ability. Pp. 77—94 in: H.G. Baker and G.L. Stebbins (eds.), The genetics of colonizing species. Academic Press, N.Y. Lord, R.D. 1960. Litter size and latitude in North Amer- ican mammals. Amer. Midl. Nat. 64:488-499. Lusty, J.A. and B. Seaton. 1978. Oestrus and ovulation in the casiragua Proechim s ouairae (Rodentia, Hystrico- morpha). J. 2001. London 182:255-265. MacArthur, R.H. 1962. Some generalized theorems of natu- ral selection. Proc. Nat. Acad. Sci. 48:1893—1897. MacArthur, R.H. and E.O. Wilson. 1967. Island Biogeo- graphy. Princeton Univ. Press, Princeton, N.J. Maliniak, E. and J.F. Eisenberg. 1971. Breeding spiny rats, Proechimys semispinosus, in captivity. Int. Zoo Yearb. 11:93-98. 96 Mares, M.A. and R.A. Ojeda. 1982. Patterns of diversity and adaptation in South American hystricognath rodents. Pp. 393-432 in: M.A. Mares and H.H. Genoways (eds.) , Mammalian Biology in South America. Special Publ. Pyma— tuning Lab. Ecol. #6. Margalef, R. 1959. Mode of evolution of species in re— lation to their places in ecological succession. Proc. Xv Int. Congr. 2001. 10:787-789. McLaren, A. 1963. The distribution of eggs and embryos between sides in the mouse. J. Endocrin. 27:157-181. Moojen, Joao. 1948. Speciation in the Brazilian spiny rats (Genus Proechimys, Family Echimydae). U. Kansas Publ. Mus. Nat. Hist. 1:301—406. Moore, J.C. 1961. Geographic variation in some reproduc- tive characteristics of diurnal squirrels. Bull. Amer. Mus. Nat. Hist. 12221-32. Morrison, D.W. 1978. Foraging ecology and energetics of the frugivorous bat Artibeus jamaicensis. Ecology 59: 716-723. Orsini, M.W. 1962. Technique of preparation, study and photography of benzyl-benzoate cleared material for embry- ological studies. J. Reprod. Fertil. 3:283-287. Parry, G.D. 1981. The meanings of r- and K-lelection. Oecologia 48:260—264. Pearson, O.P. 1949. Reproduction of a South American rodent, the mountain Viscacha. J. Anat. 84:143-167. Pengelley, E.T. 1966. Differential developmental pat- terns and their adaptive value in various species of the genus Citellus. Growth 30:137-142. Petersen, M.K. 1970. Competition between the cotton rats, Sigmodon fulfiventer and S. hispidus. Ph.D. Thesis, Mich. State Univ., East Lansing. Pocock, R.I. 1922. On the external characters of some hystricomorph rodents. Proc. 2001. Soc. London 1922:365- 427. Reig, O.A., M. Aguilera, M.A. Barros, and M. Useche. 1980. Chromosomal Speciation in a Rassenkreis of Venezue- lan spiny rats (Genus Proechimys, Rodentia, Echimyidae). Pp. 291-312 in: N.N. Vorontsov and J.M. Van Brink (eds.), Animal Genetics and Evolution. Dr. W. Junk B.V. Pub— lishers, The Hague. 97 Rensch, B. 1960. Evolution above the species level. Columbia Univ. Press, N.Y. Rich, P.V. and T.H. Rich. 1983. The Central American dispersal route: Biotic history and paleogeography. Pp. 12-34 in: D.H. Janzen (ed.), Costa Rican Natural History. Univ. Chicago Press, Chicago. Schaffer, W.M. 1974. Optimal reproductive effort in fluctuating environments. Am. Nat. 108:783-790. Schmalhausen, 1.1. 1949. Factors of evolution. Blakis— ton, Philadelphia. Smith, F.E. 1954. Quantitative aspects of population growth. Pp. 277-294 in: E.J. Goell (ed.), Dynamics of population growth. Princeton Univ. Press, Princeton, N.J. Smith, M.H. and J.T. McGinnis. 1968. Relationships of latitude, altitude, and body size to litter size and mean annual production of offspring in Peromyscus. Res. Pop. Ecol. 22115-126. Smythe, N.D.E. 1970a. Ecology and behavior of the agouti (Dasyprocta punctata) and related species on Barro Colora- do Island, Panama. Ph.D. Thesis. University of Maryland, College Park, Maryland. Smythe, N. 1970b. Relationships between fruiting seasons and seed dispersal methods in a neotropical forest. Am. Nat. 104225-35. Smythe, N. 1978. The natural history of the Central American agouti (Dasyprocta punctata). Smith. Contr. 2001. 257 1-52. Smythe, N. 1982. Population regulation in some terres- trial frugivores. Pp. 227—238 in: E.G. Leigh, Jr., A. Stanley Rand, and D.M. Windsor (eds.), The ecology of a tropical forest: Seasonal rhythms and long-term changes. Smithsonian Inst. Press, Washington, D.C. Sokal, R.R. and F.J. Rohlf. 1969. Biometry. W.H. Free- man and Company, San Francisco. Spencer, A.W. and H.W. Steinhoff. 1968. An explanation of geographic variation in litter size. J. Mamm. 49:281- 86. Stearns, S.C. 1977. The evolution of life history traits: A critique of the theory and a review of the da- ta. Ann. Rev. Ecol. Syst. 8:145—171. 98 Stearns, S.C. 1980. A new view of life-history evolu- tion. Oikos 35:266—281. Stearns, S.C. 1983a. The influence of size and phylogeny on patterns of covariation among life-history traits in the mammals. Oikos 41:173-187. Stearns, S.C. 1983b. A natural experiment in life- history evolution: Field data on the introduction of mos- quitofish (Gambusia affinis to Hawaii. Evolution 37(3): 601-617. Stearns, S.C. 1984a. The effects of size and phylogeny on patterns of covariation in the life history traits of lizards and snakes. Am. Nat. 123:56-72. Stearns, S.C. 1984b. Models in evolutionary ecology. In population biology and evolution, ed. K.W6hrmann and V. Loeschcke, Pp. 261—265. Springer-Verlag, Berlin, Heidelberg. Steel, R. and J. Torrie. 1960. Principles and procedures of statistics. McGraw—Hill, N.Y. Tinbergen, N. 1959. Comparative studies of the behavior of gulls (Laridae): A progress report. Behavior 15:1- 70. Tinbergen, N. 1963. On aims and methods of ethology. Z. Tierpsychol. 20:410—433. Tinbergen, N. 1967. Adaptive features of the blackhead- ed gull, Larus ridibundus L. Pp. 43-59 in: D.W. Snow (ed.), Proc XIV Int. Orn. Congr. Tomich, P.Q., N. Wilson, and C.H. Lamoureux. 1968. Eco— logical factors on Manana Island, Hawaii. Pacif. Sci. 22:352-368. Webb, S.D. and L.C. Marshall. 1982. Historical biogeo— graphy of recent South American land mammals. Pp. 39-52 in: M.A. Mares and H.H. Genoways (eds.), Mammalian Biolo- gy in South America. Special Publ. Pymatuning Lab. Ecol. #6. Weir, B.J. 1973. Another Hystricomorph rodent: Keeping casiragua (Proechimys guairae) in captivity. Lab. Animals 7:125-134. Weir, B.J. 1974. Reproductive characteristics of Hy- stricomorph rodents. Symp. Zool. Soc. London. 34:265-301. 99 Western, D. 1979. Size, life history and ecology in mam— mals. Afr. J. Ecol. 17:185-204. Wilbur, H.M., D.W. Tinkle, J.P. Collins. 1974. Environ- mental certainty, trophic level, and resource availability in life history evolution. Amer. Nat. 108:805-817. Williams, G.C. 1966. Adaptation and natural selection. Princeton Univ. Press, Princeton, N.J. Williams, R.H., J.L. Carmon, and F.B. Golley. 1965. Ef- fect of sequence of pregnancy on litter size and growth in Peromyscus polionotus. J. Reprod. Fertil. 9:257-260. Zimmerman, K. 1950. Die Randformen der Mitteleuropai- schen Wuhlmause. Pp. 454-471 in: A. Jordans and F. Peus (eds.), Syllegomena Biologica Festschrift. Leipzig. M'°ll'l1[llllfl[lllflllflljljl'lllllfllllls