Morphologic lesions in type 2 BVDV infections experimentally induced by strain BVDV2-1373 recovered from a field case

Widespread outbreaks of severe acute BVDV, some associated with hemorrhagic syndrome (HS), were reported in Quebec and Ontario in 1993. These outbreaks caused significant economic hardship in infected herds. In the Ontario outbreak 150 dairy, 600 beef and 100 milk and grain fed veal herds were affected with losses estimated at $40000-$10000 per herd in lost animals, milk production, abortions and genetics. Fever, pneumonia, diarrhea, and sudden death occurred in all age groups of cattle. Abortions were frequently observed in pregnant cattle. The viruses associated with this outbreak were determined to be noncytopathic BVDV from the type 2 genotype. All BVDV2 associated with these outbreaks were noncytopathic. One of the viruses isolated from the Ontario outbreak, BVDV2-1373, was used to experimentally induce HS in 5-6 weeks old colostrum deprived, seronegative calves. All animals developed leukopenia and thrombocytopenia within 6-10 days with some developing bloody diarrhea and becoming moribund. Animals were killed for necropsy between 6 and 11 days postinfection. Histopathologically lesions were similar, but more severe, to those seen early on (within first 9 days after superinfection) in animals with experimentally induced mucosal disease (MD). There were no erosions and ulcerations present in the upper digestive tract. In hemorrhages in the mucosa, virus antigen (VA) was present in macrophages of both the lamina propria and the submucosa and in basal epithelial cells. Cells containing VA were vacuolated and separated from each other. The most severe lesions observed in the digestive tract were in the Peyers patches and were characterized by depletion of lymphocytes and proliferation of crypt cells resulting in crypthyperplasia. Apoptotic cells were present in crypts and areas of lymph follicles where viral antigen was detected. Out of the six animals, VA was present in four animals in the pancreas, three animals in the pituitary and in two animals in the adrenal glands. The results suggest that the pathology resulting from acute infection with a highly virulent noncytopathic BVDV2 differs from the pathology observed in classic mucosal disease.     

Changes in distribution and numbers of CD4+ and CD8+ T-lymphocytes in lymphoid tissues and intestinal mucosa in the early phase of experimentally induced early onset mucosal disease in cattle

Mucosal disease (MD), one sequelae of bovine virus diarrhoea virus (BVDV) infection, causes severe lesions in lymphoid tissues and mucosal surfaces. Lesions are associated with the presence of cytopathogenic (cp) BVDV and initially characterized by apoptotic cell death. The objective of this investigation was to determine if this cell death is mediated only by the cp BVDV, which is known to induce apoptosis in cell culture or if immune-mediated host reactions might also contribute. Early onset MD was experimentally induced in calves by inoculation of persistently viremic calves with a closely related cp BVDV. Calves were euthanized in the early phase of infection between days 5 and 13 post-inoculation and tissues from tonsils, lymph nodes, Peyer's patches, jejunum and colon were collected. Presence of cp BVDV antigen was correlated with distribution of lymphocyte subpopulations in consecutive cryostat sections. In the lymphoid tissues, cp BVDV antigen was predominantly found in the lymphoid follicles. The increase of infected cells with time post-inoculation was paralleled by a decrease of B-lymphocytes and an increase of CD4+ T-lymphocytes. An increased number of CD8+ T-lymphocytes was seen in progressed lesions only. In the intestinal mucosa, initially multifocal, later diffuse infection with cp BVDV was accompanied by a multifocal or diffuse increase of CD4+ T-lymphocytes, respectively. Numbers of IgA+ plasma cells and CD8+ T-lymphocytes were decreased. The common change observed in lymphoid tissues and mucosa was the increase of CD4+ T-lymphocytes in sites with lesions. This might indicate a cell-mediated immune response to the cp BVDV. Besides their helper function to other cells of the immune system, activated CD4+ T-lymphocytes might also exert cytotoxic activity, induce apoptosis in target cells via Fas/Fas ligand binding and thus contribute to the severity of tissue lesions in MD.     

Changes in Distribution and Numbers of CD4+ and CD8+ T‐lymphocytes in Lymphoid Tissues and Intestinal Mucosa in the Early Phase of Experimentally Induced Early Onset Mucosal Disease in Cattle

Mucosal disease (MD), one sequelae of bovine virus diarrhoea virus (BVDV) infection, causes severe lesions in lymphoid tissues and mucosal surfaces. Lesions are associated with the presence of cytopathogenic (cp) BVDV and initially characterized by apoptotic cell death. The objective of this investigation was to determine if this cell death is mediated only by the cp BVDV, which is known to induce apoptosis in cell culture or if immune‐mediated host reactions might also contribute. Early onset MD was experimentally induced in calves by inoculation of persistently viremic calves with a closely related cp BVDV. Calves were euthanized in the early phase of infection between days 5 and 13 post‐inoculation and tissues from tonsils, lymph nodes, Peyer's patches, jejunum and colon were collected. Presence of cp BVDV antigen was correlated with distribution of lymphocyte subpopulations in consecutive cryostat sections. In the lymphoid tissues, cp BVDV antigen was predominantly found in the lymphoid follicles. The increase of infected cells with time post‐inoculation was paralleled by a decrease of B‐lymphocytes and an increase of CD4+ T‐lymphocytes. An increased number of CD8+ T‐lymphocytes was seen in progressed lesions only. In the intestinal mucosa, initially multifocal, later diffuse infection with cp BVDV was accompanied by a multifocal or diffuse increase of CD4+ T‐lymphocytes, respectively. Numbers of IgA+ plasma cells and CD8+ T‐lymphocytes were decreased. The common change observed in lymphoid tissues and mucosa was the increase of CD4+ T‐lymphocytes in sites with lesions. This might indicate a cell‐mediated immune response to the cp BVDV. Besides their helper function to other cells of the immune system, activated CD4+ T‐lymphocytes might also exert cytotoxic activity, induce apoptosis in target cells via Fas/Fas ligand binding and thus contribute to the severity of tissue lesions in MD.     

Distribution of bovine virus diarrhoea viral antigens in the central nervous system of cattle with various congenital manifestations.

Distribution of bovine viral diarrhoea virus (BVDV) antigens in the central nervous system (CNS) of 26 cattle persistently BVDV infected, 11 cattle with mucosal disease (MD), and 32 calves with congenital brain malformations was studied using monoclonal antibodies against BVDV epitopes. In persistently infected cattle and in cattle with MD, a widespread infection of neurons was present. Predilection sites for BVDV antigens were the cerebral cortex and the hippocampus. In calves with congenital encephalopathies, viral antigen-containing neurons could only be detected in the CNS of four animals. From the topographical distribution of BVDV antigens in these four postnatal cases with end-stage lesions, no conclusions could be drawn concerning the pathogenesis of BVDV-induced encephalopathies.     

Distribution of viral antigen and development of lesions after experimental infection of calves with a BVDV 2 strain of low virulence.

To examine the virus-host interaction in subclinical bovine viral diarrhea virus (BVDV) infections, the spread of a BVDV 2 strain of low virulence to different organs and the development of lesions were investigated. Eight colostrum-deprived, clinically healthy, 2-3-month-old calves were intranasally inoculated with 10(6) tissue culture infective dose of the naturally occurring BVDV 2 strain 28508-5 of low virulence, and 2 served as controls. Two calves each were euthanized at days 3, 6, 9, and 13 postinoculation (pi). Representative tissues were processed for histology and immunohistology. Signs of overt clinical disease were absent. However, a mild temperature elevation at days 7 or 8 pi and a moderate decrease of circulating lymphocytes occurred in all inoculated calves. The BVDV antigen was detected at day 3 pi in several lymphoid tissues. At day 6 pi, BVDV antigen was found widespread in lymphoid tissues and multifocally in intestinal epithelial cells but was associated with no or subtle lesions only. At day 9 pi, much less BVDV antigen was detectable, but there was severe depletion of lymphoid tissues. At day 13 pi, BVDV antigen had been cleared from most lymphoid tissues that were at variable phases of depletion and recovery. In conclusion, the BVDV strain of low virulence spread to lymphoid tissues and intestinal epithelial cells but was rapidly eliminated. Transient depletion of lymphoid tissues was followed by recovery.     

Ultrastructural changes in follicles of small-intestinal aggregated lymphoid nodules in early and advanced phases of experimentally induced mucosal diseases in calves

OBJECTIVE: To investigate ultrastructural changes in follicles of small-intestinal aggregated lymphoid nodules (Peyer's patches) of calves with early and advanced phases of experimentally induced mucosal disease (MD). ANIMALS: Twenty 2.5- to 7-month-old Holstein-Friesian calves (11 females, 9 males). PROCEDURE: MD was induced in 13 of 18 calves that were persistently viremic with bovine viral diarrhea virus (BVDV). Eight of the 13 calves were euthanatized before the onset of clinical signs of MD, and 5 were euthanatized after becoming moribund with MD. Five persistently viremic calves and 2 calves without BVDV served as controls. Specimens of small-intestinal aggregated lymphoid nodules were prepared for transmission electron microscopy. RESULTS: The ultrastructure of follicles of small-intestinal aggregated lymphoid nodules from healthy calves was consistent with that in sheep. In the early phase of MD, changes were characterized by numerous apoptotic lymphocytes and macrophages with apoptotic bodies. In more advanced lesions, affected lymphoid follicles consisted of macrophages and variable numbers of follicular dendritic cells (FDC), whereas others did not contain FDC. In moribund calves, small follicles consisting predominantly of FDC and follicles with central cavities surrounded by macrophages, and few neutrophils were observed. CONCLUSIONS AND CLINICAL RELEVANCE: The ultrastructural changes in lymphoid follicles of small-intestinal aggregated lymphoid nodules indicate apoptosis of lymphocytes as an initial event. The development of small follicles consisting predominantly of FDC or the complete loss of follicular architecture in advanced phases of MD is determined by the intensity of apoptosis of lymphocytes, the capacity of the macrophages for uptake, and the reorganization of a stromal network.     

Apoptosis inhibitors delay the cytopathic effect of bovine viral diarrhoea virus (BVDV)

Based on their action in cell culture, two biotypes of bovine viral diarrhoea virus (BVDV) can be distinguished. The noncytopathic (ncp) BVDV isolated from persistently infected animals cause no visible damage to cultured bovine cells. In contrast, cytopathic (cp) BVDV induces severe damage and apoptosis in cell cultures. Cp BVDV can be isolated from cattle suffering from mucosal disease (MD) and is associated with the severe lesions that primarily affect the gastrointestinal tract. To get an insight into the molecular events during BVDV induced cytopathic effect (CPE), the effect of three chemical reagents (3-aminobenzamide, ascorbic acid and N-acetyl-leucyl-leucyl-methional) with completely different mode of actions in infected cells was analysed. All three substances were able to delay the cytopathic effect induced in permissive bovine cells.     

Viral antigen distribution in the respiratory tract of cattle persistently infected with bovine viral diarrhea virus subtype 2a

Tissues were obtained at necropsy from the nasal vestibule, turbinates, nasopharynx, trachea, tracheobronchial bifurcation, and lung from each of 10 clinically healthy calves persistently infected (PI) with bovine viral diarrhea virus (BVDV) serotype 2a. Tissues from the nasal vestibule were obtained by biopsy from five additional PI calves. Formalin-fixed tissues were processed for immunohistochemistry to localize the distribution of BVDV throughout the respiratory tract. Antigen distribution and intensity were subjectively evaluated. Throughout the respiratory tract, mononuclear leukocytes, vascular smooth muscle, and endoneural and perineural cells had BVDV immunoreactivity (BVDV-IR). Multifocally, squamous and ciliated columnar epithelium throughout the respiratory tract contained weak to moderate BVDV antigen. Viral antigen was not seen in goblet cells. BVDV-IR in mixed tubuloalveolar glands of the nasal cavity was weak to strong in serous secretory cells and ductular epithelium. Chondrocytes of the concha often contained BVDV antigen diffusely. Nasal mucus-secreting and tracheobronchial glands multifocally contained weak viral signal. In all cases, alveolar macrophages had moderate to strong BVDV-IR, whereas BVDV-IR in alveolar epithelial cells was weak to moderate. BVDV was present in interalveolar leukocytes and mesenchymal cells. Results indicate that serous secretions of the nasal cavity, productive viral infection of epithelium, and infected leukocytes in respiratory secretions are likely major sources of infectious BVDV from PI calves. The presence of BVDV antigen in respiratory epithelium is, at least, indirect support for the notion that this virus predisposes PI cattle to secondary microbial infections.     

Distribution of viral antigen and tissue lesions in persistent and acute infection with the homologous strain of noncytopathic bovine viral diarrhea virus.

Viral distribution and lesions were compared between calves born with persistent infection (PI) and calves acutely infected with the same bovine viral diarrhea virus (BVDV) isolate. Two PI calves from 1 dairy herd were necropsied. The PI viruses from these calves were isolated, characterized by sequencing, and found to be identical. This virus strain, designated BVDV2-RS886, was characterized as a noncytopathic (ncp) type 2 BVDV. To establish acute infections, BVDV2-RS886 was used to inoculate clinically healthy, seronegative calves which were 3 weeks to 3 months old. Nine calves received 10(6)-10(7) tissue culture infective dose of BVDV2-RS886 intranasally. Four additional age-matched animals served as noninfected controls. Infected calves were necropsied at 3, 6, 9, or 13 days postinoculation (dpi). Viral antigen was detected by immunohistochemistry in frozen sections, and lesions were evaluated in hematoxylin eosin-stained paraplast sections. In the PI calves, a wide distribution of viral antigen was found in all tissues and was not associated with lesions. In the acutely infected calves, viral antigen was widespread in lymphoid tissues at 6 dpi but had been mostly eliminated at 9 and 13 dpi. Depletion of lymphoid tissues was seen at 6, 9, and 13 dpi and repopulation at 9 and 13 dpi. In 1 of the calves at 13 dpi, severe arteritis was present in lymph nodes and myocardium. This comparison shows that an ncp BVDV strain that causes no lesions in PI animals is able to induce marked depletion of lymphoid tissues in calves with acute infection. Therefore, the failure to eliminate PI cattle from a herd causes problems not only in pregnant cattle but may also affect other age groups.     

Experimental infection of calves with bovine viral diarrhea virus genotype II (NY-93).

To ascertain the virulence of bovine viral diarrhea virus (BVDV) genotype II, isolate NY-93 was inoculated intranasally into 3 calves, 2 of which were treated with a synthetic glucocorticoid prior to and after virus inoculation. Anorexia, fever (up to 42 C), dyspnea, and hemorrhagic diarrhea developed 6 days after intranasal inoculation with BVDV NY-93. The condition of all calves deteriorated further until the end of the study on day 14 postinoculation. The most significant postmortem macroscopic changes in all calves were limited to the gastrointestinal tract and consisted of moderate to severe congestion of the mucosa with multifocal hemorrhages. Microscopic lesions found in the gastrointestinal tract were similar to those observed in mucosal disease, including degeneration and necrosis of crypt epithelium and necrosis of lymphoid tissue throughout the ileum, colon, and rectum. The basal stratum of the epithelium of tongue, esophagus, and rumen had scattered individual necrotic cells. Spleen and lymph nodes had lymphocytolysis and severe lymphoid depletion. Severe acute fibrinous bronchopneumonia was present in dexamethasone-treated calves. Abundant viral antigen was detected by immunohistochemistry in the squamous epithelium of tongue, esophagus, and forestomachs. BVDV antigen was prominent in cells of the media of small arteries and endothelial cells. The presence of infectious virus in tissues correlated with an absence of circulating neutralizing antibodies. These findings highlight the potential of BVDV genotype II to cause severe disease in normal and stressed cattle.     

BOVINE HERPESVIRUS 1 - INFECTION OF THE UPPER ALIMENTARY TRACT OF CATTLE AND ITS ASSOCIATION WITH A SEVERE MORTALITY

SUMMARY A severe cattle mortality in which 132 out of 340 animals died on a property in southern Queensland was investigated. Clinical signs shown by affected animals included fever, inappetance, depression, lethargy, salivation, diarrhoea, ataxia, and ulceration of the oral cavity. The most common lesions seen at autopsy of 6 affected animals were ulceration of the tongue, gums, dental pad and buccal mucosa, linear ulceration of the caudal third of the oesophagus, mild catarrhal enteritis and necrosis of lymph nodes draining areas of ulceration. Bovine herpesvirus type 1 (BHV 1) was isolated from 3 out of 5 animals from which virus isolation was attempted. BHV 1 was recovered from oesophageal ulcers, retropharyngeal lymph nodes, blood clot, and swabs from ulcers in the oral cavity but not from spleen, liver or mesenteric lymph node. Serum neutralising (SN) antibody to BHV 1 was detected in 4 out of 12 affected animals in the second of paired serum samples but not in the first. Mucosal Disease (MD) virus was not recovered from any of 17 animals from which isolation was attempted but moderate MD SN titres, without a rise on paired sera, were detected in affected animals. Fever, depression, inappetance, ulceration of the upper alimentary tract, and adrenal necrosis were produced in 2 susceptible animals following inoculation with third passage cell culture fluid containing BHV 1. A serological response to BHV 1, but not to MD virus was detected in one of the cattle infected experimentally.     

Development of a neutralising antibody response to an inoculated cytopathic strain of bovine virus diarrhoea virus

Fourteen clinically healthy cattle that were persistently infected with non-cytopathic bovine virus diarrhoea virus (BVDV) and three BVDV-free cattle were inoculated with one of three cytopathic BVDV strains. Mucosal disease developed in 12 of the viraemic cattle, resulting in a moribund condition 17 to 99 days after inoculation. Two of the viraemic cattle remained clinically healthy until the end of the experiment, 14 months after inoculation. The BVDV-free cattle did not develop clinical signs after inoculation. From each cow with mucosal disease a noncytopathic and a cytopathic BVDV strain were isolated from tissue specimens collected post mortem. All the cattle developed moderate to high levels of neutralising antibodies against the cytopathic BVDV strain with which they were inoculated. The antibodies from 10 of the 12 cattle with mucosal disease did not react with the cytopathic BVDV strains isolated post mortem, and antibodies from none of them reacted with the non-cytopathic BVDV isolates. Antibody responses to the inoculated BVDV strains developed earlier in the viraemic cattle than in the BVDV-free cattle.     

Detection of bovine viral diarrhoea virus in specimens from cattle in South Africa and possible association with clinical disease

Studies covering all aspects of bovine viral diarrhoea virus (BVDV) have been conducted in several countries in Europe, Asia and America. In southern Africa, more information is required about the nature of BVDV infection, the prevalence of different strains and the economic importance of the disease. The presence of BVDV in southern Africa has been known since the early 1970s through serological surveys but few reports confirming its presence by virus isolation and correlation with clinical disease are available. Specimens (n = 312) collected in 1998/99, from live and dead cattle from different farming systems, were obtained from private practitioners, feedlot consultants and abattoirs throughout the country. Specimens (n = 37) from African buffaloes (Syncerus caffer) in the Kruger National Park were also included. All specimens were processed for virus isolation in cell culture with confirmation by means of immunofluorescent antibody tests and some also by means of an antigen capture ELISA. BVDV was isolated from 15 (4.7%) cattle and were all noncytopathic biotypes. BVDV was not detected in 37 lymph nodes obtained from buffaloes in the Kruger National Park. Of the clinical signs in cattle from which virus were isolated, respiratory signs was the most frequent (10/15), followed by diarrhoea (5/15). Abortion, congenital malformations, haemorrhagic diarrhoea and poor growth were also included as criteria for selection of animals for specimen collection, but no BVD viruses were isolated from cattle manifesting these clinical signs.     

Response of cattle persistently infected with bovine virus diarrhoea virus to bovine leukosis virus

Six cattle persistently infected with bovine virus diarrhoea virus (BVDV) and seronegative, and two control, virus negative seropositive cattle were inoculated with lymphocytes infected with bovine leukosis virus (BLV). The two controls produced a normal immune response to BLV, developing antibodies at four and five weeks after inoculation. Two of the six cattle persistently infected with BVDV developed a strong antibody response by six weeks after inoculation with BLV. Four developed a depressed response to BLV, characterised in three by a 'hooking' reaction in the immunodiffusion test which persisted in successive bleedings but was interspersed occasionally by a weak positive reaction. In one of these animals, a series of 'hooking' reactions was followed by a number of negative results. The fourth animal remained serologically negative until 16 weeks after inoculation when a 'hooking' reaction was observed followed by a series of negative results. BLV was isolated from all the cattle persistently infected with BVDV at 42 or 58 weeks after inoculation regardless of whether the serum samples gave negative, 'hooking', weak positive or positive reactions in the immunodiffusion test. BLV was consistently isolated from the nasal secretions of a steer which was BVDV negative but seropositive. The possibility of decreased immune responsiveness to BLV in animals persistently infected with BVDV should be considered when formulating regulations governing the testing of animals for freedom from BLV.     

Experimental infection of rabbits with bovine viral diarrhoea virus by a natural route of exposure

Bovine viral diarrhoea virus (BVDV) is an important pathogen of cattle that can naturally infect a wide range of even-toed ungulates. Non-bovine hosts may represent reservoirs for the virus that have the potential to hamper BVDV eradication programs usually focused on cattle. Rabbits are very abundant in countries such as the United Kingdom or Australia and are often living on or near livestock pastures. Earlier reports indicated that rabbits can propagate BVDV upon intravenous exposure and that natural infection of rabbits with BVDV may occur but experimental proof of infection of rabbits by a natural route is lacking. Therefore, New Zealand White rabbits were exposed to a Scottish BVDV field strain intravenously, oro-nasally and by contaminating their hay with virus. None of the animals showed any clinical signs. However, the lymphoid organs from animals sacrificed at day five after exposure showed histological changes typical of transient infection with pestivirus. Most organ samples and some buffy coat samples were virus positive at day five but saliva samples remained negative. Development of antibodies was observed in all intravenously challenged animals, in all of the nebulised group and in four of six animals exposed to contaminated hay. To our knowledge this is the first report of BVDV propagation in a species other than ruminants or pigs after exposure to the virus by a natural route. However, to assess the role of rabbits as a potential reservoir for BVDV it remains to be determined whether persistent infection caused by intra-uterine infection is possible and whether BVDV is circulating in wild rabbit populations.     

Cytopathic bovine viral diarrhea viruses (BVDV): emerging pestiviruses doomed to extinction

Bovine viral diarrhea virus (BVDV), a Flaviviridae pestivirus, is arguably one of the most widespread cattle pathogens worldwide. Each of its two genotypes has two biotypes, non-cytopathic (ncp) and cytopathic (cp). Only the ncp biotype of BVDV may establish persistent infection in the fetus when infecting a dam early in gestation, a time point which predates maturity of the adaptive immune system. Such fetuses may develop and be born healthy but remain infected for life. Due to this early initiation of fetal infection and to the expression of interferon antagonistic proteins, persistently infected (PI) animals remain immunotolerant to the infecting viral strain. Although only accounting for some 1% of all animals in regions where BVDV is endemic, PI animals ensure the viral persistence in the host population. These animals may, however, develop the fatal mucosal disease, which is characterized by widespread lesions in the gastrointestinal tract. Cp BVD virus, in addition to the persisting ncp biotype, can be isolated from such animals. The cp viruses are characterized by unrestrained genome replication, and their emergence from the persisting ncp ones is due to mutations that are unique in each virus analyzed. They include recombinations with host cell mRNA, gene translocations and duplications, and point mutations. Cytopathic BVD viruses fail to establish chains of infection and are unable to cause persistent infection. Hence, these viruses illustrate a case of “viral emergence to extinction” – irrelevant for BVDV evolution, but fatal for the PI host.     

Detection of viral antigen in placenta and fetus of cattle acutely infected with bovine viral diarrhea virus.

The reproductive organs and fetuses of seven Norwegian Red heifers were investigated for the presence of bovine viral diarrhea virus (BVDV) antigen during the time of initial transplacental transmission of the virus. The heifers were inoculated with a noncytopathogenic BVDV at day 85/86 of gestation and were slaughtered at day 7, 10, 14, 18, or 22 postinoculation (pi). Cryostat sections of uterus, ovaries, placentomes, intercotyledonary fetal membranes, and fetal organs were examined using immunohistochemical techniques. A double immunofluorescence technique was used to identify cells that showed staining with antibodies against the leukocyte common antigen CD45 or the intermediate filament vimentin and BVDV antigens. The earliest stage of infection at which BVDV antigen could be detected in the fetuses was 14 days pi. At this stage, BVDV antigen was detected in cells of mesenchymal origin in the lungs and in large cells that morphologically resembled immature megakaryocytes in the liver. In the intercotyledonary fetal membranes and in the placentomes, BVDV antigen was not detected until 18 and 22 days pi, respectively. BVDV antigen was not detected in maternal tissue from any of the heifers. The present results indicate that fetal infection with BVDV can take place without preceding or simultaneous high concentrations of BVDV in uterus or placenta of acutely infected heifers.