Microtus Californicus Research Articles

Of mice and mallards: positive indirect effects of coexisting prey on waterfowl nest success

Coexisting prey species interact indirectly via their shared predators when one prey type influences predation rates of the second prey type. In a temperate system where the predominant shared predator is a generalist, I studied the indirect effects of rodent populations on waterfowl nest success, both within the nesting season among sites and among years. Among six to ten upland fields (14 to 27 ha), mallard (Anas platyrhynchos) nest success was positively correlated with rodent abundance in all three years of the study. After removing year effects, mallard nest success remained positively correlated with the relative abundance of rodents. Of the rodent species present, California voles (Microtus californicus) were the most important coexisting prey type influencing nest success. Among years, mallard nest success was positively correlated with vole abundance; the asymptotic relationship suggests a threshold response to vole abundance, beyond which predators become satiated and additional voles do little to affect nest success. I tested and rejected three alternative explanations for the observed positive correlation between mallard nest success and rodent abundance that do not involve an indirect effect of coexisting prey populations. The influences of dense nesting cover, nesting density, and predator activity did not explain the observed patterns of nest success. These results suggest that rodent populations buffer predation on waterfowl nests, both within and among years, via the behavioral responses of shared predators to coexisting prey.

Population dynamics of adult fleas (Siphonaptera) on hosts and in nests of the California vole.

Microtus californicus (Peale, 1848) were live trapped or retrapped 887 times, and fleas were collected over a 2.5-yr period in the San Francisco Watershed and Wildlife Refuge. Also, 179 M. californicus nests were collected monthly with observations to identify the environment of fleas. The ratio of the mean number of fleas per nest to the mean number collected on voles was 4.7:1 for Malaraeus telchinus (Rothschild, 1905), 13:1 for Hystrichopsylla occidentalis linsdalei Holland, 1957,9:1 for Atyphloceras multidentatus multidentatus (C. Fox, 1909), and 76:1 for Catallagia wymani (C. Fox, 1909). The proportion of each flea species per nest to those per host was not closely associated seasonally or spatially. The average nest contained 37.5% moisture content in relation to its total weight and ranged between 3 and 92%. No relationship was observed between relative humidity of air within nests and flea number, but a significant relationship existed between the moisture of nesting materials and flea number. Malaraeus telchinus and A. m. multidentatus were collected in greatest numbers from nests having 30-39% moisture content by weight, and H. o. linsdalei and C. wymani were most numerous in nests that had 40-49% moisture content. Catallagia wymani had the greatest tolerance for high moisture and was severely affected by lack of moisture and nearly absent in drier nests. No ectoparasites were collected from nests that had <12% moisture content, and nests with >50% moisture content had few fleas. A static concept of nest fleas and host fleas as suggested by averages and often used in literature is questionable. In seasonal comparison of populations of fleas on hosts and in nests, M. telchinus and H. o. linsdalei reversed so that more of the flea population was on the host than in the nest for a short time during fall.

Modelling the influence of seasonal inter-habitat movements by an ecotone rodent

I assessed the long-term effects of seasonal inter-habitat movements for a common rodent, Microtus californicus, using matrix models tailored to demographic data obtained on populations living at the marsh–grassland ecotone. Three stochastic models were used to simulate the seasonal and spatial dynamics of this system, with the purpose of determining: (1) the growth rates of populations living in single versus a two-habitat landscape; and (2) if the amount, direction, and seasonality of rodent movements between the two habitats may impact these growth rates. The results indicated that grassland populations could be self-sustaining, but that isolated marsh populations would rapidly decline. Importantly, 98% of the two-habitat simulations yielded growth rates greater than those seen in the grassland model, with populations benefiting the most by movements from marsh into grassland at the end of the dry season. This study demonstrated that movements between marsh ‘sink’ and grassland ‘source’ habitats can potentially increase long-term population growth due to seasonal, inter-habitat differences in resource availability.

Using stable isotopes to assess seasonal patterns of avian predation across a terrestrial-marine landscape.

In this study, we used the stable isotope ratios 15N/14N and 13C/12C to clarify the spatial and temporal patterns of small mammal derpredation by wide-ranging raptors across a salt marsh and grassland landscape. To determine whether clear isotope signals existed for the two adjoining habitats, and if these differences could allow for an assessment of the seasonal habitat origins of depredated rodents, primarily the vole, Microtus californicus, we investigated the nitrogen and carbon isotope ratios for three sets of data. First, to establish potential differences in isotope signatures between the two habitats, we collected plant species consumed by the vole. Second, we analyzed bone collagen from voles originating from known locations along a gradient from marsh to grassland. Finally, the signatures of these unpredated voles were then used to categorize isotope values of vole remains found in the pellets of their raptor predators. Results indicated that mean δ15N and δ13C were both higher in marsh plants than in grassland species. Although both isotopes showed a trophic level increase from plant to bone collagen, nitrogen provided the more informative signature when attempting to distinguish the habitat origins of voles. We found that the δ15N values of unpredated voles were significantly related to the distance of the specimen along the marsh-grassland interface, with higher values in the marsh habitat. Analysis of the depredated bones found in raptor pellets indicated that a greater percentage of voles originated in grassland during the spring/summer season (88%), as compared with the fall/winter period (60%). Finally, we compared this pattern of habitat and season-specific predation to another measure of predation risk, the relative encounter rates of raptors foraging for voles, and found that the two estimates were similar except during the winter season. This study demonstrates the usefulness of stable isotope analysis for elucidating seasonal patterns of predation across a complex landscape.

Pneumocystosis in wild small mammals from California.

Cyst forms of the opportunistic fungal parasite Pneumocystis carinii were found in the lungs of 34% of the desert shrew, Notiosorex crawfordi (n = 59), 13% of the ornate shrew, Sorex ornatus (n = 55), 6% of the dusky-footed wood rat, Neotoma fuscipes (n = 16), 2.5% of the California meadow vole, Microtus californicus (n = 40), and 50% of the California pocket mouse, Chaetodipus californicus (n = 2) caught from southern California between February 1998 and February 2000. Cysts were not found in any of the harvest mouse, Reithrodontomys megalotis (n = 21), California mouse, Peromyscus californicus (n = 20), brush mouse, Peromyscus boylii (n = 7) or deer mouse, Peromyscus maniculatus (n = 4) examined. All infections were mild; extrapulmonary infections were not observed. Other lung parasites detected were Hepatozoon sp./spp. from M. californicus and Notiosorex crawfordi, Chrysosporium sp. (Emmonsia) from M. californicus, and a nematode from S. ornatus.

Life with fur and without: experimental field energetics and survival of naked meadow voles.

Hair is considered to be a basic mammalian feature that provides protection and insulation, promoting energy conservation and survival. To quantify the functional significance of mammalian pelage, we tested the short-term experimental effects of fur removal in a natural population of the California vole, Microtus californicus, in winter. The daily energy expenditure (DEE) of seven voles was directly measured in the field using stable isotopes, first with the animals in their natural condition and then again after experimental removal of pelage by shaving. The initial mean±SD DEE of 96.0±23.1 kJ/day increased by only about 10%, to 106.3±21.4 kJ/day, following shaving. The voles showed a body mass loss of 5%, about half of which was due to fur removal. Comparing slightly larger samples of all animals whose local survival could be documented, we found 100% survival over the 5 initial days of the experimental manipulation in 16 control animals and nine shaved animals; over the following 3 weeks the survival of shaved mice did not differ significantly from that of controls. We were surprised that the average increase in energy expenditure of voles without fur was so modest, though the range of individual values was great, and likewise we were surprised that shaved voles survived as well as they did. M. californicus survives naturally in winter under conditions of social aggregation that include huddling together of individuals in nests; this situation probably provided our experimentally shaved voles an opportunity to minimize the energetic disadvantages of pelage loss. They may also have employed a variety of compensatory physiological and behavioral responses, including reduction in activity time and food intake, and perhaps a related small decline in body mass. Our limited sample sizes made it difficult to detect subtle differences that may have been biologically significant in the system we studied. Nonetheless, we can reaffirm that fur has an insulative value that promotes energy economy and survival. However, we also conclude that mammalian physiology and behavior are sufficiently complex and flexible that a variety of responses can be deployed to promote survival under unusual circumstances such as those of our experimental test.

Hantavirus (Bunyaviridae) infections in rodents from Orange and San Diego counties, California.

During a screening program to determine the extent of hantavirus activity in Orange and San Diego Counties, California, serum samples from 2,365 rodents representing nine genera and 15 species were tested for hantavirus antibodies. A reverse transcription-polymerase chain reaction on selected seropositive rodents was used to identify the specific hantavirus. Rodents positive for Sin Nombre virus (SNV) antibodies by Western blot included 86 (9.1%) of 948 deer mice (Peromyscus maniculatus), four (1.5%) of 275 California mice (Peromyscus californicus), one (0.5%) of 196 cactus mice (Peromyscus eremicus), 51 (12.2%) of 417 harvest mice (Reithrodontomys megalotis), and five (12.5%) of 40 California voles (Microtus californicus). All other specimens tested were negative for hantavirus antibodies. There was a correlation between age and sex of the reservoir host and prevalence of SNV antibody, especially among male deer mice and harvest mice. Few seasonal trends in antibody prevalence were observed and continued maintenance of SNV and El Moro Canyon virus was found at several foci over a 4-5-year period. Isla Vista virus was also found in voles and represents the first recorded in Orange County. Microhabitat selection on the part of these rodents based on plant density, plant height, and availability of food plants may explain, to some extent, all of the hantavirus-positive foci throughout the study area over a broad geographic range and the lack of antibody-positive rodents in dense chaparral, woodland, and riparian areas. The majority of rodents positive for SNV was identified from localities along coastal bluffs and the foothills of the Santa Ana Mountains, where trap success was high and P. maniculatus represented 43% of all rodents collected. Several residential, commercial, and industrial sites exist in these areas and the potential health risk should not be overlooked. This study represents an in-depth analysis of the prevalence, host distribution, and characteristics of rodent populations infected by three hantaviruses within a small, well-defined, geographic area.

Role of small mammals in the ecology of Borrelia burgdorferi in a peri-urban park in north coastal California.

The role of small mammals other than woodrats in the enzootiology of the Lyme disease spirochete, Borrelia burgorferi, was assessed in the peri-urban park. Mammals were collected monthly from September through to April. Following tick removal, the animals were tested for B. burgdorferi by culture of ear-punch biopsies. Larvae and nymphs that were intermediate in morphology between Ixodes spinipalpis and Ixodes neotomae occurred on several species of rodents (Peromyscus truei, Peromyscus californicus, Microtus californicus, Rattus rattus and Reithrodontomys megalotis) and the brush rabbit (Sylvilagus bachmani). Morphometric analyses of these I. spinipalpis-like ticks and the offspring from two I. neotomae females from the site suggest that I. neotomae may bo conspecific with I. spinipalpis. Borrelia burgdorferi was isolated from eight out of 109 (7.3%), three out of 16 (18.8%), two out of 38 (5.3%) and two out of six (33.3%) P. truei, P. maniculatus, M. californicus and R. rattus, respectively. One bush rabbit yielded the first isolate of B. burgdorferi from a lagomorph in western North America. This isolate and three others derived from unfed I. spinipalpis-like nymphs failed to produce infection when inoculated intradermally into 11-12 P. maniculatus each. Likewise, no spirochetes were detected in 420 Ixodes pacificus nymphs derived from larvae fed on animals inoculated with these isolates. An additional isolate, derived from an I. spinipalpis-like nymph, was recovered by ear-punch biopsies from five our of 12 (42%) needle-inoculated P. maniculatus. However, spirochetes were not detected in 20 I. pacificus nymphs fed as larvae on each of five mice (two infected and three uninfected) inoculated with this isolate. We conclude that brush rabbits and several species of rodents besides woodrats may contribute to the maintenance of B. burgdorferi because they harbour the spirochete and are fed upon by competent enzootic vectors.

Effect of ectoparasite removal procedures on recapture of Microtus californicus.

We tested the null hypothesis that anesthetizing meadow voles (Microtus californicus) and brushing them vigorously for ectoparasites would have no effect on the later recapture of these voles. Voles were trapped from 6 December 1993 to 10 January 1994 at Faye Slough Wildlife Area near Eureka, Humboldt County, California (USA). Alternate trapped voles were anesthetized with ethyl ether and brushed vigorously for ectoparasites. There was no significant difference in the frequency of recapture nor in the time to first recapture between those voles anesthetized and brushed, compared to control animals.

Isla Vista virus: a genetically novel hantavirus of the California vole Microtus californicus.

Prospect Hill virus (PH) was isolated from a meadow vole (Microtus pennsylvanicus) in 1982, and much of its genome has been sequenced. Hantaviruses of other New World microtine rodents have not been genetically characterized. We show that another Microtus species (the California vole M. californicus) from the United States is host to a genetically distinct PH-like hantavirus, Isla Vista virus (ILV). The nucleocapsid protein of ILV differs from that of PH by 11.1% and a portion of the G2 glycoprotein differs from that of PH by 19.6%. ILV antibodies were identified in five of 33 specimens of M. californicus collected in 1975 and 1994-1995. Enzymatic amplification studies showed that 1975 and 1994-1995 ILV genomes were highly similar. Secondary infection of Peromyscus californicus was identified in Santa Barbara County, California. A long-standing enzootic of a genetically distinct hantavirus lineage is present in California voles.

Social organization and space use in California voles: seasonal, sexual, and age-specific strategies.

We intensively monitored space use and movement in Microtus californicus over a 2-year period that included 1 year of high density (maximum 618/ha) and one of low (minimum 5/ha); historically this population has exhibited cycles of 2 or 4 years. Adults of both sexes dispersed at the start of the breeding season, culminating in the establishment of intrasexually exclusive territories. In females, these territories persisted throughout life, except that many young females recruiting during the breeding season established contiguous, overlapping, or adjacent home ranges with their mothers. This female philopatry explains the conclusion of previous workers that females of this species are non-territorial. In the dry (non-breeding) season, females had smaller ranges that often overlapped and were clustered. Adult males moved breeding territories at a modal interval of 6 weeks; this is consistent with their avoidance of inbreeding with philopatric daughters. Ranges overlapped 1-4 adult females at any one time, and a cohort of 7 long-lived males overlapped an average of 16.4 females during their tenure on the grid. The period of maximum overlap with adult females varied among individual males, and did not correlate with the time of maximum body weight. Ranges of males in the dry season overlapped extensively, with persistent associations among some individuals. In the lowdensity year, ranges of some adults failed to overlap intersexually. Juvenile males dispersed gradually between 3 and 13 weeks of age (half before 9 weeks), with some leaving after reaching sexual maturity; a few remained philopatric. Of juvenile females, 47% remained philopatric with the rest disappearing before 9 weeks of age. New understanding of vole social behavior, dispersal, and space use is achieved by focusing on the seasonal dynamics of spatial relationships among individuals with respect to age, sex, and relatedness.

Short Report: Prevalence of Hantavirus Infection in Rodents Associated with Two Fatal Human Infections in California

Rodents living near two fatal human cases of hantavirus pulmonary syndrome in California were surveyed for evidence of hantavirus infection. Seventeen (15%) (14 Peromyscus maniculatus and one each of P. truei, Eutamias minimus, and Microtus californicus) of 114 rodents tested had evidence (enzyme-linked immunosorbent assay or polymerase chain reaction) of hantavirus infection. This suggests that Peromyscus mice, and P. maniculatus in particular, may be the reservoir for the virus causing this newly recognized disease in California, as previously reported for New Mexico and Arizona.

Modifications of a cholinesterase method for determination of erythrocyte cholinesterase activity in wild mammals.

A method to determine erythrocyte cholinesterase (ChE) activity was modified for use in wild mammals. Erythrocyte ChE of California voles (Microtus californicus) was primarily acetylcholinesterase (AChE), which was similar to the brain and unlike plasma which was primarily butyrylcholinesterase (BChE). Triplicate erythrocyte AChE analyses from individual animals of several species of wild rodents revealed a mean coefficient of variation of 8.7% (SD = 4.3%). Erythrocyte ChE activity of several wild mammals of California revealed that mule deer (Odocoileus hemionus) had the highest erythrocyte AChE activity (1,514.5 mU/ml) and dusky-footed woodrats (Neotoma fuscipes) had the lowest activity (524.3 mU/ml). No ChE activity was found in erythrocytes of several species of birds and fish.

Uptake of environmental contaminants by small mammals in pickleweed habitats at San Francisco Bay, California.

Small mammals were live-trapped in pickleweed (Salicornia virginica) habitats near San Francisco Bay, California in order to measure the uptake of several contaminants and to evaluate the potential effects of these contaminants on the endangered salt marsh harvest mouse (Reithrodontomys raviventris). Tissues of house mice (Mus musculus), deer mice (Peromyscus maniculatus), and California voles (Microtus californicus) from nine sites were analyzed for chemical contaminants including mercury, selenium, cadmium, lead, and polychlorinated biphenyls (PCBs). Concentrations of contaminants differed significantly among sites and species. Mean concentrations at sites where uptake was greatest were less than maximum means for the same or similar species recorded elsewhere. Harvest mice (Reithrodontomys spp.) were captured only at sites where concentrations of mercury or PCBs were below specific levels in house mice. Additional studies aimed at the protection of the salt marsh harvest mouse are suggested. These include contaminant feeding studies in the laboratory as well as field monitoring of surrogate species and community structure in salt marsh harvest mouse habitats.

A Response to Solving the Enigma of Population Cycles with a Multifactorial Perspective

We believe that Lidicker's (1988) paper on the enigma of microtine cycles, based on a plenary address presented at the 1987 annual meeting of the American Society of Mammalogists, needs response because it contains several unclear points and conclusions of uncertain validity. Lidicker (1988) argued that the enigma of microtine cycles will not be solved by a single-factor approach. He reiterated his earlier claim that a multi-factorial perspective (MFP) is absolutely essential if we are going to understand microtine demography (Lidicker, 1988:226). In contrast, we believe that the multifactorial perspective in its present form is difficult to comprehend; our objective is to address in detail difficulties with it. If these difficulties can be overcome, the multifactorial perspective may be acceptable to more researchers in the field. Three major aspects in Lidicker's (1988) presentation of the multifactorial perspective that could lead to confusion are imprecise terminology, misinterpretation of empirical results, and philosophy of science. Imprecise terminology.-One of the greatest potential sources of confusion in science is imprecise terminology. Presaturation dispersal and dispersal sinks are two terms used by Lidicker (1988) that need to be clarified. The best way to eliminate confusion is to define these terms operationally. Lidicker (1988) asserted that presaturation dispersal is one of eight major components in the multifactorial perspective having a role in the density-regulation machinery of California voles (Microtus californicus). Unfortunately, Lidicker (1975, 1978, 1988) has not defined presaturation dispersal operationally. Certainly, the general characteristics of these animals could not be relied upon because Lidicker (1975:106) defined them so broadly as ... in relatively good condition, and may include any sex and age group including pregnant females. Another approach would be to use the density of the resident population as an indicator of presaturation dispersal. Animals emigrating from the resident population at carrying capacity would be saturation dispersers, whereas animals emigrating before the resident population reaches carrying capacity would be presaturation dispersers. However, numbers of voles and abundances of resources vary over time, thus, it is impossible to measure the carrying capacity. If the operationality of presaturation dispersal is dependent on the operationality of carrying capacity, it behooves Lidicker to suggest how carrying capacity in fluctuating populations of voles is measured. On the positive side, Lidicker's (1975) concept of presaturation dispersal brought attention to the fact that animals may leave a population before food, space, and mates become limiting. Interestingly, Krebs et al. (1973) recognized the effect of dispersal during phases of increasing density on the population dynamics of voles in fenced enclosures (fence effect), but their observations were not widely appreciated at the time. A solution to the problem is to use the rate of population instead of carrying capacity in distinguishing between the two types of dispersal. Presaturation dispersal would occur when population-growth rates are positive, whereas saturation dispersal occurs when population-growth rates are zero or negative. At least the rates of the population can be measured. Lidicker (1988:229) stated the growth rate of a local population is affected strongly by the availability of a dispersal sink (Lidicker, 1975, 1980). Peak densities occur when sinks are filled and dispersal becomes frustrated. These statements are confusing because of Lidicker's ambiguous definition of dispersal sink. Lidicker (1975:117) defined dispersal sink as some empty or unfilled suitable habitat, or perhaps marginal or even unsuitable habitat in which at least short-term survival is possible. If all spatially accessible unsaturated habitats are dispersal sinks, it may be impossible to find a habitat that is not a dispersal sink. Furthermore, if dispersal sinks are unsuitable or marginal habitats where dispersers presumedly suffer high mortality rates, it is difficult to understand how the sinks will become saturated and frustrate dispersal in

Extra-Large Body Size in California Voles: Causes and Fitness Consequences

Extra-large body size in microtine rodents is a ubiquitous feature of peak population densities, and it has been hypothesized that these giant individuals represent a genetically based morphotype that has high fitness under increasing and high densities, and may represent a key element of a genetic polymorphism driving multi-annual cycles (Chitty/Krebs model). We examine this large-size phenomenon (Chitty Effect) in the California vole utilizing three approaches : analyzing the weight distribution in a non-cyclic population over a 13 yr period with comparisons to cyclic populations, analyzing body composition, especially fat content, as a function of body size, and observing the microhabitat distribution of extra-large males in two populations, one of which cycles

Extrafloral Nectaries: A Defense against Ant-Homoptera Mutualisms?

microtine and shrew populations in western Finland. Oikos 46: 124-127. 1986b. Gradients in population fluctuations of Tengmalm's Owl Aegolius funereus in Europe. Oecologia (Berl.) 69: 195-201. 1989. Breeding performance of Tengmalm's Owl Aegolius funereus: effects of supplementary feeding in a peak vole year. Ibis 131 (in press). and Norrdahl, K. 1987. Low proportion of shrews in the diet of small mustelids in western Finland. Z. Saugetierkunde 52: 257-260. and Norrdahl, K. 1989a. Predation of Tengmalm's Owls: numerical responses, functional responses and dampening impact on population fluctuations of microtines. Oikos 54: 154-164. and Norrdahl, K. 1989b. Avian predation on mustelids in Europe 1: occurrence and effects on body size variation and life traits. Oikos 55: 205-215. and Sulkava, S. 1987. Diet and breeding performance of Ural Owls Strix uralensis under fluctuating food conditions. Ornis Fennica 64: 57-66. Kotler, B. P., Brown, J. S., Smith, R. J. and Wirtz II, W. 0. 1988. The effects of morphology and body size on rates of owl predation on desert rodents. Oikos 53: 145-152. Latham, R. M. 1952. The fox as a factor in the control of weasel populations. J. Wildl. Manage. 16: 516-517. 986a. Predation causing synchronous declin phases in icrotine and shrew populations in western Finland. Liberg, 0. 1984. Food habits and prey impact by feral and house-based Domestic cats in a rural area in southern Sweden. J. Mammal. 65: 424-432. M cLean, S. F. Jr., Fitzgerald, B. M. and Pitelka, F. A. 1974. Population cycles in arctic lemmings: winter reproduction and predation by weasels. Arct. Alp. Res 6: 1-12. Maher, W. J. 1967. Predation by weasels on a winter population of lemmings, Banks Island, Northwest Territories. Can. Field-Nat. 81: 248-250. 1970. The pomarine jaeger as a brown lemming predator in northern Alaska. Wilson Bull. 82: 130-157. Norrdahl, K. 1985. The population fluctuations of small mammals in Suomenselka and southern Ostrobothnia, western Finland, in 1969-84. Suomenselain Linnut 20: 57-68 (In Finnish with summary in English). Pearson, 0. P. 1966. The prey of carnivores during one cycle of mouse abundance. J. Anim. Ecol. 35: 217-233. 1971. Additional measurements of the impact of carnivores on California voles (Microtus californicus). J. Mammal. 52: 41-49.

Food habits of California voles, Microtus californicus, in artichokes, Cynara scolymus

The California vole ( Microtus californicus) is the most serious vertebrate pest of artichokes ( Cynara scolymus) in California. A microhistological analysis of the stomach contents of 45 voles collected at intervals from November 1983 to July 1985 from seven different artichoke fields indicated that artichoke was a significant and the major component of their diet during all seasons. In the spring (April) collections, the amount of artichoke consumed was somewhat reduced and about equal to all other food items combined due to increased consumption of grasses (wild oat, Avena fatua, and common foxtail, Hordeum leporinum) and/or oxalis ( Oxalis pes-caprae) at that time of year. Other items consumed in relatively minor amounts by voles in artichoke fields were cheeseweed ( Malva parviflora), chickweed ( Stellaria media), mustard ( Brassica spp.), and insects. This study clearly illustrates that artichoke may comprise the bulk of the diet of California voles and in part explains the severity of the damage they often cause. The reasons for apparent seasonal dietary shifts are discussed.

The Ecology of the Salt-Marsh Harvest Mouse (Reithrodontomys raviventris) in a Diked Salt Marsh

The salt-marsh harvest mouse (Reithrodontomys raviventris) was a fugitive species in this diked marsh. These mice were able to use more open and saltier pickleweed (Salicornia virginica) when numbers of meadow mice (Microtus californicus) were high and they moved into deeper and less salty pickleweed as the population of meadow mice declined. In this first extensive study of a diked marsh, salt-marsh harvest mice were shown to swim and run across larger open areas than they had been thought to cross previously. The salt-marsh harvest mouse (Reithrodontomys raviventris) endemic to the marshes of San Francisco Bay (Fisler, 1965), has become endangered primarily because of destruction and modification of its habitat. Approximately 80% of the historic tidal marshes have been destroyed and most of those that remain support few to no mice because of backfilling, subsidence, or vegetation changes (Shellhammer, 1982). Approximately 30% of the historic tidal marshes of San Francisco Bay remain as diked marshes. Eighty percent of the diked marshes are located in Suisun Bay, the northeasterly component of the San Francisco Bay system. The habitat is primarily managed for waterfowl use; mouse populations are small and separated by large areas of inappropriate habitat. Approximately 13 km2 of diked marshes remain in the South San Francisco Bay and constitute potential refuges for the more endangered of the two harvest mouse subspecies, R. r. raviventris, within a ring of large cities and extensive bayland development. Little, however, is known of the biology of the mouse in diked marshes subject to great changes in moisture and salinity throughout the year. Some are flooded periodically by rain waters or by bay water after breaks in outboard levees. Most surface water evaporates each summer. This study was conducted to ascertain mouse numbers, distribution, and interactions with other species during the summer when moisture is low and salinities are high; the microhabitat requirements of these mice in diked marshes; and how barriers such as sloughs and open areas affect movements of the mice.

An application of common principal component analysis to cranial morphometry of Microtus californicus and M. ochrogaster (Mammalia, Rodentia): further remarks

We welcome Dr Thorpe's interesting discussion (Thorpe, 1988), and we would like to take this opportunity to clarify some points.Both MGPCA (multiple group principal component analysis) and CPCA (common principal component analysis) serve essentially the same purpose, namely estimation of principal components simultaneously in several groups, based on the assumption of equality of principal component directions across groups, while eigenvalues may differ between groups. However, CPCA has the distinct advantage that this assumption can actually be tested, using the (CPC) statistic. In analyses involving more than two variables, it is usually difficult to decide, without a formal test, whether or not the assumption of common directions of principal components is reasonable.There is also a conceptual difficulty with MGPCA. In statistical terms, both methods assume that:(a) a certain set of parameters (namely those determining the eigenvectors) are common to all groups(b) there are sets of parameters (namely p eigenvalues per group) which are specific to each group.CPCA sets up a model that reflects this structure and estimates the parameters accordingly. MGPCA, on the other hand, ignores part (b), at least temporarily, by pooling the variance‐covariance matrices and extracting eigenvectors from the single pooled matrix. This may lead to reasonable results, but there is no guarantee that it will indeed do so. The reader may find a more familiar analog in the fitting of regression lines when data are in groups. If it is assumed that all regression lines are parallel, one should set up an appropriate model based on a single slope parameter common to all groups, and groupwise intercepts. One should then estimate the parameters of this model, and not simply apply a technique which is appropriate in the one‐group case only.