POPULATION TRENDS AND HABITAT ASSESSMENT OF SNAKES AND LIZARDS IN SOUTH CENTRAL KANSAS

Transcription

1 FINAL REPORT ON FISH AND GAME AGENCY CONTRACT # 8 POPULATION TRENDS AND HABITAT ASSESSMENT OF SNAKES AND LIZARDS IN SOUTH CENTRAL KANSAS Dwight R. Platt Department of Biology Bethel College North Newton, Kansas October, 1985

2 ABSTRACT A two-year study of snake and lizard populations in south central Kansas was done by live trapping from mid-may to mid-july to assess the use of agricultural habitats by snakes and lizards and to compare present populations with those sampled in earlier studies. Trapping was done at the Sand Prairie site in western Harvey County in natural sand prairie and borders of cultivated fields and pastures; and at a predominantly cultivated site and a predominantly pastured site in the clay loam soils of eastern Harvey County. A total of 89~ captures of 11 species of snakes and 265 captures of four species of lizards were made in the two years. Snakes and lizards were much more abundant and di verse at the Sand Prairie site. Plains garter snakes, western hognose snakes, eastern hognose snakes, glossy snakes, bullsnakes, prairie-lined racerunners, prairie lizards and slender glass lizards were only caught at the Sand Prairie site and yellowbelly racers and red-sided garter snakes were more abundant there. Black rat snakes and great plains skinks were only caught at the two eastern sites and prairie king snakes were more abundant there. Most snakes and lizards were caught in larger numbers in natural grasslands with numbers being reduced in both pastured and cultivated areas. Few snakes were caught in a wooded creekland site, although black rat snakes and prairie king snakes had higher capture rates in this habitat. 1

3 Some fluctuations in populations of snakes over a 27 ~ear period on the western stud~ site can be related to fluctuations in the availabilit~ of pre~, but snake populations also went through periods of decline and increase which cannot be obviously related to e xtrinsic factors. The ~ellowbell~ racer, red-sided garter snake and plains garter snake went through periods of population declines and increases with the highest rate of capture being more than ten times the lowest. Bullsnake populations fluctuated less strongly and prairie k ing snake populations were relatively low for the entire period. The western hognose snake, eastern hognose snake and gloss~ snake populations went through precipitous declines in the late 196 ' s and have remained in a depressed or declining state to the present time. Lizard populations have maintained themselves at the Sand Prairie site, although some have undergone substantial fluctuation over the 27 period. 2

4 INTRODUCTION This two-year study of snake and lizard populatlonb in Harvey County in south central Kansas had two objectives : 1 ) to assess the use of agricultural habitats by snakes and lizards; and 2) to compare present populations with those sampled in earlier studies in order to measure fluctuations and trends in population numbers of species of snakes and lizards on sites in western Harvey County, Kansas. 1 Traps trap used to capture snakes and lizards were installed in stations at three study sites. Each trap station included a low metal drift fence, eight to fifteen feet long, to intercept movement. Funnel traps were fitted on each end of the drift fence. To operate the trap station, the ends of the traps were closed but no bait was used to attract animals. Traps were usually checked twice a week. CFor more information on trapping methods see Platt, 1969.) Fifty trap stations were installed at the Sand Prairie site in the sandhills of the Hutchinson Dune Tracts in western Harvey County and 5 trap stations were installed at two sites in clay loam soils east of Newton in Harvey County, the Sand Creek site in an area of cultivated land and the Torline site in a predominantly pasture area. Snakes and lizards were trapped and studied at these sites from May 1~ to July 16, 198~ and from May 1~ to July 15, at each site was measured in trap station days. Trapping effort A trap station day is the operation of one trap station for a 2~-hour period. Snakes that were caught were usually taken to the Bethel 3

5

6

7 prairie CEl / 2 SEl l ~ Sec. 36 T22S R3W) managed as a natural area by the Biology Department at Bethel College. This tract had been pastured by cattle prior to the fall of 1965 when it was acquired by Bethel College. The trapping site included approximately 72 acres of the reservation. Fourteen trap stations were constructed on adjacent land--six on a sandy pasture immediately north of Sand Prairie Reservation (approximately 1 acres ) and eight bordering cultivated fields and grassland east of the reservation ( approx imately 2 acres). The longest snake movement between captures at this study site was 8 meters, and snakes moved between the reservation and the other two trapping sites. The trap stations were mostly in the same locations used in a more e xtensive snake trapping study done in the late 196 ' s and early 197's. This study site was in an area with a high proportion of cultivated land, less than one mile east of North Newton in the south half of section ~ and the north half of section 9, T23S, RlE. In the square mile surrounding this trapping site, 76% of the land was under cultivation (approximately 1 / 5 in alfalfa and the rest in wheat, milo or kept fallow during the weeks of this study ). Approx imately 1% of the land was grassland, mostly domestic pasture, ~% in wooded creek land and ~% in farmsteads and barnyards. An interstate highway passes through this section and approximately 6% of the land is used for the highway and adjoining right-of-way. Sand Creek runs along the west side of this site and a small tributary runs through the southern half. 6

8 Fifteen trap stations were placed between cultivated fields. Six trap stations were placed under solitary trees in two dry waterways with brome grass or weedy cover ( Meadowlark location). These trap stations were surrounded by wheat fields and fallow fields. The soil type was Clime silty clay. Nine trap stations were placed in two grassy fencerows with scattered tree or shrubby cover CEberle location). These fencerows bordered on wheat fields, milo fields and fallow fields. Soil types were Clime silty clay and Farnum loam. Ten trap stations were constructed in the wooded drainageway of Sand Creek. The tree canopy was at least 9% complete over these trap stations in the bottom or on the sides of the drainage-way. The soils were alluvial sands and silts. Traps at the Eberle location were less than 1 / ~ mile from those along Sand Creek and approx imately 1 / 2 mile from those at the Meadowlark location, moving between locations. but there were no recaptures of snakes The longest distance that a recaptured snake had moved at this site was 13 meters. Torline site: This site was in an area with a high proportion of pastured grassland. Twenty-five trap stations were constructed on a ~-acre property Ctraps sampled approx imately 25 acres ) eight miles east of North Newton CSl / 2 Sl / 2 NEl l ~ Sec. 5 T23S R2E). Passing through this site is a tributary of the west branch of Whitewater Creek which is impounded as a large pond just south ( downstream) of the study site. The study site consisted of a horse pasture, a small area of unpastured grass ( cultivated at one time but most of it now a rather sparse weedy unpastured grassland with scattered trees ), a small pond, 7

9 creekland and a brome grass field that was mowed for hay. Sixty-five per cent of section 5 was in native grassland, most of it pastured. In addition, 28% was cultivated, ~% was in brome grass hay meadow, 1% in farmyards and 2% in wooded creekland. Twelve trap stations were placed along the sides of the creek drainage-way. Most of these were on grassy slopes under solitary trees or shrubs, but others were under the canopy of the woodland in the bottom of the drainage-way. The brome grass hay meadow adjoined this creekland to the east. The soils were primarily alluvial in the creekland. Thirteen trap stations were placed in grassland, six in pastured grassland and seven in unpastured grassland with scattered trees and shrubs. The soil types were Clime silty clay and Irwin silty clay loam. The longest distance that a recaptured snake had moved at this site was 19 meters. RESULTS The traps were operated for a total of 12,129 trap station days during this study, 6,~1 in 198~ and 6,88 in Table 1 shows the trapping effort in various habitats on the three study sites in the two years. A total of 89~ captures of 11 species of snakes were made in the 1 trap stations in the two years. A total of 265 captures of four species of lizards were made. Table 2 summarizes the captures of snakes and lizards. 8

10 TABLE 1 : Trapping effort measured in trap station days Ctsd ) in vari ous habitats o n the three 5tudy 5itee 1n Hervey County, Kansas, in 198~ and Study site/ habitat No. of 198~ trap 1985 trap Total trap stations - -=d_ays._~_ "- days tion.~ays_., station station trap sta- Sand Prairie site 5 Sand prairie natural area 36 Cultivated f i elds 8 Pasture 6 2,982 2, 136 ~ ,1 2,232 ~ , 82 ~ Sand Creek site 25 Creekland 1 Cultivated fields 15 1, ~1 1,~ , 22 1, 1 8~ 1, 838 Torline site 25 Creekland 12 Ungrazed grassland 7 Pasture 6 1, ~ , ~99 72 ~ , 2~ 1, ~52 8~ TOTAL 1 6,~1 6,88 12,129 TABLE 2 : Total captures of snakes and lizards in the 1 trap stations on the three study sites in Harvey County, Kansas, in 198~ and SNAKES Species 198~ Both years ~ Yellowbelly racer Plains garter snake Red-sided garter snake Prairi e king snake Bull snake Black rat snake Western hognose snake Eastern hognose snake Blotched water snake Diamondback water snake Glossy snake LIZARDS Prairie-lined racerunner Prairie lizard Slender glass lizard Great plains s k ink ~9 9 '± ~OS ~ so 15 ~3 7 1'± ~ ~ '±

11 The catches of both snakes and lizards were lower in 1985 than in 198~. The largest reductions were in captures of garter snakes. There were man~ ponds and marshes on the Sand Prairie Reservation during the trapping period of 198~ and there were moderate populations of amphibians, the principal pre~ of garter snakes. A drought followed for the rest of the summer and dried out the ponds and marshes. There was onl~ one pond with water for the entire trapping period in 1985 and active amphibians were scarce. However the streams in eastern Har v e~ Count~ were running throughout the trapping period in The plains garter snake was found onl~ at the Sand Prairie site. The mean number of captures of the plains garter snake per semiwee k l~ trapping period was significant!~ lower in 1985 C2.9 ) than in 198~ (7.1 ; t = 3.8 ; p < 1%). This was probabl~ due to the scarcit~ of amphibians. Captures of red-sided garter snakes were also significant!~ reduced in 1985 at the Sand Prairie site. Mean captures per trapping period were 3.6 in 1 98~ and 1.3 in 1985 Ct ; P < 1% ). However red-sided garter snakes were caught in larger numbers in 1985 at the Torline site where frog populations were higher in 1985 than in 198~. There were substantial but not statistical!~ significant reductions in catches of ~ellowbell~ racers and bullsnakes in Total catches of racers were reduced at all three stud~ sites but the number of individual snakes captured was similar in the two ~ears Ctable ~ ) and the population estimate for racers at the Sand Prairie site i n 1985 is higher than for 198~ ( table~ ). A greater proportion of ~oung snakes migrating to and from the stud~ site in 1985 rna~ have biased the population estimate. 1

12 Younge snakes Cless than 6 mm. sv length) comp ised 31% of the captured sample in 198S and only 1~% in 198~. Of the 123 ace s caught in 198~, ~ we e ecaptured late in the 198~ t apping period an ave age of 1.6 times while only 28 of the 119 ace s caught in 198S we e ecaptu ed an ave age of 1.2 times. Rep oductive activity was significantly highe in 198~ than in 198S. Of ~1 adult female ace s caught befo e June 2 in 198~, 31 C76%) we e g avid while of 37 adult female race s caught before June 2 in 198S, only 2 CS~%) were gravid Cchi square 9.~; P < O.S% ). The reduced reproduction of racers in 198S may have esulted from the stress of drought in 198~. This drought may also have increased mortality in late summe, fall and winte of 198~-8S. The sample of bullsnakes is too low to make any reliable interp etations. The proportion of large bullsnakes in the sample is much lower in 198S. The bullsnake population may also have been affected by the drought but see the discussion of the bullsnake in the section on Population Trends. The eduction in catches of the prairie-lined racerunner in 198S may have been related to the drought in the summer of 198~. Habitat assessment: One study site with SO t ap stations was located in the sandy soil of western Harvey County and two study sites with a total of SO t ap stations were located on p edominantly clay loam soils in eastern Harvey County. In addition. to the difference in soil types, the eastern sites were more heavily disturbed by human use and settlement. Table 3 shows the ave age ate of captu e of snakes and lizards on the western site and the two eastern sites du ing the two-year study 11

13 and table ~ shows the number of individual snakes and lizards captured on these sites and Jolly-Seber population estimates for these sites and species with sufficient data. TABLE 3: Average rate of capture Cnumber of captures per 6 trap station days) of snakes and lizards from the middle of May to the middle of July in 198~ and 1985 on the western sandy study site CSand Prairie site) and the two eastern clay loam study sites CSand Creek and Torline sites) in Harvey County, Kansas. Species Ng. of captures per 6 trap station days SNAKES Western site Eastern sites Yellowbelly racer Plains garter snake Red-sided garter snake Prairie king snake Bull snake Black rat snake Western hognose snake Eastern hognose snake Glossy snake ~ ~ LIZARDS Prairie-lined racerunner Prairie lizard Slender glass lizard Great plains skink Snakes were much more abundant and diverse on the western sandy site ( almost four times the rate of capture on the eastern sites ). Five species were caught only on the sandy site--plains garter snake, western hognose snake, eastern hognose snake, glossy snake and bullsnake. The bullsnake has been observed in eastern Harvey County, but the other four species are probably restricted to the western sand prairies in Harvey County. Two of the most common species, yellowbelly racer and red-sided garter 12

14 snake, although found on all three sites, were much more abundant on the Sand Prairie site. One species, the black rat snake, was found only on the two eastern sites while another species, the prairie king snake, was more common there. The difference in lizard populations in the two parts of the county was even more e xtreme, and lizards were very rare in the eastern part. Three species, the Prairie-lined racerunner, prairie lizard and slender glass lizard, were only caught at the western site, and the great plains skink was only caught at the two eastern sites (although it has rarely been caught at the western site in past years). The differences in abundance and diversity of snake and lizard populations at these study sites are primarily due to three factors : 1) the sandy nature of the soil in the western site which facilitates burrowing Cat least four of the species limited to this region are burrowers and all seek shelter in burrows); 2) the extensive standing water at certain seasons of the year in the western site because of the lack of welldeveloped surface drainage is favorable for the development of large prey populations for some of these predators; and 3 ) the greater disturbance and human settlement in the eastern part of the county. Snake and lizard populations were probably more abundant and diverse in the sand prairies before European settlement of the area, this difference. but human disturbance has exacerbated Snake populations appear to be quite mobile with some snakes continually moving in and out of areas. However at times when conditions are favorable, snakes may remain resident in a local 13

15 a ea fo an extended pe iod of time, at least as long as thei needs a e being met. Table 5 shows ecaptu e pe centages fo those species of liza ds and snakes on the weste n study site and the two easte n study sites fo which the e a e sufficient eco ds. These data indicate that a population of ma ked snakes was building up and ca ied ove f om 198~ to 1985 in significant numbe s only fo the yellowbelly ace in both a eas and the edsided ga te snake in the easte n a ea. Fo the othe populations, the pe centage of ma ked individuals late ecaptu ed emained less than 15~, the pe centage of captu es that we e ecaptu es emained less than 2~ andl o the e was little ca yove of ma ked animals f om 198~ to Fo the yellowbelly ace and the p ai ie king snake, ecaptu e pe centages we e highe on the weste n study site, indicating a population with lowe tu nove. The p ai ie king snake showed some buildup of a ma ked population in each yea but no ecaptu es in 1985 of snakes ma ked in 198~. The ed-sided ga te snake had highe ecaptu e pe centages on the To line site whe e the population was inc easing than on the othe two sites. The bullsnake had much lowe ecaptu es in this study than in most p evious yea s, indicating a small esident population at this time. Movement eco dsindicate that liza ds, such as the p ai ielined ace unne~ emain in a mo e est icted home a ea. Howeve mo tality may have been quite high in the p ai ie-lined ace unne, as ecaptu e eco ds we e low and population tu nove high. 1~

16 TABLE ~= Number of individual snakes and lizards caught in 198~ and 1985 on each of the three study sites in Harvey County, Kansas. These are minimum estimates for the number of each species that used the site during the trapping period in that year. For those species with sufficient data, average Jolly Seber estimates of population size are included in parenthesis. SP - western Sand Prairie site; SC - eastern Sand Creek site; T - eastern Torline site. Species Number of individuals (Jolly-Seber population estimate~ ~ ~ _?. SP sc T SP sc T SNAKES 337 ~2 ~ ~8 Yellowbelly racer 123 ( 1~1) Plains garter snake 119 ( 198) Red-sided garter snake 62 9~) Prairie king snake 5 Bullsnake 26 2~ ) Black rat snake 13 ( 8 ) 1 1~ 5 21 C2 ) 1~ ( 31) ~ C176) 52 C86 ) 26 ( ~3) 1 C1) 1~ C S~ ) ~ Western hognose snake 7 ( 9 ) Eastern hognose snake ~ 3 Gl ossy snake 1 LIZARDS 1~ Prairi e-lined racerunner 13 (167* ) Prairie lizard 9 78 ( 129* ) ~ Slender glass lizard ~ 5 Great plains skink *These estimates are for Sand Prairie Reservation only because of the discrepancy in recapture rates on adjacent cultivated and pastured sites C19% of captures were recaptures on Sand Prairie and only~% on adjacent sites. ) 15

17 TABLE 5: Recaptuce peccentages foe macked snakes and lizacds caught in 198~ and 1985 on the westecn sandy study site and the two eastecn clay loam study sites in Hacvey County, Kansas. Only those species ace included foe which at least 1 individuals wece caught in one yeac. A. 198~ peccentage of individuals caught in 198~ that wece cecaptuced one De moce times peccentage of individuals caught in 1985 that wece cecaptuced one De moce times. 198~-85 peccentage of individuals caught in 198~ that wece cecaptuced in B. Peccentage of captuces that wece cecaptuces of macked individuals A. Species Pee cent of individuals that w,ece_lat~-~-r:-~ca.p._.t l:j_r::.~_c! Westecn sandy site f:_c!_st~;:ij_cla_y._l_q_c!_o}_?)_t::_~.. ~ 198~ ~ ~ ~ SNAKES Yellowbelly cacec ~2% 2~% 2% 29% 9% 15% Plains gactec snake 9% 1% 2% not pcesent Red-sided gactec snake 1% 8% 3% 17% 15% 12% Pcaicie king snake 2% 2% 11% 9% Bull snake 12% 7% Lf% not pcesent LIZARDS Pcaicie-lined cacecunnec 17% 9% 6% not pcesent B. Species Pee cent of captuces that wece cecaptuces Westecn sa~_i t~- Eastec_n cla_y_tq... r:r! !.. t~.s 198' ~ 1985 SNAKES Yellowbelly cacec 35% 38% 17% 29% Plains gactec snake 11% 12% not pcesent Red-sided gactec snake 7% 17% Lf% 18% Pcaicie king snake 17% 17% 1% 8% Bullsnake 7% 13% not pcesent LIZARDS Pcaicie-lined cacecunnec 1~% 2% not pcesent

18 Table 6 shows the capture rates of snakes and lizards in the different habitats in which traps were located on the western study site and table 7 shows the capture rates of snakes and lizards in different habltats on the eastern study sites. TABLE 6: Capture rates (numbers of captures per 6 station days) of snakes and lizards in the western sandy area CSand Prairie site), Harvey County, Kansas SNAKES Species trap study Undisturbed Disturbed areas Sand prairie Cultivated fastured ~~tb_~~?~s Yellowbelly racer Plains garter snake Red-sided garter snake Prairie king snake Bull snake Western hognose snake Eastern hognose snake Glossy snake '± '± '±.6 5.'± '±.6 '±.2 1. '±.9.7.'± LIZARDS Prairie-lined racerunner 28.3 Prairie lizard 1.5 Slender glass lizard The Yellowbelly Racer was caught at the highest rate in undisturbed sand praire, although it was caught in all habitats sampled. In the sandy area, the capture rate was reduced approx imately 5% in both cultivated and pastured areas. On the eastern sites, the ungrazed grassland, the fencerows and waterways between cultivated fields and the open grassy creekland had similar capture rates, but captures in the pasture were significantly reduced and few racers were caught in the heavily wooded creekland. The capture rates in the cultivated habitat at 17

19 the Sand Creek site and the creekland at the Torline site may be somewhat biased by immigrations of snakes caused by mowing, combining, burning and plowing of neighboring land. TABLE 7 : Capture rates Cnumbers of captures per 6 trap station days ) of snakes and lizards in the various habitats in the two eastern clay loam study sites CSand Creek and Torline sites ), Harvey County, Kansas, in 198~ and Habitat areas sampled were wooded creekland in cultivated areas and fencerows and waterways between cultivated fields in Sand Creek site and open or wooded creekland in grassland area, pastured grassland and ungrazed grassland in Torline site. See table 1 for trapping effort in the various habitat types. Species Sand Creek site Torline site Creekland Cultivated Creekland Pasture Ungrazed SNAKES Yellowbelly racer ~ Red-sided garter snake ~ Prairie king snake ~.6 ~ Black rat snake ~ LIZARDS Great plains skink The plains garter snake had the highest capture rate in the undisturbed sand prairie. The capture rate for the traps in the pasture was much higher than for those in the cultivated area where few plains garter snakes were caught. However the sand prairie probably had higher amphibian populations, at least in 198~, because surface water was more available and amphibians were the most important prey of these snakes. The red-sided garter snake had highest capture rates in the undisturbed sand prairie of the Sand Pra~e site and in the open creekland and ungrazed grassland of the Torline site. It was caught in very low numbers in the wooded creekland and cultivated 18

20 land of the Sand Creek site and in the pastured sand prairies. The highest capture rates of the prairie king snake were in the cultivated fields and nearby wooded creekland of the Sand Creek site. It was caught in all of the other habitat types but in low numbers. The bullsnake was caught in both the natural sand prairie and in pastured and cultivated sandy are~s at similar rates. However it is interesting that the capture rate in the cultivated area was slightly higher than that in the natural sand prairie or the pasture. The black rat snake was caught at the highest rate in the wooded creekland of the Sand Creek site and was cuaght in lower numbers in the cultivated areas of that site. At the Torline site it was only caught in the traps along the creekland. The highest capture rate for the prairie-lined racerunner was in the undisturbed sand prairie. These lizards were more abundantly caught on the pasture than in the cultivated area. The captures of the other three species of snakes and the other three species of lizards were too low to assess their use of different habitats. Agricultural practices affect snake and lizard populations. Practices such as mowing and baling may cause mortality. Plowing and disking often turn up nests of eggs that are destroyed. However of even more importance may be the changes created in the habitat by these practices. Harvesting, mowing, burning and plowing may destroy, at least temporaily, favorable habitat for these reptiles. In this study, migration out of cultivated fields that were harvested or mowed was observed in a number 19 of

21 instances. In 198~. after the mowing of the brome grass ha~field at the Torline site in late June, racers were caught in larger numbers in the area bordering this field. Later these snakes were caught across the creek in the nati ve unmowed grassland. Greater snake activit~ was evident for the rest of the trapping period after the completion of the mowing. In 1985, the migration was not so evident, possibl~ because the mowing occurred over a longer period of time and was interrupted b~ a period of rain. At the Sand Creek site, traps were placed between cultivated fields. When harvest occurred, there was sometimes an immediate increase in snake activit~ in the fencerow. There were usual!~ few captures later. Snakes did not remain in the fencerows and either did not survive or moved out into larger areas of favorable habitat. We were not able to ascertaion where these snakes went. However unharvested fencerows and other areas of vegetated land are probabl~ important to the surv i val of snake populations when land is periodical!~ cultivated and harvested or mowed. Fencerows serve as routes for travel of snakes in cultivated areas, especial!~ at times when the cultivated areas are bare. At the Eberle location, traps were placed in two fencerows appro x imate!~ 155 meters apart, separated b~ a wheatfield. At the Meadowlark location, traps were placed in two waterwa~s appro x imate!~ 18 meters apart, separated b~ a wheatfield. In eight of the nine recaptures that occurred at these two locations, the recapture was in the same fencerow or waterwa~ as 2

22 the previous capture. In only one instance had the snake mo ved across the field. Snakes and lizards have been trapped in the sand prairies of western Harvey County in 16 of the last 27 years. From 1959 to 1963, trapping was done in Harvey County West Park, approximately one mile southeast of Sand Prairie Natural History Reservation. From 196 to 1963, trapping was also done in a pasture which adjoins Sand Prairie Reservation to the south. These study sites were described i n Platt C1969). From 1966 to 197~ and 198~ to 1985, trapping was done on Sand Prairie Reservation and cultivated areas and pastures adjoining it to the north and east. For this analysis of population trends comparisons are made of the trapping results in the period between the middle of May and the middle of July ( e xcept for 1966 when trapping was not begun until early June). Table 8 shows the length of the periods being compared, the trapping effort in trap station days and the capture rates for certain prey of snakes for each year. Tables 1 and 11 show the capture rates of snakes for each year. Although capture rates of snakes may be affected by weather conditions such as temperature, cloudiness, or precipitation, these environmental conditions do not make a large difference in capture rates for a two-month period Ctable 9 ). The mean capture rate for the five cool years,with mean temperatures during the trapping months more than 1#F below norma~was ~6. captures/ 6 tsd, for the five years that were within= ff of normal was ~3.1 and for the six warm years C1 F or more above normal) was ~

23 TABLE 8: Numbers of da~s and total trap station da~s Ctsd ) for trapping periods being compared and capture rates Ccaptures per 6 trap station da~s) of pre~ animals for some species of snakes in each ~ear of a snake population stud~ in western Harvey Count~, Kansas. Year Number Total CaQture rate of pre!d animals C ~! uresj_ _Q_9~\i.l. of da~s tsd Frogs Lizards Rodents '± '± '± '± '± '±.'± '± '± '± '± '± '± '± '±.9 197'± 62 71'± '± '± For the eight ~ears in which precipitation during the trapping period was more than seven inches, the capture rate was '±'±.6 captures/ 6 tsd; while for the eight years with precipitation less than seven inches, it was '±'±.. Flooding of the area does affect capture rate. For the seven ~ears in which the sum of the mid-ma~ and mid-jul~ flooding indices is less than seven, the capture rate is 33. 2, while i n the four years in which the sum of the indices was seven or more, the capture rate was 55.'±. This effect was probabl~ due to a larger population of some snakes attracted by increased availability of pre~ because of the surface water. The main factor affecting capture rates at comparable seasons in different ~ears is the size of the snake 22

24 TABLE 9 : Captu e ates of all snakes Ccaptu es pe 6 t ap stati on days ) and conditions of tempe atu e, p ecipitation and su face wate fa the t apping pe iods being compa ed in each yea of a snake population study in weste n Ha vey County, Kansas. Tempe atu e conditions a e e xp essed as the mean depa tu e f om no mal of monthly ave age tempe atu es (Of ) fa May, June and July. P ecipitation data ma ked with an aste isk C ) we e measu ed on Sand P ai ie Natu al Histo y Rese vation. Othe p ecipitati on and tempe atu e data we e taken f om _Climatological Oat~ = Kansas CNOAA ) fa Newton, Kansas. Su face wate on Sand P ai ie Natu al Histo y Rese vati on is e xp essed on a scale f om Cno su face water ) to 5 ( ma x imum f l oodi ng ). Compa able surface wate data a e not a vailable fo 1959 to Yea Captu e ate Tempe atu e P ecipitation --- ~l,!f"_f a c_e!_ ~-'!!:_~L----- all snakes co F depa t. My 15-Jly15 Mid-My Mid-Jly Ma x. f om no mal ) Cinches ) '± '± '±.7'± '± '± '± '± '± '± '± '± '± 6.'± '± '± 5'± '± 3 '± 198'± '± 3 '± population. The efo e have used captu e ates as the best index of population size in snakes and liza ds du ing this study. Th ee of the most abundant species of snakes, the yellowbelly ace, the ed-sided ga te snake and the plains ga te snake, went th ough pe iods of population decline and inc ease that lasted fo a numbe of yea s wi th the highest ate of captu e being mo e than ten times the lowest Ctable 1 ). The 36 t ap stations used on Sand P ai ie Rese vation in 198'± and 23

25 TABLE 1: Rates of capture of some species of snakes, yellowbelly racer, plains garter snake, red-sided garter snake and bullsnake, from mid-may to mid-july in western Harvey County, Kansas. Rate of capture is expressed as number of captures per 1 trap stations operated for 6 days C6 trap station days) Year Yellowbelly Plains Red-sided Bull snake racer garter snake garter snake 'L '± '± '± 5.6 '± '± '± '± '± '± '± 1973 '±.9 28.'± '± 197'± '± '± 1'± were among the trap stations used in the period from 1969 to 197'±. In order to test the significance of the fluctuations in capture rate, calculated the mean capture rate of racers in these 36 trap stations per semiweekly trapping period for the 18 trapping periods used in the study for some of these years. then tested the statistical significance of the differences in these means. From 198'± to 1985, the total capture rate for the two months declined approximately 2~ Ctable 1). The mean capture rate pe~ trapping period declined from 8.67 racers to 6.78 racers. This difference was not statistically significant Ct- 1.'±32; P- 1-2~). The capture rate in 1985 was '±. 7 times the capture rate in 1972 Ctable 1). 24 The mean capture rate per

26 trapping period declined from 6.78 racers to 1.39 racers. This difference was statistically significant Ct - 5.9~5; p <.1% ). Although all the have not yet checked the statistical significance of differences between yearly means, it appears that a capture rate increase of three times or more Cor a comparable decrease ) is significant. The yellowbelly racer had high populations in , 1966 and 198~-85, with the rate of capture in 198~ being more than twice any previous high. Since the racer eats a varied diet, food was probably a limiting resource in few years. Although the high points in and 1966 coincide with highs in lizard populations, an important prey, the lowest populations from 1969 to 1971 coincide with very high frog populations, a prey also eaten by the racer. The abundance of one of the most important diet items, grasshoppers, was not monitored. The fluctuations in racer populations do not appear to be caused by changes in prey availability but are multi-year periods of decline followed by periods of increase. The fluctuations of the population of plains garter snakes can be related in a general way to fluctuations in the population of its principal prey, the leopard frog Ctables 8 and 1). Frog populations are directly related to the amount of standing water on the reservation, and so captures fluctuate greatly from year to year Ctable 8). The population of plains garter snakes on the reservation increased in the early 196's during a few years when frog populations were high. It declined to its lowest point from 1966 to 1968 when three consecutive years had rather low 25

27 populations of frogs. The plains garter snake population increased from 1969 to 1973 when four of five years had high frog populations. However the garter.snake population did not always respond in the same way to changes in frog populations, and there seems to be some inertia in increases and decreases. When frog populations were low in 1972 and 197~, there was little or no decrease in snake populations, while a decline in frog populations in 1985 was accompanied by a significant decline in the plains garter snake population. Large increases in garter snake population often occurred in a year of high frog populations following a year of low frog populations Csee 196, 1969, and 1973 in table 1), Since trapping was done from mid- May to mid-july, before any young had been born, these increases must have resulted from the migration of snakes onto the reservation in times of food abundance. However the very high numbers in 1973 and 197~ must be caused primarily by reproductive recruitment into the population over a period of years when food was abundant. Fluctuations in the population of the red-sided garter snake were similar to those of the plains garter snake, with increases and decreases Cfrom the previous year) in the populations occurring in the same years for 12 out of 15 years. Leopard frogs were also the principal prey for the red-sided garter snake on the study area. There was a reversal in the relative abundance of the two species over the years of this study. From 1959 through 1962 the captures of red-sided garter snakes were substantially higher than those of the plains garter snake. 26 In

28 ten of the twelve years in which trapping was done from 1963 through 1985 the capture rate of the plains garter snake was higher than that of the red-sidedd species and in the other two years the capture rates were similar. In 198~-85 the capture rate of the plains garter snake was almost twice that of the redsided garter snake. Bul1snake populations fluctuated less strongly than populations of racers and garter snakes, with the highesccapture rates being slightly more than five times the lowest Ctable 1). Changes in the bullsnake population usually came a year later than similar changes in the population of rodents, their principal prey. The highest bullsnake population was in 1967, following a high peak in rodent populations in 1966, and the lowest numbers in 1972 followed the minimum rodent populations in The effect of prey availability on reproductive success of bullsnakes probably accounts for a substantial part of these population increases and decreases. The precipitous drop in bullsnake captures in 1969 is probably the result of e xtensive flooding on Sand Prairie Reservation which lasted for much of that year. There is evidence from recapture records that some bullsnakes left the reservation for that year and came back in the subsequent year. The cause of the decline in captures of bullsnakes in 1985 is not evident, since rodent captures were relatively high in 198~ and 1985 (tables 8 and 1). The sample in 1985 included no snakes with a snout-vent length of more than 1 mmm., while in 198~, 12 of the 26 snakes captured were of that length. The recapture rates of bullsnakes were very low in 198~-85 compared to previous years. 27 In a previous study (Platt,

29 198~) recaptures amounted to 3~% of captures while in this study they amounted to only 9% of captures. No bullsnakes were captured in the eastern part of the county during this study, although they have been observed there previously. The bullsnake population needs to be monitored to see what changes occur in the nex t few years. The prairie king snake had relatively low populations in the sand prairies during this 27 year study Ctable 11 ). The capture rate was particularly low from 1963 through 197~, with somewhat higher rates from 1959 through 1962 and in 198~ and Although rodents are the principal prey of the prairie king snake TABLE 11 : Rates of capture of some species of snakes, prairie king snake, western hognose snake, eastern hognose snake and glossy snake, from mid-may to mid-july in western Harvey County, Kansas. Rate of capture is e xpressed as number of captures per 1 trap stations operated for 6 days C6 trap station days ). Year Prairie king snake Western hognose snake Eastern hognose snake Glossy snake ~ 1.2.~ ~. 5 1.~ ~ '± '± '± '±.1.'± 198~ b

30 on the study area, there was little positive response in the prairie king snake population to the extremely large rodent populations from 1966 to The population appears to have been in a period of decline at that time. Three species of snakes show a significant continuing decline in population size in recent years, the western and eastern hognose snakes and the glossy snake Ctable 11). A precipitous decline in captures occurred after 1967 Ctable 11). of western hognose snakes From 1959 through 1967 the mean capture rate per 6 trap station days for the seven years of trapping was 8.8 western hognose snakes and the lowest capture rate was ~.5 snakes. From 1968 through 1985 the mean capture rate for the nine years of trapping was 1.5 snakes, with the highest capture rate being 1.8 snakes. The eastern hognose snake was caught in lower numbers than the western species, earlier ( table 11 ). and the decline in captures occurred a year From 1959 through 1966 the mean capture rate per 6 trap station days for the six years of trapping was 2.8 eastern hognose snakes with the lowest capture rate 1.~ snakes. From 1967 through 1985 the mean capture rate for ten years of trapping was.~ snake and the highest capture rate was.7 snake. For the glossy snake the major decline came in 197. In the nine years of trapping prior to 197 the mean capture rate per 6 trap station days was.9 glossy snake, with none being caught in the mean In the seven years of trapping from 197 through capture rate was.2 snake, with the highest 29

31 capture rate being.~ snake. The decreased population levels in these three species have lasted from 16 to 19 ~ears. Population levels of these three species need to continue to be monitored here and in other parts of their range in the state. The long period of low populations in these species could be part of a long term c~cle. The prairie king snake showed a similar period of low population numbers for at least 11 ~ears and possibl~ longer, but its population has recovered in 198~ and However the hognose snakes and the gloss~ snake have shown no evidence of an upturn through These three species are vulnerable to decline and e xtirpation from the state because of their specialized habitat requirements and ver~ The local distribution. populations of three species of lizards are maintaining themselves on the stud~ area Ctable 12). Two of the species, the prairie-lined racerunner and the prairie lizard, have undergone substantial fluctuations over the 27 ~ear period. The~ were caught in large numbers in 1966, when trapping was first begun on Sand Prairie Reservation. The prairie-lined racerunner also had high capture rates from 1959 through 1963 and a ver~ low capture rate in 197~. The slender glass lizard had ver~ low capture rates for all sixteen ~ears that trapping was done. JO

32 TABLE 12: Rates of prairie-lined racerunner, lizard, from mid-ma~ to Kansas. Rate of capture 1 trap stations operated capture of some species of lizards, prairie lizard and slender glass mid-jul~ in western Harve~ Count~, is expressed as number of captures per for 6 da~s C6 trap station da~s). Year Prairie-lined Prairie lizard Slender glass racerunner lizard '± '±5.2 '± '± '±.'± '± '± J1

33 CONCLUSIONS The following conclusions a~e based on the data collected in Ha~vey County. Howeve~ it is assumed that they can be applied mo~e gene~ally to these species in Kansas. 1. Snakes and liza~ds a~e much mo~e abundant and dive~se in the sand dunes ( Hutchinson Dune T~acts) of weste~n Ha~vey County than in easte~n Ha~vey County with clay loam sails. Rate of captu~e of snakes an the weste~n site was almost fau~ times that on the easte~n sites. Oiffe~ences in captu~es of liza~ds we~e even ma~e e x t~eme, with few liza~ds being found in easte~n Ha~vey County. These diffe~ences a~e p~abably ~elated to sail type, amount of standing wate~, and land use and distu~bance. 2. Plains ga~ter snakes, weste~n hagnase snakes, easte~n hagnase snakes, glassy snakes and bullsnakes we~e only caught at the Sand P~airie site in this study, and all e xcept the bullsnake a~e p~obably rest~icted to the sandy part of the county. Yellawbelly racers and ~ed-sided ga~te~ snakes are much ma~e abundant on the Sand P~ai~ie site than an the easte~n sites. 3. The black ~at snake was captured only on the two easte~n sites, and the p~ai~ie king snake is ma~e abundant an the eastern sites. ~. The p~ai~ie-lined ~ace~unne~. p~airie liza~d and slender glass liza~d a~e only found in the weste~n sandy part of the county, and the g~eat plains skink is p~ima~ily ~est~icted to the easte~n pa~t. 5. Recaptu~e pe~centages of the yellawbelly ~ace~ and the p~ai~ie king snake we~e highe~ an the Sand P~ai~ie site, 32

34 indicating that these populations have lower turnover rates. The red-sided garter snake had higher recapture percentages on the Torline site where the population was increasing. 6. Yellowbelly racers are found in all habitats sampled but were caught at the highest rate in undisturbed sand prairie; captures were reduced in both cultivated and pastured areas. Few racers were caught in wooded creekland. 7. Red-sided garter snakes are found in all habitat types sampled but had highest capture rates in the natural sand prairie of the western site and in the open creekland and ungrazed grassland of the Torline site. It was only caught in low numbers in the wooded creekland of the Sand Creek site. 8. Prairie king snakes are found in all habitat types sampled, but had highest capture rates in the cultivated fields and wooded creekland of the Sand Creek site. 9. Bullsnakes were caught in natural sand prairie, cultivated sandy area and sandy pastures at similar rates of capture. Although they were only caught in the western sand dunes in this study, they have been regularly observed in the past in the eastern part of the county. 1. The black rat snake was caught at the highest rate in the wooded creekland of the Sand Creek site. 11. The prairie-lined racerunners are found in all sandy habitats sampled but were caught in highest rates in natural sand prairie and at lowest rates in the cultivated area. 12. Agricultural practices reduce snake and lizard populations, and reduction on pastured lands is similar to that on cultivated lands. Snakes can readily move out of cropland 33

35 when it is temporarily disturbed by mowing or harvest into more ravorable habitat ir it is available. more greatly reduced by agriculture. Lizard populations are 13. Fencerows and waterways serve as routes for travel and areas ror escape or snakes in cultivated areas, especially at times when cultivated fields are harvested, mowed or plowed. 1~. Some rluctuations in populations of snakes can be related to fluctuations in the availability of prey, but snake populations also go through periods or decline and increase which cannot be obviously related to extrinsic factors. 15. The yellowbelly racer, red-sided garter snake and plains garter snake went through periods of population decline and increase that lasted for a number or years in the last 27 years, with the highest rate of capture being more than ten times the lowest. Fluctuations in the populations of garter snakes can be related in a general way to fluctuations in populations of frogs, but the fluctuations in racers could not be correlated with any e xtrinsic conditions. 16. Bullsnake populations fluctuated less strongly than populations of racers and garter snakes, with the highest capture rates being slightly more than five times the lowest. These changes can be related to changes in populations of rodents and other extrinsic conditions. 17. Garter snake populations showed some increase in years of high frog populations as a result of migration to areas of high prey availability. Changes in bullsnake populations were one year later than changes in rodent populations, indicating 34

36 that inc<eases and dec<eases in bullsnake populations we<e p<imarily due to reproductive recruitment. 18. The p<ai<ie king snake population at the Sand P<airie site was relatively low for the entire 27 year period. The population was in a ve<y dep<essed state fo< at least 11 yea<s but had recovered in 198~ The western hognose snake, eastern hognose snake and glossy snake populations went through precipitous declines in the late 196's and have remained in a depressed or declining state to the present time. These species are vu l nerable to extirpation, although this decline could also be a normal part of a long te<m cycle in population abundance. 2. Prairie-lined race<unner and prai<ie liza<d populations a<e maintaining themselves at the Sand Prairie site, although they have unde<gone substantial fluctuation ave< the 27 year pe<iod. Captures of the slender glass lizard have been very low but stable ave< the 27 year pe<iod. RECOMMENDATIONS 1. In cultivated and pastu<ed <egions, grassy fencerows and wate<ways and natu<al grassy wildlife areas need to be maintained in o<de< to have good populations of snakes. These vegetated areas should not be mowed or burned at the same time that cultivated fields a<e bare or provide unfavo<able habitat for snakes. 2. Populations of lizards and snakes should continue to be monitored on a periodic basis in south central Kansas by trapping at some of the study sites used in this study, possibly in two 35

37 years out of each five. Similar monitoring sites need to be established in western Kansas and in eastern Kansas. 3. The western hognose snake, eastern hognose snake and glossy snake should be placed on the list of species in need of conservation. The highest priority should be for studies to determine if the depressed populations of these species are typical of other parts of their range in Kansas or whether this is a local phenomenon. Also the population needs to be monitored for signs of an upturn. If the depressed state of these populations proves to be quite general and if there is no sign of an increase in the next five years, these species need to be placed on the threatened list and studies initiated to try to determine the cause for the depressed state of their populations. ~. Bullsnake populations need to be watched. Although their numbers are still adequate in the sand prairies of western Harvey County, the results from this study indicate that their populations are low in the agricultural regions of the eastern part of the county. Also there is evidence that they may not be maintaining self-sustaining populations at the Sand Prairie site at the present time. ACKNOWLEDGMENTS Jennifer Hiebert and Andrea Schmidt Andres served as research assistants helping with all aspects of this study. would like to thank the following landowners who allowed us to trap snakes and lizards on their property: Joe W. Goering, Harvey Holzrichter, Torline. Cecil Eberle, Meadowlark Homestead and John J6