1
|
Neupane R, Aryal A, Haeussermann A, Hartung E, Pinedo P, Paudyal S. Evaluating machine learning algorithms to predict lameness in dairy cattle. PLoS One 2024; 19:e0301167. [PMID: 39024328 PMCID: PMC11257334 DOI: 10.1371/journal.pone.0301167] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [MESH Headings] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 03/11/2024] [Accepted: 07/05/2024] [Indexed: 07/20/2024] Open
Abstract
Dairy cattle lameness represents one of the common concerns in intensive and commercial dairy farms. Lameness is characterized by gait-related behavioral changes in cows and multiple approaches are being utilized to associate these changes with lameness conditions including data from accelerometers, and other precision technologies. The objective was to evaluate the use of machine learning algorithms for the identification of lameness conditions in dairy cattle. In this study, 310 multiparous Holstein dairy cows from a herd in Northern Colorado were affixed with a leg-based accelerometer (Icerobotics® Inc, Edinburg, Scotland) to obtain the lying time (min/d), daily steps count (n/d), and daily change (n/d). Subsequently, study cows were monitored for 4 months and cows submitted for claw trimming (CT) were differentiated as receiving corrective claw trimming (CCT) or as being diagnosed with a lameness disorder and consequent therapeutic claw trimming (TCT) by a certified hoof trimmer. Cows not submitted to CT were considered healthy controls. A median filter was applied to smoothen the data by reducing inherent variability. Three different machine learning (ML) models were defined to fit each algorithm which included the conventional features (containing daily lying, daily steps, and daily change derived from the accelerometer), slope features (containing features extracted from each variable in Conventional feature), or all features (3 simple features and 3 slope features). Random forest (RF), Naive Bayes (NB), Logistic Regression (LR), and Time series (ROCKET) were used as ML predictive approaches. For the classification of cows requiring CCT and TCT, ROCKET classifier performed better with accuracy (> 90%), ROC-AUC (> 74%), and F1 score (> 0.61) as compared to other algorithms. Slope features derived in this study increased the efficiency of algorithms as the better-performing models included All features explored. However, further classification of diseases into infectious and non-infectious events was not effective because none of the algorithms presented satisfactory model accuracy parameters. For the classification of observed cow locomotion scores into severely lame and moderately lame conditions, the ROCKET classifier demonstrated satisfactory accuracy (> 0.85), ROC-AUC (> 0.68), and F1 scores (> 0.44). We conclude that ML models using accelerometer data are helpful in the identification of lameness in cows but need further research to increase the granularity and accuracy of classification.
Collapse
Affiliation(s)
- Rajesh Neupane
- Department of Animal Science, Texas A&M University, College Station, Texas, United States of America
| | - Ashrant Aryal
- Department of Construction Science, Texas A&M University, College Station, Texas, United States of America
| | | | - Eberhard Hartung
- Department of Agricultural Engineering, Kiel University, Kiel, Germany
| | - Pablo Pinedo
- Department of Animal Sciences, Colorado State University, Fort Collins, Colorado, United States of America
| | - Sushil Paudyal
- Department of Animal Science, Texas A&M University, College Station, Texas, United States of America
| |
Collapse
|
2
|
Stoddard G, Cramer G. Observational study evaluating the association of hoof trimming with dairy cattle behavior and milk yield in Canada and the United Kingdom. JDS COMMUNICATIONS 2024; 5:322-326. [PMID: 39220847 PMCID: PMC11365343 DOI: 10.3168/jdsc.2023-0493] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Track Full Text] [Figures] [Subscribe] [Scholar Register] [Received: 09/15/2023] [Accepted: 12/20/2023] [Indexed: 09/04/2024]
Abstract
The objective of this observational study was to describe the associations between preventive hoof trimming (HT) of cows with no lesions and several behavior- and productivity-outcome measures including activity, lying behavior, and milk yield. A convenience sample of 4 freestall farms in the United Kingdom (n = 2) and Canada (n = 2) were recruited for the study. Inclusion criteria for herds included having a regular trim schedule and use of a specific pedometer system. Cows were trimmed as per their farm's regular protocol and schedule and all HT events were recorded. Only cows with no lesions present on the first recorded HT event were used in the analysis. Activity (steps/d), resting time (min/d), and resting bouts (bouts/d) were recorded daily by pedometers and total daily milk yield (kg/d) was recorded by the farm's milk recording system. Cow-level data such as DIM, breed, and parity were collected from the farm's herd management software. A generalized linear mixed repeated measures model with an exchangeable correlation structure was built to compare the outcome for 7 d following HT to baseline measures for each outcome of interest. For each outcome the average from the 5 d before HT was used as a baseline. All models included the baseline status, lactation number, season, farm, and DIM as covariates. Days from HT and its interaction with farm were included in all models. A total of 1,573 cows with no lesion recorded in their first HT were included in the analysis. Activity and resting time were inversely related on the day of HT, with activity increasing on 3 out of 4 farms (+59 to +84 steps/d) and resting time decreasing (-28 to -52 min/d). Following the day of HT, activity decreased to levels lower than those reported at baseline and resting time increased compared with baseline on most farms. There was a reduction in milk yield on the day of HT on 3 out of 4 farms and a reduction in average yield across the 7 d following trimming for all farms (range: 0.6 to 1.3 kg/d). Results were found to be farm dependent. The impact of the degree of change from baseline in behaviors and milk yield reported in this study on welfare of the cow is unclear, but the results suggest that alterations in the outcomes evaluated can persist beyond the day of trimming compared with baseline, necessitating further investigation.
Collapse
Affiliation(s)
- G. Stoddard
- Department of Veterinary Population Medicine, University of Minnesota, Saint Paul, MN 55108
- Purina Animal Nutrition, Arden Hills, MN 55126
| | - G. Cramer
- Department of Veterinary Population Medicine, University of Minnesota, Saint Paul, MN 55108
| |
Collapse
|
3
|
Roche SM, Renaud DL, Saraceni J, Kelton DF, DeVries TJ. Invited review: Prevalence, risk factors, treatment, and barriers to best practice adoption for lameness and injuries in dairy cattle-A narrative review. J Dairy Sci 2024; 107:3347-3366. [PMID: 38101730 DOI: 10.3168/jds.2023-23870] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Key Words] [MESH Headings] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 06/14/2023] [Accepted: 11/08/2023] [Indexed: 12/17/2023]
Abstract
Lameness and leg injuries are both painful and prevalent across the dairy industry, and are a major welfare concern. There has been a considerable amount of research focused on investigating the risk factors associated with lameness and injuries and how they might be prevented and treated. The objectives of this narrative review were to summarize herd-level prevalence estimates, risk factors, strategies for prevention, control, and treatment of these conditions, and the barriers to best practice adoption for lameness and injuries on dairy farms. There is a relatively high within-herd prevalence of lameness on dairy farms globally, with a recent systematic review estimating the mean prevalence at 22.8%. Similarly, there is a relatively high prevalence of hock injuries, with within-herd estimates ranging from 12% to 81% of cows affected. Knee and neck injuries have been reported to be less common, with 6% to 43% and 1% to 33%, respectively. Numerous risk factors have been associated with the incidence of lameness, notably housing (e.g., access to pasture, bedding depth, bedding type, flooring type, stall design), management (e.g., stall cleanliness, frequency of trimming, holding times, stocking density), and cow-level (e.g., body condition, parity, injured hocks) factors. Risk factors associated with hock injuries can be similarly classified into housing (e.g., bedding type and depth, outdoor access, parlor type, stall design), management (e.g., bedding depth, cleanliness), and cow (e.g., parity, days in milk, lameness) factors. Key preventative approaches for lameness include routine preventative and corrective hoof trimming, improving hoof cushioning and traction through access to pasture or adding rubber flooring, deep-bedded stalls, sand bedding, ensuring appropriate stocking densities, reduced holding times, and the frequent use of routine footbaths. Very little research has been conducted on hock, knee, and neck injury prevention and recovery. Numerous researchers have concluded that both extrinsic (e.g., time, money, space) and intrinsic (e.g., farmer attitude, perception, priorities, and mindset) barriers exist to addressing lameness and injuries on dairy farms. There are many diverse stakeholders in lameness and injury management including the farmer, farm staff, veterinarian, hoof trimmer, nutritionist, and other advisors. Addressing dairy cattle lameness and injuries must, therefore, consider the people involved, as it is these people who are influencing and implementing on-farm decisions related to lameness prevention, treatment, and control.
Collapse
Affiliation(s)
- S M Roche
- Department of Population Medicine, University of Guelph, Guelph, ON, Canada, N1G 2W1; ACER Consulting Ltd., Guelph, ON, Canada, N1G 5L3
| | - D L Renaud
- Department of Population Medicine, University of Guelph, Guelph, ON, Canada, N1G 2W1
| | - J Saraceni
- ACER Consulting Ltd., Guelph, ON, Canada, N1G 5L3
| | - D F Kelton
- Department of Population Medicine, University of Guelph, Guelph, ON, Canada, N1G 2W1
| | - T J DeVries
- Department of Animal Biosciences, University of Guelph, Guelph, ON, Canada, N1G 2W1.
| |
Collapse
|
4
|
Werema CW, Hoekstra F, Laven LJ, Müller KR, Gifford D, Laven RA. Investigating the effect of prophylactic claw trimming on the interval between calving and first observed elevated locomotion score in pasture-based dairy cows. N Z Vet J 2023; 71:295-305. [PMID: 37492960 DOI: 10.1080/00480169.2023.2238654] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Key Words] [MESH Headings] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 09/23/2022] [Accepted: 06/22/2023] [Indexed: 07/27/2023]
Abstract
AIMS To evaluate, in a pasture-based dairy herd, the response to a three-time point hoof trimming regime on lameness incidence and time from calving to observation of an elevated locomotion score (LS). METHODS This study was conducted on a 940-cow spring-calving herd in New Zealand's North Island between May 2018 and May 2019. Cows (n = 250) were randomly allocated to the hoof trimming group, with the remainder assigned to the non-trim cohort. One trained professional hoof trimmer used the five-step Dutch method to trim the hind feet of the trimming group. Throughout the subsequent production season, the whole herd was locomotion-scored fortnightly using the 4-point (0-3) Dairy NZ lameness score. Kaplan-Meier survival curves were used to assess the univariable effect of trimming on the interval between calving and first LS of ≥ 2 and first LS ≥ 1. A multivariable Cox proportional hazards regression was used to further evaluate the effect of trimming on time to elevated LS. RESULTS Mean lameness (LS ≥ 2) prevalence was 2.6%, with 30% of cows having ≥ 4 observations during the study period when at least one LS was ≥ 2. For LS ≥ 1, mean prevalence was 40%, with 98.6% of cows having ≥ 4 observations during the study period when at least one LS was ≥ 1 during lactation. Hoof trimming had no apparent effect on the incidence of clinical lameness (LS ≥ 2) (trimmed vs. non-trimmed: 33.2% vs. 28.8%, respectively), but for LS ≥ 1, there was a small decrease in the incidence of LS ≥ 1 (trimmed vs. non-trimmed: 96.9% vs. 99.3%, respectively). The hazard of a cow having a first observed LS ≥ 2 in the control group was 0.87 (95% CI = 0.66-1.14) times that of the trimmed group; however, the hazard of a cow having a first LS ≥ 1 was 1.60 (95% CI = 1.37-1.88) times higher in the control than in the trimmed group. CONCLUSION AND CLINICAL RELEVANCE On this farm, prophylactic hoof trimming had no clinically relevant impact on the incidence of clinical lameness and was not associated with clinically beneficial reductions in time to first observed LS ≥ 2. This may be because claw horn imbalance was not pronounced on this farm, with 53% of cows needing no trim on either hind limb on the first trimming occasion. Further research on the response to prophylactic trimming in pasture-based dairy cattle is required.
Collapse
Affiliation(s)
- C W Werema
- Tāwharau Ora - School of Veterinary Science, Massey University, Palmerston North, New Zealand
| | - F Hoekstra
- Dairy Hoofcare Institute, Ashburton, New Zealand
| | - L J Laven
- Tāwharau Ora - School of Veterinary Science, Massey University, Palmerston North, New Zealand
| | - K R Müller
- Tāwharau Ora - School of Veterinary Science, Massey University, Palmerston North, New Zealand
| | - D Gifford
- Tāwharau Ora - School of Veterinary Science, Massey University, Palmerston North, New Zealand
| | - R A Laven
- Tāwharau Ora - School of Veterinary Science, Massey University, Palmerston North, New Zealand
| |
Collapse
|
5
|
A modified functional hoof trimming technique reduces the risk of lameness and hoof lesion prevalence in housed dairy cattle. Prev Vet Med 2021; 195:105463. [PMID: 34416653 DOI: 10.1016/j.prevetmed.2021.105463] [Citation(s) in RCA: 6] [Impact Index Per Article: 2.0] [Reference Citation Analysis] [Abstract] [Key Words] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 06/01/2021] [Revised: 07/29/2021] [Accepted: 08/09/2021] [Indexed: 12/26/2022]
Abstract
The objective of this study was to compare the impact of the functional hoof trimming (HT) method to an adaptation that results in increased modelling of the weight bearing claw on time to lameness and lesion prevalence in housed dairy cows. A total of 418 non-lame cows at early and late lactation were enrolled in the study and randomly allocated into 3 groups: functional HT (TRIM1), adaptation method (TRIM2), and an untrimmed group (CON). Locomotion scores (LS; 5-point scale) and animal-based welfare measures were assessed monthly from enrolment until the following 270 days in milk. Cows were considered lame when 2 consecutive LS = 3, or any assessment with a score of 4 or 5. Kaplan-Meier analysis was used to evaluate the difference in time to lameness between the groups. Multivariable CC Cox regression models were fitted to assess associations between covariates and time to lameness event, while associated factors for lesion prevalence were estimated using logistic regression models. The incidence rate of lameness was 29.3 cases/100/month, with corresponding rates of 28.7, 15.8 and 42.8 cases/100/months, in TRIM1, TRIM2 and CON, respectively. Time to first lameness event was higher in TRIM2 (mean ± S.E; 8.26 ± 0.16, P = 0.03) than CON (7.32 ± 0.2) and tended to be higher than TRIM1 (7.83 ± 0.19; P = 0.07). The risk of new lameness event was higher in cows with low BCS (≤2.5) (Hazard ratio; HR = 1.5; 95 % CI 1.1-2.1), while trimming all feet was a protective factor (HR = 0.58; 95 % CI 0.38-0.90). The prevalence of hoof lesions in the study population was 36.3 %, with the highest prevalence in CON (45.6 %), followed by TRIM1 (34.6 %) and TRIM2 (28.1 %). Seventy-eight per cent of hoof lesions were non-infectious and the majority of cows with hoof lesion in TRIM1 (72.3 %), CON (74.6 %) and TRIM2 (55.6 %) were lame. The odds of having a non-infectious hoof lesion were higher in CON (OR = 2.1; 95 % CI 1.2-3.7) compared to TRIM2, and in cows with hock injury (OR = 3.1; 95 % 1.1-9.7) relative to healthy hock, and tended to be higher in cows with low BCS (P = 0.09) relative to those with moderate BCS. Our results suggest that the adaptation to the functional method reduced the risk of lameness during lactation, however, proper management of under-conditioned cows is equally important.
Collapse
|
6
|
Lüchinger I, Pieper L, Nuss K. Functional foot trimming to balance load distribution between the paired forelimb claws in dairy cows: An experimental study. J Dairy Sci 2021; 104:4803-4812. [PMID: 33516555 DOI: 10.3168/jds.2020-19066] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Key Words] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 06/10/2020] [Accepted: 11/05/2020] [Indexed: 11/19/2022]
Abstract
In standing dairy cows, the medial claws carry most of the load of the forelimb. The goal of our experimental study was to investigate how even load distribution can be achieved between the paired forelimb claws through foot trimming. We measured vertical ground reaction forces, mean and maximum pressures, and contact areas under the sound forelimb claws of 28 nonlame dairy cows using a force plate and a pressure plate. The experimental setup consisted of 3 measurement situations: (1) baseline measurements made after routine foot trimming, (2) measurements made with plywood plates, 3 mm in thickness, attached to both lateral claws, and (3) measurements made with the same plywood plates attached to both medial claws. After routine foot trimming, the medial claws carried 70.1% of the entire forelimb load with its palmar zone bearing 54.0% of the entire forelimb load. Even load distribution between the paired claws was achieved when the plywood plates were attached to the lateral claws, whereas plates on the medial claws increased the load discrepancy. The pressures on the medial claws were reduced by attaching the plywood plates to the lateral claws but remained uneven between paired claws and claw zones. The palmar zone of the medial claws was the region subjected to the greatest force and highest pressures in all situations. A drawback of the study was that the plywood plates increased the contact area and therefore reduced the mean and maximum pressures in the claws they were attached to. We concluded that equal load distribution between the forelimb claws was achieved in our setting when the lateral claws were left 3 mm higher than the medial claws.
Collapse
Affiliation(s)
- Isabelle Lüchinger
- Department of Farm Animals, Vetsuisse Faculty, University of Zurich, CH-8057 Zurich, Switzerland
| | - Laura Pieper
- Department of Farm Animals, Vetsuisse Faculty, University of Zurich, CH-8057 Zurich, Switzerland
| | - Karl Nuss
- Department of Farm Animals, Vetsuisse Faculty, University of Zurich, CH-8057 Zurich, Switzerland.
| |
Collapse
|
7
|
Matias BF, Filho LFCC, Rego FCA, Barreto JVP, Reis LSLS, Queiroz ATZ, Queiroz GR. Prevalence of foot disorders in lactating Jersey cows raised in semi-confinement. Vet World 2020; 13:2612-2617. [PMID: 33487978 PMCID: PMC7811547 DOI: 10.14202/vetworld.2020.2612-2617] [Citation(s) in RCA: 1] [Impact Index Per Article: 0.3] [Reference Citation Analysis] [Abstract] [Key Words] [Track Full Text] [Download PDF] [Journal Information] [Subscribe] [Scholar Register] [Received: 06/23/2020] [Accepted: 11/02/2020] [Indexed: 12/04/2022] Open
Abstract
Background and Aim: To establish prevention strategies, recording the prevalence of foot injuries within a herd should be the starting point in determining the risk factors involved in digital diseases. This study aimed to assess the prevalence of claw disorders in lactating Jersey cows raised in a semi-confinement system. Materials and Methods: Five hundred and eighty-four digits were examined from 73 lactating Jersey cows. The lameness score system was used to assess each cow while walking and on standing position, and digital lesions were evaluated with the cows restrained in a hydraulic cattle chute. Results: The prevalence of digital lesions was 93.1%. Among the 68 affected cows, only 21 were lame. Of the 584 digits examined, 970 lesions were recorded, corresponding to 13.3% lesions per cow and 1.66% injuries per digit. Forty-eight cows (65.7%) had lesions in all digits, and 92.8% of digits had at least one lesion. Lesions in digits of fore limbs were more common (55.3%) (p<0.0001) than those of hind limbs (44.7%). Foot injuries in medial digits of the fore limbs were more prevalent (56.8%) (p<0.001) than in the lateral digits (43.2%). The lesions’ occurrence was similar in both medial and lateral digits of the hind limbs (p=0.8347). The primary diseases observed were heel horn erosion (53.8%), white line disease (19.3%), and double sole (12.4%), which together accounted for 92.4% and 84.9% of foot disorders diagnosed in the fore and hind limbs, respectively. Other digital diseases occurred less frequently. Conclusion: The prevalence of foot disorders in lactating Jersey cows raised in semi-confinement systems was high. This could be due to the lack of preventive trimming, infectious diseases, and nutritional problems.
Collapse
Affiliation(s)
- B F Matias
- Master´s Degree in Animal Health and Production, Pitágoras Unopar University, Arapongas, PR, Brazil
| | - L F C Cunha Filho
- Master´s Degree in Animal Health and Production, Pitágoras Unopar University, Arapongas, PR, Brazil
| | - F C A Rego
- Master´s Degree in Animal Health and Production, Pitágoras Unopar University, Arapongas, PR, Brazil
| | - J V P Barreto
- Master´s Degree in Animal Health and Production, Pitágoras Unopar University, Arapongas, PR, Brazil
| | - L S L S Reis
- University of Western Paulista, Sao Paulo, Brazil
| | - A T Z Queiroz
- Master´s Degree in Animal Health and Production, Pitágoras Unopar University, Arapongas, PR, Brazil
| | - G R Queiroz
- Master´s Degree in Animal Health and Production, Pitágoras Unopar University, Arapongas, PR, Brazil
| |
Collapse
|
8
|
Stilwell GT, Ferrador AM, Santos MS, Domingues JM, Carolino N. Use of topical local anesthetics to control pain during treatment of hoof lesions in dairy cows. J Dairy Sci 2019; 102:6383-6390. [PMID: 31030913 DOI: 10.3168/jds.2018-15820] [Citation(s) in RCA: 10] [Impact Index Per Article: 2.0] [Reference Citation Analysis] [Abstract] [Key Words] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 10/09/2018] [Accepted: 03/12/2019] [Indexed: 11/19/2022]
Abstract
Hoof pathologies in dairy cows have a major effect on both production and animal welfare. Trimming of excess or diseased hoof tissue is essential for the treatment of many of these conditions. Trimming hoof lesions can cause severe pain, resulting in adverse behavioral responses with risk for animal and human safety. Interventions are usually carried out by nonveterinary technicians in the absence of pain management training. Pain control during trimming is not only an ethical obligation but also allows for better manipulation and more meticulous treatment. The aim of this study was to test the efficacy of Tri-Solfen (Bayer Australia Ltd., Pymble, NSW, Australia), a combination of local anesthetics in a topical gel form, containing lidocaine, bupivacaine, adrenaline, and cetrimide, for the treatment of pain associated with trimming of hoof lesions. Sixty-two Holstein-Frisian cows were selected for trimming at the drying-off period and were visually scored for lameness before entering the chute. After diagnosis of the hoof lesion but before deep trimming was initiated, each animal was randomly distributed to 2 groups: C, usual trimming with no pain control, and T, trimming with a local anesthetic formulation being applied immediately after live corium was exposed. During curative trimming, behavior observation was conducted by 2 observers blind to treatment. In 27 cows, algometry measurements were performed before and after the procedure to assess animal reaction to pressure. Lameness scoring was again performed as the cow left the chute. Nonparametric tests and ANOVA were performed. Results showed that use of the topical anesthetic formulation significantly reduced reaction to trimming and lameness score after trimming when compared with nontreated animals. Algometry values showed increased pressure threshold after application of topical anesthetics. This study suggests that the use of topical local anesthesia with lidocaine and bupivacaine helps reduce pain associated with corrective trimming of severe hoof lesions, enhancing animal welfare and potentially ensuring safety of trimmers.
Collapse
Affiliation(s)
- G T Stilwell
- Animal Behavior and Welfare Laboratory, Center for Interdisciplinary Research in Animal Health, Faculty of Veterinary Medicine, Lisbon University, Lisboa, Lisboa 1300-477, Portugal.
| | - A M Ferrador
- Animal Behavior and Welfare Laboratory, Center for Interdisciplinary Research in Animal Health, Faculty of Veterinary Medicine, Lisbon University, Lisboa, Lisboa 1300-477, Portugal
| | - M S Santos
- Animal Behavior and Welfare Laboratory, Center for Interdisciplinary Research in Animal Health, Faculty of Veterinary Medicine, Lisbon University, Lisboa, Lisboa 1300-477, Portugal
| | - J M Domingues
- Animal Behavior and Welfare Laboratory, Center for Interdisciplinary Research in Animal Health, Faculty of Veterinary Medicine, Lisbon University, Lisboa, Lisboa 1300-477, Portugal
| | - N Carolino
- Instituto Nacional de Investigação Agrária e Veterinária, I.P., Unidade Estratégica de Investigação e Serviços de Biotecnologia e Recursos Genéticos, Polo de Investigação da Fonte Boa, Vale de Santarem 2005-048, Portugal
| |
Collapse
|